Botanical Journal of the Linnean Society

Shefferson, Richard, P;Jacquemyn,, Hans;Kull,, Tiiu;Hutchings, Michael, J

doi: 10.1093/botlinnean/boz084pmid: N/A

Abstract Terrestrial orchid life-cycles are complex and dependent on pollinators and mycorrhizal associates. Worldwide, orchid populations are declining because of urbanization, atmospheric nitrogen deposition and climate change. To advance understanding of the factors determining orchid population viability, we review knowledge about orchid demography, life histories and population dynamics. Orchids can produce thousands of seeds, although few survive to reach maturity, with mortality rates declining from juvenile to adult life states. Flowering and fruiting rates vary widely between years, and many populations, especially of deceptive species, are pollen- and seed-limited. Many species have long lifespans and periods of vegetative dormancy and exhibit costs associated with reproduction, sprouting, vegetative dormancy, growth and size. Population growth rates range from 0.50–2.92 (mean: 0.983 ± 0.026). Although vital rates can fluctuate widely between years and be strongly correlated, these correlations have little impact on population dynamics. Variation in spatial density of fungi and microsite quality, limited dispersal and competition generate density dependence in vital rates. Future research should elucidate the roles of biotic and abiotic factors on population dynamics to underpin effective management for conservation. Understanding the impact of idiosyncratic individual plant behaviour on population dynamics will also improve demographic parameter estimation, including population growth rate and net reproductive rate. dormancy, evolution, Orchidaceae, plant mutualisms INTRODUCTION Orchidaceae are one of the most species-rich flowering plant families (Christenhusz & Byng, 2016). The family has attracted significant research attention from botanists, evolutionary biologists and ecologists. Pollination of the flowers of orchids can involve complex mechanisms, including pseudo-copulation (van der Pijl & Dodson, 1966; van der Cingel, 1995; Alcock, 2006; Claessens & Kleynen, 2011), and interest in the evolution of the relationships between orchids and their pollinators has a very long history (Darwin, 1862). Terrestrial orchids are unusual in many respects beyond their reproductive structures and pollination mechanisms. Their life histories involve a prolonged underground juvenile stage termed the protocorm, which, outside the Orchidaceae, is only found in the family Pyrolaceae (Hashimoto et al., 2012; Johansson & Eriksson, 2013). All orchids are perennial, and many species are capable of long life (Tamm, 1972; Inghe & Tamm, 1988). In addition, orchids have a unique mycorrhiza, termed the orchid mycorrhiza (OM). This symbiosis differs from other mycorrhizae, in particular because the orchid uses it to obtain carbon from its fungal associate, rather than providing carbon to the fungus in exchange for mineral nutrition (Rasmussen, 1995). Orchids use their mycorrhizal associate parasitically during the protocorm phase, but are able to continue to acquire carbon from this source throughout their lives (Gebauer, Preiss & Gebauer, 2016). In extreme cases, some species have evolved to utilize the mycorrhiza as their only source of carbon (Bidartondo, 2005; Jacquemyn & Merckx, 2019). Despite the high number of orchid species, many have small geographical ranges, and decreases in range, in number of populations and in size of populations have been documented in many studies (see Gale et al., 2018), particularly in species occurring in Europe (Kull et al., 2016) and Australia (Swarts & Dixon, 2009). In this paper, we review current knowledge about the population biology of orchids, in particular to seek explanations for the diminished and precarious state of many species and populations. We begin with a general introduction to the unique features of orchid ecology and demography, and then focus on important recent developments in research into orchid population ecology and conservation. Finally, we discuss the current status of orchid populations worldwide, and key directions in research that we believe are of particular importance in providing essential information to promote the conservation of these species. The information presented has been assembled from a large dataset documenting the demographic characteristics of all terrestrial plant species that have been reported to exhibit vegetative dormancy. This dataset, results from which were presented in Shefferson et al. (2018a), currently contains data on 125 dormancy-prone species. Of these 125 species, 73 are orchids. TERRESTRIAL ORCHID LIFE HISTORIES Terrestrial orchid life histories have at least six main stages: seed, protocorm, juvenile, dormant adult, vegetative adult and flowering individual. Juvenile plants include potentially mycotrophic underground structures, that are developmentally more advanced than the protocorm, and seedlings, which often have the physical appearance of adult plants but are typically small and lacking in sufficient resources to produce flowers (Tatarenko & Vakhrameeva, 1998) (Fig. 1). Each stage has its own specific vital rates, including characteristic probabilities of mortality and of transition to other stages, and other characteristics that exert an influence on population size from year to year. Vital rates vary with size, age and environmental conditions within species and differ between species. These vital rates affect population-level traits, such as the discrete population growth rate λ, which is typically estimated using either stage-based matrix projection models (Kull, 1995; Nicolè, Brzosko & Till-Bottraud, 2005) or integral projection models (Jacquemyn, Brys & Jongejans, 2010a; Miller et al., 2012). Here, we review the demography of key life stages. Figure 1. Open in new tabDownload slide Typical life cycle of a perennial terrestrial orchid (O. purpurea). Here, the juvenile portion of life following the protocorm stage includes a tuber, which only occurs in some orchid species, and a seedling. Arrows represent possible transitions between stages, and values are elasticity values of the mean matrix that was constructed with vital rates averaged over all years and sites. Vegetative dormancy is not noted in this figure, but in a typical orchid population, juveniles, non-reproductive adults and reproductive adults may all transition to this non-emergent stage (from Jacquemyn et al., 2010a). Figure 1. Open in new tabDownload slide Typical life cycle of a perennial terrestrial orchid (O. purpurea). Here, the juvenile portion of life following the protocorm stage includes a tuber, which only occurs in some orchid species, and a seedling. Arrows represent possible transitions between stages, and values are elasticity values of the mean matrix that was constructed with vital rates averaged over all years and sites. Vegetative dormancy is not noted in this figure, but in a typical orchid population, juveniles, non-reproductive adults and reproductive adults may all transition to this non-emergent stage (from Jacquemyn et al., 2010a). Seed and fruit production An orchid begins life as a dust seed. Although a single orchid plant may produce thousands of seeds in one reproductive season (Arditti & Ghani, 2000), few seeds will become plants that survive to adulthood. Deterministic matrix analyses suggest that seed production and seedling recruitment are of minor importance in determining the demographic behaviour and long-term viability of populations of long-lived plant species, including orchids (Franco & Silvertown, 2004). However, meta-analysis of seed addition experiments has shown that addition of seeds to populations of many plant species results in increased seedling recruitment (Poulsen et al., 2007), suggesting that most plant populations are seed-limited. For example, increasing fruit and seed set by hand-pollination increased overall seed germination in the tropical epiphytic orchid Tolumnia variegata (Sw.) Braem (Ackerman, Sabat & Zimmerman, 1996). In a population of the terrestrial orchid Orchis purpurea Huds., open habitat yielded greater recruitment and a higher population growth rate than shaded habitat, resulting in double the average fruit production and five times higher net reproductive rate in populations occurring in open habitat (Jacquemyn, Brys & Jongejans, 2010b). The availability of suitable microsites for germination may also limit number of seedlings recruited into populations (Kull, 1998; Rasmussen et al., 2015). Fruit and seed production in orchids differ not only between orchid species (Tremblay et al., 2005; Trunschke, Sletvold & Ågren, 2017), but also between different populations within species (Chapurlat, Ågren & Sletvold, 2015) and between years within single populations (Firmage & Cole, 1988; Jacquemyn et al., 2010a). Fruit set per flower is generally > 50% in species that provide rewards for pollinating species, but usually < 20% in deceptive, non-rewarding species (Dafni & Ivri, 1981; Gill, 1989; Neiland & Wilcock, 1998). Direct observations and pollen addition experiments have shown that the low reproductive success in deceptive orchids is in most cases the result of pollinator limitation (Nilsson, 1992; Johnson & Bond, 1997;Sletvold, Tye & Ågren, 2017): adding pollen to flowers almost always increases fruit set (Tremblay et al., 2005). Likewise, populations with low pollinator abundance generally exhibit low fruiting success (Ackerman, Meléndez‐Ackerman & Salguero‐Faria, 1997). Local density of flowering plants can also influence fruit set in both rewarding and deceptive orchids (Sabat & Ackerman, 1996; Jacquemyn et al., 2009a). In a French population of Dactylorhiza sambucina (L.) Soó and in populations of eight Australian orchid species [Caladenia arenicola Hopper & A.P.Br., C. discoidea Lindl., C. latifolia R.Br., Diuris magnifica D.L.Jones, Leporella fimbriata (Lindl.) A.S.George, Pterostylis ectypha D.L.Jones & C.J.French, P. sanguinea D.L.Jones & M.A.Clem. and Thelymitra macrophylla Lindl.], fruit set was highest at low densities and decreased with increasing density (Internicola et al., 2006; Brundrett, 2019). Conversely, fruit set in populations of the deceptive orchids Cypripedium candidum Muhl. ex Willd. and C. parviflorum Salisb. was positively dependent on the local density of conspecific plants (Shefferson, Mizuta & Hutchings, 2017). However, density did not affect fruiting in two populations of the deceptive species O. purpurea (Jacquemyn & Brys, 2010). More spatial analyses are required to understand whether effects of density are typically negative but deceptive species seem to produce more seeds per fruit than rewarding species (Sonkoly et al., 2016). Early life and demography of juvenile plants Germination of orchid seeds generally requires the establishment of a mycorrhiza (Rasmussen, 1995). Achieving this successfully leads to the protocorm stage, during which the orchid is a juvenile plant living heterotrophically, acquiring its nutrition directly from its mycorrhizal fungus. The protocorm stage occupies the time between the dust seed and the seedling stages, and it may last from months to years depending on the species (Ziegenspeck, 1936). The demographic properties of the earliest orchid life stages are not well understood because of the difficulties of working with dust seeds and obtaining data on subterranean life stages. Survival of germinated seeds through the protocorm stage to the seedling stage is typically low. Some species are estimated to have germination as low as one to five seeds per billion seeds (Rasmussen, 1995). Germination of orchid seeds under natural conditions is typically estimated using seed enclosed in mesh packets, containing tens or hundreds of seeds per packet, that are implanted into the substrate. However, the level of germination achieved under such conditions can be affected by both density of seeds and the material from which the packet is made, especially mesh size, as this determines the extent to which seeds are in contact with the environment external to the packet (Rasmussen & Whigham, 1993, 1998). The local abundance of appropriate mycorrhizal fungi also strongly determines germination in at least some species (McCormick et al., 2016), and germination generally requires abundant soil moisture, high organic matter content in the soil and low exposure to light (Rasmussen et al., 2015). Thus, germination depends strongly on both biotic community and abiotic ecosystem characteristics. The impact of fungal density in the local environment can influence patterns in germination and establishment. Although adult plants of Anacamptis morio (L.) R.M.Bateman, Pridgeon & M.W.Chase, Gymnadenia conopsea (L.) R.Br. and Orchis mascula L. occupied microsites in which appropriate fungi were naturally more abundant, germination declined rapidly with distance from those plants (Jacquemyn et al., 2012), suggesting that the mycorrhizal fungi needed to induce germination declined in abundance with increasing distance from adult plants. Using quantitative PCR analyses, Waud et al. (2016) showed clear relationships between the probability of seedling establishment and fungal abundance in the terrestrial orchids O. purpurea and O. mascula. Similarly, greater abundance of mycorrhizal fungi was associated with increased seed germination and more successful protocorm development in four North American terrestrial orchid species (McCormick et al., 2009, 2012). However, fungal density in the substrate is not constant at any location, and some populations exhibit no discernable spatial pattern in germination relative to established adults. It is possible that orchid germination may be no more successful near adult plants than further away in some species if established individuals of those species act as parasites on their mycorrhizal fungi through much of their lives, rather than promoting the growth of the fungus, or if there is simply no net benefit for the fungus in these cases (Rasmussen, 1995). The effects of other factors, including weather conditions, can also overwhelm the impact of fungal density on seed germination (R.P. Shefferson, unpublished data). Mortality during the earliest years of life appears to be high in most orchid species. Orchids are sometimes consumed by the mycorrhizal fungi that they typically parasitize during germination and as protocorms (Rasmussen, 1995). Mortality across early life stages may be similar to that of trees and other perennial species, many of which start life with high mortality levels that decline with age (Jones et al., 2014). However, accurate data are scarce, and early life mortality is often estimated indirectly. For example, Nicolè et al. (2005) estimated protocorm recruitment and mortality in Cypripedium calceolus L. from sightings of new seedlings and knowledge of the early development of related species of Cypripedium L. from North America (Curtis, 1943). Obtaining reliable data on early life demography in orchids is one of the biggest challenges in understanding their population biology. Demography of mature orchids In demographic terms, the adult stages of orchids are profoundly different from the juvenile stages. Adult orchids generally exhibit much lower annual mortality rates and, in many species, have the potential to achieve long lifespans. Indeed, O. purpurea has a mean life expectancy from the seedling stage of > 60 years (Jacquemyn et al., 2010b), and the Australian species Caladenia orientalis (G.W.Carr) Hopper & A.P.Br. has a mean life expectancy from the seedling stage of 522 years (Shefferson et al., 2018a). Although some orchid species are capable of long life, nonetheless mean life expectancy differs considerably between species, and some are typically short-lived (Shefferson et al., 2018a). Our dataset shows that mean (±SE) life expectancy from the seedling stage across all terrestrial orchid species for which we have reliable data is 16.3 ± 5.5 years, with a median of 6.0 years and a range from one to 522 years (Fig. 2). In a 24-year demographic study of a population of the North American Cypripedium parviflorum, observed longevity of adults ranged from one to 24 years (the maximum observable), with an average of 13.22 ± 0.07 and a median of 13 years (Shefferson et al., 2018a). In contrast, a 32-year study of Ophrys sphegodes Mill. estimated mean life expectancy at 1.27 years, with a range of one to 29 years (Hutchings, 2010; Shefferson et al., 2018a). Although mean life expectancy from germination has never been rigorously quantified in any orchid species, it is likely that most young plants do not survive to the seedling stage (Batty et al., 2001). In comparison with other dormancy-prone herbaceous perennials, mean life expectancy does not differ significantly (likelihood ratio test: χ2 = 0.057, d.f. = 1, P = 0.811; Fig. 2). However, the data on which this comparison is based do not include the years of life spent by orchids in subterranean stages. Figure 2. Open in new tabDownload slide Histogram of mean life expectancy in orchids vs. non-orchids, with one value (522 years) removed. Figure 2. Open in new tabDownload slide Histogram of mean life expectancy in orchids vs. non-orchids, with one value (522 years) removed. Analyses of mean life expectancy in different orchid species suggest a strong influence of environmental factors. For example, mean life expectancy increases with increasing distance from the Equator, and with increasing mean annual temperature corrected for latitude (Fig. 3A, B; see Supporting Information, Appendix S1). However, it is also negatively correlated with length of study (Fig. 3C; see Supporting Information, Appendix S1), suggesting that short demographic studies may over-estimate the mean longevity of the species under study. Smaller individuals of many herbaceous plant species typically exhibit higher annual mortality. Although orchids appear to follow this general trend (Bierzychudek, 1982; Shefferson, 2006), there are many exceptions, including populations of Calypso bulbosa (L.) Oakes, Cleistesiopsis bifaria (Fernald) Pansarin & F. Barros, Cleistesiopsis divaricata (Fernald) Pansarin & F. Barros, Cephalanthera longifolia (L.) Fritsch, Cephalanthera damasonium (Mill.) Druce, Corallorhiza odontorhiza (Willd.) Poir., Cypripedium spp. (C. acaule Aiton, C. calceolus, C. candidum, C. parviflorum), Dactylorhiza cruenta (O.F.Müller) Soó, D. incarnata (L.) Soó, D. lapponica (Laest. ex Hartm.) Soó, D. maculata (L.) Soó, D. sambucina (L.) Soó, G. conopsea, Neotinea tridentata (Scop.) R.M.Bateman, Pridgeon & M.W.Chase, Neottia ovata (L.) Bluff & Fingerh., O. sphegodes, Platanthera bifolia (L.) L.C.Rich., P. ciliaris (L.) Lindl. and P. praeclara Sheviak & M.L.Bowles (Shefferson et al., 2018a). Our own analysis shows that the primary determinants of whether a population exhibits a cost of size on survival are environmental rather than intrinsic: these costs become more frequent with increasing distance from the Equator, and with decreasing mean annual precipitation (see Supporting Information, Appendix S1). Figure 3. Open in new tabDownload slide Estimated mean life expectancy in terrestrial orchids as a function of: A, absolute value of latitude, B, mean annual temperature in degrees Celsius and C, duration of study. Figure 3. Open in new tabDownload slide Estimated mean life expectancy in terrestrial orchids as a function of: A, absolute value of latitude, B, mean annual temperature in degrees Celsius and C, duration of study. Probability and rate of flowering (i.e. number of flowers or inflorescences produced in a growing season) in orchids are typically related to size, with larger individuals generally producing more flowers and flowering more frequently across years (Primack & Stacy, 1998; Jacquemyn et al., 2010a). Flowering is also often influenced by weather conditions (Molnár et al., 2012; Brundrett, 2016). For example, the proportion of emergent plants that were flowering in any year in a population of O. sphegodes was negatively correlated with the temperature of the previous year, and inflorescence extension was negatively correlated with the temperature of the previous year and sunshine hours (Hutchings, 2010). Clonal reproduction is also important in some orchid species, although its impacts on population dynamics and life histories are rarely studied. For example, Cypripedium japonicum Thunb. and C. guttatum Sw. have long rhizomes capable of growing 0.5 m per year (Cribb, 1997), leading to the potential for single individuals to dominate large areas within populations. Genetic analysis in a population of Cremastra appendiculata (D.Don) Makino in South Korea suggests that effective population size may be only 20% that of the above-ground number of ramets in clonal orchids (Chung, Nason & Chung, 2004). Vegetative dormancy Among the most interesting differences between orchids and many other flowering plants is the propensity for vegetative dormancy in many species. This is a phenomenon whereby adult plants fail to produce above-ground structures, but remain physiologically active under the ground during one or more growing seasons (Lesica & Steele, 1994; Shefferson et al., 2001). Vegetative dormancy has been observed in species from 24 plant families (Shefferson et al., 2018a). However, whereas the ability to display vegetative dormancy is unusual in some of these families, it is common in orchids. In the analysis of vegetative dormancy in Shefferson et al. (2018a), over half of the species in which it was recorded were orchids, and we are unaware of terrestrial orchid species in which vegetative dormancy does not occur. Vegetative dormancy is unlikely to be functionally equivalent in all species in which it occurs, at least in physiological terms. Although it is often assumed to be a period during which the plant is metabolically inactive, this has never been demonstrated. Root growth appears to be possible during vegetative dormancy, and it is possible that other activities remain functional. For example, buds are produced in some species but they do not sprout during the growing season due to slow and lengthy development (Tatarenko & Kondo, 2003). In C. bifaria, dormancy often occurs when a bud has been removed by herbivory, after which the plant survives without producing a replacement sprout in that year (Gregg, 2011). In this example, herbivory prevents sprouting in the year in which it takes place, whereas in Cypripedium acaule and C. parviflorum herbivory increases the probability of vegetative dormancy occurring in future years (Primack & Stacy, 1998; Shefferson et al., 2018a). The demographic characteristics of vegetative dormancy differ between species. In some orchids, it is strongly associated with higher mortality, and longer episodes of vegetative dormancy are associated with lower chances of surviving to resprout at a later date (Hutchings, 1987). In other species, longer periods of vegetative dormancy are associated with greater stress or smaller size and do not appear to be related to increased mortality risk (Shefferson, 2006). Indeed, in some species, vegetative dormancy appears to be associated with lower mortality (Shefferson et al., 2018a). For example, in the mycoheterotrophic orchid C. odontorhiza, consecutive years of sprouting are associated with higher mortality than consecutive years of dormancy, perhaps because of the high cost of sprouting in non-photosynthetic species (Shefferson et al., 2011). Furthermore, the survival of dormant plants and their probability of transition to other life stages often depends on their history before becoming dormant and on the length of time they stay dormant (Primack & Stacy, 1998; Alahuhta et al., 2017; Hurskainen et al., 2018). Some characteristics of vegetative dormancy appear to be widespread across many orchid species. The maximum duration of dormancy recorded in monitoring studies on orchids tends to decrease with distance of the population from the Equator and with mean annual precipitation, suggesting strong climatic drivers of the condition (Fig. 4A, B). Rhizomatous species are also capable of longer vegetative dormancy than bulbous or tuberous orchids (Fig. 4). Potentially high costs of sprouting, such as herbivory or stomatal dysfunction leading to high evapotranspiration (Shefferson et al., 2011, 2016; Roy et al., 2013), are also associated with longer maximum durations of vegetative dormancy (Shefferson et al., 2018a). The mean proportion of orchids that are dormant in a population in any year is significantly different in orchids with differing modes of nutrition, with partial mycoheterotrophs exhibiting lower proportions of plants in vegetative dormancy than mycoheterotrophs. Figure 4. Open in new tabDownload slide The maximum years of vegetative dormancy noted in orchid monitoring studies is influenced by: A, absolute value of latitude, B, mean annual precipitation in mm and C, the number of years over which monitoring was conducted. Figure 4. Open in new tabDownload slide The maximum years of vegetative dormancy noted in orchid monitoring studies is influenced by: A, absolute value of latitude, B, mean annual precipitation in mm and C, the number of years over which monitoring was conducted. Vegetative dormancy has evolved and been lost on many independent occasions and in a range of families and circumstances throughout evolutionary time (Shefferson et al., 2018a). The frequent observation that dormant individuals exhibit higher mortality risk than sprouting individuals (Shefferson et al., 2003) has led to the hypothesis that it is a bet-hedging strategy. In evolutionary theory, bet-hedging traits increase geometric mean fitness by reducing variance in fitness across time (Seger & Brockman, 1987). Operationally, a bet-hedging trait should appear maladaptive in the short-term because it reduces or even eliminates the probabilities of high and catastrophically low fitness. Studies of vegetative dormancy using adult survival as a proxy for fitness have demonstrated lower survival during dormancy, supporting the idea that it is either maladaptive or a bet-hedging strategy (Shefferson et al., 2003). Some work explicitly modelling the evolutionary fitness of vegetative dormancy also supports this view, although the species studied were not members of Orchidaceae (Gremer, Crone & Lesica, 2012). However, a game theoretical analysis showed that in a population of C. parviflorum, the mean proportion of plants that were observed to be in a dormant state was evolutionarily optimal, given measured costs associated with sprouting and growth (Shefferson, Warren & Pulliam, 2014). This suggests that vegetative dormancy is simply adaptive and not a bet-hedge, because survival and reproduction should suffer if the mean proportion of the population that sprouted exceeded the optimum level for the size, age and previous history of each individual. It remains to be seen whether further evidence will support the interpretation of vegetative dormancy as typically a bet-hedging trait or as a trait that is more generally adaptive. Trade-offs Trade-offs (typically negative correlations between different vital rates such as survival probability and rate of flower production that suggest costs to life history traits; Stearns, 1976) have been found in many orchid species. The view that all species express trade-offs (Cohen, Isaksson & Salguero-Gómez, 2017) was reinforced recently by an analysis that showed that failure to observe trade-offs was strongly associated with small sample size and short study duration (Shefferson et al., 2018a). However, although trade-offs may be virtually ubiquitous across species, the types of trade-off that occur differ between species and even between populations of the same species. Five categories of trade-off have been widely observed, namely costs associated with reproduction, sprouting, dormancy, size and growth. Costs may be discernible within one year of the life state or stature of an individual plant having been recorded, or they may not become evident until more than one year has elapsed. These rapidly and more slowly incurred costs are categorized as ahistoric and historic, respectively (Shefferson et al., 2014). Many analyses of trade-offs in studies of orchids focus on costs of reproduction (Fig. 5), and they may include measurements to identify a cost of flowering or fruiting to survival (Primack & Stacy, 1998), future sprouting (Shefferson et al., 2011), future size (Miller et al., 2012), future flowering (Snow & Whigham, 1989; Primack & Hall, 1990; Jacquemyn et al., 2010a) or future fruiting (Shefferson et al., 2018a). The most commonly reported cost of reproduction in orchids is a reduction in future size (observed in 44.6% of studies seeking such a cost), and the least common (24.7% of studies) is a lower probability of future sprouting. Although the frequency of occurrence of these costs does not appear to differ significantly between studies conducted on orchids and non-orchids, some costs of reproduction have strong, unique impacts in orchids. In particular, significantly shorter mean life expectancy has been observed as a reproductive cost in orchids, whereas an equivalent effect is not apparent in other dormancy-prone species (Shefferson et al., 2018a). Figure 5. Open in new tabDownload slide Prevalence of life history trade-offs in a dataset of 73 orchid species and 180 populations (Shefferson et al., 2018a). Ahistoric costs refer to costs occurring within one year of life history events, historic costs refer to those incurred two or more years later. Figure 5. Open in new tabDownload slide Prevalence of life history trade-offs in a dataset of 73 orchid species and 180 populations (Shefferson et al., 2018a). Ahistoric costs refer to costs occurring within one year of life history events, historic costs refer to those incurred two or more years later. Orchids also exhibit costs as a consequence of sprouting. Because sprouting involves the production of above-ground tissue, which is usually photosynthetic, these costs may be the result of mineral nutrition limitation, incurred in defending these tissues against herbivory (Shefferson et al., 2011) or caused by stomatal dysfunction leading to uncontrolled transpiration, as in some mycoheterotrophic orchids (Roy et al., 2013; Shefferson et al., 2016). The most common cost of sprouting is a reduction in survival (found in 72.1% of orchid species examined), and the least common is a reduction in the probability of future sprouting (occurring in 15.8% of cases). Orchids also exhibit costs of vegetative dormancy, which can be viewed as the logical inverse of sprouting. The most common costs of dormancy are evident as reductions in the probability of future sprouting (observed in 72.5% of cases examined) and future flowering (70.1% of cases), and the least common are reductions in future size (45.6% of cases) and in future fruiting (46.7% of cases). Costs of growth and of larger size are also evident in orchids. These costs have been far less commonly reported than reproductive costs, but this may be because such costs do not usually become apparent for several years (Shefferson et al., 2014). Costs of growth result from plants increasing their above-ground biomass between years in ways that lead to increased mortality or reductions in sprouting probability, size or future reproduction (Shefferson et al., 2014). For example, growing to a large size in one year may result in increased mortality in the next if environmental conditions rapidly deteriorate, particularly if growth involves the use of stored reserves that would otherwise be used to endure more difficult periods, or if growth results in greater visibility to herbivores or greater exposure to pathogens (Shefferson & Roach, 2010). Since the typical orchid monitoring study involves only one monitoring session per year, this cost of growth would be observed only if the plant grew measurably larger than in the previous year, meaning that a cost of growth to survival in some year requires knowledge of the size of the individual in the preceding two years. Costs of large size may also occur when growth is physiologically costly or when large size leads to greater mortality risks imposed by the environment (Shefferson et al., 2018a). Although larger plants are typically expected to have higher survival than smaller plants (Bierzychudek, 1982), a deteriorating environment or changing conditions may lead to larger plants being more exposed to herbivory or dehydration and, hence, lead to higher mortality in larger plants. Each category of cost is significantly more common in orchids than in other dormancy-prone plant species (growth costs have been observed in 46.2% of non-orchid species vs. 81.0% of orchid species; size costs have been observed in 41.8% of non-orchid species vs. 61.4% of orchid species). In orchids, a reduction in survival is the most commonly observed cost of growth (51.2% of cases), whereas reduction in future size is the least common cost (13.8% of cases). The most common cost of large size is a reduction in survival (60.6% of cases), and the least common is reduction in future size (8.6% of cases). A cost of reproduction on survival is evident if flowering or fruiting plants exhibit higher mortality in the following year or later than non-flowering or non-fruiting plants. Methods to assess the importance of longer-term (historic) costs were developed by Ehrlén (2000). The complexity of the methods and an initial case study showing few differences between the results of historic and ahistoric analyses suggested that such exploration was unwarranted (Ehrlén, 2000). However, a more detailed study by Shefferson et al. (2018a) demonstrated that longer-term trade-offs (i.e. historic costs) are just as common as ahistoric costs and that their patterns often differ from those of ahistoric costs (Fig. 5). Long-term costs of growth appear to even drive the evolution of dormancy in some orchid species (Shefferson et al., 2014). Life history trade-offs are driven partially by environmental factors, but some environmental effects exert different impacts in orchids and in species from other families (Shefferson et al., 2018a). These differences may be caused by variation in the availability of resources governing the expression of trade-offs, and they suggest that trade-offs are highly plastic. For example, in Orchidaceae, costs of reproduction, sprouting and size are less common in species and populations located in sites with higher precipitation. A similar effect is not apparent in dormancy-prone species from other families (Shefferson et al., 2018a). Different environmental conditions may also alter the expression of trade-offs via altered gene regulation (Stearns & Magwene, 2003), although we are unaware of studies exploring this phenomenon in orchids. Population dynamics The population dynamics of orchids have been explored most commonly through analyses of the long-term behaviour of individual plants in different stage classes (Tamm, 1972; Hutchings, 1987; Brzosko, 2002; Kull, 2002), although deterministic analyses of population projection matrices are also common (Gregg, 1991; Tremblay & Hutchings, 2002; Nicolè et al., 2005; Pfeifer et al., 2006; Sletvold et al., 2013). The latter methodology is widely used in conservation biology to compare the effects of different management scenarios and to make inferences about the life history evolution of populations (Beissinger & Westphal, 1998). Software packages, code and instructions are widely available (Caswell, 2001; Merow et al., 2014; Griffith et al., 2016), and most information about orchid population dynamics comes from such analyses. The growth rate of orchid populations shows wide variation between species and between different populations of the same species. The 180 populations of 73 orchid species in our dataset have a mean deterministic population growth rate of 0.983 ± 0.026, which is not significantly different from stasis. However, growth rate ranged from 0.5 in a Massachusetts population of C. acaule to 2.92 in a Minnesota population of Platanthera praeclara. Population growth rate can change profoundly from year to year in the same population, regardless of whether the long-term average is near stasis or more dramatic, and this variation can cause population size to vary widely between years (Light & MacConaill, 2006; Jacquemyn & Hutchings, 2010). The probability of emerging can also vary dramatically between years, potentially yielding biased estimates of population size because a substantial proportion of the population may be dormant in any given year (Shefferson et al., 2001; Shefferson & Tali, 2007). Deterministic analysis suggests that the survival of adult plants is particularly important for the maintenance of population growth rate in orchids, as it is in species from other families (Franco & Silvertown, 2004; Tremblay, Raventós & Ackerman, 2015). Such results have led to general calls for management strategies that protect large, mature individuals (Pfeifer et al., 2006). Orchid populations have a tendency to be composed of seemingly close proportions of adult flowering, adult vegetative and adult dormant plants (32.8 ± 1.3%, 42.9 ± 1.3% and 23.8 ± 1.0%). Although variability exists, we have not found any evidence of population structure varying systematically with population size. However, because dormancy is generally not measurable in the beginning and end years of a monitoring study, the proportion of dormant plants may actually be higher in many populations, particularly those studied for only several years (Shefferson, 2002). Population dynamics across time Orchids can display considerable variation in mortality, fecundity and other vital rates between years (Wells, 1967; Hutchings, 2010). For example, in a study of an English population of the early spider orchid (O. sphegodes), mortality and reproduction varied between years in which the number of recruits exceeded the number of deaths (net annual gain) and years with a net loss of individuals (Hutchings, 2010). In a population of Spiranthes spiralis (L.) Chevall. in the Netherlands, the percentage of plants flowering varied from 0 to 100% (Jacquemyn et al., 2007a). The percentage of dormant plants can also vary strongly between years (Shefferson & Tali, 2007; Hutchings, 2010; Shefferson et al., 2018a). Variation in recruitment may to some extent be related to variation in fruit and seed set (Ackerman et al., 1996), which themselves fluctuate from year to year. In a grassland population of the lady orchid (O. purpurea) in eastern Belgium, for example, average fruit set across all flowering individuals within the population ranged from 5.0 to 19.9% (mean 10.2%) (Jacquemyn & Brys, 2010). When populations that differed in the number of flowering individuals were compared, large populations set more fruit per flowering plant on average than small populations and they also experienced smaller annual variation in fruit set, suggesting that small populations may be much more vulnerable to variation in fruit set than large populations (Jacquemyn, Brys & Honnay, 2009b). Taken together, these and other instances of temporal variability in vital rates can have a profound impact on long-term population dynamics in orchids. The main way in which this variability has been analysed is through stochastic population models, which assume that vital rates vary randomly between years. The growth rate in a stochastic environment is called the stochastic growth rate and is commonly denoted by logλs (Caswell, 2001) or a (Tuljapurkar, 1990). It can be calculated by simulation, or approximated as: logλs≈logλ1−V(logλ1)2λ12 (1) where λ1 is the deterministic population growth rate, and V(log λ1) refers to the variance in the natural logarithm of λ1 (Lewontin & Cohen, 1969; Caswell, 2001). The inclusion of the variance in λ1 as a key factor determining stochastic growth rate strongly signifies that any increase in temporal variability in vital rates will decrease the long-term population growth rate (Lewontin & Cohen, 1969; Tuljapurkar & Orzack, 1980). Davison et al. (2010) and Caswell (2010) developed stochastic life table response experiment (SLTRE) analysis to assess the impacts of life stages, matrix transition probabilities and their variance and other factors on the stochastic population growth rate. These analyses provide more accurate insights than deterministic analyses into the way in which variation in climate, for example, affects population dynamics, and which vital rates (or variability in vital rates) contribute most to variation in stochastic growth rate. For example, in an in situ experiment on wild populations of C. calceolus and C. longifolia increased vegetative dormancy following defoliation, and shading was found to buffer population growth against the negative effects of lower levels of flowering and survival (Shefferson et al., 2012). The sign and magnitude of correlations between vital rates are important factors determining the effect of temporal variability on population dynamics (Tuljapurkar, 1982). Positive correlations between vital rates through time are expected to increase the variability in population growth, making the effect of temporal variance on long-term stochastic growth rate more negative than if the same vital rates varied independently. Negative correlations between vital rates through time, on the other hand, are expected to decrease the variability of population growth, thereby buffering the negative effects of temporal variation in vital rates on long-term stochastic growth rate. However, quantifying the contributions of different types of correlation between vital rates to stochastic population dynamics is not always straightforward, and few studies have attempted to assess their effects on population dynamics in stochastic environments (Davison et al., 2013; Compagnoni et al., 2016). Empirical assessments of vital rate correlations suggest that they are common in orchid populations, but that they may not lead to large impacts on population dynamics. Using long-term demographic data for the lady orchid, O. purpurea, Compagnoni et al. (2016) found a significant positive correlation between the probability of flowering and the number of flowers produced per flowering plant, indicating that years that are good for flowering are generally good for investment in large numbers of flowers in those plants that flower. They also found a significant negative correlation between growth and the number of flowers produced per flowering plant, suggesting a trade-off between growth and reproduction. The effect of these vital rate correlations on long-term stochastic population growth rate was small, suggesting that vital rate correlations had negligible effects on population dynamics in this long-lived species. Using a different approach, Davison et al. (2013) showed that, in C. calceolus, correlations between vital rates accounted for only 3.7% of all effects on the stochastic growth rate, whereas mean vital rates accounted for 77.1%. Variability in vital rates accounted for 12.6% and elasticities, representing the relative influence of proportional changes in vital rates on stochastic growth rate, accounted for 6.6%. Conservation management and impacts of disturbance on population dynamics Conservation management practices, in particular prescribed burns and livestock grazing, have significant effects on vital rates and population dynamics. In Australia, for example, whereas individuals of the terrestrial orchid C. orientalis were more likely to increase in size and to flower immediately after prescribed fire (Coates & Duncan, 2009), over the long term, fire favoured long-lived shrubs that competed with this orchid species, resulting in lower population growth rate. In populations of Prasophyllum correctum D.L.Jones, intervals of less than three years between successive fires also promoted more flowering, but grazing in the first two years following a fire reduced population size (Coates, Lunt & Tremblay, 2006). Populations of many orchid species rely on grazing of associated species to keep population growth rates high. These include European species such as O. sphegodes (Hutchings, 1987), S. spiralis (Jacquemyn & Hutchings, 2010) and G. conopsea (Meekers et al., 2012). In these cases, grazers remove seedlings of shrubs, trees and larger herbaceous species that compete with the orchids for light or other resources. In contrast, some North American C. candidum populations suffered from grazing because herbivores explicitly targeted orchids more than their primary competitors (Falb & Leopold, 1993). The impacts of management and discrete or periodic events are increasingly assessed by using transient analysis, which examines the dynamics of populations that are perturbed away from their equilibrium state (Koons et al., 2005; Stott, Townley & Hodgson, 2011). Eliminating the assumption of equilibrium allows for a more natural interpretation of population dynamics, since the impacts of climate change, habitat alteration and other environmental changes probably keep populations at least somewhat out of equilibrium, and probably even drive natural selection for life history evolution (Koons et al., 2016). In transient analysis, fluctuations in population size are assumed to decay in magnitude before a return to the equilibrium state. However, the amount of time it takes to return to equilibrium may be substantial, and if perturbations keep occurring it may never happen. Ecologists using transient analysis techniques often estimate resilience, which is defined as the projected rate at which population dynamics return to equilibrium, and reactivity, which is the maximum rate at which small perturbations grow in magnitude (Caswell, 2001). The latter is of particular interest in conservation, because reactive populations may be more extinction-prone when undergoing more extreme population fluctuations in response to disturbance. Inertia, or the degree to which perturbations result in long-term deviations from equilibrium, is also fundamentally important (Stott et al., 2011). Transient analyses of orchid population dynamics are not common, but they have recently started to increase in number. Raventós et al. (2015) used such analyses to assess the impacts of hurricanes on the epiphytic orchids Broughtonia cubensis (Lindl.) Cogn. and Dendrophylax lindenii (Lindl.) Benth. ex Rolfe. They found that the latter was more reactive than the former, and they predicted that hurricanes could cause a 16% decline in population size within one year, with a worst case scenario of 50% if adults were eliminated from the population. They also used these analyses to suggest that translocating adult orchids from felled trees after a hurricane would have a more positive impact on population growth than translocating seedlings. An analysis of the dynamics of six populations of Lepanthes rubripetala Stimson produced similar results for the effects of shifts in stage structure, with populations composed of only seedlings shrinking dramatically while populations composed of adults increased in size moderately (Tremblay et al., 2015). The authors speculated that most cases of natural dispersal of orchids are likely to be short-lived, since dispersal by seed should lead to populations composed of only seedlings, which are likely to shrink. They also warned that deterministic methods used alone could lead to the adoption of inappropriate management practices (Tremblay et al., 2015). Spatial dynamics and density Individuals within orchid populations tend to be spatially clumped (Gillman & Dodd, 1998; Jacquemyn et al., 2007b; Brundrett, 2016). Spatial aggregation may be a consequence of vegetative propagation, limited dispersal of seeds, or spatially variable microsite quality (Phillips & MacMahon, 1981). Both genetic and ecological studies show that orchid populations are typically clumped because of all three (Kull & Tuulik, 1994; Jacquemyn et al., 2007b; ,Kartzinel, Shefferson & Trapnell, 2013), but typically low rates of clonal propagation keeping most ramets near the ramets of other genets, the potential for long distance seed dispersal to yield seedling establishment away from parents and high variability in key resources within orchid population habitats suggest that spatial variability in microsite quality may often be the key driver (Fay et al., 2009; Waud et al., 2016). Microsite quality is likely to depend on many factors including soil chemistry, light availability, moisture and biotic aspects such as the densities of plant competitors, facilitators, herbivores and fungal symbionts. In an analysis of 16 populations of C. acaule in Georgia, USA, soil chemistry was found to be highly variable both within and between the locations of of the populations, with calcium concentration being the most variable measured chemical factor and soil pH the least (Bunch et al., 2013). In a population of C. parviflorum, the most densely populated areas were emergent tussocks in a submerged meadow, whereas the least densely populated areas were drier with a sandy substrate (Shefferson et al., 2001). Variable microsite quality is among the dominant factors causing positive density dependence in vital rates of orchid populations. In O. purpurea, variable microsite quality causes higher recruitment in more densely occupied habitat patches (Jacquemyn et al., 2007b). In C. candidum and C. parviflorum, survival and fruiting were also positively density dependent. In O. sphegodes, the probabilities of sprouting and flowering were positively density dependent (Shefferson et al., 2017). Variable microsite quality may also be behind the five-year lagged density dependence in population density itself that was observed in a 27-year study of A. morio (L.) R.M.Bateman, Pridgeon & M.W.Chase (Gillman & Dodd, 2000). In orchids as in other species, negative density dependence is thought to be driven primarily by competition (Gunton & Kunin, 2007). This has made negative density dependence useful as a means of inferring competition in models of the mode of action of natural selection in orchid populations. Shefferson et al. (2014) tested evolutionary hypotheses for the persistence of vegetative dormancy in a population of the North American C. parviflorum. They tested for density dependence in survival, sprouting, growth, flowering and fruiting, and they found negative density dependence in sprouting and flowering. Game theoretical models were developed to assess optimal sprouting frequencies that incorporated negative density dependence in the proportion of the population that sprouted. The resulting models predicted optimal sprouting rates that corresponded closely with those observed in the populations over the year of monitoring. Williams et al. (2015) used negative density dependence as a competition proxy to investigate the way in which weather conditions under climate change scenarios might affect the optimal size at flowering in a population of O. purpurea. There was no evidence of a relationship between seed production and recruitment, suggesting that the population was close to carrying capacity. This led to the assumption that intraspecific competition would be strongest for recruitment sites. A game theoretical model was then developed in which seed production and recruitment were positively correlated at low density, but uncorrelated at high density. This model accurately predicted optimal plant size at flowering under ambient climatic conditions, and it predicted smaller optimal flowering size under climatic conditions with more frequent droughts. There is mounting evidence that orchid population dynamics depend on the abundance of mycorrhizal fungi in the soil. For example, long-term observational studies combined with assessments of fungal abundance in the soil using quantitative real-time PCR analyses have shown that the number of orchid individuals is significantly and positively correlated with local fungal abundance (McCormick et al., 2009; McCormick, Whigham & Canchani‐Viruet, 2018). In studies in which mycorrhizal fungi have been added to the soil, orchid seeds only germinate when the fungi are abundant (McCormick et al., 2012). There is also evidence that fungal abundance affects the probabilities of survival and sprouting in established populations. For example, Rock-Blake et al. (2017) showed that the abundance of Russulaceae in the soil affected emergence of the terrestrial orchid Isotria medeoloides (Pursh) Raf. McCormick et al. (2009) showed that only individuals that formed associations with a particular operational taxonomic unit (OTU) of Tomentella were able to withstand severe drought. Plants associating with other OTUs failed to sprout or died. This suggests that both fungal abundance and fungal identity have impacts on orchid population dynamics (McCormick et al., 2018). Evidence of mycorrhizal impacts on orchid density strongly suggest an important role for plant-soil feedback in orchid demography. Negative feedbacks occur when increasing plant density causes pathogens and herbivores specializing on the spreading plant species to spread as well, leading to lower conspecific plant density (Bever, Westover & Antonovics, 1997; van der Heijden, Bardgett & van Straalen, 2008). Positive feedbacks occur when increasing plant density causes plant mutualists, such as mycorrhizal fungi, to spread as well, leading to higher densities of mutualistic soil fungi and higher plant germination and survival (Bever et al., 2010). It is likely that orchid populations typically encounter positive plant-soil feedbacks given their reliance on mycorrhizal fungal density for recruitment. Declining orchid populations Orchids are of prime conservation concern. Their ranges are known to be shrinking in Europe, and they are consistently on Red Lists worldwide (Kull & Hutchings, 2006; Kull et al., 2016). Two important factors resulting in shrinking ranges are habitat destruction as urban environments spread and land use changes (Swarts & Dixon, 2009). Harvest for the medicinal trade is also a major concern in East Asia (Liu et al., 2015). However, because of the direct nature of human involvement in these causes of decline, they are unlikely to account for declines uniformly across entire ranges of species. The factors that have been responsible for the decline of orchid species worldwide may function mechanistically by first reducing the growth rates of existing populations, potentially via increasing individual mortality. In extreme situations this can lead to the loss of entire populations, and a contraction of range. Decreases in orchid population growth rates are likely to be due to factors operating over much of the world, or even globally. For example, populations of S. spiralis have been declining in Europe for many years, with a greater decline in Western Europe than to the east (Jacquemyn & Hutchings, 2010). The east–west gradient in decline is thought to be due to factors linked to the higher human population density in Western Europe, including greater urbanization and probably atmospheric nitrogen deposition (Kull et al., 2016). Climate change has also increased the severity and frequency of hurricanes, which can also cause declines in orchid populations through extreme mortality (Raventós et al., 2015). Range contraction may be driven by disproportionately low population growth rate in populations at distribution limits. Although we estimated the mean deterministic population growth rate across all the populations in our database as not significantly different from 1, virtually none of the data recorded the position of the study population relative to species distribution range. We suspect that orchid populations at range edges may be more susceptible to decline than more central populations. Although human activity is probably the ultimate cause of this decline, the mechanisms are mostly indirect and include factors such as succession resulting from altered management regimes, such as the cessation of grazing in many parts of Europe and elsewhere (Pykälä, 2000; Shefferson et al., 2018a) and climate change, which may disrupt necessary synchronies between flowering time and pollinator emergence (Hutchings et al., 2018). Atmospheric nitrogen deposition resulting from industrial and automobile pollutants is also a long-term threat that has already reduced the diversity of ectomycorrhizal fungi in temperate forests (Lilleskov et al., 2002; Cox et al., 2010; van der Linde et al., 2018), causing problems for orchids that rely on these fungi for their nutrition. Invasive species also alter patterns of interspecific competition and community structure in ways that affect long-term orchid persistence (Swarts & Dixon, 2009). FUTURE DIRECTIONS We suggest a number of pressing issues for further research. First and foremost, we need more research on the roles of environmental variability, including the impacts of environmental factors themselves, and of the roles that density plays in determining population dynamics. Stochastic analyses are currently being undertaken on orchid populations (Davison et al., 2013; Compagnoni et al., 2016), but transient analyses, which do not make any assumptions about long-term stationarity in vital rates or environmental factors (Caswell, 1978), and density dependent analyses are still rarely used. However, the data-hungry nature of these analyses and their complexity are obstacles to their widespread adoption; although such attempts would lead to the most accurate, predictive models of population dynamics and, hence, to the best management plans for the conservation of populations. Second, little work has been done to assess to the size of the typical orchid population, and particularly on whether small populations exhibit different dynamics than larger populations. Although many studies of orchid population ecology exist, they are typically focused on larger populations [mean sample size of 351 ± 61 individuals in the dataset of Shefferson et al., (2018a)], which provide statistical power that smaller populations lack and which are relatively easily detected in the landscape. What is the typical distribution of population sizes in orchid species? If typical orchid populations are small and relatively ephemeral, then we actually know less about orchid population dynamics than we think we do, despite the relative wealth of long-term studies on sizeable populations. Third, it is unclear to what extent orchids are evolving in response to changing environmental conditions in the wild. Rapid evolution is common under natural conditions, but it is often temporally random, in terms of both its strength and direction (Schoener, 2011). Nonetheless, at least some of this evolution is probably rapid, beneficial and consistent enough to influence population dynamics and demographic patterns (Shefferson & Salguero-Gómez, 2015). Urbanization and other anthropogenic influences may also be causing evolutionary domestication in many species, as has to some extent already occurred in species reliant on herbivore-dependent management regimes in Europe (Johnson & Munshi-South, 2017; Shefferson et al., 2018b). Game theoretical approaches have been used to test for the potential for rapid evolution in some orchid populations (Shefferson et al., 2014; Williams et al., 2015), but validation of predicted trends and observation of natural selection in situ will both be necessary in the future. Fourth, it is unclear to what extent orchid populations are influenced by differences in the behaviour of the individual plants that constitute populations. Matrix population projections assume no individual heterogeneity in behaviour within a population beyond variation by life stage and/or age (Caswell, 2001; Legendre & Legendre, 2012). However, theory suggests that individual differences strongly determine senescence patterns at the population level (Vaupel, Manton & Stallard, 1979), and senescence influences population dynamics, as it is generally documented in populations via increasing age-specific mortality in older individuals. The assumption of no individual heterogeneity is relaxed in historical matrix projections on non-orchid species (Ehrlén, 2000), which have shown that matrix projections mask important variation affecting population dynamics that are caused by long-term trends in individual growth (Shefferson et al., 2018a). These issues may also affect orchid populations. Fifth, we believe that further research should focus on the impacts of trade-offs on population dynamics. Although it may seem that much information already exists on this subject, this is not the case. Although trade-offs are ultimately expressed physiologically or genetically within individuals, evidence for their existence is typically assessed using patterns of expression across all the individuals of a population (Reznick, 1985). Assessing trade-offs at the population level ignores physiological processes that do not yield strong inter-individual differences in expression patterns between individuals. A well-designed breeding study might at least discover trade-offs that would be susceptible to natural selection (Reznick, Nunney & Tessier, 2000). We still do not understand how physiological trade-offs within individuals impact on demographic variables at the population level. Finally, as dynamic systems, orchid populations are typically resilient to some degree of perturbation, such as climate change and habitat modification (Scheffer et al., 2012). Systemic resilience may prevent extinction for some time in the face of such challenges, but it may also exacerbate decline once a critical tipping point is reached and population dynamics shift more dramatically toward decline. Such critical tipping points can sometimes be predicted through the observation of slow recovery from perturbations, e.g. by following management actions such as prescribed burning or the removal of invasive species (Caswell & Kaye, 2001; McEachern, Thomson & Chess, 2009). Critical tipping points are currently being investigated across many ecological systems (Brook et al., 2013; Allen, Breshears & McDowell, 2015), and we encourage further work assessing their importance in orchid population dynamics. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Appendix S1. Supplementary methods and results. AUTHOR CONTRIBUTIONS R.P.S. handled the analysis, organization, and primary writing of the manuscript. H.J., T.K. and M.J.H. all contributed to the writing and organization of the manuscript. ACKNOWLEDGEMENTS RPS thanks J. Nagata for support during the writing of this manuscript. Funding was provided by the Japan Society for the Promotion of Science Grant-in-Aid 16K07503. This study was also supported by institutional research funding IUT21-1 from the Estonian Ministry of Education and Research. We thank two reviewers for their helpful suggestions and critiques. REFERENCES Ackerman JD , Meléndez‐Ackerman EJ , Salguero‐Faria J . 1997 . Variation in pollinator abundance and selection on fragrance phenotypes in an epiphytic orchid . American Journal of Botany 84 : 1383 – 1390 . Google Scholar Crossref Search ADS PubMed WorldCat Ackerman JD , Sabat A , Zimmerman JK . 1996 . Seedling establishment in an epiphytic orchid: an experimental study of seed limitation . Oecologia 106 : 192 – 198 . Google Scholar Crossref Search ADS PubMed WorldCat Alahuhta K , Crone E , Ettinger A , Hens H , Jäkäläniemi A , Tuomi J . 2017 . Instant death, slow death and the consequences of assumptions about prolonged dormancy for plant population dynamics . Journal of Ecology 105 : 471 – 483 . Google Scholar Crossref Search ADS WorldCat Alcock J . 2006 . An enthusiasm for orchids: sex and deception in plant evolution. Oxford : Oxford University Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Allen CD , Breshears DD , McDowell NG . 2015 . On underestimation of global vulnerability to tree mortality and forest die-off from hotter drought in the Anthropocene . Ecosphere 6 : art129 . Google Scholar Crossref Search ADS WorldCat Arditti J , Ghani AKA . 2000 . Tansley Review No. 110. Numerical and physical properties of orchid seeds and their biological implications . New Phytologist 145 : 367 – 421 . Google Scholar Crossref Search ADS WorldCat Batty AL , Dixon KW , Brundrett M , Sivasithamparan K . 2001 . Constraints to symbiotic germination of terrestrial orchid seed in mediterranean bushland . New Phytologist 152 : 511 – 520 . Google Scholar Crossref Search ADS WorldCat Beissinger SR , Westphal MI . 1998 . On the use of demographic models of population viability in endangered species management . Journal of Wildlife Management 62 : 821 – 841 . Google Scholar Crossref Search ADS WorldCat Bever JD , Dickie IA , Facelli E , Facelli JM , Klironomos J , Moora M , Rillig MC , Stock WD , Tibbett M , Zobel M . 2010 . Rooting theories of plant community ecology in microbial interactions . Trends in Ecology & Evolution 25 : 468 – 478 . Google Scholar Crossref Search ADS PubMed WorldCat Bever JD , Westover KM , Antonovics J . 1997 . Incorporating the soil community into plant population dynamics: the utility of the feedback approach . Journal of Ecology 85 : 561 – 573 . Google Scholar Crossref Search ADS WorldCat Bidartondo MI . 2005 . Tansley Review: the evolutionary ecology of myco-heterotrophy . New Phytologist 167 : 335 – 352 . Google Scholar Crossref Search ADS PubMed WorldCat Bierzychudek P . 1982 . Life histories and demography of shade-tolerant temperate forest herbs: a review . New Phytologist 90 : 757 – 776 . Google Scholar Crossref Search ADS WorldCat Brook BW , Ellis EC , Perring MP , Mackay AW , Blomqvist L . 2013 . Does the terrestrial biosphere have planetary tipping points? Trends in Ecology & Evolution 28 : 396 – 401 . Google Scholar Crossref Search ADS PubMed WorldCat Brundrett MC . 2016 . Using vital statistics and core-habitat maps to manage critically endangered orchids in the Western Australian wheatbelt . Australian Journal of Botany 64 : 51 – 64 . Google Scholar Crossref Search ADS WorldCat Brundrett MC . 2019 . A comprehensive study of orchid seed production relative to pollination traits, plant density and climate in an urban reserve in Western Australia . Diversity 11 : 123 . Google Scholar Crossref Search ADS WorldCat Brzosko E . 2002 . Dynamics of island populations of Cypripedium calceolus in the Biebrza river valley (north-east Poland) . Botanical Journal of the Linnean Society 139 : 67 – 77 . Google Scholar Crossref Search ADS WorldCat Bunch WD , Cowden CC , Wurzburger N , Shefferson RP . 2013 . Geography and soil chemistry drive the distribution of fungal associations in lady’s slipper orchid, Cypripedium acaule . Botany 91 : 850 – 856 . Google Scholar Crossref Search ADS WorldCat Caswell H . 1978 . Transient behavior and life history analysis of teasel (Dipsacus sylvestris Huds.) . Ecology 59 : 53 – 66 . Google Scholar Crossref Search ADS WorldCat Caswell H . 2001 . Matrix population models: construction, analysis, and interpretation. Sunderland : Sinauer Associates . Google Preview WorldCat COPAC Caswell H . 2010 . Life table response experiment analysis of the stochastic growth rate . Journal of Ecology 98 : 324 – 333 . Google Scholar Crossref Search ADS WorldCat Caswell H , Kaye TN . 2001 . Stochastic demography and conservation of an endangered perennial plant (Lomatium bradshawii) in a dynamic fire regime . Advances in Ecological Research 32: 1 – 51 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Chapurlat E , Ågren J , Sletvold N . 2015 . Spatial variation in pollinator-mediated selection on phenology, floral display and spur length in the orchid Gymnadenia conopsea . New Phytologist 208 : 1264 – 1275 . Google Scholar Crossref Search ADS PubMed WorldCat Christenhusz MJM , Byng JW . 2016 . The number of known plants species in the world and its annual increase . Phytotaxa 261 : 201 – 217 . Google Scholar Crossref Search ADS WorldCat Chung MY , Nason JD , Chung MG . 2004 . Implications of clonal structure for effective population size and genetic drift in a rare terrestrial orchid, Cremastra appendiculata . Conservation Biology 18 : 1515 – 1524 . Google Scholar Crossref Search ADS WorldCat van der Cingel NA . 1995 . An atlas of orchid pollination, European orchids. Rotterdam : A.A. Balkema . Google Preview WorldCat COPAC Claessens J , Kleynen J . 2011 . The flower of the European orchid. Schrijen-Lippertz : Voerendall/Stein . Google Preview WorldCat COPAC Coates F , Duncan M . 2009 . Demographic variation between populations of Caladenia orientalis – a fire-managed threatened orchid . Australian Journal of Botany 57 : 326 – 339 . Google Scholar Crossref Search ADS WorldCat Coates F , Lunt ID , Tremblay RL . 2006 . Effects of disturbance on population dynamics of the threatened orchid Prasophyllum correctum D.L. Jones and implications for grassland management in south-eastern Australia . Biological Conservation 129 : 59 . Google Scholar Crossref Search ADS WorldCat Cohen AA , Isaksson C , Salguero-Gómez R . 2017 . Co-existence of multiple trade-off currencies shapes evolutionary outcomes . PLoS One 12 : e0189124 . Google Scholar Crossref Search ADS PubMed WorldCat Compagnoni A , Bibian AJ , Ochocki BM , Rogers HS , Schultz EL , Sneck ME , Elderd BD , Iler AM , Inouye DW , Jacquemyn H , Miller TEX . 2016 . The effect of demographic correlations on the stochastic population dynamics of perennial plants . Ecological Monographs 86 : 480 – 494 . Google Scholar Crossref Search ADS WorldCat Cox F , Barsoum N , Lilleskov EA , Bidartondo MI . 2010 . Nitrogen availability is a primary determinant of conifer mycorrhizas across complex environmental gradients . Ecology Letters 13 : 1103 – 1113 . Google Scholar Crossref Search ADS PubMed WorldCat Cribb P . 1997 . The genus Cypripedium. Portland : Timber Press . Google Preview WorldCat COPAC Curtis JT . 1943 . Germination and seedling development in five species of Cypripedium L . American Journal of Botany 30 : 199 – 206 . Google Scholar Crossref Search ADS WorldCat Dafni A , Ivri Y . 1981 . The flower biology of Cephalanthera longifolia (Orchidaceae) - pollen limitation and facultative floral mimicry . Plant Systematics and Evolution 137 : 229 – 240 . Google Scholar Crossref Search ADS WorldCat Darwin C . 1862 . The various contrivances by which orchids are fertilised by insects. London : John Murray . Google Preview WorldCat COPAC Davison R , Jacquemyn H , Adriaens D , Honnay O , de Kroon H , Tuljapurkar S . 2010 . Demographic effects of extreme weather events on a short-lived calcareous grassland species: stochastic life table response experiments . Journal of Ecology 98 : 255 – 267 . Google Scholar Crossref Search ADS WorldCat Davison R , Nicolè F , Jacquemyn H , Tuljapurkar S . 2013 . Contributions of covariance: decomposing the components of stochastic population growth in Cypripedium calceolus . The American Naturalist 181 : 410 – 420 . Google Scholar Crossref Search ADS PubMed WorldCat Ehrlén J . 2000 . The dynamics of plant populations: does the history of individuals matter? Ecology 81 : 1675 – 1684 . Google Scholar Crossref Search ADS WorldCat Falb DL , Leopold DJ . 1993 . Population dynamics of Cypripedium candidum Muhl. ex Willd., small white ladyslipper, in a western New York fen . Natural Areas Journal 13 : 76 – 86 . WorldCat Fay MF , Bone R , Cook P , Kahandawala I , Greensmith J , Harris S , Pedersen HÆ , Ingrouille MJ , Lexer C . 2009 . Genetic diversity in Cypripedium calceolus (Orchidaceae) with a focus on north-western Europe, as revealed by plastid DNA length polymorphisms . Annals of Botany 104 : 517 – 525 . Google Scholar Crossref Search ADS PubMed WorldCat Firmage DH , Cole FR . 1988 . Reproductive success and inflorescence size of Calopogon tuberosus (Orchidaceae) . American Journal of Botany 75 : 1371 – 1377 . Google Scholar Crossref Search ADS WorldCat Franco M , Silvertown J . 2004 . A comparative demography of plants based upon elasticities of vital rates . Ecology 85 : 531 – 538 . Google Scholar Crossref Search ADS WorldCat Gale SW , Fischer GA , Cribb PJ , Fay MF . 2018 . Orchid conservation: bridging the gap between science and practice . Botanical Journal of the Linnean Society 186 : 425 – 434 . Google Scholar Crossref Search ADS WorldCat Gebauer G , Preiss K , Gebauer AC . 2016 . Partial mycoheterotrophy is more widespread among orchids than previously assumed . New Phytologist 211 : 11 – 15 . Google Scholar Crossref Search ADS PubMed WorldCat Gill DE . 1989 . Fruiting failure, pollinator inefficiency, and speciation in orchids . In: Otte D , Endler JA , eds. Speciation and its consequences. Sunderland : Sinauer Associates , 458 – 481 . Google Preview WorldCat COPAC Gillman MP , Dodd ME . 1998 . The variability of orchid population size . Botanical Journal of the Linnean Society 126 : 65 – 74 . Google Scholar Crossref Search ADS WorldCat Gillman MP , Dodd M . 2000 . Detection of delayed density dependence in an orchid population . Journal of Ecology 88 : 204 – 212 . Google Scholar Crossref Search ADS WorldCat Gregg KB . 1991 . Variation in behaviour of four populations of the orchid Cleistes divaricata, an assessment using transition matrix models . In: Wells TCE , Willems JH , eds. Population ecology of terrestrial orchids. The Hague : SPB Academic Publishing , 139 – 159 . Google Preview WorldCat COPAC Gregg KB . 2011 . Recovery from bud disappearance explains prolonged dormancy in Cleistes bifaria (Orchidaceae) . American Journal of Botany 98 : 326 – 330 . Google Scholar Crossref Search ADS PubMed WorldCat Gremer JR , Crone EE , Lesica P . 2012 . Are dormant plants hedging their bets? Demographic consequences of prolonged dormancy in variable environments . The American Naturalist 179 : 315 – 327 . Google Scholar Crossref Search ADS PubMed WorldCat Griffith AB , Salguero-Gómez R , Merow C , McMahon S . 2016 . Demography beyond the population . Journal of Ecology 104 : 271 – 280 . Google Scholar Crossref Search ADS WorldCat Gunton RM , Kunin WE . 2007 . Density effects at multiple scales in an experimental plant population . Journal of Ecology 95 : 435 – 445 . Google Scholar Crossref Search ADS WorldCat Hashimoto Y , Fukukawa S , Kunishi A , Suga H , Richard F , Sauve M , Selosse MA . 2012 . Mycoheterotrophic germination of Pyrola asarifolia dust seeds reveals convergences with germination in orchids . New Phytologist 195 : 620 – 630 . Google Scholar Crossref Search ADS PubMed WorldCat van der Heijden MGA , Bardgett RD , van Straalen NM . 2008 . The unseen majority: soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems . Ecology Letters 11 : 296 – 310 . Google Scholar Crossref Search ADS PubMed WorldCat Hurskainen S , Alahuhta K , Hens H , Jäkäläniemi A , Kull T , Shefferson RP , Tuomi J . 2018 . Vegetative dormancy in orchids incurs absolute and relative demographic costs in large but not in small plants . Botanical Journal of the Linnean Society 188 : 426 – 437 . Google Scholar Crossref Search ADS WorldCat Hutchings MJ . 1987 . The population biology of the early spider orchid, Ophrys sphegodes Mill. I. A demographic study from 1975 to 1984 . Journal of Ecology 75 : 711 – 727 . Google Scholar Crossref Search ADS WorldCat Hutchings MJ . 2010 . The population biology of the early spider orchid Ophrys sphegodes Mill. III. Demography over three decades . Journal of Ecology 98 : 867 – 878 . Google Scholar Crossref Search ADS WorldCat Hutchings MJ , Robbirt KM , Roberts DL , Davy AJ . 2018 . Vulnerability of a specialized pollination mechanism to climate change revealed by a 356-year analysis . Botanical Journal of the Linnean Society 186 : 498 – 509 . Google Scholar Crossref Search ADS WorldCat Inghe O , Tamm CO . 1988 . Survival and flowering of perennial herbs. V Patterns of flowering . Oikos 51 : 203 – 219 . Google Scholar Crossref Search ADS WorldCat Internicola AI , Juillet N , Smithson A , Gigord LDB . 2006 . Experimental investigation of the effect of spatial aggregation on reproductive success in a rewardless orchid . Oecologia 150 : 435 – 441 . Google Scholar Crossref Search ADS PubMed WorldCat Jacquemyn H , Brys R . 2010 . Temporal and spatial variation in flower and fruit production in a food-deceptive orchid: a five-year study . Plant Biology 12 : 145 – 153 . Google Scholar Crossref Search ADS PubMed WorldCat Jacquemyn H , Brys R , Hermy M , Willems JH . 2007a . Long-term dynamics and population viability in one of the last populations of the endangered Spiranthes spiralis (Orchidaceae) in the Netherlands . Biological Conservation 134 : 14 – 21 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Brys R , Honnay O . 2009b . Large population sizes mitigate negative effects of variable weather conditions on fruit set in two spring woodland orchids . Biology Letters 5 : 495 – 498 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Brys R , Jongejans E . 2010a . Size-dependent flowering and costs of reproduction affect population dynamics in a tuberous perennial woodland orchid . Journal of Ecology 98 : 1204 – 1215 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Brys R , Jongejans E . 2010b . Seed limitation restricts population growth in shaded populations of a perennial woodland orchid . Ecology 91 : 119 – 129 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Brys R , Lievens B , Wiegand T . 2012 . Spatial variation in below-ground seed germination and divergent mycorrhizal associations correlate with spatial segregation of three co-occurring orchid species . Journal of Ecology 100 : 1328 – 1337 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Brys R , Vandepitte K , Honnay O , Roldán‐Ruiz I , Wiegand T . 2007b . A spatially explicit analysis of seedling recruitment in the terrestrial orchid Orchis purpurea . New Phytologist 176 : 448 – 459 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Hutchings MJ . 2010 . Biological flora of the British Isles: Spiranthes spiralis (L.) Chevall . Journal of Ecology 98 : 1253 – 1267 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Merckx VSFT . 2019 . Mycorrhizal symbioses and the evolution of trophic modes in plants . Journal of Ecology 107 : 1567 – 1581 . Google Scholar Crossref Search ADS WorldCat Jacquemyn H , Wiegand T , Vandepitte K , Brys R , Roldán-Ruiz I , Honnay O . 2009a . Multigenerational analysis of spatial structure in the terrestrial, food-deceptive orchid Orchis mascula . Journal of Ecology 97 : 206 – 216 . Google Scholar Crossref Search ADS WorldCat Johansson VA , Eriksson O . 2013 . Recruitment limitation, germination of dust seeds, and early development of underground seedlings in six Pyroleae species . Botany 91 : 17 – 24 . Google Scholar Crossref Search ADS WorldCat Johnson MTJ , Munshi-South J . 2017 . Evolution of life in urban environments . Science 358 : eaam8327 . Google Scholar Crossref Search ADS PubMed WorldCat Johnson SD , Bond WJ . 1997 . Evidence for widespread pollen limitation of fruiting success in Cape wildflowers . Oecologia 109 : 530 – 534 . Google Scholar Crossref Search ADS PubMed WorldCat Jones OR , Scheuerlein A , Salguero-Gómez R , Camarda CG , Schaible R , Casper BB , Dahlgren JP , Ehrlén J , Garcia MB , Menges E , Quintana-Ascencio PF , Caswell H , Baudisch A , Vaupel JW . 2014 . Diversity of ageing across the tree of life . Nature 505 : 169 – 173 . Google Scholar Crossref Search ADS PubMed WorldCat Kartzinel TR , Shefferson RP , Trapnell DW . 2013 . Relative importance of pollen and seed dispersal across a Neotropical mountain landscape for an epiphytic orchid . Molecular Ecology 22 : 6048 – 6059 . Google Scholar Crossref Search ADS PubMed WorldCat Koons DN , Grand JB , Zinner B , Rockwell RF . 2005 . Transient population dynamics: relations to life history and initial population state . Ecological Modelling 185 : 283 – 297 . Google Scholar Crossref Search ADS WorldCat Koons DN , Iles DT , Schaub M , Caswell H . 2016 . A life-history perspective on the demographic drivers of structured population dynamics in changing environments . Ecology Letters 19 : 1023 – 1031 . Google Scholar Crossref Search ADS PubMed WorldCat Kull T . 1995 . Genet and ramet dynamics of Cypripedium calceolus in different habitats . Abstracta Botanica 19 : 95 – 104 . WorldCat Kull T . 1998 . Fruit-set and recruitment in populations of Cypripedium calceolus L. in Estonia . Botanical Journal of the Linnean Society 126 : 27 – 38 . WorldCat Kull T . 2002 . Population dynamics of north temperate orchids . In: Kull T , Arditti J , eds. Orchid biology: reviews and perspectives, VIII. Dordrecht : Kluwer Academic Publishers , 139 – 165 . Google Preview WorldCat COPAC Kull T , Hutchings MJ . 2006 . A comparative analysis of decline in the distribution ranges of orchid species in Estonia and the United Kingdom . Biological Conservation 129 : 31 – 39 . Google Scholar Crossref Search ADS WorldCat Kull T , Selgis U , Villoslada Peciña M , Metsare M , Ilves A , Tali K , Sepp K , Kull K , Shefferson RP . 2016 . Factors influencing IUCN threat levels to orchids across Europe on the basis of national red lists . Ecology and Evolution 6 : 6245 – 6265 . Google Scholar Crossref Search ADS PubMed WorldCat Kull T , Tuulik T . 1994 . Orchid studies on permanent plots . In: Kull T , ed. Orchid ecology and protection in Estonia. Tartu : Tartu University Press , 35 – 42 . Google Preview WorldCat COPAC Legendre P , Legendre L . 2012 . Numerical ecology. Amsterdam : Elsevier Science . Google Preview WorldCat COPAC Lesica P , Steele BM . 1994 . Prolonged dormancy in vascular plants and implications for monitoring studies . Natural Areas Journal 14 : 209 – 212 . WorldCat Lewontin RC , Cohen D . 1969 . On population growth in a randomly varying environment . Proceedings of the National Academy of Sciences of the USA 62 : 1056 – 1060 . Google Scholar Crossref Search ADS PubMed WorldCat Light MHS , MacConaill M . 2006 . Appearance and disappearance of a weedy orchid, Epipactis helleborine . Folia Geobotanica 41 : 77 – 94 . Google Scholar Crossref Search ADS WorldCat Lilleskov EA , Fahey TJ , Horton TR , Lovett GM . 2002 . Belowground ectomycorrhizal fungi community change over a nitrogen deposition gradient in Alaska . Ecology 83 : 104 – 115 . Google Scholar Crossref Search ADS WorldCat van der Linde S , Suz LM , Orme CDL , Cox F , Andreae H , Asi E , Atkinson B , Benham S , Carroll C , Cools N , Vos BD , Dietrich HP , Eichhorn J , Gehrmann J , Grebenc T , Gweon HS , Hansen K , Jacob F , Kristöfel F , Lech P , Manninger M , Martin J , Meesenburg H , Merilä P , Nicolas M , Pavlenda P , Rautio P , Schaub M , Schröck HW , Seidling W , Šrámek V , Thimonier A , Thomsen IM , Titeux H , Vanguelova E , Verstraeten A , Vesterdal L , Waldner P , Wijk S , Zhang Y , Žlindra D , Bidartondo MI . 2018 . Environment and host as large-scale controls of ectomycorrhizal fungi . Nature 558 : 243 – 248 . Google Scholar Crossref Search ADS PubMed WorldCat Liu Q , Chen J , Corlett RT , Fan XL , Yu DL , Yang HP , Gao JY . 2015 . Orchid conservation in the biodiversity hotspot of southwestern China . Conservation Biology 29 : 1563 – 1572 . Google Scholar Crossref Search ADS PubMed WorldCat McCormick MK , Taylor DL , Juhaszova K , Burnett RK , Whigham DF , O’Neill JP . 2012 . Limitations on orchid recruitment: not a simple picture . Molecular Ecology 21 : 1511 – 1523 . Google Scholar Crossref Search ADS PubMed WorldCat McCormick MK , Taylor DL , Whigham DF , Burnett RK . 2016 . Germination patterns in three terrestrial orchids relate to abundance of mycorrhizal fungi . Journal of Ecology 104 : 744 – 754 . Google Scholar Crossref Search ADS WorldCat McCormick MK , Whigham DF , Canchani‐Viruet A . 2018 . Mycorrhizal fungi affect orchid distribution and population dynamics . New Phytologist 219 : 1207 – 1215 . Google Scholar Crossref Search ADS PubMed WorldCat McCormick MK , Whigham DF , O’Neill JP , Becker JJ , Werner S , Rasmussen HN , Bruns TD , Taylor DL . 2009 . Abundance and distribution of Corallorhiza odontorhiza reflects variations in climate and ectomycorrhizae . Ecological Monographs 79 : 619 – 635 . Google Scholar Crossref Search ADS WorldCat McEachern AK , Thomson DM , Chess KA . 2009 . Climate alters response of an endemic island plant to removal of invasive herbivores . Ecological Applications 19 : 1574 – 1584 . Google Scholar Crossref Search ADS PubMed WorldCat Meekers T , Hutchings MJ , Honnay O , Jacquemyn H . 2012 . Biological Flora of the British Isles: Gymnadenia conopsea s.l. Journal of Ecology 100 : 1269 – 1288 . Google Scholar Crossref Search ADS WorldCat Merow C , Dahlgren JP , Metcalf CJE , Childs DZ , Evans MEK , Jongejans E , Record S , Rees M , Salguero‐Gómez R , McMahon SM . 2014 . Advancing population ecology with integral projection models: a practical guide . Methods in Ecology and Evolution 5 : 99 – 110 . Google Scholar Crossref Search ADS WorldCat Miller TEX , Williams JL , Jongejans E , Brys R , Jacquemyn H . 2012 . Evolutionary demography of iteroparous plants: incorporating non-lethal costs of reproduction into integral projection models . Proceedings of the Royal Society B: Biological Sciences 279 : 2831 – 2840 . Google Scholar Crossref Search ADS WorldCat Molnár A , Tökölyi J , Végvári Z , Sramkó G , Sulyok J , Barta Z . 2012 . Pollination mode predicts phenological response to climate change in terrestrial orchids: a case study from central Europe . Journal of Ecology 100 : 1141 – 1152 . Google Scholar Crossref Search ADS WorldCat Neiland MRM , Wilcock CC . 1998 . Fruit set, nectar reward, and rarity in the Orchidaceae . American Journal of Botany 85 : 1657 – 1671 . Google Scholar Crossref Search ADS PubMed WorldCat Nicolè F , Brzosko E , Till-Bottraud I . 2005 . Population viability analysis of Cypripedium calceolus in a protected area: longevity, stability and persistence . Journal of Ecology 93 : 716 – 726 . Google Scholar Crossref Search ADS WorldCat Nilsson LA . 1992 . Orchid pollination biology . Trends in Ecology and Evolution 7 : 255 – 259 . Google Scholar Crossref Search ADS PubMed WorldCat Pfeifer M , Wiegand K , Heinrich W , Jetschke G . 2006 . Long-term demographic fluctuations in an orchid species driven by weather: implications for conservation planning . Journal of Applied Ecology 43 : 313 – 324 . Google Scholar Crossref Search ADS WorldCat Phillips DL , MacMahon JA . 1981 . Competition and spacing patterns in desert shrubs . Journal of Ecology 69 : 97 – 115 . Google Scholar Crossref Search ADS WorldCat van der Pijl L , Dodson CH . 1966 . Orchid flowers: their pollination and evolution. Coral Gables : Fairchild Tropical Garden and Miami University Press . Google Preview WorldCat COPAC Poulsen JR , Osenberg CW , Clark CJ , Levey DJ , Bolker BM . 2007 . Plants as reef fish: fitting the functional form of seedling recruitment . The American Naturalist 170 : 167 – 183 . Google Scholar Crossref Search ADS PubMed WorldCat Primack R , Stacy E . 1998 . Cost of reproduction in the pink lady’s slipper orchid (Cypripedium acaule, Orchidaceae): an eleven-year experimental study of three populations . American Journal of Botany 85 : 1672 – 1679 . Google Scholar Crossref Search ADS PubMed WorldCat Primack RB , Hall P . 1990 . Costs of reproduction in the pink lady’s slipper orchid: a four-year experimental study . The American Naturalist 136 : 638 – 656 . Google Scholar Crossref Search ADS WorldCat Pykälä J . 2000 . Mitigating human effects on European biodiversity through traditional animal husbandry . Conservation Biology 14 : 705 – 712 . Google Scholar Crossref Search ADS WorldCat Rasmussen HN . 1995 . Terrestrial orchids: from seed to mycotrophic plant. Cambridge : Cambridge University Press. Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Rasmussen HN , Dixon KW , Jersáková J , Těšitelová T . 2015 . Germination and seedling establishment in orchids: a complex of requirements . Annals of Botany 116 : 391 – 402 . Google Scholar Crossref Search ADS PubMed WorldCat Rasmussen HN , Whigham DF . 1993 . Seed ecology of dust seeds in situ: a new study technique and its application in terrestrial orchids . American Journal of Botany 80 : 1374 – 1378 . Google Scholar Crossref Search ADS WorldCat Rasmussen HN , Whigham DF . 1998 . The underground phase: a special challenge in studies of terrestrial orchid populations . Botanical Journal of the Linnean Society 126 : 49 – 64 . Google Scholar Crossref Search ADS WorldCat Raventós J , González E , Mújica E , Bonet A . 2015 . Transient population dynamics of two epiphytic orchid species after Hurricane Ivan: implications for management . Biotropica 47 : 441 – 448 . Google Scholar Crossref Search ADS WorldCat Reznick DN . 1985 . Costs of reproduction: an evaluation of the empirical evidence . Oikos 44 : 257 – 267 . Google Scholar Crossref Search ADS WorldCat Reznick DN , Nunney L , Tessier A . 2000 . Big houses, big cars, superfleas and the costs of reproduction . Trends in Ecology and Evolution 15 : 421 – 425 . Google Scholar Crossref Search ADS PubMed WorldCat Rock-Blake R , McCormick MK , Brooks HEA , Jones CS , Whigham DF . 2017 . Symbiont abundance can affect host plant population dynamics . American Journal of Botany 104 : 72 – 82 . Google Scholar Crossref Search ADS PubMed WorldCat Roy M , Gonneau C , Rocheteau A , Berveiller D , Thomas JC , Damesin C , Selosse MA . 2013 . Why do mixotrophic plants stay green? A comparison between green and achlorophyllous orchid individuals in situ . Ecological Monographs 83 : 95 – 117 . Google Scholar Crossref Search ADS WorldCat Sabat AM , Ackerman JD . 1996 . Fruit set in a deceptive orchid: the effect of flowering phenology, display size, and local floral abundance . American Journal of Botany 83 : 1181 – 1186 . Google Scholar Crossref Search ADS WorldCat Scheffer M , Carpenter SR , Lenton TM , Bascompte J , Brock W , Dakos V , Koppel J van de , Leemput IA van de , Levin SA , Nes EH van , Pascual M , Vandermeer J . 2012 . Anticipating critical transitions . Science 338 : 344 – 348 . Google Scholar Crossref Search ADS PubMed WorldCat Schoener TW . 2011 . The newest synthesis: understanding the interplay of evolutionary and ecological dynamics . Science 331 : 426 – 429 . Google Scholar Crossref Search ADS PubMed WorldCat Seger J , Brockman HJ . 1987 . What is bet-hedging? Oxford Surveys in Evolutionary Biology 4 : 182 – 211 . WorldCat Shefferson RP . 2002 . Dormancy and survival in rare, terrestrial orchids . In: Kindlmann P , Willems JH , Whigham DF , eds. Trends and fluctuations and underlying mechanisms in terrestrial orchid populations. Leiden : Backhuys Publishers , 53 – 64 . Google Preview WorldCat COPAC Shefferson RP . 2006 . Survival costs of adult dormancy and the confounding influence of size in lady’s slipper orchids, genus Cypripedium . Oikos 115 : 253 – 262 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Kull T , Hutchings MJ , Selosse MA , Jacquemyn H , Kellett KM , Menges ES , Primack RB , Tuomi J , Alahuhta K , Hurskainen S , Alexander HM , Anderson DS , Brys R , Brzosko E , Dostálik S , Gregg KB , Ipser Z , Jäkäläniemi A , Jersáková J , Kettle WD , McCormick MK , Mendoza A , Miller MT , Moen A , Øien DI , Püttsepp Ü , Roy M , Sather N , Sletvold N , Štipkova Z , Tali K , Warren RJ , II , Whigham DF . 2018a . Drivers of vegetative dormancy across herbaceous perennial plant species . Ecology Letters 21 : 724 – 733 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Kull T , Tali K , Kellett KM . 2012 . Linking vegetative dormancy to fitness in two long-lived herbaceous perennials . Ecosphere 3 : 13 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Mason CM , Kellett KM , Goolsby EW , Coughlin E , Flynn RW . 2018b . The evolutionary impacts of conservation actions . Population Ecology 60 : 49 – 59 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , McCormick MK , O’Neill JP , Whigham DF . 2011 . Life history strategy in herbaceous perennials: inferring demographic patterns from the aboveground dynamics of a primarily subterranean, myco-heterotrophic orchid . Oikos 120 : 1291 – 1300 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Mizuta R , Hutchings MJ . 2017 . Predicting evolution in response to climate change: the example of sprouting probability in three dormancy-prone orchid species . Royal Society Open Science 4 : 160647 . Google Scholar Crossref Search ADS PubMed WorldCat Shefferson RP , Proper J , Beissinger SR , Simms EL . 2003 . Life history trade-offs in a rare orchid: the costs of flowering, dormancy, and sprouting . Ecology 84 : 1199 – 1206 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Roach DA . 2010 . Longitudinal analysis of Plantago: adaptive benefits of iteroparity in a short-lived, herbaceous perennial . Ecology 91 : 441 – 447 . Google Scholar Crossref Search ADS PubMed WorldCat Shefferson RP , Roy M , Puttsepp Ü , Selosse MA . 2016 . Demographic shifts related to mycoheterotrophy and their fitness impacts in two Cephalanthera species . Ecology 97 : 1452 – 1462 . Google Scholar Crossref Search ADS PubMed WorldCat Shefferson RP , Salguero-Gómez R . 2015 . Eco-evolutionary dynamics in plants: interactive processes at overlapping timescales and their implications . Journal of Ecology 103 : 789 – 797 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Sandercock BK , Proper J , Beissinger SR . 2001 . Estimating dormancy and survival of a rare herbaceous perennial using mark-recapture models . Ecology 82 : 145 – 156 . WorldCat Shefferson RP , Tali K . 2007 . Dormancy is associated with decreased adult survival in the burnt orchid, Neotinea ustulata . Journal of Ecology 95 : 217 – 225 . Google Scholar Crossref Search ADS WorldCat Shefferson RP , Warren RJ II , Pulliam HR . 2014 . Life history costs make perfect sprouting maladaptive in two herbaceous perennials . Journal of Ecology 102 : 1318 – 1328 . Google Scholar Crossref Search ADS WorldCat Sletvold N , Dahlgren JP , Øien DI , Moen A , Ehrlén J . 2013 . Climate warming alters effects of management on population viability of threatened species: results from a 30-year experimental study on a rare orchid . Global Change Biology 19 : 2729 – 2738 . Google Scholar Crossref Search ADS PubMed WorldCat Sletvold N , Tye M , Ågren J . 2017 . Resource- and pollinator-mediated selection on floral traits . Functional Ecology 31 : 135 – 141 . Google Scholar Crossref Search ADS WorldCat Snow AA , Whigham DF . 1989 . Costs of flower and fruit production in Tipularia discolor (Orchidaceae) . Ecology 70 : 1286 – 1293 . Google Scholar Crossref Search ADS WorldCat Sonkoly J , Vojtkó AE , Tökölyi J , Török P , Sramkó G , Illyés Z , Molnár V A . 2016 . Higher seed number compensates for lower fruit set in deceptive orchids . Journal of Ecology 104 : 343 – 351 . Google Scholar Crossref Search ADS WorldCat Stearns SC . 1976 . Life-history tactics: a review of the ideas . Quarterly Review of Biology 51 : 3 – 47 . Google Scholar Crossref Search ADS PubMed WorldCat Stearns SC , Magwene P . 2003 . The naturalist in a world of genomics . American Naturalist 161 : 171 – 180 . Google Scholar Crossref Search ADS PubMed WorldCat Stott I , Townley S , Hodgson DJ . 2011 . A framework for studying transient dynamics of population projection matrix models . Ecology Letters 14 : 959 – 970 . Google Scholar Crossref Search ADS PubMed WorldCat Swarts ND , Dixon KW . 2009 . Terrestrial orchid conservation in the age of extinction . Annals of Botany 104 : 543 – 556 . Google Scholar Crossref Search ADS PubMed WorldCat Tamm CO . 1972 . Survival and flowering of some perennial herbs. II. The behaviour of some orchids on permanent plots . Oikos 23 : 23 – 38 . Google Scholar Crossref Search ADS WorldCat Tatarenko IV , Kondo K . 2003 . Seasonal development of annual shoots in some terrestrial orchids from Russia and Japan . Plant Species Biology 18 : 43 – 55 . Google Scholar Crossref Search ADS WorldCat Tatarenko IV , Vakhrameeva MG . 1998 . On vegetative propagation in orchids . Byulleten’ Botanicheskogo Sada IS Kosenko 7 : 155 – 157 . WorldCat Tremblay RL , Ackerman JD , Zimmerman JK , Calvo RN . 2005 . Variation in sexual reproduction in orchids and its evolutionary consequences: a spasmodic journey to diversification . Biological Journal of the Linnean Society 84 : 1 – 54 . Google Scholar Crossref Search ADS WorldCat Tremblay RL , Hutchings MJ . 2002 . Population dynamics in orchid conservation: a review of analytical methods, based on the rare species Lepanthes eltoroensis . In: Dixon KW , Kell SP , Barrett RL , Cribb PJ , eds. Orchid conservation. Kota Kinabalu : Natural History Publications , 163 – 183 . Google Preview WorldCat COPAC Tremblay RL , Raventos J , Ackerman JD . 2015 . When stable-stage equilibrium is unlikely: integrating transient population dynamics improves asymptotic methods . Annals of Botany 116 : 381 – 390 . Google Scholar Crossref Search ADS PubMed WorldCat Trunschke J , Sletvold N , Ågren J . 2017 . Interaction intensity and pollinator-mediated selection . New Phytologist 214 : 1381 – 1389 . Google Scholar Crossref Search ADS PubMed WorldCat Tuljapurkar S . 1982 . Population dynamics in variable environments. III. Evolutionary dynamics of r-selection. Theoretical Population Biology 21 : 141 – 165 . Google Scholar Crossref Search ADS WorldCat Tuljapurkar S . 1990 . Population dynamics in variable environments. Berlin : Springer-Verlag . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Tuljapurkar SD , Orzack SH . 1980 . Population dynamics in variable environments I. Long-run growth rates and extinction . Theoretical Population Biology 18 : 314 – 342 . Google Scholar Crossref Search ADS WorldCat Vaupel JW , Manton KG , Stallard E . 1979 . The impact of heterogeneity in individual frailty on the dynamics of mortality . Demography 16 : 439 – 454 . Google Scholar Crossref Search ADS PubMed WorldCat Waud M , Wiegand T , Brys R , Lievens B , Jacquemyn H . 2016 . Nonrandom seedling establishment corresponds with distance-dependent decline in mycorrhizal abundance in two terrestrial orchids . New Phytologist 211 : 255 – 264 . Google Scholar Crossref Search ADS PubMed WorldCat Wells TCE . 1967 . Changes in a population of Spiranthes spiralis (L.) Chevall. at Knocking Hoe National Nature Reserve, Bedfordshire, 1962–65 . Journal of Ecology 55 : 83 – 99 . Google Scholar Crossref Search ADS WorldCat Williams JL , Jacquemyn H , Ochocki BM , Brys R , Miller TEX . 2015 . Life history evolution under climate change and its influence on the population dynamics of a long-lived plant . Journal of Ecology 103 : 798 – 808 . Google Scholar Crossref Search ADS WorldCat Ziegenspeck H . 1936 . Orchidaceae . In: von Kirchner O , Loew E , Schröter C , eds. Lebensgeschichte der Blütenpflanzen Mitteleuropas. Stuttgart : Verlagsbuchhandlung Eugenulmer , 1 – 840 . Google Preview WorldCat COPAC © 2019 The Linnean Society of London, Botanical Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Yu, Ren-Yong; Van Welzen, Peter C

