Abstract Background and Aims Dioecious plants are of particular concern in view of global environmental changes because reproductive females are more sensitive to abiotic stresses, thus compromising population viability. Positive interactions with other plants may counteract the direct effects of any abiotic environmental stress, allowing them to thrive and maintain a viable population in suboptimal habitats, although this process has not been tested for dioecious species. Furthermore, almost no data are available on the outcome of such species interactions and their link with local spatial patterns and sex ratios. Methods We set up a field experiment with Poa ligularis, a dioecious native grass from the arid grasslands of South America. We studied the interaction of male and female plants with cushion shrubs of contrasting ecological strategies. We experimentally limited direct shrub–grass competition for soil moisture and transplanted plants to evaluate the amelioration of abiotic stress by shrub canopies (i.e. sun and wind) on grasses. We also studied the distribution of naturally established female and male plants to infer process–pattern relationships. Key Results Positive canopy effects as well as negative below-ground effects were more intense for females than for males. Deep-rooted Mulinum spinosum shrubs strongly facilitated survival, growth and reproduction of P. ligularis females. Naturally established female plants tended to distribute more closely to Mulinum than co-occurring males. Female growth suffered intensive negative root competition from the shallow-rooted Senecio filaginoides shrub. Conclusions Interactions with other plants may reduce or enhance the effect of abiotic stresses on the seemingly maladapted sex to arid environments. We found that these biased interactions are evident in the current organization of sexes in the field, confirming our experimental findings. Therefore, indirect effects of climate change on population sex ratios may be expected if benefactor species abundances are differentially affected. Arid ecosystems, competition, dioecy, facilitation, female plants, male plants, Mulinum spinosum, Patagonian steppe, plant–plant interactions, Poa ligularis, positive interactions, Senecio filaginoides INTRODUCTION It has been proposed that dioecious plant populations in harsh environments should be male-biased due to the higher susceptibility of female plants to abiotic stress (Dawson and Ehleringer, 1993; Delph, 1999; Obeso, 2002). However, the expected pattern of male preponderance in stressful habitats is not always found (Verdú and García-Fayos, 1998; Graff et al., 2013; Juvany and Munné-Bosch, 2015; Moreno and Bertiller, 2016). While there are numerous physiological and genetic explanations for female-biased sex ratios in some populations (Field et al., 2013), little attention has been given to the effect of biotic interactions on the performance of dioecious species (but see Verdu et al., 2004; Montesinos et al., 2007; Lortie and Reid, 2012; Graff et al., 2013). Plant–plant interactions (facilitation and exploitation competition) may buffer or increase the stress experienced within a site and largely drive population vital rates (Goldberg, 1990; Liancourt et al., 2013; López et al., 2016), and thus sex ratios (Graff et al., 2013). In harsh environments, microhabitat differences may allow dioecious species to maximize fitness by restricting more stress-sensitive females to low-stress microsites and allowing males to stock resource-poorer microsites. Shrubs are frequently responsible for the microhabitat heterogeneity in sparsely vegetated arid lands (Noy-Meir, 1973; Aguiar and Sala, 1999; Yang et al., 2015). Some studies have documented the spatial segregation of sexes (Bierzychudek and Eckhart, 1988) associated with the microhabitat heterogeneity generated by shrub patches. Females were found to be distributed close to shrubby patches, whereas males were more common in open areas (Delph, 1999; Bertiller et al., 2000, 2002). Although not emphasized in the literature, this pattern of spatial segregation could be interpreted as a sex-biased facilitation of shrubs for females via habitat amelioration (but see Moreno and Bertiller, 2016). By providing shelter from wind and from extreme temperatures and by increasing soil fertility, facilitation mediated by shrubs could allow stress-sensitive females to grow and persist in environments where they would otherwise be limited (Palmer and Dixon, 1990; Bruno et al., 2003). In this situation, the population sex ratio expected based on aridity could be different from that expected based on shrub cover (Moreno and Bertiller, 2016). If species interactions are key drivers of the spatial pattern, the net balance should influence vital rates and the strength of the spatial segregation of sexes (Graff et al., 2013). Morphological variation in the canopy and root architecture among dominant overstorey plants can determine understorey microenvironment as well as the magnitude of the interactive effects of canopy facilitation and below-ground interference (Callaway et al., 1991; Michalet et