doi: 10.1093/botlinnean/boz075pmid: N/A

Abstract Trigonostemon and Dimorphocalyx are two morphologically similar genera in tropical Asia. We estimated their divergence times through a Bayesian clock analysis and reconstructed the historical biogeography using a likelihood analysis under the dispersal-extinction-cladogenesis (DEC) model and a statistical dispersal-vicariance analysis (S-DIVA). We have found that the two genera differ in their historical biogeography: Trigonostemon originated on the South-East Asian mainland, but one section dispersed to the Malay Peninsula and Borneo, where rapid speciation events occurred during the Pleistocene, whereas Dimorphocalyx originated on and extended to its current distribution from Borneo. The dispersal routes of both genera are well supported by the tectonic history and are comparable to the conclusions in previous case studies. Long-distance dispersals across Wallace’s line are of particular interest in biogeography. We compared the patterns of historical distribution and dispersal of our taxa and other comparable taxa in this area. Our data support the hypothesis that the Philippines is the most common stepping stone for crossing Wallace’s line. Furthermore, we consider that the frequent change of sea levels during the Pleistocene propelled the diversification of Trigonostemon section Trigonostemon in Borneo and the Malay Peninsula. dated phylogeny, dispersal-extinction-cladogenesis, Pleistocene, Quaternary, speciation, statistical dispersal-vicariance analysis, Tritaxis, Wallace’s line INTRODUCTION South-East Asia ranges from the Alpine-Himalayan mountain belt south-eastwards to a massive collection of islands across the Equator between 95°E and 140°E (Hall, 2009), which is commonly known as the Malay Archipelago (Wallace, 1869) or Malesia (Zollinger, 1857, van Steenis, 1948, 1950, Raes & van Welzen, 2009). The South-East Asian mainland (Sundaland continental core) and western Malesia (Sunda Shelf) were already attached in the Mesozoic, whereas the much younger eastern Malesia was formed by continental fragments that mainly rifted from the Australian part of the splitting Gondwana since the Eocene (Hall, 2009). Malesia has a rich biodiversity. This area harbours c. 42 000 plant species (Roos, 1993), 70% of which are endemic (van Welzen, 2005). Plant dispersals in this area have facilitated the floristic exchange between the continents of Asia and Australia (e.g. Sirichamorn et al., 2014, Crayn, Costion & Harrington, 2015, Buerki et al., 2016, Hauenschild et al., 2018). The Malesian islands form a more or less linear pathway of stepping stones for plant dispersals (van Welzen, Slik & Alahuhta, 2005), but significant water barriers (e.g. Makassar Strait and Banda Sea) exist in Wallacea, the central part of Malesia (most of Java, Lesser Sunda Islands, Sulawesi, Philippines, Moluccas; van Welzen, Parnell & Slik, 2011), and the stepping stones (e.g. Moluccas) have only emerged above sea level in the last 10 Myr (Hall, 2009). At present, the climate in Malesia is ever-wet in the west (Sunda Shelf: Malay Peninsula, Sumatra, Borneo, south-western Java) and the east (Sahul Shelf: New Guinea), whereas in most parts of Wallacea, there is a yearly dry monsoon. Because climate (especially the yearly precipitation and temperature) is an important regulator of species occurrences (Boucher-Lalonde, Morin & Curri, 2012, Araújo et al., 2013), it may have overruled the historical dispersal pathways. Relatively few studies have attempted to reveal the exact migration route for plants across Wallace’s line. The most common route between western and eastern Malesia is via the Philippines [Nauheimer, Boyce & Renner, 2012; Denduangboripant, Mendum & Cronk, 2001; Thomas et al., 2012; Jønsson et al., 2010 (birds)], but other routes have also been found, e.g. via Borneo and Sulawesi [Grudinski et al., 2014; Thomas et al., 2012; Evans et al., 2003 (frogs)] or between Java and the Lesser Sunda Islands (e.g. the route proposed by Su & Saunders, 2009, fig. 7; Chantarasuwan et al., 2016). General migration patterns can only be discovered when a sufficient number of case studies are performed. In this paper, we add an example of the historical biogeography of Trigonostemon Blume and Dimorphocalyx Thwaites, by investigating their geographical origin and dispersal routes, particularly those across Wallace’s line, and comparing these to the previous studies. Trigonostemon and Dimorphocalyx are two closely related genera of Euphorbiaceae, comprising c. 60 (Yu & van Welzen, 2018) and c. 13 species (van Welzen & van Oostrum, 2015), respectively. They are classified in tribe Codiaeae (Pax) Hutch. of subfamily Crotonoideae (Webster, 2014) and are morphologically similar in the connate stamens and the presence of petals in both staminate and pistillate flowers. Both genera have more or less the same South-East Asian distribution, ranging from India to southern China, throughout continental South-East Asia and Malesia to North-East Australia, and for Trigonostemon, even to the western Pacific (Govaerts, Frodin & Radcliff-Smith, 2000; van Welzen & van Oostrum, 2015; Yu & van Welzen, 2018). The main differences between the two genera lie in the sexual system and the number and arrangement of stamens: Trigonostemon spp. are monoecious shrubs or trees with only one whorl of three or five connate stamens, whereas most Dimorphocalyx spp. are dioecious trees and often have ten to 15 stamens arranged in two or three whorls, of which only the inner whorls are partly connate. Dimorphocalyx (including Tritaxis Baill., the two genera will be merged under the name Tritaxis; see Yu, Slik & van Welzen, 2019) was once treated as a section of Trigonostemon s.l. (Müller, 1865, 1866; Beddome, 1873), but our recent molecular phylogenetic study (Yu, Slik & van Welzen, 2019) demonstrated that Trigonostemon and Dimorphocalyx are two separate monophyletic groups in the C2 clade of Crotonoideae in the backbone phylogenetic tree of Euphorbiaceae (Wurdack, Hoffmann & Chase, 2005). Therefore, they are treated as separate genera as in their original generic circumscriptions (Blume, 1825; Thwaites, 1861). West Malesia is the distribution centre for Trigonostemon. In the genus, section Trigonostemon contains most species (c. 50% of the species), and these species are probably the result of a series of recent speciation events (which is reflected in the short branches in the phylogram; Yu, Slik & van Welzen, 2019). Particular factors in the tectonic and climatological history of this region may have promoted these events. Therefore, another aim of this paper is to provide an explanation for the rapid diversification in section Trigonostemon from a historical biogeographic angle. MATERIAL AND METHODS Divergence time estimation The same specimens and molecular markers as in Yu, Slik & van Welzen (2019) were used to estimate the divergence times of Trigonostemon and Dimorphocalyx. Forty-two species (out of c. 60) and five varieties of Trigonostemon, 12 species (out of c. 13) of Dimorphocalyx, one species of Tritaxis (probably monotypic; Dimorphocalyx and Tritaxis will be combined into one genus, see Yu, Slik & van Welzen, 2019), and one species of Ostodes (out of c. 2) were included. The sampling covers the whole distribution of the two genera, except that Trigonostemon section Pycnanthera Benth. (T. diplopetalus Thwaites and T. nemoralis Thwaites from southern India and Sri Lanka) is missing. Voucher and locality information is provided in the Supporting Information, Appendix. The molecular dating analysis was performed on matrix 1 of Yu, Slik & van Welzen (2019) using Bayesian inference (BI) via BEAST v.1.10.1 (Suchard et al., 2018). Five molecular markers were used: the nuclear ITS and the plastid trnK intron, trnT-L, trnL-F and rbcL. In the data matrix, sequences of all these five markers obtained for Trigonostemon and Dimorphocalyx are aligned with the dataset for Euphorbiaceae (based on trnL-F and rbcL) of Wurdack et al. (2005). The substitution model of the nuclear and plastid markers was selected based on the lowest Akaike information criterion scores detected by Modeltest-NG v.0.1.5, which showed the same model for the nuclear and plastid markers. The input file was created by BEAUti v.1.10.1 (part of the BEAST package) with the following settings: the dataset contained five partitions (each partition for a single molecular marker); the substitution rates were calculated under the general time reversal (GTR) model with a discrete Gamma distribution (+Γ, 4 categories) of evolutionary rates among the sites and a certain number of invariable sites (+I); the divergence times were estimated using the uncorrelated relaxed clock model (Drummond et al., 2006) with a lognormal distribution of rates; the Yule process was selected as tree prior (Yule, 1925; Gernhard, 2008), and a random starting tree was used; a total of 1.25 × 108 generations of Markov chain Monte Carlo were performed and trees were sampled every 1 × 104 generations. The results were examined for effective sampling size (ESS > 200) using Tracer v.1.7.1 (Rambaut et al., 2018). The first 20% of the sampled trees was discarded as burn-in, and the maximum clade credibility (MCC) tree was found using TreeAnnotator v.1.7.5 (part of the BEAST package). Three calibration points were used to estimate the divergence times. As no distribution type of the fossil ages was known, the calibration priors were coded as uniform distributions (Ho, 2007) with an upper and lower boundary (presented between parentheses). Because no fossil records directly identified as Trigonostemon or Dimorphocalyx could be found, we used two secondary calibrations and a fossil record of Hippomaneoidea warmanensis Crepet & Daghlian instead: The crown node of Euphorbiaceae s.s. (fig. 2 in Yu, Slik & van Welzen, 2019; the node joining Acalyphoideae s.s. and Euphorbioideae) was assigned a mean age of 89.9 Myr (97.4–81.2 Myr). The age was taken from a phylogenomic study of Malpighiales (Xi et al., 2012). In our phylogenetic tree, this node refers to most of the taxa; only Peroideae (Wurdack et al., 2005; presently Peraceae), Pandaceae and the outgroup, Humiriaceae, are excluded. The crown node of Trewia nudiflora L. [presently Mallotus nudiflorus (L.) Kulju & Welzen] and Mallotus japonicus (Spreng.) Müll.Arg. (see Fig. 3 in Wurdack et al., 2005, in the A1 clade of subfamily Acalyphoideae s.s. in Wurdack et al., 2005) was assigned a mean age of 34.31 Mya (44.79–32.35 Mya). This is the inferred age of the ancestral species of Mallotus Lour. (Fig. 3, node 113, in van Welzen et al., 2014). The crown node of hippomanoids (equivalent to tribe Hippomaneae in Webster, 2014; see Fig. 3 in Wurdack et al., 2005, including H1 and H2 clades of subfamily Euphorbioideae) was assigned a mean age of 42.8 Mya (47.8–37.8 Mya). The data was obtained from a fossil collected from the Claiborne Formation of the Mid Eocene (Crepet & Daghlian, 1982). The geological timescale of the Mid Eocene (Lutetian and Bartonian) was thus used to determine the upper and lower bounds, of which the average value was assigned as the mean age. The fossil was named Hippomaneoidea warmanensis Crepet & Daghlian. It was selected because it contained well-preserved inflorescences and pollen, and we consider the identification credible. The closest affinities of the fossil seemed to be tribe Hippomaneae, as the floral and palynological characters are shared among several genera (from H1 and H2 clades in Wurdack et al., 2005) in the tribe (Crepet & Daghlian, 1982). Therefore, we assigned this calibration point to the ancestral taxon of the whole tribe. Ancestral area reconstruction Fifteen areas (Fig. 1; Table 1) were delimited based on the distribution of extant taxa, the tectonic history and the presence of endemic species. For example, Palawan (area I in Fig. 1) was set apart from the rest of the Philippine islands (area J) due to its different geological origin (Hall, 2002) and the presence of two endemic species [Trigonostemon sp. 2 (to be described as T. montanus R.Y.Yu; Yu & van Welzen 2019) and T. victoriae R.Y.Yu & Welzen]; north-eastern India (including Bangladesh and a small part of western Myanmar) is separate from South-East Asian mainland because of two endemic species [Trigonostemon sp. 3 (unidentified at this time) and T. semperflorens (Roxb.) Müll.Arg.]. Figure 1. Open in new tabDownload slide Area delimitation used for the historical biogeographic reconstruction of Trigonostemon and Dimorphocalyx: A, Southern India and Sri Lanka; B, South-East Asian mainland including the Andaman and Nicobar Islands; C, North-eastern India, Bangladesh and western Myanmar; D, Southern China (Guangdong and Hainan); E, Southern Peninsular Thailand and the Malay Peninsula; F, Sumatra; G, Java and Bali; H, Borneo; I, Palawan; J, the Philippine Islands, excluding Palawan; K, Lesser Sunda Islands excluding Bali; L, Northern Moluccas; M, New Guinea; N, New Caledonia and O, North-eastern Australia. Table 1. Distribution of the sampled taxa. The authorities for the taxa are also shown. The abbreviations in the column ‘distributions’ refer to the areas discriminated in Figure 1. Trigonostemon sp. 1 will be described as T. palustris R.Y.Yu & Welzen (Yu et al., 2019); Trigonostemon sp. 2 will be described as T. montanus R.Y.Yu & Welzen (Yu & van Welzen 2019); we were unable to identify Trigonostemon sp. 3 for the time being Taxon . Distribution . Trigonostemon semperflorens (Roxb.) Müll.Arg. C Trigonostemon polyanthus Merr. J Trigonostemon villosus Hook.f. var. borneensis (Merr.) Airy Shaw FHJ Trigonostemon filiformis Quisumb. HJ Trigonostemon detritiferus R.I.Milne H Trigonostemon villosus Hook.f. var. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon cf. filiformis Quisumb. HJ Trigonostemon villosus Hook.f. var. cordatus R.Y.Yu & Welzen H Trigonostemon lychnos (R.I.Milne) R.Y.Yu & Welzen H Trigonostemon villosus Hook.f. var. aff. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon verticillatus (Jack) Pax var. verticillatus BE Trigonostemon flavidus Gagnep. BD Trigonostemon diffusus Merr. H Trigonostemon sp. 1 I Trigonostemon victoriae R.Y.Yu & Welzen I Trigonostemon longipes (Merr.) Merr. J Trigonostemon malaccanus Müll.Arg. BEF Trigonostemon murtonii Craib B Trigonostemon beccarii Ridl. F Trigonostemon pentandrus Pax & K.Hoffm. E Trigonostemon verticillatus (Jack) Pax var. salicifolius (Ridl.) Whitmore E Trigonostemon rufescens Jabl. E Trigonostemon balgooyi R.Y.Yu & Welzen E Trigonostemon villosus Hook.f. var. villosus E Trigonostemon capillipes (Hook.f.) Airy Shaw E Trigonostemon magnificus R.I.Milne F Trigonostemon merrillii Elmer IJ Trigonostemon sandakanensis Jabl. H Trigonostemon longifolius Baill. BCDEFHJ Trigonostemon reidioides (Kurz) Craib B Trigonostemon viridissimus (Kurz) Airy Shaw var. viridissimus BCDEFGHJK Trigonostemon verrucosus J.J.Sm. B Trigonostemon quocensis Gagnep. BCE Trigonostemon laevigatus Müll.Arg. var. laevigatus BEFGHIJ Trigonostemon sp. 3 B Trigonostemon albiflorus Airy Shaw BD Trigonostemon wui H.S.Kiu BD Trigonostemon sp. 2 C Trigonostemon cherrieri J.M.Veillon N Trigonostemon inopinatus Airy Shaw O Trigonostemon aurantiacus (Kurz ex Teijsm. & Binn.) Boerl. BEFG Trigonostemon xyphophylloides (Croizat) L.K.Dai & T.L.Wu D Trigonostemon hartleyi Airy Shaw M Trigonostemon philippinensis Stapf BEFHJ Trigonostemon bonianus Gagnep. B Trigonostemon adenocalyx Gagnep. B Trigonostemon lii Y.T.Chang B Ostodes paniculata Blume BDEFG Dimorphocalyx moluccensis Welzen & Oostrum L Dimorphocalyx sp. H Dimorphocalyx malayanus Hook.f. EHJ Dimorphocalyx glabellus Thwaites A Dimorphocalyx beddomei (Benth.) Airy Shaw A Tritaxis gaudichaudii Baill. B Dimorphocalyx ixoroides (C.B.Rob.) Airy Shaw J Dimorphocalyx trichocarpus (Airy Shaw) Welzen & Oostrum H Dimorphocalyx denticulatus Merr. EFHIJ Dimorphocalyx pauciflorus (Merr.) Airy Shaw H Dimorphocalyx muricatus (Hook.f.) Airy Shaw EFH Dimorphocalyx australiensis C.T.White KMO Dimorphocalyx cumingii (Müll. Arg.) Airy Shaw J Taxon . Distribution . Trigonostemon semperflorens (Roxb.) Müll.Arg. C Trigonostemon polyanthus Merr. J Trigonostemon villosus Hook.f. var. borneensis (Merr.) Airy Shaw FHJ Trigonostemon filiformis Quisumb. HJ Trigonostemon detritiferus R.I.Milne H Trigonostemon villosus Hook.f. var. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon cf. filiformis Quisumb. HJ Trigonostemon villosus Hook.f. var. cordatus R.Y.Yu & Welzen H Trigonostemon lychnos (R.I.Milne) R.Y.Yu & Welzen H Trigonostemon villosus Hook.f. var. aff. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon verticillatus (Jack) Pax var. verticillatus BE Trigonostemon flavidus Gagnep. BD Trigonostemon diffusus Merr. H Trigonostemon sp. 1 I Trigonostemon victoriae R.Y.Yu & Welzen I Trigonostemon longipes (Merr.) Merr. J Trigonostemon malaccanus Müll.Arg. BEF Trigonostemon murtonii Craib B Trigonostemon beccarii Ridl. F Trigonostemon pentandrus Pax & K.Hoffm. E Trigonostemon verticillatus (Jack) Pax var. salicifolius (Ridl.) Whitmore E Trigonostemon rufescens Jabl. E Trigonostemon balgooyi R.Y.Yu & Welzen E Trigonostemon villosus Hook.f. var. villosus E Trigonostemon capillipes (Hook.f.) Airy Shaw E Trigonostemon magnificus R.I.Milne F Trigonostemon merrillii Elmer IJ Trigonostemon sandakanensis Jabl. H Trigonostemon longifolius Baill. BCDEFHJ Trigonostemon reidioides (Kurz) Craib B Trigonostemon viridissimus (Kurz) Airy Shaw var. viridissimus BCDEFGHJK Trigonostemon verrucosus J.J.Sm. B Trigonostemon quocensis Gagnep. BCE Trigonostemon laevigatus Müll.Arg. var. laevigatus BEFGHIJ Trigonostemon sp. 3 B Trigonostemon albiflorus Airy Shaw BD Trigonostemon wui H.S.Kiu BD Trigonostemon sp. 2 C Trigonostemon cherrieri J.M.Veillon N Trigonostemon inopinatus Airy Shaw O Trigonostemon aurantiacus (Kurz ex Teijsm. & Binn.) Boerl. BEFG Trigonostemon xyphophylloides (Croizat) L.K.Dai & T.L.Wu D Trigonostemon hartleyi Airy Shaw M Trigonostemon philippinensis Stapf BEFHJ Trigonostemon bonianus Gagnep. B Trigonostemon adenocalyx Gagnep. B Trigonostemon lii Y.T.Chang B Ostodes paniculata Blume BDEFG Dimorphocalyx moluccensis Welzen & Oostrum L Dimorphocalyx sp. H Dimorphocalyx malayanus Hook.f. EHJ Dimorphocalyx glabellus Thwaites A Dimorphocalyx beddomei (Benth.) Airy Shaw A Tritaxis gaudichaudii Baill. B Dimorphocalyx ixoroides (C.B.Rob.) Airy Shaw J Dimorphocalyx trichocarpus (Airy Shaw) Welzen & Oostrum H Dimorphocalyx denticulatus Merr. EFHIJ Dimorphocalyx pauciflorus (Merr.) Airy Shaw H Dimorphocalyx muricatus (Hook.f.) Airy Shaw EFH Dimorphocalyx australiensis C.T.White KMO Dimorphocalyx cumingii (Müll. Arg.) Airy Shaw J Open in new tab Table 1. Distribution of the sampled taxa. The authorities for the taxa are also shown. The abbreviations in the column ‘distributions’ refer to the areas discriminated in Figure 1. Trigonostemon sp. 1 will be described as T. palustris R.Y.Yu & Welzen (Yu et al., 2019); Trigonostemon sp. 2 will be described as T. montanus R.Y.Yu & Welzen (Yu & van Welzen 2019); we were unable to identify Trigonostemon sp. 3 for the time being Taxon . Distribution . Trigonostemon semperflorens (Roxb.) Müll.Arg. C Trigonostemon polyanthus Merr. J Trigonostemon villosus Hook.f. var. borneensis (Merr.) Airy Shaw FHJ Trigonostemon filiformis Quisumb. HJ Trigonostemon detritiferus R.I.Milne H Trigonostemon villosus Hook.f. var. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon cf. filiformis Quisumb. HJ Trigonostemon villosus Hook.f. var. cordatus R.Y.Yu & Welzen H Trigonostemon lychnos (R.I.Milne) R.Y.Yu & Welzen H Trigonostemon villosus Hook.f. var. aff. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon verticillatus (Jack) Pax var. verticillatus BE Trigonostemon flavidus Gagnep. BD Trigonostemon diffusus Merr. H Trigonostemon sp. 1 I Trigonostemon victoriae R.Y.Yu & Welzen I Trigonostemon longipes (Merr.) Merr. J Trigonostemon malaccanus Müll.Arg. BEF Trigonostemon murtonii Craib B Trigonostemon beccarii Ridl. F Trigonostemon pentandrus Pax & K.Hoffm. E Trigonostemon verticillatus (Jack) Pax var. salicifolius (Ridl.) Whitmore E Trigonostemon rufescens Jabl. E Trigonostemon balgooyi R.Y.Yu & Welzen E Trigonostemon villosus Hook.f. var. villosus E Trigonostemon capillipes (Hook.f.) Airy Shaw E Trigonostemon magnificus R.I.Milne F Trigonostemon merrillii Elmer IJ Trigonostemon sandakanensis Jabl. H Trigonostemon longifolius Baill. BCDEFHJ Trigonostemon reidioides (Kurz) Craib B Trigonostemon viridissimus (Kurz) Airy Shaw var. viridissimus BCDEFGHJK Trigonostemon verrucosus J.J.Sm. B Trigonostemon quocensis Gagnep. BCE Trigonostemon laevigatus Müll.Arg. var. laevigatus BEFGHIJ Trigonostemon sp. 3 B Trigonostemon albiflorus Airy Shaw BD Trigonostemon wui H.S.Kiu BD Trigonostemon sp. 2 C Trigonostemon cherrieri J.M.Veillon N Trigonostemon inopinatus Airy Shaw O Trigonostemon aurantiacus (Kurz ex Teijsm. & Binn.) Boerl. BEFG Trigonostemon xyphophylloides (Croizat) L.K.Dai & T.L.Wu D Trigonostemon hartleyi Airy Shaw M Trigonostemon philippinensis Stapf BEFHJ Trigonostemon bonianus Gagnep. B Trigonostemon adenocalyx Gagnep. B Trigonostemon lii Y.T.Chang B Ostodes paniculata Blume BDEFG Dimorphocalyx moluccensis Welzen & Oostrum L Dimorphocalyx sp. H Dimorphocalyx malayanus Hook.f. EHJ Dimorphocalyx glabellus Thwaites A Dimorphocalyx beddomei (Benth.) Airy Shaw A Tritaxis gaudichaudii Baill. B Dimorphocalyx ixoroides (C.B.Rob.) Airy Shaw J Dimorphocalyx trichocarpus (Airy Shaw) Welzen & Oostrum H Dimorphocalyx denticulatus Merr. EFHIJ Dimorphocalyx pauciflorus (Merr.) Airy Shaw H Dimorphocalyx muricatus (Hook.f.) Airy Shaw EFH Dimorphocalyx australiensis C.T.White KMO Dimorphocalyx cumingii (Müll. Arg.) Airy Shaw J Taxon . Distribution . Trigonostemon semperflorens (Roxb.) Müll.Arg. C Trigonostemon polyanthus Merr. J Trigonostemon villosus Hook.f. var. borneensis (Merr.) Airy Shaw FHJ Trigonostemon filiformis Quisumb. HJ Trigonostemon detritiferus R.I.Milne H Trigonostemon villosus Hook.f. var. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon cf. filiformis Quisumb. HJ Trigonostemon villosus Hook.f. var. cordatus R.Y.Yu & Welzen H Trigonostemon lychnos (R.I.Milne) R.Y.Yu & Welzen H Trigonostemon villosus Hook.f. var. aff. merrillianus (Airy Shaw) R.Y.Yu & Welzen HIJ Trigonostemon verticillatus (Jack) Pax var. verticillatus BE Trigonostemon flavidus Gagnep. BD Trigonostemon diffusus Merr. H Trigonostemon sp. 1 I Trigonostemon victoriae R.Y.Yu & Welzen I Trigonostemon longipes (Merr.) Merr. J Trigonostemon malaccanus Müll.Arg. BEF Trigonostemon murtonii Craib B Trigonostemon beccarii Ridl. F Trigonostemon pentandrus Pax & K.Hoffm. E Trigonostemon verticillatus (Jack) Pax var. salicifolius (Ridl.) Whitmore E Trigonostemon rufescens Jabl. E Trigonostemon balgooyi R.Y.Yu & Welzen E Trigonostemon villosus Hook.f. var. villosus E Trigonostemon capillipes (Hook.f.) Airy Shaw E Trigonostemon magnificus R.I.Milne F Trigonostemon merrillii Elmer IJ Trigonostemon sandakanensis Jabl. H Trigonostemon longifolius Baill. BCDEFHJ Trigonostemon reidioides (Kurz) Craib B Trigonostemon viridissimus (Kurz) Airy Shaw var. viridissimus BCDEFGHJK Trigonostemon verrucosus J.J.Sm. B Trigonostemon quocensis Gagnep. BCE Trigonostemon laevigatus Müll.Arg. var. laevigatus BEFGHIJ Trigonostemon sp. 3 B Trigonostemon albiflorus Airy Shaw BD Trigonostemon wui H.S.Kiu BD Trigonostemon sp. 2 C Trigonostemon cherrieri J.M.Veillon N Trigonostemon inopinatus Airy Shaw O Trigonostemon aurantiacus (Kurz ex Teijsm. & Binn.) Boerl. BEFG Trigonostemon xyphophylloides (Croizat) L.K.Dai & T.L.Wu D Trigonostemon hartleyi Airy Shaw M Trigonostemon philippinensis Stapf BEFHJ Trigonostemon bonianus Gagnep. B Trigonostemon adenocalyx Gagnep. B Trigonostemon lii Y.T.Chang B Ostodes paniculata Blume BDEFG Dimorphocalyx moluccensis Welzen & Oostrum L Dimorphocalyx sp. H Dimorphocalyx malayanus Hook.f. EHJ Dimorphocalyx glabellus Thwaites A Dimorphocalyx beddomei (Benth.) Airy Shaw A Tritaxis gaudichaudii Baill. B Dimorphocalyx ixoroides (C.B.Rob.) Airy Shaw J Dimorphocalyx trichocarpus (Airy Shaw) Welzen & Oostrum H Dimorphocalyx denticulatus Merr. EFHIJ Dimorphocalyx pauciflorus (Merr.) Airy Shaw H Dimorphocalyx muricatus (Hook.f.) Airy Shaw EFH Dimorphocalyx australiensis C.T.White KMO Dimorphocalyx cumingii (Müll. Arg.) Airy Shaw J Open in new tab The MCC tree from the dated phylogenetic tree was first modified, whereby the redundant taxa (mainly the species that are not part of Trigonostemon or Dimorphocalyx) were removed from the tree and only the same taxa (i.e. 42 species and five varieties of Trigonostemon, 12 species of Dimorphocalyx, Tritaxis gaudichaudii Baill., Ostodes paniculata Blume and Jatropha gossypifolia L.) as in matrix 2 in Yu, Slik & van Welzen (2019) remained. This modified MCC tree was used for the likelihood analyses (DEC) via the package RASP 4.0 (Reconstruct Ancestral State in Phylogenies; Yu et al., 2015). Prior dispersal constraints between the areas in four timeframes (Table 2) were defined based on the geographical distances at the relevant time (Hall, 2002, 2009). An extra Bayesian analysis was performed based on the matrix 2 (Yu, Slik & van Welzen, 2019) via BEAST v.1.10.1 using the same parameters as above (but without calibration points). The last obtained 10 000 trees were used as input in the S-DIVA analyses. A maximum of two or three areas were optimized per node in the DEC analyses, and two to four in the S-DIVA analyses; higher numbers of areas took much longer computation and often gave more ambiguous results (i.e. many possibilities, but each with a low probability and with often tectonically unlikely combinations of areas). Table 2. Prior dispersal constraints between the areas (A‒O, see Fig. 1) for DEC analysis. The dispersal constraints are estimated by the absolute distance (Hall, 2009) between the relevant areas in the concerning time frames. Lower values indicate stronger constraints 10 Mya–present A B C D E F G H I J K L M N O A 0.8 0.8 0.7 0.7 0.6 0.5 0.6 0.4 0.4 0.3 0.1 0.1 0.1 0.1 B 0.6 1 1 1 0.9 0.8 0.9 0.7 0.6 0.6 0.3 0.2 0.1 0.1 C 0.7 1 0.9 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 D 0.5 1 1 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 E 0.5 1 1 1 1 0.9 1 0.7 0.6 0.7 0.3 0.2 0.1 0.1 F 0.4 0.7 0.6 0.6 0.8 1 0.9 0.7 0.6 0.8 0.4 0.3 0.1 0.1 G 0.2 0.5 0.4 0.4 0.6 0.8 0.9 0.6 0.6 0.9 0.5 0.4 0.2 0.2 H 0.5 0.9 0.8 0.8 1 0.8 0.8 0.9 0.9 0.7 0.6 0.4 0.2 0.2 I 0.3 0.7 0.7 0.7 0.7 0.6 0.5 0.8 0.9 0.5 0.4 0.2 0.1 0.1 J 0.3 0.5 0.6 0.6 0.6 0.6 0.6 0.9 0.9 0.6 0.6 0.5 0.3 0.3 K 0.1 0.3 0.3 0.3 0.4 0.4 0.8 0.5 0.4 0.4 0.6 0.6 0.4 0.4 L 0.1 0.2 0.2 0.2 0.3 0.3 0.4 0.4 0.3 0.3 0.4 0.9 0.7 0.7 M 0.1 0.2 0.2 0.2 0.2 0.2 0.4 0.4 0.3 0.3 0.5 0.9 0.8 0.7 N 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.8 0.7 O 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.7 0.7 20–10 Mya 30–20 Mya A B C D E F G H I J K L M N O A 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0 0 0 0 0 B 0 1 1 1 1 0.8 0.9 0.7 0.6 0.1 0.1 0.1 0.1 0.1 C 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 D 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 E 0 1 1 1 1 0.8 1 0.7 0.6 0.1 0.1 0.1 0.1 0.1 F 0 1 1 1 1 1 0.9 0.6 0.7 0.1 0.1 0.1 0.1 0.1 G 0 0.9 0.8 0.8 1 1 0.9 0.5 0.7 0.3 0.1 0.1 0.1 0.1 H 0 1 1 1 0.9 1 1 0.7 0.9 0.1 0.2 0.2 0.1 0.1 I 0 0.7 0.8 0.8 1 1 0.5 0.6 0.8 0.1 0.1 0.1 0.1 0.1 J 0 0.5 0.4 0.4 0.6 0.5 0.7 0.8 0.5 0.1 0.2 0.2 0.1 0.1 K 0 0 0 0 0.5 0.6 0.2 0.1 0 0 0.2 0.2 0.1 0.1 L 0 0 0 0 0 0 0.1 0.1 0 0 0.1 0.8 0.6 0.6 M 0 0 0 0 0 0 0 0 0 0 0 0.1 0.9 0.7 N 0 0 0 0 0 0 0 0 0 0 0 0 1 0.7 O 0 0 0 0 0 0 0 0 0 0 0 0 0.7 0.7 51–30 Mya 10 Mya–present A B C D E F G H I J K L M N O A 0.8 0.8 0.7 0.7 0.6 0.5 0.6 0.4 0.4 0.3 0.1 0.1 0.1 0.1 B 0.6 1 1 1 0.9 0.8 0.9 0.7 0.6 0.6 0.3 0.2 0.1 0.1 C 0.7 1 0.9 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 D 0.5 1 1 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 E 0.5 1 1 1 1 0.9 1 0.7 0.6 0.7 0.3 0.2 0.1 0.1 F 0.4 0.7 0.6 0.6 0.8 1 0.9 0.7 0.6 0.8 0.4 0.3 0.1 0.1 G 0.2 0.5 0.4 0.4 0.6 0.8 0.9 0.6 0.6 0.9 0.5 0.4 0.2 0.2 H 0.5 0.9 0.8 0.8 1 0.8 0.8 0.9 0.9 0.7 0.6 0.4 0.2 0.2 I 0.3 0.7 0.7 0.7 0.7 0.6 0.5 0.8 0.9 0.5 0.4 0.2 0.1 0.1 J 0.3 0.5 0.6 0.6 0.6 0.6 0.6 0.9 0.9 0.6 0.6 0.5 0.3 0.3 K 0.1 0.3 0.3 0.3 0.4 0.4 0.8 0.5 0.4 0.4 0.6 0.6 0.4 0.4 L 0.1 0.2 0.2 0.2 0.3 0.3 0.4 0.4 0.3 0.3 0.4 0.9 0.7 0.7 M 0.1 0.2 0.2 0.2 0.2 0.2 0.4 0.4 0.3 0.3 0.5 0.9 0.8 0.7 N 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.8 0.7 O 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.7 0.7 20–10 Mya 30–20 Mya A B C D E F G H I J K L M N O A 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0 0 0 0 0 B 0 1 1 1 1 0.8 0.9 0.7 0.6 0.1 0.1 0.1 0.1 0.1 C 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 D 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 E 0 1 1 1 1 0.8 1 0.7 0.6 0.1 0.1 0.1 0.1 0.1 F 0 1 1 1 1 1 0.9 0.6 0.7 0.1 0.1 0.1 0.1 0.1 G 0 0.9 0.8 0.8 1 1 0.9 0.5 0.7 0.3 0.1 0.1 0.1 0.1 H 0 1 1 1 0.9 1 1 0.7 0.9 0.1 0.2 0.2 0.1 0.1 I 0 0.7 0.8 0.8 1 1 0.5 0.6 0.8 0.1 0.1 0.1 0.1 0.1 J 0 0.5 0.4 0.4 0.6 0.5 0.7 0.8 0.5 0.1 0.2 0.2 0.1 0.1 K 0 0 0 0 0.5 0.6 0.2 0.1 0 0 0.2 0.2 0.1 0.1 L 0 0 0 0 0 0 0.1 0.1 0 0 0.1 0.8 0.6 0.6 M 0 0 0 0 0 0 0 0 0 0 0 0.1 0.9 0.7 N 0 0 0 0 0 0 0 0 0 0 0 0 1 0.7 O 0 0 0 0 0 0 0 0 0 0 0 0 0.7 0.7 51–30 Mya Open in new tab Table 2. Prior dispersal constraints between the areas (A‒O, see Fig. 1) for DEC analysis. The dispersal constraints are estimated by the absolute distance (Hall, 2009) between the relevant areas in the concerning time frames. Lower values indicate stronger constraints 10 Mya–present A B C D E F G H I J K L M N O A 0.8 0.8 0.7 0.7 0.6 0.5 0.6 0.4 0.4 0.3 0.1 0.1 0.1 0.1 B 0.6 1 1 1 0.9 0.8 0.9 0.7 0.6 0.6 0.3 0.2 0.1 0.1 C 0.7 1 0.9 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 D 0.5 1 1 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 E 0.5 1 1 1 1 0.9 1 0.7 0.6 0.7 0.3 0.2 0.1 0.1 F 0.4 0.7 0.6 0.6 0.8 1 0.9 0.7 0.6 0.8 0.4 0.3 0.1 0.1 G 0.2 0.5 0.4 0.4 0.6 0.8 0.9 0.6 0.6 0.9 0.5 0.4 0.2 0.2 H 0.5 0.9 0.8 0.8 1 0.8 0.8 0.9 0.9 0.7 0.6 0.4 0.2 0.2 I 0.3 0.7 0.7 0.7 0.7 0.6 0.5 0.8 0.9 0.5 0.4 0.2 0.1 0.1 J 0.3 0.5 0.6 0.6 0.6 0.6 0.6 0.9 0.9 0.6 0.6 0.5 0.3 0.3 K 0.1 0.3 0.3 0.3 0.4 0.4 0.8 0.5 0.4 0.4 0.6 0.6 0.4 0.4 L 0.1 0.2 0.2 0.2 0.3 0.3 0.4 0.4 0.3 0.3 0.4 0.9 0.7 0.7 M 0.1 0.2 0.2 0.2 0.2 0.2 0.4 0.4 0.3 0.3 0.5 0.9 0.8 0.7 N 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.8 0.7 O 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.7 0.7 20–10 Mya 30–20 Mya A B C D E F G H I J K L M N O A 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0 0 0 0 0 B 0 1 1 1 1 0.8 0.9 0.7 0.6 0.1 0.1 0.1 0.1 0.1 C 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 D 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 E 0 1 1 1 1 0.8 1 0.7 0.6 0.1 0.1 0.1 0.1 0.1 F 0 1 1 1 1 1 0.9 0.6 0.7 0.1 0.1 0.1 0.1 0.1 G 0 0.9 0.8 0.8 1 1 0.9 0.5 0.7 0.3 0.1 0.1 0.1 0.1 H 0 1 1 1 0.9 1 1 0.7 0.9 0.1 0.2 0.2 0.1 0.1 I 0 0.7 0.8 0.8 1 1 0.5 0.6 0.8 0.1 0.1 0.1 0.1 0.1 J 0 0.5 0.4 0.4 0.6 0.5 0.7 0.8 0.5 0.1 0.2 0.2 0.1 0.1 K 0 0 0 0 0.5 0.6 0.2 0.1 0 0 0.2 0.2 0.1 0.1 L 0 0 0 0 0 0 0.1 0.1 0 0 0.1 0.8 0.6 0.6 M 0 0 0 0 0 0 0 0 0 0 0 0.1 0.9 0.7 N 0 0 0 0 0 0 0 0 0 0 0 0 1 0.7 O 0 0 0 0 0 0 0 0 0 0 0 0 0.7 0.7 51–30 Mya 10 Mya–present A B C D E F G H I J K L M N O A 0.8 0.8 0.7 0.7 0.6 0.5 0.6 0.4 0.4 0.3 0.1 0.1 0.1 0.1 B 0.6 1 1 1 0.9 0.8 0.9 0.7 0.6 0.6 0.3 0.2 0.1 0.1 C 0.7 1 0.9 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 D 0.5 1 1 0.9 0.8 0.7 0.7 0.7 0.7 0.6 0.4 0.2 0.1 0.1 E 0.5 1 1 1 1 0.9 1 0.7 0.6 0.7 0.3 0.2 0.1 0.1 F 0.4 0.7 0.6 0.6 0.8 1 0.9 0.7 0.6 0.8 0.4 0.3 0.1 0.1 G 0.2 0.5 0.4 0.4 0.6 0.8 0.9 0.6 0.6 0.9 0.5 0.4 0.2 0.2 H 0.5 0.9 0.8 0.8 1 0.8 0.8 0.9 0.9 0.7 0.6 0.4 0.2 0.2 I 0.3 0.7 0.7 0.7 0.7 0.6 0.5 0.8 0.9 0.5 0.4 0.2 0.1 0.1 J 0.3 0.5 0.6 0.6 0.6 0.6 0.6 0.9 0.9 0.6 0.6 0.5 0.3 0.3 K 0.1 0.3 0.3 0.3 0.4 0.4 0.8 0.5 0.4 0.4 0.6 0.6 0.4 0.4 L 0.1 0.2 0.2 0.2 0.3 0.3 0.4 0.4 0.3 0.3 0.4 0.9 0.7 0.7 M 0.1 0.2 0.2 0.2 0.2 0.2 0.4 0.4 0.3 0.3 0.5 0.9 0.8 0.7 N 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.8 0.7 O 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.7 0.7 0.7 20–10 Mya 30–20 Mya A B C D E F G H I J K L M N O A 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0 0 0 0 0 B 0 1 1 1 1 0.8 0.9 0.7 0.6 0.1 0.1 0.1 0.1 0.1 C 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 D 0 1 1 1 1 0.7 0.8 0.8 0.5 0.1 0.1 0.1 0.1 0.1 E 0 1 1 1 1 0.8 1 0.7 0.6 0.1 0.1 0.1 0.1 0.1 F 0 1 1 1 1 1 0.9 0.6 0.7 0.1 0.1 0.1 0.1 0.1 G 0 0.9 0.8 0.8 1 1 0.9 0.5 0.7 0.3 0.1 0.1 0.1 0.1 H 0 1 1 1 0.9 1 1 0.7 0.9 0.1 0.2 0.2 0.1 0.1 I 0 0.7 0.8 0.8 1 1 0.5 0.6 0.8 0.1 0.1 0.1 0.1 0.1 J 0 0.5 0.4 0.4 0.6 0.5 0.7 0.8 0.5 0.1 0.2 0.2 0.1 0.1 K 0 0 0 0 0.5 0.6 0.2 0.1 0 0 0.2 0.2 0.1 0.1 L 0 0 0 0 0 0 0.1 0.1 0 0 0.1 0.8 0.6 0.6 M 0 0 0 0 0 0 0 0 0 0 0 0.1 0.9 0.7 N 0 0 0 0 0 0 0 0 0 0 0 0 1 0.7 O 0 0 0 0 0 0 0 0 0 0 0 0 0.7 0.7 51–30 Mya Open in new tab RESULTS Dated molecular phylogenetic tree (FIG. 2, TABLE 3) Table 3. Summary of the dated phylogeny and ancestral areas. Shown from left to right are node numbers, posterior probabilities, mean ages of the nodes, 95% height of the posterior density intervals, DEC reconstructions and relative probability, S-DIVA reconstructions and marginal probability and remarks. In either analysis, only the optimization with the highest probability is shown. When the analyses result in different reconstructions with different maximum numbers of areas, extra results are indicated in the remarks. The abbreviations (A‒O) indicate the areas shown in Figure 1. The cross icon ‘†’ refers to many possibilities Node . Posterior . Age . 95% HPD . DEC . RP . S-DIVA . MP . Remarks . 63 0.71 1.42 3.68–0.8 H 0.76 H 0.96 64 0.38 0.52 1.78–0.09 H 0.85 H 1.0 65 0.95 0.82 2.63–0.28 H 1.0 H 1.0 66 0.18 1.53 4.21–1.11 H 1.0 H 1.0 67 0.14 1.72 4.81–1.66 H 0.82 H 1.0 68 1.0 0.81 2.5–0.11 H 1.0 H 1.0 DEC 3: HIJ=0.52 69 0.52 1.83 5.4–1.9 H 0.62 H 0.98 DEC 3: HJ=0.52 70 0.50 4.27 5.67–2.07 J 0.55 H 0.85 DEC 2: HJ=0.41 71 0.31 5.81 7.97–2.84 H 0.18 CH 0.82 DEC 3: CJ=0.16 72 0.98 2.94 6.1–1.62 B 0.6 B 1.0 73 0.48 5.61 7.55–2.94 BH 0.46 BH 0.99 74 0.17 8.85 8.59–4.2 B 0.62 BH 0.56 75 1.0 2.71 4.19–0.94 I 1.0 I 1.0 76 1.0 3.96 5.9–1.98 IJ 0.62 IJ 0.9 77 0.16 7.67 7.59–3.05 B 0.54 BI 0.47 S-DIVA 3: BEI=BI=BFI=0.14; 4: BEI=BI=BFI=BEFI=0.09 78 0.79 9.11 8.03–3.36 B 0.88 BI 0.39 79 0.52 9.23 9.98–5.17 B 0.52 B 0.28 80 0.94 2.28 2.68–0.37 E 1.0 E 1.0 81 1.0 3.25 3.82–0.82 E 1.0 E 1.0 82 1.0 4.43 6.36–1.99 EF 0.59 EF 1.0 83 0.66 5.63 6.91–1.8 E 1.0 E 1.0 84 0.67 7.11 8.47–3.34 E 0.87 E 1.0 85 0.27 10.38 11.83–6.44 BE 0.47 BE 0.4 86 1.0 11.06 12.66–6.81 BE 0.46 BE 0.66 87 1.0 12.9 16.05–8.22 BE 0.33 BF 0.66 DEC 3: B=0.25; S-DIVA 3: BEF=BF=0.25; S-DIVA 4: BEF=BF=0.18 88 1.0 3.46 8.75–1.95 H 1.0 BH=H=HJ 0.33 S-DIVA 3: †=0.08 all with H; 4: unable to calculate 89 1.0 6.46 11.91–3.75 HI 0.23 HJ 0.2 S-DIVA 3: †=0.06; 4: unable to calculate 90 1.0 14.25 18.93–8.27 B 0.72 BH=B=BJ=BI 0.25 S-DIVA 3: †=0.14 all with B; 4: † all with B 91 0.96 18.02 23.96–13.73 B 1.0 B 0.57 92 0.95 7.84 11.93–4.78 B 1.0 B 1.0 93 1.0 4.4 10.3–2.97 B 0.76 B 1.0 94 1.0 10.87 13.8–6.48 B 1.0 B 1.0 95 1.0 0.84 3.02–0.09 BD 0.83 B 1.0 96 1.0 5.95 11.93–3.79 BC 0.43 BC 1.0 DEC 2: B=0.44 97 0.52 9.91 14.31–6.43 B 1.0 B 1.0 98 0.99 8.42 11.8–3.17 O 0.28 NO 1.0 99 0.98 10.56 16.28–8.56 B 0.37 B 0.69 100 1.0 13.46 18.42–10.66 B 1.0 B 1.0 101 1.0 7.35 11.33–2.97 D 0.58 BD 1.0 S-DIVA 3: BDE=BDG=BD=BDF=0.25; 4: † 0.14 all with BD 102 0.99 10.87 15.22–5.96 B 0.47 BM 1.0 DEC 3: M=0.42; S-DIVA 3: †=0.2, all with BM, 4: †=0.09, all with BM 103 1.0 12.31 17.93–9.23 B 0.43 B 1.0 104 1.0 16.75 21.14–12.92 B 1.0 B 1.0 105 1.0 2.14 1.93–0.04 B 1.0 B 1.0 106 1.0 5.06 6.9–1.15 B 1.0 B 1.0 107 0.99 17.82 24.06–14.65 B 1.0 B 1.0 108 1.0 18.8 26.35–16.33 B 1.0 B 0.97 Divergence of Trigonostemon 109 0.16 25.77 35.82–22.87 B 1.0 B 0.96 110 0.96 3.35 6.22–0.03 HL 0.82 HL 1.0 111 1.0 0.23 1.55–0 A 1.0 A 1.0 112 0.96 2.78 8.27–2.36 A 0.56 AH 0.99 DEC 3: AEH=0.34; S-DIVA 3: AH=AEH=AHJ=0.33; 4: AH=AHJ=AEH=AEHJ=0.25 113 0.86 6.01 10.02–3.65 H 0.62 BH 0.99 S-DIVA 3: ABH=BHJ=BH=BEH=0.25; 4: †=0.14 all with BH 114 0.86 5.19 5.19–0.99 H 1.0 H 0.99 115 0.97 6.36 6.89–1.75 H 0.8 HJ 0.92 116 0.88 6.97 11.19–4.61 H 0.71 H 0.9 117 1.0 10.02 13.66–5.58 H 1.0 H 0.98 118 1.0 1.75 8.9–0.92 H 1.0 H 1.0 DEC 3: EFH=0.59 119 1.0 10.69 20.06–8.49 H 0.84 H 0.99 120 1.0 12.78 14.43–2.74 J 0.5 JM=JO=JK 0.33 S-DIVA 3: †=0.17 all with J; 4: †=0.14 all with J 121 0.99 23.72 30.45–13.54 H 0.38 HM=HO=HK=HJ 0.25 Divergence of Dimorphocalyx; S-DIVA 3: †=0.1 all with H; 4: †=0.07 all with H Node . Posterior . Age . 95% HPD . DEC . RP . S-DIVA . MP . Remarks . 63 0.71 1.42 3.68–0.8 H 0.76 H 0.96 64 0.38 0.52 1.78–0.09 H 0.85 H 1.0 65 0.95 0.82 2.63–0.28 H 1.0 H 1.0 66 0.18 1.53 4.21–1.11 H 1.0 H 1.0 67 0.14 1.72 4.81–1.66 H 0.82 H 1.0 68 1.0 0.81 2.5–0.11 H 1.0 H 1.0 DEC 3: HIJ=0.52 69 0.52 1.83 5.4–1.9 H 0.62 H 0.98 DEC 3: HJ=0.52 70 0.50 4.27 5.67–2.07 J 0.55 H 0.85 DEC 2: HJ=0.41 71 0.31 5.81 7.97–2.84 H 0.18 CH 0.82 DEC 3: CJ=0.16 72 0.98 2.94 6.1–1.62 B 0.6 B 1.0 73 0.48 5.61 7.55–2.94 BH 0.46 BH 0.99 74 0.17 8.85 8.59–4.2 B 0.62 BH 0.56 75 1.0 2.71 4.19–0.94 I 1.0 I 1.0 76 1.0 3.96 5.9–1.98 IJ 0.62 IJ 0.9 77 0.16 7.67 7.59–3.05 B 0.54 BI 0.47 S-DIVA 3: BEI=BI=BFI=0.14; 4: BEI=BI=BFI=BEFI=0.09 78 0.79 9.11 8.03–3.36 B 0.88 BI 0.39 79 0.52 9.23 9.98–5.17 B 0.52 B 0.28 80 0.94 2.28 2.68–0.37 E 1.0 E 1.0 81 1.0 3.25 3.82–0.82 E 1.0 E 1.0 82 1.0 4.43 6.36–1.99 EF 0.59 EF 1.0 83 0.66 5.63 6.91–1.8 E 1.0 E 1.0 84 0.67 7.11 8.47–3.34 E 0.87 E 1.0 85 0.27 10.38 11.83–6.44 BE 0.47 BE 0.4 86 1.0 11.06 12.66–6.81 BE 0.46 BE 0.66 87 1.0 12.9 16.05–8.22 BE 0.33 BF 0.66 DEC 3: B=0.25; S-DIVA 3: BEF=BF=0.25; S-DIVA 4: BEF=BF=0.18 88 1.0 3.46 8.75–1.95 H 1.0 BH=H=HJ 0.33 S-DIVA 3: †=0.08 all with H; 4: unable to calculate 89 1.0 6.46 11.91–3.75 HI 0.23 HJ 0.2 S-DIVA 3: †=0.06; 4: unable to calculate 90 1.0 14.25 18.93–8.27 B 0.72 BH=B=BJ=BI 0.25 S-DIVA 3: †=0.14 all with B; 4: † all with B 91 0.96 18.02 23.96–13.73 B 1.0 B 0.57 92 0.95 7.84 11.93–4.78 B 1.0 B 1.0 93 1.0 4.4 10.3–2.97 B 0.76 B 1.0 94 1.0 10.87 13.8–6.48 B 1.0 B 1.0 95 1.0 0.84 3.02–0.09 BD 0.83 B 1.0 96 1.0 5.95 11.93–3.79 BC 0.43 BC 1.0 DEC 2: B=0.44 97 0.52 9.91 14.31–6.43 B 1.0 B 1.0 98 0.99 8.42 11.8–3.17 O 0.28 NO 1.0 99 0.98 10.56 16.28–8.56 B 0.37 B 0.69 100 1.0 13.46 18.42–10.66 B 1.0 B 1.0 101 1.0 7.35 11.33–2.97 D 0.58 BD 1.0 S-DIVA 3: BDE=BDG=BD=BDF=0.25; 4: † 0.14 all with BD 102 0.99 10.87 15.22–5.96 B 0.47 BM 1.0 DEC 3: M=0.42; S-DIVA 3: †=0.2, all with BM, 4: †=0.09, all with BM 103 1.0 12.31 17.93–9.23 B 0.43 B 1.0 104 1.0 16.75 21.14–12.92 B 1.0 B 1.0 105 1.0 2.14 1.93–0.04 B 1.0 B 1.0 106 1.0 5.06 6.9–1.15 B 1.0 B 1.0 107 0.99 17.82 24.06–14.65 B 1.0 B 1.0 108 1.0 18.8 26.35–16.33 B 1.0 B 0.97 Divergence of Trigonostemon 109 0.16 25.77 35.82–22.87 B 1.0 B 0.96 110 0.96 3.35 6.22–0.03 HL 0.82 HL 1.0 111 1.0 0.23 1.55–0 A 1.0 A 1.0 112 0.96 2.78 8.27–2.36 A 0.56 AH 0.99 DEC 3: AEH=0.34; S-DIVA 3: AH=AEH=AHJ=0.33; 4: AH=AHJ=AEH=AEHJ=0.25 113 0.86 6.01 10.02–3.65 H 0.62 BH 0.99 S-DIVA 3: ABH=BHJ=BH=BEH=0.25; 4: †=0.14 all with BH 114 0.86 5.19 5.19–0.99 H 1.0 H 0.99 115 0.97 6.36 6.89–1.75 H 0.8 HJ 0.92 116 0.88 6.97 11.19–4.61 H 0.71 H 0.9 117 1.0 10.02 13.66–5.58 H 1.0 H 0.98 118 1.0 1.75 8.9–0.92 H 1.0 H 1.0 DEC 3: EFH=0.59 119 1.0 10.69 20.06–8.49 H 0.84 H 0.99 120 1.0 12.78 14.43–2.74 J 0.5 JM=JO=JK 0.33 S-DIVA 3: †=0.17 all with J; 4: †=0.14 all with J 121 0.99 23.72 30.45–13.54 H 0.38 HM=HO=HK=HJ 0.25 Divergence of Dimorphocalyx; S-DIVA 3: †=0.1 all with H; 4: †=0.07 all with H Open in new tab Table 3. Summary of the dated phylogeny and ancestral areas. Shown from left to right are node numbers, posterior probabilities, mean ages of the nodes, 95% height of the posterior density intervals, DEC reconstructions and relative probability, S-DIVA reconstructions and marginal probability and remarks. In either analysis, only the optimization with the highest probability is shown. When the analyses result in different reconstructions with different maximum numbers of areas, extra results are indicated in the remarks. The abbreviations (A‒O) indicate the areas shown in Figure 1. The cross icon ‘†’ refers to many possibilities Node . Posterior . Age . 95% HPD . DEC . RP . S-DIVA . MP . Remarks . 63 0.71 1.42 3.68–0.8 H 0.76 H 0.96 64 0.38 0.52 1.78–0.09 H 0.85 H 1.0 65 0.95 0.82 2.63–0.28 H 1.0 H 1.0 66 0.18 1.53 4.21–1.11 H 1.0 H 1.0 67 0.14 1.72 4.81–1.66 H 0.82 H 1.0 68 1.0 0.81 2.5–0.11 H 1.0 H 1.0 DEC 3: HIJ=0.52 69 0.52 1.83 5.4–1.9 H 0.62 H 0.98 DEC 3: HJ=0.52 70 0.50 4.27 5.67–2.07 J 0.55 H 0.85 DEC 2: HJ=0.41 71 0.31 5.81 7.97–2.84 H 0.18 CH 0.82 DEC 3: CJ=0.16 72 0.98 2.94 6.1–1.62 B 0.6 B 1.0 73 0.48 5.61 7.55–2.94 BH 0.46 BH 0.99 74 0.17 8.85 8.59–4.2 B 0.62 BH 0.56 75 1.0 2.71 4.19–0.94 I 1.0 I 1.0 76 1.0 3.96 5.9–1.98 IJ 0.62 IJ 0.9 77 0.16 7.67 7.59–3.05 B 0.54 BI 0.47 S-DIVA 3: BEI=BI=BFI=0.14; 4: BEI=BI=BFI=BEFI=0.09 78 0.79 9.11 8.03–3.36 B 0.88 BI 0.39 79 0.52 9.23 9.98–5.17 B 0.52 B 0.28 80 0.94 2.28 2.68–0.37 E 1.0 E 1.0 81 1.0 3.25 3.82–0.82 E 1.0 E 1.0 82 1.0 4.43 6.36–1.99 EF 0.59 EF 1.0 83 0.66 5.63 6.91–1.8 E 1.0 E 1.0 84 0.67 7.11 8.47–3.34 E 0.87 E 1.0 85 0.27 10.38 11.83–6.44 BE 0.47 BE 0.4 86 1.0 11.06 12.66–6.81 BE 0.46 BE 0.66 87 1.0 12.9 16.05–8.22 BE 0.33 BF 0.66 DEC 3: B=0.25; S-DIVA 3: BEF=BF=0.25; S-DIVA 4: BEF=BF=0.18 88 1.0 3.46 8.75–1.95 H 1.0 BH=H=HJ 0.33 S-DIVA 3: †=0.08 all with H; 4: unable to calculate 89 1.0 6.46 11.91–3.75 HI 0.23 HJ 0.2 S-DIVA 3: †=0.06; 4: unable to calculate 90 1.0 14.25 18.93–8.27 B 0.72 BH=B=BJ=BI 0.25 S-DIVA 3: †=0.14 all with B; 4: † all with B 91 0.96 18.02 23.96–13.73 B 1.0 B 0.57 92 0.95 7.84 11.93–4.78 B 1.0 B 1.0 93 1.0 4.4 10.3–2.97 B 0.76 B 1.0 94 1.0 10.87 13.8–6.48 B 1.0 B 1.0 95 1.0 0.84 3.02–0.09 BD 0.83 B 1.0 96 1.0 5.95 11.93–3.79 BC 0.43 BC 1.0 DEC 2: B=0.44 97 0.52 9.91 14.31–6.43 B 1.0 B 1.0 98 0.99 8.42 11.8–3.17 O 0.28 NO 1.0 99 0.98 10.56 16.28–8.56 B 0.37 B 0.69 100 1.0 13.46 18.42–10.66 B 1.0 B 1.0 101 1.0 7.35 11.33–2.97 D 0.58 BD 1.0 S-DIVA 3: BDE=BDG=BD=BDF=0.25; 4: † 0.14 all with BD 102 0.99 10.87 15.22–5.96 B 0.47 BM 1.0 DEC 3: M=0.42; S-DIVA 3: †=0.2, all with BM, 4: †=0.09, all with BM 103 1.0 12.31 17.93–9.23 B 0.43 B 1.0 104 1.0 16.75 21.14–12.92 B 1.0 B 1.0 105 1.0 2.14 1.93–0.04 B 1.0 B 1.0 106 1.0 5.06 6.9–1.15 B 1.0 B 1.0 107 0.99 17.82 24.06–14.65 B 1.0 B 1.0 108 1.0 18.8 26.35–16.33 B 1.0 B 0.97 Divergence of Trigonostemon 109 0.16 25.77 35.82–22.87 B 1.0 B 0.96 110 0.96 3.35 6.22–0.03 HL 0.82 HL 1.0 111 1.0 0.23 1.55–0 A 1.0 A 1.0 112 0.96 2.78 8.27–2.36 A 0.56 AH 0.99 DEC 3: AEH=0.34; S-DIVA 3: AH=AEH=AHJ=0.33; 4: AH=AHJ=AEH=AEHJ=0.25 113 0.86 6.01 10.02–3.65 H 0.62 BH 0.99 S-DIVA 3: ABH=BHJ=BH=BEH=0.25; 4: †=0.14 all with BH 114 0.86 5.19 5.19–0.99 H 1.0 H 0.99 115 0.97 6.36 6.89–1.75 H 0.8 HJ 0.92 116 0.88 6.97 11.19–4.61 H 0.71 H 0.9 117 1.0 10.02 13.66–5.58 H 1.0 H 0.98 118 1.0 1.75 8.9–0.92 H 1.0 H 1.0 DEC 3: EFH=0.59 119 1.0 10.69 20.06–8.49 H 0.84 H 0.99 120 1.0 12.78 14.43–2.74 J 0.5 JM=JO=JK 0.33 S-DIVA 3: †=0.17 all with J; 4: †=0.14 all with J 121 0.99 23.72 30.45–13.54 H 0.38 HM=HO=HK=HJ 0.25 Divergence of Dimorphocalyx; S-DIVA 3: †=0.1 all with H; 4: †=0.07 all with H Node . Posterior . Age . 95% HPD . DEC . RP . S-DIVA . MP . Remarks . 63 0.71 1.42 3.68–0.8 H 0.76 H 0.96 64 0.38 0.52 1.78–0.09 H 0.85 H 1.0 65 0.95 0.82 2.63–0.28 H 1.0 H 1.0 66 0.18 1.53 4.21–1.11 H 1.0 H 1.0 67 0.14 1.72 4.81–1.66 H 0.82 H 1.0 68 1.0 0.81 2.5–0.11 H 1.0 H 1.0 DEC 3: HIJ=0.52 69 0.52 1.83 5.4–1.9 H 0.62 H 0.98 DEC 3: HJ=0.52 70 0.50 4.27 5.67–2.07 J 0.55 H 0.85 DEC 2: HJ=0.41 71 0.31 5.81 7.97–2.84 H 0.18 CH 0.82 DEC 3: CJ=0.16 72 0.98 2.94 6.1–1.62 B 0.6 B 1.0 73 0.48 5.61 7.55–2.94 BH 0.46 BH 0.99 74 0.17 8.85 8.59–4.2 B 0.62 BH 0.56 75 1.0 2.71 4.19–0.94 I 1.0 I 1.0 76 1.0 3.96 5.9–1.98 IJ 0.62 IJ 0.9 77 0.16 7.67 7.59–3.05 B 0.54 BI 0.47 S-DIVA 3: BEI=BI=BFI=0.14; 4: BEI=BI=BFI=BEFI=0.09 78 0.79 9.11 8.03–3.36 B 0.88 BI 0.39 79 0.52 9.23 9.98–5.17 B 0.52 B 0.28 80 0.94 2.28 2.68–0.37 E 1.0 E 1.0 81 1.0 3.25 3.82–0.82 E 1.0 E 1.0 82 1.0 4.43 6.36–1.99 EF 0.59 EF 1.0 83 0.66 5.63 6.91–1.8 E 1.0 E 1.0 84 0.67 7.11 8.47–3.34 E 0.87 E 1.0 85 0.27 10.38 11.83–6.44 BE 0.47 BE 0.4 86 1.0 11.06 12.66–6.81 BE 0.46 BE 0.66 87 1.0 12.9 16.05–8.22 BE 0.33 BF 0.66 DEC 3: B=0.25; S-DIVA 3: BEF=BF=0.25; S-DIVA 4: BEF=BF=0.18 88 1.0 3.46 8.75–1.95 H 1.0 BH=H=HJ 0.33 S-DIVA 3: †=0.08 all with H; 4: unable to calculate 89 1.0 6.46 11.91–3.75 HI 0.23 HJ 0.2 S-DIVA 3: †=0.06; 4: unable to calculate 90 1.0 14.25 18.93–8.27 B 0.72 BH=B=BJ=BI 0.25 S-DIVA 3: †=0.14 all with B; 4: † all with B 91 0.96 18.02 23.96–13.73 B 1.0 B 0.57 92 0.95 7.84 11.93–4.78 B 1.0 B 1.0 93 1.0 4.4 10.3–2.97 B 0.76 B 1.0 94 1.0 10.87 13.8–6.48 B 1.0 B 1.0 95 1.0 0.84 3.02–0.09 BD 0.83 B 1.0 96 1.0 5.95 11.93–3.79 BC 0.43 BC 1.0 DEC 2: B=0.44 97 0.52 9.91 14.31–6.43 B 1.0 B 1.0 98 0.99 8.42 11.8–3.17 O 0.28 NO 1.0 99 0.98 10.56 16.28–8.56 B 0.37 B 0.69 100 1.0 13.46 18.42–10.66 B 1.0 B 1.0 101 1.0 7.35 11.33–2.97 D 0.58 BD 1.0 S-DIVA 3: BDE=BDG=BD=BDF=0.25; 4: † 0.14 all with BD 102 0.99 10.87 15.22–5.96 B 0.47 BM 1.0 DEC 3: M=0.42; S-DIVA 3: †=0.2, all with BM, 4: †=0.09, all with BM 103 1.0 12.31 17.93–9.23 B 0.43 B 1.0 104 1.0 16.75 21.14–12.92 B 1.0 B 1.0 105 1.0 2.14 1.93–0.04 B 1.0 B 1.0 106 1.0 5.06 6.9–1.15 B 1.0 B 1.0 107 0.99 17.82 24.06–14.65 B 1.0 B 1.0 108 1.0 18.8 26.35–16.33 B 1.0 B 0.97 Divergence of Trigonostemon 109 0.16 25.77 35.82–22.87 B 1.0 B 0.96 110 0.96 3.35 6.22–0.03 HL 0.82 HL 1.0 111 1.0 0.23 1.55–0 A 1.0 A 1.0 112 0.96 2.78 8.27–2.36 A 0.56 AH 0.99 DEC 3: AEH=0.34; S-DIVA 3: AH=AEH=AHJ=0.33; 4: AH=AHJ=AEH=AEHJ=0.25 113 0.86 6.01 10.02–3.65 H 0.62 BH 0.99 S-DIVA 3: ABH=BHJ=BH=BEH=0.25; 4: †=0.14 all with BH 114 0.86 5.19 5.19–0.99 H 1.0 H 0.99 115 0.97 6.36 6.89–1.75 H 0.8 HJ 0.92 116 0.88 6.97 11.19–4.61 H 0.71 H 0.9 117 1.0 10.02 13.66–5.58 H 1.0 H 0.98 118 1.0 1.75 8.9–0.92 H 1.0 H 1.0 DEC 3: EFH=0.59 119 1.0 10.69 20.06–8.49 H 0.84 H 0.99 120 1.0 12.78 14.43–2.74 J 0.5 JM=JO=JK 0.33 S-DIVA 3: †=0.17 all with J; 4: †=0.14 all with J 121 0.99 23.72 30.45–13.54 H 0.38 HM=HO=HK=HJ 0.25 Divergence of Dimorphocalyx; S-DIVA 3: †=0.1 all with H; 4: †=0.07 all with H Open in new tab The molecular dating analysis (Fig. 2) gave a similar phylogenetic tree to that in Yu, Slik & van Welzen (2019). Compared to the results of MrBayes (fig. 3 in Yu, Slik & van Welzen, 2019), the MCC tree generated by BEAST was more similar to the results from the maximum likelihood (ML) analyses (suppl. fig. S6 in Yu, Slik & van Welzen, 2019): three species (Trigonostemon adenocalyx Gagnep., T. bonianus Gagnep. and T. lii Y.T.Chang; distributed in south-western China and Indochina) were placed at the base in the T3 clade (in the BI analysis via MrBayes they were one node higher in the clade), and T. magnificus R.I.Milne (endemic to Sumatra) was at the base of section Trigonostemon (with MrBayes it was sister to T. beccarii Ridl., also endemic to Sumatra). Figure 2. Open in new tabDownload slide Chronogram (MCC tree) for Trigonostemon and Dimorphocalyx generated by a dated Bayesian analysis using BEAST. The redundant taxa in the analysis were removed. The posterior probabilities of important nodes are shown below the relevant branches. The blue node bars indicate the 95% height of the posterior density. Dashed lines mean the branch length has been modified. Trigonostemon was estimated to have diverged from other crotonoids with inaperturate pollen between 25.77 [95% highest posterior density interval (HPD): 35.82–22.87] Mya and 18.8 (26.35–16.33) Mya. The ages of the crown group of sections Trigonostemon, Spinipollen R.Y.Yu & Welzen and Tylosepalum (Kurz ex Teijsm. & Binn.) Pax & K.Hoffm. were estimated to be 12.90 (16.05–8.22) Mya, 14.25 (18.93–8.27) Mya and 17.82 (24.06–14.65) Mya, respectively. The divergence times of the extant taxa in section Trigonostemon were often relatively younger than those in the other two sections. More than half of the extant taxa of section Trigonostemon were estimated to have probably diverged during the Pleistocene (i.e. lower bound of HPD younger than 2.58 Myr). The divergence of Dimorphocalyx from other inaperturate crotonoids was inferred between 29.04 (38.11–24.11) Mya and 23.72 (30.45–13.54) Mya. Historical biogeography (FIG. 3A, B, TABLE 3) The most probable ancestral area of Trigonostemon (node 108, Figs 2, 3B) was inferred to be the South-East Asian mainland in both DEC (relative probability, RP = 1.0) and S-DIVA (marginal probability, MP = 0.97) analyses. Section Tylosepalum (crown node 107, Figs 2, 3B) diverged first but still remained on the South-East Asian mainland (RP = 1.0, MP = 1.0), where it further diversified. Some species then radiated or dispersed eastwards and southwards, resulting in a wider distribution. Two long-distance dispersal events involved crossing Wallace’s line [between nodes 99 and 98, and between nodes 103 and 102 (S-DIVA) or between node 102 and T. hartleyi Airy Shaw (DEC), Figs 2, 3B], but it is unclear via which route these dispersals happened. The DEC analysis (three maximum areas) also indicated a westwards dispersal event from New Guinea to west Malesia (from node 102 to T. philippinensis Stapf, Figs 2, 3B), but in the S-DIVA analysis, it was more likely to be a vicariance event. Section Spinipollen diverged on the South-East Asian mainland (node 90, Figs 2, 3A, RP = 0.72, four possibilities in S-DIVA, but all containing the South-East Asian mainland) and dispersed to Borneo and the Philippines (either Palawan or the Philippine islands, but both with low support; node 89, Figs 2, 3A, RP = 0.23, MP = 0.2). A possible vicariance event followed, leading to the split of T. merrillii Elmer and the ancestral species of T. longifolius Baill. and T. sandakanensis Jabl. Figure 3A. Open in new tabDownload slide Open in new tabDownload slide Ancestral area reconstructions for Trigonostemon and Dimorphocalyx. Distributions of taxa are shown at the end of the branches (see legend for Fig. 1 for abbreviations). Only optimizations with the highest probability are shown in the figure (if the optimizations are different in analyses with different numbers of maximum areas). Results from the DEC analyses are shown at each node; when S-DIVA analyses yield a different optimization, the results are shown next to the relevant nodes. Blue arrows refer to dispersals from west to east; red arrows refer to dispersals from east to west. Section Trigonostemon radiated to the Malay Peninsula (node 87, Figs 2, 3A, RP = 0.33) or Sumatra (node 87, MP = 0.66; but this was less likely because node 87 represents a major difference in the results between BEAST and MrBayes, see above; the Malay Peninsula would be the only optimized area if the analyses were based on cladograms obtained using MrBayes and not BEAST; not shown). After T. magnificus and T. capillipes (Hook.f.) Airy Shaw split off, the crown group remained on the South-East Asian mainland and Malay Peninsula (node 85, Figs 2, 3A, RP = 0.47, MP = 0.4). After a vicariance event, a lineage (crown node 84, Figs 2, 3A, RP = 0.87, MP = 1.0) diverged and diversified on the Malay Peninsula; the other lineage (three clades can be recognized) dispersed (probably through Palawan, node 77, Figs 2, 3A, MP = 0.47) to the Philippines and to Borneo. Rapid speciation events in this section were inferred to have occurred on the Malay Peninsula and Borneo (i.e. nodes 63–70 and 80–84, Figs 2, 3A). The posterior probabilities (PP) in section Trigonostemon (nodes 63–86, Figs 2, 3A) were relatively low, and the ancestral area reconstructions may be less definite (discussed below). Borneo was inferred as the most probable ancestral area of Dimorphocalyx (node 121, Figs 2, 3B, RP = 0.38, four possibilities in S-DIVA, but all containing Borneo). A lineage dispersed (node 121 to 120, Figs 2, 3B) first to the Philippine islands (node 120, Figs 2, 3B, RP = 0.5, three possibilities in S-DIVA, but all containing the Philippines; node age 12.78 [14.43–2.74] Ma) and then to eastern Malesia and north-eastern Australia (D. australiensis C.T.White). The other lineage remained on Borneo and diversified. The ancestral species of Dimorphocalyx sp. and D. moluccensis Welzen & Oostrum (node 110, Figs 2, 3B) reached the north Moluccan islands through radiation at 10.02–3.35 Mya, and this was followed by a vicariance event, giving rise to these two species. Around 6.01–2.78 Mya, a lineage dispersed to Southern India (including Sri Lanka) probably through the South-East Asian mainland (discussed below). Other modern Dimorphocalyx spp. were the result of radiation or dispersal in various directions from Borneo after 7 Mya. DISCUSSION Tree compatibility The condensed (MCC) tree and the 10 000 background trees used in the DEC and S-DIVA analyses were generated based on two different data matrixes. Because direct fossil records are lacking, the estimation of the divergence times in Trigonostemon and Dimorphocalyx was realized using other taxa of Euphorbiaceae. We therefore used a combined dataset (matrix 1 from Yu, Slik & van Welzen, 2019), in which our molecular data for Trigonostemon and Dimorphocalyx are embedded into the dataset for Euphorbiaceae (Wurdack et al., 2005), for a dated phylogenetic analysis. Taxa irrelevant to the analysis of Trigonostemon and Dimorphocalyx were then removed from the MCC tree, and only the same taxa as in matrix 2 (Yu, Slik & van Welzen, 2019) remained. The background trees for the S-DIVA analysis, however, are based on our molecular data only (i.e. matrix 2 of Yu, Slik & van Welzen, 2019). The MCC tree (Fig. 1) and the background trees (the 50% consensus tree is shown in fig. 3 in Yu, Slik & van Welzen, 2019) have the same topology for the main clades, and the ancestral areas reconstructed by the DEC and S-DIVA methods are highly similar. Matrix 1 includes eight additional rbcL sequences compared with matrix 2, but the marker contains little variation in Trigonostemon (Yu, Slik & van Welzen, 2019); it only helps to determine the phylogenetic position of Trigonostemon in Euphorbiaceae, and it has almost no impact on the internal relationships in the genus. Therefore, it is reasonable to display the probabilities calculated based on the 10 000 background trees in the S-DIVA analysis on the MCC tree. Using this strategy, we also avoided an unnecessary secondary Bayesian clock analysis. The divergence times were already inferred by indirect calibration points, a secondary dating phylogeny would have caused more bias. Because of the missing data (see Appendix 1 in Yu, Slik & van Welzen, 2019) and high speciation rates (discussed below) of the taxa in section Trigonostemon (i.e. nodes 63–86, Figs 2, 3A), several nodes have a relatively low (< 0.9) PP in the dated molecular phylogenetic tree. This may affect the credibility of the ancestral area reconstructions of these clades. However, the four main lineages (i.e. crown nodes 71, 73, 77, 84, Figs 2, 3B) appear relatively stable in the phylogenetic trees reconstructed by different methods [maximum parsimony (MP), ML and BI; see fig. 3 and suppl. figs S5, S6 in Yu, Slik & van Welzen, 2019]; the taxa more or less remained in the same lineage, although the position might slightly change within the lineages. Moreover, most area reconstructions in these three lineages have high relative and marginal probabilities; therefore, they are still considered credible. In addition, the phylogenetic relationships of Trigonostemon, Dimorphocalyx and other genera in the C2 clade of crotonoids are still unsolved (fig. 2 in Yu, Slik & van Welzen, 2019). Jatropha L. was used as the outgroup and was assigned a distribution of all areas. This may have made the reconstructions of the crown nodes (ancestral areas) of Trigonostemon and Dimorphocalyx less realistic, but the analyses gave strong signals for both crown nodes (for Trigonostemon, node 108, RP = 1.0, MP = 0.97; for Dimorphocalyx, node 121, RP = 0.38, four possibilities in S-DIVA, all containing Borneo, total MP = 1.0), which were reconstructed based on the internal taxa. Therefore, these reconstructions can be trusted. Migration routes Trigonostemon is inferred to have originated on the geologically old South-East Asian mainland (it has formed a promontory of the Eurasian continent since the Early Mesozoic; Hall, 2009). Historical migrations mainly occurred in section Trigonostemon; these: (1) radiated to the adjacent Malay Peninsula at c. 18.02–12.90 Mya (node 91 to 87; then to 85, Figs 2, 3A); and (2) migrated eastwards from the South-East Asian mainland to Borneo [occurring in two linages: 1. from node 79 to 74 (S-DIVA) or from node 74 to 71 (DEC); 2. from node 74 to 73] and to the Philippines [occurring in one lineage, possibly through Palawan: from node 79 to 78 (S-DIVA) or from node 77 to 76 (DEC)] c. 9.23–3.96 Mya (Figs 2, 3A). The former is easy to explain: the Malay Peninsula has always been connected to South-East Asia, and the ever-wet climate in this part, also during the Pleistocene, surely worked to the advantage of the plants (11 extant species in the Malay Peninsula; Yu & van Welzen, 2018). This also appears to be a commonly used route [vs. a less common route through Taiwan to the Philippines, e.g. Matuszak et al., 2015; Condamine et al., 2013 (butterflies)] for plants of Asian origin to colonize parts of Malesia (e.g. Thomas et al., 2012, Denduangboripant et al., 2001, Ridder-Numan, 1998; or the other way round, e.g. Haegens, 2000, Grudinski et al., 2014). In the latter case, the radiation to Borneo was possible as the land was connected to the Malay Peninsula before at least 5 Mya (Hall, 2009), and a similar scenario has also been recorded (Su & Saunders, 2009); the dispersal (or radiation) via Palawan appears to be a new route: Palawan may have been a stepping stone for the dispersal (node 78 in S-DIVA), but no collection data is available for the islands to the west (e.g. the Spratly Islands); alternatively, Palawan, as the only microplate that moved from Asia to the east (starting c. 30 Mya), may also have acted as an ‘Ark of Noah’, transporting ancestral species of Trigonostemon to the Philippines. The results also indicate a possible dispersal from the South-East Asian mainland to Borneo and the Philippines in section Spinipollen c. 14.25–6.46 Mya (from node 90 to 89, Figs 2, 3A), but no concrete conclusions can be made, because the support values for node 89 are low (RP = 0.23, MP = 0.2). Dimorphocalyx probably originated c. 29.04–23.72 Mya on Borneo. This was followed by dispersal of one lineage from Borneo to the Philippines (node 121 to 120, Figs 2, 3B) c. 23.72–12.78 Mya. Similar dispersals have been recorded in several case studies [e.g. Nauheimer et al., 2012, Thomas et al., 2012, Su & Saunders, 2009, Sheldon et al., 2012 (birds)], but these mostly occurred < 12 Mya. Our results indicate that the exchange was already possible in an earlier time, probably up to the Early Miocene, when the Sulu Archipelago and Tawi-Tawi islands were already in place (Hall, 2009). The dispersal from Borneo to southern India and Sri Lanka (node 113 to 112, Figs 2, 3B) at 6.01–2.78 Mya is somewhat difficult to explain: it is less likely that the plant dispersed (e.g. via birds) directly from western Malesia to southern India, because the geographical distance is too long, but at that time the Indian plate had partly collided with the South-East Asian mainland (Patriat & Achache, 1984), and long-distance dispersals between Borneo and southern India/Sri Lanka (via land, wind, water, birds etc.) were indeed possible (e.g. Repetur, van Welzen & de Vogel, 1997; Sumatra was then mainly below water and had no impact on the dispersal; Hall, 2009). Long-distance dispersal may sound like an erratic occurrence, but the shape and size of India prior to its collision with South-East Asia are unknown; northern India may have been broad and large, thus greatly reducing the distance with Borneo. Alternatively, the South-East Asian mainland may have played a role, as before the plants reached Southern India (i.e. one node higher, at node 113, Figs 2, 3B) an endemic species in Vietnam (Tritaxis gaudichaudii) diverged, and the S-DIVA analyses indicate a combination of Borneo and the South-East Asian mainland for node 113. Second, although the largest part of India always had a long dry season (Singhvi & Krishnan, 2014), the coastal areas could have been wetter. Therefore, a possible dispersal route is via first the South-East Asian mainland (comparable hypotheses see, e.g., Denduangboripant et al., 2001; Haegens, 2000) and then north-eastern India and eastern coastal India, where the plants may have become extinct later during drier periods (distribution over land during wetter periods is also used to explain disjunct distributions between Africa and Asia, e.g. Chantarasuwan et al., 2016). If the missing section Pycnanthera (including two species from southern India and Sri Lanka) had been included in the analyses, it would probably have revealed a migration route between western Malesia and southern India (but the optimizations of the crown group of Trigonostemon would not change because the signal is strong; see above). About half of the extant Dimorphocalyx spp. reached their current location through dispersal [e.g. from node 113 to Tritaxis gaudichaudii and from node 115 to D. ixoroides (C.B.Rob.) Airy Shaw, Figs 2, 3B] or radiation [e.g. from node 114 to D. denticulatus Merr. and from node 118 to D. muricatus (Hook.f.) Airy Shaw, Figs 2, 3B] since the Late Miocene (6.97–0 Mya). In that period, Borneo and the South-East Asian mainland were still connected (Hall, 2009) and stepping stones from Borneo to the Philippine Islands were already present (see above). If these events occurred during the Pleistocene when Sundaland was a connected landmass (Morley and Flenley, 1987), radiation was even more probable (e.g. from node 118 to D. muricatus, Figs 2, 3B). Wallace’s line Wallace’s line (1860) divides the floristic region of the Malay Archipelago into two parts (van Steenis, 1950). The tectonic and climatic history of this region has caused different floras between western and eastern Malesia (van Welzen et al., 2005). Wallacea, a good phytogeographic region (about one third of the species are endemic), is also a transition zone between western and eastern Malesia (van Welzen et al., 2011) and plays an important role in the floristic exchanges (van Welzen et al., 2005). The probably most commonly used migration route across Wallace’s line is through the Philippines to eastern Wallacea/New Guinea (Nauheimer et al., 2012, Denduangboripant et al., 2001) or vice versa [e.g. Thomas et al., 2012, Su & Saunders, 2009, Jønsson et al., 2010 (birds)]. The dispersal of the eastern Malesian species D. australiensis (node 121 to 120 then to eastern Malesia, Figs 2, 3B) adds to these examples. Node 120 had an optimization of the Philippine islands in the DEC analysis. The other analyses yielded equal (S-DIVA) optimizations, but all of them contained the Philippine islands. Node 120 is the first divergence in Dimorphocalyx. It is inferred at an age (12.78 Mya) when the stepping stones in Wallacea were already in place (Hall, 2009). The other three dispersal events across Wallace’s line do not show clear routes in the analyses. The most likely dispersal route for D. moluccensis (node 117 to 110, Figs 2, 3B; divergence time 10.02–3.35 Mya) is also via the Philippines. On the one hand its sister species Dimorphocalyx sp. is endemic (?) to Sandakan (north-eastern Borneo), which is close to the Philippines and the Sulu Archipelago and Tawi-Tawi islands could have acted as stepping stones; on the other hand the Makassar Strait was always present and had a strong current (Hall, 2009). There is no record for Dimorphocalyx on Sulawesi, thus dispersal via Sulawesi seems unlikely. In Trigonostemon, the inferred ages of the two dispersal events (10.56–8.42 Mya between nodes 99 and 98, Figs 23B, 12.31–10.87 Mya between nodes 103 and 102, Figs 2, 3B) are already possible for exchanges across Wallace’s line (e.g. van Welzen et al., 2014). A possible westward dispersal is inferred after 10.87 Mya (node 103 to T. philippinensis, Figs 2, 3B). Westward dispersals across Wallace’s line are less frequent (Richardson, Costion & Muellner, 2012, Crayn et al., 2015) and often occurred after 8 Mya (Thomas et al., 2012, Su & Saunders, 2009), but our results indicate that this could occur even earlier. Rapid diversification Trigonostemon section Trigonostemon experienced fast diversification in western Malesia during the Pleistocene, especially in two lineages — one on the Malay Peninsula (crown node 84, Figs 2, 3A) and one on Borneo (crown node 71, Figs 2, 3A). More than two-thirds of the taxa (19 out of 26) probably had their last speciation event < 2.58 Mya (Fig. 2), which is hardly the case in the other two sections (their diversification mainly occurred in the South-East Asian mainland). Trigonostemon spp. are mostly small trees growing in lowland ever-wet forests characterised by dipterocarp trees (Whitmore, 1984) along rivers or coast lines (Yu & van Welzen, 2018). These habitats (Heaney, 1991) are to a great extent prone to the rise and fall of sea levels [repeated c. 50 times in the last 2.7 Myr, even up to −120 m during the Last Glacial Maximum (LGM), Woodruff, 2010]: a rapid expansion of the populations during the Pleistocene when sea levels were lower, after which the populations retreated to the islands and became isolated when the vast Sundaland broke up into fragments during interglacials, and consequently speciation events driven by vicariance occurred (Rand, 1948). The next question is why the rapid speciation events almost only took place on the Malay Peninsula and Borneo, but not in the South-East Asian mainland or elsewhere? Western Malesia was a connected landmass (Sundaland) during many Quaternary glacial maxima (Morley & Flenley, 1987). This area is inferred to have been covered by dipterocarp forests, particularly between southern Malaya and Borneo (based on species distribution modelling for the LGM via the CCSM4 model; Raes et al., 2014). Trigonostemon, a typical rainforest group (Yu & van Welzen, 2018), is also likely to have had the highest species richness in this area. After the Pleistocene, probably a great proportion of species became extinct as most of Sundaland submerged, and the surviving ones remained on the Malay Peninsula in the west and Borneo in the east. The distribution of the extant species also corroborates this: the Malay Peninsula, northern Borneo (Sarawak, Sabah and Brunei) and part of eastern Kalimantan (also with high postulated dipterocarp forest density; Raes et al., 2014) represent most of the diversity in the genus (Yu & van Welzen, 2018). As Trigonostemon and Dimorphocalyx mainly grow in ever-wet surroundings, diversification on the South-East Asian mainland was probably reduced by climate, as the further north from the Equator, the more pronounced (longer and drier) the dry season becomes. CONCLUSIONS The dated phylogenetic tree enables us to interpret the historical biogeography of the morphologically similar genera Trigonostemon and Dimorphocalyx in the light of South-East Asian tectonic history. Both genera are estimated to have diverged during the Oligocene, but probably in different geographical locations: the South-East Asian mainland for Trigonostemon and Borneo for Dimorphocalyx. Compared to the hypothesized tectonic history of Malesia, all long-distance dispersals appear to have taken place when the stepping stones were in place. Most dispersal routes are supported by previous studies. The Philippines are inferred to have been the stepping stone for one of the more important dispersal routes across Wallace’s line. Similarly, the historical biogeography also allows us to further interpret the molecular phylogenetic tree. The frequent change of the sea levels during the Pleistocene is considered to have facilitated the diversification in section Trigonostemon, and this is supported by the likely distribution of historical dipterocarp forests in central Sundaland (Raes et al., 2014) and the extant Trigonostemon spp. As a result, after leaving the South-East Asian mainland, one lineage of Trigonostemon succeeded in radiating or dispersing to the Malay Peninsula and Borneo where it could diversify. In contrast, Dimorphocalyx mainly remained in its area of origin, Borneo, which is a possible reason why it diverged earlier than Trigonostemon but has fewer species. ACKNOWLEDGEMENTS We thank Dr Kenneth Wurdack for providing a molecular dataset for Euphorbiaceae. The last author (PvW) thanks the Treub-Maatschappij for supporting the Ornstein chair in Tropical Plant Biogeography. This research was financially supported by the Alberta Mennega Stichting, the Treub-Maatschappij and the Leids Universiteits Fonds. REFERENCES Araújo MB , Ferri-Yáñez F, Bozinovic F, Marquet PA, Valladares F, Chown SL. 2013 . Heat freezes niche evolution . Ecology Letters 16 : 1206 – 1219 . Google Scholar Crossref Search ADS PubMed WorldCat Beddome RH . 1873 . The flora sylvatica for southern India 2 . Madras : Gantz Brothers . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Blume CL . 1825 . Bijdragen tot de Flora van Nederlandsch Indië . Batavia : Lands Drukkerij . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Boucher-Lalonde V , Morin A, Curri DJ. 2012 . How are tree species distributed in climatic space? A simple and general pattern . Global Ecology and Biogeography 21 : 1157 – 1166 . Google Scholar Crossref Search ADS WorldCat Buerki S , Gallaher T, Booth T, Brewer G, Forest F, Pereira JT, Callmander MW. 2016 . Biogeography and evolution of the screw-pine genus Benstonea Callm. & Buerk (Pandanaceae) . Candollea 71 : 217 – 229 . Google Scholar Crossref Search ADS WorldCat Chantarasuwan B , Rønsted N, Kjellberg F, Sungkaew S, van Welzen PC. 2016 . Palaeotropical intercontinental disjunctions revisited using a dated phylogenetic hypothesis with nearly complete species level sampling of Ficus subsect. Urostigma (Moraceae) . Journal of Biogeography 43 : 384 – 397 . Google Scholar Crossref Search ADS WorldCat Condamine FL , Toussainta EFA, Cottonb AM, Gensona GS, Sperlingc FAH, Kergoata GJ. 2013 . Fine-scale biogeographical and temporal diversification processes of peacock swallowtails (Papilio subgenus Achillides) in the Indo-Australian Archipelago . Cladistics 29 : 88 – 111 . Google Scholar Crossref Search ADS WorldCat Crayn DM , Costion C, Harrington MG. 2015 . The Sahul-Sunda floristic exchange: dated molecular phylogenies document Cenozoic intercontinental dispersal dynamics . Journal of Biogeography 42 : 11 – 24 . Google Scholar Crossref Search ADS WorldCat Crepet WL , Daghlian , CP. 1982 . Euphorbioid inflorescences from the Middle Eocene Claiborne formation . American Journal of Botany 69 : 258 – 266 . Google Scholar Crossref Search ADS WorldCat Denduangboripant J , Mendum M, Cronk QCB. 2001 . Evolution in Aeschynanthus (Gesneriaceae) inferred from ITS sequences . Plant Systematics and Evolution 228 : 181 – 197 . Google Scholar Crossref Search ADS WorldCat Drummond AJ , Ho S, Phillips M, Rambaut A. 2006 . Relaxed phylogenetics and dating with confidence . PLoS Biology 4 : 699 – 710 . Google Scholar Crossref Search ADS WorldCat Evans BJ , Brown RM, McGuire JA, Supriatna J, Andayani N, Diesmos A, Iskandar D, Melnick DJ, Cannatella DC. 2003 . Phylogenetics of fanged frogs: testing biogeographical hypotheses at the interface of the Asian and Australian faunal zones . Systematic Biology 52 : 794 – 819 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Gernhard T . 2008 . The conditioned reconstructed process . Journal of Theoretical Biology 253 : 769 ‒ 778 . Google Scholar Crossref Search ADS PubMed WorldCat Govaerts R , Frodin DG, Radcliff-Smith A. 2000 . World checklist and bibliography of Euphorbiaceae (and Pandaceae) 4. Kew : Royal Botanic Gardens . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Grudinski M , Wanntorp L, Pannell CM, Muellner-Riehl AN. 2014 . West to east dispersal in a widespread animal-dispersed woody angiosperm genus (Aglaia, Meliaceae) across the Indo-Australian Archipelago . Journal of Biogeography 41 : 1149 – 1159 . Google Scholar Crossref Search ADS WorldCat Haegens RMAP . 2000 . Taxonomy, phylogeny and biogeography of Baccaurea, Distichirhops, and Nothobaccaurea (Euphorbiaceae) . Blumea Supplement 12 : 1 – 218 . Google Scholar OpenURL Placeholder Text WorldCat Hall R . 2002 . Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations . Journal of Asian Earth Sciences 20 : 353 – 431 . Google Scholar Crossref Search ADS WorldCat Hall R . 2009 . Southeast Asia’s changing palaeogeography . Blumea 54 : 148 – 161 . Google Scholar Crossref Search ADS WorldCat Hauenschild F , Favre A, Schulz M, Muellner-Riehl AN. 2018 . Biogeographic analyses support an Australian origin for the Indomalesian-Australasian wet forest-adapted tropical tree and shrub genus Alphitonia and its close allies (Rhamnaceae) . Botanical Journal of the Linnean Society 188 : 1 – 20 . Google Scholar Crossref Search ADS WorldCat Heaney LR . 1991 . A synopsis of climatic and vegetational change in Southeast Asia . Climate Change 19 : 53 – 61 . Google Scholar Crossref Search ADS WorldCat Ho SYW . 2007 . Calibrating molecular estimates of substitution rates and divergence times in birds . Journal of Avian Biology 38 : 409 – 414 . Google Scholar Crossref Search ADS WorldCat Jønsson KA , Bowie RCK, Nylander JAA, Christidis L, Norman JA, Haegens J. 2010 . Biogeographical history of cuckoo-shrikes (Aves: Passeriformes): transoceanic colonization of Africa from Australo-Papua . Journal of Biogeography 37 : 1767 – 1781 . Google Scholar Crossref Search ADS WorldCat Matuszak S , Muellner‐Riehl AN, Sun H, Favre A. 2015 . Dispersal routes between biodiversity hotspots in Asia: the case of the mountain genus Tripterospermum (Gentianinae, Gentianaceae) and its close relatives . Journal of Biogeography 43 : 580 – 590 . Google Scholar Crossref Search ADS WorldCat Morley RJ , Flenley , JR. 1987 . Late Cainozoic vegetational and environmental changes in the Malay Archipelago. In: Whitmore TC, ed. Biogeographical evolution of the Malay Archipelago. Oxford : Clarendon Press , 50 – 59 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Müller Argoviensis J . 1865 . Euphorbiaceae . Linnaea 34 : 1 ‒ 224 . Google Scholar OpenURL Placeholder Text WorldCat Müller Argoviensis J . 1866 . Euphorbiaceae excl . Euphorbieae . In: de Candolle A, ed. Prodromus systematis naturalis regni vegetabilis 15, 2 . Paris : Masson & Filii , 189 – 1260 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Nauheimer L , Boyce PC, Renner SS. 2012 . Giant taro and its relatives: a phylogeny of the large genus Alocasia (Araceae) sheds light on Miocene floristic exchange in the Malesian region . Molecular Phylogenetics and Evolution 63 : 43 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat Patriat P , Achache J. 1984 . India-Eurasia collision chronology has implications for crustal shortening and driving mechanism of plates . Nature 311 : 615 – 621 . Google Scholar Crossref Search ADS WorldCat Rambaut A , Drummond AJ, Xie D, Baele G, Suchard MA. 2018 . Posterior summarization in Bayesian phylogenetics using Tracer 1.7 . Systematic Biology 67 : 901 – 904 . Google Scholar Crossref Search ADS PubMed WorldCat Raes N , Cannon CH, Hijmans RJ, Piessens T, Saw LG, van Welzen PC, Slik JWF. 2014 . Historical distribution of Sundaland’s dipterocarp rainforests at Quaternary glacial maxima . Proceedings of the National Academy of Sciences of the United States of America 111 : 16790 – 16795 . Google Scholar Crossref Search ADS PubMed WorldCat Raes N , van Welzen PC. 2009 . The demarcation and internal division of Flora Malesiana: 1857 – present . Blumea 54 : 6 – 8 . Google Scholar Crossref Search ADS WorldCat Rand AL . 1948 . Glaciation, an isolating factor in speciation . Evolution; International Journal of Organic Evolution 2 : 314 – 321 . Google Scholar Crossref Search ADS PubMed WorldCat Repetur CP , van Welzen PC, de Vogel EF. 1997 . Phylogeny and historical biogeography of the genus Bromheadia (Orchidaceae) . Systematic Botany 22 : 465 – 477 . Google Scholar Crossref Search ADS WorldCat Richardson JE , Costion CM, Muellner AN. 2012 . The Malesian floristic interchange: plant migration patterns across Wallace’s Line . In: Gower DJ, Johnson KG, Richardson JE, Rosen BR, Rüber L, Williams ST, eds. Biotic evolution and environmental change in Southeast Asia . Cambridge : Cambridge University Press, 138 – 163 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ridder-Numan , JWA . 1998 . Historical biogeography of Spatholobus (Leguminosae-Papilionoideae) and allies in SE Asia . In: Hall R, Holloway JD, eds. Biogeography and geological evolution of SE Asia. Leiden : Backhuys Publishers , 259 – 277 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Roos MC . 1993 . State of affairs regarding Flora Malesiana: progress in revision work and publication schedule . Flora Malesiana Bulletin 11 : 133 – 142 . Google Scholar OpenURL Placeholder Text WorldCat Sheldon FH , Oliveros CH, Taylor SS, McKay B, Lim HC, Abdul Rahman M, Mays H, Moyle RG. 2012 . Molecular phylogeny and insular biogeography of the lowland tailorbirds of Southeast Asia (Cisticolidae: Orthotomus) . Molecular Phylogenetics and Evolution 65 : 54 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat Sirichamorn Y , Thomas DC, Adema FACB, van Welzen PC. 2014 . Historical biogeography of Aganope, Brachypterum and Derris (Fabaceae, tribe Millettieae): insights into the origins of Palaeotropical intercontinental disjunctions and general biogeographical patterns in Southeast Asia . Journal of Biogeography 41 : 882 – 893 . Google Scholar Crossref Search ADS WorldCat Singhvi AK , Krishnan R. 2014 . Past and the present climate of India . In: Kale VS, ed. Landscapes and landforms of India, World geomorphological landscapes. Dordrecht, Heidelberg, New York, London : Springer , 15 – 23 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC van Steenis CGGJ . 1948 . Hoofdlijnen van de plantengeografie van de Indische archipel op grond van de verspreiding van de Phanerogamen geslachten . Tijdschrift van het Koninklijke Nederlandse Aardrijkskundige Genootschap 65 : 193 – 208 . Google Scholar OpenURL Placeholder Text WorldCat van Steenis CGGJ . 1950 . The delimitation of Malaysia and its main plant geographical divisions . Flora Malesiana, Ser. I, Vol. 1 : lxx – lxxv . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Su YCF , Saunders MK. 2009 . Evolutionary divergence times in the Annonaceae: evidence of a late Miocene origin of Pseuduvaria in Sundaland with subsequent diversification in New Guinea . BMC Evolutionary Biology 9 : 153 . Google Scholar Crossref Search ADS PubMed WorldCat Suchard MA , Lemey P, Baele G, Ayres DL, Drummond AJ, Rambaut A. 2018 . Bayesian phylogenetic and phylodynamic data integration using BEAST 1.10 . Virus Evolution 4 : vey016 . Google Scholar Crossref Search ADS PubMed WorldCat Thomas DC , Hughes M, Phutthai T, Ardi WH, Rajbhandary S, Rubite R, Twyford AD, Richardson JE. 2012 . West to east dispersal and subsequent rapid diversification of the mega-diverse genus Begonia (Begoniaceae) in the Malesian archipelago . Journal of Biogeography 39 : 98 – 113 . Google Scholar Crossref Search ADS WorldCat Thwaites GHK . 1861 . Enumeratio plantarum Zeylaniae 4 . London : Dulau & Co . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Wallace AR . 1860 . On the zoological geography of the Malay Archipelago . Journal of the Linnean Society of London 4 : 172 – 184 . Google Scholar OpenURL Placeholder Text WorldCat Wallace AR . 1869 . The Malay Archipelago: the land of the orang-utan, and the bird of paradise. A narrative of travel, with sketches of man and nature. London : Macmillan . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Webster GL . 2014 . Euphorbiaceae. In: Kubitzki K, ed. The families and genera of vascular plants 11 . Heidelberg, New York, Dordrecht, London : Springer , 51 ‒ 216 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC van Welzen PC , van Oostrum AF. 2015 . Revision of the Malesian species of Dimorphocalyx (Euphorbiaceae) . Blumea 53 : 191 – 201 . Google Scholar Crossref Search ADS WorldCat van Welzen PC , Parnell JAN, Slik , JWF. 2011 . Wallace’s Line and plant distributions: two or three phytogeographical areas and where to group Java? Biological Journal of the Linnean Society 103 : 531 ‒ 545 . Google Scholar Crossref Search ADS WorldCat van Welzen PC , Slik JWF, Alahuhta J. 2005 . Plant distribution patterns and plate tectonics in Malesia . Biologiske Skrifter 55 : 199 ‒ 217 . Google Scholar OpenURL Placeholder Text WorldCat van Welzen PC , Strijk JS, van Konijnenburg-van Cittert JH, Nucete M, Merckx VS. 2014 . Dated phylogenies of the sister genera Macaranga and Mallotus (Euphorbiaceae): congruence in historical biogeographic patterns? PLoS One 9 : e85713 . Google Scholar Crossref Search ADS PubMed WorldCat Whitmore TC . 1984 . Tropical rain forests of the Far East, 2nd edn Oxford : Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Woodruff DS . 2010 . Biogeography and conservation in Southeast Asia: how 2.7 million years of repeated environmental fluctuations affect today’s patterns and the future of the remaining refugial-phase biodiversity . Biodiversity and Conservation 19 : 919 – 941 . Google Scholar Crossref Search ADS WorldCat Wurdack KJ , Hoffmann P, Chase MW. 2005 . Molecular phylogenetic analysis of uniovulate Euphorbiaceae (Euphorbiaceae sensu stricto) using plastid rbcL and trnL-F DNA sequences . American Journal of Botany 92 : 1397 – 1420 . Google Scholar Crossref Search ADS PubMed WorldCat Xi Z , Ruhfel BR, Schaefer H, Amorim AM, Sugumaran M, Wurdack KJ, Endress PK, Matthews ML, Stevens PF, Mathews S, Davis CC. 2012 . Phylogenomics and a posteriori data partitioning resolve the Cretaceous angiosperm radiation Malpighiales . Proceedings of the National Academy of Sciences of the United States of America 109 : 17519 – 17524 . Google Scholar Crossref Search ADS PubMed WorldCat Yu R-Y , van Welzen PC. 2018 . A taxonomic revision of Trigonostemon Blume (Euphorbiaceae) in Malesia . Blumea 62 : 179 – 229 . Google Scholar Crossref Search ADS WorldCat Yu R-Y , Slik FJW, van Welzen PC. 2019 . In press. Molecular phylogeny of Trigonostemon Blume and its relatives (Euphorbiaceae) . Taxon . Google Scholar OpenURL Placeholder Text WorldCat Yu R-Y , Agoo EMG, Callado JR, van Welzen PC. 2019 . In press. Taxonomic notes on Trigonostemon (Euphorbiaceae) in the Philippines . Blumea . Google Scholar OpenURL Placeholder Text WorldCat Yu R-Y , van Welzen PC. 2019 . In press. A taxonomic revision of Trigonostemon (Euphorbiaceae) outside Malesia . Blumea . Google Scholar OpenURL Placeholder Text WorldCat Yu Y , Harris , AJ, Blair C, He XJ. 2015 . RASP (Reconstruct Ancestral State in Phylogenies): a tool for historical biogeography . Molecular Phylogenetics and Evolution 87 : 46 ‒ 49 . Google Scholar Crossref Search ADS PubMed WorldCat Yule GU . 1925 . A mathematical theory of evolution, based on the conclusions of Dr. J. C. Willis, F.R.S . Philosophical Transactions of the Royal Society London: Series B . 213 : 21 ‒ 87 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Zollinger H . 1857 . Over het begrip en den omvang eener Flora Malesiana . Natuurkundig Tijdschrift voor Nederlandsch-Indië 13 : 293 – 322 . Google Scholar OpenURL Placeholder Text WorldCat APPENDIX. VOUCHER INFORMATION Taxon . Collector . Field number . Location . Herbarium barcode . Dimorphocalyx australiensis Gray 7859 Australia L.2211804 D. beddomei Ridsdale 388 India L.2211814 D. cumingii Mendoza PNH 42406 Philippines L.2211815 D. denticulatus Church 1819 W. Kalimantan L.2204228 D. glabellus Kosterman 26317 India L.2204358 D. ixoroides Ruffo PNH 41828 Philippines L.2204362 D. malayanus Sidisunthorn ST 1640 Thailand L.2204317 D. moluccensis Ramlanto 905 Moluccas L.2204173 D. muricatus Pooma 4499 Thailand L.2204330 D. pauciflorus Forman 906 Brunei L.2211818 D. sp. Yu SAN 158473 Sabah L D. trichocarpus Anderson 20974 Sarawak L.2182719 Jatropha gossypiifolia Snellius 11150 Lesser Sunda Islands L.2236164 Ostodes paniculata Yu 172 Java L Trigonostemon adenocalyx Huang H 09427 Guangxi, China IBK 00214246 T. albiflorus Gillespie 7405 Vietnam L.2260160 T. aurantiacus Yu 160 Java* L T. balgooyi K. Hisham FRI 73820 Malay Peninsula KEP 217032 T. beccarii Yu 169 Java* L T. bonianus Huang NG 293 Guangxi, China IBK00214243 T. capillipes Gardner ST 2836 Thailand L.2260155 T. cherrieri Veillon 7420 New Caledonia L.2260157 T. detritiferus Yu 91 Brunei L T. diffusus Pureseglove P 4702 Sarawak L.2260164 T. filiformis Yu 243 Philippines L T. flavidus Yu 264 Guangdong, China* L T. hartleyi Hartley TGH 11087 New Guinea L 0043309 T. inopinatus Forster PIF 30012 Australia L.2260532 T. laevigatus var. laevigatus Yu SAN 158478 Sabah L T. lii Huang NG 306 Guangxi, China IBK 00214245 T. longifolius Yu SAN 158474 Brunei L T. longipes Yu 227 Philippines L T. lychnos Ali Ahmad BRUN 19908 Brunei SING 0179065 T. magnificus de Wilde 19441 Sumatra L.2260380 T. malaccanus Yu 176 Malay Peninsula L T. merrillii Yu 255 Philippines L T. murtonii Webb WA 207 Cambodia L.2260393 T. pentandrus Yu FRI 86653 Malay Peninsula L T. philippinensis Bouman RWB 67 Yunnan, China* XTBG T. polyanthus Lagrimas PNH 39433 Philippines L.2258802 T. quocensis Middleton 4046 Thailand L.2258834 T. reidioides Cheng CL 1238 Cambodia L.3784729 T. rufescens Yu FRI 86664 Malay Peninsula L T. sandakanensis Yu SAN 158473 Sabah L T. semperflorens Koelz 27866 India L.2258764 T. sp. 1 Yu 260 Philippines L T. sp. 2 Koelz 27903 India L.2260526 T. sp. 3 Leong-Škorničková OS 6243 Laos SING 0190181 T. sp. cf. filiformis Yu SAN 158462 Sabah L T. verticillatus var. salicifolius Worthington 13073 Malay Peninsula L.2258554 T. verticillatus var. verticillatus Hai HN-NY 503 Vietnam L T. verrucosus Chase 1274 Java* K T. victoriae Yu 258 Philippines L T. villosus var. borneensis Yu SAN 158461 Sabah L T. villosus var. cordatus Yu SAN 158479 Sabah L T. villosus var. aff. merrillianus Yu 107 Brunei L T. villosus var. merrillianus Yu 254 Philippines L T. villosus var. villosus Yu 178 Malay Peninsula L T. viridissimus var. viridissimus Julius FRI 56285 Malay Peninsula L.3795794 T. wui Yu 266 Guangdong, China* L T. xyphophylloides Yu 265 Guangdong, China* L Tritaxis gaudichaudii D. Soejarto DDS 14143 Vietnam L.3784654 Taxon . Collector . Field number . Location . Herbarium barcode . Dimorphocalyx australiensis Gray 7859 Australia L.2211804 D. beddomei Ridsdale 388 India L.2211814 D. cumingii Mendoza PNH 42406 Philippines L.2211815 D. denticulatus Church 1819 W. Kalimantan L.2204228 D. glabellus Kosterman 26317 India L.2204358 D. ixoroides Ruffo PNH 41828 Philippines L.2204362 D. malayanus Sidisunthorn ST 1640 Thailand L.2204317 D. moluccensis Ramlanto 905 Moluccas L.2204173 D. muricatus Pooma 4499 Thailand L.2204330 D. pauciflorus Forman 906 Brunei L.2211818 D. sp. Yu SAN 158473 Sabah L D. trichocarpus Anderson 20974 Sarawak L.2182719 Jatropha gossypiifolia Snellius 11150 Lesser Sunda Islands L.2236164 Ostodes paniculata Yu 172 Java L Trigonostemon adenocalyx Huang H 09427 Guangxi, China IBK 00214246 T. albiflorus Gillespie 7405 Vietnam L.2260160 T. aurantiacus Yu 160 Java* L T. balgooyi K. Hisham FRI 73820 Malay Peninsula KEP 217032 T. beccarii Yu 169 Java* L T. bonianus Huang NG 293 Guangxi, China IBK00214243 T. capillipes Gardner ST 2836 Thailand L.2260155 T. cherrieri Veillon 7420 New Caledonia L.2260157 T. detritiferus Yu 91 Brunei L T. diffusus Pureseglove P 4702 Sarawak L.2260164 T. filiformis Yu 243 Philippines L T. flavidus Yu 264 Guangdong, China* L T. hartleyi Hartley TGH 11087 New Guinea L 0043309 T. inopinatus Forster PIF 30012 Australia L.2260532 T. laevigatus var. laevigatus Yu SAN 158478 Sabah L T. lii Huang NG 306 Guangxi, China IBK 00214245 T. longifolius Yu SAN 158474 Brunei L T. longipes Yu 227 Philippines L T. lychnos Ali Ahmad BRUN 19908 Brunei SING 0179065 T. magnificus de Wilde 19441 Sumatra L.2260380 T. malaccanus Yu 176 Malay Peninsula L T. merrillii Yu 255 Philippines L T. murtonii Webb WA 207 Cambodia L.2260393 T. pentandrus Yu FRI 86653 Malay Peninsula L T. philippinensis Bouman RWB 67 Yunnan, China* XTBG T. polyanthus Lagrimas PNH 39433 Philippines L.2258802 T. quocensis Middleton 4046 Thailand L.2258834 T. reidioides Cheng CL 1238 Cambodia L.3784729 T. rufescens Yu FRI 86664 Malay Peninsula L T. sandakanensis Yu SAN 158473 Sabah L T. semperflorens Koelz 27866 India L.2258764 T. sp. 1 Yu 260 Philippines L T. sp. 2 Koelz 27903 India L.2260526 T. sp. 3 Leong-Škorničková OS 6243 Laos SING 0190181 T. sp. cf. filiformis Yu SAN 158462 Sabah L T. verticillatus var. salicifolius Worthington 13073 Malay Peninsula L.2258554 T. verticillatus var. verticillatus Hai HN-NY 503 Vietnam L T. verrucosus Chase 1274 Java* K T. victoriae Yu 258 Philippines L T. villosus var. borneensis Yu SAN 158461 Sabah L T. villosus var. cordatus Yu SAN 158479 Sabah L T. villosus var. aff. merrillianus Yu 107 Brunei L T. villosus var. merrillianus Yu 254 Philippines L T. villosus var. villosus Yu 178 Malay Peninsula L T. viridissimus var. viridissimus Julius FRI 56285 Malay Peninsula L.3795794 T. wui Yu 266 Guangdong, China* L T. xyphophylloides Yu 265 Guangdong, China* L Tritaxis gaudichaudii D. Soejarto DDS 14143 Vietnam L.3784654 Open in new tab Taxon . Collector . Field number . Location . Herbarium barcode . Dimorphocalyx australiensis Gray 7859 Australia L.2211804 D. beddomei Ridsdale 388 India L.2211814 D. cumingii Mendoza PNH 42406 Philippines L.2211815 D. denticulatus Church 1819 W. Kalimantan L.2204228 D. glabellus Kosterman 26317 India L.2204358 D. ixoroides Ruffo PNH 41828 Philippines L.2204362 D. malayanus Sidisunthorn ST 1640 Thailand L.2204317 D. moluccensis Ramlanto 905 Moluccas L.2204173 D. muricatus Pooma 4499 Thailand L.2204330 D. pauciflorus Forman 906 Brunei L.2211818 D. sp. Yu SAN 158473 Sabah L D. trichocarpus Anderson 20974 Sarawak L.2182719 Jatropha gossypiifolia Snellius 11150 Lesser Sunda Islands L.2236164 Ostodes paniculata Yu 172 Java L Trigonostemon adenocalyx Huang H 09427 Guangxi, China IBK 00214246 T. albiflorus Gillespie 7405 Vietnam L.2260160 T. aurantiacus Yu 160 Java* L T. balgooyi K. Hisham FRI 73820 Malay Peninsula KEP 217032 T. beccarii Yu 169 Java* L T. bonianus Huang NG 293 Guangxi, China IBK00214243 T. capillipes Gardner ST 2836 Thailand L.2260155 T. cherrieri Veillon 7420 New Caledonia L.2260157 T. detritiferus Yu 91 Brunei L T. diffusus Pureseglove P 4702 Sarawak L.2260164 T. filiformis Yu 243 Philippines L T. flavidus Yu 264 Guangdong, China* L T. hartleyi Hartley TGH 11087 New Guinea L 0043309 T. inopinatus Forster PIF 30012 Australia L.2260532 T. laevigatus var. laevigatus Yu SAN 158478 Sabah L T. lii Huang NG 306 Guangxi, China IBK 00214245 T. longifolius Yu SAN 158474 Brunei L T. longipes Yu 227 Philippines L T. lychnos Ali Ahmad BRUN 19908 Brunei SING 0179065 T. magnificus de Wilde 19441 Sumatra L.2260380 T. malaccanus Yu 176 Malay Peninsula L T. merrillii Yu 255 Philippines L T. murtonii Webb WA 207 Cambodia L.2260393 T. pentandrus Yu FRI 86653 Malay Peninsula L T. philippinensis Bouman RWB 67 Yunnan, China* XTBG T. polyanthus Lagrimas PNH 39433 Philippines L.2258802 T. quocensis Middleton 4046 Thailand L.2258834 T. reidioides Cheng CL 1238 Cambodia L.3784729 T. rufescens Yu FRI 86664 Malay Peninsula L T. sandakanensis Yu SAN 158473 Sabah L T. semperflorens Koelz 27866 India L.2258764 T. sp. 1 Yu 260 Philippines L T. sp. 2 Koelz 27903 India L.2260526 T. sp. 3 Leong-Škorničková OS 6243 Laos SING 0190181 T. sp. cf. filiformis Yu SAN 158462 Sabah L T. verticillatus var. salicifolius Worthington 13073 Malay Peninsula L.2258554 T. verticillatus var. verticillatus Hai HN-NY 503 Vietnam L T. verrucosus Chase 1274 Java* K T. victoriae Yu 258 Philippines L T. villosus var. borneensis Yu SAN 158461 Sabah L T. villosus var. cordatus Yu SAN 158479 Sabah L T. villosus var. aff. merrillianus Yu 107 Brunei L T. villosus var. merrillianus Yu 254 Philippines L T. villosus var. villosus Yu 178 Malay Peninsula L T. viridissimus var. viridissimus Julius FRI 56285 Malay Peninsula L.3795794 T. wui Yu 266 Guangdong, China* L T. xyphophylloides Yu 265 Guangdong, China* L Tritaxis gaudichaudii D. Soejarto DDS 14143 Vietnam L.3784654 Taxon . Collector . Field number . Location . Herbarium barcode . Dimorphocalyx australiensis Gray 7859 Australia L.2211804 D. beddomei Ridsdale 388 India L.2211814 D. cumingii Mendoza PNH 42406 Philippines L.2211815 D. denticulatus Church 1819 W. Kalimantan L.2204228 D. glabellus Kosterman 26317 India L.2204358 D. ixoroides Ruffo PNH 41828 Philippines L.2204362 D. malayanus Sidisunthorn ST 1640 Thailand L.2204317 D. moluccensis Ramlanto 905 Moluccas L.2204173 D. muricatus Pooma 4499 Thailand L.2204330 D. pauciflorus Forman 906 Brunei L.2211818 D. sp. Yu SAN 158473 Sabah L D. trichocarpus Anderson 20974 Sarawak L.2182719 Jatropha gossypiifolia Snellius 11150 Lesser Sunda Islands L.2236164 Ostodes paniculata Yu 172 Java L Trigonostemon adenocalyx Huang H 09427 Guangxi, China IBK 00214246 T. albiflorus Gillespie 7405 Vietnam L.2260160 T. aurantiacus Yu 160 Java* L T. balgooyi K. Hisham FRI 73820 Malay Peninsula KEP 217032 T. beccarii Yu 169 Java* L T. bonianus Huang NG 293 Guangxi, China IBK00214243 T. capillipes Gardner ST 2836 Thailand L.2260155 T. cherrieri Veillon 7420 New Caledonia L.2260157 T. detritiferus Yu 91 Brunei L T. diffusus Pureseglove P 4702 Sarawak L.2260164 T. filiformis Yu 243 Philippines L T. flavidus Yu 264 Guangdong, China* L T. hartleyi Hartley TGH 11087 New Guinea L 0043309 T. inopinatus Forster PIF 30012 Australia L.2260532 T. laevigatus var. laevigatus Yu SAN 158478 Sabah L T. lii Huang NG 306 Guangxi, China IBK 00214245 T. longifolius Yu SAN 158474 Brunei L T. longipes Yu 227 Philippines L T. lychnos Ali Ahmad BRUN 19908 Brunei SING 0179065 T. magnificus de Wilde 19441 Sumatra L.2260380 T. malaccanus Yu 176 Malay Peninsula L T. merrillii Yu 255 Philippines L T. murtonii Webb WA 207 Cambodia L.2260393 T. pentandrus Yu FRI 86653 Malay Peninsula L T. philippinensis Bouman RWB 67 Yunnan, China* XTBG T. polyanthus Lagrimas PNH 39433 Philippines L.2258802 T. quocensis Middleton 4046 Thailand L.2258834 T. reidioides Cheng CL 1238 Cambodia L.3784729 T. rufescens Yu FRI 86664 Malay Peninsula L T. sandakanensis Yu SAN 158473 Sabah L T. semperflorens Koelz 27866 India L.2258764 T. sp. 1 Yu 260 Philippines L T. sp. 2 Koelz 27903 India L.2260526 T. sp. 3 Leong-Škorničková OS 6243 Laos SING 0190181 T. sp. cf. filiformis Yu SAN 158462 Sabah L T. verticillatus var. salicifolius Worthington 13073 Malay Peninsula L.2258554 T. verticillatus var. verticillatus Hai HN-NY 503 Vietnam L T. verrucosus Chase 1274 Java* K T. victoriae Yu 258 Philippines L T. villosus var. borneensis Yu SAN 158461 Sabah L T. villosus var. cordatus Yu SAN 158479 Sabah L T. villosus var. aff. merrillianus Yu 107 Brunei L T. villosus var. merrillianus Yu 254 Philippines L T. villosus var. villosus Yu 178 Malay Peninsula L T. viridissimus var. viridissimus Julius FRI 56285 Malay Peninsula L.3795794 T. wui Yu 266 Guangdong, China* L T. xyphophylloides Yu 265 Guangdong, China* L Tritaxis gaudichaudii D. Soejarto DDS 14143 Vietnam L.3784654 Open in new tab © The Botanical Journal of the Linnean Society 2019. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. © The Botanical Journal of the Linnean Society 2019.

Wang,, Zhi-Wei;Zhang,, Ti-Cao;Luo,, Dong;Sun,, Wen-Guang;Sun,, Hang

doi: 10.1093/botlinnean/boz080pmid: N/A

Abstract It has been hypothesized that geological and climatic changes in the Sino–Himalayan region played a significant role in evolutionary history. In this study, we tested this hypothesis by investigating the phylogeography of Excoecaria acerifolia (Euphorbiaceae), a riparian plant species that is widely distributed in the hot/warm-dry river valleys of the Sino–Himalayan region. Spatial analysis of molecular variance, a median-joining network and a Bayesian phylogenetic tree based on plastid DNA, all suggested three major lineages corresponding to the Jialing-Min-Dadu (JMD lineage), Jinsha-Yalong-Salween (JY lineage) and Yarlung Tsangpo-Mekong-Red-Nanpan (YMRN lineage) drainage basins. This was also generally supported by the results based on nuclear DNA. The divergence times of these three major lineages based on both datasets fell in the early Pleistocene, coinciding with the period of drainage reorganization events in the Sino–Himalayan region. The diversification times in the lineages were, however, dated back to the mid Pleistocene, corresponding to the Naynayxungla glaciation (0.72–0.50 Mya) and the penultimate glaciation (0.30–0.13 Mya), which were the most and second most severe glaciations in this region, respectively. Furthermore, mismatch analyses, neutrality test and ecological niche modelling suggest long-term demographic stability of the JY and JMD populations, with expansion only in the YMRN populations during the period(s) from the late penultimate glaciation (138.12 Kya) to the Last Interglacial (95.79 Kya), probably because of less extensive glaciations since the late Pleistocene and the gradually warming interglacial stage. Our study provides one of the few pieces of evidence indicating that combined historical drainage reorganization and climatic change since the Pleistocene might also have acted as important factors in the evolutionary history of riparian plants in the region. climatic change, drainage reorganization events, glaciation, Pleistocene INTRODUCTION The Sino–Himalayan region, which extends from the northern parts of the eastern Himalayas through adjacent mountain ranges (e.g. the Hengduan Mountains) further east to south-western China (Qiu, Fu & Comes, 2011), has diverse terrain and landforms. Hot/warm-dry river valleys are one of the most prominent natural landforms in this region (Tang, Xie & Sun, 2004; Wang et al., 2015b), which, historically, has undergone dramatic geological and climatic changes. The principal events were the reorganization of major drainage systems and the glacial-interglacial cycles (Li et al., 1979; An et al., 1990; Li & Fang, 1999; Clark et al., 2004; Zhou et al., 2006; Royden, Burchfiel & van der Hilst, 2008). For the reorganization of drainage systems, major rivers in the Sino–Himalayan region were considered to be tributaries of a single, south-flowing system, namely the Palaeo-Red River (Brookfield, 1998; Clark et al., 2004; Clift, Blusztajn & Nguyen, 2006; Ming, Shi & Zhang, 2006; Ming, Shi & Dong, 2007; Fig. 1A). Subsequently, great drainage reorganization events associated with the uplift of the Qinghai-Tibet Plateau took place in these rivers, leading to the formation of the modern river drainage basins (Clark et al., 2004; Ming et al., 2006, 2007; Fig. 1B). The exact timing of drainage reorganization events has been controversial, ranging from the Miocene to Pleistocene, but not later than the mid Pleistocene (c. 0.781 Mya) based on geological studies (Brookfield, 1998; Clark et al., 2004; Clift et al., 2006; Ming et al., 2006, 2007). Apart from these events, in particular, river valleys in the Sino–Himalayan region have also been subject to dramatic climatic changes (Ruddiman & Kutzbach, 1989; Zheng, Xu & Shen, 2002). Four major glaciations (Xixiabangma, 800–1170 Kya; Naynayxungla, 500–720 Kya; the penultimate glaciation, 130–300 Kya; the last glaciation, 10–70 Kya) took place during the Pleistocene, and gave rise to many piedmont glaciers and great valley glaciers in this region (Zheng et al., 2002). In particular, the climate became quite severe after the Naynayxungla glaciation, the most extensive glaciation in the Sino–Himalayan and adjacent regions (Zheng et al., 2002). Figure 1. Open in new tabDownload slide Drainage systems in the Sino–Himalayan region (adapted from Clark et al., 2004, and Zhang et al., 2011). A, the palaeo-drainage system before drainage reorganization events, and B, the modern drainage system after drainage reorganization events. Figure 1. Open in new tabDownload slide Drainage systems in the Sino–Himalayan region (adapted from Clark et al., 2004, and Zhang et al., 2011). A, the palaeo-drainage system before drainage reorganization events, and B, the modern drainage system after drainage reorganization events. Population processes (e.g. population divergence, diversity and demography) are strongly linked with geological and/or climatic changes (Hamrick, Godt & Sherman-Broyles, 1992; Hewitt, 2004; Qiu et al., 2011; Sun et al., 2017). The combined effect of these processes can result in barriers to gene flow, create different niches for genetic divergence and provide pressures for population demography (Liu et al., 2013; Luo et al., 2016). This effect has received considerable attention and been examined over recent years, and it demonstrates a variety of phylogeographic patterns on (sub)alpine plants (Chen et al., 2008; Wang et al., 2008; Yang et al., 2008; Cun & Wang, 2010; Li et al., 2011; Qiu et al., 2011; Luo et al., 2016). In contrast, few phylogeographic studies have been conducted on the characteristic riparian plants in this region, and those had been done mainly focused on problems such as how riparian plants respond to the drainage reorganization events [e.g. Terminalia franchetii Gagnep. (Zhang, Comes & Sun, 2011; Zhang & Sun, 2011), Buddleja crispa Benth. (Yue et al., 2012) and Salweenia Baker f. (Yue et al., 2011)]. In contrast, topics such as how riparian plants respond to the impact of the climatic changes have been rarely approached. Since the drainage reorganization events are older than the severe climatic change during the Pleistocene, plants responding to the pattern of drainage reorganization events would show more significant genetic divergence among disconnected rivers (e.g. the Jinsha, Yalong and Mekong), particularly from the Palaeo-Red River, and with divergence time earlier than the mid Pleistocene. Studies of T. franchetii, B. crispa and Salweenia have shown evidence of significant isolation from the Palaeo-Red River (or among disconnected rivers), correlation with historical drainage reorganization events during the late Pliocene to mid Pleistocene (Yue et al., 2011; Zhang et al., 2011; Zhang & Sun, 2011; Yue et al., 2012; Sun et al., 2017). More recently, a study of a riparian plant species (Osteomeles schwerinae C.K.Schneid.) was conducted to examine the impact of climatic changes in river valleys, the results of which only indicated population genetic divergence between the Jialing-Min rivers and other rivers, linked with fluctuations in the Asian monsoons since the late Pleistocene (Wang et al., 2015b), but still leaving the impact of glacial-interglacial climatic change and especially the combined effects of historical drainage reorganization events and climatic changes on riparian plants rarely studied. Excoecaria acerifolia Didr. (Euphorbiaceae) is a characteristic riparian species widely distributed along the hot/warm-dry river valleys of the Sino–Himalayan region (Jin, 1998; Li & Esser, 2008), including the majority of the major rivers in this region (Fig. 2). According to the Flora of China (Li & Esser, 2008) and our field observations, it commonly grows as a small tree species on the slopes of river valleys between 800 and 3000 m elevation (Table 1) and never reaches or extends over the top of mountain ranges. Its seeds are usually gravity-dispersed, accompanied by their capsules, and/or occasionally subject to secondary dispersal by flood waters, resulting in long-distance movement. It rarely experiences long-distance dispersal by animals (e.g. birds) since it probably contains poisonous chemical constituents (Webster, 1975; Das et al., 2011; Lin et al., 2016; Rawat, Parasad & Bisht, 2017; Yeragi & Mendhulkar, 2017). Moreover, its non-fleshy fruits are less attractive to animals, probably making it less suitable for long-distance dispersal by animals (Li & Esser, 2008; Fleming & Kress, 2011; Wotton & Mcalpine, 2015). Excoecaria acerifolia is entomophilous, its pollen being only rarely dispersed by the wind (Webster, 1975). However, the dispersal distance of insect-mediated pollination is usually limited, commonly skewed towards short distances of just a few metres to hundreds of metres, rarely > 5 km (Sophia et al., 2009; García, Jordano & Godoy, 2010; Krauss et al., 2017). Given these, we predict that the seed and pollen of E. acerifolia seldom cross the barriers (e.g. the Wumeng, Yunling and Nushan mountains, all > 4000 m in elevation) between the disconnected rivers after the drainage reorganization events and the gene flow within this species is more likely to occur along rivers rather than between them (Yue et al., 2011; Liu et al., 2013). In addition, E. acerifolia is a riparian species endemic to the hot/warm-dry river valleys, and it probably prefers warm conditions (Jin, 1998; Li & Esser, 2008). Therefore, it is likely that E. acerifolia would also respond to climatic change, particularly the climatic change during the glacial–interglacial periods. Thus, it is probably an ideal subject for studying the combined effects of historical drainage reorganization events and climatic changes on population evolutionary history in river valleys of the Sino–Himalayan region. Here, we use an integrative approach by combining molecular approaches with climatic data to elucidate the population divergence, diversity and demographic history of E. acerifolia. Our specific objectives are: (1) to quantify whether there is genetic divergence among the populations of E. acerifolia in different river valleys; (2) to date the divergences and explore whether these divergences relate to geological changes (e.g. historical drainage reorganization events) and/or climatic changes (e.g. the glacial-interglacial climatic change) and (3) to unravel the demographic history of E. acerifolia. Table 1. Details of sampled populations of Excoecaria acerifolia (312 and 347 individuals for plastid DNA and nuclear DNA, respectively), haplotypes and ribotypes (number of individuals) detected per population, vouchers and the assignment of each population to drainage basins Code . Location . Drainages . Latitude (°N) . Longitude (°E) . Elevation (m) . Haplotype frequencies . Ribotype frequencies . Voucher . 1 Wenxian Jialing River 32.762 105.013 804 H6 (12) R5 (22) WZW4, KUN 2 Longnan Jialing River 33.269 105.111 1005 H13 (11) R5 (24) WZW26, KUN 3 Baoxing Min River 30.387 102.813 956 H2 (10) R2 (22) WZW153, KUN 4 Kangding Dadu River 30.063 102.174 1410 H5 (9), H11 (3) R2 (2), R4 (1), R5 (21) WZW129, KUN 5 Danba Dadu River 31.081 101.871 2018 H5 (11), H6 (1) R2 (16), R4 (2), R5 (6) WZW173, KUN 6 Jiulong Yalong River 28.617 101.664 2130 H9 (7) R9 (14) WZW110, KUN 7 Muli Yalong River 27.926 101.287 2152 H9 (6) R1 (5), R6 (11), R9 (2) MCQ037, KUN 8 Yongsheng Jinsha River 26.697 100.719 2015 H9 (5) R1 (11), R6 (5) 15659, KUN 9 Batang Jinsha River 29.814 99.014 2474 H1 (11) R1 (14) WZW181, KUN 10 Tuoding Jinsha River 27.954 99.414 1967 H3 (8), H12 (1) R1 (20) WZW252, KUN 11 Derong Jinsha River 28.857 99.306 2540 H3 (11) R1 (22) WZW194, KUN 12 Daocheng Jinsha River 28.277 100.191 2328 H7 (12) R1 (8), R6 (4), R7 (7), R8 (1) WZW208, KUN 13 Jinjiang Jinsha River 26.996 99.950 1841 H3 (4) R1 (22) 15703, KUN 14 Judian Jinsha River 27.282 99.672 1964 H1 (9), H12 (2) R1 (24) WZW254, KUN 15 Benzilan Jinsha River 28.402 99.141 2613 H1 (1), H3 (11) R1 (24) WZW236, KUN 16 Yunling Mekong 28.348 98.901 2858 H8 (4) R3 (5), R9 (3), R10 (8) WZW243, KUN 17 Yanmen Mekong 28.075 98.929 1935 H8 (10) R3 (4), R9 (1), R10 (5), R11 (1), R18 (13) WZW248, KUN 18 Feilaishi Mekong 28.478 98.847 2977 H8 (11) R3 (9),R10 (11), R11 (2) WWW044, KUN 19 Rumei Mekong 29.659 98.368 2794 H8 (6) R10 (24) WWW048, KUN 20 Yanjing Mekong 29.065 98.614 2672 H8 (10) R10 (20), R15 (2), R17 (2) WWW047, KUN 21 Yezhi Mekong 27.662 99.014 1739 H8 (5) R3 (4), R10 (4), R11 (3), R15 (10), R18 (1) WWW131, KUN 22 Lanping Mekong 26.504 99.256 2100 H8 (9) R3 (2), R10 (14) WWW137, KUN 23 Gongshan Salween 27.845 98.684 1508 H4 (12) R3 (22) WWW147, KUN 24 Bingzhongluo Salween 27.962 98.661 1518 H4 (11) R3 (24) WWW161, KUN 25 Tongmai Yarlung Tsangpo 30.102 95.095 2115 H8 (12) R15 (4), R16 (20) WWW095, KUN 26 Yigong Yarlung Tsangpo 30.271 94.793 2204 H8 (7) R16 (24) WWW105, KUN 27 Qiubei Nanpan River 23.768 104.211 1565 H8 (12) R3 (16), R10 (7), R11 (1) 15310, KUN 28 Yanshan Nanpan River 23.714 104.018 1489 H10 (12) R3 (16), R10 (8) WSG0033, KUN 29 Jiangchuan Nanpan River 24.373 102.812 1745 H8 (12) R3 (24) WSG0005, KUN 30 Kaiyuan Nanpan River 23.786 103.616 1478 H8 (10) R3 (20), R9 (4) WSG0038, KUN 31 Jianshui Red River 23.918 102.880 1399 H8 (7) R3 (14) 15279, KUN 32 Dali Red River 25.812 100.225 1977 H8 (6) R9 (8) 15744, KUN 33 Mengzi Red River 23.379 103.449 1391 H14 (10), H15 (1) R1 (2), R3 (8), R9 (2), R12 (4), R13 (4), R14 (4) WSG0011, KUN Code . Location . Drainages . Latitude (°N) . Longitude (°E) . Elevation (m) . Haplotype frequencies . Ribotype frequencies . Voucher . 1 Wenxian Jialing River 32.762 105.013 804 H6 (12) R5 (22) WZW4, KUN 2 Longnan Jialing River 33.269 105.111 1005 H13 (11) R5 (24) WZW26, KUN 3 Baoxing Min River 30.387 102.813 956 H2 (10) R2 (22) WZW153, KUN 4 Kangding Dadu River 30.063 102.174 1410 H5 (9), H11 (3) R2 (2), R4 (1), R5 (21) WZW129, KUN 5 Danba Dadu River 31.081 101.871 2018 H5 (11), H6 (1) R2 (16), R4 (2), R5 (6) WZW173, KUN 6 Jiulong Yalong River 28.617 101.664 2130 H9 (7) R9 (14) WZW110, KUN 7 Muli Yalong River 27.926 101.287 2152 H9 (6) R1 (5), R6 (11), R9 (2) MCQ037, KUN 8 Yongsheng Jinsha River 26.697 100.719 2015 H9 (5) R1 (11), R6 (5) 15659, KUN 9 Batang Jinsha River 29.814 99.014 2474 H1 (11) R1 (14) WZW181, KUN 10 Tuoding Jinsha River 27.954 99.414 1967 H3 (8), H12 (1) R1 (20) WZW252, KUN 11 Derong Jinsha River 28.857 99.306 2540 H3 (11) R1 (22) WZW194, KUN 12 Daocheng Jinsha River 28.277 100.191 2328 H7 (12) R1 (8), R6 (4), R7 (7), R8 (1) WZW208, KUN 13 Jinjiang Jinsha River 26.996 99.950 1841 H3 (4) R1 (22) 15703, KUN 14 Judian Jinsha River 27.282 99.672 1964 H1 (9), H12 (2) R1 (24) WZW254, KUN 15 Benzilan Jinsha River 28.402 99.141 2613 H1 (1), H3 (11) R1 (24) WZW236, KUN 16 Yunling Mekong 28.348 98.901 2858 H8 (4) R3 (5), R9 (3), R10 (8) WZW243, KUN 17 Yanmen Mekong 28.075 98.929 1935 H8 (10) R3 (4), R9 (1), R10 (5), R11 (1), R18 (13) WZW248, KUN 18 Feilaishi Mekong 28.478 98.847 2977 H8 (11) R3 (9),R10 (11), R11 (2) WWW044, KUN 19 Rumei Mekong 29.659 98.368 2794 H8 (6) R10 (24) WWW048, KUN 20 Yanjing Mekong 29.065 98.614 2672 H8 (10) R10 (20), R15 (2), R17 (2) WWW047, KUN 21 Yezhi Mekong 27.662 99.014 1739 H8 (5) R3 (4), R10 (4), R11 (3), R15 (10), R18 (1) WWW131, KUN 22 Lanping Mekong 26.504 99.256 2100 H8 (9) R3 (2), R10 (14) WWW137, KUN 23 Gongshan Salween 27.845 98.684 1508 H4 (12) R3 (22) WWW147, KUN 24 Bingzhongluo Salween 27.962 98.661 1518 H4 (11) R3 (24) WWW161, KUN 25 Tongmai Yarlung Tsangpo 30.102 95.095 2115 H8 (12) R15 (4), R16 (20) WWW095, KUN 26 Yigong Yarlung Tsangpo 30.271 94.793 2204 H8 (7) R16 (24) WWW105, KUN 27 Qiubei Nanpan River 23.768 104.211 1565 H8 (12) R3 (16), R10 (7), R11 (1) 15310, KUN 28 Yanshan Nanpan River 23.714 104.018 1489 H10 (12) R3 (16), R10 (8) WSG0033, KUN 29 Jiangchuan Nanpan River 24.373 102.812 1745 H8 (12) R3 (24) WSG0005, KUN 30 Kaiyuan Nanpan River 23.786 103.616 1478 H8 (10) R3 (20), R9 (4) WSG0038, KUN 31 Jianshui Red River 23.918 102.880 1399 H8 (7) R3 (14) 15279, KUN 32 Dali Red River 25.812 100.225 1977 H8 (6) R9 (8) 15744, KUN 33 Mengzi Red River 23.379 103.449 1391 H14 (10), H15 (1) R1 (2), R3 (8), R9 (2), R12 (4), R13 (4), R14 (4) WSG0011, KUN Open in new tab Table 1. Details of sampled populations of Excoecaria acerifolia (312 and 347 individuals for plastid DNA and nuclear DNA, respectively), haplotypes and ribotypes (number of individuals) detected per population, vouchers and the assignment of each population to drainage basins Code . Location . Drainages . Latitude (°N) . Longitude (°E) . Elevation (m) . Haplotype frequencies . Ribotype frequencies . Voucher . 1 Wenxian Jialing River 32.762 105.013 804 H6 (12) R5 (22) WZW4, KUN 2 Longnan Jialing River 33.269 105.111 1005 H13 (11) R5 (24) WZW26, KUN 3 Baoxing Min River 30.387 102.813 956 H2 (10) R2 (22) WZW153, KUN 4 Kangding Dadu River 30.063 102.174 1410 H5 (9), H11 (3) R2 (2), R4 (1), R5 (21) WZW129, KUN 5 Danba Dadu River 31.081 101.871 2018 H5 (11), H6 (1) R2 (16), R4 (2), R5 (6) WZW173, KUN 6 Jiulong Yalong River 28.617 101.664 2130 H9 (7) R9 (14) WZW110, KUN 7 Muli Yalong River 27.926 101.287 2152 H9 (6) R1 (5), R6 (11), R9 (2) MCQ037, KUN 8 Yongsheng Jinsha River 26.697 100.719 2015 H9 (5) R1 (11), R6 (5) 15659, KUN 9 Batang Jinsha River 29.814 99.014 2474 H1 (11) R1 (14) WZW181, KUN 10 Tuoding Jinsha River 27.954 99.414 1967 H3 (8), H12 (1) R1 (20) WZW252, KUN 11 Derong Jinsha River 28.857 99.306 2540 H3 (11) R1 (22) WZW194, KUN 12 Daocheng Jinsha River 28.277 100.191 2328 H7 (12) R1 (8), R6 (4), R7 (7), R8 (1) WZW208, KUN 13 Jinjiang Jinsha River 26.996 99.950 1841 H3 (4) R1 (22) 15703, KUN 14 Judian Jinsha River 27.282 99.672 1964 H1 (9), H12 (2) R1 (24) WZW254, KUN 15 Benzilan Jinsha River 28.402 99.141 2613 H1 (1), H3 (11) R1 (24) WZW236, KUN 16 Yunling Mekong 28.348 98.901 2858 H8 (4) R3 (5), R9 (3), R10 (8) WZW243, KUN 17 Yanmen Mekong 28.075 98.929 1935 H8 (10) R3 (4), R9 (1), R10 (5), R11 (1), R18 (13) WZW248, KUN 18 Feilaishi Mekong 28.478 98.847 2977 H8 (11) R3 (9),R10 (11), R11 (2) WWW044, KUN 19 Rumei Mekong 29.659 98.368 2794 H8 (6) R10 (24) WWW048, KUN 20 Yanjing Mekong 29.065 98.614 2672 H8 (10) R10 (20), R15 (2), R17 (2) WWW047, KUN 21 Yezhi Mekong 27.662 99.014 1739 H8 (5) R3 (4), R10 (4), R11 (3), R15 (10), R18 (1) WWW131, KUN 22 Lanping Mekong 26.504 99.256 2100 H8 (9) R3 (2), R10 (14) WWW137, KUN 23 Gongshan Salween 27.845 98.684 1508 H4 (12) R3 (22) WWW147, KUN 24 Bingzhongluo Salween 27.962 98.661 1518 H4 (11) R3 (24) WWW161, KUN 25 Tongmai Yarlung Tsangpo 30.102 95.095 2115 H8 (12) R15 (4), R16 (20) WWW095, KUN 26 Yigong Yarlung Tsangpo 30.271 94.793 2204 H8 (7) R16 (24) WWW105, KUN 27 Qiubei Nanpan River 23.768 104.211 1565 H8 (12) R3 (16), R10 (7), R11 (1) 15310, KUN 28 Yanshan Nanpan River 23.714 104.018 1489 H10 (12) R3 (16), R10 (8) WSG0033, KUN 29 Jiangchuan Nanpan River 24.373 102.812 1745 H8 (12) R3 (24) WSG0005, KUN 30 Kaiyuan Nanpan River 23.786 103.616 1478 H8 (10) R3 (20), R9 (4) WSG0038, KUN 31 Jianshui Red River 23.918 102.880 1399 H8 (7) R3 (14) 15279, KUN 32 Dali Red River 25.812 100.225 1977 H8 (6) R9 (8) 15744, KUN 33 Mengzi Red River 23.379 103.449 1391 H14 (10), H15 (1) R1 (2), R3 (8), R9 (2), R12 (4), R13 (4), R14 (4) WSG0011, KUN Code . Location . Drainages . Latitude (°N) . Longitude (°E) . Elevation (m) . Haplotype frequencies . Ribotype frequencies . Voucher . 1 Wenxian Jialing River 32.762 105.013 804 H6 (12) R5 (22) WZW4, KUN 2 Longnan Jialing River 33.269 105.111 1005 H13 (11) R5 (24) WZW26, KUN 3 Baoxing Min River 30.387 102.813 956 H2 (10) R2 (22) WZW153, KUN 4 Kangding Dadu River 30.063 102.174 1410 H5 (9), H11 (3) R2 (2), R4 (1), R5 (21) WZW129, KUN 5 Danba Dadu River 31.081 101.871 2018 H5 (11), H6 (1) R2 (16), R4 (2), R5 (6) WZW173, KUN 6 Jiulong Yalong River 28.617 101.664 2130 H9 (7) R9 (14) WZW110, KUN 7 Muli Yalong River 27.926 101.287 2152 H9 (6) R1 (5), R6 (11), R9 (2) MCQ037, KUN 8 Yongsheng Jinsha River 26.697 100.719 2015 H9 (5) R1 (11), R6 (5) 15659, KUN 9 Batang Jinsha River 29.814 99.014 2474 H1 (11) R1 (14) WZW181, KUN 10 Tuoding Jinsha River 27.954 99.414 1967 H3 (8), H12 (1) R1 (20) WZW252, KUN 11 Derong Jinsha River 28.857 99.306 2540 H3 (11) R1 (22) WZW194, KUN 12 Daocheng Jinsha River 28.277 100.191 2328 H7 (12) R1 (8), R6 (4), R7 (7), R8 (1) WZW208, KUN 13 Jinjiang Jinsha River 26.996 99.950 1841 H3 (4) R1 (22) 15703, KUN 14 Judian Jinsha River 27.282 99.672 1964 H1 (9), H12 (2) R1 (24) WZW254, KUN 15 Benzilan Jinsha River 28.402 99.141 2613 H1 (1), H3 (11) R1 (24) WZW236, KUN 16 Yunling Mekong 28.348 98.901 2858 H8 (4) R3 (5), R9 (3), R10 (8) WZW243, KUN 17 Yanmen Mekong 28.075 98.929 1935 H8 (10) R3 (4), R9 (1), R10 (5), R11 (1), R18 (13) WZW248, KUN 18 Feilaishi Mekong 28.478 98.847 2977 H8 (11) R3 (9),R10 (11), R11 (2) WWW044, KUN 19 Rumei Mekong 29.659 98.368 2794 H8 (6) R10 (24) WWW048, KUN 20 Yanjing Mekong 29.065 98.614 2672 H8 (10) R10 (20), R15 (2), R17 (2) WWW047, KUN 21 Yezhi Mekong 27.662 99.014 1739 H8 (5) R3 (4), R10 (4), R11 (3), R15 (10), R18 (1) WWW131, KUN 22 Lanping Mekong 26.504 99.256 2100 H8 (9) R3 (2), R10 (14) WWW137, KUN 23 Gongshan Salween 27.845 98.684 1508 H4 (12) R3 (22) WWW147, KUN 24 Bingzhongluo Salween 27.962 98.661 1518 H4 (11) R3 (24) WWW161, KUN 25 Tongmai Yarlung Tsangpo 30.102 95.095 2115 H8 (12) R15 (4), R16 (20) WWW095, KUN 26 Yigong Yarlung Tsangpo 30.271 94.793 2204 H8 (7) R16 (24) WWW105, KUN 27 Qiubei Nanpan River 23.768 104.211 1565 H8 (12) R3 (16), R10 (7), R11 (1) 15310, KUN 28 Yanshan Nanpan River 23.714 104.018 1489 H10 (12) R3 (16), R10 (8) WSG0033, KUN 29 Jiangchuan Nanpan River 24.373 102.812 1745 H8 (12) R3 (24) WSG0005, KUN 30 Kaiyuan Nanpan River 23.786 103.616 1478 H8 (10) R3 (20), R9 (4) WSG0038, KUN 31 Jianshui Red River 23.918 102.880 1399 H8 (7) R3 (14) 15279, KUN 32 Dali Red River 25.812 100.225 1977 H8 (6) R9 (8) 15744, KUN 33 Mengzi Red River 23.379 103.449 1391 H14 (10), H15 (1) R1 (2), R3 (8), R9 (2), R12 (4), R13 (4), R14 (4) WSG0011, KUN Open in new tab Figure 2. Open in new tabDownload slide Geographical distributions of A, 15 haplotypes (H1–15) and B, 18 ribotypes (R1–18) identified in 33 populations of Excoecaria acerifolia. The dashed red line delineates the boundary between the lineages corresponding to the JMD, JY and YMRN drainage basins. Figure 2. Open in new tabDownload slide Geographical distributions of A, 15 haplotypes (H1–15) and B, 18 ribotypes (R1–18) identified in 33 populations of Excoecaria acerifolia. The dashed red line delineates the boundary between the lineages corresponding to the JMD, JY and YMRN drainage basins. MATERIAL AND METHODS Sampling, DNA extraction and sequencing For the phylogeographic survey of plastid and nuclear DNA sequence variation, 312 and 347 individuals, respectively were sampled from 33 populations of E. acerifolia covering most of its geographical range (Fig. 2; Table 1). Individuals were randomly sampled and spaced at least 100 m apart within each population (approximately ten individuals per population). Fresh green leaves were collected from each individual and dried in silica gel for DNA extraction. Voucher specimens from each sampled population are deposited at the Herbarium of the Kunming Institute of Botany, Chinese Academy of Sciences (Table 1). Total genomic DNA was extracted from c. 20 mg dried leaf material using a modified CTAB method (Doyle & Doyle, 1990). Based on preliminary screening of plastid fragments (Taberlet et al., 1991; Shaw et al., 2005, 2007), we chose the petL-psbE and psbA-trnH intergenic spacer regions for the full phylogeographic survey because they showed sufficient polymorphism and worked well for amplification and sequencing purposes. In addition, one nuclear DNA region (ITS) was also chosen and surveyed in this study, amplified using the PCR procedure described in Olsen & Schaal (1999). All PCR products in this study were purified and sequenced by the Sangon Corporation, Shanghai, China. DNA sequences were revised manually and aligned in BioEdit v.7.0 (Hall, 1999). Population genetic diversity and structure Plastid haplotypes and ribotypes were produced in DnaSP v.5.0 (Librado & Rozas, 2009). The method of Clark (1990) was used to identify the ribotypes for ITS sequences. In this method, chromatograms of ITS with ‘double peaks’ at polymorphic sites were further analysed by inferring the identity of ribotypes of heterogeneous individuals through ribotype subtraction (Clark, 1990; Christe et al., 2014; Qi et al., 2014). ArcMap v.10.0 (ESRI, Inc.) was used to plot the distribution of haplotypes and ribotypes on a relief map covering the range of E. acerifolia. Total haplotype diversity (HT) and average within-population diversity (HS) were estimated to evaluate the level of genetic variation (Nei, 1975). GST and NST were used to calculate the genetic differentiation between populations, and compared using U-statistics. Usually, a significantly higher NST than GST indicates the presence of a phylogeographic structure (Pons & Petit, 1996). All these parameters were calculated using the program Permut v.1.0 (available at http://www.pierroton.inra.fr/genetics/labo/Software/Permut) with 1000 permutations. Spatial analysis of molecular variance (SAMOVA) was conducted to examine the spatial genetic structure of haplotypes and ribotypes using the program SAMOVA v.1.0 (Dupanloup, Schneider & Excoffier, 2002). This program uses a simulated annealing approach to define K groups of populations that are geographically homogeneous and maximally differentiated from each other. The program was run for 1000 iterations for K = (2, 3,…, 12) from each of 100 random initial conditions, and a corresponding FCT index, which measures the proportion of genetic differentiation among the K groups was calculated. The configuration with the FCT value that reaches a plateau first is usually considered the optimal grouping of populations. A hierarchical analysis of molecular variance (AMOVA) was conducted in Arlequin v.3.5 (Excoffier & Lischer, 2010) with significance tests of variance components based on 1000 permutations, to elucidate the genetic differentiation between groups (as identified by SAMOVA) further, as well as revealing differences between populations within groups and individuals within populations. Genealogical relationships between haplotypes and ribotypes were inferred from an unrooted statistical parsimony network within Network v.5.0 (Bandelt, Forster & Röhl, 1999) using a median-joining method. Indels (gaps) were treated separately as binary states, namely single events, following Caicedo & Schaal (2004). Mononucleotide repeats and duplicated indels of sequences in this study were removed as they introduced homoplasious reticulations into the network, due to their high levels of bidirectional mutation (Xia et al., 2009). Observed and inferred haplotypes and ribotypes were nested following the rules defined by Posada & Crandall (2001). Phylogenetic analysis of haplotypes and ribotypes was reconstructed by Bayesian inference. This analysis was conducted in MrBayes v.3.1.2 (Ronquist & Huelsenbeck, 2003) using substitution models (GTR+G) explored by Modeltest v.3.7 (Posada & Crandall, 1998), according to the Akaike information criterion (AIC). Four independent Markov chain Monte Carlo analyses with runs of ten million generations were carried out, sampling every 1000 generations. The first 20% of generations were discarded as burn-in, and the remaining data were used to construct a 50% majority rule consensus tree. The posterior probabilities found in this consensus tree were used to evaluate the robustness of the trees. Divergence times estimation The divergence times of lineages identified in the phylogenetic analysis was estimated using a Bayesian approach as implemented in BEAST v.1.8.4 (Drummond & Rambaut, 2007). Estimations of divergence times for each data set in BEAST were run under the assumption of a strict clock, using the following criteria: (1) the GTR substitution model with estimated base frequencies and a site heterogeneity model of gamma; and (2) a coalescent tree prior that assumes a flexible model ‘skyline plot’ (Pybus, Rambaut & Harvey, 2000). No fossil records or taxon-specific substitution rates were available to calibrate our plastid or nuclear DNA regions. Therefore, we used the substitution rates of plastid DNA in angiosperms and that of ITS in woody perennial plants to estimate the divergence times. In BEAST, we assumed a normal distribution with mean 2.0 × 10–9 substitutions per site per year (s/s/y) and a standard deviation of 0.51 × 10–9 s/s/y for plastid DNA. This will give a central 95% range of c. 1.0 × 10–9 to 3.0 × 10–9 s/s/y, and correspond to the estimated substitution rates of plastid DNA in angiosperms (Wolfe, Li & Sharp, 1987; Richardson et al., 2001). Similarly, we also assumed a normal distribution centred at 4.11 × 10–9 s/s/y with a standard deviation of 1.9 × 10–9 s/s/y, which will include a central 95% range of c. 0.38 × 10–9 to 7.83 × 10–9 s/s/y corresponding to the estimated substitution rates of ITS in woody perennial plants (Kay et al., 2006). Results from BEAST were compiled and visualized in the program Tracer v.1.7.1 (available at https://github.com/beast-dev/tracer/releases/tag/v1.7.1). Demographic history Mismatch distribution analysis (Rogers & Harpending, 1992) based on plastid DNA was carried out to infer the demographic history of E. acerifolia as a whole and for identified population genetic groups. In this analysis, an expected distribution was generated from 1000 parametric bootstrap replicates to test demographic expansion and spatial expansion models, respectively. Generally, a unimodal mismatch distribution curve indicates a recent population expansion, whereas a multimodal distribution usually suggests demographic stability (Rogers & Harpending, 1992). The goodness of fit of observed-to-theoretical mismatch distributions under a sudden (stepwise) expansion model was tested with the sum of squared deviations (SSD) and Harpending’s raggedness index (HRag; Rogers & Harpending, 1992; Harpending, 1994) based on 1000 parametric bootstrap replicates. In addition, neutrality test parameters, namely Tajima’s D (Tajima, 1989, 1996) and Fu’s Fs (Fu, 1997), were also calculated for the total populations and each group of populations to further elucidate the demographic history of E. acerifolia. All the aforementioned analyses were performed in Arlequin v.3.5 (Excoffier & Lischer, 2010). In the populations for which the hypothesis of expansion was not rejected, the value for the mode of the mismatch distribution (τ) was assessed based on 1000 parametric bootstrap replicates and converted into estimates of time since expansion (t) using t = τ/2u (Rogers & Harpending, 1992), where u = μkg. In this formula, μ is the mutation rate of the sequence (2.0 × 10–9 s/s/y in this study), k is the number of nucleotides (1305 bp in this study) and g is the generation time in years (assumed to be six years based on personal observations of E. acerifolia cultivated at Kunming Botanical Garden). Ecological niche modelling Ecological niche modelling (ENM) can simulate past distributions of species and further elucidate demographic history through time (Wang et al., 2015a). Thus, it was performed in this study to examine the suitable distribution of E. acerifolia at the Last Interglacial (LIG, resolution 30 s), Last Glacial Maximum (LGM, resolution 2.5 min, CCSM4) and in the current period (resolution 2.5 min). Initially, to eliminate the effects of multi-colinearity, Spearman correlation analysis was employed to calculate the correlation coefficient among 19 bioclimatic variables (Supporting Information, Table S1). Then, predictors with correlation coefficient < 0.85 were selected as the input environmental layers for ENM. Ultimately, eight candidate bioclimatic variables (bio1, 2, 3, 7, 10, 12, 14, 15) were selected in this study for modelling habitat shift of this species. Based on the populations included in this study (Table 1) and herbarium records (IBSC, KUN and PE; Table S2) of E. acerifolia, we conducted an ENM analysis in Maxent v.3.3.3k (Phillips, Anderson & Schapire, 2006) using the eight candidate bioclimatic variables from the WorldClim database (Hijmans et al., 2005) with the default parameters for number of iterations (5000) and convergence threshold (10–5). To ensure more reliable results, we set up 15 replicate runs for each analysis. Meanwhile, 25% of the data in each run was randomly chosen by Maxent and compared with the model output created with the remaining data, to test the performance of each model. In addition, to reduce the effects of spatial autocorrelation, duplicate records from the same locality were removed. The area under the receiver operating characteristic curves (AUC) was used to compare model performance. An AUC value of 0.5 indicates that the performance of the model is no better than random, while values closer to 1.0 indicate better model performance (Phillips et al., 2006). Finally, the output of the logistic layer produced from Maxent was reclassified into a binary prediction map (unsuitable and suitable) with a threshold of ten percentile training presence. We quantified the potential range size by summing the total number of suitable cell with probability value above the threshold of ten percentile training presence. The suitable cell of LIG was converted to the same level as LGM (or current) based on the resolution of layers, using formula suitable cell = original suitable cell/25. All suitable range-size estimation and geographical plotting were conducted in ArcGIS v.10.0. RESULTS Population genetic diversity Variations in the length of two plastid DNA regions were detected (598–648 bp, trnH-psbA; 750–756 bp, petL-psbE) in our samples. Removal of mononucleotide repeats and duplicated indels for the alignment of the plastid DNA regions from the 312 samples resulted in an overall alignment length of 1305 bp and produced 15 haplotypes (H1–15; Table S3). The aligned sequences of the ITS from the 347 samples were 679 bp long and produced 18 ribotypes (R1-18; Table S4). Newly generated sequences have been deposited in GenBank (accession numbers MK084701-MK084744). The total haplotype and ribotype diversity (plastid DNA: HT = 0.806 ± 0.0636; nuclear DNA: HT = 0.870 ± 0.0235) was high, and the average within-population diversity was low (plastid DNA: HS = 0.045 ± 0.0181; nuclear DNA: HS = 0.244 ± 0.0498). Both the inter-population differentiation parameters were also high (plastid DNA: NST = 0.988 ± 0.0050, GST = 0.945 ± 0.0211; nuclear DNA: NST = 0.882 ± 0.0325, GST = 0.720 ± 0.0561). The U-test showed that NST was significantly higher than GST (plastid DNA: U = 2.55, P < 0.05; nuclear DNA: U = 2.50, P < 0.05), indicating significant phylogeographic structure of the haplotype and ribotype distribution. Population genetic structure and divergence times Based on the SAMOVA analysis of the plastid DNA sequences, the FCT value first reached a peak when K = 3 (Supporting Information, Fig. S1A), with the three groups of populations corresponding to the JMD, JY and YMRN drainage basins (Fig. 2A). The FCT value reached a plateau when K = 6 (Supporting Information, Fig. S1A), but the case of six groups only made the JY and YMRN groups of populations further divide into two and three subgroups, respectively, when compared to the case of three groups. Although the FCT value was not the highest when K = 3, the AMOVA revealed high genetic variation partitioned between the groups (78.03%), and low variation among populations within groups (20.97%) and within populations (1.00%). Moreover, genealogical analysis (network) and the Bayesian phylogenetic tree based on plastid DNA revealed three clusters/lineages of haplotypes, also corresponding to the JMD, JY and YMRN drainage basins (Figs 3A, 4A). In the network, the JMD, JY and YMRN clusters of haplotypes were all derived from one internal point, but showed different numbers of mutation sites from this point (Fig. 3A). In the phylogenetic tree, the JY lineage of haplotypes differentiated first and formed the sister clade to the large clade formed by the JMD and YMRN lineages of haplotypes. In this large clade, haplotypes from JMD drainage basins formed a clade, whereas the YMRN haplotypes formed two (Fig. 4A). This suggests a close genetic link within populations of the JMD, JY and YMRN drainage basins. Given this, we chose the case of three groups based on SAMOVA for consistency with the genealogical and Bayesian phylogenetic analysis in this study. Figure 3. Open in new tabDownload slide Median-joining networks of A, 15 haplotypes (H1–15) and B, 18 ribotypes (R1–18) identified in Excoecaria acerifolia in this study. The size of circles corresponds to the frequency of each haplotype and ribotype. Short lines and red dots (A) represent haplotypes and ribotypes missing from the dataset and each represents one mutation. Figure 3. Open in new tabDownload slide Median-joining networks of A, 15 haplotypes (H1–15) and B, 18 ribotypes (R1–18) identified in Excoecaria acerifolia in this study. The size of circles corresponds to the frequency of each haplotype and ribotype. Short lines and red dots (A) represent haplotypes and ribotypes missing from the dataset and each represents one mutation. Figure 4. Open in new tabDownload slide A, Bayesian phylogenetic trees of 15 haplotypes (H1–15) and B, 18 ribotypes (R1–18) identified in Excoecaria acerifolia in this study. Numbers above the branch indicate posterior probabilities. Numbers next to the nodes indicate the divergence times and 95% highest posterior density (HPD) intervals. Figure 4. Open in new tabDownload slide A, Bayesian phylogenetic trees of 15 haplotypes (H1–15) and B, 18 ribotypes (R1–18) identified in Excoecaria acerifolia in this study. Numbers above the branch indicate posterior probabilities. Numbers next to the nodes indicate the divergence times and 95% highest posterior density (HPD) intervals. In the SAMOVA analysis based on ITS sequences, the FCT value reached a plateau and fluctuated little with the increasing number of groups when K = 7 (Supporting Information, Fig. S1B). Therefore, seven groups were identified in this scenario, also largely consistent with the three group case based on plastid DNA, with four groups located in the JMD, one group in the JY and two groups in the YMRN drainage basin, respectively (Fig. 2B). The AMOVA also revealed relatively high genetic variation partitioned between groups (83.35%), compared with 8.09% between populations and 8.55% within populations. Genealogical analysis and the Bayesian phylogenetic tree of ribotypes had a similar structure, generally agreeing with the results based on plastid DNA (Figs 3B, 4B). According to our BEAST-derived age estimates, the divergence time of haplotypes among the three lineages in the JMD, JY and YMRN drainage basins fell in the early Pleistocene [1.77 (0.71–3.52) Mya, 1.18 (0.437–2.39) Mya; Fig. 4A], whereas the diversification within lineages generally fell in the mid Pleistocene [JY lineage, 0.621 (0.157–1.40) Mya, 0.187 (0.0088–0.526) Mya; YMRN lineage, 0.193(0.0015–0.633) Mya, 0.099(0.000014–0.455) Mya; JMD lineage, 0.412 (0.083–0.954) Mya, 0.219 (0.037–0.541) Mya; Fig. 4A]. The divergence time of ribotypes was estimated to be 2.45 (0.577–8.82) Mya, also in the early Pleistocene, and the diversification within lineages was also dated back to the mid Pleistocene [JY lineage, 0.481 (0.0653–1.84) Mya, 0.276 (0.021–1.14) Mya, 0.120 (0.001–0.567) Mya; YMRN lineage: 0.174 (0.0012–0.809) Mya, 0.290 (0.0325–1.19) Mya, 0.123 (0.001–0.577) Mya; JMD lineage: 0.530 (0.0463–2.18) Mya, 0.156 (0.001–0.72) Mya; Fig. 4B], which further supports the results based on plastid DNA. Demographic history and ENM The mismatch distribution curves for the total populations were multimodal (Fig. 5A, B) based on both demographic expansion and spatial expansion models, which seem to suggest the stability of the total populations. Under the demographic expansion model, this result was further supported by the significant SSD and HRag values, and the positive and non-significant Tajima’s D and Fu’s Fs values (Table 2). Under the spatial expansion model, however, the stability was not supported by all the values (not by the non-significant SSD and HRag). Figure 5. Open in new tabDownload slide A, B, Mismatch distribution curves of the total; C, D, JY; E, F, YMRN and G, H, JMD populations of Excoecaria acerifolia. Left: demographic expansion model, right: spatial expansion model. Figure 5. Open in new tabDownload slide A, B, Mismatch distribution curves of the total; C, D, JY; E, F, YMRN and G, H, JMD populations of Excoecaria acerifolia. Left: demographic expansion model, right: spatial expansion model. Table 2. Mismatch distribution analysis (MDA) of Excoecaria acerifolia for all populations and the three groups (JMD, JY and YMRN) of populations, with the sum of squared deviations (SSD), Harpending’s (1994) raggedness index (HRag), Tajima’s D, Fu’s Fs and their P values. Upper and lower 95% confidence limits around estimates of τ value and associated expansion time of (t; Kya) are in parentheses Populations . Expansion types . SSD . P . HRag . P . Tajima’s D . P . Fu’s Fs . P . τ . MDA . t (Kya) . JMD Demographic expansion 0.01001 0.087ns 0.10927 0.040* 0.46237 0.659 ns 0.1927 0.598 ns 1.344 (0.906–1.877) unimodal 42.91 （28.93– 59.93） Spatial expansion 0.01001 0.032* 0.10927 0.042* 0.46237 0.685 ns 0.1927 0.641 ns 1.344 (0.717–1.752) unimodal 42.91 （22.89– 55.93） JY Demographic expansion 0.11938 0.041* 0.40989 0.000** 6.34424 0.999 ns 1.65139 0.952 ns 7.373（0.223– 12.385） multimodal NC Spatial expansion 0.11228 0.005** 0.40989 0.060ns 6.34424 0.998 ns 1.65139 0.933 ns 6.632（3.787–9.840 ） multimodal NC YMRN Demographic expansion 0.01895 0.220 ns 0.37270 0.570 ns 0.71792 0.658 ns −0.36612 0.383 ns 3.000 (0.402–3.188) unimodal 95.79 （12.84– 101.79 ） Spatial expansion 0.01013 0.398 ns 0.37270 0.656 ns 0.71792 0.669 ns −0.36612 0.347 ns 4.326 (0.000–63.312) unimodal 138.12 （0.0– 2021.46 ） Total Demographic expansion 0.03480 0.010* 0.04156 0.014* 6.24835 0.983 ns 2.16239 0.973 ns 10.131（5.090– 13.502） multimodal NC Spatial expansion 0.01223 0.747 ns 0.04156 0.860 ns 6.24835 0.992 ns 2.16239 0.97 ns 8.898 （5.780–12.600 multimodal NC Populations . Expansion types . SSD . P . HRag . P . Tajima’s D . P . Fu’s Fs . P . τ . MDA . t (Kya) . JMD Demographic expansion 0.01001 0.087ns 0.10927 0.040* 0.46237 0.659 ns 0.1927 0.598 ns 1.344 (0.906–1.877) unimodal 42.91 （28.93– 59.93） Spatial expansion 0.01001 0.032* 0.10927 0.042* 0.46237 0.685 ns 0.1927 0.641 ns 1.344 (0.717–1.752) unimodal 42.91 （22.89– 55.93） JY Demographic expansion 0.11938 0.041* 0.40989 0.000** 6.34424 0.999 ns 1.65139 0.952 ns 7.373（0.223– 12.385） multimodal NC Spatial expansion 0.11228 0.005** 0.40989 0.060ns 6.34424 0.998 ns 1.65139 0.933 ns 6.632（3.787–9.840 ） multimodal NC YMRN Demographic expansion 0.01895 0.220 ns 0.37270 0.570 ns 0.71792 0.658 ns −0.36612 0.383 ns 3.000 (0.402–3.188) unimodal 95.79 （12.84– 101.79 ） Spatial expansion 0.01013 0.398 ns 0.37270 0.656 ns 0.71792 0.669 ns −0.36612 0.347 ns 4.326 (0.000–63.312) unimodal 138.12 （0.0– 2021.46 ） Total Demographic expansion 0.03480 0.010* 0.04156 0.014* 6.24835 0.983 ns 2.16239 0.973 ns 10.131（5.090– 13.502） multimodal NC Spatial expansion 0.01223 0.747 ns 0.04156 0.860 ns 6.24835 0.992 ns 2.16239 0.97 ns 8.898 （5.780–12.600 multimodal NC NC, not calculated, nsNon-significant, *P < 0.05, **P < 0.01 Open in new tab Table 2. Mismatch distribution analysis (MDA) of Excoecaria acerifolia for all populations and the three groups (JMD, JY and YMRN) of populations, with the sum of squared deviations (SSD), Harpending’s (1994) raggedness index (HRag), Tajima’s D, Fu’s Fs and their P values. Upper and lower 95% confidence limits around estimates of τ value and associated expansion time of (t; Kya) are in parentheses Populations . Expansion types . SSD . P . HRag . P . Tajima’s D . P . Fu’s Fs . P . τ . MDA . t (Kya) . JMD Demographic expansion 0.01001 0.087ns 0.10927 0.040* 0.46237 0.659 ns 0.1927 0.598 ns 1.344 (0.906–1.877) unimodal 42.91 （28.93– 59.93） Spatial expansion 0.01001 0.032* 0.10927 0.042* 0.46237 0.685 ns 0.1927 0.641 ns 1.344 (0.717–1.752) unimodal 42.91 （22.89– 55.93） JY Demographic expansion 0.11938 0.041* 0.40989 0.000** 6.34424 0.999 ns 1.65139 0.952 ns 7.373（0.223– 12.385） multimodal NC Spatial expansion 0.11228 0.005** 0.40989 0.060ns 6.34424 0.998 ns 1.65139 0.933 ns 6.632（3.787–9.840 ） multimodal NC YMRN Demographic expansion 0.01895 0.220 ns 0.37270 0.570 ns 0.71792 0.658 ns −0.36612 0.383 ns 3.000 (0.402–3.188) unimodal 95.79 （12.84– 101.79 ） Spatial expansion 0.01013 0.398 ns 0.37270 0.656 ns 0.71792 0.669 ns −0.36612 0.347 ns 4.326 (0.000–63.312) unimodal 138.12 （0.0– 2021.46 ） Total Demographic expansion 0.03480 0.010* 0.04156 0.014* 6.24835 0.983 ns 2.16239 0.973 ns 10.131（5.090– 13.502） multimodal NC Spatial expansion 0.01223 0.747 ns 0.04156 0.860 ns 6.24835 0.992 ns 2.16239 0.97 ns 8.898 （5.780–12.600 multimodal NC Populations . Expansion types . SSD . P . HRag . P . Tajima’s D . P . Fu’s Fs . P . τ . MDA . t (Kya) . JMD Demographic expansion 0.01001 0.087ns 0.10927 0.040* 0.46237 0.659 ns 0.1927 0.598 ns 1.344 (0.906–1.877) unimodal 42.91 （28.93– 59.93） Spatial expansion 0.01001 0.032* 0.10927 0.042* 0.46237 0.685 ns 0.1927 0.641 ns 1.344 (0.717–1.752) unimodal 42.91 （22.89– 55.93） JY Demographic expansion 0.11938 0.041* 0.40989 0.000** 6.34424 0.999 ns 1.65139 0.952 ns 7.373（0.223– 12.385） multimodal NC Spatial expansion 0.11228 0.005** 0.40989 0.060ns 6.34424 0.998 ns 1.65139 0.933 ns 6.632（3.787–9.840 ） multimodal NC YMRN Demographic expansion 0.01895 0.220 ns 0.37270 0.570 ns 0.71792 0.658 ns −0.36612 0.383 ns 3.000 (0.402–3.188) unimodal 95.79 （12.84– 101.79 ） Spatial expansion 0.01013 0.398 ns 0.37270 0.656 ns 0.71792 0.669 ns −0.36612 0.347 ns 4.326 (0.000–63.312) unimodal 138.12 （0.0– 2021.46 ） Total Demographic expansion 0.03480 0.010* 0.04156 0.014* 6.24835 0.983 ns 2.16239 0.973 ns 10.131（5.090– 13.502） multimodal NC Spatial expansion 0.01223 0.747 ns 0.04156 0.860 ns 6.24835 0.992 ns 2.16239 0.97 ns 8.898 （5.780–12.600 multimodal NC NC, not calculated, nsNon-significant, *P < 0.05, **P < 0.01 Open in new tab The mismatch distribution curves for the JY populations were multimodal (Fig. 5C, D), with most of the SSD and HRag values significant (Table 2). The Tajima’s D and Fu’s Fs values were positive and non-significant. All of these indicated the recent stability of JY populations. The mismatch distribution curve for the YMRN populations was unimodal (Fig. 5E, F), and was well supported by the non-significant SSD and HRag values as well as negative Fu’s Fs values (Table 2). Therefore, the hypothesis of population expansion for the YMRN populations could not be rejected. Based on the τ value, the spatial and demographic expansion times for this group of populations were estimated to be late in the penultimate glaciation and the Last Interglacial period, respectively (138.12 Kya, 95.79 Kya; Table 2). The mismatch distribution curves for the JMD populations was also unimodal (Fig. 5H, I), but with mostly significant SSD and HRag values. Furthermore, Tajima’s D and Fu’s Fs values were positive and non-significant (Table 2). Under these circumstances, the hypothesis of population expansion for the JMD populations can probably be rejected. However, to provide a time reference for the ENM, we also estimated the expansion times for the JMD populations, falling in the last glaciation (42.91 Kya; Table 2). The AUCs of the ENM for the YMRN, JMD and JY populations under the climate scenarios were ≥ 0.972, indicating that the ENM performed better than random. Palaeo-distribution modelling for the YMRN populations indicated a wider distribution during the LIG than the current distribution (suitable cell 837639/25 = 33 506 vs. 27 716, 5790 more), particularly significant in the more southerly Red-Nanpan drainage basins (Fig. 6A, B). Compared to the YMRN populations, palaeo-distribution modelling for the JMD and the JY populations showed much less difference in the distribution ranges during the LGM with the Current period (JMD, suitable cell 8048 vs. 8612, 564 less; JY, suitable cell 13 553 vs. 12 569, 984 more; Fig. 6C–F), further supporting the demographic stability of the JMD and JY populations. Figure 6. Open in new tabDownload slide Ecological niche modelling for Excoecaria acerifolia using bioclimatic factors for periods (LIG; LGM; current) based on the YMRN, JMD and JY populations. Red colour indicates areas with suitability (suitable cell; YMRN, A vs. B: 33506 vs. 27716, 5790 more; JMD, C vs. D: 8048 vs. 8612, 564 less; JY, E vs. F:13553 vs. 12569, 984 more). Figure 6. Open in new tabDownload slide Ecological niche modelling for Excoecaria acerifolia using bioclimatic factors for periods (LIG; LGM; current) based on the YMRN, JMD and JY populations. Red colour indicates areas with suitability (suitable cell; YMRN, A vs. B: 33506 vs. 27716, 5790 more; JMD, C vs. D: 8048 vs. 8612, 564 less; JY, E vs. F:13553 vs. 12569, 984 more). DISCUSSION Major lineage divergence A major finding of the present study is that E. acerifolia comprises three geographically distinct lineages, as inferred from the SAMOVA and the phylogenetic and network analysis of haplotypes, corresponding to the JMD, JY and YMRN drainage basins (Figs 2A, 3A, 4A), which was further supported by the significantly higher NST than GST (plastid DNA: U = 2.55, P < 0.05). This significant phylogeographic structure is indicative of long-term impediments to gene flow among the regional populations. Moreover, the estimated differentiation value (GST) was higher than the mean genetic differentiation calculated for chloroplast sequences in angiosperms (0.945 vs. 0.637; Petit et al., 2005) and that calculated for riparian plants [e.g. Terminalia franchetii, 0.841 (Zhang & Sun, 2011) and Buddleja crispa, 0.89 (Yue et al., 2012)], also indicating the existence of strong barriers to seed flow between the three major lineages of E. acerifolia. Seeds do not seem to be able to cross the mountain barriers between the JMD, JY and YMRN drainage basins. This accords with the species’ habitat being restricted to deep river valleys and usually gravity-dispersed seeds (Jin, 1998; Li & Esser, 2008; Das et al., 2011; Mondal, Ghosh & Ramakrishna, 2016; Rawat et al., 2017). The haplotype (H4) from two populations of the Salween (populations 23–24) drainage basin clustered with the haplotypes from the Jinsha and Yalong rivers (Figs 2A, 3A). This could be explained by either long-distance dispersal or the historic drainage link between the Jinsha, Yalong and Salween rivers. Long-distance dispersal is unlikely because of the lack of a suitable mechanism in E. acerifolia. The capsules of E. acerifolia cannot be dispersed by wind since it is not winged (Li & Esser, 2008), and cannot be spread by birds because it is poisonous and non-fleshy (Webster, 1975; Das et al., 2011; Rawat et al., 2017; Yeragi & Mendhulkar, 2017). Given this, it is difficult to believe that the capsules and seeds could be transported over huge mountains (e.g. the Yunling and Nushan mountains, both > 4500 m in elevation) between the Yalong, Jinsha and Salween rivers (Yue et al., 2011; Liu et al., 2013). In contrast, the historic drainage link seems a more plausible explanation for this genetic relationship and disjunction. Although these rivers are now separated by high mountain ridges and each has its own drainage basin, they were once connected to the single southward-flowing Palaeo-Red River system (Brookfield, 1998; Clark et al., 2004; Clift et al., 2006; Ming et al., 2006, 2007). This could have allowed seed-based gene flow between the Jinsha, Yalong and Salween Rivers through the palaeo-river courses. Such a scenario is not unique and was found in the riparian genus Salweenia (discontinuously distributed in Yalong and Salween drainage basins), also indicating close genetic relationship and historic drainage link between the Yalong and Salween rivers (Yue et al., 2011). The SAMOVA, phylogenetic and network analysis of the biparentally inherited nuclear DNA in this study support the significant phylogeographic structure based on plastid DNA in general, also showing significant genetic divergence among the JMD, JY and YMRN drainage basins (Figs 2B, 3B, 4B). Significantly higher NST than GST (nuclear DNA: U = 2.50, P < 0.05) and a high level of nuclear variation (83.35%) due to differences between nuclear DNA groups based on the AMOVA, both suggest that pollen is not dispersed over wide geographical distances. Therefore, we conclude that the sharing of the common nuclear ribotype (R9) between the JY and YMRN drainage basins (Fig. 2B) is most probably due to the retention of ancestral polymorphisms, that is, there has been insufficient time for lineage sorting at the nuclear DNA level, rather than a high contemporary pollen flow (Neigel & Avise, 1986; Chiang & Schaal, 2006). Geological and climatic changes are considered to be two major precursors of population genetic divergence (Andrew et al., 2010; Yan et al., 2012; Liu et al., 2013; Wang, Glor & Losos, 2013; Luo et al., 2016). Our divergence times based on both datasets suggest that the three major lineages diverged in the early Pleistocene [plastid DNA: 1.77 (0.77–3.52) Mya, 1.18 (0.437–2.39) Mya; nuclear DNA: 2.45 (0.577–8.82) Mya; Fig. 4A, B], all before the earliest glaciation (1.17–0.80 Myr) in the Pleistocene (Zheng et al., 2002). The climate in the Sino–Himalayan and adjacent regions became quite severe only after the Naynayxungla glaciation (0.72–0.50 Mya) and was warm before that (Zheng et al., 2002). Therefore, climatic changes in the Pleistocene are unlikely to explain the divergences of the three major lineages of E. acerifolia. In contrast, our divergence times correspond better with the drainage reorganization events during the late Pliocene to mid Pleistocene (Brookfield, 1998; Li, Xie & Kuang, 2001; Ming et al., 2006, 2007; Kong et al., 2009). Meanwhile, all haplotypes were derived from one internal point (probably the ancestral but missing haplotype; Fig. 3A), probably just suggests the palaeo-drainage link between these rivers. In addition, our genetic divergence times coincide with recent findings relating to riparian plants in the Sino–Himalayan region [e.g. T. franchetii, 1.44–4.24 Mya (Zhang et al., 2011); B. crispa, 1.228–3.683 Mya (Yue et al., 2012); Salweenia, 3.03–6.06 Mya (Yue et al., 2011)] and are broadly congruent with results from studies of freshwater fishes, e.g. 0.42–8.0 Mya for schizothoracine fishes (He et al., 2004), 4.4–6.8 Mya for Schizothorax (He & Chen, 2006) and 1.4–10.9 Mya for glyptosternoid catfishes (Peng et al., 2006). These further support the explanation of drainage reorganization events since the early Pleistocene for the genetic divergence of the three major lineages of E. acerifolia. Finally, our results demonstrated a closer genetic relationship between the YMRN and JMD populations, even with JY populations lying between these two groups (Fig. 4A). This probably implies that the JY two rivers (Jinsha and Yalong rivers) were probably the first to be separated from the Palaeo-Red River, whereas the other rivers (the JMD rivers and YMRN rivers) separated later. According to geological evidence, the Jinsha (upper Yangtze River) and Yalong rivers changed their direction and flowed eastward after the disconnection with the Palaeo-Red River (Brookfield, 1998; Clark et al., 2004; Clift et al., 2006; Ming et al., 2006, 2007; Fig. 1B). If we infer that the separation of the JY rivers from the Palaeo-Red River was earlier than that of the JMD rivers (located to the east of the JY rivers; Fig. 1A), however, it may face a problem of how to store the water of Jinsha and Yalong rivers since the still southward-flowing JMD rivers would hinder the eastward flow of the JY rivers. Several studies shed some light on this problem. Lacustrine sediments hundreds of metres thick widely spread along the middle Yangtze River, suggesting that an event blocked the river and confirming the existence of this palaeo-lake during the late Pliocene to early Pleistocene (Jiang et al., 1999; Kong et al., 2009; Wang et al., 2011). These studies are in agreement with our results and probably solve the problem. However, more studies on riparian plants endemic to this region are needed to further verify this result and inference. Genetic diversification in lineages Another finding of the present study is that the genetic diversification in lineages of E. acerifolia was probably affected by climatic changes starting from the mid Pleistocene. Our genetic divergence times for the haplotypes and ribotypes in lineages fell into the mid Pleistocene (Fig. 4A, B) and broadly coincide with the Naynayxungla glaciation (0.72–0.50 Mya) and the penultimate glaciation (0.30–0.130 Mya; Zheng et al., 2002). The Naynayxungla glaciation is regarded as the most extensive glaciation in the Sino–Himalayan and adjacent regions: there were many large ice caps, glacier complexes and great valley glaciers, which led to the coldest climate in the Pleistocene (Zheng et al., 2002). The penultimate glaciation was the second most severe glaciation in these regions. Although the extent of glaciers during the penultimate glaciation on the inner plateau was smaller than during the Naynayxungla glaciation, the glacier extent was larger than that of the Naynayxungla glaciation in the southern Himalayas and south-eastern Tibet. In particular, this glaciation was characterized by the formation of many valley glaciers [e.g. Lijiang Valley glacier (259–316 Kya) and Jilongshi Valley glacier (200 Kya)], including several glaciers in river valleys that were even larger than those present during the Naynayxungla glaciation (Zheng et al., 2002). Thus, it is feasible that these two glaciations and their induced climatic changes resulted in considerable fragmentation of subnival habitat, which in turn may have triggered the genetic diversification within the lineages of E. acerifolia since the mid Pleistocene. Such a similar impact of glaciations on diversification is reported in a growing body of phylogeographic studies of (sub)alpine plants in the Sino–Himalayan and adjacent regions in recent years (Chen et al., 2008; Wang et al., 2008; Li et al., 2011; Qiu et al., 2011; Luo et al., 2016), but limited evidence pertaining to riparian plants in these regions (Yang et al., 2014). Diversification of E. acerifolia in this study provides one of the few pieces of evidence indicating that glaciations may have also acted on riparian plants of the hot/warm-dry river valleys. In summary, the above results suggest a combined effect of geological and climatic changes on the diversity of lower latitude plants in the Sino–Himalayan region, one of the key biodiversity hotspots of in the world (Myers et al., 2000). Such a scenario has also been reported in other parts of the world. For instance, a combined effect of Neogene tectonic events and Pleistocene climatic changes was also proposed to account for the biodiversity of the Neotropical region, another one of the most biodiverse regions of the world (Myers et al., 2000; Rull, 2011, 2018; Antonelli et al., 2018a, b), and this has been confirmed by many meta-analyses of Neotropical species (e.g. Rull, 2008; Turchetto-Zolet et al., 2013; Antonelli et al., 2018b). The discovery of our study, along with other previously studies of high latitude plants in the Sino–Himalayan region (e.g. Liu et al., 2013; Luo et al., 2016) and the studies performed on species of the Neotropical region (Antonelli et al., 2018a; Rull, 2018), suggested the commonness and importance of the combined effects of geological and climatic changes on the diversity of the key biodiversity hotspots in the world (Myers et al., 2000). Demographic history and ENM Multimodal mismatch distribution curves (Fig. 5A, B) were identified for the total populations of E. acerifolia under demographic expansion and spatial expansion models. Non-significant and positive Tajima’s D and Fu’s Fs values further confirmed these (Table 2). All of these seem to imply the stability of the total populations. Significant SSD and HRag were found for the demographic expansion model, but not for the spatial expansion model (Table 2). This inconsistency may result from the different demographic histories within the three groups of populations of E. acerifolia. For the JY populations, multimodal mismatch distribution curves with mostly significant SSD and HRag values were found (Table 2; Fig. 5C, D) based on both models, which probably imply its stability. It was supported by positive and non-significant Tajima’s D and Fu’s Fs values (Table 2), and was further confirmed by the ENM showing not much difference in the distribution ranges during the LGM with the Current period (suitable cell 13 553 vs. 12 569; Fig. 6E, F). Unimodal pairwise mismatch distribution curves for the JMD and YMRN populations were identified in this study (Fig. 5E–I), but these probably need to be treated separately. For the JMD populations, although the mismatch distribution curve was unimodal with expansion times estimated to be in the last glaciation (42.91 Kya), this was neither supported by the mostly significant SSD and HRag values nor the non-significant and positive Tajima’s D and Fu’s Fs values (Table 2). Moreover, the ENM analysis for the JMD populations showed little difference in distribution ranges during the LGM with the Current period (suitable cell 8048 vs. 8612; Fig. 6C, D). Hence, the scenario for the JMD population is more similar to stability than expansion. For the YMRN populations, in contrast, the unimodal mismatch distribution curve was better supported by the non-significant SSD and HRag values and negative Fu’s Fs values (Table 2), indicating the recent expansion of the YMRN populations. The expansion times fell in the range from the late part of the penultimate glaciation (138.12 Kya) to the Last Interglacial (95.79 Kya; Table 2), which was supported by ENM analysis for YMRN populations, showing larger distribution ranges during the LIG than during the Current period (suitable cell 33 506 vs. 27 716, 5790 more; Fig. 6A, B). For the expansion of the YMRN populations, one explanation is that glaciations in the Pleistocene became progressively less extensive after the most severe Naynayxungla glaciation in this region, particularly after 0.17 Mya (Schäfer et al., 2002; Zheng et al., 2002; Owen & Lewis, 2008). Therefore, during the period from 138.12 Kya to 95.79 Kya, the intensity of the penultimate glaciation may have weakened further and the climate probably became warm through the Last Interglacial period. Given the E. acerifolia is a drought-tolerant riparian species that prefers warm conditions located in the hot/warm-dry river valleys (Jin, 1998; Li & Esser, 2008), it therefore seems plausible that this species expanded its range during the gradually warming period from 138.12 Kya to 95.79 Kya. Moreover, it is worth mentioning that the expansion of the YMRN group probably occurred in the more southerly regions, which was supported by the ENM analysis, showing significant larger distribution ranges in more southerly Red-Nanpan drainage basins (Fig. 6A, B). This probably just explains the derived (tip) haplotypes (H10 and H14; Fig. 3A) only existing in the two populations (populations 28 and 33; Fig. 2A; Table 1) of the more southerly Red-Nanpan Rivers within the YMRN group. Finally, the expansion of the YMRN populations might also explain the inconsistency of scenarios under demographic expansion and spatial expansion models of the total populations. Spatial expansion is quite different from demographic expansion and is more responsive to rapid climatic changes (Excoffier, Foll & Petit, 2009; Oberle & Dirzo, 2011). In the presence of the expansion of the YMRN populations due to the interglacial-glacial climatic change, therefore, the spatial stability of the total populations is more likely to be affected and showed non-significant SSD and HRag values. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher's web-site: Table S1. What the 19 bioclimatic variables represent and the correlation analysis among them. Table S2. Information of the collection records of Excoecaria acerifolia obtained from the three main herbaria in China (IBSC, KUN and PE). Table S3. Polymorphic sites in the three aligned plastid DNA sequences and defining features (substitutions and indels) of the 15 haplotypes (H1–15). Table S4. Polymorphic sites in the aligned nuclear DNA sequences, and defining features (substitutions) of the 18 ribotypes (R1–18). Figure S1. Distributions of the FCT values for indicated numbers of groups (K) of Excoecaria acerifolia populations based on A, plastid DNA and B, nuclear DNA sequences. Figure S2. Ecological niche modelling using bioclimatic factors for other periods of the YMRN, JMD and JY populations. Red colour indicates areas with suitability (suitable cell; YMRN, A: 12 879; JMD, B: 24 031; JY, C: 5796). ACKNOWLEDGEMENTS The authors thank Qiangbang Gong, Haoran Wang and Dr Xinxin Zhu for help with the field survey and leaf collection. We also thank Miss Minshu Song, Yuran Dong and Dr Deli Peng for assistance with the data analysis. In addition, we are grateful to Sees-editing Ltd for English editing. This study was funded by the Regional Science Foundation Program of the National Natural Science Foundation of China (31760053), the National Key R&D Program of China (2017YFC0505200 to H.S.), the Major Program of the National Natural Science Foundation of China (31590823 to H.S.), the NSFC joint fund of Yunnan Province (U1802232 to H.S.), the Natural Science Foundation of Guizhou Department of Education (QianJiaoHe KY Zi [2018]089) and the Foundation of Guizhou Administration of Traditional Chinese Medicine (QZYY–2018–097). REFERENCES An Z , Liu T , Lu Y , Porter SC , Kukla G , Wu X , Hua Y . 1990 . The long-term paleomonsoon variation recorded by the loess-paleosol sequence in Central China . Quaternary International 7 : 91 – 95 . OpenURL Placeholder Text WorldCat Andrew RL , Wallis IR , Harwood CE , Foley WJ . 2010 . Genetic and environmental contributions to variation and population divergence in a broad-spectrum foliar defence of Eucalyptus tricarpa . Annals of Botany 105 : 707 – 717 . Google Scholar Crossref Search ADS PubMed WorldCat Antonelli A , Ariza M , Albert J , Andermann T , Azevedo J , Bacon C , Faurby S , Guedes T , Hoorn C , Lohmann LG , Matos-Maraví P , Ritter CD , Sanmartín I , Silvestro D , Tejedor M , Steege Ht , Tuomisto H , Werneck FP , Zizka A , Edwards SV . 2018a . Conceptual and empirical advances in Neotropical biodiversity research . PeerJ 6 : e5644 . Google Scholar Crossref Search ADS WorldCat Antonelli A , Zizka A , Carvalho FA , Scharn R , Bacon CD , Silvestro D , Condamine FL . 2018b . Amazonia is the primary source of Neotropical biodiversity . Proceedings of the National Academy of Sciences of the United States of America 115 : 6034 – 6039 . Google Scholar Crossref Search ADS WorldCat Bandelt H-J , Forster P , Röhl A . 1999 . Median-joining networks for inferring intraspecific phylogenies . Molecular Biology and Evolution 16 : 37 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat Brookfield ME . 1998 . The evolution of the great river systems of southern Asia during the Cenozoic India-Asia collision: rivers draining southwards . Geomorphology 22 : 285 – 312 . Google Scholar Crossref Search ADS WorldCat Caicedo AL , Schaal BA . 2004 . Population structure and phylogeography of Solanum pimpinellifolium inferred from a nuclear gene . Molecular Ecology 13 : 1871 – 1882 . Google Scholar Crossref Search ADS PubMed WorldCat Chen S , Wu G , Zhang D , Gao Q , Duan Y , Zhang F , Chen S . 2008 . Potential refugium on the Qinghai-Tibet Plateau revealed by the chloroplast DNA phylogeography of the alpine species Metagentiana striata (Gentianaceae) . Botanical Journal of the Linnean Society 157 : 125 – 140 . Google Scholar Crossref Search ADS WorldCat Chiang TY , Schaal BA . 2006 . Phylogeography of plants in Taiwan and the Ryukyu Archipelago . Taxon 55 : 31 – 41 . Google Scholar Crossref Search ADS WorldCat Christe C , Caetano S , Aeschimann D , Kropf M , Diadema K , Naciri Y . 2014 . The intraspecific genetic variability of siliceous and calcareous Gentiana species is shaped by contrasting demographic and re-colonization processes . Molecular Phylogenetics and Evolution 70 : 323 – 336 . Google Scholar Crossref Search ADS PubMed WorldCat Clark AG . 1990 . Inference of haplotypes from PCR-amplified samples of diploid populations . Molecular Biology and Evolution 7 : 111 – 122 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Clark MK , Schoenbohm LM , Royden LH , Whipple KX , Burchfiel BC , Zhang X , Tang W , Wang E , Chen L . 2004 . Surface uplift, tectonics, and erosion of eastern Tibet from large-scale drainage patterns . Tectonics 23 : TC1006 . Google Scholar Crossref Search ADS WorldCat Clift PD , Blusztajn J , Nguyen AD . 2006 . Large-scale drainage capture and surface uplift in eastern Tibet-SW China before 24 Ma inferred from sediments of the Hanoi Basin, Vietnam . Geophysical Research Letters 33 : 1 – 4 . Google Scholar Crossref Search ADS WorldCat Cun Y-Z , Wang X-Q . 2010 . Plant recolonization in the Himalaya from the southeastern Qinghai-Tibetan Plateau: geographical isolation contributed to high population differentiation . Molecular Phylogenetics and Evolution 56 : 972 – 982 . Google Scholar Crossref Search ADS PubMed WorldCat Das AB , Jena S , Pradhan C , Chand PK . 2011 . Genetic variability among male populations of a minor mangrove Excoecaria agallocha L. as evident by chromosome morphology and DNA markers . Nucleus 54 : 39 – 47 . Google Scholar Crossref Search ADS WorldCat Doyle JJ , Doyle JL . 1990 . Isolation of plant DNA from fresh tissue . Focus 12 : 13 – 15 . OpenURL Placeholder Text WorldCat Drummond AJ , Rambaut A . 2007 . BEAST: Bayesian evolutionary analysis by sampling trees . BMC Evolutionary Biology 7 : 214 . Google Scholar Crossref Search ADS PubMed WorldCat Dupanloup I , Schneider S , Excoffier L . 2002 . A simulated annealing approach to define the genetic structure of populations . Molecular Ecology 11 : 2571 – 2581 . Google Scholar Crossref Search ADS PubMed WorldCat Excoffier L , Foll M , Petit RJ . 2009 . Genetic consequences of range expansions . Annual Review of Ecology Evolution & Systematics 40 : 481 – 501 . Google Scholar Crossref Search ADS WorldCat Excoffier L , Lischer HE . 2010 . Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows . Molecular Ecology Resources 10 : 564 – 567 . Google Scholar Crossref Search ADS PubMed WorldCat Fleming TH , Kress WJ . 2011 . A brief history of fruits and frugivores . Acta Oecologica 37 : 0 – 530 . Google Scholar Crossref Search ADS WorldCat Fu YX . 1997 . Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection . Genetics 147 : 915 – 925 . Google Scholar PubMed OpenURL Placeholder Text WorldCat García C , Jordano P , Godoy JA . 2010 . Contemporary pollen and seed dispersal in a Prunus mahaleb population: patterns in distance and direction . Molecular Ecology 16 : 1947 – 1955 . Google Scholar Crossref Search ADS WorldCat Hall TA . 1999 . BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT . Nucleic Acids Symposium Series 41 : 95 – 98 . OpenURL Placeholder Text WorldCat Hamrick JL , Godt MJW , Sherman-Broyles SL . 1992 . Factors influencing levels of genetic diversity in woody plant species . New Forests 6 : 95 – 124 . Google Scholar Crossref Search ADS WorldCat Harpending HC . 1994 . Signature of ancient population growth in a low-resolution mitochondrial DNA mismatch distribution . Human Biology 66 : 591 – 600 . Google Scholar PubMed OpenURL Placeholder Text WorldCat He D , Chen Y . 2006 . Biogeography and molecular phylogeny of the genus Schizothorax (Teleostei: Cyprinidae) in China inferred from cytochrome b sequences . Journal of Biogeography 33 : 1448 – 1460 . Google Scholar Crossref Search ADS WorldCat He D , Chen Y , Chen Y , Chen Z . 2004 . Molecular phylogeny of the specialized schizothoracine fishes (Teleostei: Cyprinidae), with their implications for the uplift of the Qinghai-Tibetan Plateau . Chinese Science Bulletin 49 : 39 – 48 . Google Scholar Crossref Search ADS WorldCat Hewitt G . 2004 . Genetic consequences of climatic oscillations in the Quaternary . Philosophical Transactions of the Royal Society of London B: Biological Sciences 359 : 183 – 195 . Google Scholar Crossref Search ADS WorldCat Hijmans RJ , Cameron SE , Parra JL , Jones PG , Jarvis A . 2005 . Very high resolution interpolated climate surfaces for global land areas . International Journal of Climatology 25 : 1965 – 1978 . Google Scholar Crossref Search ADS WorldCat Jiang F , Wu X , Xiao H , Zhao Z . 1999 . On the age of the Xigeda formation in Luding, Sichuan, and its Neotectonic significance . Acta Geologica Sinica 73 : 1 – 6 Google Scholar Crossref Search ADS WorldCat Jin Z . 1998 . Study on the floristic elements of seed plant in the dry-warm valleys of Yunnan and Sichuan . Guihaia 18 : 313 – 321 . OpenURL Placeholder Text WorldCat Kay KM , Whittall JB , Hodges SA . 2006 . A survey of nuclear ribosomal internal transcribed spacer substitution rates across angiosperms: an approximate molecular clock with life history effects . BMC Evolutionary Biology 6 : 36 . Google Scholar Crossref Search ADS PubMed WorldCat Kong P , Granger DE , Wu FY , Caffee MW , Wang YJ , Zhao XT , Zheng Y . 2009 . Cosmogenic nuclide burial ages and provenance of the Xigeda paleo-lake: implications for evolution of the Middle Yangtze River . Earth & Planetary Science Letters 278 : 131 – 141 . Google Scholar Crossref Search ADS WorldCat Krauss SL , Phillips RD , Karron JD , Johnson SD , Roberts DG , Hopper SD . 2017 . Novel consequences of bird pollination for plant mating . Trends in Plant Science 22 : 395 – 410 . Google Scholar Crossref Search ADS PubMed WorldCat Li B , Esser H-J . 2008 . Excoecaria L. In: Wu ZY , Raven PH , eds. Flora of China . Beijing/Saint Louis : Science Press/Missouri Botanical Garden Press , 11 : 280 – 281 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Li J , Fang X . 1999 . Uplift of the Tibetan Plateau and environmental changes . Chinese Science Bulletin 44 : 2117 – 2124 . Google Scholar Crossref Search ADS WorldCat Li J , Xie S , Kuang M . 2001 . Geomorphic evolution of the Yangtze Gorges and the time of their formation . Geomorphology 41 : 125 – 135 . Google Scholar Crossref Search ADS WorldCat Li JJ , Wen SX , Zhang QS , Wang FB , Zheng BX , Li BY . 1979 . A discussion on the period, amplitude and type of the uplift of the Qinghai-Xizang Plateau . Scientia Sinica 22 : 1314 – 1328 . OpenURL Placeholder Text WorldCat Li Y , Zhai S-N , Qiu Y-X , Guo Y-P , Ge X-J , Comes HP . 2011 . Glacial survival east and west of the ‘Mekong-Salween Divide’ in the Himalaya-Hengduan Mountains region as revealed by AFLPs and cp DNA sequence variation in Sinopodophyllum hexandrum (Berberidaceae) . Molecular Phylogenetics and Evolution 59 : 412 – 424 . Google Scholar Crossref Search ADS PubMed WorldCat Librado P , Rozas J . 2009 . DnaSP v5: a software for comprehensive analysis of DNA polymorphism data . Bioinformatics 25 : 1451 – 1452 . Google Scholar Crossref Search ADS PubMed WorldCat Lin BD , Zhou B , Dong L , Wu Y , Yue JM . 2016 . Formosins A-F: diterpenoids with anti-microbial activities from Excoecaria formosana . Natural Products and Bioprospecting 6 : 57 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat Liu J , Möller M , Provan J , Gao LM , Poudel RC , Li DZ . 2013 . Geological and ecological factors drive cryptic speciation of yews in a biodiversity hotspot . New Phytologist 199 : 1093 – 1108 . Google Scholar Crossref Search ADS PubMed WorldCat Luo D , Yue JP , Sun WG , Xu B , Li ZM , Comes HP , Sun H . 2016 . Evolutionary history of the subnival flora of the Himalaya-Hengduan Mountains: first insights from comparative phylogeography of four perennial herbs . Journal of Biogeography 43 : 31 – 43 . Google Scholar Crossref Search ADS WorldCat Ming QZ , Shi ZT , Dong M . 2007 . The inquiry on genesis and formation times of the first bend of Yangtze River . Progress in Geography 26 : 119 – 126 . OpenURL Placeholder Text WorldCat Ming QZ , Shi ZT , Zhang HC . 2006 . The evolution of the landform and environment in the region of the three parallel rivers . Tropical Geography 26 : 119 – 122 . OpenURL Placeholder Text WorldCat Mondal S , Ghosh D , Ramakrishna K . 2016 . A complete profile on blind-your-eye mangrove Excoecaria agallocha L. (Euphorbiaceae): ethnobotany, phytochemistry, and pharmacological aspects . Pharmacognosy Reviews 10 : 123 – 138 . Google Scholar Crossref Search ADS PubMed WorldCat Myers N , Mittermeier RA , Mittermeier CG , Da Fonseca GA , Kent J . 2000 . Biodiversity hotspots for conservation priorities . Nature 403 : 853 – 858 . Google Scholar Crossref Search ADS PubMed WorldCat Nei M . 1975 . Molecular population genetics and evolution . New York : North-Holland Publishing Company , 208 – 209 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Neigel JE , Avise JC . 1986 . Phylogenetic relationships of mitochondrial DNA under various demographic models of speciation . Evolutionary Processes and Theory 515 – 534 . OpenURL Placeholder Text WorldCat Oberle B , Dirzo R . 2011 . Responses to historical climate change identify contemporary threats to diversity in Dodecatheon . Proceedings of the National Academy of Sciences of the United States of America 108 : 5655 – 5660 . Google Scholar Crossref Search ADS PubMed WorldCat Olsen KM , Schaal BA . 1999 . Evidence on the origin of cassava: phylogeography of Manihot esculenta . Proceedings of the National Academy of Sciences of the United States of America 96 : 5586 – 5591 . Google Scholar Crossref Search ADS PubMed WorldCat Owen, Lewis A. 2008 . Latest Pleistocene and Holocene glacier fluctuations in the Himalaya and Tibet . Quaternary Science Reviews 28 : 2150 – 2164 . OpenURL Placeholder Text WorldCat Peng Z , Ho SY , Zhang Y , He S . 2006 . Uplift of the Tibetan plateau: evidence from divergence times of glyptosternoid catfishes . Molecular Phylogenetics & Evolution 39 : 568 – 572 . Google Scholar Crossref Search ADS WorldCat Petit RJ, Duminil J, Fineschi S, Hampe A, Salvini D, Vendramin GG. 2005 . Invited review: comparative organization of chloroplast, mitochondrial and nuclear diversity in plant populations . Molecular Ecology 14 : 689 – 701 . PubMed OpenURL Placeholder Text WorldCat Phillips SJ , Anderson RP , Schapire RE . 2006 . Maximum entropy modeling of species geographic distributions . Ecological Modelling 190 : 231 – 259 . Google Scholar Crossref Search ADS WorldCat Pons O , Petit R . 1996 . Measuring and testing genetic differentiation with ordered versus unordered alleles . Genetics 144 : 1237 – 1245 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Posada D , Crandall KA . 1998 . Modeltest: testing the model of DNA substitution . Bioinformatics 14 : 817 – 818 . Google Scholar Crossref Search ADS PubMed WorldCat Posada D , Crandall KA . 2001 . Intraspecific gene genealogies: trees grafting into networks . Trends in Ecology & Evolution 16 : 37 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat Pybus OG , Rambaut A , Harvey PH . 2000 . An integrated framework for the inference of viral population history from reconstructed genealogies . Genetics 155 : 1429 – 1437 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Qi XS , Yuan N , Comes HP , Sakaguchi S , Qiu YX . 2014 . A strong ‘filter’ effect of the East China Sea land bridge for East Asia’s temperate plant species: inferences from molecular phylogeography and ecological niche modelling of Platycrater arguta (Hydrangeaceae) . BMC Evolutionary Biology 14 : 41 . Google Scholar Crossref Search ADS PubMed WorldCat Qiu Y-X , Fu C-X , Comes HP . 2011 . Plant molecular phylogeography in China and adjacent regions: tracing the genetic imprints of Quaternary climate and environmental change in the world’s most diverse temperate flora . Molecular Phylogenetics and Evolution 59 : 225 – 244 . Google Scholar Crossref Search ADS PubMed WorldCat Rawat K , Parasad K , Bisht G . 2017 . Chemical composition of essential oil of Excoecaria acerifolia collected from Kumaun Himalayas . World Journal of Pharmaceutical Research 6 : 1261 – 1266 . Google Scholar Crossref Search ADS WorldCat Richardson JE , Pennington RT , Pennington TD , Hollingsworth PM . 2001 . Rapid diversification of a species-rich genus of Neotropical rain forest trees . Science 293 : 2242 – 2245 . Google Scholar Crossref Search ADS PubMed WorldCat Rogers AR , Harpending H . 1992 . Population growth makes waves in the distribution of pairwise genetic differences . Molecular Biology and Evolution 9 : 552 – 569 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Ronquist F , Huelsenbeck JP . 2003 . MrBayes 3: Bayesian phylogenetic inference under mixed models . Bioinformatics 19 : 1572 – 1574 . Google Scholar Crossref Search ADS PubMed WorldCat Royden LH , Burchfiel BC , van der Hilst HD . 2008 . The geological evolution of the Tibetan Plateau . Science 321 : 1054 – 1058 . Google Scholar Crossref Search ADS PubMed WorldCat Ruddiman WF , Kutzbach JE . 1989 . Forcing of late Cenozoic Northern Hemisphere climate by plateau uplift in southern Asia and the American West . Journal of Geophysical Research 94 : 18409 – 18427 . Google Scholar Crossref Search ADS WorldCat Rull V . 2008 . Speciation timing and Neotropical biodiversity: the Tertiary-Quaternary debate in the light of molecular phylogenetic evidence . Molecular Ecology 17 : 2722 – 2729 . Google Scholar Crossref Search ADS PubMed WorldCat Rull V . 2011 . Neotropical biodiversity: timing and potential drivers . Trends in Ecology & Evolution 26 : 508 – 513 . Google Scholar Crossref Search ADS PubMed WorldCat Rull V . 2018 . Neotropical diversification: historical overview and conceptual insights . PeerJ Preprints 6 : e27294v27291 . OpenURL Placeholder Text WorldCat Schäfer JM , Tschudi S , Zhao Z , Wu X , Ivy-Ochs S , Wieler R , Baur H , Kubik PW , Schlüchter C . 2002 . The limited influence of glaciations in Tibet on global climate over the past 170 000 yr . Earth & Planetary Science Letters 194 : 287 – 297 . Google Scholar Crossref Search ADS WorldCat Shaw J , Lickey EB , Beck JT , Farmer SB , Liu W , Miller J , Siripun KC , Winder CT , Schilling EE , Small RL . 2005 . The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis . American Journal of Botany 92 : 142 – 166 . Google Scholar Crossref Search ADS PubMed WorldCat Shaw J , Lickey EB , Schilling EE , Small RL . 2007 . Comparison of whole chloroplast genome sequences to choose noncoding regions for phylogenetic studies in angiosperms: the tortoise and the hare III . American Journal of Botany 94 : 275 – 288 . Google Scholar Crossref Search ADS PubMed WorldCat Sophia A , Compton SG , Butlin RK , Gilmartina PM . 2009 . Wind-borne insects mediate directional pollen transfer between desert fig trees 160 kilometers apart . Proceedings of the National Academy of Sciences of the United States of America 106 : 20342 – 20347 . Google Scholar Crossref Search ADS PubMed WorldCat Sun H , Zhang J , Deng T , Boufford DE . 2017 . Origins and evolution of plant diversity in the Hengduan Mountains, China . Plant Diversity 39 : 161 – 166 . Google Scholar Crossref Search ADS PubMed WorldCat Taberlet P , Gielly L , Pautou G , Bouvet J . 1991 . Universal primers for amplification of three non-coding regions of chloroplast DNA . Plant Molecular Biology 17 : 1105 – 1109 . Google Scholar Crossref Search ADS PubMed WorldCat Tajima F . 1989 . Statistical method for testing the neutral mutation hypothesis by DNA polymorphism . Genetics 123 : 585 – 595 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Tajima F . 1996 . The amount of DNA polymorphism maintained in a finite population when the neutral mutation rate varies among sites . Genetics 143 : 1457 – 1465 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Tang Y , Xie J , Sun H . 2004 . Revisiting sustainable development of dry valleys in Hengduan Mountains region . Journal of Mountain Science 1 : 38 – 45 . Google Scholar Crossref Search ADS WorldCat Turchetto-Zolet AC , Pinheiro F , Salgueiro F , Palma-Silva C . 2013 . Phylogeographical patterns shed light on evolutionary process in South America . Molecular Ecology 22 : 1193 – 1213 . Google Scholar Crossref Search ADS PubMed WorldCat Wang F-Y , Gong X , Hu C-M , Hao G . 2008 . Phylogeography of an alpine species Primula secundiflora inferred from the chloroplast DNA sequence variation . Journal of Systematics and Evolution 46 : 13 – 22 . OpenURL Placeholder Text WorldCat Wang IJ , Glor RE , Losos JB . 2013 . Quantifying the roles of ecology and geography in spatial genetic divergence . Ecology Letters 16 : 175 – 182 . Google Scholar Crossref Search ADS PubMed WorldCat Wang P , Li J , Wang J , Liu C , Han F , Gao L . 2011 . Quartz Ti-center in ESR dating of Xigeda formation in Sichuan and contrast with magnetic stratigraphic profiles . Nuclear Techniques 34 : 110 – 115 . OpenURL Placeholder Text WorldCat Wang YH , Jiang WM , Comes HP , Hu FS , Qiu YX , Fu CX . 2015a . Molecular phylogeography and ecological niche modelling of a widespread herbaceous climber, Tetrastigma hemsleyanum (Vitaceae): insights into Plio–Pleistocene range dynamics of evergreen forest in subtropical China . New Phytologist 206 : 852 – 867 . Google Scholar Crossref Search ADS WorldCat Wang ZW , Chen ST , Nie ZL , Zhang JW , Zhou Z , Deng T , Sun H . 2015b . Climatic factors drive population divergence and demography: insights based on the phylogeography of a riparian plant species endemic to the Hengduan Mountains and adjacent regions . PLoS One 10 : e0145014 . Google Scholar Crossref Search ADS WorldCat Webster GL . 1975 . Classification of the Euphorbiaceae . Annals of the Missouri Botanical Garden 81 : 3 – 32 . Google Scholar Crossref Search ADS WorldCat Wolfe KH , Li W-H , Sharp PM . 1987 . Rates of nucleotide substitution vary greatly among plant mitochondrial, chloroplast, and nuclear DNAs . Proceedings of the National Academy of Sciences of the United States of America 84 : 9054 – 9058 . Google Scholar Crossref Search ADS PubMed WorldCat Wotton DM , McAlpine KG . 2015 . Seed dispersal of fleshy-fruited environmental weeds in New Zealand . New Zealand Journal of Ecology 39 : 155 – 169 . OpenURL Placeholder Text WorldCat Xia K , Su T , Liu Y-SC , Xing Y-W , Jacques FM , Zhou Z-K . 2009 . Quantitative climate reconstructions of the late Miocene Xiaolongtan megaflora from Yunnan, southwest China . Palaeogeography, Palaeoclimatology, Palaeoecology 276 : 80 – 86 . Google Scholar Crossref Search ADS WorldCat Yan H-F , Zhang C-Y , Wang F-Y , Hu C-M , Ge X-J , Hao G . 2012 . Population expanding with the phalanx model and lineages split by environmental heterogeneity: a case study of Primula obconica in subtropical China . PLoS One 7 : e41315 . Google Scholar Crossref Search ADS PubMed WorldCat Yang FS , Li YF , Ding X , Wang XQ . 2008 . Extensive population expansion of Pedicularis longiflora (Orobanchaceae) on the Qinghai-Tibetan Plateau and its correlation with the Quaternary climate change . Molecular Ecology 17 : 5135 – 5145 . Google Scholar Crossref Search ADS PubMed WorldCat Yang ZY , Yi TS , Zeng LQ , Gong X . 2014 . The population genetic structure and diversification of Aristolochia delavayi (Aristolochiaceae), an endangered species of the dry hot valleys of the Jinsha River, southwestern China . Botany 92 : 579 – 587 . Google Scholar Crossref Search ADS WorldCat Yeragi LA , Mendhulkar VD . 2017 . Seasonal variations in antibacterial and antifungal activity of Excoecaria agallocha . Asian Journal of Microbiology Biotechnology & Environmental Sciences 19 : 105 – 109 . OpenURL Placeholder Text WorldCat Yue LL , Chen G , Sun WB , Sun H . 2012 . Phylogeography of Buddleja crispa (Buddlejaceae) and its correlation with drainage system evolution in southwestern China . American Journal of Botany 99 : 1726 – 1735 . Google Scholar Crossref Search ADS PubMed WorldCat Yue XK , Yue JP , Yang LE , Li ZM , Sun H . 2011 . Systematics of the genus Salweenia (Leguminosae) from southwest China with discovery of a second species . Taxon 60 : 1366 – 1374 . Google Scholar Crossref Search ADS WorldCat Zhang TC , Comes HP , Sun H . 2011 . Chloroplast phylogeography of Terminalia franchetii (Combretaceae) from the eastern Sino-Himalayan region and its correlation with historical river capture events . Molecular Phylogenetics and Evolution 60 : 1 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat Zhang TC , Sun H . 2011 . Phylogeographic structure of Terminalia franchetii (Combretaceae) in southwest China and its implications for drainage geological history . Journal of Plant Research 124 : 63 – 73 . Google Scholar Crossref Search ADS PubMed WorldCat Zheng B , Xu Q , Shen Y . 2002 . The relationship between climate change and Quaternary glacial cycles on the Qinghai-Tibetan Plateau: review and speculation . Quaternary International 97 : 93 – 101 . Google Scholar Crossref Search ADS WorldCat Zhou S , Wang X , Wang J , Xu L . 2006 . A preliminary study on timing of the oldest Pleistocene glaciation in Qinghai-Tibetan Plateau . Quaternary International 154–155 : 44 – 51 . Google Scholar Crossref Search ADS WorldCat © 2019 The Linnean Society of London, Botanical Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Doostmohammadi,, Moslem;Malekmohammadi,, Maryam;Djamali,, Morteza;Akhani,, Hossein

doi: 10.1093/botlinnean/boz079pmid: N/A

Abstract Pteropyrum is a small genus of Polygonaceae with four species from the arid regions of Iran and adjacent countries. Pteropyrum spp. are not precisely delimitated and are difficult to identify because of their high plasticity in morphological characters. Pteropyrum (C3) has a close affinity to Calligonum (C4) and is therefore a suitable case for C4 evolutionary studies. We investigated the morphology and micromorphology (including pollen morphology) of Pteropyrum and elucidated the phylogenetic relationships with Atraphaxis and Calligonum using nuclear ITS sequences. Characteristics of the photosynthetic tissues such as volume and number of layers of primary carbon assimilation tissues (PCA) and photosynthetic carbon reduction tissue (PCR) were studied. In addition, the leaf and cotyledon anatomical characters of Pteropyrum (C3), Atraphaxis (C3) and Calligonum (C4), and their δ 13C values were compared to look for evolutionary changes in assimilating organs. The molecular phylogenetic tree identifies two strongly supported clades in Pteropyrum and its close relationship with Calligonum, confirming previous studies. Some morphologically similar species belong to different clades, which is probably due to convergent evolution and homoplasy. Leaf anatomical studies show that Atraphaxis has a multilayered mesophyll tissue, whereas Calligonum has one-layered mesophyll cells. The volume and layer number of mesophyll tissue cells decreases, whereas water storage tissue area significantly increases from Atraphaxis to Pteropyrum and Calligonum. This phenomenon confirms previous studies in other lineages with C4 salsoloid anatomy that have evolved through increasing of water storage tissue and succulence of assimilating organs. In the taxonomic part of the paper, a key to identification of accepted taxa of Pteropyrum, description of species and distribution maps are presented based on numerous herbarium specimens and our own rich collections from the field. Four new species are described based on a combination of morphology of seedlings and mature plants, pollen morphology and molecular data. A subspecific classification is suggested to show morpho-geographical variation of Pteropyrum aucheri s.l. Atraphaxis, C4 photosynthesis, Caryophyllales, cryptic speciation, desert flora, Irano–Turanian flora, molecular phylogeny, pollen morphology, proto-Kranz anatomy INTRODUCTION Polygonaceae belong to the non-core Caryophyllales and have a close affinity with Plumbaginaceae. They are well circumscribed with synapomorphies including an ochrea sheath, swollen nodes, five or six petaloid tepals, a single orthotropous ovule and usually trigonal achenes (Cuénoud et al., 2002; Hernández-Ledesma et al., 2015). Recently, Sanchez et al. (2011) revised the tribal classification of subfamily Polygonoideae using molecular data and proposed five newly circumscribed tribes of which Calligoneae consist of two poorly studied genera, Calligonum L. and Pteropyrum Jaub. & Spach. Close relationships between these two genera have been suggested in previous molecular (Sanchez, Schuster & Kron, 2009; Tavakkoli, Kazempour & Maassoumi, 2010; Schuster et al., 2015) and non-molecular (Haraldson, 1978; Ronse-Decraene & Akeroyd, 1988; Hong, 1995; Hong, Ronse-Decraene & Smets, 1998; Tavakkoli, Kazempour & Maassoumi, 2008) studies. Calligonum is the only C4 genus in Polygonaceae (Sage, Christin & Edwards, 2011), and species are leafless shrubby plants with assimilating stems with salsoloid anatomy (Muhaidat, Sage & Dengler, 2007). Calligonum is a species-rich genus with c. 80 species, widely distributed in sand dunes and desert areas of south-western and Central Asia, south-eastern Europe and North Africa. In contrast, Pteropyrum is a small genus with a limited distribution in Iran and adjacent countries (Turkmenistan, Afghanistan, Pakistan, Oman, UAE, and Iraq) (Brandbyge, 1993). Atraphaxis L. is a genus of C3 shrubs; it was classified in tribe Polygoneae by Sanchez et al. (2011) and occurs widely in Irano–Turanian montane steppes. All these genera appear to have formed an important floristic component of the Irano–Turanian shrub steppes during the Quaternary glacial intervals (Djamali et al., 2008), the long persistence of which in these landscapes may have had a significant role in their evolutionary history. Pteropyrum spp., with their small leaves (microphylls), are successful xerophytic shrubs usually growing near seasonal water runnels or dry riverbeds or on gypsum hills in desert areas and, rarely, on rocky slopes. Pteropyrum was first described with three species categorized into two sections according to fruit characters: Pteropyrum section Streptocarya Jaub. & Spach (P. aucheri Jaub. & Spach, P. olivieri Jaub. & Spach) and Pteropyrum section Orthocarya Jaub. & Spach (P. scoparium Jaub. & Spach) (Jaubert & Spach, 1846). In most cases, subsequent additional described species (Boissier, 1853; Meisner, 1856; Gilli, 1963) were not significantly different from the original three species; therefore, Rechinger & Schiman-Czeika (1968) accepted the occurrence of only three species in the Flora Iranica area: P. aucheri, P. noeanum Boiss. and P. olivieri. In studies following Flora Iranica, P. naufelum Al-Khayat was introduced from Iraq (Al-Khayat, 1993) and Akhani (2004) reported this species in Iran and synonymized P. noeanum with P. aucheri. Species boundaries in Pteropyrum are not precise, and species growing in different habitats are not easily distinguishable based on their morphology. Leaf size and shape are important diagnostic characters for distinguishing Pteropyrum spp., but these traits are often affected by environmental conditions, and there are many apparently intermediate forms that complicate species boundaries. From an evolutionary point of view, it has been accepted that in some phylogenetic groups, C3 sister species of C4 plants exhibit some anatomical traits and subcellular events that are considered to be the initial phase of C4 evolution. Anatomical traits such as enlarged bundle sheath (BS) cells and centripetal localization of mitochondria in BS cells form the proto-Kranz anatomy (Muhaidat et al., 2011; Sage, Busch & Johnson, 2013; Voznesenskaya et al., 2013; Sage, Khoshravesh & Sage, 2014). Proto-Kranz anatomy has some adaptive benefits and facilitates the glycine shuttle development towards C4 evolution. Since Pteropyrum is the sister group of Calligonum distributed in hot/arid regions, it is not unexpected that some Pteropyrum spp. may show a proto-Kranz condition or even have C2 photosynthesis. The aims of this paper are: (1) to revise the taxonomy of Pteropyrum and clarify the species circumscription using morphological, pollen morphological, anatomical and molecular traits; (2) to reconstruct phylogenetic relationships in Pteropyrum spp. and the sister genera using nuclear ITS (internal transcribed spacer) sequence data and testing the accuracy of traditional classifications and (3) to compare the anatomy of Pteropyrum with Atraphaxis (C3) and Calligonum (C4) to test our hypothesis that Pteropyrum as a sympatric desert plant with Calligonum undergoes anatomical changes towards C4 photosynthesis that finally evolved in Calligonum. MATERIAL AND METHODS Taxonomy, morphology and growth condition Many natural populations of Pteropyrum have been studied in the field in different parts of its range in Iran and Oman by MD and HA. Herbarium specimens including type specimens (or photographs of these) have been studied in herbaria in Iran and Europe including B, G, P, IRAN, TARI, TUH, MIR, W, E and LE (abbreviations after Thiers, 2018+) and the private herbarium of the senior author (Herb. H. Akhani). The holotype specimens of the newly described taxa are preserved in IRAN. Seeds collected during field studies have been cultivated in the laboratory under the same condition. They were grown in a mix of one part sand, one part clay and one part commercial peat. Seedling morphological traits were studied and measured under an Olympus BZ 21 stereomicroscope and photographed after emergence of the first true leaves using a Sony DSC-HX5Vi digital camera. Studied morphological characters of seedlings include hypocotyl height, cotyledon length, cotyledon width and hypocotyl indumentum. Pollen morphology Pollen grains of 23 specimens of Pteropyrum were sampled from dried herbarium specimens and examined using a light microscope (LM) and scanning electron microscope (SEM). Pollen grains were first acetolysed following Erdtman (1960) and Dehghani & Akhani (2009). For LM studies, the pollen grains were mounted in glycerine jelly, and then observed by a Nikon Optiphot-2 light microscope and photographed with a Motic Image Plus 2.0 digital camera. Polar axis, equatorial axis, length of colpi, maximum distance between colpi in mesocolpium, apocolpium diameter and the thickness of exine, sexine and nexine were measured using calibrated Motic Image Plus 2.0 software. In general, the terminology follows Hong (1995) and Mondal (1997), taking into account the standardized pollen and spore terminology proposed by Punt et al. (2007). For SEM studies, acetolysed pollen grains were gold coated and examined using a Zeiss DSM 960 SEM microscope in the Electron Microscopy Laboratory, College of Science, University of Tehran, Iran. Anatomy of assimilating organs Anatomical studies were conducted on cotyledons, leaf blades and young stem cross sections. Fresh leaf sections from cultivated plants or during field trips were fixed in FAA (one part formaldehyde, one part glacial acetic acid, 18 parts 70% ethanol), then rinsed in distilled water twice and dehydrated in a sequence of ethanol solutions. Acetone was used as the transition solvent to improve the infiltration. Samples were then embedded in pure epoxy resin and incubated for 24–48 h at 60 °C (Davies, 1999). Cross sections were made by Leica Ultracut UCT ultramicrotome (housed in College of Science, University of Tehran), stained with 0.5% (w/v) Toluidine blue in 1% (w/v) Na2CO3 and studied with an Olympus BX51 LM. Cross sections from green stems of Calligonum were made by hand using a razor blade. Sections were then stained in aqueous methylene blue/carmine and photographed before dehydration. Dehydrated sections were permanently mounted in Euparal. To compare the assimilating structures, the volume of primary carbon assimilation tissues (PCA or mesophyll cells), PCR tissues (PCR or BS), their ratios (PCA/PCR), percentages of epidermis and the percentages of aqueous tissues were measured for assimilating stems of seven Calligonum spp., mature leaves of six Calligonum spp., nine Pteropyrum taxa and four Atraphaxis spp. and the cotyledons of six Calligonum spp. and six Pteropyrum taxa. Our sampling strategy for comparative anatomy was based on multiple species of the compared genera rather than multiple sampling of the same species. The cell volume of each specimen was measured for one section using ImageJ software (Rasband, 1997–2018). DNA sequencing Plant material and taxon sampling All of the Pteropyrum samples used in this study were herbarium material mostly deposited in Herb. H. Akhani. The selected specimens cover almost all geographical distribution range and morphological diversity of Pteropyrum. Our ITS dataset includes 25 accessions representing all ten taxa of Pteropyrum. Four ITS sequences of Atraphaxis and Calligonum were taken from GenBank. Voucher information and GenBank accession numbers are provided in Appendix 1. In addition to ITS, we also sequenced two plastid markers (rpl32 and petD) for several accessions. Due to failure to complete our data set and low resolution of the obtained trees, however, we prefer not to publish incomplete data. DNA isolation, amplification and sequencing Total genomic DNA was extracted using the modified CTAB method of Doyle & Doyle (1987). To amplify the ITS region (ITS1 and ITS2 spacers plus the 5.8S gene), the universal primer combination ITS4/ITS5 (White et al., 1990) was used following the protocol of Tavakkoli et al. (2010). Reactions were carried out in the volume of 25 μl, containing 10.5 μl deionized water, 12.5 μl Taq DNA polymerase master mix Red [Ampliqon; Tris-HCl pH 8.5, (NH4)2SO4, 4 mM MgCl2, 0.2% Tween 20, 0.4 mM each of dNTP, 0.2 units/µl Ampliqon Taq DNA polymerase, inert red dye and stabilizer], 0.5 μl each primer and 1 μl template DNA. Cycle sequencing reactions (PCR) were performed on a Techne TC-3000 thermocycler (Cole-Parmer, UK). Amplification conditions were an initial denaturation (1 min 30 s at 96 °C), 34 cycles of denaturation (30 s at 95 °C), annealing (1 min at 48 °C), extension (1 min 30 s at 72 °C) and a final extension step (20 min at 72 °C). PCR products were purified using the Avegene PCR cleanup kit following the manufacturer’s protocols and sequenced via Macrogen Inc. (Seoul, South Korea). The same primers were used for amplification and sequencing. Sequence alignment and phylogenetic reconstruction PhyDE v.0.9971 (Müller, Müller & Quandt, 2010) was used to edit the electropherograms. Dataset was aligned automatically using Muscle (Edgar, 2004) and edited manually using PhyDE. The ITS matrix was analysed using Bayesian Inference (BI), Maximum Likelihood (ML) and Maximum Parsimony (MP). Atraphaxis spinosa Jaub. & Spach and A. suaedifolia L. were used to root trees based on the results of a previous phylogenetic study by Tavakkoli et al. (2010). The best model of molecular evolution was found using jModelTest v.2.1.10 (Darriba et al., 2012). The GTR+G model was found to fit best with the ITS region according to the Akaike information criterion (AIC). MrBayes v.3.2.6 (Ronquist et al., 2012) was used to conduct BI. Two parallel runs of four MCMC chains including three heated and one cold chain was run simultaneously for five million generations for each matrix. The first 25% of generations of each run were discarded and only the trees after the ‘burn-in’ sampled at a frequency of 1000. When the standard deviation of split frequencies was well below 0.01 the analyses were stopped. RAxML v.8.2.8 (Stamatakis, 2014) was used to conduct ML tree inference and bootstrapping (BS). For this purpose, the model was set to GTRGAMMA and bootstrap analyses were carried out with 1000 replicates. MP analyses were conducted using heuristic searches and tree lengths and homoplasy indices [consistency (CI), retention (RI), rescaled consistency (RC) and homoplasy indices (HI)] were calculated in PAUP v.4.a164 (Swofford, 1991). Bootstrapping (BS) was estimated in PAUP by conducting a heuristic search with 10 000 replicates using TBR branch swapping. Trees were visualized in TreeGraph2 v.2.10.1-641 beta (Stöver & Müller, 2010). All BI, ML and MP analyses were conducted on the Cipres portal (www.phylo.org). Carbon isotope values (δ 13C) The carbon isotope ratio was determined on green stems or leaves of herbarium materials or dried cotyledons of seedlings (Osmond et al., 1975). The samples were first ground to a fine powder and placed in a tin capsule, then combusted in a Heraeus CHN Eurovector elemental analyser (GSF National Research Center for Environment and Health, Munich, Germany for four samples and Iso-Analytical Limited, Cheshire, UK, for other samples). The resulting N2 and CO2 were separated by gas chromatography and admitted into the inlet of a Finnigan MAT Delta S (GSF) or Micromass Isoprime Isotope Ratio Mass Spectrometer (IRMS) (Iso-Analytical Limited) for determination of 13C/12C ratios (R). δ 13C values were calculated using the formula: d=1000×(Rsample/Rstandard−1). RESULTS Taxonomy and nomenclature An updated revision of Pteropyrum is provided in the taxonomic treatment at the end of this paper. Based on our multidisciplinary approach using morpho-molecular data and field observations seven species and ten taxa are recognized in Pteropyrum. Selected figures are provided for the habitat and close up of studied Pteropyrum spp. and two species each of Atraphaxis and Calligonum (Figs 1–17). Widespread species P. aucheri and P. olivieri are merged and the morpho-geographical varieties are classified as subspecies: P. aucheri subsp. aucheri, P. aucheri subsp. olivieri (Jaub. & Spach) Doostmohammadi & Akhani, P. aucheri subsp. ericoides (Boiss.) Doostmohammadi & Akhani and P. aucheri subsp. noeanum (Boiss.) Doostmohammadi & Akhani. Pteropyrum scoparium Jaub. & Spach is restricted to Oman and the UAE and is well-distinguishable by its unique upper part of achenes that are not twisted as in other species. The other five species (P. jakdanense Doostmohammadi, P. macrocarpum Doostmohammadi & Akhani, P. naufelum, P. gypsaceum Akhani & Doostmohammadi and P. zagricum Doostmohammadi & Akhani) are rare plants usually growing on gypsum hills, shale or rocky habitats in a geographical arc from eastern Iraq to south-eastern Iran in Sistan-va Baluchestan Province. Figures 1–8. Open in new tabDownload slide Natural habitat of several species of Pteropyrum and one species of Calligonum and Atraphaxis: Fig. 1. P. naufelum community growing on gypsum hills; Fig. 2. P. zagricum on limestone rocks; Fig. 3. P. jakdanense on shale hills; Fig. 4. P. aucheri subsp. ericoides along dry riverbank; Fig. 5. P. scoparium on gravelly dry rivulet beds; Fig. 6. P. macrocarpum in seasonal water runnel; Fig. 7. C. amoenum, sand dunes in Lut Desert; Fig. 8. A. spinosa on foothills and mountain slopes. Photographs: H. Akhani: 1, 5, 7, 8; M. Doostmohammadi 2, 3, 4, 6. Figures 1–8. Open in new tabDownload slide Natural habitat of several species of Pteropyrum and one species of Calligonum and Atraphaxis: Fig. 1. P. naufelum community growing on gypsum hills; Fig. 2. P. zagricum on limestone rocks; Fig. 3. P. jakdanense on shale hills; Fig. 4. P. aucheri subsp. ericoides along dry riverbank; Fig. 5. P. scoparium on gravelly dry rivulet beds; Fig. 6. P. macrocarpum in seasonal water runnel; Fig. 7. C. amoenum, sand dunes in Lut Desert; Fig. 8. A. spinosa on foothills and mountain slopes. Photographs: H. Akhani: 1, 5, 7, 8; M. Doostmohammadi 2, 3, 4, 6. Figures 9–17. Open in new tabDownload slide Close up of nine taxa of Pteropyrum in their natural habitats; Fig. 9. P. aucheri subsp. ericoides; Fig. 10. P. aucheri subsp. aucheri; Fig. 11. P. aucheri subsp. olivieri; Fig. 12. P. jakdanense; Fig. 13. P. macrocarpum; Fig. 14. P. scoparium; Fig. 15. P. zagricum; Fig. 16. P. naufelum; Fig. 17. P. gypsaceum. Photographs: H. Akhani: 9, 10, 11, 14, 16; M. Doostmohammadi 12, 13, 15, 17. Figures 9–17. Open in new tabDownload slide Close up of nine taxa of Pteropyrum in their natural habitats; Fig. 9. P. aucheri subsp. ericoides; Fig. 10. P. aucheri subsp. aucheri; Fig. 11. P. aucheri subsp. olivieri; Fig. 12. P. jakdanense; Fig. 13. P. macrocarpum; Fig. 14. P. scoparium; Fig. 15. P. zagricum; Fig. 16. P. naufelum; Fig. 17. P. gypsaceum. Photographs: H. Akhani: 9, 10, 11, 14, 16; M. Doostmohammadi 12, 13, 15, 17. Molecular phylogeny For phylogenetic studies, we generated 21 ITS sequences, and four ITS sequences for Atraphaxis and Calligonum were taken from GenBank (see Appendix 1 for GenBank accessions). The length of the multiple sequence alignment of the ITS region is 800 characters that is reduced to 693 characters after trimming the ends. The analysed ITS matrix includes 112 potentially parsimony-informative characters. Table 1 shows the sequence statistics of the different matrices. Table 1. Sequence and tree statistics for the ITS matrix Complete dataset  Number of taxa 27  Number of Pteropyrum taxa 23  Aligned length (bp) 800  Length range (bp) 364–703  Mean length (SD) 664.8 (44.4)  GC (%) 61.15 Dataset after trimming  Aligned length (bp) 693  Length range (bp) 394–620  Mean length (SD) 591.3 (67.59)  Transition-transversion ratio (Ti/Tv) 6.020  Divergence (%) 4.65  GC (%) 62.93  Percentage variable characters 18.61  Percentage potentially informative sites 16.01 Tree statistics  Constant characters 564  Variable uninformative characters 17  Potentially parsimony-informative characters 112  Number of shortest trees 1056  Tree length 148  Consistency index (CI) 0.97  Retention index (RI) 0.98  Rescaled consistency index (RC) 0.95  Homoplasy index (HI) 0.03 Complete dataset  Number of taxa 27  Number of Pteropyrum taxa 23  Aligned length (bp) 800  Length range (bp) 364–703  Mean length (SD) 664.8 (44.4)  GC (%) 61.15 Dataset after trimming  Aligned length (bp) 693  Length range (bp) 394–620  Mean length (SD) 591.3 (67.59)  Transition-transversion ratio (Ti/Tv) 6.020  Divergence (%) 4.65  GC (%) 62.93  Percentage variable characters 18.61  Percentage potentially informative sites 16.01 Tree statistics  Constant characters 564  Variable uninformative characters 17  Potentially parsimony-informative characters 112  Number of shortest trees 1056  Tree length 148  Consistency index (CI) 0.97  Retention index (RI) 0.98  Rescaled consistency index (RC) 0.95  Homoplasy index (HI) 0.03 Open in new tab Table 1. Sequence and tree statistics for the ITS matrix Complete dataset  Number of taxa 27  Number of Pteropyrum taxa 23  Aligned length (bp) 800  Length range (bp) 364–703  Mean length (SD) 664.8 (44.4)  GC (%) 61.15 Dataset after trimming  Aligned length (bp) 693  Length range (bp) 394–620  Mean length (SD) 591.3 (67.59)  Transition-transversion ratio (Ti/Tv) 6.020  Divergence (%) 4.65  GC (%) 62.93  Percentage variable characters 18.61  Percentage potentially informative sites 16.01 Tree statistics  Constant characters 564  Variable uninformative characters 17  Potentially parsimony-informative characters 112  Number of shortest trees 1056  Tree length 148  Consistency index (CI) 0.97  Retention index (RI) 0.98  Rescaled consistency index (RC) 0.95  Homoplasy index (HI) 0.03 Complete dataset  Number of taxa 27  Number of Pteropyrum taxa 23  Aligned length (bp) 800  Length range (bp) 364–703  Mean length (SD) 664.8 (44.4)  GC (%) 61.15 Dataset after trimming  Aligned length (bp) 693  Length range (bp) 394–620  Mean length (SD) 591.3 (67.59)  Transition-transversion ratio (Ti/Tv) 6.020  Divergence (%) 4.65  GC (%) 62.93  Percentage variable characters 18.61  Percentage potentially informative sites 16.01 Tree statistics  Constant characters 564  Variable uninformative characters 17  Potentially parsimony-informative characters 112  Number of shortest trees 1056  Tree length 148  Consistency index (CI) 0.97  Retention index (RI) 0.98  Rescaled consistency index (RC) 0.95  Homoplasy index (HI) 0.03 Open in new tab ITS phylogenetic trees The BI, ML and MP analyses of the ITS dataset produced congruent trees without any major difference in topology. Therefore, only the results from the Bayesian analyses are shown here, along with posterior probabilities as well as ML and MP bootstrap values. The phylogenetic tree (Fig. 18) strongly supports the monophyly of Pteropyrum (node B) and confirms Calligonum as its sister group (node C). Two strongly supported clades of Pteropyrum are detectable: (1) the P. aucheri clade [node D; PP = 0.99, BS (ML) = 100, BS (MP) = 100] and (2) the P. naufelum clade [node E; PP = 1, BS (ML) = 100, BS (MP) = 100]. The first clade (node D) is an unresolved group comprising several accessions of the highly variable and widely distributed P. aucheri s.l. (i.e. P. aucheri subspp. aucheri, ericoides, olivieri and noeanum) with P. scoparium. In the second clade (node E), P. gypsaceum is sister to the rest of the species. Three accessions of P. naufelum form a subclade [node H; PP = 0.96, BS (ML) = 91, BS (MP) = 63] that is placed along with another group of species including P. macrocarpum, P. jakdanense and P. zagricum [node G; PP = 0.93, BS (ML) = 82, BS (MP) = 63]. Figure 18. Open in new tabDownload slide Fifty percent majority-rule phylogenetic tree of Pteropyrum inferred from ITS dataset with Bayesian inference. Posterior probabilities obtained from BI (regular) and bootstrap support values for the same nodes found in ML analysis (boldface) are shown above branches, and bootstrap support values from MP analysis (italics) are indicated below branches. Figure 18. Open in new tabDownload slide Fifty percent majority-rule phylogenetic tree of Pteropyrum inferred from ITS dataset with Bayesian inference. Posterior probabilities obtained from BI (regular) and bootstrap support values for the same nodes found in ML analysis (boldface) are shown above branches, and bootstrap support values from MP analysis (italics) are indicated below branches. Carbon isotope composition Carbon isotope compositions of 24 samples of leaves, cotyledons or assimilating stems belonging to four Atraphaxis spp., six Calligonum spp. and nine taxa of Pteropyrum are given in Table 2. All assimilating shoots of Calligonum spp. have C4-type carbon isotope ratios ranging from −15.13 to −13.96‰ (average of −14.55). Leaves of all studied Atraphaxis spp. have δ 13C ranging from −29.35 to −26.17‰ (average of −27.65) and leaves of all Pteropyrum taxa have δ 13C rations ranging from −29.12 to −24.96‰ (average of −26.85), indicative of non-C4 plants (C3, proto-Kranz or C2 species). The values for three cotyledons of Pteropyrum are −29.06‰ (P. aucheri subsp. ericoides), −30.94‰ (P. aucheri subsp. olivieri) and −27.98‰ (P. jakdanense). Table 2. Carbon isotope values (δ 13C) of some species of Atraphaxis and Calligonum and all known Pteropyrum species. Voucher specimens of those samples with only collector name and number are listed in the studied specimens of taxonomic treatment. Species . Voucher . Analysed plant part . δ 13C (‰) . Atraphaxis seravshanica Pavlov Iran: Golestan National Park, Akhani 11039 Leaf −26.78 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 9832 Leaf −29.35 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 11702 Leaf −26.80 Atraphaxis suaedifolia Jaub. & Spach Iran: 38 km NW of Zanjan, Akhani et al. 22515 Leaf −29.04 Atraphaxis suaedifolia Jaub. & Spach Iran: East Azerbayjan: c. 20 km to Ahar, Wendelbo & Assadi 17119 Leaf −26.17 Atraphaxis tournefortii Jaub. & Spach Iraq: Sulaimaniya, Rechinger 10437 Leaf −27.80 Average of Atraphaxis 27.65 Calligonum bungei Boiss. Iran: Kerman, Doostmohammadi 4658 Green stem −14.88 Calligonum caput-medusae Schrenk Iran: 22 km N Aranbidgol, Akhani et al. 19769 Green stem −13.96 Calligonum crinitum Boiss. Iran: 11 km NE Aranbidgol, Akhani 21379-b Green stem −15.13 Calligonum paletzkianum Litv. Iran: South Khorassan, Yazdan, Doostmohammadi 4838 Green stem −14.65 Calligonum persicum Boiss. Gilan: Near Rudbar, Akhani & Salimain 14345 Green stem −14.62 Calligonum polygonoides L. Azarbaijan, 44 km NW Poldasht, Akhani 18951 Green stem −14.07 Average of Calligonum −14.55 Pteropyrum aucheri subsp. aucheri Dehghani et al. 4788 Leaf −25.14 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4826 Leaf −28.40 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4830 Cotyledon leaf −29.06 Pteropyrum aucheri subsp. olivieri Akhani 18571 Leaf −25.47 Pteropyrum aucheri subsp. olivieri Doostmohammadi 4673 Cotyledon leaf −30.94 Pteropyrum gypsaceum Doostmohammadi 4896 Leaf −25.62 Pteropyrum jakdanense Doostmohammadi 4861 Leaf −24.96 Pteropyrum jakdanense Doostmohammadi 4861 Cotyledon leaf −27.98 Pteropyrum macrocarpum Doostmohammadi 4729 Leaf −27.02 Pteropyrum naufelum Doostmohammadi & Noormohammadi 4875 Leaf −28.21 Pteropyrum scoparium Akhani 24638 Leaf −29.12 Pteropyrum zagricum Doostmohammadi 4898 Leaf −27.76 Average of Pteropyrum (excluding cotyledon values) −26.85 Species . Voucher . Analysed plant part . δ 13C (‰) . Atraphaxis seravshanica Pavlov Iran: Golestan National Park, Akhani 11039 Leaf −26.78 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 9832 Leaf −29.35 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 11702 Leaf −26.80 Atraphaxis suaedifolia Jaub. & Spach Iran: 38 km NW of Zanjan, Akhani et al. 22515 Leaf −29.04 Atraphaxis suaedifolia Jaub. & Spach Iran: East Azerbayjan: c. 20 km to Ahar, Wendelbo & Assadi 17119 Leaf −26.17 Atraphaxis tournefortii Jaub. & Spach Iraq: Sulaimaniya, Rechinger 10437 Leaf −27.80 Average of Atraphaxis 27.65 Calligonum bungei Boiss. Iran: Kerman, Doostmohammadi 4658 Green stem −14.88 Calligonum caput-medusae Schrenk Iran: 22 km N Aranbidgol, Akhani et al. 19769 Green stem −13.96 Calligonum crinitum Boiss. Iran: 11 km NE Aranbidgol, Akhani 21379-b Green stem −15.13 Calligonum paletzkianum Litv. Iran: South Khorassan, Yazdan, Doostmohammadi 4838 Green stem −14.65 Calligonum persicum Boiss. Gilan: Near Rudbar, Akhani & Salimain 14345 Green stem −14.62 Calligonum polygonoides L. Azarbaijan, 44 km NW Poldasht, Akhani 18951 Green stem −14.07 Average of Calligonum −14.55 Pteropyrum aucheri subsp. aucheri Dehghani et al. 4788 Leaf −25.14 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4826 Leaf −28.40 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4830 Cotyledon leaf −29.06 Pteropyrum aucheri subsp. olivieri Akhani 18571 Leaf −25.47 Pteropyrum aucheri subsp. olivieri Doostmohammadi 4673 Cotyledon leaf −30.94 Pteropyrum gypsaceum Doostmohammadi 4896 Leaf −25.62 Pteropyrum jakdanense Doostmohammadi 4861 Leaf −24.96 Pteropyrum jakdanense Doostmohammadi 4861 Cotyledon leaf −27.98 Pteropyrum macrocarpum Doostmohammadi 4729 Leaf −27.02 Pteropyrum naufelum Doostmohammadi & Noormohammadi 4875 Leaf −28.21 Pteropyrum scoparium Akhani 24638 Leaf −29.12 Pteropyrum zagricum Doostmohammadi 4898 Leaf −27.76 Average of Pteropyrum (excluding cotyledon values) −26.85 Open in new tab Table 2. Carbon isotope values (δ 13C) of some species of Atraphaxis and Calligonum and all known Pteropyrum species. Voucher specimens of those samples with only collector name and number are listed in the studied specimens of taxonomic treatment. Species . Voucher . Analysed plant part . δ 13C (‰) . Atraphaxis seravshanica Pavlov Iran: Golestan National Park, Akhani 11039 Leaf −26.78 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 9832 Leaf −29.35 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 11702 Leaf −26.80 Atraphaxis suaedifolia Jaub. & Spach Iran: 38 km NW of Zanjan, Akhani et al. 22515 Leaf −29.04 Atraphaxis suaedifolia Jaub. & Spach Iran: East Azerbayjan: c. 20 km to Ahar, Wendelbo & Assadi 17119 Leaf −26.17 Atraphaxis tournefortii Jaub. & Spach Iraq: Sulaimaniya, Rechinger 10437 Leaf −27.80 Average of Atraphaxis 27.65 Calligonum bungei Boiss. Iran: Kerman, Doostmohammadi 4658 Green stem −14.88 Calligonum caput-medusae Schrenk Iran: 22 km N Aranbidgol, Akhani et al. 19769 Green stem −13.96 Calligonum crinitum Boiss. Iran: 11 km NE Aranbidgol, Akhani 21379-b Green stem −15.13 Calligonum paletzkianum Litv. Iran: South Khorassan, Yazdan, Doostmohammadi 4838 Green stem −14.65 Calligonum persicum Boiss. Gilan: Near Rudbar, Akhani & Salimain 14345 Green stem −14.62 Calligonum polygonoides L. Azarbaijan, 44 km NW Poldasht, Akhani 18951 Green stem −14.07 Average of Calligonum −14.55 Pteropyrum aucheri subsp. aucheri Dehghani et al. 4788 Leaf −25.14 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4826 Leaf −28.40 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4830 Cotyledon leaf −29.06 Pteropyrum aucheri subsp. olivieri Akhani 18571 Leaf −25.47 Pteropyrum aucheri subsp. olivieri Doostmohammadi 4673 Cotyledon leaf −30.94 Pteropyrum gypsaceum Doostmohammadi 4896 Leaf −25.62 Pteropyrum jakdanense Doostmohammadi 4861 Leaf −24.96 Pteropyrum jakdanense Doostmohammadi 4861 Cotyledon leaf −27.98 Pteropyrum macrocarpum Doostmohammadi 4729 Leaf −27.02 Pteropyrum naufelum Doostmohammadi & Noormohammadi 4875 Leaf −28.21 Pteropyrum scoparium Akhani 24638 Leaf −29.12 Pteropyrum zagricum Doostmohammadi 4898 Leaf −27.76 Average of Pteropyrum (excluding cotyledon values) −26.85 Species . Voucher . Analysed plant part . δ 13C (‰) . Atraphaxis seravshanica Pavlov Iran: Golestan National Park, Akhani 11039 Leaf −26.78 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 9832 Leaf −29.35 Atraphaxis spinosa L. Iran: Golestan National Park, Akhani 11702 Leaf −26.80 Atraphaxis suaedifolia Jaub. & Spach Iran: 38 km NW of Zanjan, Akhani et al. 22515 Leaf −29.04 Atraphaxis suaedifolia Jaub. & Spach Iran: East Azerbayjan: c. 20 km to Ahar, Wendelbo & Assadi 17119 Leaf −26.17 Atraphaxis tournefortii Jaub. & Spach Iraq: Sulaimaniya, Rechinger 10437 Leaf −27.80 Average of Atraphaxis 27.65 Calligonum bungei Boiss. Iran: Kerman, Doostmohammadi 4658 Green stem −14.88 Calligonum caput-medusae Schrenk Iran: 22 km N Aranbidgol, Akhani et al. 19769 Green stem −13.96 Calligonum crinitum Boiss. Iran: 11 km NE Aranbidgol, Akhani 21379-b Green stem −15.13 Calligonum paletzkianum Litv. Iran: South Khorassan, Yazdan, Doostmohammadi 4838 Green stem −14.65 Calligonum persicum Boiss. Gilan: Near Rudbar, Akhani & Salimain 14345 Green stem −14.62 Calligonum polygonoides L. Azarbaijan, 44 km NW Poldasht, Akhani 18951 Green stem −14.07 Average of Calligonum −14.55 Pteropyrum aucheri subsp. aucheri Dehghani et al. 4788 Leaf −25.14 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4826 Leaf −28.40 Pteropyrum aucheri subsp. ericoides Doostmohammadi 4830 Cotyledon leaf −29.06 Pteropyrum aucheri subsp. olivieri Akhani 18571 Leaf −25.47 Pteropyrum aucheri subsp. olivieri Doostmohammadi 4673 Cotyledon leaf −30.94 Pteropyrum gypsaceum Doostmohammadi 4896 Leaf −25.62 Pteropyrum jakdanense Doostmohammadi 4861 Leaf −24.96 Pteropyrum jakdanense Doostmohammadi 4861 Cotyledon leaf −27.98 Pteropyrum macrocarpum Doostmohammadi 4729 Leaf −27.02 Pteropyrum naufelum Doostmohammadi & Noormohammadi 4875 Leaf −28.21 Pteropyrum scoparium Akhani 24638 Leaf −29.12 Pteropyrum zagricum Doostmohammadi 4898 Leaf −27.76 Average of Pteropyrum (excluding cotyledon values) −26.85 Open in new tab C3 and C4 anatomy of assimilating parts Selected cross sections of assimilating parts of the studied plants are illustrated in Figs 19–24 and 25–31. The percentages of PCA, PCR, PCA/PCR, epidermal cells (EP) and water storage tissue (WST) for each plant are given in Supporting Information, Appendix S1). The averages of WST and PCA for leaves or assimilating stems of mature plants are shown in Figs 32 and 33. Based on configuration of photosynthetic cells, percentages of PCA, PCR and WST and the succulence of assimilating organs, four categories are recognized: (1) four Atraphaxis spp.; (2) five Pteropyrum taxa with flattened leaves; (3) four Pteropyrum taxa with terete leaves and (4) six Calligonum spp. having only assimilating stems as mature plants. Figures 19–24. Open in new tabDownload slide Cross sections of leaves (19–22 and 23–24) and cotyledons (21–22) of Atraphaxis and Pteropyrum; Fig. 19. A. suaedifolia, Dehghani et al. 4814; Fig. 20. A. spinosa, Akhani 24156; Fig. 21 P. naufelum, seed grown from from Doostmohammadi & Noormohammadi 4876; Fig. 22. P. olivieri, seeds grown from Doostmohammadi 4673; Fig. 23. P. jakdanense, Doostmohammadi 4879; Fig. 24. P. naufelum, Doostmohammadi & Noormohammadi 4876. Arrows represent the enlarged bundle sheath cells. Figures 19–24. Open in new tabDownload slide Cross sections of leaves (19–22 and 23–24) and cotyledons (21–22) of Atraphaxis and Pteropyrum; Fig. 19. A. suaedifolia, Dehghani et al. 4814; Fig. 20. A. spinosa, Akhani 24156; Fig. 21 P. naufelum, seed grown from from Doostmohammadi & Noormohammadi 4876; Fig. 22. P. olivieri, seeds grown from Doostmohammadi 4673; Fig. 23. P. jakdanense, Doostmohammadi 4879; Fig. 24. P. naufelum, Doostmohammadi & Noormohammadi 4876. Arrows represent the enlarged bundle sheath cells. Figures 25–31. Open in new tabDownload slide Cross-sections of leaves (25–26), cotyledon leaves (27–38) and green stems (29–31) of Calligonum. Fig. 25. C. bungei, Doostmohammadi 4654; Fig. 26. C. polygonoides, Doostmohammadi 4669; Fig. 27. C. crinitum, Doostmohammadi 4675; Fig. 28. C. polygonoides, Doostmohammadi 4669; Fig. 29. C. denticulatum, Doostmohammadi 4669; Fig. 30. C. crinitum, Doostmohammadi 4678; Fig. 31. A single peripheral vascular bundle oriented with its xylem side facing towards the Kranz cells. Pa – palisade cells; kr – Kranz cells; xy – xylem; ph – phloem; ws – water storage cells; cb – central bundles; pb – peripheral bundles; hy – hypodermis; ep – epidermis. Figures 25–31. Open in new tabDownload slide Cross-sections of leaves (25–26), cotyledon leaves (27–38) and green stems (29–31) of Calligonum. Fig. 25. C. bungei, Doostmohammadi 4654; Fig. 26. C. polygonoides, Doostmohammadi 4669; Fig. 27. C. crinitum, Doostmohammadi 4675; Fig. 28. C. polygonoides, Doostmohammadi 4669; Fig. 29. C. denticulatum, Doostmohammadi 4669; Fig. 30. C. crinitum, Doostmohammadi 4678; Fig. 31. A single peripheral vascular bundle oriented with its xylem side facing towards the Kranz cells. Pa – palisade cells; kr – Kranz cells; xy – xylem; ph – phloem; ws – water storage cells; cb – central bundles; pb – peripheral bundles; hy – hypodermis; ep – epidermis. Figures 32–33. Open in new tabDownload slide Proportion of WST (Fig. 32) and PCA (or C3 mesophyll, Fig. 33) among leaves of Atraphaxis, flat leaved species group of Pteropyrum, cylindrical leaved species group of Pteropyrum, Calligonum leaf and Calligonum green stems. Vertical lines indicate standard deviations. Figures 32–33. Open in new tabDownload slide Proportion of WST (Fig. 32) and PCA (or C3 mesophyll, Fig. 33) among leaves of Atraphaxis, flat leaved species group of Pteropyrum, cylindrical leaved species group of Pteropyrum, Calligonum leaf and Calligonum green stems. Vertical lines indicate standard deviations. All four Atraphaxis spp. (category 1) investigated here were characterized by flattened ovate to elliptic leaves (Fig. 20), except A. suaedifolia having thickened linear leaves (Fig. 19). The leaves are bifacial with palisade cells on the adaxial and roundish spongy cells on the abaxial side. All species have a typical C3 structure in leaf sections lacking a Kranz layer and BS. They also lack hypodermis with well-developed mesophyll cells in two to five layers. The number of mesophyll layers on the adaxial side is highest in A. suaedifolia (four or five layers, Fig. 19) and lowest in A. intricata (two or three layers); the other two species have three or four layers. The number of mesophyll layers of the spongy cells ranges from five or six layers in A. spinosa (Fig. 20) to two or three layers in A. intricata; the other two species have three to five layers. WST is not well-developed, consisting of only 2–6% among different species (Fig. 32). The mesophyll volume of mature leaves of Atraphaxis spp. is the highest among three studied genera with an average of 68.1% (Fig. 33; Supporting Information, Appendix S1). Mature leaves of nine taxa of Pteropyrum and cotyledons of six Pteropyrum spp. (categories 2 and 3) were anatomically studied (Figs 21–24, Supporting Information, Appendix S1). Five taxa have flat leaves (P. aucheri subsp. aucheri, P. olivieri, P. naufelum and P. gypsaceum and P. zagricum) and the other four taxa (P. aucheri subsp. ericoides, P. scoparium, P. jakdanense and P. macrocarpum) have more-or-less succulent leaves that are terete or semi-terete in cross section (Figs 9–17). All species lack a hypodermis layer with a ± equifacial leaves having two or three palisade layers on adaxial and abaxial sides. The palisade cells on the abaxial surface are slightly smaller than the adaxial ones. BS cells are enlarged in all species without chloroplast orientation, but no Kranz layer is developed in the species studied here. In species with cylindrical leaves the vascular bundles scattered around the WST each with a crescent-like layer of BS cells present both in adaxial and abaxial leaf sides (Fig. 21). The vascular bundles of taxa with flat leaves are horizontally scattered with BS cells occurring only on adaxial leaf side (Fig. 22). The PCA volume of both groups is slightly different on average (40.2% in terete leaves vs. 45.6% in flat leaves) and considering the high standard deviation, this feature is not statistically important (Fig. 33). The two groups are well distinguished in the volume of their WST. In flat leaves, WST occupies on average 24.8% of the leaf volume and in the cylindrical leaves it reaches 55.3% in P. scoparium, with an average of 34.7% among all species (Fig. 32; Supporting Information, Appendix S1). The cotyledons of all studied Pteropyrum spp. are flat, even those with cylindrical mature leaves. Similar to mature leaves, the cotyledons of Pteropyrum lack hypodermis, but differ markedly in the bifacial configuration of mesophyll cells. The palisade layers consist of three to five layers on the adaxial side and two to five layers of spongy cells in the abaxial side (Figs 23, 24). Another main feature of the cotyledons is a well-developed BS tissue surrounding vascular bundles in both abaxial and adaxial sides (Figs 23, 24). Anatomy of young stem, mature leaves and cotyledon leaves of seven Calligonum species from three different sections [category 4; Calligonum sections Calligonum, Pterococcus (Pall.) Endl. and Calliphyssa (Fisch. & C.A.Mey.) Endl.] were studied. Leaves and green stems are succulent, terete with a salsoloid Kranz anatomy and a distinct layer of hypodermis in cross section (Figs 25, 26, 29–31). The leaves appear only in young plants or in the early part of the growing season in spring. They soon die and the assimilating green stems take over the CO2 fixation role. The WST in leaves of Calligonum, with an average of 39%, is higher than that in Atraphaxis and Pteropyrum, and the mesophyll volume, ranging from 22–35% (average of 28%), is the lowest among the three genera (Figs 32, 33, Supporting Information, Appendix S1). Small peripheral vascular bundles are oriented with their xylem side facing towards the BS layer. The anatomy of cotyledons is similar to true leaves, but differs from them in lacking the peripheral vascular bundles (Figs 27, 28). The anatomy of the stem is markedly different from leaves and cotyledons in the formation of well-structured supportive tissues of cholenchyma and sclerenchyma cells (Fig. 29–31). The cholenchyma bundles interrupt the hypodermis layer. The sclerenchyma tissue occurs on the outer and, in some cases, alternatively the inner side of central vascular bundles. The water storage cells occupy the largest volume of the stem both between sclerenchyma cells and the whole central parts of the stem. The Kranz ring encircles the complex composed of vascular bundles, WST and the sclerenchyma fibres (Fig. 30). Seedling morphology In Pteropyrum, the hypocotyl length varies from 1.9–2.6 cm in P. naufelum to 5.2–5.6 cm in P. jakdanense (Figs 34–41). Apart from these two extremes, the hypocotyl height of most species is c. 3–4 cm on average (Table 3). The indumentum of hypocotyl and young stem is different among studied species. Pteropyrum aucheri subsp. olivieri, P. macrocarpum and P. zagricum have glabrous hypocotyls, whereas other taxa are either sparsely papillose or, in the case of P. jakdanense (described here from southern Iran), densely papillose (Fig. 34). Table 3. The length and indumentum of hypocotyls of eight taxa of Pteropyrum. The minimum, maximum and average values and their standard deviations (±) are given based on ten measurements Taxa . Hypocotyl length (cm) . Hypocotyl indumentum . P. aucheri subsp. aucheri 3.2–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. ericoides 3.4–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. olivieri 3.2–3.6 (3.3 ± 0.3) Glabrous P. gypsaceum 3.5–4.4 (4.2 ± 0.1) Sparse papillose P. jakdanense 5.2–5.6 (5.3 ± 0.5) Dense papillose P. macrocarpum 3.9–4.3 (4.1 ± 0.2) Glabrous P. naufelum 1.9–2.6 (2.3 ± 0.2) Sparse papillose P. zagricum 3.5–4.3 (3.6 ± 0.2) Glabrous Taxa . Hypocotyl length (cm) . Hypocotyl indumentum . P. aucheri subsp. aucheri 3.2–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. ericoides 3.4–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. olivieri 3.2–3.6 (3.3 ± 0.3) Glabrous P. gypsaceum 3.5–4.4 (4.2 ± 0.1) Sparse papillose P. jakdanense 5.2–5.6 (5.3 ± 0.5) Dense papillose P. macrocarpum 3.9–4.3 (4.1 ± 0.2) Glabrous P. naufelum 1.9–2.6 (2.3 ± 0.2) Sparse papillose P. zagricum 3.5–4.3 (3.6 ± 0.2) Glabrous Open in new tab Table 3. The length and indumentum of hypocotyls of eight taxa of Pteropyrum. The minimum, maximum and average values and their standard deviations (±) are given based on ten measurements Taxa . Hypocotyl length (cm) . Hypocotyl indumentum . P. aucheri subsp. aucheri 3.2–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. ericoides 3.4–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. olivieri 3.2–3.6 (3.3 ± 0.3) Glabrous P. gypsaceum 3.5–4.4 (4.2 ± 0.1) Sparse papillose P. jakdanense 5.2–5.6 (5.3 ± 0.5) Dense papillose P. macrocarpum 3.9–4.3 (4.1 ± 0.2) Glabrous P. naufelum 1.9–2.6 (2.3 ± 0.2) Sparse papillose P. zagricum 3.5–4.3 (3.6 ± 0.2) Glabrous Taxa . Hypocotyl length (cm) . Hypocotyl indumentum . P. aucheri subsp. aucheri 3.2–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. ericoides 3.4–3.8 (3.6 ± 0.1) Sparse papillose P. aucheri subsp. olivieri 3.2–3.6 (3.3 ± 0.3) Glabrous P. gypsaceum 3.5–4.4 (4.2 ± 0.1) Sparse papillose P. jakdanense 5.2–5.6 (5.3 ± 0.5) Dense papillose P. macrocarpum 3.9–4.3 (4.1 ± 0.2) Glabrous P. naufelum 1.9–2.6 (2.3 ± 0.2) Sparse papillose P. zagricum 3.5–4.3 (3.6 ± 0.2) Glabrous Open in new tab Figure 34–41. Open in new tabDownload slide Seedlings of several species of Pteropyrum. Fig. 34. P. jakdanense, inset shows the pubescent hypocotyl; Fig. 35. P. macrocarpum; Fig. 36. P. aucheri subsp. aucheri; Fig. 37. P. gypsaceum; Fig. 38. P. aucheri subsp. olivieri; Fig. 39. P. zagricum; Fig. 40. P. aucheri subsp. ericoides; Fig. 41. P. naufelum. Figure 34–41. Open in new tabDownload slide Seedlings of several species of Pteropyrum. Fig. 34. P. jakdanense, inset shows the pubescent hypocotyl; Fig. 35. P. macrocarpum; Fig. 36. P. aucheri subsp. aucheri; Fig. 37. P. gypsaceum; Fig. 38. P. aucheri subsp. olivieri; Fig. 39. P. zagricum; Fig. 40. P. aucheri subsp. ericoides; Fig. 41. P. naufelum. Pollen morphology of Pteropyrum Micrographs (LM and SEM) for ten taxa are shown in Figures 42–91. Measured pollen characters are presented in Table 4. Pollen grains of Pteropyrum are prolate-spheroidal to subprolate and trizonocolporate. The pollen size varies between different taxa with many overlapping (Fig. 92). The smallest grains were measured in one population of P. aucheri subsp. aucheri from Birjand (Doostmohammadi 4832) with average polar and equatorial lengths of 23.47 and 20.46 μm, respectively. The largest grains were measured in one population of P. naufelum collected from northern Khuzestan (Akhani 9064-b) with average polar and equatorial lengths of 39.58 and 33.17 μm, respectively. Among the pollen grains of different subspecies of P. aucheri examined here, those of P. aucheri subsp. aucheri had the smallest grains with an average polar length (Pl) of 24.4 μm, followed by its close relatives P. aucheri subspp. ericoides (Pl = 30.2 μm), olivieri (Pl = 30.8 μm) and noeanum (Pl = 35.1 μm). The polar/equatorial ratio (P/E) of all taxa ranges from 1.1 to 1.3. Other pollen characters such as colpus length, maximum length of mesocolpium and apocolpus diameter show more-or-less similar patterns as pollen size (Table 4). The exine thickness among different species is not significantly different with a range of 1.8–2.4 μm thick. The SEM micrographs of pollen ornamentation provide evidence of two distinct pollen types with different exine sculptures: P. naufelum is characterized by a suprareticulate surface deviating from all other taxa that have a semi-striate tectum perforatum (Figs 82–91). Pteropyrum naufelum also shows other morphological features that make it distinct from other Pteropyrum spp., and we provide a more detailed description here. Table 4. Pollen morphological data for Pteropyrum with mean and standard deviations based on 30 pollen grains. All measurements are in µm Taxon name . Polar length . Equatorial length . P/E . Colpus length . Mesocolpium Maximum distance . Apocolpus diameter . Exine thickness . P. aucheri subsp. aucheri Doostmohammadi 4832 23.47 ± 1.7 20.46 ± 1.7 1.14 18.07 ± 1.6 17.1 ± 2.5 14.55 ± 3.1 1.94 ± 0.1 Dehghani et al. 4788 25.09 ± 1.2 22.25 ± 1.1 1.12 19.74 ± 1.5 16.69 ± 0.9 12.73 ± 1.6 1.83 ± 0.1 Dehghani et al. 4782 25.72 ± 1.5 23.21 ± 1.4 1.11 20.92 ± 1.5 18.99 ± 2.4 16.37 ± 2.9 2.07 ± 0.1 Akhani et al. 17728 23.38 ± 1.3 21.82 ± 1.1 1.1 20.09 ± 1.3 16.86 ± 1.7 14.35 ± 1.9 1.9 ± 0.1 Average 24.4 21.9 1.11 19.7 17.4 14.5 1.93 P. aucheri subsp. noeanum Noë 1002 35.5±2.8 28.1±1.7 1.26 29.9±2.9 19.46±0.1 18.06±0.4 2.03±0.1 Dinarvand 10142 34.7±2.1 28.3±2.3 1.23 27±1.3 17.5±0.1 14.9±3.5 2.2±0.1 Average 35.1 28.2 1.24 28.4 18.5 16.5 2.1 P. aucheri subsp. ericoides Doostmohammadi 4824 29.5 ± 1.3 24.84 ± 1.4 1.2 22.63 ± 1.3 18.58 ± 1.2 16.48 ± 1.2 2.16 ± 0.2 Doostmohammadi 4825 30.62 ± 1.5 26.29 ± 1.6 1.16 24.71 ± 1.6 20.05 ± 2.4 17.58 ± 3.5 2.02 ± 0.1 Doostmohammadi 4826 30.61 ± 1.9 25.36 ± 1.5 1.2 24.24 ± 1.9 18.82 ± 1.5 15.77 ± 1.8 1.91 ± 0.2 Average 30.2 25.5 1.18 23.9 19.2 16.6 2.03 P. aucheri subsp. olivieri Doostmohammadi 4852 28.07 ± 1.3 24.73 ± 1.1 1.13 22.81 ± 1.3 17.57 ± 1.8 16.5 ± 2 1.9 ± 0.18 Akhani & Zarrinpor 14797 28.32 ± 1.1 23.65 ± 1 1.2 23.1 ± 1.2 18.97 ± 1.3 16.25 ± 2.5 1.98 ± 0.2 Akhani 1221 34.81 ± 2.7 26.26 ± 1.7 1.32 28.89 ± 1.8 20.39 ± 2.6 17.95 ± 2.6 2.14 ± 0.2 Doostmohammadi 4730 30.58 ± 2.1 26.95 ± 2.2 1.13 24.57 ± 1.9 21.1 ± 2.5 18.52 ± 2.8 1.89 ± 0.2 Doostmohammadi 4673 31.5±1.6 26±1.1 1.21 25.4±1.9 21±2.8 20.5±4.3 1.9±0.2 Akhani et al. 18571 31.75 ± 1.9 26.09 ± 1.2 1.21 26.17 ± 1.8 20.37 ± 2.5 18.63 ± 2.8 1.98 ± 0.2 Average 30.8 25.6 1.2 25.2 19.9 18.1 1.96 P. gypsaceum Akhani 16001 28.77 ± 1.2 25.63 ± 1.5 1.12 23.61 ± 1.5 25.57 ± 3.9 16.95 ± 1.6 1.91 ± 0.2 P. jakdanense Doostmohammadi 4879 27.4±1.8 22.7±1.4 1.2 21.3±1.5 21.7±2.7 20.5±2.4 1.8±0.15 P. macrocarpum Doostmohammadi 4729 38.7 ± 2.4 33.8 ± 3.5 1.14 31.74 ± 2.5 30.05 ± 0.5 27.6 ± 1.2 2.37 ± 0.1 P. naufelum Akhani 9064-b 39.58 ± 2.4 33.17 ± 1.4 1.2 32.46 ± 2.1 25.4 ± 2.9 24.13 ± 3.4 2.3 ± 0.1 Akhani 9049 32.07 ± 1.8 28.39 ± 1.8 1.13 26.48 ± 1.7 23.08 ± 2.6 22.08 ± 3.4 1.8 ± 0.2 Akhani 21580 37.25 ± 2.3 32.94 ± 2.3 1.13 31.27 ± 2.1 23.64 ± 1.3 22.46 ± 1.7 1.94 ± 0.2 Average 36.3 31.5 1.15 30.1 24.1 22.9 2.01 P. scoparium Kürschner 99-33 30.1 ± 2.7 25 ± 1.9 1.2 24.4 ± 3.1 23 ± 3.1 21.5 ± 2.8 2.1 ± 0.2 P. zagricum Bagheri et al. 4780 30 ± 1.5 26.4 ± 1.4 1.14 23.6 ± 1.5 20.9 ± 2.4 19.7 ± 2.8 1.81 ± 0.11 Taxon name . Polar length . Equatorial length . P/E . Colpus length . Mesocolpium Maximum distance . Apocolpus diameter . Exine thickness . P. aucheri subsp. aucheri Doostmohammadi 4832 23.47 ± 1.7 20.46 ± 1.7 1.14 18.07 ± 1.6 17.1 ± 2.5 14.55 ± 3.1 1.94 ± 0.1 Dehghani et al. 4788 25.09 ± 1.2 22.25 ± 1.1 1.12 19.74 ± 1.5 16.69 ± 0.9 12.73 ± 1.6 1.83 ± 0.1 Dehghani et al. 4782 25.72 ± 1.5 23.21 ± 1.4 1.11 20.92 ± 1.5 18.99 ± 2.4 16.37 ± 2.9 2.07 ± 0.1 Akhani et al. 17728 23.38 ± 1.3 21.82 ± 1.1 1.1 20.09 ± 1.3 16.86 ± 1.7 14.35 ± 1.9 1.9 ± 0.1 Average 24.4 21.9 1.11 19.7 17.4 14.5 1.93 P. aucheri subsp. noeanum Noë 1002 35.5±2.8 28.1±1.7 1.26 29.9±2.9 19.46±0.1 18.06±0.4 2.03±0.1 Dinarvand 10142 34.7±2.1 28.3±2.3 1.23 27±1.3 17.5±0.1 14.9±3.5 2.2±0.1 Average 35.1 28.2 1.24 28.4 18.5 16.5 2.1 P. aucheri subsp. ericoides Doostmohammadi 4824 29.5 ± 1.3 24.84 ± 1.4 1.2 22.63 ± 1.3 18.58 ± 1.2 16.48 ± 1.2 2.16 ± 0.2 Doostmohammadi 4825 30.62 ± 1.5 26.29 ± 1.6 1.16 24.71 ± 1.6 20.05 ± 2.4 17.58 ± 3.5 2.02 ± 0.1 Doostmohammadi 4826 30.61 ± 1.9 25.36 ± 1.5 1.2 24.24 ± 1.9 18.82 ± 1.5 15.77 ± 1.8 1.91 ± 0.2 Average 30.2 25.5 1.18 23.9 19.2 16.6 2.03 P. aucheri subsp. olivieri Doostmohammadi 4852 28.07 ± 1.3 24.73 ± 1.1 1.13 22.81 ± 1.3 17.57 ± 1.8 16.5 ± 2 1.9 ± 0.18 Akhani & Zarrinpor 14797 28.32 ± 1.1 23.65 ± 1 1.2 23.1 ± 1.2 18.97 ± 1.3 16.25 ± 2.5 1.98 ± 0.2 Akhani 1221 34.81 ± 2.7 26.26 ± 1.7 1.32 28.89 ± 1.8 20.39 ± 2.6 17.95 ± 2.6 2.14 ± 0.2 Doostmohammadi 4730 30.58 ± 2.1 26.95 ± 2.2 1.13 24.57 ± 1.9 21.1 ± 2.5 18.52 ± 2.8 1.89 ± 0.2 Doostmohammadi 4673 31.5±1.6 26±1.1 1.21 25.4±1.9 21±2.8 20.5±4.3 1.9±0.2 Akhani et al. 18571 31.75 ± 1.9 26.09 ± 1.2 1.21 26.17 ± 1.8 20.37 ± 2.5 18.63 ± 2.8 1.98 ± 0.2 Average 30.8 25.6 1.2 25.2 19.9 18.1 1.96 P. gypsaceum Akhani 16001 28.77 ± 1.2 25.63 ± 1.5 1.12 23.61 ± 1.5 25.57 ± 3.9 16.95 ± 1.6 1.91 ± 0.2 P. jakdanense Doostmohammadi 4879 27.4±1.8 22.7±1.4 1.2 21.3±1.5 21.7±2.7 20.5±2.4 1.8±0.15 P. macrocarpum Doostmohammadi 4729 38.7 ± 2.4 33.8 ± 3.5 1.14 31.74 ± 2.5 30.05 ± 0.5 27.6 ± 1.2 2.37 ± 0.1 P. naufelum Akhani 9064-b 39.58 ± 2.4 33.17 ± 1.4 1.2 32.46 ± 2.1 25.4 ± 2.9 24.13 ± 3.4 2.3 ± 0.1 Akhani 9049 32.07 ± 1.8 28.39 ± 1.8 1.13 26.48 ± 1.7 23.08 ± 2.6 22.08 ± 3.4 1.8 ± 0.2 Akhani 21580 37.25 ± 2.3 32.94 ± 2.3 1.13 31.27 ± 2.1 23.64 ± 1.3 22.46 ± 1.7 1.94 ± 0.2 Average 36.3 31.5 1.15 30.1 24.1 22.9 2.01 P. scoparium Kürschner 99-33 30.1 ± 2.7 25 ± 1.9 1.2 24.4 ± 3.1 23 ± 3.1 21.5 ± 2.8 2.1 ± 0.2 P. zagricum Bagheri et al. 4780 30 ± 1.5 26.4 ± 1.4 1.14 23.6 ± 1.5 20.9 ± 2.4 19.7 ± 2.8 1.81 ± 0.11 Open in new tab Table 4. Pollen morphological data for Pteropyrum with mean and standard deviations based on 30 pollen grains. All measurements are in µm Taxon name . Polar length . Equatorial length . P/E . Colpus length . Mesocolpium Maximum distance . Apocolpus diameter . Exine thickness . P. aucheri subsp. aucheri Doostmohammadi 4832 23.47 ± 1.7 20.46 ± 1.7 1.14 18.07 ± 1.6 17.1 ± 2.5 14.55 ± 3.1 1.94 ± 0.1 Dehghani et al. 4788 25.09 ± 1.2 22.25 ± 1.1 1.12 19.74 ± 1.5 16.69 ± 0.9 12.73 ± 1.6 1.83 ± 0.1 Dehghani et al. 4782 25.72 ± 1.5 23.21 ± 1.4 1.11 20.92 ± 1.5 18.99 ± 2.4 16.37 ± 2.9 2.07 ± 0.1 Akhani et al. 17728 23.38 ± 1.3 21.82 ± 1.1 1.1 20.09 ± 1.3 16.86 ± 1.7 14.35 ± 1.9 1.9 ± 0.1 Average 24.4 21.9 1.11 19.7 17.4 14.5 1.93 P. aucheri subsp. noeanum Noë 1002 35.5±2.8 28.1±1.7 1.26 29.9±2.9 19.46±0.1 18.06±0.4 2.03±0.1 Dinarvand 10142 34.7±2.1 28.3±2.3 1.23 27±1.3 17.5±0.1 14.9±3.5 2.2±0.1 Average 35.1 28.2 1.24 28.4 18.5 16.5 2.1 P. aucheri subsp. ericoides Doostmohammadi 4824 29.5 ± 1.3 24.84 ± 1.4 1.2 22.63 ± 1.3 18.58 ± 1.2 16.48 ± 1.2 2.16 ± 0.2 Doostmohammadi 4825 30.62 ± 1.5 26.29 ± 1.6 1.16 24.71 ± 1.6 20.05 ± 2.4 17.58 ± 3.5 2.02 ± 0.1 Doostmohammadi 4826 30.61 ± 1.9 25.36 ± 1.5 1.2 24.24 ± 1.9 18.82 ± 1.5 15.77 ± 1.8 1.91 ± 0.2 Average 30.2 25.5 1.18 23.9 19.2 16.6 2.03 P. aucheri subsp. olivieri Doostmohammadi 4852 28.07 ± 1.3 24.73 ± 1.1 1.13 22.81 ± 1.3 17.57 ± 1.8 16.5 ± 2 1.9 ± 0.18 Akhani & Zarrinpor 14797 28.32 ± 1.1 23.65 ± 1 1.2 23.1 ± 1.2 18.97 ± 1.3 16.25 ± 2.5 1.98 ± 0.2 Akhani 1221 34.81 ± 2.7 26.26 ± 1.7 1.32 28.89 ± 1.8 20.39 ± 2.6 17.95 ± 2.6 2.14 ± 0.2 Doostmohammadi 4730 30.58 ± 2.1 26.95 ± 2.2 1.13 24.57 ± 1.9 21.1 ± 2.5 18.52 ± 2.8 1.89 ± 0.2 Doostmohammadi 4673 31.5±1.6 26±1.1 1.21 25.4±1.9 21±2.8 20.5±4.3 1.9±0.2 Akhani et al. 18571 31.75 ± 1.9 26.09 ± 1.2 1.21 26.17 ± 1.8 20.37 ± 2.5 18.63 ± 2.8 1.98 ± 0.2 Average 30.8 25.6 1.2 25.2 19.9 18.1 1.96 P. gypsaceum Akhani 16001 28.77 ± 1.2 25.63 ± 1.5 1.12 23.61 ± 1.5 25.57 ± 3.9 16.95 ± 1.6 1.91 ± 0.2 P. jakdanense Doostmohammadi 4879 27.4±1.8 22.7±1.4 1.2 21.3±1.5 21.7±2.7 20.5±2.4 1.8±0.15 P. macrocarpum Doostmohammadi 4729 38.7 ± 2.4 33.8 ± 3.5 1.14 31.74 ± 2.5 30.05 ± 0.5 27.6 ± 1.2 2.37 ± 0.1 P. naufelum Akhani 9064-b 39.58 ± 2.4 33.17 ± 1.4 1.2 32.46 ± 2.1 25.4 ± 2.9 24.13 ± 3.4 2.3 ± 0.1 Akhani 9049 32.07 ± 1.8 28.39 ± 1.8 1.13 26.48 ± 1.7 23.08 ± 2.6 22.08 ± 3.4 1.8 ± 0.2 Akhani 21580 37.25 ± 2.3 32.94 ± 2.3 1.13 31.27 ± 2.1 23.64 ± 1.3 22.46 ± 1.7 1.94 ± 0.2 Average 36.3 31.5 1.15 30.1 24.1 22.9 2.01 P. scoparium Kürschner 99-33 30.1 ± 2.7 25 ± 1.9 1.2 24.4 ± 3.1 23 ± 3.1 21.5 ± 2.8 2.1 ± 0.2 P. zagricum Bagheri et al. 4780 30 ± 1.5 26.4 ± 1.4 1.14 23.6 ± 1.5 20.9 ± 2.4 19.7 ± 2.8 1.81 ± 0.11 Taxon name . Polar length . Equatorial length . P/E . Colpus length . Mesocolpium Maximum distance . Apocolpus diameter . Exine thickness . P. aucheri subsp. aucheri Doostmohammadi 4832 23.47 ± 1.7 20.46 ± 1.7 1.14 18.07 ± 1.6 17.1 ± 2.5 14.55 ± 3.1 1.94 ± 0.1 Dehghani et al. 4788 25.09 ± 1.2 22.25 ± 1.1 1.12 19.74 ± 1.5 16.69 ± 0.9 12.73 ± 1.6 1.83 ± 0.1 Dehghani et al. 4782 25.72 ± 1.5 23.21 ± 1.4 1.11 20.92 ± 1.5 18.99 ± 2.4 16.37 ± 2.9 2.07 ± 0.1 Akhani et al. 17728 23.38 ± 1.3 21.82 ± 1.1 1.1 20.09 ± 1.3 16.86 ± 1.7 14.35 ± 1.9 1.9 ± 0.1 Average 24.4 21.9 1.11 19.7 17.4 14.5 1.93 P. aucheri subsp. noeanum Noë 1002 35.5±2.8 28.1±1.7 1.26 29.9±2.9 19.46±0.1 18.06±0.4 2.03±0.1 Dinarvand 10142 34.7±2.1 28.3±2.3 1.23 27±1.3 17.5±0.1 14.9±3.5 2.2±0.1 Average 35.1 28.2 1.24 28.4 18.5 16.5 2.1 P. aucheri subsp. ericoides Doostmohammadi 4824 29.5 ± 1.3 24.84 ± 1.4 1.2 22.63 ± 1.3 18.58 ± 1.2 16.48 ± 1.2 2.16 ± 0.2 Doostmohammadi 4825 30.62 ± 1.5 26.29 ± 1.6 1.16 24.71 ± 1.6 20.05 ± 2.4 17.58 ± 3.5 2.02 ± 0.1 Doostmohammadi 4826 30.61 ± 1.9 25.36 ± 1.5 1.2 24.24 ± 1.9 18.82 ± 1.5 15.77 ± 1.8 1.91 ± 0.2 Average 30.2 25.5 1.18 23.9 19.2 16.6 2.03 P. aucheri subsp. olivieri Doostmohammadi 4852 28.07 ± 1.3 24.73 ± 1.1 1.13 22.81 ± 1.3 17.57 ± 1.8 16.5 ± 2 1.9 ± 0.18 Akhani & Zarrinpor 14797 28.32 ± 1.1 23.65 ± 1 1.2 23.1 ± 1.2 18.97 ± 1.3 16.25 ± 2.5 1.98 ± 0.2 Akhani 1221 34.81 ± 2.7 26.26 ± 1.7 1.32 28.89 ± 1.8 20.39 ± 2.6 17.95 ± 2.6 2.14 ± 0.2 Doostmohammadi 4730 30.58 ± 2.1 26.95 ± 2.2 1.13 24.57 ± 1.9 21.1 ± 2.5 18.52 ± 2.8 1.89 ± 0.2 Doostmohammadi 4673 31.5±1.6 26±1.1 1.21 25.4±1.9 21±2.8 20.5±4.3 1.9±0.2 Akhani et al. 18571 31.75 ± 1.9 26.09 ± 1.2 1.21 26.17 ± 1.8 20.37 ± 2.5 18.63 ± 2.8 1.98 ± 0.2 Average 30.8 25.6 1.2 25.2 19.9 18.1 1.96 P. gypsaceum Akhani 16001 28.77 ± 1.2 25.63 ± 1.5 1.12 23.61 ± 1.5 25.57 ± 3.9 16.95 ± 1.6 1.91 ± 0.2 P. jakdanense Doostmohammadi 4879 27.4±1.8 22.7±1.4 1.2 21.3±1.5 21.7±2.7 20.5±2.4 1.8±0.15 P. macrocarpum Doostmohammadi 4729 38.7 ± 2.4 33.8 ± 3.5 1.14 31.74 ± 2.5 30.05 ± 0.5 27.6 ± 1.2 2.37 ± 0.1 P. naufelum Akhani 9064-b 39.58 ± 2.4 33.17 ± 1.4 1.2 32.46 ± 2.1 25.4 ± 2.9 24.13 ± 3.4 2.3 ± 0.1 Akhani 9049 32.07 ± 1.8 28.39 ± 1.8 1.13 26.48 ± 1.7 23.08 ± 2.6 22.08 ± 3.4 1.8 ± 0.2 Akhani 21580 37.25 ± 2.3 32.94 ± 2.3 1.13 31.27 ± 2.1 23.64 ± 1.3 22.46 ± 1.7 1.94 ± 0.2 Average 36.3 31.5 1.15 30.1 24.1 22.9 2.01 P. scoparium Kürschner 99-33 30.1 ± 2.7 25 ± 1.9 1.2 24.4 ± 3.1 23 ± 3.1 21.5 ± 2.8 2.1 ± 0.2 P. zagricum Bagheri et al. 4780 30 ± 1.5 26.4 ± 1.4 1.14 23.6 ± 1.5 20.9 ± 2.4 19.7 ± 2.8 1.81 ± 0.11 Open in new tab Figures 42–61. Open in new tabDownload slide Light microscopy images of pollen grains of five taxa of Pteropyrum. Figs 42–45. P. aucheri subsp. aucheri, Dehghani et al. 4788; Figs 46–49. P. aucheri subsp. noeanum, type specimen of P. noeanum, Noe 1002; Figs 50–53. P. aucheri subsp. ericoides, Doostmohammadi 4825; Figs 54–57. P. aucheri subsp. olivieri, Akhani & Zarrinpour 14979; Figs 58–61. P. scoparium, Kurschner 99-33. Scale bar = 10 μm. Figures 42–61. Open in new tabDownload slide Light microscopy images of pollen grains of five taxa of Pteropyrum. Figs 42–45. P. aucheri subsp. aucheri, Dehghani et al. 4788; Figs 46–49. P. aucheri subsp. noeanum, type specimen of P. noeanum, Noe 1002; Figs 50–53. P. aucheri subsp. ericoides, Doostmohammadi 4825; Figs 54–57. P. aucheri subsp. olivieri, Akhani & Zarrinpour 14979; Figs 58–61. P. scoparium, Kurschner 99-33. Scale bar = 10 μm. Figures 62–81. Open in new tabDownload slide Light microscopy images of pollen grains of five species of Pteropyrum. Figs 62–65. P. naufelum, Akhani 21580; Figs 66–69. P. macrocarpum, Doostmohammadi 4729; Figs 70–73. P. jakdanense, Doostmohammadi 4879; Figs 74–77. P. gypsaceum, Akhani 16001; Figs 78–81. P. zagricum, Bagheri et al. 4780. Scale bar = 10 μm. Figures 62–81. Open in new tabDownload slide Light microscopy images of pollen grains of five species of Pteropyrum. Figs 62–65. P. naufelum, Akhani 21580; Figs 66–69. P. macrocarpum, Doostmohammadi 4729; Figs 70–73. P. jakdanense, Doostmohammadi 4879; Figs 74–77. P. gypsaceum, Akhani 16001; Figs 78–81. P. zagricum, Bagheri et al. 4780. Scale bar = 10 μm. Figures 82–91. Open in new tabDownload slide Scanning electron micrographs of pollen grains of ten taxa of Pteropyrum. Fig. 82. P. aucheri subsp. aucheri, Dehghani et al. 4788; Fig. 83. P. naufelum, Akhani 21580; Fig. 84. P. aucheri subsp. noeanum, type specimen of P. noeanum, Noe 1002; Fig. 85. P. macrocarpum, Doostmohammadi 4729; Fig. 86. P. aucheri subsp. olivieri, Akhani & Zarrinpour 14979; Fig. 87. P. gypsaceum, Akhani 16001; Fig. 88. P. aucheri subsp. ericoides, Doostmohammadi 4825; Fig. 89. P. jakdanense, Doostmohammadi 4879; Fig. 90. P. scoparium, Kürschner 99-33; Fig. 91. P. zagricum, Bagheri et al. 4780. Scale bars = 5 µm (A); 1 µm (B). Figures 82–91. Open in new tabDownload slide Scanning electron micrographs of pollen grains of ten taxa of Pteropyrum. Fig. 82. P. aucheri subsp. aucheri, Dehghani et al. 4788; Fig. 83. P. naufelum, Akhani 21580; Fig. 84. P. aucheri subsp. noeanum, type specimen of P. noeanum, Noe 1002; Fig. 85. P. macrocarpum, Doostmohammadi 4729; Fig. 86. P. aucheri subsp. olivieri, Akhani & Zarrinpour 14979; Fig. 87. P. gypsaceum, Akhani 16001; Fig. 88. P. aucheri subsp. ericoides, Doostmohammadi 4825; Fig. 89. P. jakdanense, Doostmohammadi 4879; Fig. 90. P. scoparium, Kürschner 99-33; Fig. 91. P. zagricum, Bagheri et al. 4780. Scale bars = 5 µm (A); 1 µm (B). Figure 92. Open in new tabDownload slide Variation in polar axis length of pollen grains among different populations of Pteropyrum. The middle point of each horizontal line indicates the average of 30 counts. Different species are separated by solid lines. Figure 92. Open in new tabDownload slide Variation in polar axis length of pollen grains among different populations of Pteropyrum. The middle point of each horizontal line indicates the average of 30 counts. Different species are separated by solid lines. The exine of P. naufelum shows a constant thickness all around the pollen. The ectocolpi are bordered by costae (costa ectocolpi) that thin out towards the colpus ends. Endopori are not always present, but when present are mostly circular to elliptic with the longer axis of the ellipse oriented along the polar axis. Endopori also have costae (costa endopori). Observation under phase contrast at 1000× magnification reveals that the columellae mostly occur in groups of five or six. The exine is reticulate with muri formed by fusion of the apices of columellae and organized in a real reticulate and not semi-striate pattern. DISCUSSION Diversity of Pteropyrum Pteropyrum is a small exclusively Irano–Turanian genus with several xeromorphic traits, namely sturdy rigid branches and small, more-or-less succulent leaves that often fall in extremely hot and dry seasons (Figs 9–17, 94–101). The species have a limited range in Iran and neighbouring countries including Iraq, Afghanistan, Pakistan, Turkmenistan, the UAE and Oman (Rechinger & Schiman-Czeika, 1968; Nyberg & Miller, 1996; Breckle, Hedge & Rafiqpoor, 2013; Edmondson & Akeroyd, 2016) (Figs 102–107). Occurrence of the genus in Turkey is not confirmed as the plant material cited in Flora of Turkey (Nâbelek 463, Cullen, 1967) belongs to Atraphaxis. So far, only four species have been accepted in the mentioned standard floras (P. aucheri, P. naufelum, P. olivieri and P. scoparium). Our multidisciplinary approach revealed presence of additional species particularly in the southern distribution area. These species have been unknown for the botanists due to (1) limited collection and the overlooking of small populations in southern and western Iran and (2) the limitation of classical morphological characters in delineating species diversity in oligospecific genera. In particular, the molecular techniques provide unequivocal insight into species diversity and uncovering cryptic species with scarce morphological characters (Bickford et al., 2006; Shneyer & Kotseruba, 2015). Furthermore, relying only on morphological data from herbarium specimens is not informative enough to distinguish species diversity in critical taxa. Therefore, we examined applicability of other characters such as anatomy, micromorphology of epidermis (not presented in this paper), pollen morphology and seedling structure. This approach recently led to the discovery of two additional species of Bienertia Bunge with a similar distribution range, which previously was considered as monotypic (Akhani et al., 2005, 2012). We found the importance of habitat, geography and climate as crucial bases for understanding the species area and species delimitation in closely related species groups with minor distinguishing characters in the Irano–Turanian region. Based on combined morpho-molecular data we describe here four new species (P. gypsaceum, P. macrocarpum, P. jakdanense and P. zagricum). The most widespread species in the area is the P. aucheri complex, treated here in a broad sense to include P. olivieri. We hypothesized that species boundaries in sympatric microspecies are not established and therefore suggested a subspecific classification with four subspecies representing geographical races correlated with some morphological characters (see taxonomic enumeration). Phylogeny and classification Our sequence analysis of the nuclear (ITS) marker supports the monophyly of Pteropyrum as sister to Calligonum (Fig. 18), confirming previous studies (Tavakkoli et al., 2010; Sanchez et al., 2009; Schuster et al., 2015). The ITS tree resulted in two highly supported clades in Pteropyrum. The P. aucheri clade (node D) includes four subspecies of P. aucheri recognized in this paper in addition to P. scoparium, an endemic species of easternmost parts of the Arabian Peninsula (Fig. 107). Jaubert & Spach (1846) recognized two sections in Pteropyrum that are distinguished based on fruit morphology: P. section Orthocarya characterized by non-rotated upper lobes and P. section Streptocarya with rotated upper fruit lobes. Pteropyrum scoparium is the only representative of section Orthocarya and the remaining species belong to section Streptocarya. This sectional classification based on fruit morphology is not confirmed in our phylogenetic tree (Fig. 18). There is no obvious morphological trait characterizing the P. aucheri and P. naufelum clades. Ecologically, all members of P. aucheri clade inhabit dry riverbeds, but most of the species of P. naufelum clade inhabit gypsum, rocky or shale habitats with the exception of P. macrocarpum that grows on riverbeds, similar to members of the P. aucheri clade. All species of P. naufelum clade have restricted ranges in an arc extending from the Zagros foothills of the Iran–Iraq border to the south-eastern parts of Iran with a similar climate as an ecotone between the Irano–Turanian and tropical climates (Djamali et al., 2011). Occurrence of terete and flat leaf types in both clades supports the convergent evolution of terete leaves from flat leaves pushed by ecological pressures (Nyffeler et al., 2008; Ogburn & Edwards, 2013). Leaf shape diversity in P. aucheri is an example of the role of habitat and climate on leaf morphology: P. aucheri subsp. olivieri with flat leaves inhabits mountain areas with moderate climate and more water supplies; P. aucheri subsp. aucheri with narrower leaves grows in drier habitats partly sympatric with P. aucheri subsp. olivieri (Fig. 102) and P. aucheri subsp. ericoides with almost terete leaves penetrates further into desert areas in central and eastern parts of Iran (Fig. 104). C4 evolution in polygonaceae Based on leaf anatomy and carbon isotope values, all examined members of Calligonum show C4 photosynthesis, in agreement with previous studies (Winter et al., 1977; Winter, 1981; Muhaidat et al., 2007). The green assimilating shoots including young leaves early in the growing season and green annual stems in Calligonum exhibit C4 photosynthesis. Calligonum seeds have a large amount of endosperm, serve as storage organs and the tiny cotyledons have no role in the storage of organic substances. However, as the seeds germinate, cotyledons grow rapidly and, in most cases, they grow to > 5 cm in length. The cotyledons perform C4 photosynthesis and have an important function in seedling establishment and early carbon gain prior to emergence of true leaves. Anatomy of cotyledons and mature leaves are not always the same and these two organs may have different anatomical or even different photosynthetic types (Akhani & Ghasemkhani, 2007; Akhani & Khoshravesh, 2013; Pyankov et al., 1999; Pyankov et al., 2000). Anatomy of cotyledons in Calligonum is also different from salsoloid anatomy of leaves and/or green stems and is similar to salsinoid anatomy. Salsinoid anatomy, which was first introduced by Carolin, Jacobs & Vesk (1975) and then by Jacobs (2001) under the name of Kranz-suaedoid, is different from salsoloid anatomy in lacking the peripheral vascular bundles (Edwards & Voznesenskaya, 2011) and has only previously been detected in Suaeda section Salsina Moq. (Schütze, Freitag & Weising, 2003). Anatomy of cotyledons in Calligonum can be regarded as a variant of this type containing a distinct layer of hypodermis. In most cases, vein density decreases with an increase in leaf succulence, and consequently interveinal distances increases. Vein density decrease leads to a longer transport distance between veins and mesophyll cells that can affect photosynthesis and restrict whole plant growth (Ogburn & Edwards, 2013). Therefore, evolution of succulent plants is restricted, and these plants have to evolve an efficient leaf venation system prior to or while becoming succulent. As a solution it is proposed that, in multiple phylogenetic groups, three-dimensional (3D) venation patterns can maintain the transport pathway in a moderate distance that allows tissues to store a large amount of water and become succulent (Griffiths, 2013; Ogburn & Edwards, 2013). Ogburn & Edwards (2013) identified two types of 3D venation (types І and ІІ), which differ in the orientation of xylem and phloem strands in abaxial veins. They proposed that 3D type І evolved through the adaxialization procedure. This process is well represented in Pteropyrum. Mean WST percentage in Atraphaxis, planar leaved Pteropyrum, terete leaved Pteropyrum and Calligonum are 4.9, 24.8, 34.7 and 39.2%, respectively (Fig. 32). Atraphaxis leaves have little WST, and vascular bundles are normally arranged in the middle of the leaf with their xylem strands facing towards the adaxial surface. Planar leaved Pteropyrum spp. are more succulent than Atraphaxis and WS cells are placed between the veins and abaxial mesophyll. Therefore, vascular bundles contact the adaxial mesophyll from their xylem side while their phloem side is facing towards the WST. By diminishing the abaxial leaf surface and increase in WST, planar leaves gradually convert to terete leaves with 3D vein systems in which main veins are in the middle of WST and peripheral veins are oriented with their xylem side facing the mesophyll (Fig. 93). Terete leaves in Pteropyrum exhibit a sympegmoid anatomy with two or three layers of mesophyll cells. It has been suggested that salsoloid anatomy has evolved from sympegmoid anatomy in ancestral plants (Voznesenskaya et al., 2013). Kadereit, Askerly & Pirie (2012) showed that in the evolution of succulent C4 plants (including salsoloid anatomy), succulence and salt tolerance have evolved prior to C4. Hence, succulence is a precondition in the evolution of salsoloid anatomy and on the other hand an efficient vein system (like 3D venation) is required for development of succulence. Occurrence of 3D venation and the consequential improvement of succulence in salsoloid anatomy have occurred via an adaxialization procedure that is also demonstrated in the leaf anatomy of Pteropyrum. It is also proposed here that something opposite to adaxialization (abaxialization) probably occurred during the evolution of tecticornioid anatomy, as the peripheral veins are oriented in the opposite direction to that of salsoloid anatomy. Figure 93. Open in new tabDownload slide Evolutionary transition of planar leaves to terete leaves along with conversion of 2D venation type to 3D. A. Atraphaxis spinosa; B. P. aucheri subsp. aucheri and C. P. jakdanense. Vascular bundles in planar leaves are oriented with their xylem cells (blue cells) contacting the adaxial leaf mesophyll and phloem cells (red cells) facing towards the abaxial side (Fig. 93B). In terete leaves in both adaxial and abaxial sides, xylem sides of the peripheral vascular bundles are connected to mesophyll cells and their phloem sides are facing towards the central aqueous tissue (Fig. 93C). Figure 93. Open in new tabDownload slide Evolutionary transition of planar leaves to terete leaves along with conversion of 2D venation type to 3D. A. Atraphaxis spinosa; B. P. aucheri subsp. aucheri and C. P. jakdanense. Vascular bundles in planar leaves are oriented with their xylem cells (blue cells) contacting the adaxial leaf mesophyll and phloem cells (red cells) facing towards the abaxial side (Fig. 93B). In terete leaves in both adaxial and abaxial sides, xylem sides of the peripheral vascular bundles are connected to mesophyll cells and their phloem sides are facing towards the central aqueous tissue (Fig. 93C). The polyphyletic origin of C4 photosynthesis in a few families indicates that there are some preconditioning traits in distinct C3 plants that increase the potential for C4 evolution (McKown, Mocalvo & Dengler, 2005; Khoshravesh et al., 2012). Among these preconditions are some anatomical enablers including high vein density and enlarged BS cells (Sage, 2001; Christin et al., 2012; Griffiths et al., 2013). It is largely accepted that in the evolutionary trajectory of C4 plants, C2 Kranz has evolved from C3 ancestors with enlarged BS cells (known as proto-Kranz) (Muhaidat et al., 2011; Sage et al., 2014). In Pteropyrum spp., M tissue area is relatively low (42.9%) compared to Atraphaxis (68.1%) (Fig. 33), and BS cells are enlarged, which may reflect the presence of some anatomical preconditions for C4 photosynthesis in these species. All the studied species have almost enlarged BS cells; however, PCR percentage is the highest in P. aucheri subsp. aucheri (4.1%) and P. jakdanense (3.8%) among planar leaved and terete leaved species, respectively. These initial steps in the evolution of C4 are not generally detectable by carbon isotope observations, however, close phylogenetic relationships of Pteropyrum with Calligonum and anatomical evidence, such as enlarged BS cells, increased WST percentage and presence of 3D venation in some species, implies that some Pteropyrum spp. more probably exhibit proto-Kranz or C2 conditions. However, final confirmation of this hypothesis will require additional studies. Palynological application Pollen morphological characters indicate that P. nafelum pollen approaches the Fagopyrum esculentum Moench type (Leeuwen, Punt & Hoen, 1988), but with the reticulate structure being more visible and the pollen size variation being more significant in the latter pollen type. In a pollen morphological point of view, the examined pollen of Pteropyrum can thus be easily divided into two pollen types, the P. aucheri and P. naufelum types. The latter type is mainly distinguished by its larger size and suprareticulate ornamentation. Our palynological results are important in ecological interpretation of fossil pollen data in that they help to distinguish between three pollen types in the Calligonum-Pteropyrum group. This group includes the Calligonum type, indicator of desertification and sand dune expansion, the P. aucheri type, an indicator of desertic areas with well-drained soils and the P. naufelum type that can provide two sorts of information, i.e. hot desert conditions with soils containing a high content of gypsum. This information can thus be useful for palaeoecologists working on the Quaternary plant ecology of interior south-west Asia. Taxonomic revision of Pteropyrum Morphology and nomenclature An updated synopsis of the genus including identification key to species, descriptions and distribution maps of all accepted Pteropyrum spp. is provided. The importance of habitat and climate is discussed above. Some morphological characters used in delimitation and circumscription of the species are presented in Figures 94–101. Figures 94–101. Open in new tabDownload slide Morphological traits used for species delimitation of Pteropyrum. Fruit shapes. Twisted upper wing lobes in P. aucheri subsp. aucheri (Fig. 94) and P. aucheri subsp. olivieri (Fig. 95) and straight upper wing lobes in P. scoparium (Fig. 96). Three different leaf shapes: spathulate (Fig. 97), linear (Fig. 98) and cylindrical (Fig. 99). Leaf indumentum: glabrous leaf in P. jakdanense (Fig. 100, scale bar = 100 μm); pubescent leaf in P. macrocarpum (Fig. 101, scale bar = 150 μm). Figures 94–101. Open in new tabDownload slide Morphological traits used for species delimitation of Pteropyrum. Fruit shapes. Twisted upper wing lobes in P. aucheri subsp. aucheri (Fig. 94) and P. aucheri subsp. olivieri (Fig. 95) and straight upper wing lobes in P. scoparium (Fig. 96). Three different leaf shapes: spathulate (Fig. 97), linear (Fig. 98) and cylindrical (Fig. 99). Leaf indumentum: glabrous leaf in P. jakdanense (Fig. 100, scale bar = 100 μm); pubescent leaf in P. macrocarpum (Fig. 101, scale bar = 150 μm). Pteropyrum Jaub. & Spach, Ill. Pl. Orient. 2: 7. 1844. Type: Pteropyrum aucheri Jaub. & Spach. Shrubby plants; leaves fasciculate; flowers in cluster, pedicels jointed near the base or in the middle, associated with leaf clusters, bisexual, perianth segments five, connate at base, petaloid, white inside, three inner segments slightly larger than two outer segments, inner segments slightly accrescent in fruit, each segment with a greenish central part; stamens eight (three + five), three longer ones opposite to the larger perianth segments, five shorter ones alternate with perianth segments, filament barbellate at base, anthers reddish, versatile; carpels three, style short, with three capitate red stigma; fruit achene, with three usually reddish-membranous wings, wings bilobed, upper parts of fruit in the same direction of lower or twisted at 120°. Key to identification of the species 1. Leaves cylindrical, linear-filiform, < 1 mm wide, terete to semi-terete in cross section ……..…...… 2 1. Leaves flat, oblanceolate to obovate, linear, > 1 mm wide, not terete in cross section ……………………. 5 2. Leaves pubescent, achenes (8–) 10–12 mm long …......................................…..……. P. macrocarpum 2. Leaves glabrous or rarely subglabrous, achenes up to 7.5 mm long ….……......................................…… 3 3. Upper and lower lobes of achenes in the same direction, distribution in Oman and UAE ………... ……………………................................. P. scoparium 3. Upper lobes and lower lobes not in the same direction, rotated in 120°, plants not occurring in Oman and UAE, distribution in Iran and surrounding countries .......................................... 4 4. Gemmae present on leafy branches, up to 5 mm long, formed by the remnants of dead leaves; hypocotyl elongate, 5–6 cm long, densely papillose; habitat on shale; local endemic in south-eastern Iran in Hormozgan Province, near Jakdan .......................................................................... P. jakdanense 4. Gemmae on leafy branches absent, if present short and occurring only in lower leafy branches; hypocotyl 3–4 cm long, glabrous ………………….......................................... P. aucheri subsp. ericoides 5. Achenes 9–10 × 8–9 mm; stem light brown; plant chasmophytic growing near Oak forests of Zagros mountains ….…………......……..……… P. zagricum 5. Achenes < 7.5 mm long; desert plants growing on dry river beds or gypsum slopes, never chasmophytic ….......……………………...………… 6 6. Leaves narrowly linear, leaf length more than six times as long as wide .....................................…… 7 6. Leaves usually obovate to lanceolate, leaf length less than four times as long as wide ……...…..…. 8 7. The polar length of pollen grains 23.3–25.7 µm on average, distribution in central and eastern Iran …...………………………… P. aucheri subsp. aucheri 7. The polar length of pollen grains 34.7–35.5 µm on average, distribution in south-western parts of Zagros in Iran and adjacent Iraq …………...…….....…..… P. aucheri subsp. noeanum 8.  Erect shrub, up to 1.5 m height; leaves oblanceolate, acute at apex; habitat on dry water runnels ...…………………..………. P. aucheri subsp. olivieri 8.  Small shrubs, up to 50 cm height; leaves obovate, elliptic, obtuse or round at apex; habitat on gypsum hills or saline-gypsum beds …………………………........………….……………... 9 9. Leaves 8–18 × 3–8 mm, obovate to subspathulate, not mucronate, leaves and young branches glabrous or pubescent-papillose, flowers occurring in autumn, pollen grain ornamentation suprareticulate, polar length 32.10–39.58 μm on average, habitat on gypsum hills of western Iran and adjacent Iraq …...……..…….......…………………….... P. naufelum 9. Leaves 8–12 mm × 3–4 mm, subacute to minutely mucronate, leaves and young branches always densely pubescent-papillose, flowers occurring in winter or early spring, pollen grain ornamentation punctate, polar length 28.77 μm on average, on saline-gypsum habitats near the Persian Gulf shores …….....……………….……...…. P. gypsaceum Pteropyrum aucheri Jaub. & Spach, Ill. Pl. Or. 2: 8 tab. 107 (1844) Type In collibus aridis Persiae septentrionalis, ad amnem Kizil Ouzein, Aucher 5269 (Lectotype: P-00734256!, selected here; isotypes: P-00734257, P-00734258, P-00734259. K000830354, K!, LE!). Description Shrub, 0.5–1.7 m height, richly branched, rigid, bark of young and mature branches greyish white, older branches darker, young twigs parallel striate; leaves three to seven (to 12) in each cluster, on a short elevated base, the remnants of dead leaves on a very short 0.5–1.0 (2.0) mm gemmae; leaf blades 5–16 × 1–6 mm, glabrous or pubescent, variable in shape, linear-filiform, linear subulate, oblong, lanceolate, elliptic to obovate, ± succulent, terete to oval in cross section, flat leaves slightly revolute at margin, acute to obtuse at apex, ± sessile or with a short petiole; ochrea membranous, 0.5–0.7 mm long; flowers bisexual, clustered, pedicles capillary, 2.0–5.5 mm long, jointed near the base and rarely up to the middle; perianth segments white at the margins with a green part in the middle, slightly accrescent in fruit, three internal segments 2.0–2.8 × 1.5–2.0 mm, two external segments 1.7–2.4 × 1.0–1.4 mm; stamens eight, the three longer ones opposite to the internal perianth segments 1.4–1.8 mm long, the five shorter ones alternating with tepals 1.0–1.5 mm long, filaments persistent; ovary reddish, styles three, minute, stigma capitate, reddish; achenes coriaceous, red, pale red to pinkish, 6.0–7.5 mm long, 5.0–7.0 mm wide, three-winged, wings two-parted, upper lobes are twisted clockwise or anti-clockwise, upper and lower lobes are overlapped or distant, lower lobes 1.5–2.5 mm wide, upper lobes 1.2–1.5 mm wide. Note: In Flora Iranica (Rechinger & Schiman-Czeika, 1968) P. aucheri, P. olivieri and P. noeanum are treated as separate species with the explanation of difficulties to distinguish many specimens due to overlapping characters. Pteropyrum aucheri was characterized by its linear and narrower leaves that are often six times longer than wide. In contrast, P. olivieri and P. noeanum were distinguished by their flat leaves at most six times longer than wide. Checking the type specimen of P. noeanum carefully and field studies in western Iran adjacent to Iraq border revealed that these populations have linear to linear-lanceolate leaves matching the type of P. aucheri, and therefore it was reduced as synonymy of P. aucheri (Akhani, 2004). The only cited specimen of P. noeanum (Rechinger, 9640) belongs to P. naufelum (see below). Intensive field studies over the distribution range of P. aucheri and the P. olivieri complex and study of numerous herbarium specimens integrated with our anatomical and micromorphological studies and analysis of nuclear ITS sequences convinced us that (1) distinction between P. aucheri and P. olivieri is not always possible because of intermediate traits in many cases; (2) a more-or-less geographical correlation is distinguishable with three leaf forms: populations with P. olivieri leaf types (flat-wide) distributed in central and southern parts of Iran (Fig. 103); populations with P. aucheri leaf types (flat-linear) occurring in central and eastern parts of Iran (Fig. 102) and populations with P. ericoides leaf types (cylindrical terete) distributed in eastern and south-eastern parts of Iran (Fig. 104). Leaves of populations in the western parts of Iran and adjacent Iraq are morphologically similar to P. aucheri leaf type but differ markedly in their much larger pollen grains; they are treated here as P. aucheri subsp. noeanum. All these species fall clearly into the P. aucheri clade in ITS tree (Fig. 18). The common character of most ITS sequences is the presence of several polymorphic nucleotide sites that might indicate interpopulation hybrids. This is an indication that over the distribution range of these taxa, some microspecies are interbreeding each other. We can hypothesis that these populations might have undergone several range expansion and range restriction events that have repeatedly mixed different populations during some climatic and geological times. Therefore, we decided to suggest a subspecific categorization of these species into four subspecies. Although this study covers most parts of Iran, but the lack of opportunity to study enough material and field studies in Iraq, Afghanistan and Pakistan requires reassessment of our classification. Figures 102–107. Open in new tabDownload slide Distribution maps of Pteropyrum species. Figures 102–107. Open in new tabDownload slide Distribution maps of Pteropyrum species. Pteropyrum aucheri Jaub. & Spach. subsp. aucheri (Fig. 10) Description: Leaves linear, narrowly oblanceolate, acute, 7.0–16.0 × 1.5–2.0 mm, glabrous; internal perianth segments 2.5–2.8 × 1.7–2.0 mm, external ones 2.2–2.4 × 1.3–1.5 mm; achenes, 6.0–7.0 × 5.0–5.5 mm long, upper and lower lobes are overlapped or distant, lower lobes 1.4–1.7 mm wide, upper lobes 1.0–1.4 mm wide; hypocotyl sparse papillose, ± 3.5 cm height. Habitat: Gravelly desert steppes along dry riverbeds and water runnels Distribution: Flat plains in southern slopes of Alborz mountain range, central and eastern parts of Iran and adjacent Afghanistan and Pakistan (Fig. 102). Additional examined material (Supporting Information, Appendix S2) Pteropyrum aucheri Jaub. & Spach. subsp. noeaum (Boiss.) Doostmohammadi & Akhani comb. & stat. nov. ≡ P. noeanum Boiss. ex Meisn. in DC., Prodr. 14: 31 (1856). Type: Ad ripam Dialae et Tigridis, prope Mossul, Noë 1002 (Lectotype: G-DC!, selected here; isotypes: G-BOIS.! G00330368, LE! B100294865!) Description: Leaves linear, narrowly oblanceolate, acute, 3.0– 15.0 × 1.2–2.0 mm, glabrous or pubescent; internal perianth segments 2.0–2.5 × 1.2–2.0 mm, external ones 1.8–2.2 × 1.2–1.3 mm; achenes 7.0–7.5 × 7.0–7.5 mm, lower lobes 1.5–1.8 mm wide, upper lobes 1.2–1.5 mm wide. Habitat: along dry riverbeds and adjacent low slopes. Distribution: West of Iran close to the Iraqi border and northern and eastern Iraq (Fig. 102). Additional examined material (Supporting Information, Appendix S2) Note: Pteropyrum noeanum was synonymized with P. aucheri by Akhani (2004) and P. olivieri by Edmondson & Akeroyd (2016). The pollen morphology of two specimens checked in this study is clearly larger than several examined samples from central Iran (Table 4, Fig. 92). Therefore, we decided to classify the westernmost range of P. aucheri complex as a separate subspecies, considering the fact that the leaves of some specimens tend to be closer to P. aucheri leaf type and some specimens are more similar to P. olivieri leaf type. Pteropyrum aucheri Jaub. & Spach subsp. olivieri (Jaub. & Spach) Doostmohammadi & Akhani comb. nov. (Fig. 11). ≡ Pteropyrum olivieri Jaub. & Spach, Ill. Pl. Or. 2: 9 tab. 108 (1844). Type: In ‘Persia septentrionali, circa urben Teheran’, Olivier & Bruguiere (Lectotype: P00734250, selected here; isotype: P00734252). = P. gracile Boiss., Diagn. Pl. Or. Nov. Ser. 1, 12. 102 (1853). ≡ P. olivieri Jaub. & Spach var. gracile (Boiss.) Boiss. Fl. Or. 4: 1002 (1879). Type: Prope Sonak in monte Elburz, 1843, Kotschy 664 (Lectotype: G-BOIS!, selected here; isotypes: G00330367, P00734245-photograph!, LE!). = P. griffithii Meisn. in DC., Prodr. 14: 31 (1856) Type: Afghanistan: ln collibus circa Sharkabad, Griffith 721 (Lectotypus: K001097690 -photograph!, selected here; isotypes: G-BOIS, P00762383 -photograph!). Description: Leaves oblanceolate, elliptic, oblong, glabrous or densely pubescent, 6.0–16.0 × 2.5–6.0 mm, acute or rarely obtuse, summer leaves smaller; internal perianth segments 2.0–2.5 × 1.5–1.8 mm, external ones 1.7–2.0 × 1.0–1.4 mm; achenes 6–7 × 5–6 mm, lower lobes 1.5–2.0 mm and upper lobes 1.2–1.5 mm wide, wing margin entire; hypocotyl glabrous, ± 3.5 cm height. Habitat: Desert steppes, flat plains and foothills on dry water runnels. Distribution: Central and southern Iran, Afghanistan and Pakistan (Fig. 103). Additional examined material (Supporting Information, Appendix S2) Pteropyrum aucheri Jaub. & Spach subsp. ericoides (Boiss.) Doostmohammadi & Akhani comb. nov. (Figs 4, 9) ≡ P. ericoides Boiss., Fl. Or. 4: 1002 (1879). Type: Belutschia inferior, Stocks 419 (Lectotype: G-BOIS00330366, selected here). Description: Branches greyish white, milky; leaves fasciculate, linear, subterete, terete, 6–12 mm long, usually < 1 mm wide, glabrous or occasionally sparsely papillose; internal perianth segments 2.4–2.8 × 1.5–1.8 mm, external perianth segments 1.9–2.2 × 1.0–1.4 mm; achenes 6.5–7.5 × 6.0–7.0 mm, lower lobes 2.0–2.5 mm, upper lobes 1.5–1.8 mm wide, wing margin entire, denticulate or sinuate;. hypocotyl glabrous, ± 3.5 cm height; cotyledons straight. Note: Boissier (1879) described P. ericoides based on Stocks’ collection from Baluchistan (Stocks 419, G-BOIS00330366) mentioning that the sterile collection of Haussknecht (Persiae australis monte Sawers in glareosis fl. Chyrsan. JE00021268, JE00021269) from south-western Iran is also the same. The leaves of Haussknecht’s collection are not completely cylindrical and its distribution in the Khersan Valley in western Iran is far from the range of P. aucheri subsp. ericoides (Fig. 104). We studied pollen morphology of another specimen from the Khersan Valley (10142) that confirms this population belongs to P. aucheri subsp. noeanum. Pteropyrum aucheri subsp. ericoides is the most xerophytic taxon of the P. aucheri complex distributed in desertic parts of central and eastern Iran and adjacent parts of Afghanistan, Pakistan and Turkmenistan (Fig. 104). It is characterized by usually linear and terete leaves. However, in many cases as in the type specimen the leaves are linear with revolute margins that cause difficulties in distinguishing it from P. aucheri subsp. aucheri. The species protologue of Boissier (1879) described the species with densely pubescent leaves. This character is not constant in this subspecies. Habitat: Widespread in desert areas of Iran and surrounding area in Afghanistan and Pakistan, commonly grow on temporary water runnels, flood contexts and sometimes on sand hills. Distribution: Central and eastern parts of Iran, Afghanistan, Pakistan and Turkmenistan (Fig. 104). Additional examined material (Supporting Information, Appendix S2) Pteropyrum naufelum Al-Khayat, Nordic. J. Bot. 13 (1): 33 (1993) (Figs 1, 16). Type: Iraq: Hashima, on the Iraqi-Persian border, near Badra, Muthanna Province, dry stony hill, 2 Dec. 1962; A. D. Q. Agnew, W. El-Hashimi & S. Haddad (Holotype: BUH; isotype: BAG, K). Description: Small shrubs, up to 50 cm height, stems blackish grey; leaves leathery, obovate, obtuse, rounded at apex, mostly densely papillose, sometimes glabrous, 8–18 × 3–8 mm, two to seven in each cluster; ochrea short, 1.0–1.5 mm long, hyaline; flowers axillary, pedicles capillary, minutely papillose, 4–8 mm long, jointed above the base; perianth segments 2.2–2.8 × 1.2–1.6 mm; achene coriaceous, red becoming brown when ripe, 6–7 × 6–7 mm, three-winged, wings two-parted, upper lobes twisted clockwise or anti-clockwise, lower lobes 1.2–2.0 mm, upper lobes 1.0–1.4 mm wide; hypocotyl sparsely papillose, relatively short, ± 2.5 cm height. Habitat: gypsum hills (see Akhani, 2004). Distribution: Western and south-western Iran in Khuzestan, Ilam and Lorestan Provinces and adjacent gypsum hills in Iraq (Fig. 105). Additional examined material (Supporting Information, Appendix S2) Note: Pteropyrum naufelum is a typical gypsophyte frequently growing on low gypsum hills along the Iran–Iraq border; its distribution extends further eastwards in Lorestan and southwards in Khuzestan Provinces in Iran. On gypsum hills between Mehran and Dehloran it forms a community with many gypsophilous plants, including Euphorbia acanthodes Akhani, Gypsophila linearifolia (Fisch. & C.A.Mey.) Boiss., Diplotaxis harra (Forssk.) Boiss., Albraunia fugax (Boiss. & Noë) Speta, Crepis aspera L., Onobrychis gypsicola Rech.f., Scabiosa leucactis Patzak, Cleome glaucescens DC., Convolvulus gonocladus Boiss. Astrgalus akhanii Podlech and Rumex ephedroides Bornm. (Akhani, 2004). The field observations in this area suggest possible hybridization of P. naufelum growing on the hills and P. aucheri subsp. noeanum growing along the rivers. Pteropyrum scoparium Jaub. & Spach, Ill. Pl. Orient 2. 10 tab: 109 (Figs 5, 14) Type: In Arabiae regno Mascate, Aucher 5270 (Lectotype: P04927957!, left specimen, selected here; isotypes: P04927956, K000830353!, LE!). Description: Shrub, up to 170 cm height, bark of mature branches milky-white, smooth, much branched, young twigs striate; leaves linear, succulent, glabrous, 3.0–12.0 × 0.5–1.0 mm, terete, acutish at apex, three to ten in each cluster; ochrea short (0.8–1.2 cm), membranous; flowers bisexual; pedicles capillary, 3–5 mm long, jointed above the base; perianth segments five, petaloid, white to cream, oblong-obovate 1.5–2.5 × 1.0–1.5 mm, in fruiting time slightly accrescent; stigma capitate; achenes coriaceous, 6.0–7.5 × 6.0–7.5 mm, three-winged, wings two-parted, upper lobes not twisted, lower lobes 2.5–3.0 mm, upper lobes 1.5–2.0 mm wide; seeds rounded-cordate in outline, not sinuate at the base of rostrum. Habitat: Gravelly wadis, dry water runnels and dry riverbeds in desert areas, Acacia Mill. shrubland. Distribution: Northern Oman and UAE (Fig. 107). Additional examined material (Supporting Information, Appendix S2) Pteropyrum jakdanense Doostmohammadi sp. nov. (Figs 3, 12). Holotype: Iran: Hormozgan: Bashagard, 5 km after Jakdan towards Sardasht, 26°24’33”N, 57°47’52”E, 833 m, 16.6.2013, Doostmohammadi 4860 (Holotype: IRAN; isotype: Hb. Akhani) Diagnosis: P. jakdanense is superficially similar to P. aucheri subsp. ericoides. It differs in the presence of a well-developed leaf base branchlet up to 5 mm long (absent or short in P. aucheri), 5.2–5.6 mm long hairy hypocotyls (3.2–3.8 mm long and glabrous or sparsely pubescent in P. aucheri) and the habitat on shale. Based on the ITS tree (Fig. 18), P. jakdanense shows affinity with species of the P. naufelum clade. From P. naufelum, P. gypsaceum and P. zagricum it differs in its linear terete leaves (not flat) and from P. macrocarpum in its smaller fruits 7.0–7.5 × 6.0–7.0 mm (9.0–13.0 × 8.0–10.5 mm in P. macrocarpum). Description: Robust and rigid shrub, up to 150 cm height, branches erect, bark blackish grey, young branches whitish striate; leaves fasciculate, clustered on well-developed leaf base branchlets 2–5 mm long, linear-filiform, 4.0–10.0 × 0.4–0.7 mm, three to 12 leaves in each cluster, glabrous, terete in cross section, acute; ochrea 0.8–1.5 mm long, membranous; flowers bisexual, pedicles capillary, 3.0–5.5 mm long, jointed above the base; perianth segments 2.2–2.6 × 1.3–2.2 mm; stigmas capitate; achenes coriaceous, 7.0–7.5 × 6.0–7.0 mm, three-winged, wings two-parted, upper lobes are twisted clockwise or anti-clockwise, lower lobes 1.5–2.5 mm wide, upper lobes 1.0–1.5 mm wide; cotyledons straight. Habitat: Lowland undulating hills with shale substrate. Distribution: Endemic to southern Iran, Hormozgan Province, Bashagard Area (Fig. 107). Additional examined material (Supporting Information, Appendix S2) Etymology: The specific epithet ‘jakdanense’ refers to the geographical distribution of this species near Jakdan. Pteropyrum zagricum Doostmohammadi & Akhani sp. nov. (Figs 2, 15). = P. aucheri Jaub. & Spach subsp. caespitosum Ghahreman & Mozaffarian in Fl. Iran [Ghahreman] 21: number 2593. 2001; nom. inval. Type: Fars: c. 17 km N of Mamasani (Noorabad) on the road towards Yasuj, 30°22’05’’N, 51°28’31’’E, 1410m, 23.11.2012, F. Bagheri, T. Chatrenoor, M. Dehghgani & M. Doostmohammadi 4780 (Holotype: IRAN; isotype: Hb. Akhani). Diagnosis: Pteropyrum zagricum is unique in its chasmophytic habitat on rocky crevices of oak (Quercus brantii Lindl.) forests; the other species growing on dry riverbeds or gypsum or shale habitats. It differs from P. aucheri s.l. by its larger achenes (8.5–10.0 × 7.5–9.0 mm vs. 6.0–7.5 × 5.0–7.0 mm). In the ITS tree P. zagricum belongs to the P. naufelum clade (Fig. 18). From P. macrocarpum and P. jakdanense it differs in its flat or semi-terete leaves (not cylindrical and terete) and from P. naufelum and P. gypsaceum in its unique habitat and different achene sizes (8.5–10.0 × 7.5–9.0 mm vs. 6.5–7.5 × 6.5–7.0 mm) and light brown bark (not blackish, greyish or whitish as in other species). Description: Prostrate-ascending shrubs, up to 50 cm height, spreading on rocks forming large mats to 2 m diameter, bark light brown; leaves fasciculate, in clusters of three to nine, dark green, variable, linear, linear-lanceolate to oblanceolate, 2.0–12.0 × 1.5–4.0 mm, pubescent, margin revolute, ±semi-terete in cross section, acute at apex; ochrea 1.0–1.5 mm long, membranous; flowers clustered, pedicle 4–8 mm long, jointed above the base; perianth segments 1.5–2.3 × 1.0–1.7 mm, achenes coriaceous, pink and becoming red when ripe, 8.5–10.0 × 7.5–9.0 mm, three-winged, wings two-parted, upper lobes are twisted clockwise or anti-clockwise, lower lobes 2–3 mm and upper lobes 1.5–2.0 mm wide; hypocotyl glabrous, ± 3.5 cm height; cotyledons curved and convex. Habitat: Steep limestone rocks in Quercus brantii steppe forest. Distribution: Western Iran (Fig. 106). Additional examined material (Supporting Information, Appendix S2) Etymology: The epithet ‘zagricum’ refers to the type locality of the species in Zagros Mountains Range, western Iran. Note: Pteropyrum zagricum is a unique species in its habitat on limestone cliffs and spreading habit. It shows leaf variability similar to P. aucheri in which linear to linear-lanceolate leaves occur on the same plant. The leaves of studied specimens are pubescent, despite glabrous cultivated seedlings based on one population. Whether the seeds of a plant with pubescent leaves produce glabrous plants or the indumentum develops latter in mature plants should be checked. The illegitimate taxon P. aucheri Jaub. & Spach subsp. caespitosum Ghahreman & Mozaffarian (Ghahreman 2001) was published with photographs and descriptions in Persian, French and English, but without designation of type or mention of other herbarium vouchers. The locality was given between Andimeskh (Iran: Khuzestan Prov.) and Malavi (Iran: Lorestan Prov.). Herbarium specimens cited in Flore de LʼIran are preserved in TUH. Our efforts to see a specimen of the photographed plant in TUH was not successful. The attempt by one of the authors (H.A.) to collect this plant between Andimeshk and Malavi was not successful. Pteropyrum macrocarpum Doostmohammadi & Akhani sp. nov. (Figs 6, 13). Type: Iran: Kerman Province: 42 km on the road from Kahnuj towards Rudan, 27°33’44’’N, 57°35’59’’E, 479 m, 16.3.2013, Doostmohammadi 4722 (Holotype: IRAN; isotype: Hb. Akhani). Diagnosis: P. macrocarpum is superficially similar to P. aucheri subsp. ericoides but differs clearly in much larger achenes (9.0–12.0 × 8.0–10.5 mm vs. 7.0–7.5 × 6.5–7.0 mm) and densely pubescent leaves (not glabrous or sparse pubescent) and presence of short leaf base branchlet to 4 mm. In the ITS tree (Fig. 18) it belongs to the P. naufelum clade. This species differs from P. naufelum, P. gypsaceum and P. zagricum in its cylindrical and terete leaves (not flat) and from P. jakdanense with larger achenes (9.0–12.0 × 8.0–10.5 mm vs. 7.0–7.5 × 6.0–7.0 mm) and its larger pollen grains (Fig. 94). Description: Small shrubs, with rigid erect branches, up to 100 cm height, barks grey, young twigs parallel striate; leaves fasciculate, on well-developed leaf base branchlets up to 4 mm long, three to 12 in each cluster, linear, 3.0–11.0 × 0.8–1.0 mm, densely pubescent, terete in cross section, acute at apex; ochrea 0.5–1.5 mm long, membranous; flowers bisexual, pedicles 4–6 mm long, papillose, jointed above the base; perianth segments 2.8–3.0 × 1.8–2.2 mm; achenes coriaceous, 9.0–13.0 × 8.0–10.5 mm, three-winged, wings two-parted, lower lobes almost completely overlap the upper ones and reach the fruit apex, upper lobes are twisted clockwise or anti-clockwise, lower lobes 3–4 mm, upper lobes 2.0–3.5 mm wide; hypocotyl glabrous, ± 4 cm height. Habitat: along water runnels and gravelly alluvial plains associated with Ziziphus spina-christii (L.) Desf. and Calligonum denticulatum Bunge ex Boiss., C. bungei Boiss. and C. comosum L’Hér. Distribution: South-eastern Iran; Kerman, Hormozgan and Sistano va Baluchestan Provinces (Fig. 106). Additional examined material (Supporting Information, Appendix S2) Etymology: The epithet name of ‘macrocarpum’ refers to large fruits of this remarkable species. Pteropyrum gypsaceum Akhani & Doostmohammadi sp. nov. (Fig. 17). Type: Hormozgan: 56 km E of Bandare-Lengeh towards Bandare-Khamir, 26°47’17’’N, 55°19’42’’E, gypsum hills near the coasts, 5 m, 20.12.2001, H. Akhani 16001 (Holotype: IRAN; isotype: Hb. Akhani). Diagnosis: Pteropyrum gypsaceum is superficially similar to P. aucheri subsp. olivieri but differs in the presence of a well-developed leaf base branchlet up to 5 mm (absent or short in P. aucheri s.l.), and its smaller leaves (8.0–12.0 × 3.0–4.0 mm vs. 6.0–16.0 × 2.5–6.0 mm). Based on the ITS tree it has a close affinity with P. naufelum but shows some differences in leaf sizes (8–12 × 3–4 mm vs. 8–18 × 3–8 mm) and leaf apex (mucronate vs. not mucronate) and also micromorphological characters such as the polar length of pollen grain (average of 28.77 μm vs. 36.30 μm in P. naufelum; Fig. 94) and ornamentation (punctate, not obscurely reticulate as in P. naufelum). From P. macrocarpum and P. jakdanense it differs in its flat leaves (not linear terete) and from P. zagricum in achene size (7.0–7.5 × 6.5–7.0 mm vs. 8.5–10.0 × 7.5–9.0 mm). Description: Dwarf shrub, up to 40 cm height, spreading-ascendant, branches light grey; leaves fasciculate, clusters on a small branchlet up to 5 mm long, two to seven in each cluster, elliptic, 8–12 × 3–4 mm wide, mucronate at apex, densely pubescent-papillose; ochrea 0.5–1.0 mm long, membranous; flowers bisexual, pedicles capillary, 4.5–5.5 mm long, jointed above the base; perianth segments 1.5–2.5 × 1.0–1.5 mm; achenes coriaceous, 6.5–7.5 mm long, 6–7 mm wide, three-winged; wings two-parted, upper lobes twisted clockwise or anti-clockwise, lower lobes 1.5–2.5 mm, upper lobes 1–2 mm wide; hypocotyl sparsely papillose, ±4 cm height; cotyledons straight. Habitat: A halo-gypsophytic shrub confined to gypsum-saline hills near the Persian Gulf coasts. Distribution: Southern Iran; Hormozgan Province along the Persian Gulf coasts (Fig. 105). Additional examined material (Supporting Information, Appendix S2) Etymology: The epithet ‘gypsaceum’ refers to the habitat of the species mostly on gypsum soils. Doubtful taxa and specimens from Afghanistan and Pakistan Pteropyrum olivieri Jaub. & Spach var. scabrida Gilli, Feddes Repert. 68: 242 (1963) Type: Bei Kabul, Tangi Gharu, Felsen und Schutthalden oberhalb des Wasserfalles Maipar, 1540 m, 11. V. 1951, fol. 2215, A. Gilli 2215 (W). We have only seen a photograph of this taxon available at http://herbarium.univie.ac.at/database/detail.php?ID=1160180. It superficially resembles P. naufelum and other related species of the P. naufelum clade described from western and southern Iran. We hypothesized that these specimens and many other immature plants seen in different herbaria from eastern Afghanistan and adjacent Pakistan (particularly from Quetta) represent a distinct species. To clarify its identity, recent collections will need to be sequenced. Here, we list a number of specimens that should be further reconsidered. Examined specimens studied based on virtual herbarium images are indicated by an asterisks (*). Afghanistan: 5 km W of Sarobi, gravelled plains and slopes, 34°36’35”N, 69°42’32”E, 1200 m, 7.11.1967, Toncev 4066 (W*); Sarobi, 25 km SW at road to Lataband pass, 34°28’58”N, 69°37’35”E, 1300 m, 16.6.1967, on vertical rocks of conglomerates, Freitag 6174 (W*); Kabul: c. 5 km W Sarobi, 1200 m, 24.10.1970, Podlech 20018 (W*); Hills 15 km SW Kandahar, desertic to semi-desertic hill sides, 31°33’44”N, 65°31’04”E, 1100 m, 22.6.1967, Freitag 6340 (W*); Kabul: In the mid of Tange-Gharu, 2 km to Mahipar, 1250 m, 22.5.1970, Anders 3718 (W*); Between Sarobi and Jalalabad, 900 m, 26.3.1967, Breckle 6580 (W*); Nangarha: between and Jalalabad, 820 m, 8.10.1969, Sharifi 449 (W*). Pakistan: Quetta: Murgha Kibzai to Fort Sandeman, 12–25 km from Murgha Kibazai, c. 1600 m, 19.5.1965, J. Lamond 1434 (E); Quetta: Mekhtar to Murgha Kibazi, stony and sandy plain, c. 1500 m, 18.5.1965, J. Lamond 1418 (E00191345); Quetta, Murdar Range, 14.9.1942, J. Sinclair 2465 (E); Ornach cross, c. 30 miles from Wadh on road to Karachi, 30.4.85, Abdolghafoor & Rizwan Yousuf 983 (B). Furthermore, delimitation of specimens of P. aucheri s.l. collected from Baluchistan and Sind provinces in western Pakistan are also controversial. They appear to have an intermediate leaf shape between P. aucheri subsp. aucheri and P. aucheri subsp. olivieri for which their identification requires additional studies: Pakistan: Baluchistan, 47 km from Awaran on way to Jhal Ihao and Las-Bella, sedimentary rocks, erect shrub, c. 40 cm tall, flower white; fruit winged, common, 23.3.1990, A. Ghafoor & S. M. Goodman 4862 (E); Karachi: Khadeji, c. 46 km ENE of Karachi, rocks at side of river, 29.4.1965, J. Lamond 759 (E); Baluchistan: Kalat; way to Bela, variations in colour, 1100–1250 m, 2.4.1965, J. Lamond 216 (E); Baluchistan: Kalat; Jhal Jhao to Awaran, roadside, plant of mountain areas of loose shale, 10.4.1965, J. Lamond 331, 323, 339 (E); Kachal, c. 5 miles from Naigaj to Karach Mountain, 8.11.1982, K. A. Malik, M. Qaiser, S. Omar & G. Khan. 2138 (B); 4 miles from Hubchauki on way to Bande-Murad, 24.3.1973, Sultan Alabedin et al. 9809 (B). SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher's web-site: Appendix S1. Measurements of stem, leaf and cotyledon leaf anatomical characters for Calligonum, Pteropyrum and Atraphaxis. PCA – primary carbon assimilation, M –mesophyll, PCR – photosynthetic carbon reduction, BS – bundle sheath, EP – epidermis, MCL – mesophyll cell layers, HY – hypodermis, AB – abaxial, AD – adaxial. Appendix S2. List of additional examined specimens: Examined specimens studied based on virtual herbarium images are indicated by an asterisks (*). ACKNOWLEDGEMENTS This study is the result of a long-term research started in 1990 by H. Akhani. M. Doostmohammadi continued the research as his Master’s thesis. The field and herbarium visits have been supported through different projects and organizations. The expeditions supported by the Department of Environment (Natural History Museum of Iran) and Research Council University of Tehran. The herbarium visits to B, E, G, LE, P and W were supported by German Academic Exchange Service (DAAD), Sibbald Trust (Edinburgh Botanical Garden), Alexander von Humboldt Foundation and University of Tehran. Some of the sequences granted by the Botanical Garden and Botanical Museum, Berlin, Dahlem with great help of Prof. Thomas Borsch. The stable isotope analyses performed at the ISO Analytical where funded by LIA HAOMA ‘Human adaptation to environmental constraints on the Iranian plateau since the late glacial’ project of the CNRS (Centre National de REcherche Scientifique) in France. We appreciate the Central Laboratory of College of Science for SEM and ultramicrotome facilities. The improving suggestions by Alexander Suchorukov (Moscow) and an anonymous reviewer are much appreciated. REFERENCES Akhani H . 2004 . A new spiny, cushion-like Euphorbia (Euphorbiaceae) from south-west Iran with special reference to the phytogeographic importance of local endemic species . Botanical Journal of the Linnean Society 146 : 107 – 121 . Google Scholar Crossref Search ADS WorldCat Akhani H , Barroca J , Koteeva N , Voznesenskaya E , Franceschi V , Edwards G , Ghaffari SM , Ziegler H . 2005 . Bienertia sinuspersici (Chenopodiaceae): a new species from southwest Asia, and discovery of a third terrestrial C4 plant without Kranz anatomy . Systematic Botany 30 : 290 – 330 . Google Scholar Crossref Search ADS WorldCat Akhani H , Chatrenoor T , Dehghani M , Khoshravesh R , Mahdavi P , Matinzadeh Z . 2012 . A new species of Bienertia (Chenopodiaceae) from Iranian salt deserts: a third species of the genus and discovery of a fourth terrestrial C4 plant without Kranz anatomy . Plant Biosystems 146 : 550 – 559 . OpenURL Placeholder Text WorldCat Akhani H , Ghasemkhani M . 2007 . Diversity of photosynthetic organs in Chenopodiaceae from Golestan National Park NE Iran based on carbon isotope composition and anatomy of leaves and cotyledones . Nova Hedwigia 131 : 256 – 277 . OpenURL Placeholder Text WorldCat Akhani H , Khoshravesh R . 2013 . The relationship and different C4 Kranz anatomy of Bassia eriantha and Bassia eriophora, two often confused Irano – Turanian and Saharo-Sindian species . Phytotaxa 93 : 1 – 24 . Google Scholar Crossref Search ADS WorldCat Al-Khayat AH . 1993 . Pteropyrum naufelum (Polygonaceae), a new species from Iraq . Nordic Journal of Botany 13 : 33 – 35 . Google Scholar Crossref Search ADS WorldCat Bickford D , Lohman D , Sodhi N , Ng P , Meier R , Winter K , Ingram K , Das I . 2006 . Cryptic species as a window on diversity and conservation . Trends in Ecology and Evolution 22 : 148 – 155 . Google Scholar Crossref Search ADS PubMed WorldCat Boissier E . 1853 . Diagnoses plantarum orientalium novarum . Geneva : Typographia Ferd. Ramboz . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Boissier PE . 1879 . Flora Orientalis 4 . Basel, Geneva : H. Georg . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Brandbyge J . 1993 . Polygonaceae. In: Kubitzki K , Rohver J , Bittrich V , eds. The families and genera of vascular plants 2 . Berlin : Springer-Verlag , 531 – 544 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Breckle SW , Hedge IC , Rafiqpoor MD . 2013 . Vascular plants of Afghanistan: an augmented checklist. Bonn : Scientia Bonnensis . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Carolin R , Jacobs S , Vesk M . 1975 . Leaf structure in Chenopodiaceae . Botanische Jahrbücher fur Systematik, Pflanzengeschichte und Pflanzengeographie 95 : 226 – 255 . OpenURL Placeholder Text WorldCat Christin PA , Osborne CP , Chatelet DS , Columbus JT , Besnard G , Hodkinson TR , Garrison LM , Vorontsova MS , Edwards EJ . 2012 . Anatomical enablers and the evolution of C4 photosynthesis in grasses . Proceedings of the National Academy of Sciences of the United States of America 110 : 1381 – 1386 . Google Scholar Crossref Search ADS PubMed WorldCat Cuénoud P , Savolainen V , Chatrou LW , Powell M , Grayer RJ , Chase MW . 2002 . Molecular phylogenetics of Caryophyllales based on nuclear 18S rDNA and plastid rbcL, atpB, and matK DNA sequences . American Jourbal of Botany 89 : 132 – 144 . Google Scholar Crossref Search ADS WorldCat Cullen J . 1967 . Pteropyrum. In: Davis P , ed. Flora of Turkey and the east Aegean islands 2 . Edinburgh : Edinburgh University Press , 268 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Darriba D , Taboada GL , Doallo R , Posada D . 2012 . jModelTest 2: more models, new heuristics and parallel computing . Nature Methods 9 : 772 . Google Scholar Crossref Search ADS PubMed WorldCat Davies AH . 1999 . General preparation of material and staining of sections. In: Nasser Hajibagheri MA , ed. Electro microscopy methods and protocols . Totowa : Humana Press , 1 – 12 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Dehghani M , Akhani H . 2009 . Pollen morphological studies in subfamily Suaedoideae (Chenopodiaceae) . Grana 48 : 70 – 101 . Google Scholar Crossref Search ADS WorldCat Djamali M , Akhani H , Khoshravesh R , Andrieu-Ponel V , Ponel V , Brewer S . 2011 . Application of the global bioclimatic classifcation to Iran: implications for understanding the modern vegetation and biogeography . Ecologia Mediterranea 37 : 91 – 114 . OpenURL Placeholder Text WorldCat Djamali M , de Beaulieu J-L , Shah-Hosseini M , Andrieu-Ponel V , Amini A , Akhani H, Leroy SAG , Stevens L , Alizadeh H , Ponel P , Brewer S . 2008 . A late Pleistocene long pollen record from Lake Urmia, NW Iran . Quaternary Research 69 : 413 – 420 . Google Scholar Crossref Search ADS WorldCat Doyle JJ , Doyle JL . 1987 . A rapid DNA isolation procedure for small quantities of fresh leaf tissue . Phytochemical Bulletin, Botanical Society of America 19 : 11 – 15 . OpenURL Placeholder Text WorldCat Edgar RC . 2004 . MUSCLE: multiple sequence alignment with high accuracy and high throughput . Nucleic Acid Research 32 : 1792 – 1797 . Google Scholar Crossref Search ADS WorldCat Edmondson JR , Akeroyd JR . 2016 . Polygonaceae. In: Ghazanfar SA , Edmondson JR , eds. Flora of Iraq , Vol. 5 . Tehran : Iraq Ministry of Agriculture ; Kew : Royal Botanic Gardens , 124 – 163 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Edwards G , Voznesenskaya E . 2011 . C4 Photosynthesis: Kranz forms and single-cell C4 in terrestrial plants. In: Raghavendra A , Sage R , eds. C4 Photosynthesis and related CO2 concentrating mechanisms. Advances in photosynthesis and respiration , Vol. 32 . Heidelberg-Berlin : Springer Verlag , 29 – 61 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Erdtman G . 1960 . The acetolysis method. A revised description . Svensk Botanisk Tidskrift 54 : 135 – 141 . OpenURL Placeholder Text WorldCat Ghahreman A . 2001 . Flore de LʼIran 21, No. 2593 . Tehran : Research Institute of Forests and Rangelands . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Gilli A . 1963 . Beiträge zur Flora Afghanistans III . Repertorium novarum specierum regni vegetabilis 68 : 239 – 259 . Google Scholar Crossref Search ADS WorldCat Griffiths H . 2013 . Plant venation: from succulence to succulents . Current Biology 23 : 340 – 341 . Google Scholar Crossref Search ADS WorldCat Griffiths H , Weller G , Toy L , Dennis RJ . 2013 . You’re so vein: bundle sheath physiology, phylogeny and evolution in C3 and C4 plants . Plant, Cell and Environment 36 : 249 – 261 . Google Scholar Crossref Search ADS WorldCat Haraldson K . 1978 . Anatomy and taxonomy in Polygonaceae subfam. Polygonoideae Miesn. emend. Jaretzky . Symbolae Botanicae Upsalienses 22 : 1 – 95 . OpenURL Placeholder Text WorldCat Hernández-Ledesma P , Berendsohn WG , Borsch T , Mering SV , Akhani H , Arias S , Castañeda-Noa I , Eggli U , Eriksson R , Flores-Olvera H , Fuentes-Bazán S , Kadereit G , Klak C , Korotkova N , Nyffeler R , Ocampo G , Ochoterena H , Oxelman B , Rabeler RK , Sanchez A , Schlumpberger BO , Uotila P . 2015 . A taxonomic backbone for the global synthesis of species diversity in the angiosperm order Caryophyllales . Willdenowia 45 : 281 – 383 . Google Scholar Crossref Search ADS WorldCat Hong SP . 1995 . Pollen morphology of Parapteropyrum and some putatively related genera (Polygonaceae – Atraphaxideae) . Grana 34 : 153 – 159 . Google Scholar Crossref Search ADS WorldCat Hong SP , Ronse-Decraene LP , Smets E . 1998 . Systematic significance of tepal surface morphology in tribes Persicarieae and Polygoneae (Polygonaceae) . Botanical Journal of the Linnean Society 127 : 91 – 116 . Google Scholar Crossref Search ADS WorldCat Jacobs S . 2001 . Review of leaf anatomy and ultrastructure in the Chenopodiaceae (Caryophyllales) . The Journal of the Torrey Botanical Society 128 : 236 – 253 . Google Scholar Crossref Search ADS WorldCat Jaubert C , Spach E . 1846 . Illustrationes plantarum orientalium 2 . Paris : Roret , 7 – 10 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kadereit G , Ackerly D , Pirie MD . 2012 . A broader model for C4 photosynthesis evolution in plants inferred from the goosefoot family (Chenopodiaceae s.s.) . Proceedings of the Royal Society of London B 279 : 3304 – 3311 . Google Scholar Crossref Search ADS WorldCat Leeuwen P van , Punt W , Hoen PP . 1988 . The Northwest European pollen flora, 43: Polygonaceae . Review of Palaeobotany and Palynology 57 : 81 – 151 Google Scholar Crossref Search ADS WorldCat Khoshravesh R , Akhani H , Sage TL , Nordenstam B , Sage RF . 2012 . Phylogeny and photosynthetic pathway distribution in Anticharis Endl. (Scrophulariaceae) . Journal of Experimental Botany 65 : 5645 – 5658 . Google Scholar Crossref Search ADS WorldCat McKown AD , Moncalvo JM , Dengler NG . 2005 . Phylogeny of Flaveria (Asteraceae) and inference of C4 photosynthesis evolution . American Journal of Botany 92 : 1911 – 1928 . Google Scholar Crossref Search ADS PubMed WorldCat Meisner C . 1856 . Polygonaceae . In: De Candolle A , ed. Prodromus systematis naturalis , Vol. 14 . Paris : Victor Masson , 30 – 31 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Mondal M . 1997 . Pollen morphology and systematic relationship of the family Polygonaceae . Calcutta : Botanical Survey of India . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Muhaidat R , Sage RF , Dengler NG . 2007 . Diversity of Kranz anatomy and biochemistry in C4 eudicots . American Journal of Botany 94 : 362 – 381 . Google Scholar Crossref Search ADS PubMed WorldCat Muhaidat R , Sage TL , Frohlich MW , Dengler NG , Sage RF . 2011 . Characterization of C3-C4 intermediate species in the genus Heliotropium L. (Boraginaceae): anatomy, ultrastructure and enzyme activity . Plant, Cell and Environment 34 : 1723 – 1736 . Google Scholar Crossref Search ADS WorldCat Müller K , Müller J , Quandt D . 2010 . PhyDE®: phylogenetic data editor, version 0.9971 . Available at: http://www.phyde.de/index.html. Nyberg JA , Miller AG . 1996 . Polygonaceae. In: Miller AG , Cope TA , eds. Flora of the Arabian Peninsula and Socotra , 127 – 143 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Nyffeler R , Eggli U , Ogburn M , Edwards E . 2008 . Variations on a theme: repeated evolution of succulent life forms in the Portulacineae (Caryophyllales) . Haseltonia 14 : 26 – 36 . Google Scholar Crossref Search ADS WorldCat Ogburn RM , Edwards EJ . 2013 . Repeated origin of three-dimensional leaf venation releases constraints on the evolution of succulence in plants . Current Biology 23 : 722 – 726 . Google Scholar Crossref Search ADS PubMed WorldCat Osmond CB , Ziegler H , Stichler W , Trimborn P . 1975 . Carbon isotope discrimination in alpine succulent plants supposed to be capable of crassulacean acid metabolism (CAM) . Oecologia 18 : 226 – 231 . Google Scholar Crossref Search ADS WorldCat Punt W , Hoen PP , Blackmore S , Nilsson S , Le Thomas A . 2007 . Glossary of pollen and spore terminology . Review of Palaeobotany and Palynology 143 : 1 – 81 . Google Scholar Crossref Search ADS WorldCat Pyankov V , Black C , Artyusheva E , Voznesenskaya E , Ku M , Edwards EJ . 1999 . Features of photosynthesis in Haloxylon species of Chenopodiaceae that are dominant plants in Central Asian deserts . Plant and Cell Physiology 40 : 125 – 134 . Google Scholar Crossref Search ADS WorldCat Pyankov V , Voznesenskaya E , Kuzmin A , Ku M , Ganko E , Franceschi V , Black C , Edwards G . 2000 . Occurrence of C3 and C4 photosynthesis in cotyledons and leaves of Salsola species (Chenopodiaceae) . Photosynthesis Research 63 : 69 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat Rasband WS . 1997–2018 . ImageJ . Bethesda : U.S. National Institutes of Health . imagej.nih.gov/ij. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Rechinger K , Schiman-Czeika H . 1968 . Polygonaceae. In: Rechinger K , ed. Flora Iranica 56 . Graz : Akademische Druck und Verlagsanstalt , 1 – 158 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ronse-Decraene L , Akeroyd J . 1988 . Generic limits in Polygonum and related genera (Polygonaceae) on the basis of floral characters . Botanical Journal of the Linnean Society 98 : 321 – 371 . Google Scholar Crossref Search ADS WorldCat Ronquist F , Teslenko M , Van der Mark P , Ayres DL , Darling A , Höhna S , Larget B , Liu L , Suchard MA , Huelsenbeck JP . 2012 . MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space . Systematic Biology 61 : 539 – 542 . Google Scholar Crossref Search ADS PubMed WorldCat Sage RF . 2001 . Environmental and evolutionary preconditions for the origin and diversification of the C4 photosynthetic syndrome . Plant Biology 3 : 202 – 213 . Google Scholar Crossref Search ADS WorldCat Sage RF , Christin P , Edwards E . 2011 . The C4 plant lineages of planet Earth . Journal of Experimental Botany 62 : 3155 – 3169 . Google Scholar Crossref Search ADS PubMed WorldCat Sage RF , Khoshravesh R , Sage TL . 2014 . From proto-Kranz to C4 Kranz: building the bridge to C4 photosynthesis . Journal of Experimental Botany 65 : 3341 – 3356 . Google Scholar Crossref Search ADS PubMed WorldCat Sage TL , Busch F , Johnson D . 2013 . Initial events during the evolution of C4 photosynthesis in C3 species of Flaveria . Plant Physiology 163 : 1266 – 1276 . Google Scholar Crossref Search ADS PubMed WorldCat Sanchez A , Schuster T , Burke J , Kron K . 2011 . Taxonomy of Polygonoideae (Polygonaceae): a new tribal classification . Taxon 60 : 151 – 160 . Google Scholar Crossref Search ADS WorldCat Sanchez A , Schuster T , Kron K . 2009 . A large-scale phylogeny of Polygonaceae based on molecular data . International Journal of Plant Sciences 170 : 1044 – 1055 . Google Scholar Crossref Search ADS WorldCat Schuster TM , Reveal JL , Bayly MJ , Kron KA . 2015 . An updated molecular phylogeny of Polygonoideae (Polygonaceae): relationships of Oxygonum, Pteroxygonum, and Rumex, and a new circumscription of Koenigia . Taxon 64 : 1188 – 1208 . Google Scholar Crossref Search ADS WorldCat Schütze P , Freitag H , Weising K . 2003 . An integrated molecular and morphological study of the subfamily Suaedoideae Ulbr. (Chenopodiaceae) . Plant Systematics and Evolution 239 : 257 – 286 . Google Scholar Crossref Search ADS WorldCat Shneyer VS , Kotseruba VV . 2015 . Cryptic species in plants and their detection by genetic differentiation between populations . Russian Journal of Genetics: Applied Research 5 : 528 – 542 . Google Scholar Crossref Search ADS WorldCat Stamatakis A . 2014 . RAxML Version 8: a tool for phylogenetic analysis and postanalysis of large phylogenies . Bioinformatics 30 : 1312 – 1313 . Google Scholar Crossref Search ADS PubMed WorldCat Stöver BC , Müller KF . 2010 . TreeGraph 2: combining and visualizing evidence from different phylogenetic analyses . BMC Bioinformatics 11 : 7 . Google Scholar Crossref Search ADS PubMed WorldCat Swofford DL . 1991 . PAUP: phylogenetic analysis using parsimony, version 3.1 . Computer program distributed by the Illinois Natural History Survey, Champaign, Illinois . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tavakkoli S , Kazempour S , Maassoumi A . 2008 . Morphological cladistic analysis of Calligonum and Pteropyrum (Polygonaceae) in Iran . Iranian Jouranl of Botany 14 : 117 – 125 . OpenURL Placeholder Text WorldCat Tavakkoli S , Kazempour S , Maassoumi A . 2010 . The phylogeny of Calligonum and Pteropyrum (Polygonaceae) based on nuclear ribosomal DNA ITS and chloroplast trnL-F sequences . Iranian Journal of Biotechnology 8 : 7 – 15 . OpenURL Placeholder Text WorldCat Thiers B . 2018 . continuously updated. Index herbariorum: a global directory of public herbaria and associated staff, New York Botanical Garden’s Virtual Herbarium . Available at: http://sweetgum.nybg.org/ih/. Voznesenskaya E , Koteyeva N , Akhani H , Roalson EH , Edwards G . 2013 . Structural and physiological analyses in Salsoleae (Chenopodiaceae) indicate multiple transitions among C3, intermediate, and C4 photosynthesis . Journal of Experimental of Botany 64 : 3583 – 3604 . Google Scholar Crossref Search ADS WorldCat White TJ , Bruns T , Lee S , Taylor J . 1990 . Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA , Gelfand DH , Sninsky JJ , White TJ , eds. PCR protocols: a guide to methods and applications . San Diego : Academic Press 18 : 315 – 322 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Winter K . 1981 . C4 plants of high biomass in arid regions of Asia. Occurrence of C4 photosynthesis in Chenopodiaceae and Polygonaceae from the Middle East and USSR . Oecologia 48 : 100 – 106 . Google Scholar Crossref Search ADS PubMed WorldCat Winter K , Kramer D , Troughton J , Card K , Fischer K . 1977 . C4 pathway of photosynthesis in a member of the Polygonaceae: Calligonum persicum (Boiss. & Buhse) Boiss . Pflanzenphysiologie 81 : 341 – 346 . Google Scholar Crossref Search ADS WorldCat APPENDIX 1. GENBANK ACCESSION NUMBER OF ITS SEQUENCES STUDIED IN THIS PAPER. DETAILS OF HERBARIUM VOUCHERS OF ALL NEW SEQUENCES ARE AVAILABLE IN THE TAXONOMIC ENUMERATION. Taxon . Voucher collector(s) and herbarium number . GenBank accession numbers . Pteropyrum aucheri subsp. aucheri M. Doostmohammadi 4832 MN437492 Pteropyrum aucheri subsp. aucheri H. Akhani 24127 MN437491 Pteropyrum aucheri subsp. noeanum H. Akhani 8397 MN437490 Pteropyrum aucheri subsp. ericoides H. Akhani et al. 23487 MN437495 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4679 MN437493 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4826 MN437494 Pteropyrum aucheri subsp. olivieri H. Akhani 24224 MN437502 Pteropyrum aucheri subsp. olivieri H. Doostmohammadi 4852 MN437503 Pteropyrum aucheri subsp. olivieri H. Akhani 7941 MN437504 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4673 MN437505 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4730 MN437506 Pteropyrum aucheri subsp. olivieri V. Mozaffarian 49456 MN437507 Pteropyrum gypsaceum H. Akhani 16001 MN437509 Pteropyrum jakdanense M. Doostmohammadi 4861 MN437498 Pteropyrum macrocarpum M. Doostmohammadi 4722 MN437496 Pteropyrum macrocarpum M. Doostmohammadi 4729 MN437497 Pteropyrum naufelum M. Doostmohammadi & A. Noormohammadi 4875 MN437500 Pteropyrum naufelum H. Akhani 21580 MN437499 Pteropyrum naufelum H. Akhani 9064-b MN437501 Pteropyrum scoparium Kürschner 99-33 MN437508 Pteropyrum zagricum F. Bagheri et al. 4780 MN437510 Calligonum polygonoides Tavakoli et al. 2010 AB542779 Calligonum persicum Tavakoli et al. 2010 AB542777 Atraphaxis spinosa Tavakoli et al. 2010 AB542772 Atraphaxis suaedifolia V. Mozaffarian 87202 AB542773 Taxon . Voucher collector(s) and herbarium number . GenBank accession numbers . Pteropyrum aucheri subsp. aucheri M. Doostmohammadi 4832 MN437492 Pteropyrum aucheri subsp. aucheri H. Akhani 24127 MN437491 Pteropyrum aucheri subsp. noeanum H. Akhani 8397 MN437490 Pteropyrum aucheri subsp. ericoides H. Akhani et al. 23487 MN437495 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4679 MN437493 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4826 MN437494 Pteropyrum aucheri subsp. olivieri H. Akhani 24224 MN437502 Pteropyrum aucheri subsp. olivieri H. Doostmohammadi 4852 MN437503 Pteropyrum aucheri subsp. olivieri H. Akhani 7941 MN437504 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4673 MN437505 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4730 MN437506 Pteropyrum aucheri subsp. olivieri V. Mozaffarian 49456 MN437507 Pteropyrum gypsaceum H. Akhani 16001 MN437509 Pteropyrum jakdanense M. Doostmohammadi 4861 MN437498 Pteropyrum macrocarpum M. Doostmohammadi 4722 MN437496 Pteropyrum macrocarpum M. Doostmohammadi 4729 MN437497 Pteropyrum naufelum M. Doostmohammadi & A. Noormohammadi 4875 MN437500 Pteropyrum naufelum H. Akhani 21580 MN437499 Pteropyrum naufelum H. Akhani 9064-b MN437501 Pteropyrum scoparium Kürschner 99-33 MN437508 Pteropyrum zagricum F. Bagheri et al. 4780 MN437510 Calligonum polygonoides Tavakoli et al. 2010 AB542779 Calligonum persicum Tavakoli et al. 2010 AB542777 Atraphaxis spinosa Tavakoli et al. 2010 AB542772 Atraphaxis suaedifolia V. Mozaffarian 87202 AB542773 Open in new tab Taxon . Voucher collector(s) and herbarium number . GenBank accession numbers . Pteropyrum aucheri subsp. aucheri M. Doostmohammadi 4832 MN437492 Pteropyrum aucheri subsp. aucheri H. Akhani 24127 MN437491 Pteropyrum aucheri subsp. noeanum H. Akhani 8397 MN437490 Pteropyrum aucheri subsp. ericoides H. Akhani et al. 23487 MN437495 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4679 MN437493 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4826 MN437494 Pteropyrum aucheri subsp. olivieri H. Akhani 24224 MN437502 Pteropyrum aucheri subsp. olivieri H. Doostmohammadi 4852 MN437503 Pteropyrum aucheri subsp. olivieri H. Akhani 7941 MN437504 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4673 MN437505 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4730 MN437506 Pteropyrum aucheri subsp. olivieri V. Mozaffarian 49456 MN437507 Pteropyrum gypsaceum H. Akhani 16001 MN437509 Pteropyrum jakdanense M. Doostmohammadi 4861 MN437498 Pteropyrum macrocarpum M. Doostmohammadi 4722 MN437496 Pteropyrum macrocarpum M. Doostmohammadi 4729 MN437497 Pteropyrum naufelum M. Doostmohammadi & A. Noormohammadi 4875 MN437500 Pteropyrum naufelum H. Akhani 21580 MN437499 Pteropyrum naufelum H. Akhani 9064-b MN437501 Pteropyrum scoparium Kürschner 99-33 MN437508 Pteropyrum zagricum F. Bagheri et al. 4780 MN437510 Calligonum polygonoides Tavakoli et al. 2010 AB542779 Calligonum persicum Tavakoli et al. 2010 AB542777 Atraphaxis spinosa Tavakoli et al. 2010 AB542772 Atraphaxis suaedifolia V. Mozaffarian 87202 AB542773 Taxon . Voucher collector(s) and herbarium number . GenBank accession numbers . Pteropyrum aucheri subsp. aucheri M. Doostmohammadi 4832 MN437492 Pteropyrum aucheri subsp. aucheri H. Akhani 24127 MN437491 Pteropyrum aucheri subsp. noeanum H. Akhani 8397 MN437490 Pteropyrum aucheri subsp. ericoides H. Akhani et al. 23487 MN437495 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4679 MN437493 Pteropyrum aucheri subsp. ericoides M. Doostmohammadi 4826 MN437494 Pteropyrum aucheri subsp. olivieri H. Akhani 24224 MN437502 Pteropyrum aucheri subsp. olivieri H. Doostmohammadi 4852 MN437503 Pteropyrum aucheri subsp. olivieri H. Akhani 7941 MN437504 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4673 MN437505 Pteropyrum aucheri subsp. olivieri M. Doostmohammadi 4730 MN437506 Pteropyrum aucheri subsp. olivieri V. Mozaffarian 49456 MN437507 Pteropyrum gypsaceum H. Akhani 16001 MN437509 Pteropyrum jakdanense M. Doostmohammadi 4861 MN437498 Pteropyrum macrocarpum M. Doostmohammadi 4722 MN437496 Pteropyrum macrocarpum M. Doostmohammadi 4729 MN437497 Pteropyrum naufelum M. Doostmohammadi & A. Noormohammadi 4875 MN437500 Pteropyrum naufelum H. Akhani 21580 MN437499 Pteropyrum naufelum H. Akhani 9064-b MN437501 Pteropyrum scoparium Kürschner 99-33 MN437508 Pteropyrum zagricum F. Bagheri et al. 4780 MN437510 Calligonum polygonoides Tavakoli et al. 2010 AB542779 Calligonum persicum Tavakoli et al. 2010 AB542777 Atraphaxis spinosa Tavakoli et al. 2010 AB542772 Atraphaxis suaedifolia V. Mozaffarian 87202 AB542773 Open in new tab © 2019 The Linnean Society of London, Botanical Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Cruz-Nicolás,, Jorge;Giles-Pérez,, Gustavo;González-Linares,, Eréndira;Múgica-Gallart,, Julia;Lira-Noriega,, Andrés;Gernandt, David, S;Eguiarte, Luis, E;Jaramillo-Correa, Juan, P

doi: 10.1093/botlinnean/boz077pmid:

Bisang,, Irene;Ehrlén,, Johan;Hedenäs,, Lars

doi: 10.1093/botlinnean/boz063pmid: N/A

Abstract Sex ratio variation is common among organisms with separate sexes. In bryophytes, sex chromosome segregation at meiosis suggests a balanced progeny sex ratio. However, most bryophyte populations exhibit female-biased phenotypic sex ratios based on the presence of reproductive structures on gametophytes. Many bryophyte populations do not form sexual organs, and genotypic sex ratio variation in such populations is mostly unknown. We tested sex expression, and phenotypic and genotypic sex ratios against environmental parameters in natural populations of the unisexual wetland moss Drepanocladus lycopodiodes at 11 sites in each of three regions in southern Sweden. We identified sex in 660 individual ramets, based on sexual structures, when present, or with a specifically designed molecular marker, when absent. All regions exhibited a female bias in phenotypic and genotypic sex ratios. Sex ratio biases and sex expression differed between regions. Sex ratios were less female-biased in larger patches. Wetter patches exhibited a stronger female bias in genotypic sex ratio and lower sex expression. This is the first evidence of environmental effects on genotypic sex ratio in mosses. A higher frequency of females in wet patches could be due to higher female resource demands for sporophyte production or higher male sensitivity to wetness. A higher incidence of females than males in moister sites aligns with female flowering plants, but differs from reproductive bryophytes in drier environments. Taken together with previous results, our data indicate that sex ratio variation and its drivers differ among species, their life histories and environments. biased genotypic sex ratios, biased phenotypic sex ratios, dioicous bryophyte, environmental effects, molecular sex identification, southern Baltic area INTRODUCTION Deviations from a balanced sex ratio are common among animal and plant species with separate sexes and genotypic sex determination (Hardy, 2002; Bisang & Hedenäs, 2005; West, 2009; Sinclair, Emlen & Freeman, 2012; Székely, Weissing & Komdeur, 2014). Sex ratio differences occur both among populations within species and among species. In plants with a sessile habit, a sex ratio bias may arise among seeds (e.g. as a consequence of sex ratio distorters or lethal maternal effects; de Jong & Klinkhamer, 2002) or establish at later stages in the life cycle. To date, differential reproductive investment between males and females is considered one of the primary drivers of biased sex ratios in adult plants (e.g. Obeso, 2002; Barrett et al., 2010; Graff, Rositano & Aguiar, 2013; Tonnabel, David & Pannell, 2017). Within species, sex-related differences in life history traits might interact with environmental factors and lead to spatial segregation of the sexes, which is well documented for flowering plants (Bierzychudek & Eckhart, 1988; Bertiller et al., 2002; Field, Pickup & Barrett, 2013a). In addition, the successful establishment of sexes at a given locality can be influenced by random processes during the recruitment phase (e.g. Field, Pickup & Barrett, 2013b). Thus, life histories, environmental conditions and stochastic processes are expected to contribute to between-population sex ratio variation for species occupying heterogeneous environments. In seed plants, dioecy is manifested in the diploid sporophyte, which is the dominant life cycle phase. Five to six per cent of angiosperm species exhibit sporophytic dioecy, and their gametophytes are always unisexual (Renner, 2014). In contrast, bryophytes (mosses, liverworts, hornworts; Wickett et al., 2014) exhibit gametophytic dioicy (for terminology of dioecy vs. dioicy and monoecy vs. monoicy, see Wyatt, 1985). The haploid, dominant and free-living gametophyte of bryophytes is either dioicous (i.e. the gametophyte individual is sexually specialized or unisexual) or monoicous (bisexual), whereas their diploid sporophytes do not specialize as females and males; they produce both male and female (or bisexual) spores (Jesson & Garnock-Jones, 2012). More than half of bryophyte species are dioicous (McDaniel & Perroud, 2012; Villarreal & Renner, 2013). The mechanism of bryophyte sex determination in the sporophyte at meiosis is expected to result in a balanced progeny sex ratio (by segregation of sex-determining loci on the heteromorphic U and V chromosomes; Bachtrog et al., 2011; McDaniel & Perroud, 2012; Charlesworth, 2015; Renner, Heinrichs & Sousa, 2017). Nevertheless, the majority of populations and species of dioicous bryophytes are characterized by a strong female bias in gametophytic sex ratios (Bisang & Hedenäs, 2005; Bisang et al., 2014). So far, bryophyte sex ratios have generally been inferred from counts of male and female individual plants with sexual organs in field patches or herbarium specimens (phenotypic sex ratios; reviewed by Bisang & Hedenäs, 2005; and, for example, Alvarenga, Pôrto & Zartman, 2013; Blackstock, 2015). However, many bryophyte populations include large numbers of individuals without sexual organs, and in some species sexual organs are unknown (Longton & Schuster, 1983; Bisang & Hedenäs, 2005). This implies that the observed phenotypic sex ratios, estimated from reproductive ramets, might be different from genotypic sex ratios (Fig. 1). Despite the early observation of skewed phenotypic sex ratios in bryophyte populations (Bedford, 1938) and increased recent attention to patterns of reproductive allocation in bryophytes (e.g. Rydgren, Halvorsen & Cronberg, 2010; Brzyski, Taylor & McLetchie, 2014; Holá et al., 2014), intraspecific sex ratio variation and its drivers remain poorly explored. For a few dioicous bryophytes, phenotypic sex ratio has been reported to vary with environmental conditions in natural habitats (e.g. Benassi et al., 2011; Blackstock, 2015; Pereira, Dambros & Zartman, 2016), whereas other studies failed to find such a relationship (e.g. Groen et al., 2010). Genotypic sex ratio variation in bryophytes has, however, rarely been addressed due to the difficulty in assessing sex in non-reproductive ramets. Figure 1. Open in new tabDownload slide Schematic representation of the analysed reproductive parameters in Drepanocladus lycopodioides. ‘Genotypic sex ratio’ is the number of genotypic male ramets irrespective of their reproductive state divided by the total number of ramets. ‘Phenotypic sex ratio’ is the number of reproductive male ramets (reproductive males) divided by the total number of reproductive ramets. ‘Reproductive males’ are male ramets bearing sexual organs (perigonia). ‘Reproductive females’ are female ramets bearing sexual organs (sporophytes and/or perichaetia). ‘Non-reproductive males’ and ‘Non-reproductive females’ are ramets without sexual organs (to the right in the figure). Figure 1. Open in new tabDownload slide Schematic representation of the analysed reproductive parameters in Drepanocladus lycopodioides. ‘Genotypic sex ratio’ is the number of genotypic male ramets irrespective of their reproductive state divided by the total number of ramets. ‘Phenotypic sex ratio’ is the number of reproductive male ramets (reproductive males) divided by the total number of reproductive ramets. ‘Reproductive males’ are male ramets bearing sexual organs (perigonia). ‘Reproductive females’ are female ramets bearing sexual organs (sporophytes and/or perichaetia). ‘Non-reproductive males’ and ‘Non-reproductive females’ are ramets without sexual organs (to the right in the figure). Studies of plant sex ratio variation focusing on dioecious seed plants identified relationships with several environmental factors (Barrett et al., 2010; Field et al., 2013a). However, plant sexual systems are highly diverse (Barrett et al., 2010; Petersen & Burd, 2018). Beyond bryophyte dioicy being expressed in the haploid gametophyte, the sexual system of their gametophytes has repeatedly changed over evolutionary time to an extent that is unparalleled in other land plant lineages (McDaniel, Atwood & Burleigh, 2013; Laenen et al., 2016; Renner et al., 2017). Data on bryophyte sex ratio variation will thus significantly complement existing information and contribute to a more general understanding of plant sex ratio biases. Only recently, sex-linked or sex-associated molecular markers were developed to identify genotypic sex in non-reproductive life cycle stages in bryophyte populations (Korpelainen et al., 2008, and references therein; Norrell et al., 2014; Bisang, Hedenäs & Cronberg, 2017). For the long-lived dioicous moss Drepanocladus lycopodioides (Brid.) Warnst., we have developed a reliable and efficient molecular method to identify sex in individuals without sexual structures, based on a female-targeting molecular marker (Korpelainen et al., 2008; Bisang, Korpelainen & Hedenäs, 2010). The consistency of the molecular marker for sex assignment was previously tested and successfully applied in D. lycopodioides and a related species (Hedenäs et al., 2010; Bisang & Hedenäs, 2013). Drepanocladus lycopodioides exhibits an overall female bias in both phenotypic and genotypic sex ratios in its European distribution range (Bisang & Hedenäs, 2013). Sporophytes are produced occasionally, and sporophyte frequency varies over time and space (Bisang et al., 2014, and unpubl. data IB, LH). Using this molecular method, we recently uncovered a balanced meiotic spore sex ratio, suggesting that the observed sex ratio bias in natural populations is explained by other factors (Bisang et al., 2017). The aim of this study was to explore the regional variation in reproductive parameters in natural populations of D. lycopodioides and to identify the drivers of this variation. We investigated phenotypic and genotypic ramet-based sex ratios and the proportion of ramets with sexual organs (Fig. 1) in 11 sites in each of three geographically separated regions in its European distribution range. We examined variation in these reproductive traits, sporophyte frequency and microhabitat diversity in terms of habitat wetness and patch size and addressed the following questions. (1) How do the proportions of male and female reproductive ramets and phenotypic and genotypic sex ratios vary between regions? Is this variation related to the level of sporophyte production in the three regions? We expect regional differences in sex ratio biases and in the proportions of reproductive ramets. We base this expectation on previous field observations of regional differences in sporophyte production, and because sporophyte formation is affected by sex expression and sex ratios (Bisang et al., 2014). (2) Are differences in sex ratios and proportions of reproductive ramets between sites related to patch size and habitat wetness? We hypothesize that larger patch size should be associated with a relatively higher frequency of the rarer sex because of increased establishment possibilities and potentially more varied microhabitats. We anticipate that moist conditions may favour females due to high resource demands for sporophyte production, including a lower risk for drought-induced sporophyte abortion. (3) Do phenotypic and genotypic sex ratios differ from each other, indicating sex-related differences in sexual organ formation? We have no evidence that the sexes of D. lycopodioides differ in their relative proportions of sex organ formation in the European distribution range. We thus expect corresponding genotypic and phenotypic sex ratios also in the current study regions. MATERIAL AND METHODS Study species, study system and terminology Bryophyte gametophytes may produce sexual organs, gametangia. These are surrounded by specialized leaves and form male perigonia or female perichaetia in mosses. Following fertilization of the sessile egg cell by a spermatozoid with limited motility (Bisang, Ehrlén & Hedenäs, 2004), a sporophyte is formed that remains attached to the gametophyte during its lifetime. Spores, either male and female spores or spores carrying both sexes, form in the terminal sporangium. They develop into male, female or bisexual gametophytes, respectively. Even in gametophytes that do not produce sexual organs (during periods or during their life span), the individual ramet carries a genotypic sex regardless of sex expression. In this paper, we term ramets without sexual organs, i.e. without sex expression, ‘non-reproductive’ rather than ‘sterile’ (Fig. 1). Most bryophytes lack evident secondary morpho-anatomical sex characteristics, which prevents sex identification without molecular methods in non-reproductive plants (Bisang & Hedenäs, 2005; Korpelainen et al., 2008). Because non-reproductive ramets are common, we need to distinguish between phenotypic and genotypic sex ratios (Fig. 1). We use ‘phenotypic sex ratio’ as the number of reproductive male ramets divided by the total number of reproductive ramets. ‘Genotypic sex ratio’ is the number of genotypic male ramets irrespective of their state of sex expression divided by the total number of ramets. The calculation of sex ratios as the proportion of male plants of the total of males plus females follows a widely applied practice in plant biology (de Jong & Klinkhamer, 2002; Barrett et al., 2010). We refer to sex ratios in adult gametophytic individuals if not otherwise specified. This is critical because sex ratio at the spore stage or in juvenile gametophytes may differ from adult sex ratios (Norrell et al., 2014; Bisang et al., 2017). The ‘proportion of reproductive males’ is the proportion of genotypic male ramets that bear sexual structures (perigonia). The ‘proportion of reproductive females’ denotes the proportion of genotypic female ramets bearing sexual structures (sporophytes and/or perichaetia) (Fig. 1). Drepanocladus lycopodioides (Amblystegiaceae; Fig. 2) is a dioicous pleurocarpous moss that grows in wet, calcareous habitats. It occurs in western Eurasia and is near-endemic to Europe (Hedenäs & Bisang, 2015). The highest abundance of the species is in south-east Sweden and neighbouring Baltic regions (core distribution area; Hedenäs & Bisang, 2015). In this, and adjacent areas, the species grows in depressions in natural landscapes or landscapes that have been extensively managed as pastures for at least 2500 years (Petterson, 1958). Outside its core distribution area, the species occurs mainly in strongly managed agricultural ecosystems, where it has significantly declined due to drainage and agricultural intensification (Šoltés, Kubinská & Janovicová, 2002; Hedenäs, Bisang & Schnyder, 2003;,Caspari et al., 2018). Phenotypic and genotypic sex ratios in the total European distribution area of the species were found to be 0.304 (N = 115) and 0.278 (N = 102), respectively, and did not differ from each other (Bisang & Hedenäs, 2013; and unpubl. data IB; sporophytic samples excluded). Overall, 51% of 265 herbarium samples bore sexual organs (i.e. they were reproductive), and 23% of the reproductive female samples (N = 24 out of 104) carried sporophytes (Bisang et al., 2014; and unpubl. data IB). Specialized vegetative diaspores are unknown in this species. Figure 2. Open in new tabDownload slide Distribution of (A) the study species Drepanocladus lycopodioides, and (B) the three study regions in the southern Swedish Baltic area in the centre of the core distribution area of the species (see Material and Methods for details on the extant and historical distribution). Photographs depict a sampling plot in a patch of D. lycopodioides in a typical habitat, a periodically wet depression on calcareous ground (lower left corner) and a number of typical ramets (upper left corner). Figure 2. Open in new tabDownload slide Distribution of (A) the study species Drepanocladus lycopodioides, and (B) the three study regions in the southern Swedish Baltic area in the centre of the core distribution area of the species (see Material and Methods for details on the extant and historical distribution). Photographs depict a sampling plot in a patch of D. lycopodioides in a typical habitat, a periodically wet depression on calcareous ground (lower left corner) and a number of typical ramets (upper left corner). The study was conducted within the core distribution area of the species, on two islands (Öland, Gotland) and an island region (the Stockholm Archipelago) in the southern Swedish Baltic area (56.5–59.6°N, 16.4–19.1°E) (Fig. 2), which will hereafter be referred to as regions. The study area has a maritime climate with an annual mean temperature of 6–7 °C (January −2 to 0 °C; July 16 °C), annual precipitation in the range 500–700 mm, with ~70% falling from July to January, and an annual vegetation period (days with mean temperatures >5 °C) of 180–210 days (http://www.smhi.se/klimatdata; accessed 4 July 2015; normal values 1961–1990). The three regions include large exposures of calcium-rich rock (limestone or marl) or thin soil that lack dense vascular plant vegetation, due to natural conditions or human management (Hedenäs & Bisang, 2015). Drepanocladus lycopodioides occurs abundantly in small wetland depressions of various depths and extensions, and with intermittent moisture availability (i.e. they dry out periodically) (Hedenäs, 2003). The depressions are situated in a matrix of dry, unsuitable habitats. In these depressions, the species forms distinct patches (see Field data collection for ‘patch’). The size of a D. lycopodioides patch is determined by small-scale topography, which reflects the local conditions rather than the age of the patch. A typical suitable depression is likely to be fully occupied by the species considerably more quickly than a typical life span of a patch. Patch size serves as a surrogate for microhabitat diversity, with larger patches likely to include more open spots and more diverse conditions, which may affect the sexes and their establishment differently. During the well-aerated summer dry periods in the study regions, organic matter decomposes and mineral soil erodes through wind and other disturbance effects (e.g. grazing cattle), implying that these depressions are semi-permanent landscape structures. The maximum depth of a depression is an appropriate proxy for overall moisture availability during the year (wetness). Patch size and wetness were not correlated with each other (Pearson’s R = 0.1969, P = 0.272, log-transformed values). The occurrence of D. lycopodioides patches in isolated depressions implies that each patch must have been colonized by at least one diaspore or fragment, respectively. Given the long distances between the sampled patches (see below, Field data collection), it is likely that colonization of these occurred by spores and not by ramet fragments. A strong genetic differentiation between D. lycopodioides patches was suggested by the fact that 62% of the total haplotype variation was between patches, although identical haplotypes across patches occurred (Hedenäs & Bisang, 2015). We have previously observed in the field that the frequency of sporophyte production seems to differ between regions. The regions thus provide an excellent study system to examine the distribution of the sexes and the variation in phenotypic and genotypic sex ratios and to test for associations with environmental factors and sporophyte formation. Field data collection Fieldwork was carried out in May and the first half of June 2011. In each of the three regions, we selected one patch of D. lycopodioides at each of 11 pre-defined sites (Supporting Information, Table S1) as described by Hedenäs & Bisang (2015). A patch was defined as a largely continuous occurrence of D. lycopodioides in a small wetland depression, where the distance between neighbouring ramets (i.e. small open spots) was at most 2 m. Drepanocladus lycopodioides patches were easy to recognize in the field. Individual patches were selected as being separated from other patches by ≥20 m. Actual measured minimum and mean inter-patch distances were much longer in most situations: 80 m and 16.7 km (SD = 10.2 km) on Gotland, 60 m and 7.4 km (13.5 km) in the Stockholm Archipelago, and 21 m and 15.9 km (13.5 km) on Öland. We sampled one plot in each patch using a metal frame of 50 × 40 cm with a centimetre-grid (Fig. 2). In each plot we sampled 20 individual ramets of D. lycopodioides as to a pre-defined scheme based on random numbers assigned to the centimetre-coordinates of the metal frame. The sampled ramets were kept in glassine bags, dried as quickly as possible and stored at room temperature until further treatment in the laboratory. Vouchers of one ramet of D. lycopodioides from each patch are kept in the bryophyte herbarium at the Swedish Museum of Natural History (S) (Table S2). We noted the geographical position of each patch to an accuracy of 5–8 m with a GPS (latitude, longitude). We estimated patch size as a surrogate for microhabitat diversity to the nearest 0.25 m2 for patches up to 1 m2, to the nearest 1 m2 for patches up to 15 m2 and to nearest 5 m2 for larger patches. As a proxy for overall patch wetness, we measured the maximum depth of the patch-containing depression. We first identified the deepest point of the patch. Over this point and along the longest extension of the depression, we spanned a horizontal line at the height of the high-water level as indicated by the vegetation at the margins of the depression. We then measured the depth to the nearest centimetre from this line to the firm ground at the deepest point, using a wooden stick and a measuring tape. Laboratory techniques and molecular analyses We examined all 660 (20 per patch, 11 patches, three regions) sampled ramets under a dissecting microscope to assess the occurrence of perigonia, perichaetia and sporophytes. Sexual organ development follows a specific seasonal phenology (Arnell, 1875; Stark, 2002). In D. lycopodioides, perigonia and perichaetia remain for the entire season on the gametophytic ramets, even after their function has ceased, and sometimes longer. We are therefore confident that we were able to reliably assess sex expression (176 reproductive ramets). The sex of the 484 non-reproductive ramets was identified using a molecular sex marker as described below. We calculated proportions of reproductive males, proportions of reproductive females and phenotypic and genotypic sex ratios as explained above (Study species, study system and terminology; Fig. 1). Based on the sampled ramets, we categorized each patch as either (0) for patches without reproductive ramets or as (1) for patches with at least one reproductive ramet (either male or female). To be able to explore potential associations of sex expression and sex ratio with sporophyte occurrence and, eventually, with the dispersal capacity of the species, we estimated sporophyte frequency in the three regions. Drepanocladus lycopodioides forms sporophytes only occasionally (see above). It is relatively unlikely that sporophytic ramets would be collected with the field method used in this study, i.e. by randomly sampling 20 single ramets from the many ramets occurring in a patch. We therefore estimated relative regional sporophyte frequency by investigating herbarium specimens of D. lycopodioides kept at S. Herbarium specimens normally consist of tens of ramets and represent single patches. We are aware of sporophytic samples being potentially over-represented among herbarium specimens. However, we see no reason to assume that this potential bias in sporophyte collecting should differ between the three study regions. Herbarium specimens can thus be regarded as legitimate replicates of frequencies of sporophyte production in different regions. We quantified sporophyte frequency as the proportion of specimens bearing sporophytes in each region (Öland N = 59, Gotland N = 36, Stockholm Archipelago N = 90). We sampled the top portion of the non-reproductive ramets and extracted total DNA using the DNeasy Plant Mini Kit for DNA isolation from plant tissue (Qiagen). We identified sex in these ramets (N = 484) following the method developed by Korpelainen et al. (2008). We ran a PCR with the primers PT-3f and PT-3r (Korpelainen et al., 2008) with 2 µL of template and using the following protocol: 4 min at 94 °C for initial denaturation, followed by 35 cycles of 45 s at 94 °C, 45 s at 53 °C and 30 s at 72 °C, ending with a final extension step of 8 min at 72 °C. Amplification products were separated on agarose gels, where bands present indicate female ramets. For ramets that did not yield products in the first PCR with primers PT-3f and PT-3r, we ran a second PCR with primers PT-1f and PT-2r to test DNA quality (as to Korpelainen et al., 2008; and Bisang & Hedenäs, 2013). A random selection of the ramets of D. lycopodioides used in this study was included in a parallel study to explore the regional intraspecific genetic variation (Hedenäs & Bisang, 2015). In these ramets, the successful DNA amplification by Hedenäs & Bisang (2015) confirmed that the method for sex identification in the present study is adequate, and they were not subjected to the second PCR. Because the molecular marker targets females, a lack of PCR products (no band on the gels) in the first PCR indicate males. The sex marker operationality depends on the second portion in the sex-differentiating region, which strongly differs between the sexes [GenBank accession numbers EU368956–EU368958, and Korpelainen et al. (2008) for details] and the developed primers amplify the female region. The reliability of the molecular marker for sex assignment in this study has previously been tested using in total 55 individual ramets bearing documented male or female sexual organs in D. lycopodioides and the related D. trifarius (F.Weber & D.Mohr) Broth. (Bisang et al., 2010; Hedenäs et al., 2010). All reproductive females produced a PCR band with primers PT-3f and PT-3r, whereas the reproductive males never did. This strong correlation does not necessarily imply that our female-specific marker is completely sex-linked (Charlesworth, 2015). To demonstrate the latter would require a different methodological approach, for example a genetic linkage map or a genealogical study. However, based on the strong correlation, we are confident that our method reliably identified sex in our study species. All laboratory work for this study and the amplifications in Hedenäs & Bisang (2015) were conducted in autumn and winter following fieldwork to ensure high DNA quality. Data analyses We tested first (1) for overall differences in the reproductive traits between regions, and then (2) whether the differences were associated with differences in patch size and wetness, representing environmental conditions. (1) To examine overall differences in ‘phenotypic sex ratio’, ‘genotypic sex ratio’, the ‘proportion of reproductive males’ and the ‘proportion of reproductive females’ (Fig. 1) between regions we used generalized linear models with binomial error distributions and logit link functions. We included region as a fixed factor and grouped observations of sex (phenotypic males vs. phenotypic females or genotypic males vs. genotypic females, respectively) and sex expression (reproductive vs. non-reproductive males or reproductive vs. non-reproductive females, respectively) per patch using the command ‘cbind’. We used multiple comparisons of means by Tukey contrasts to test for differences between individual regions. We tested whether regional mean phenotypic and genotypic sex ratios differed from an unbiased sex ratio [M/(F + M) = 0.5] by checking if the 95% confidence intervals of the estimates from these generalized linear models included zero (=logit 0.5). Sporophyte frequencies between regions were compared with Pearson’s chi-squared test. (2) To examine to what extent differences between regions in phenotypic sex ratio, genotypic sex ratio, the proportion of reproductive males and the proportion of reproductive females were mediated by differences in environmental conditions, we ran similar models including also patch size and wetness. Lastly, we examined whether region, wetness and patch size affected the presence of reproductive structures (regardless of sex) at the patch level, and the presence of one or both sexes in patches using generalized linear models with binomial error distribution and logit link functions. Patch size and wetness were log-transformed prior to all analyses. In the subsample of 24 patches containing reproductive ramets of either sex, i.e. with perigonia or perichaetia with or without sporophytes, we tested for differences between phenotypic and genotypic sex ratios by a paired t-test. Generalized linear models and related tests were conducted in the R 3.1.1 environment (generalized linear models; R Development Core Team, 2014). For Pearson’s chi-squared tests and t-tests we used STATISTICA 10 (StatSoft, 2011). RESULTS There was large variation in sex expression for both sexes, i.e. the proportions of male and female reproductive ramets, and in phenotypic and genotypic sex ratios between patches (Supporting Information, Table S2; Fig. S1A–D). In total, nine of 33 patches (27%) completely lacked ramets with sexual structures (two to four patches in each region). In individual patches, the proportions of male reproductive ramets (of the male total) and female reproductive ramets (of the female total) ranged from 0 to 88% and 0 to 85%, respectively. All regions included phenotypic and genotypic purely female, purely male and mixed sex patches. Both phenotypic and genotypic sex ratios differed from a balanced sex ratio in all regions and were significantly skewed towards the female sex (i.e. <0.5, Table 1; Table S2). Table 1. A, reproductive traits in the dioicous moss Drepanocladus lycopodioides in three regions in the southern Swedish Baltic area (Fig. 2) and differences between regions in these traits. B, effects of region, patch size and wetness on these traits. See Figure 1 for further explanations of the reproductive traits . . Proportion of reproductive females . Proportion of reproductive males . Phenotypic sex ratio . Genotypic sex ratio . A Öland 0.250a (0.069) 0.141a (0.086) 0.167a (0.118) 0.173a (0.104) Gotland 0.178a (0.084) 0.306a (0.207) 0.214a (0.149) 0.155a (0.094) Stockholm Archipelago 0.338b (0.138) 0.313a (0.159) 0.375a (0.183) 0.296b (0.138) B Region 8.728 4.924 6.806 23.168 0.013 ns 0.033 <0.001 Patch size −0.306 0.808 0.750 1.044 ns ns 0.010 <0.001 Wetness −1.703 1.982 −0.120 −2.754 0.014 ns ns <0.001 . . Proportion of reproductive females . Proportion of reproductive males . Phenotypic sex ratio . Genotypic sex ratio . A Öland 0.250a (0.069) 0.141a (0.086) 0.167a (0.118) 0.173a (0.104) Gotland 0.178a (0.084) 0.306a (0.207) 0.214a (0.149) 0.155a (0.094) Stockholm Archipelago 0.338b (0.138) 0.313a (0.159) 0.375a (0.183) 0.296b (0.138) B Region 8.728 4.924 6.806 23.168 0.013 ns 0.033 <0.001 Patch size −0.306 0.808 0.750 1.044 ns ns 0.010 <0.001 Wetness −1.703 1.982 −0.120 −2.754 0.014 ns ns <0.001 A, means and (SE) across 11 patches per region, based on 20 sampled individual ramets per patch; superscript letters a and b indicate that two regions significantly differ in the respective reproductive trait. Statistical significance (P-values in italics in Table 1B) was assessed with generalized linear models (GLM); ns indicates a non-significant difference. Phenotypic and genotypic sex ratios differ in all regions significantly from a balanced sex ratio (0.5), i.e. 95% confidence intervals of the estimates from the GLM did not include zero [logit (0.5) = 0]. B, effects: coefficient estimates from the GLM, or χ 2 for region (three levels) from analysis of deviance. Open in new tab Table 1. A, reproductive traits in the dioicous moss Drepanocladus lycopodioides in three regions in the southern Swedish Baltic area (Fig. 2) and differences between regions in these traits. B, effects of region, patch size and wetness on these traits. See Figure 1 for further explanations of the reproductive traits . . Proportion of reproductive females . Proportion of reproductive males . Phenotypic sex ratio . Genotypic sex ratio . A Öland 0.250a (0.069) 0.141a (0.086) 0.167a (0.118) 0.173a (0.104) Gotland 0.178a (0.084) 0.306a (0.207) 0.214a (0.149) 0.155a (0.094) Stockholm Archipelago 0.338b (0.138) 0.313a (0.159) 0.375a (0.183) 0.296b (0.138) B Region 8.728 4.924 6.806 23.168 0.013 ns 0.033 <0.001 Patch size −0.306 0.808 0.750 1.044 ns ns 0.010 <0.001 Wetness −1.703 1.982 −0.120 −2.754 0.014 ns ns <0.001 . . Proportion of reproductive females . Proportion of reproductive males . Phenotypic sex ratio . Genotypic sex ratio . A Öland 0.250a (0.069) 0.141a (0.086) 0.167a (0.118) 0.173a (0.104) Gotland 0.178a (0.084) 0.306a (0.207) 0.214a (0.149) 0.155a (0.094) Stockholm Archipelago 0.338b (0.138) 0.313a (0.159) 0.375a (0.183) 0.296b (0.138) B Region 8.728 4.924 6.806 23.168 0.013 ns 0.033 <0.001 Patch size −0.306 0.808 0.750 1.044 ns ns 0.010 <0.001 Wetness −1.703 1.982 −0.120 −2.754 0.014 ns ns <0.001 A, means and (SE) across 11 patches per region, based on 20 sampled individual ramets per patch; superscript letters a and b indicate that two regions significantly differ in the respective reproductive trait. Statistical significance (P-values in italics in Table 1B) was assessed with generalized linear models (GLM); ns indicates a non-significant difference. Phenotypic and genotypic sex ratios differ in all regions significantly from a balanced sex ratio (0.5), i.e. 95% confidence intervals of the estimates from the GLM did not include zero [logit (0.5) = 0]. B, effects: coefficient estimates from the GLM, or χ 2 for region (three levels) from analysis of deviance. Open in new tab The Stockholm Archipelago exhibited less strongly female-skewed phenotypic and genotypic sex ratios and higher proportions of female reproductive and male reproductive ramets than the other two regions (Table 1A). Regional differences in the proportions of reproductive males were not statistically significant, and phenotypic sex ratio differences between regions were significant only when patch size and wetness were included in the model (Table 1B). Differences between regions in both sex ratios and the proportion of reproductive females remained also when differences in patch size and wetness were accounted for (Table 1B). Sporophyte frequency was significantly higher in the Stockholm Archipelago (0.189) than on Gotland (0.028) (χ 2 = 5.45; P = 0.020) and marginally higher than on Öland (0.085) (χ 2 = 3.07; P = 0.080). Patches were on average larger on Öland than in the other two regions (Hedenäs & Bisang, 2015), but wetness, measured as the depth of the patch-containing depressions, did not differ between regions (Supporting Information, Fig. S1E, F; Table S2). Larger patches carried more phenotypic and genotypic male ramets relative to females than smaller patches, and thus exhibited a weaker female sex ratio bias (Table 1B). Wetter patches had a stronger female bias in genotypic sex ratio than drier patches (Table 1B). Wetter patches had a lower proportion of reproductive females than drier patches. In line with this, the overall presence of sexual structures at patch level of either sex was negatively related to wetness; i.e. ramets in wet patches were less likely to form sexual organs than those in relatively drier habitats (log likelihood = −17.2, χ 2 = 4.3, P = 0.039). Finally, both sexes co-occurred more often in larger patches than in smaller patches (log likelihood = −14.8, χ 2 = 4.4, P = 0.036). In the 24 patches that contained reproductive ramets (of either sex), phenotypic and genotypic sex ratios did not differ (d.f. = 23, t = −0.811, P = 0.426). DISCUSSION Our results demonstrate that phenotypic and genotypic sex ratios are clearly female-biased in D. lycopodioides across three different regions. Phenotypic and genotypic sex ratios and the proportions of reproductive female ramets differed between regions, which could partly be explained by patch size and wetness of the habitat. This constitutes the first evidence that the occurrence of genotypic sexes, irrespective of the formation of sexual organs, can vary with environmental conditions in natural populations of bryophytes. Do sex expression and sex ratios vary between regions? Our data indicate, in support of our first hypothesis, higher phenotypic and genotypic male availability (less female-biased sex ratios), and higher proportions of reproductive ramets in the Stockholm Archipelago than in the other regions. Differences between regions were non-significant for proportions of reproductive male ramets and for phenotypic sex ratio in the models without co-variates. Regional differences in these models with a small number of male reproductive ramets were obscured by patch size and wetness variation. Including information about patch size and wetness removed variation due to non-measured factors and led to a significant region effect for phenotypic sex ratio (Table 1B). The Stockholm Archipelago, with the lowest female sex ratio bias, also had a higher sporophyte frequency compared to the other regions. It has previously been demonstrated that increased sex expression and male mate availability enhance fertilization incidences (e.g. Longton & Greene, 1969; Cronberg, 2002; Bisang et al., 2004). On the other hand, sporophyte production incurs a substantial reproductive cost for the female plant (Ehrlén, Bisang & Hedenäs, 2000; Rydgren & Økland, 2002b), which has been shown to lead to lower performance of the females and may eventually result in higher female mortality (Bisang & Ehrlén, 2002; Obeso, 2002). In a modelling approach with the perennial pleurocarpous moss Hylocomium splendens (Hedw.) Schimp., Rydgren et al. (2010) showed that both sporophyte frequency and phenotypic sex ratios affected population growth rate. Their model predicted that a female-biased sex ratio was maintained if sporophyte production was infrequent, keeping female reproductive costs limited, and that the female bias was a consequence of lower clonal growth rate in males compared to non-sporophytic females. Our results can be interpreted in the light of these findings: the lower female sex ratio biases may be a consequence of higher fertilization incidence and lower performance of sporophyte-bearing females in the Stockholm Archipelago. Moreover, the fact that meiotic spore sex ratio is balanced at the germination stage in D. lycopodioides (Bisang et al., 2017) implies that relative male provision increases with increased sporophyte production, which may influence recruitment opportunities. A higher number of males and more mating events may also contribute to the greater haplotype diversity among males in the Stockholm Archipelago compared to the other regions observed in a previous study (Hedenäs & Bisang, 2015). Unexplained regional variation remained both for sex ratios and for the proportion of reproductive female ramets after fitting patch size and wetness to the models. This implies that the factor ‘region’ includes variation in additional parameters that we did not directly measure. For example, Öland and Gotland became available for colonization after the Late Glacial Maximum at around 11–12 ky BP, in contrast to the Stockholm Archipelago where most of the habitats accessible for D. lycopodioides were exposed only 2–3 ky BP (Lundegårdh, Lundqvist & Lindström, 1974; Hedenäs & Bisang, 2015). Thus, the former experienced a considerably longer period than the Stockholm Archipelago for a potential inter-population sex ratio heterogeneity to establish (Field et al., 2013b). Furthermore, light, variation in photoperiod, temperature, precipitation events and nutrient levels have been shown to affect sex induction in bryophytes and to account for sex-differential responses in dioicous bryophytes (e.g. Benson-Evans, 1961; Chopra, 1984; Hohe et al., 2002; Rydgren & Økland, 2002a; Lee, Rosenstiel & Eppley, 2010; Maciel-Silva, Marques Valio & Rydin, 2012). Such factors might also have differed between our study regions. Can variation in sexual traits be explained by environmental factors? Large patches are likely to comprise more micro-sites suitable for diaspore colonization, increasing the probability that both sexes establish. Moreover, larger patches probably contain more diverse environmental conditions than smaller patches (e.g. Gignac & Dale, 2005), which improves the chance for sexes with different habitat requirements to co-occur. For D. lycopodioides we indeed found, in line with our second hypothesis, relatively more phenotypic and genotypic male ramets (higher sex ratios) and a higher likelihood of larger patches harbouring both sexes. This agrees with the generally higher genetic diversity in larger than in smaller patches (Hedenäs & Bisang, 2015). Although patch size did not affect sex-expression in D. lycopodioides, larger patches were more likely than small patches to contain sex-expressing thalli in Marchantia inflexa Nees & Mont. (McLetchie & Puterbaugh, 2000). Those authors attributed this to micro-habitat variation affecting the formation of sexual structures. It seems likely that environmental effects on the formation of sexual organs differ among species depending on life histories and the general environment of the target species. This may explain the different response in sex expression in M. inflexa, a ruderal thalloid liverwort, and our study species, a long-lived wetland pleurocarpous moss. Wetness was negatively associated with genotypic sex ratio in D. lycopodioides, i.e. fewer males relative to females occurred in wetter patches. This finding supports our second hypothesis. It agrees with the general pattern in flowering plants, in which females typically occur in moister or otherwise more favourable sites. Drought and resource-poor conditions tend to bias plant sex ratios in favour of males (Zhang et al., 2012; Golenberg & West, 2013). It is commonly suggested that this is driven by functional sex differences and related resource demands, i.e. sperm production vs. egg production and offspring maturation (Dudley & Galen, 2007; Barrett & Hough, 2013). Moisture availability is a key factor also for bryophyte physiology and reproduction. Cell water content is directly regulated by ambient humidity and fertilization requires liquid water for the sperm to reach the egg cell (Vanderpoorten & Goffinet, 2009). Previous bryophyte sex ratio studies in natural habitats have focused on populations with sexual reproduction, i.e. on phenotypic sex ratio. Hitherto reported environmental effects on phenotypic sex ratios may thus be due to sex-specific environmental effects on the formation of sexual organs rather than due to sex-specific habitat requirements (Longton, 1990). In some species growing in generally dry environments, reproductive males tended to inhabit, or were restricted to, more mesic sites than reproductive females, which was hypothesized to be an effect of lower desiccation tolerance of males (Stark et al., 2005; Benassi et al., 2011; Blackstock, 2015). On the other hand, reproductive females, compared to their conspecific males, occurred under more favourable conditions with respect to air humidity in Pseudoscleropodium purum (Hedw.) M.Fleisch. (Boquete et al., 2016). In laboratory experiments, female plants of different species of Splachnum Hedw. grew better under favourable light and substrate conditions (Cameroon & Wyatt, 1990). Female plants of M. inflexa from reproductive populations were shown to be more tolerant to dehydration than males under certain experimental conditions (Marks, Burton & McLetchie, 2016), whereas the sexes of the desert moss Syntrichia caninervis Mitt. did not differ in their response to desiccation stress (Stark et al., 2005). Little is known, however, about whether genotypic sexes are associated with environmental conditions in non-reproductive bryophytes. The frequency of genotypic sexes differed between two populations in S. caninervis, but a direct association with habitat factors was not established (Baughman et al., 2017). The distribution of genotypic sexes was not related to habitat parameters in a large population of the wetland moss D. trifarius (Bisang et al., 2015). Drepanocladus lycopodioides and D. trifarius are related, and both are long-lived (Hedenäs & Rosborg, 2009). They nevertheless differ in the distribution and frequency of sexual reproduction. Proportions of reproductive ramets per patch in D. lycopodioides varied regionally from 20% to 36%, and sporophytes occurred in all regions and at least in one of the study patches per region (this study). In the intensively studied population of D. trifarius, the proportion of reproductive ramets was as low as 1.4% (N = 277), and no sporophytes were formed in the study area (Bisang et al., 2015). Even at the European scale sex expression is higher in D. lycopodioides (51% of N = 265) than in D. trifarius (34% of N = 223) (Bisang et al., 2014; and unpubl. data IB). Based on the accumulated knowledge on plant sex ratio variation, we may hypothesize that females of D. lycopodioides occupy the wetter habitats because they need more resources to nurture the developing sporophytes (Ehrlén et al., 2000; Rydgren & Økland, 2002b), and that they may outcompete males under such habitat conditions. High sporophytic resource demands require that photosynthesis occurs over longer periods (Proctor, 1990), i.e. in deeper depressions that dry out later in the growing season, or less often. In species without, or with rare, sexual reproduction the evolutionary driver of differential reproductive allocation is absent (Bierzychudek & Eckhart, 1988). This could explain why we did not observe genotypic sex-related differences in habitat conditions in the large population of D. trifarius, which largely lacked sexual organs. Although genotypic females of D. lycopodioides were relatively more common than males in wetter places in this study, the proportion of reproductive female ramets and the overall presence of sexual organs per patch were negatively associated with wetness. In a perennial moss inhabiting grasslands, female sex expression rate increased with bush cover, probably improving water and nutrient availability and protection (Rosengren et al., 2014). On the other hand, mosses growing in waterlogged environments or under water, for example in lakes or in streams, rarely reproduce sexually (Glime, 1984; Blackstock, 2018; pers. observ. by IB, LH). In a strongly calcareous environment, as in our study regions, C uptake from calcareous water with high pH is limited (Proctor, 1980). This may lead to resource deficits and ensuing low performance such as lack of sex expression, during periods when the habitat depressions are submerged. Sexual organ development follows a specific seasonal phenology. In D. lycopodioides, the sexual organs mature and fertilization occurs during summer and autumn (Arnell, 1875; pers. observ. by IB, LH), whereas sporophytes mature in the following growth season from spring into early summer. Thus, critical moisture conditions occur at different seasons for the different developmental stages. We used a proxy measure for habitat wetness (the depth of the depression inhabited by D. lycopodioides patches) that does not take wetness variation between different seasons into account. Prolonged overall moisture in deeper depressions seems to generally favour females, whereas high wetness in the summer negatively affects their sex expression. Limited C uptake during submersion may also provide an alternative explanation for fewer males than females in wet habits vs. higher female moisture demand. Males could be relatively more sensitive than females against shortage of available C and thus suffer higher mortality. Does phenotypic sex reflect genotypic sex? Phenotypic and genotypic sex ratios did not differ in the subsample of patches that harboured reproductive ramets, in alignment with our third hypothesis. This suggests that overall, the two sexes form sexual branches in similar proportions of ramets in these regions. This agrees with the comparisons of the two sex ratios that have been carried out at larger geographical scales: the species investigated in this study, the related D. trifarius growing in wetlands and the ruderal Bryum argenteum Hedw. (Hedenäs et al., 2010; Stark, McLetchie & Eppley, 2010; Bisang & Hedenäs, 2013). It contradicts the ‘shy male’ hypothesis that predicts a disproportionately large reservoir of males among the non-reproductive ramets (Stark et al., 2010). In contrast, males in local populations of two other species formed sexual branches less often than females, leading to a phenotypic sex ratio with a stronger female bias than the genotypic sex ratio (Newton, 1971; Baughman et al., 2017). Comparing the sex data from this study with haplotype patterns in Hedenäs & Bisang (2015) shows that patches might harbour multiple male and female genets. A more intensive haplotype screening of the sampled ramets would probably provide important information relating to sex ratios at the genet level, to differences in recruitment patterns between sexes and to the clonal structure of the two sexes. CONCLUSIONS Studies on bryophyte sex expression and sex ratio variation are still few, in particular with respect to genotypic sex. We found evidence of geographical variation in sex expression and phenotypic and genotypic sex ratios in a long-lived wetland moss with occasional sexual reproduction. Sex expression was negatively affected by wetness, and sex ratios were more strongly female-biased in wet habitats. We postulate that bryophyte sex response to environmental conditions differs in relation to the general habitat which the species occupy, their life history traits and possible interactions between them (Field et al., 2013b; ,Barrett, 2015). It seems likely that species with regular sexual reproduction perform differently from largely non-reproductive species as only the former experience sex-specific resource demands (e.g. Rydgren et al., 2010; Barrett & Hough, 2013). ‘Stressful conditions’ vary among species depending on the species’ physiology and general habitat of the species, and sensitivity should also depend on whether the species occupies a habitat where resources (e.g. water availability) are generally limited. Sex ratios have a profound impact on demography and genetic variation within populations, which eventually affect the extinction risk of species (e.g. Bessa-Gomes, Legendre & Clobert, 2004; Hedenäs, 2015). The ongoing climate change may affect the sexes differently and, consequently, modify the frequency, spatial distribution and genetic diversity of males and females and thus alter population dynamics and fertility (Petry et al., 2016; Blackstock, 2018). Because bryophyte life cycles and sex determination fundamentally differ from those of seed plants, insights in the common female sex ratio bias in these organisms broadens significantly our general understanding of sex ratio biases in plants, which is so far largely based on flowering plants. Our results on genotypic sex ratios have exclusively been based on observational data. It is therefore important to supplement these with experimental approaches with species from different habitats and explore whether bryophyte males, in contrast to male seed plants, indeed respond more strongly to environmental stress than females (Benassi et al., 2011; Golenberg & West, 2013; this study). Existing data need to be complemented with data on genotypic sexes, by applying sex-specific molecular markers, from bryophyte species with different levels of sexual reproduction and from different natural habitats. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s website. Table S1. Geographical location and field sampling dates of the 33 study patches in the Baltic area in southern Sweden. Table S2. Sex and sex expression of the 660 individual ramets analysed, and sex ratios, patch size of the D. lycopodioides occurrence and wetness, for each of the 33 patches. Figure S1. Box and box-whisker of sexual traits, patch size and wetness across 11 patches in each of the three study regions in the southern Swedish Baltic area. ACKNOWLEDGEMENTS We thank Keyvan Mirbakhsh for efficient laboratory work, and the anonymous reviewers for constructive comments. AUTHOR CONTRIBUTIONS I.B. and L.H. conceived and designed the study. L.H. and I.B. performed fieldwork and data collection. J.E. and I.B. analysed the data. I.B. wrote the manuscript with significant input from J.E. and L.H. FUNDING This work was financially supported by Stiftelsen Oscar och Lili Lamms minne (grant number FO2010-0050). REFERENCES Alvarenga LDP , Pôrto KC , Zartman CE . 2013 . Sex ratio, spatial segregation, and fertilization rates of the epiphyllous moss Crossomitrium patrisiae (Brid.) Mull.Hal. in the Brazilian Atlantic rainforest . Journal of Bryology 35 : 88 – 95 . Google Scholar Crossref Search ADS WorldCat Arnell HW . 1875 . De skandinaviska löfmossornas kalendarium . Uppsala Universitets Årsskrift, Matematik och Naturvetenskap IV 1875 : 1 – 129 . OpenURL Placeholder Text WorldCat Bachtrog D , Kirkpatrick M , Mank JE , McDaniel SF , Pires JC , Rice W , Valenzuela N . 2011 . Are all sex chromosomes created equal? Trends in Genetics: TIG 27 : 350 – 357 . Google Scholar Crossref Search ADS PubMed WorldCat Barrett SCH . 2015 . Influences of clonality on plant sexual reproduction . Proceedings of the National Academy of Sciences USA 112 : 8859 – 8866 . Google Scholar Crossref Search ADS WorldCat Barrett SC , Hough J . 2013 . Sexual dimorphism in flowering plants . Journal of Experimental Botany 64 : 67 – 82 . Google Scholar Crossref Search ADS PubMed WorldCat Barrett SC , Yakimowski SB , Field DL , Pickup M . 2010 . Ecological genetics of sex ratios in plant populations . Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 365 : 2549 – 2557 . Google Scholar Crossref Search ADS PubMed WorldCat Baughman JT , Payton AC , Paasch AE , Fisher KM , McDaniel SF . 2017 . Multiple factors influence population sex ratios in the Mojave Desert moss Syntrichia caninervis . American Journal of Botany 104 : 733 – 742 . Google Scholar Crossref Search ADS PubMed WorldCat Bedford THB . 1938 . Sex distribution in colonies of Climacium dendroides W. & M. and its relation to fruit bearing . Northwestern Naturalist 13 : 213 – 221 . OpenURL Placeholder Text WorldCat Benassi M , Stark LR , Brinda JC , McLetchie DN , Bonine M , Mishler BD . 2011 . Plant size, sex expression and sexual reproduction along an elevation gradient in a desert moss . The Bryologist 114 : 277 – 288 . Google Scholar Crossref Search ADS WorldCat Benson-Evans K . 1961 . Environmental factors and bryophytes . Nature 191 : 255 – 260 . Google Scholar Crossref Search ADS WorldCat Bertiller MB , Sain CL , Bisigato AJ , Coronato FR , Aries JO , Graff P . 2002 . Spatial sex segregation in the dioecious grass Poa ligularis in northern Patagonia: the role of environmental patchiness . Biodiversity and Conservation 11 : 69 – 84 . Google Scholar Crossref Search ADS WorldCat Bessa-Gomes C , Legendre S , Clobert J . 2004 . Allee effects, mating systems and the extinction risk in populations with two sexes . Ecology Letters 7 : 802 – 812 . Google Scholar Crossref Search ADS WorldCat Bierzychudek P , Eckhart V . 1988 . Spatial segregation of the sexes of diecious plants . American Naturalist 132 : 34 – 43 . Google Scholar Crossref Search ADS WorldCat Bisang I , Ehrlén J . 2002 . Reproductive effort and cost of sexual reproduction in female Dicranum polysetum . The Bryologist 105 : 384 – 397 . Google Scholar Crossref Search ADS WorldCat Bisang I , Ehrlén J , Hedenäs L . 2004 . Mate limited reproductive success in two dioicous mosses . Oikos 104 : 291 – 298 . Google Scholar Crossref Search ADS WorldCat Bisang I , Ehrlén J , Korpelainen H , Hedenäs L . 2015 . No evidence of sexual niche partitioning in a dioecious moss with rare sexual reproduction . Annals of Botany 116 : 771 – 779 . Google Scholar Crossref Search ADS PubMed WorldCat Bisang I , Ehrlén J , Persson C , Hedenäs L . 2014 . Family affiliation, sex ratio and sporophyte frequency in unisexual mosses . Botanical Journal of the Linnean Society 174 : 163 – 172 . Google Scholar Crossref Search ADS WorldCat Bisang I , Hedenäs L . 2005 . Sex ratio patterns in dioicous bryophytes re-visited . Journal of Bryology 27 : 207 – 219 . Google Scholar Crossref Search ADS WorldCat Bisang I , Hedenäs L . 2013 . Males are not shy in the wetland moss Drepanocladus lycopodioides . International Journal of Plant Science 174 : 733 – 739 . Google Scholar Crossref Search ADS WorldCat Bisang I , Hedenäs L , Cronberg N . 2017 . Can the meiotic sex ratio explain the sex ratio bias in adult populations in the dioicous moss Drepanocladus lycopodioides? Journal of Bryology 39 : 115 – 120 . Google Scholar Crossref Search ADS WorldCat Bisang I , Korpelainen H , Hedenäs L . 2010 . Can the sex-specific molecular marker of Drepanocladus trifarius uncover gender in related species? Journal of Bryology 32 : 305 – 308 . Google Scholar Crossref Search ADS WorldCat Blackstock TH . 2015 . Sex expression and sporophyte frequency in Frullania tamarisci (L.) Dumort . Journal of Bryology 37 : 202 – 208 . Google Scholar Crossref Search ADS WorldCat Blackstock TH . 2018 . Apparent increase in fertility of Lunularia cruciata (L.) Lind. (Marchantiophyta) in Britain associated with climate change . Journal of Bryology 40 : 377 – 383 . Google Scholar Crossref Search ADS WorldCat Boquete MT , Fernández JA , Aboal JR , Shaw AJ . 2016 . Significance of the intraspecific morphological variability in biomonitoring studies with mosses: among-populations and between-sexes approach . Environmental and Experimental Botany 130 : 106 – 112 . Google Scholar Crossref Search ADS WorldCat Brzyski JR , Taylor W , McLetchie DN . 2014 . Reproductive allocation between the sexes, across natural and novel habitats, and its impact on genetic diversity . Evolutionary Ecology 28 : 247 – 261 . Google Scholar Crossref Search ADS WorldCat Cameroon RG , Wyatt R . 1990 . Spatial patterns and sex ratios in dioecious and monoecious mosses of the genus Splachnum . The Bryologist 93 : 161 – 166 . Google Scholar Crossref Search ADS WorldCat Caspari S , Dürhammer O , Sauer M , Schmidt C . 2018 . Rote Liste und Gesamtartenliste der Moose (Anthocerotophyta, Marchantiophyta und Bryophyta) Deutschlands . Naturschutz und Biologische Vielfalt 70 : 361 – 489 . OpenURL Placeholder Text WorldCat Charlesworth D . 2015 . Plant contributions to our understanding of sex chromosome evolution . The New Phytologist 208 : 52 – 65 . Google Scholar Crossref Search ADS PubMed WorldCat Chopra RN . 1984 . Environmental factors affecting gametangial induction in bryophytes . Journal of the Hattori Botanical Laboratory 55 : 99 – 104 . OpenURL Placeholder Text WorldCat Cronberg N . 2002 . Colonization dynamics of the clonal moss Hylocomium splendens on islands in a Baltic land uplift area: reproduction, genet distribution and genetic variation . Journal of Ecology 90 : 925 – 935 . Google Scholar Crossref Search ADS WorldCat Dudley LS , Galen C . 2007 . Stage-dependent patterns of drought tolerance and gas exchange vary between sexes in the alpine willow, Salix glauca . Oecologia 153 : 1 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Ehrlén J , Bisang I , Hedenäs L . 2000 . Costs of sporophyte production in the moss, Dicranum polysetum . Plant Ecology 149 : 207 – 217 . Google Scholar Crossref Search ADS WorldCat Field DL , Pickup M , Barrett SC . 2013a . Comparative analyses of sex-ratio variation in dioecious flowering plants . Evolution 67 : 661 – 672 . Google Scholar Crossref Search ADS WorldCat Field DL , Pickup M , Barrett SC . 2013b . Ecological context and metapopulation dynamics affect sex-ratio variation among dioecious plant populations . Annals of Botany 111 : 917 – 923 . Google Scholar Crossref Search ADS WorldCat Gignac LD , Dale MRT . 2005 . Effects of fragment size and habitat heterogeneity on cryptogam diversity in the low-boreal forest of western Canada . The Bryologist 108 : 50 – 66 . Google Scholar Crossref Search ADS WorldCat Glime J . 1984 . Physio-ecological factors relating to reproduction and phenology in Fontinalis dalecarlica . The Bryologist 87 : 17 – 23 . Google Scholar Crossref Search ADS WorldCat Golenberg EM , West NW . 2013 . Hormonal interactions and gene regulation can link monoecy and environmental plasticity to the evolution of dioecy in plants . American Journal of Botany 100 : 1022 – 1037 . Google Scholar Crossref Search ADS PubMed WorldCat Graff P , Rositano F , Aguiar MR . 2013 . Changes in sex ratios of a dioecious grass with grazing intensity: the interplay between gender traits, neighbour interactions and spatial patterns . Journal of Ecology 101 : 1146 – 1157 . Google Scholar Crossref Search ADS WorldCat Groen KE , Stieha CR , Crowley PH , McLetchie DN . 2010 . Sex-specific plant responses to light intensity and canopy openness: implications for spatial segregation of the sexes . Oecologia 162 : 561 – 570 . Google Scholar Crossref Search ADS PubMed WorldCat Hardy ICW 2002 . Sex ratios . Cambridge : Cambridge University Press . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Hedenäs L . 2003 . The European species of the Calliergon–Scorpidium–Drepanocladus complex, including some related or similar species . Meylania 28 : 1 – 116 . OpenURL Placeholder Text WorldCat Hedenäs L . 2015 . Rhytidium rugosum (Bryophyta) colonized Scandinavia from at least two glacial refugial source populations . Botanical Journal of the Linnean Society 179 : 635 – 657 . Google Scholar Crossref Search ADS WorldCat Hedenäs L , Bisang I . 2015 . Infraspecific diversity in a spore-dispersed species with limited distribution range . Systematics and Biodiversity 13 : 17 – 27 . Google Scholar Crossref Search ADS WorldCat Hedenäs L , Bisang I , Korpelainen H , Cronholm B . 2010 . The true sex ratio in European Pseudocalliergon trifarium (Bryophyta: Amblystegiaceae) revealed by a novel molecular approach . Biological Journal of the Linnean Society 100 : 132 – 140 . Google Scholar Crossref Search ADS WorldCat Hedenäs L , Bisang I , Schnyder N . 2003. The distribution of bryophytes in Switzerland and in Liechtenstein IV. Hamatocaulis and Pseudocalliergon . Botanica Helvetica 113 : 111 – 123 . OpenURL Placeholder Text WorldCat Hedenäs L , Rosborg C . 2009 . Pseudocalliergon is nested within Drepanocladus (Bryophyta: Amblystegiaceae) . Lindbergia 33 : 67 – 74 . OpenURL Placeholder Text WorldCat Hohe A , Rensing SA , Mildner M , Lang D , Reski R . 2002 . Day length and temperature strongly influence sexual reproduction and expression of a novel MADS-box gene in the moss Physcomitrella patens . Plant Biology 4 : 595 – 602 . Google Scholar Crossref Search ADS WorldCat Holá E , Vesalainen T , Těšitel J , Laaka-Lindberg S . 2014 . Sex ratio, sex-specific pattern in vegetative growth and gemma production in an aquatic liverwort, Scapania undulata (Marchantiophyta: Scapaniaceae) . Botanical Journal of the Linnean Society 175 : 229 – 241 . Google Scholar Crossref Search ADS WorldCat Jesson LK , Garnock-Jones PJ . 2012 . Can classifications of functional gender be extended to all land plants? Perspectives in Plant Ecology, Evolution and Systematics 14 : 153 – 160 . Google Scholar Crossref Search ADS WorldCat de Jong TJ , Klinkhamer PGL . 2002 . Sex ratios in dioecious plants. In: Hardy ICW ed. Sex ratios . Cambridge : Cambridge University Press , 349 – 364 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Korpelainen H , Bisang I , Hedenäs L , Kolehmainen J . 2008 . The first sex-specific molecular marker discovered in the moss Pseudocalliergon trifarium . The Journal of Heredity 99 : 581 – 587 . Google Scholar Crossref Search ADS PubMed WorldCat Laenen B , Machac A , Gradstein SR , Shaw B , Patiño J , Désamoré A , Goffinet B , Cox CJ , Shaw AJ , Vanderpoorten A . 2016 . Increased diversification rates follow shifts to bisexuality in liverworts . The New Phytologist 210 : 1121 – 1129 . Google Scholar Crossref Search ADS PubMed WorldCat Lee L , Rosenstiel TN , Eppley SM . 2010 . Variation of photoperiod response in moss gametangial formation . The Bryologist 113 : 673 – 678 . Google Scholar Crossref Search ADS WorldCat Longton RE . 1990 . Sexual reproduction in bryophytes in relation to physical factors of the environment. In: Chopra RN , Bathla SC , eds. Bryophyte development: physiology and biochemistry . Boca Raton : CRC Press , 139 – 166 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Longton RE , Greene SW . 1969 . Relationship between sex distribution and sporophyte production in Pleurozium schreberi (Brid.) Mitt . Annals of Botany 33 : 107 – 126 . Google Scholar Crossref Search ADS WorldCat Longton RE , Schuster RM . 1983 . Reproductive biology. In: Schuster RM , ed. New manual of bryology . Nichinan : The Hattori Botanical Laboratory , 386 – 462 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Lundegårdh PH , Lundqvist J , Lindström M 1974 . Berg och jord i Sverige . Stockholm : Almqvist & Wiksell Läromedel . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Maciel-Silva AS , Marques Valio IF , Rydin H . 2012 . Altitude affects the reproductive performance in monoicous and dioicous bryophytes: examples from a Brazilian Atlantic rainforest . AoB Plants 2012 : pls016 . Google Scholar Crossref Search ADS PubMed WorldCat Marks RA , Burton JF , McLetchie DN . 2016 . Sex differences and plasticity in dehydration tolerance: insight from a tropical liverwort . Annals of Botany 118 : 347 – 356 . Google Scholar Crossref Search ADS PubMed WorldCat McDaniel S , Atwood J , Burleigh J . 2013 . Recurrent evolution of dioecy in bryophytes . Evolution 67 : 567 – 572 . Google Scholar Crossref Search ADS PubMed WorldCat McDaniel S , Perroud J-P . 2012 . Invited perspective: bryophytes as models for understanding the evolution of sexual systems . The Bryologist 115 : 1 – 11 . Google Scholar Crossref Search ADS WorldCat McLetchie DN , Puterbaugh MN . 2000 . Population sex-ratios, sex-specific clonal traits and tradeoffs among these traits in the liverwort Marchantia inflexa . Oikos 90 : 227 – 237 . Google Scholar Crossref Search ADS WorldCat Newton ME . 1971 . A cytological distinction between male and female Mnium undulatum Hedw . Transactions of the British Bryological Society 6 : 230 – 243 . Google Scholar Crossref Search ADS WorldCat Norrell TE , Jones KS , Payton AC , McDaniel SF . 2014 . Meiotic sex ratio variation in natural populations of Ceratodon purpureus (Ditrichaceae) . American Journal of Botany 101 : 1572 – 1576 . Google Scholar Crossref Search ADS PubMed WorldCat Obeso JR . 2002 . Tansley review no. 139. The costs of reproduction in plants . New Phytologist 155 : 321 – 348 . Google Scholar Crossref Search ADS WorldCat Pereira MR , Dambros CS , Zartman CE . 2016 . Prezygotic resource-allocation dynamics and reproductive trade-offs in Calymperaceae (Bryophyta) . American Journal of Botany 103 : 1838 – 1846 . Google Scholar Crossref Search ADS PubMed WorldCat Petersen KB , Burd M . 2018 . The enigma of sex allocation in Selaginella . Annals of Botany 121 : 377 – 383 . Google Scholar Crossref Search ADS PubMed WorldCat Petry WK , Soule JD , Iler AM , Chicas-Mosier A , Inouye DW , Miller TE , Mooney KA . 2016 . Sex-specific responses to climate change in plants alter population sex ratio and performance . Science 353 : 69 – 71 . Google Scholar Crossref Search ADS PubMed WorldCat Petterson B . 1958 . Dynamik och konstans i Gotlands flora och vegetation . Acta Phytogeographica Suecica 40 . OpenURL Placeholder Text WorldCat Proctor MCF . 1980 . The requirement of aquatic bryophytes for free CO2 as an inorganic carbon source: some experimental evidence . New Phytologist 86 : 393 – 400 . Google Scholar Crossref Search ADS WorldCat Proctor MCF . 1990 . The physiological basis of bryophyte production . Botanical Journal of the Linnean Society 104 : 61 – 77 . Google Scholar Crossref Search ADS WorldCat R Development Core Team 2014 . R: a language and environment for statistical computing . Vienna : R Foundation for Statistical Computing . Available at: http://www.R-project.org/. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Renner SS . 2014 . The relative and absolute frequencies of angiosperm sexual systems: dioecy, monoecy, gynodioecy, and an updated online database . American Journal of Botany 101 : 1588 – 1596 . Google Scholar Crossref Search ADS PubMed WorldCat Renner SS , Heinrichs J , Sousa A . 2017 . The sex chromosomes of bryophytes: recent insights, open questions, and reinvestigations of Frullania dilatata and Plagiochila asplenioides . Journal of Systematics and Evolution 55 : 333 – 339 . Google Scholar Crossref Search ADS WorldCat Rosengren F , Cronberg N , Reitalu T , Prentice HC . 2014 . Sexual reproduction in the phyllodioicous bryophyte Homalothecium lutescens (Hedw.) H.Rob. in relation to habitat age, growth conditions and genetic variation . Journal of Bryology 36 : 200 – 208 . Google Scholar Crossref Search ADS WorldCat Rydgren K , Halvorsen R , Cronberg N . 2010 . Infrequent sporophyte production maintains a female-biased sex ratio in the unisexual clonal moss Hylocomium splendens . Journal of Ecology 98 : 1224 – 1231 . Google Scholar Crossref Search ADS WorldCat Rydgren K , Økland RH . 2002a . Sex distribution and sporophyte frequency in a population of the clonal moss Hylocomium splendens . Journal of Bryology 24 : 207 – 214 . Google Scholar Crossref Search ADS WorldCat Rydgren K , Økland RH . 2002b . Ultimate costs of sexual reproduction in the clonal moss Hylocomium splendens . Ecology 83 : 1573 – 1579 . Google Scholar Crossref Search ADS WorldCat Sinclair JP , Emlen J , Freeman DC . 2012 . Biased sex ratios in plants: theory and trends . Botanical Review 78 : 63 – 86 . Google Scholar Crossref Search ADS WorldCat Šoltés R , Kubinská A , Janovicová K . 2002 . Extinction risk to the bryophytes in Slovakia, reasons and evaluation . Portugaliae Acta Biologica 20 : 57 – 63 . OpenURL Placeholder Text WorldCat Stark LR . 2002 . Phenology and its repercussions on the reproductive ecology of mosses . The Bryologist 105 : 204 – 218 . Google Scholar Crossref Search ADS WorldCat Stark LR , McLetchie DN , Eppley SM . 2010 . Sex ratios and the shy male hypothesis in the moss Bryum argenteum (Bryaceae) . The Bryologist 113 : 788 – 797 . Google Scholar Crossref Search ADS WorldCat Stark LR , Nichols L , McLetchie ND , Bonine ML . 2005 . Do the sexes of the desert moss Syntrichia caninervis differ in desiccation tolerance? A leaf regeneration assay . International Journal of Plant Science 166 : 21 – 29 . Google Scholar Crossref Search ADS WorldCat StatSoft Inc. 2011 . STATISTICA (data analysis software system), version 10 . Available at: www.statsoft.com. Székely T , Weissing FJ , Komdeur J . 2014 . Adult sex ratio variation: implications for breeding system evolution . Journal of Evolutionary Biology 27 : 1500 – 1512 . Google Scholar Crossref Search ADS PubMed WorldCat Tonnabel J , David P , Pannell JR . 2017 . Sex-specific strategies of resource allocation in response to competition for light in a dioecious plant . Oecologia 185 : 675 – 686 . Google Scholar Crossref Search ADS PubMed WorldCat Vanderpoorten A , Goffinet B . 2009 . Introduction to bryophytes . Cambridge : Cambridge University Press . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Villarreal JC , Renner SS . 2013 . Correlates of monoicy and dioicy in hornworts, the apparent sister group to vascular plants . BMC Evolutionary Biology 13 : 239 . Google Scholar Crossref Search ADS PubMed WorldCat West SA 2009 . Sex allocation . Princeton : Princeton University Press . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Wickett NJ , Mirarab S , Nguyen N , Warnow T , Carpenter E , Matasci N , Ayyampalayam S , Barker MS , Burleigh JG , Gitzendanner MA , Ruhfel BR , Wafula E , Der JP , Graham SW , Mathews S , Melkonian M , Soltis DE , Soltis PS , Miles NW , Rothfels CJ , Pokorny L , Shaw AJ , DeGironimo L , Stevenson DW , Surek B , Villarreal JC , Roure B , Philippe H , dePamphilis CW , Chen T , Deyholos MK , Baucom RS , Kutchan TM , Augustin MM , Wang J , Zhang Y , Tian Z , Yan Z , Wu X , Sun X , Wong GK-S , Leebens-Mack J . 2014 . Phylotranscriptomic analysis of the origin and early diversification of land plants . Proceedings of the National Academy of Sciences USA 111 : E4859 – E4868 . Google Scholar Crossref Search ADS WorldCat Wyatt R . 1985 . Terminology for bryophyte sexuality: toward a unified system . Taxon 34 : 420 – 425 . Google Scholar Crossref Search ADS WorldCat Zhang S , Chen L , Duan B , Korpelainen H , Li C . 2012 . Populus cathayana males exhibit more efficient protective mechanisms than females under drought stress . Forest Ecology and Management 275 : 68 – 78 . Google Scholar Crossref Search ADS WorldCat © 2019 The Linnean Society of London, Botanical Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

, De Albuquerque, Elaine Santiago Brilhante;Tenorio,, Vitor;Braga, João Marcelo, Alvarenga;Vieira, Ricardo, Cardoso

doi: 10.1093/botlinnean/boz081pmid: N/A

Abstract Marantaceae consist of species with asymmetric leaves of two types: those with either a wider left or right half; this asymmetry is related, respectively, to clockwise or counterclockwise convolute vernation. In this study, we analysed whether anatomical differences in the leaf edges, i.e. the anatomical asymmetry, were related to the orientation of the convolute vernation and to the asymmetry of leaf morphology, and whether these differences supported the organization of the clades in the family. Transverse sections of the mid third of the leaf buds expanded to the height of the right and left edges of the blades were prepared for 19 species belonging to 11 genera, using cyto-histological techniques. Anatomical analyses of the blade edges revealed that there is a relationship between morphological asymmetry and anatomical asymmetry that has never before been ascribed to the family. The anatomical data support differences between the arrangements in two of the three Neotropical informal groups. In the Calathea clade, Calathea showed much more similarity with Goeppertia than with Ischnosiphon and Monotagma, since they are the only genera that do not present with anatomical asymmetry. In the Maranta clade, Ctenanthe, Saranthe and Stromanthe appear to be related to one another, as they share strong anatomical asymmetry and fibrous edges. These characteristics, however, are not observed in Myrosma, which in turn is more anatomically similar to Maranta. dexiotropic leaf – laeotropic leaf, medio-lateral asymmetry , Neotropics, Zingiberales INTRODUCTION Marantaceae comprise 31 genera and c. 550 species (Andersson, 1998) with a pantropical distribution, primarily in the Neotropics (Dahlgren, Clifford & Yeo, 1985; Kennedy, Andersson & Hagberg, 1988; Goldberg, 1989; Rzedowski, 2001), where 14 genera and c. 450 species can be found (Prince & Kress, 2006a, b). Brazil has the highest species richness of Marantaceae in the Neotropics (Suárez & Galeano, 1996), and although species occur in all biomes in that country, especially in the Amazon region, they have been neglected in floristic surveys conducted in this region (Costa et al., 2011). In addition, species of Marantaceae in the Neotropics are still under-represented in anatomical studies, with the classical anatomical works of Bertrand (1958) and Tomlinson (1961, 1962, 1969) based primarily on Old World species. The informal groups described by Andersson (1981, 1998) have been modified as new data is used in the systematics of the family, as demonstrated in the work of Andersson & Chase (2001), Prince & Kress (2006a, b) and Borchsenius, Suárez & Prince (2012). The distinction between genera has not always been clear (Andersson, 1998), and Calathea G.Mey., the largest genus in the Neotropics, was recently split into two, with the resurrection of Goeppertia Nees (Borchsenius et al., 2012). Anatomical data on the distribution of silica phytolith idioblasts (Albuquerque, Braga & Vieira, 2013) do not support the relationship of Myrosma L.f. with Saranthe (Regel & Körn.) Eichler, Ctenanthe Eichler and Stromanthe Sond., as shown in the work of Andersson (1981, 1998), Andersson & Chase (2001) and Prince & Kress (2006a, b). Therefore, the anatomical data on leaf edges obtained in this study will be analysed to improve understanding of the association between the genera within the informal groups. The leaves of species of Marantaceae are asymmetric (Rogers, 1984), alternate, distichous, simple and entire (Judd et al., 1999), generally with wide blades, pinnate venation, parallel tertiary veins, petiolate, a distal pulvinus and an open sheath (Delin & Kennedy, 2002). Leaf asymmetry is a marked characteristic of the family and can be of two types, with either the left half or right half of the leaf blade being wider than the other (Tomlinson, 1961; Andersson, 1981). Most of the genera of the Neotropics are considered homotropic (Andersson, 1981), as their species show only one type of leaf asymmetry: that with a wider right side. Species of Ctenanthe and Stromanthe and a few species of Maranta Plum. ex L., however, demonstrate two types of leaf asymmetry, and are therefore considered antitropic (Andersson, 1981). Leaves with a wider right side will be treated here as ‘standard leaves’, as they are observed in all species, whereas those with a wider left side will be called ‘mirror leaves’. As leaf organization is distichous, the standard leaves in antitropic species are all arranged on the same side of the branch, whereas mirror leaves are distributed on the other side, much like congruent and inverted images in a mirror. As such, mirror leaves maintain the same size and shape, and antitropic species always present this mirroring of their leaves on the same branch. This arrangement of leaves on the branch was called ‘antitropous’ by Tomlinson (1961), as opposed to ‘homotropous’, which defines an arrangement exclusively of leaves with a wider right side (Tomlinson, 1961) (which we refer to as ‘standard leaves’). In most species of Marantaceae, leaves show only counterclockwise convolute vernation (Tomlinson, 1961). This type of one-way vernation is termed homodromous (Font Quer, 2001). Species that demonstrate convolute vernation in both directions are considered antidromous (Suemoto, 1953; Font Quer, 2001). During vernation, when the leaf is rolled up (longitudinally) in a counterclockwise direction, the left half of the blade faces outward; when it is rolled up in a clockwise direction, the opposite occurs, with the left half of the blade being on the inside and enveloped by the right side. Therefore, homotropic species have leaves with a wider right side, originating from a single (counterclockwise) vernation direction (=homodromous), whereas antitropic species have two types of leaf asymmetry as a consequence of the presence of two directions of vernation (=antidromous) (Tomlinson, 1961). This relationship between morphological leaf asymmetry and the direction of the vernation was also observed by Gomes (2002) in the leaf blades of Bambusoideae (Poaceae). However, that author also stated that there is a relationship between the direction of the vernation and the anatomical changes between the right and left margins of the blade, i.e. in transverse section, the edge of the left margin is always thin, whereas the edge of the right margin is always thick. Given that the direction of convolute vernation is related to the type of morphological leaf asymmetry present in the genera of Marantaceae (Tomlinson (1961) and that the literature supports the relationship between the asymmetry and the anatomy of the leaf edges in other monocotyledons (Gomes, 2002), in this paper we seek to: (1) describe the comparative anatomy of the left and right leaf edges of 19 species of Marantaceae; (2) determine whether there is a relationship between leaf-edge anatomy and the direction of convolute vernation and between leaf-edge anatomy and the type of morphological asymmetry in the leaves of the family and (3) determine whether such anatomical data are potentially useful for the taxonomy and systematics of the family. MATERIAL AND METHODS For this study, we collected leaves from 19 species belonging to 11 out of the 13 genera of Marantaceae found in Brazil (in the Distrito Federal and the states of Bahia, Espírito-Santo, Mato Grosso do Sul, Rio de Janeiro and Pará). These species are listed in Table 1. The infrageneric classification of Maranta used here is based on the work of Schumann (1902), and the organization of the species in the tables and photographic illustrations presented here follows their grouping into the suprageneric clades (Calathea, Donax and Maranta) proposed by Andersson & Chase (2001). Table 1. Analysed species of Marantaceae, organized by informal clade, with authors and voucher information Clade . Species . Voucher . Calathea lutea (Aubl.) E.Mey. ex Schult. J. M. A. Braga 2009/E1 (RB) Goeppertia cylindrica (Roscoe) Borchs. & S.Suárez J. M. A. Braga & E. S. B. Albuquerque 2007/E1 (RB) Goeppertia mansoi (Körn.) Borchs. & S.Suárez J. M. A. Braga 2005/E1 (RB, R) Calathea clade Goeppertia wiotii (E.Morren) Borchs. & S.Suárez J. M. A. Braga 2003/E34 (RB) Goeppertia zingiberina (Körn.) Borchs. & S.Suárez J. M. A. Braga 2007/E33 (RB) Ischnosiphon ovatus Körn. J. M. A. Braga & E. S. B. Albuquerque 2007/E2 (RB) Monotagma plurispicatum (Körn.) K.Schum. J. M. A. Braga 2005/E67 (RB) Donax clade Thalia geniculata L. J. M. A. Braga 2007/E12 (RB, R) Ctenanthe sp. nov. J. M. A. Braga 8501 (RB, R) Ctenanthe setosa (Roscoe) Eichler E. S. B. Albuquerque 40830 (RFA) Koernickanthe orbiculata (Körn.) L.Andersson J. M. A. Braga 2007/E13 (RB) Maranta aff. bracteosa Petersen J. M. A. Braga 2007/E15 (RB, R) Maranta clade Maranta leuconeura E.Morren var. leuconeura. J. M. A. Braga & E. S. B. Albuquerque 2007/E4 (RB, R) Maranta subterranea J.M.A.Braga J. M. A. Braga et al. 5299 (RB) Myrosma cannifolia L.f. J. M. A. Braga 2007/E14 (RB) Saranthe composita (K.Koch) K.Schum. J. M. A. Braga 2009/E5 (RB) Saranthe leptostachya (Regel & Körn.) Eichler J. M. A. Braga & E. S. B. Albuquerque 2007/E3 (RB, R) Stromanthe porteana Gris J. M. A. Braga 2009/E6 (RB) Stromanthe schottiana (Körn.) Eichler J. M. A. Braga 2007/E11 (RB, R) Clade . Species . Voucher . Calathea lutea (Aubl.) E.Mey. ex Schult. J. M. A. Braga 2009/E1 (RB) Goeppertia cylindrica (Roscoe) Borchs. & S.Suárez J. M. A. Braga & E. S. B. Albuquerque 2007/E1 (RB) Goeppertia mansoi (Körn.) Borchs. & S.Suárez J. M. A. Braga 2005/E1 (RB, R) Calathea clade Goeppertia wiotii (E.Morren) Borchs. & S.Suárez J. M. A. Braga 2003/E34 (RB) Goeppertia zingiberina (Körn.) Borchs. & S.Suárez J. M. A. Braga 2007/E33 (RB) Ischnosiphon ovatus Körn. J. M. A. Braga & E. S. B. Albuquerque 2007/E2 (RB) Monotagma plurispicatum (Körn.) K.Schum. J. M. A. Braga 2005/E67 (RB) Donax clade Thalia geniculata L. J. M. A. Braga 2007/E12 (RB, R) Ctenanthe sp. nov. J. M. A. Braga 8501 (RB, R) Ctenanthe setosa (Roscoe) Eichler E. S. B. Albuquerque 40830 (RFA) Koernickanthe orbiculata (Körn.) L.Andersson J. M. A. Braga 2007/E13 (RB) Maranta aff. bracteosa Petersen J. M. A. Braga 2007/E15 (RB, R) Maranta clade Maranta leuconeura E.Morren var. leuconeura. J. M. A. Braga & E. S. B. Albuquerque 2007/E4 (RB, R) Maranta subterranea J.M.A.Braga J. M. A. Braga et al. 5299 (RB) Myrosma cannifolia L.f. J. M. A. Braga 2007/E14 (RB) Saranthe composita (K.Koch) K.Schum. J. M. A. Braga 2009/E5 (RB) Saranthe leptostachya (Regel & Körn.) Eichler J. M. A. Braga & E. S. B. Albuquerque 2007/E3 (RB, R) Stromanthe porteana Gris J. M. A. Braga 2009/E6 (RB) Stromanthe schottiana (Körn.) Eichler J. M. A. Braga 2007/E11 (RB, R) Open in new tab Table 1. Analysed species of Marantaceae, organized by informal clade, with authors and voucher information Clade . Species . Voucher . Calathea lutea (Aubl.) E.Mey. ex Schult. J. M. A. Braga 2009/E1 (RB) Goeppertia cylindrica (Roscoe) Borchs. & S.Suárez J. M. A. Braga & E. S. B. Albuquerque 2007/E1 (RB) Goeppertia mansoi (Körn.) Borchs. & S.Suárez J. M. A. Braga 2005/E1 (RB, R) Calathea clade Goeppertia wiotii (E.Morren) Borchs. & S.Suárez J. M. A. Braga 2003/E34 (RB) Goeppertia zingiberina (Körn.) Borchs. & S.Suárez J. M. A. Braga 2007/E33 (RB) Ischnosiphon ovatus Körn. J. M. A. Braga & E. S. B. Albuquerque 2007/E2 (RB) Monotagma plurispicatum (Körn.) K.Schum. J. M. A. Braga 2005/E67 (RB) Donax clade Thalia geniculata L. J. M. A. Braga 2007/E12 (RB, R) Ctenanthe sp. nov. J. M. A. Braga 8501 (RB, R) Ctenanthe setosa (Roscoe) Eichler E. S. B. Albuquerque 40830 (RFA) Koernickanthe orbiculata (Körn.) L.Andersson J. M. A. Braga 2007/E13 (RB) Maranta aff. bracteosa Petersen J. M. A. Braga 2007/E15 (RB, R) Maranta clade Maranta leuconeura E.Morren var. leuconeura. J. M. A. Braga & E. S. B. Albuquerque 2007/E4 (RB, R) Maranta subterranea J.M.A.Braga J. M. A. Braga et al. 5299 (RB) Myrosma cannifolia L.f. J. M. A. Braga 2007/E14 (RB) Saranthe composita (K.Koch) K.Schum. J. M. A. Braga 2009/E5 (RB) Saranthe leptostachya (Regel & Körn.) Eichler J. M. A. Braga & E. S. B. Albuquerque 2007/E3 (RB, R) Stromanthe porteana Gris J. M. A. Braga 2009/E6 (RB) Stromanthe schottiana (Körn.) Eichler J. M. A. Braga 2007/E11 (RB, R) Clade . Species . Voucher . Calathea lutea (Aubl.) E.Mey. ex Schult. J. M. A. Braga 2009/E1 (RB) Goeppertia cylindrica (Roscoe) Borchs. & S.Suárez J. M. A. Braga & E. S. B. Albuquerque 2007/E1 (RB) Goeppertia mansoi (Körn.) Borchs. & S.Suárez J. M. A. Braga 2005/E1 (RB, R) Calathea clade Goeppertia wiotii (E.Morren) Borchs. & S.Suárez J. M. A. Braga 2003/E34 (RB) Goeppertia zingiberina (Körn.) Borchs. & S.Suárez J. M. A. Braga 2007/E33 (RB) Ischnosiphon ovatus Körn. J. M. A. Braga & E. S. B. Albuquerque 2007/E2 (RB) Monotagma plurispicatum (Körn.) K.Schum. J. M. A. Braga 2005/E67 (RB) Donax clade Thalia geniculata L. J. M. A. Braga 2007/E12 (RB, R) Ctenanthe sp. nov. J. M. A. Braga 8501 (RB, R) Ctenanthe setosa (Roscoe) Eichler E. S. B. Albuquerque 40830 (RFA) Koernickanthe orbiculata (Körn.) L.Andersson J. M. A. Braga 2007/E13 (RB) Maranta aff. bracteosa Petersen J. M. A. Braga 2007/E15 (RB, R) Maranta clade Maranta leuconeura E.Morren var. leuconeura. J. M. A. Braga & E. S. B. Albuquerque 2007/E4 (RB, R) Maranta subterranea J.M.A.Braga J. M. A. Braga et al. 5299 (RB) Myrosma cannifolia L.f. J. M. A. Braga 2007/E14 (RB) Saranthe composita (K.Koch) K.Schum. J. M. A. Braga 2009/E5 (RB) Saranthe leptostachya (Regel & Körn.) Eichler J. M. A. Braga & E. S. B. Albuquerque 2007/E3 (RB, R) Stromanthe porteana Gris J. M. A. Braga 2009/E6 (RB) Stromanthe schottiana (Körn.) Eichler J. M. A. Braga 2007/E11 (RB, R) Open in new tab The fully expanded leaves of three individuals of each species were collected for the anatomical studies. The specimens were fixed in FPA (10 ml formaldehyde, 5 ml propionic acid, 50 ml 95% ethanol and 35 ml distilled water) (Ruzin, 1999) and subsequently preserved in 70° GL ethanol (Johansen, 1940). The same procedure was employed with longitudinally rolled up young leaves. Sections of the mid third of the expanded and rolled up leaf blades were prepared in order to analyse their right and left edges. The specimens were sectioned in the transverse plane using a Ranvier microtome. Semi-permanent slides were prepared and stained with a mixture of Safranin-Astra blue (Bukatsch, 1972). Some of the samples were soaked in a solution of 20% P. M. 1500 polyethylene glycol (PEG) in water and left to dry at 60 °C until all of the water evaporated. The samples were then embedded in pure PEG and transversely sectioned in a sledge microtome (with a thickness of 10–20 µm) (Burger & Richter, 1991). The sections were cleared in sodium hypochlorite, neutralized in acidified water, washed in distilled water, stained using Safranin-Astra blue (Bukatsch, 1972) and mounted as permanent slides. RESULTS The leaf blades of all species studied were asymmetrical, with one half of the blade always wider than the other (Fig. 1A–C). Among the 19 species studied here, 15 were homotropic, i.e. they produced only standard leaves, with the right half of the blade wider than the left (Fig. 1A). Only four species were antitropic, having both standard leaves (with a wider right side) and mirror leaves (with a wider left side): Ctenanthe sp. nov., Ctenanthe setosa (Roscoe) Eichler (Fig. 1C), Stromanthe porteana Gris and S. schottiana (Körn.) Eichler (Fig. 1B) (Table 2). Table 2. Types of asymmetric leaves and anatomical characteristics of its edges found in the 19 species of Marantaceae studied here Informal groups . N . Species studied . Leaves . Convolute vernation . Standard leaf . Mirror leaf . Anatomical difference between the right and left edge . Occurrence of fibres along the edges . . . . . . Curvature of the outer edge (left) . Curvature of the inner edge (right) . Curvature of the inner edge (left) . Curvature of the outer edge (right) . . . . . . . . . . . . . . LE . RE . Calathea clade 1. Calathea lutea Homotropic Homodromous Straight Upward − − − − − 2. Goeppertia cylindrica Homotropic Homodromous Upward Upward − − − − − 3. Goeppertia mansoi Homotropic Homodromous Downward Downward − − − − − 4. Goeppertia wiotii Homotropic Homodromous Downward Downward − − − − − 5. Goeppertia zingiberina Homotropic Homodromous Downward Downward − − +/− − − 6. Ischnosiphon ovatus Homotropic Homodromous Downward Straight − − + − − 7. Monotagma plurispicatum Homotropic Homodromous Straight Upward − − + − − Donax clade 8. Thalia geniculata Homotropic Homodromous Slightly Upward Upward − − + − − Maranta clade 9. Ctenanthe sp. nov. Antitropic Antidromous Straight Upward Upward Straight + +* +* 10. Ctenanthe setosa Antitropic Antidromous Straight Upward Upward Slightly Downward + +* +* 11. Koernickanthe orbiculata Homotropic Homodromous Straight Upward − − + − − 12. Maranta aff. bracteosa Homotropic Homodromous Downward Upward − − + − − 13. Maranta leuconeura Homotropic Homodromous Straight Upward − − + − − 14. Maranta subterranea Homotropic Homodromous Downward Upward − − + − − 15. Myrosma cannifolia Homotropic Homodromous Upward Upward − − +/− − − 16. Saranthe composita Homotropic Homodromous Straight Upward − − + + + 17. Saranthe leptostachya Homotropic Homodromous Downward Downward − − + + + 18. Stromanthe porteana Antitropic Antidromous Straight Upward Slightly Upward Straight + +* +* 19. Stromanthe schottiana Antitropic Antidromous Straight Upward Upward Straight + +* +* Informal groups . N . Species studied . Leaves . Convolute vernation . Standard leaf . Mirror leaf . Anatomical difference between the right and left edge . Occurrence of fibres along the edges . . . . . . Curvature of the outer edge (left) . Curvature of the inner edge (right) . Curvature of the inner edge (left) . Curvature of the outer edge (right) . . . . . . . . . . . . . . LE . RE . Calathea clade 1. Calathea lutea Homotropic Homodromous Straight Upward − − − − − 2. Goeppertia cylindrica Homotropic Homodromous Upward Upward − − − − − 3. Goeppertia mansoi Homotropic Homodromous Downward Downward − − − − − 4. Goeppertia wiotii Homotropic Homodromous Downward Downward − − − − − 5. Goeppertia zingiberina Homotropic Homodromous Downward Downward − − +/− − − 6. Ischnosiphon ovatus Homotropic Homodromous Downward Straight − − + − − 7. Monotagma plurispicatum Homotropic Homodromous Straight Upward − − + − − Donax clade 8. Thalia geniculata Homotropic Homodromous Slightly Upward Upward − − + − − Maranta clade 9. Ctenanthe sp. nov. Antitropic Antidromous Straight Upward Upward Straight + +* +* 10. Ctenanthe setosa Antitropic Antidromous Straight Upward Upward Slightly Downward + +* +* 11. Koernickanthe orbiculata Homotropic Homodromous Straight Upward − − + − − 12. Maranta aff. bracteosa Homotropic Homodromous Downward Upward − − + − − 13. Maranta leuconeura Homotropic Homodromous Straight Upward − − + − − 14. Maranta subterranea Homotropic Homodromous Downward Upward − − + − − 15. Myrosma cannifolia Homotropic Homodromous Upward Upward − − +/− − − 16. Saranthe composita Homotropic Homodromous Straight Upward − − + + + 17. Saranthe leptostachya Homotropic Homodromous Downward Downward − − + + + 18. Stromanthe porteana Antitropic Antidromous Straight Upward Slightly Upward Straight + +* +* 19. Stromanthe schottiana Antitropic Antidromous Straight Upward Upward Straight + +* +* (+) presence; (−) absence; (+/−) slight difference; (+*) presence of fibres in both the standard leaves and mirror leaves; LE – left edge; RE – right edge. Open in new tab Table 2. Types of asymmetric leaves and anatomical characteristics of its edges found in the 19 species of Marantaceae studied here Informal groups . N . Species studied . Leaves . Convolute vernation . Standard leaf . Mirror leaf . Anatomical difference between the right and left edge . Occurrence of fibres along the edges . . . . . . Curvature of the outer edge (left) . Curvature of the inner edge (right) . Curvature of the inner edge (left) . Curvature of the outer edge (right) . . . . . . . . . . . . . . LE . RE . Calathea clade 1. Calathea lutea Homotropic Homodromous Straight Upward − − − − − 2. Goeppertia cylindrica Homotropic Homodromous Upward Upward − − − − − 3. Goeppertia mansoi Homotropic Homodromous Downward Downward − − − − − 4. Goeppertia wiotii Homotropic Homodromous Downward Downward − − − − − 5. Goeppertia zingiberina Homotropic Homodromous Downward Downward − − +/− − − 6. Ischnosiphon ovatus Homotropic Homodromous Downward Straight − − + − − 7. Monotagma plurispicatum Homotropic Homodromous Straight Upward − − + − − Donax clade 8. Thalia geniculata Homotropic Homodromous Slightly Upward Upward − − + − − Maranta clade 9. Ctenanthe sp. nov. Antitropic Antidromous Straight Upward Upward Straight + +* +* 10. Ctenanthe setosa Antitropic Antidromous Straight Upward Upward Slightly Downward + +* +* 11. Koernickanthe orbiculata Homotropic Homodromous Straight Upward − − + − − 12. Maranta aff. bracteosa Homotropic Homodromous Downward Upward − − + − − 13. Maranta leuconeura Homotropic Homodromous Straight Upward − − + − − 14. Maranta subterranea Homotropic Homodromous Downward Upward − − + − − 15. Myrosma cannifolia Homotropic Homodromous Upward Upward − − +/− − − 16. Saranthe composita Homotropic Homodromous Straight Upward − − + + + 17. Saranthe leptostachya Homotropic Homodromous Downward Downward − − + + + 18. Stromanthe porteana Antitropic Antidromous Straight Upward Slightly Upward Straight + +* +* 19. Stromanthe schottiana Antitropic Antidromous Straight Upward Upward Straight + +* +* Informal groups . N . Species studied . Leaves . Convolute vernation . Standard leaf . Mirror leaf . Anatomical difference between the right and left edge . Occurrence of fibres along the edges . . . . . . Curvature of the outer edge (left) . Curvature of the inner edge (right) . Curvature of the inner edge (left) . Curvature of the outer edge (right) . . . . . . . . . . . . . . LE . RE . Calathea clade 1. Calathea lutea Homotropic Homodromous Straight Upward − − − − − 2. Goeppertia cylindrica Homotropic Homodromous Upward Upward − − − − − 3. Goeppertia mansoi Homotropic Homodromous Downward Downward − − − − − 4. Goeppertia wiotii Homotropic Homodromous Downward Downward − − − − − 5. Goeppertia zingiberina Homotropic Homodromous Downward Downward − − +/− − − 6. Ischnosiphon ovatus Homotropic Homodromous Downward Straight − − + − − 7. Monotagma plurispicatum Homotropic Homodromous Straight Upward − − + − − Donax clade 8. Thalia geniculata Homotropic Homodromous Slightly Upward Upward − − + − − Maranta clade 9. Ctenanthe sp. nov. Antitropic Antidromous Straight Upward Upward Straight + +* +* 10. Ctenanthe setosa Antitropic Antidromous Straight Upward Upward Slightly Downward + +* +* 11. Koernickanthe orbiculata Homotropic Homodromous Straight Upward − − + − − 12. Maranta aff. bracteosa Homotropic Homodromous Downward Upward − − + − − 13. Maranta leuconeura Homotropic Homodromous Straight Upward − − + − − 14. Maranta subterranea Homotropic Homodromous Downward Upward − − + − − 15. Myrosma cannifolia Homotropic Homodromous Upward Upward − − +/− − − 16. Saranthe composita Homotropic Homodromous Straight Upward − − + + + 17. Saranthe leptostachya Homotropic Homodromous Downward Downward − − + + + 18. Stromanthe porteana Antitropic Antidromous Straight Upward Slightly Upward Straight + +* +* 19. Stromanthe schottiana Antitropic Antidromous Straight Upward Upward Straight + +* +* (+) presence; (−) absence; (+/−) slight difference; (+*) presence of fibres in both the standard leaves and mirror leaves; LE – left edge; RE – right edge. Open in new tab Figure 1. Open in new tabDownload slide A, Standard leaf (Calathea cylindrica) with the right half (RH) of the blade larger than the left (LH) common to all Marantaceae (morphological asymmetry). B, C, Antitropic species with standard leaves (SL) and mirror leaves (ML) on the same individual (B, Stromanthe schottiana; C, Ctenanthe setosa). Note the left edge oriented toward the interior of the mirror leaf (C) during convolute vernation. D, Cross-section of the blade of a mirror leaf, demonstrating clockwise convolute vernation (Ctenanthe setosa). E, Cross-section of the blade of the standard leaf, demonstrating counterclockwise convolute vernation (Maranta subterranea). In D and E, observe that the internal edge is always thicker, independent of the direction of the vernation (anatomical asymmetry). AF: adaxial face; EE: external edge; IE: internal edge; LE: left edge; LH: left half; LS: leaf sheath; ML: mirror leaf; RE: right edge; RH: right half; SL: standard leaf. Scale bars: D – 100 µm; E – 350 µm. Figure 1. Open in new tabDownload slide A, Standard leaf (Calathea cylindrica) with the right half (RH) of the blade larger than the left (LH) common to all Marantaceae (morphological asymmetry). B, C, Antitropic species with standard leaves (SL) and mirror leaves (ML) on the same individual (B, Stromanthe schottiana; C, Ctenanthe setosa). Note the left edge oriented toward the interior of the mirror leaf (C) during convolute vernation. D, Cross-section of the blade of a mirror leaf, demonstrating clockwise convolute vernation (Ctenanthe setosa). E, Cross-section of the blade of the standard leaf, demonstrating counterclockwise convolute vernation (Maranta subterranea). In D and E, observe that the internal edge is always thicker, independent of the direction of the vernation (anatomical asymmetry). AF: adaxial face; EE: external edge; IE: internal edge; LE: left edge; LH: left half; LS: leaf sheath; ML: mirror leaf; RE: right edge; RH: right half; SL: standard leaf. Scale bars: D – 100 µm; E – 350 µm. When observed in transverse section, the leaf blades of the homotropic species demonstrated convolute vernation in a counterclockwise sense (Fig. 1E), whereas antitropic species demonstrated both clockwise (Fig. 1D) and counterclockwise vernation (Fig. 1E). Regardless of the direction of vernation, the leaf blades demonstrated, in general, a thicker internal edge due to a greater number of cell layers and/or larger cells when compared to the external edge (Figs 1D, E, 2I, J, 3–7), resulting in anatomical asymmetry. Therefore, due to the counterclockwise vernation of homotropic species (Fig. 1E), the thickest edge of the open standard leaf is always on the right, and the thinnest edge is always on the left (Figs 1E, 2I, J, 3–4, 6, 7). The same occurs with antitropic species due to the counterclockwise vernation of standard leaves (Figs 5A–D, G, H, K, L), although they also have mirror leaves showing clockwise vernation (Fig. 1D), which in this latter case results in their left edges being thicker than their right edges (Figs 1D, 5E, F, I, J). Therefore, homotropic species, in addition to having standard leaves with right blades wider than the left (Fig. 1A), also present thicker edges on these right blades in the transverse section (Fig. 1E). The same occurs in antitropic species with standard leaves, although their mirror leaves have a wider left side and a thicker left edge (when compared to the right edge) in transverse sections (Figs 5E, F, I, J). Ctenanthe setosa, Ctenanthe sp. nov., Stromanthe porteana and S. schottiana fall in the latter group, as they have both mirror leaves and standard leaves (Fig. 5). The relationships between this leaf-edge asymmetry and morphological asymmetry in the species studied between leaf-edge asymmetry and vernation are summarized in Table 3. Table 3. Relationship of homodromous and antidromous vernation to leaf asymmetry and edge thickness in the species of Marantaceae investigated here Types of genera Convolute vernation Sense of vernation Type of morphological asymmetry of leaves Position of widest half in leaf bud Type of anatomical asymmetry of leaves Homotropic (most genera) Homodromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section)* Antitropic (Ctenanthe and Stromanthe) Antidromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section) Clockwise Left half wider (mirror leaf) Internal (left half) Left edge thicker (mirror leaf in transverse section) Types of genera Convolute vernation Sense of vernation Type of morphological asymmetry of leaves Position of widest half in leaf bud Type of anatomical asymmetry of leaves Homotropic (most genera) Homodromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section)* Antitropic (Ctenanthe and Stromanthe) Antidromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section) Clockwise Left half wider (mirror leaf) Internal (left half) Left edge thicker (mirror leaf in transverse section) *Except Calathea lutea, Goeppertia cylindrica, G. mansoi and G. wiotii. Open in new tab Table 3. Relationship of homodromous and antidromous vernation to leaf asymmetry and edge thickness in the species of Marantaceae investigated here Types of genera Convolute vernation Sense of vernation Type of morphological asymmetry of leaves Position of widest half in leaf bud Type of anatomical asymmetry of leaves Homotropic (most genera) Homodromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section)* Antitropic (Ctenanthe and Stromanthe) Antidromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section) Clockwise Left half wider (mirror leaf) Internal (left half) Left edge thicker (mirror leaf in transverse section) Types of genera Convolute vernation Sense of vernation Type of morphological asymmetry of leaves Position of widest half in leaf bud Type of anatomical asymmetry of leaves Homotropic (most genera) Homodromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section)* Antitropic (Ctenanthe and Stromanthe) Antidromous Counter clockwise Right half wider (standard leaf) Internal (right half) Right edge thicker (standard leaf in transverse section) Clockwise Left half wider (mirror leaf) Internal (left half) Left edge thicker (mirror leaf in transverse section) *Except Calathea lutea, Goeppertia cylindrica, G. mansoi and G. wiotii. Open in new tab Figure 2. Open in new tabDownload slide A, C, E, G, I, Transverse sections of left edges and B, D, F, H, J, right edges of the standard-leaf blades of the species of the Calathea clade. Note the edge space filled with hypodermal cells in Calathea lutea (A, B), Goeppertia cylindrica (C, D) and G. mansoi (E, F), the presence of trichomes (arrows) in B, C, D and G. zingiberina (I, J) and the anatomical similarity between the left and right edges (A–B; C–D; E–F; G–H: G. wiotti; I–J). VB – vascular bundle. Scale bars: 60 µm. Figure 2. Open in new tabDownload slide A, C, E, G, I, Transverse sections of left edges and B, D, F, H, J, right edges of the standard-leaf blades of the species of the Calathea clade. Note the edge space filled with hypodermal cells in Calathea lutea (A, B), Goeppertia cylindrica (C, D) and G. mansoi (E, F), the presence of trichomes (arrows) in B, C, D and G. zingiberina (I, J) and the anatomical similarity between the left and right edges (A–B; C–D; E–F; G–H: G. wiotti; I–J). VB – vascular bundle. Scale bars: 60 µm. Figure 3. Open in new tabDownload slide A, C, Transverse sections of left edges and B, D, right edges of the blades of standard-leaf species of the Calathea clade. Note the anatomical differences in thickness between the edges of I. ovatus (A, B) and M. plurispicatum (C, D); and the presence of trichomes (D). VB – vascular bundle. Scale bars: A, B – 60 µm; C, D – 75 µm. Figure 3. Open in new tabDownload slide A, C, Transverse sections of left edges and B, D, right edges of the blades of standard-leaf species of the Calathea clade. Note the anatomical differences in thickness between the edges of I. ovatus (A, B) and M. plurispicatum (C, D); and the presence of trichomes (D). VB – vascular bundle. Scale bars: A, B – 60 µm; C, D – 75 µm. Figure 4. Open in new tabDownload slide A, Transverse sections of the left edge and B, right edge of the standard-leaf blade of Thalia geniculata (Donax clade). Note the anatomical differences and the thickness of both edges. VB – vascular bundle. Scale bars: 75 µm. Figure 4. Open in new tabDownload slide A, Transverse sections of the left edge and B, right edge of the standard-leaf blade of Thalia geniculata (Donax clade). Note the anatomical differences and the thickness of both edges. VB – vascular bundle. Scale bars: 75 µm. Figure 5. Open in new tabDownload slide A, C, E, G, I, K, Transverse sections of left edges, and B, D, F, H, J, L, right edges (of the antitropic species in the Maranta clade). Note the anatomical differences between the two edges of each species and that the right edge (RE) is thicker on the blade of the standard leaf (SL) in B (Ctenanthe sp. nov.), D (Ctenanthe setosa), H (Stromanthe porteana) and L (Stromanthe schottiana); and that the left edge (LE) is thicker on the blade of the mirror leaf (ML) in E (C. setosa) and I (S. porteana). Ctenanthe sp. nov. (A, B, standard leaf); C. setosa (C, D, standard leaf; E, F, mirror leaf); Stromanthe porteana (G, H, standard leaf; I, J, mirror leaf); S. schottiana (K, L, standard leaf). EE – external edge; FB – fibre strands; FS – fibrous sheath; IE – internal edge; VB – vascular bundle. Scale bars: 60 µm. Figure 5. Open in new tabDownload slide A, C, E, G, I, K, Transverse sections of left edges, and B, D, F, H, J, L, right edges (of the antitropic species in the Maranta clade). Note the anatomical differences between the two edges of each species and that the right edge (RE) is thicker on the blade of the standard leaf (SL) in B (Ctenanthe sp. nov.), D (Ctenanthe setosa), H (Stromanthe porteana) and L (Stromanthe schottiana); and that the left edge (LE) is thicker on the blade of the mirror leaf (ML) in E (C. setosa) and I (S. porteana). Ctenanthe sp. nov. (A, B, standard leaf); C. setosa (C, D, standard leaf; E, F, mirror leaf); Stromanthe porteana (G, H, standard leaf; I, J, mirror leaf); S. schottiana (K, L, standard leaf). EE – external edge; FB – fibre strands; FS – fibrous sheath; IE – internal edge; VB – vascular bundle. Scale bars: 60 µm. Figure 6. Open in new tabDownload slide A, C, E, Transverse sections of left edges and B, D, F, right edges of the blades of standard leaves of homotropic species in the Maranta clade. Note the anatomical differences and differences in thickness between the leaf edges of Myrosma cannifolia (A, B), Saranthe composita (C, D), and Saranthe leptostachya (E, F). FB – fibre strands; VB – vascular bundle. Scale bars: 50 μm. Figure 6. Open in new tabDownload slide A, C, E, Transverse sections of left edges and B, D, F, right edges of the blades of standard leaves of homotropic species in the Maranta clade. Note the anatomical differences and differences in thickness between the leaf edges of Myrosma cannifolia (A, B), Saranthe composita (C, D), and Saranthe leptostachya (E, F). FB – fibre strands; VB – vascular bundle. Scale bars: 50 μm. Figure 7. Open in new tabDownload slide A, C, E, G, Transverse sections of left edges and B, D, F, H, right edges of the blades of standard leaves in the Maranta clade. Note the anatomical differences, differences in thickness and curvature of the two edges of K. orbiculata (A, B), Maranta aff. bracteosa (C, D), M. leuconeura (E, F) and M. subterranea (G, H). FB - fibre strands; VB - vascular bundle. Scale bars: 50 μm. Figure 7. Open in new tabDownload slide A, C, E, G, Transverse sections of left edges and B, D, F, H, right edges of the blades of standard leaves in the Maranta clade. Note the anatomical differences, differences in thickness and curvature of the two edges of K. orbiculata (A, B), Maranta aff. bracteosa (C, D), M. leuconeura (E, F) and M. subterranea (G, H). FB - fibre strands; VB - vascular bundle. Scale bars: 50 μm. Leaf edges can be straight or curved slightly upwards or downwards in transverse section. The edge that is rolled up inward during vernation tends to be bent upwards in both standard leaves and mirror leaves. In the case of mirror leaves of antitropic species, the left edge, which is the thickest and most internal, always curves upwards (Fig. 5E, I; Table 2). In standard leaves (in both homotropic and antitropic species), the right leaf margins are generally curved upwards, as in Calathea lutea (Aubl.) Schult. (Fig. 2B), Goeppertia cylindrica (Roscoe) Borchs. & S.Suárez (Fig. 2D), Monotagma plurispicatum K.Schum. (Fig. 3D), Thalia geniculata L. (Fig. 4B), Ctenanthe sp. nov. (Fig. 5B), C. setosa (Fig. 5D), S. porteana (Fig. 5H), S. schottiana (Fig. 5L), Myrosma cannifolia L.f. (Fig. 7B), Koernickanthe orbiculata (Körn.) L.Andersson (Fig. 6B), Maranta spp. (Fig. 6D, F, H) and S. composita (K.Koch) K.Schum (Fig. 7D). In the other taxa, the leaf edges are curved downwards or remain straight. The left edge is generally curved downwards in Goeppertia mansoi (Körn.) Borchs. & S.Suárez (Fig. 2E), G. wioti (E.Morren) Borchs. & S.Suárez (Fig. 2G), G. zingiberina (Körn.) Borchs. & S.Suárez (Fig. 2I), Ischnosiphon ovatus Körn.. (Fig. 3A), Maranta aff. bracteosa Petersen (Fig. 6C), M. subterranea J.M.A.Braga (Fig. 6G) and Saranthe leptostachya (Regel & Körn.) Eichler (Fig. 7E). In the other species, it is curved upwards or remains straight. The leaf epidermis in the edge region in transverse section (of both faces), which may contain stomata, is formed by cells with straight external periclinal walls that are slightly projected. Sparse unicellular tector trichomes are present on the leaves of G. cylindrica (Fig. 2C, D), G. zingiberina (Fig. 2I, J), I. ovatus, M. plurispicatum (Fig. 3D), S. porteana (Fig. 5H), K. orbiculata (Fig. 6B), Maranta aff. bracteosa (Fig. 6C) and M. leuconeura. A layer of rounded to squarish hypodermal cells can be seen below the epidermis (Figs 1D, E, 2–7). The number of layers in the hypodermis, on the right edge, can be up to two in I. ovatus (Fig. 3B), in both adaxial and abaxial faces. The size of the cells in these layers may also be larger on the internal edge (in standard and mirror leaves) and the increase in thickness of this edge can be seen in Figs 3D, 5B, D, E, L, 6B, D, F, H, 7D. The mesophyll of the species studied here was of the dorsiventral type, although cells near the edges tend to become isodiametric, and it was impossible to differentiate between the palisade and spongy layers. Thalia geniculata was the only species that demonstrated an isolateral mesophyll, although the parenchymatous cells became isodiametric near the right leaf margin, whereas those on the left edge remained in a palisade form (Fig. 4A, B). In addition, the thickness of the mesophyll tends to be greater at the internal edge of certain species, contributing to the increase in thickness of the edge (Figs 2J, 3B, D, 4B, 5D, E, H, I, L). The vascular bundles were of the collateral type, with a fibrous sheath. In some species, the vascular bundles may be quite close to the margin and larger on the internal edge than the external edge, as in I. ovatus (Fig. 3B), M. plurispicatum (Fig. 3D), T. geniculata (Fig. 4B) and S. leptostachya (Fig. 7F). In C. lutea (Fig. 2A, B), G. cylindrica and G. mansoi (Fig. 2C–F), the leaf edges are quite narrow, being composed only of epidermal and hypodermal cells and not showing anatomical asymmetry between them. In some species, the edge is filled with fibre strands and at times even some small hypodermal cells [as in Ctenanthe sp. nov. (Fig. 5A, B), C. setosa (Fig. 5C–F), Stromanthe porteana (Fig. 5G–J), Saranthe composita (Fig. 7C, D), S. leptostachya (Fig. 7E, F)] or with fibrous vascular bundle sheaths in addition to a small number of hypodermal cells (as in Stromanthe schottiana) (Fig. 5K, L). The histological organizations of the right and left edges of the standard and mirror leaves are summarized in Tables 2, 4. Table 4. Proportion of the different types of leaf curvature in the species of Marantaceae investigated here . Standard leaves . Mirror leaves . . External edge (left) . Internal edge (right) . Internal edge (left) . External edge (right) . . . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Curved upward 3 16% 14 74% 4 100% 0 – Straight 9 47% 1 5% 0 – 4 100% Curved downward 7 37% 4 21% 0 – 0 – Species total 19 100% 19 100% 4 100% 4 100% . Standard leaves . Mirror leaves . . External edge (left) . Internal edge (right) . Internal edge (left) . External edge (right) . . . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Curved upward 3 16% 14 74% 4 100% 0 – Straight 9 47% 1 5% 0 – 4 100% Curved downward 7 37% 4 21% 0 – 0 – Species total 19 100% 19 100% 4 100% 4 100% Open in new tab Table 4. Proportion of the different types of leaf curvature in the species of Marantaceae investigated here . Standard leaves . Mirror leaves . . External edge (left) . Internal edge (right) . Internal edge (left) . External edge (right) . . . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Curved upward 3 16% 14 74% 4 100% 0 – Straight 9 47% 1 5% 0 – 4 100% Curved downward 7 37% 4 21% 0 – 0 – Species total 19 100% 19 100% 4 100% 4 100% . Standard leaves . Mirror leaves . . External edge (left) . Internal edge (right) . Internal edge (left) . External edge (right) . . . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Number of species . Proportion . Curved upward 3 16% 14 74% 4 100% 0 – Straight 9 47% 1 5% 0 – 4 100% Curved downward 7 37% 4 21% 0 – 0 – Species total 19 100% 19 100% 4 100% 4 100% Open in new tab DISCUSSION The morphological asymmetry of leaves in Marantaceae studied here represents a conservative characteristic intimately linked to their convolute vernation, as also reported by Tomlinson (1961). Körn (2006) studied leaves of species of Araceae, Marantaceae and Euphorbiaceae and reported that their asymmetry was due to the delayed growth of one side of the leaf during convolute vernation. We likewise observed differences in the organizational anatomy between the right and left edges of standard and mirror leaves. Our results also demonstrate that there is a relationship between edge anatomy and both the orientation of vernation and the morphological asymmetry of the leaf. Anatomical differences between the leaf margins were reported by Gomes (2002) for species of Merostachys Spreng. (Poaceae); Chen et al. (2007) and Yuan, Li & Peng (2015) likewise observed differences in thickness of the two halves of the asymmetric leaves of rice (Oryza sativa L.). In all of these studies, as in our study, the wider side of the leaf was also the thicker. The species studied here were distributed among three clades (the informal Maranta, Calathea and Donax groups; Prince & Kress, 2006a) with each demonstrating distinctive characteristics. In the Calathea clade, Goeppertia was segregated from Calathea based on molecular phylogenetic studies (Borchsenius et al., 2012). In the present study, Goeppertia demonstrated no anatomical differences between its two leaf edges (although a slight difference was observed between the thickness of the leaf margins in G. zingiberina). Although Calathea lutea is more closely related to Monotagma K.Schum. and Ischnosiphon Körn. (Borchsenius et al., 2012; Albuquerque et al., 2013), it was found to be similar to Goeppertia in terms of the anatomical uniformity of its left and right leaf edges and by the fact that these edges are extremely thinly tapered and filled with hypodermal cells (as seen in G. cylindrica and G. mansoi). The remaining species, I. ovatus and M. plurispicatum, show a marked anatomical difference in its edges. Based on those results, we suggest that the anatomical similarity between the left and right leaf edges may represent a relevant taxonomic feature for the genus Goeppertia, and possibly for Calathea. The Maranta clade contains the only antitropic genera in our sample (Ctenanthe and Stromanthe), plants having two types of asymmetric leaves. All of the standard and mirror leaves of Ctenanthe sp. nov., C. setosa, Stromanthe porteana and S. schottiana had thicker internal edges (i.e. the right and left edge, respectively). As such, independent of the type of leaf asymmetry, there are distinct anatomical differences between the two margins. In the Maranta clade, Ctenanthe, Stromanthe, Saranthe and Myrosma have been shown to be related by Andersson (1981, 1998), Andersson & Chase (2001) and Prince & Kress (2006a, b). All six species studied here belonging to the first three genera had fibres filling the leaf-edge region. This appears to be a taxonomic character exclusive to these genera, as identical situations were reported in Saranthe eichleri Petersen by Espírito-Santo (1998) and in Stromanthe thalia (Vell.) J.M.A.Braga by Espírito-Santo & Pugialli (1999). In this study, however, Myrosma diverges from the other taxa of that clade by not sharing that characteristic and by the fact that its leaf edges do not demonstrate significant anatomical differences. In addition, the histological organization and the contour of Myrosma cannifolia are quite similar to those of Maranta leuconeura.Albuquerque et al. (2013) reported that Myrosma differs from the other genera in the clade in the distribution and morphotype of its silica phytoliths, illustrating how dissimilar that genus is to the others. Leaves of K. orbiculata and the other Maranta spp. studied here, tend to show changes in curvature and anatomical asymmetry, but, like Myrosma, do not have fibres on their edges. Tomlinson (1961) noted that the widest (and internal) half of the leaf blade in Marantaceae demonstrates a straight edge in transverse section, whereas the narrowest (external) half tends to be curved. These differences in the two edges of the Neotropical species studied here were quite striking, but our results differ from those of Tomlinson (1961). Therefore, the internal leaf edge is generally curved, whereas almost half of the specimens analysed demonstrated a straight external edge (the customary situation for standard and mirror leaves; Tables 2 and 4). This curvature probably occurs because the internal edge is always nested inward during vernation, curving slightly as it occupies the narrow space within the still rolled up leaf. In an anatomical study of the leaves of 13 species of Bambusoideae (Poaceae), Gomes (2002) pointed out that the external edge was always straight, whereas half of the species demonstrated curved internal edges. In our study the majority of the genera of the Maranta clade showed differences in the curvatures of opposing leaf margins, but this was not observed in Goeppertia and Calathea in the Calathea clade, indicating that the presence or absence of curved margins might have diagnostic value. The Donax clade is represented in the Neotropical region only by Thalia L. The species studied here, T. geniculata, like the other species in that genus, showed large differences in the thickness of opposing leaf edges, with that anatomical asymmetry even involving differences in cell shapes in both margins of the chlorophyll parenchyma. An examination of a number of other species is still required, however, to be able to make reliable inferences about the clade. The differences between the leaf margins of Marantaceae studied here were similar to those reported by Gomes (2002) for Poaceae subfamily Bambusoideae and by Chen et al. (2007) and Yuan et al. (2015) for Oryza sativa L. Considering that most monocotyledons have convolute vernation (Skutch, 1930; Körn, 2006; Grubb & Jackson, 2007), it can be inferred that differences in thickness between the right and left edges are quite common. An anatomical study undertaken by Tenorio et al. (2014) examining the leaves of three species of Philodendron Schott (also with convolute vernation) did not, however, show any such differences. In addition, among the species studied here, anatomical asymmetry was not found in Calathea and Goeppertia. This leads to the question of whether there is actually a relationship of interference between those variables rather than one of association; i.e. does the type of vernation in fact determine the anatomical asymmetry of the edges, or is there a third determinant factor involved which leads to the concomitant appearance of these two characteristics? Finally, could there be a factor that would prevent, in some species, the manifestation of anatomical asymmetry as a consequence of the direction of the convolute vernation? Palmer (2016) pointed out that left–right asymmetries provide an attractive system for studying the role that genes, the environment and chance play in the developmental origins and subsequent evolution of novel phenotypic variation. We therefore conclude, in describing the comparative anatomy of the leaf edges of the 19 species studied here, that it was possible to determine that the anatomical asymmetry was due to the difference in the number of cell layers, the size and shape of the cells, the presence or absence of a large vascular bundle and the curvature leading to an unequal contour of the two margins. We also conclude that this asymmetry is related to the morphological asymmetry and to the direction of the convolute vernation, in both standard and mirror leaves. In addition, we may infer that, in the Maranta clade, in contrast to phylogenetic studies, Myrosma has more similarities with Maranta than with other genera. Also, in contrast to prior studies, Calathea was shown to be more similar to Goeppertia than Ischnosiphon and Monotagma in the Calathea clade, as they do not show the anatomical asymmetry so characteristic of the other genera studied here. Based on the data for this clade, it seems that leaves with anatomical asymmetry in Marantaceae necessarily follow the direction of the convolute vernation, but not all of the leaves with this type of vernation contain such asymmetry. Marantaceae have the capacity to occupy partially shaded environments and (occasionally) locations exposed to full sunlight such as forest clearings, roadsides and degraded sites (as seen with some of the species collected in the present study). As such, the constancy of the anatomical characters observed here among individuals from various populations and different habitats determines their taxonomic value. Finally, these anatomical characteristics have been shown to be conservative traits that are useful in contributing to the systematics of the group. ACKNOWLEDGEMENTS The authors declare that they have no conflict of interest. This paper is part of the PhD thesis of the first author. REFERENCES Albuquerque ESB , Braga JMA , Vieira RC . 2013 . Morphological characterisation of silica phytoliths in Neotropical Marantaceae leaves . Plant Systematics and Evolution 299 : 1659 – 1670 . Google Scholar Crossref Search ADS WorldCat Andersson L . 1981 . The Neotropical genera of Marantaceae. Circumscription and relationships . Nordic Journal of Botany 1 : 218 – 245 . Google Scholar Crossref Search ADS WorldCat Andersson L . 1998 . Marantaceae. In: Kubitzki K , ed. The families and genera of vascular plants 4 – flowering plants, monocotyledons: Alismatanae and Commelinanae (except Gramineae). Berlin : Springer-Verlag , 278 – 293 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Andersson L , Chase MW . 2001 . Phylogeny and classification of Marantaceae . Botanical Journal Linnean Society 135 : 275 – 287 . Google Scholar Crossref Search ADS WorldCat Bertrand L . 1958 . Contribution à l’étude anatomique des Marantacées africaines . Bulletin de L’Institut Centrafricaines 15–16 : 99 – 144 . OpenURL Placeholder Text WorldCat Borchsenius F , Suárez LSS , Prince LM . 2012 . Molecular phylogeny and redefined generic limits of Calathea (Marantaceae) . Systematic Botany 37 : 620 – 635 . Google Scholar Crossref Search ADS WorldCat Bukatsch F . 1972 . Bemerkungen zur Doppelfarbung Astrablau-Safranina . Mikrokosmos 61 : 225 . OpenURL Placeholder Text WorldCat Burger LM , Richter HG . 1991 . Anatomia da madeira. São Paulo : Nobel . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Chen YD , Zhang X , Zhou XQ , Chen GH . 2007 . Preliminary studies on thickness of nondestructive rice (Oryza sativa L.) leaf blade . Agricultural Sciences in China 6 : 802 – 807 . Google Scholar Crossref Search ADS WorldCat Costa FRC , Espinelli FP , Figueiredo FOG , Magnusson WE . 2011 . Guia de Zingiberales dos sítios PPBio na Amazônia ocidental brasileira. Manaus : Áttema Design Editorial . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Dahlgren RMT , Clifford HT , Yeo PF . 1985 . The families of the monocotyledons - structure, evolution, and taxonomy. Berlin : Springer-Verlag . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Delin W , Kennedy H . 2002 . Marantaceae . Flora of China 24 : 379 – 382 . OpenURL Placeholder Text WorldCat Espírito Santo A . 1998 . Anatomia foliar comparada entre indivíduos de sol e de sombra de Saranthe eichleri Petersen (Marantaceae) em dois ambientes distintos na Mata Atlântica. Unpublished Master’s Thesis, Universidade Federal do Rio de Janeiro . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Espírito Santo A , Pugialli HRL . 1999 . Estudo da plasticidade anatômica foliar de Stromanthe thalia (Vell.) J.M.A. Braga (Marantaceae) em dois ambientes de Mata Atlântica . Rodriguésia 50 : 107 – 122 . OpenURL Placeholder Text WorldCat Font Quer P . 2001 . Diccionario de botánica. 2nd edn. Barcelona : Ediciones Península . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Goldberg A . 1989 . Classification, evolution and phylogeny of the families of monocotyledons . Smithsonian Contributions of Botany 71 : 1 – 74 . OpenURL Placeholder Text WorldCat Gomes DMS . 2002 . Anatomia foliar de espécies de Merostachys Spreng. (Poaceae: Bambusoideae) no sudeste brasileiro. Unpublished D. Phil. Thesis, Universidade Federal do Rio de Janeiro . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Grubb PJ , Jackson RV . 2007 . The adaptive value of young leaves being tightly folded or rolled on monocotyledons in tropical lowland rain forest: an hypothesis in two parts . Plant Ecology 192 : 317 – 327 . Google Scholar Crossref Search ADS WorldCat Johansen DA . 1940 . Plant microtechnique. New York : McGraw-Hill Book Company Inc . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Judd WS , Campbell CS , Kellogg EA , Stevens PF, Donogue MJ . 1999 . Plant systematics – a phylogenetic approach. Sunderland : Sinauer Associates . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kennedy H , Andersson L , Hagberg M . 1988 . Marantaceae. In: Harling G, Andersson L . Flora of Ecuador 32 : 13 – 188 . OpenURL Placeholder Text WorldCat Korn RW . 2006 . Anodic asymmetry of leaves and flowers and its relationship to phyllotaxis . Annals of Botany 97 : 1011 – 1015 . Google Scholar Crossref Search ADS PubMed WorldCat Palmer AR . 2016 . What determines direction of asymmetry: genes, environment or chance? Philosophical Transactions of the Royal Society B: Biological Sciences 371 : 1 – 16 . Google Scholar Crossref Search ADS WorldCat Prince LM , Kress WJ . 2006a . Phylogenetic relationships and classification in Marantaceae: insights from plastid DNA sequence data . Taxon 55 : 281 – 296 . Google Scholar Crossref Search ADS WorldCat Prince LM , Kress WJ . 2006b . Phylogeny and biogeography of the prayer plant family: getting to the root problem in Marantaceae . Aliso 22 : 643 – 657 . Google Scholar Crossref Search ADS WorldCat Rogers GK . 1984 . The Zingiberales (Cannaceae, Marantaceae, and Zingiberaceae) in the southeastern United States . Journal of the Arnold Arboretum 65 : 5 – 55 . OpenURL Placeholder Text WorldCat Ruzin SE . 1999 . Plant microtechnique. Ames : The Iowa State College Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Rzedowski GC . 2001 . Marantaceae. In: Rzedowski J , Rzedowski GC , eds. Flora del Bajío y de Regiones Adyacentes . Pátzcuaro : Instituto de Ecología, Centro Regional del Bajío , 1 – 7 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Schumann KM . 1902 . Marantaceae . In: Engler HGA , ed. Das Pflanzenreich: Regni Vegetabilis Conspectus. Berlin : Verlag von Engelmann , 1 – 176 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Skutch AF . 1930 . Unrolling of leaves of Musa sapientum and some related plants and their relation to environmental acidity . Botanical Gazette 90 : 337 – 365 . Google Scholar Crossref Search ADS WorldCat Suárez S , Galeano G . 1996 . Las marantáceas de la region de Araracuara. Estudios en la Amazonia colombiana. Bogotá : Fundación Tropenbos-Colombia . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Suemoto H . 1953 . Studies on the right- and left- handedness of spikelets in einkorn wheats. II. Developmental process in the ear . Proceedings of the Japan Academy 29 : 362 – 367 . Google Scholar Crossref Search ADS WorldCat Tenorio V , Sakuragui CM , Ferreira RO , Albuquerque ESB , Vieira RC . 2014 . Morfoanatomia e histoquímica foliar de espécies do gênero Philodendron Schott (Araceae) . Revista de Ciência & Tecnologia (UNIG) 14 : 13 – 32 . OpenURL Placeholder Text WorldCat Tomlinson PB . 1961 . Morphological and anatomical characteristics of the Marantaceae . Botanical Journal of the Linnean Society 58 : 55 – 78 . Google Scholar Crossref Search ADS WorldCat Tomlinson PB . 1962 . Phylogeny of the Scitamineae - morphological and anatomical considerations . Evolution 16 : 192 – 213 . Google Scholar Crossref Search ADS WorldCat Tomlinson PB . 1969 . Commelinales - Zingiberales . In: Metcalfe CR , ed. Anatomy of the monocotyledons. Oxford : Clarendon Press , 341 – 359 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Yuan S , LI Y , Peng S . 2015 . Leaf lateral asymmetry in morphological and physiological traits of rice plant . PLoS One 10 : 1 – 15 . OpenURL Placeholder Text WorldCat © 2019 The Linnean Society of London, Botanical Journal of the Linnean Society This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)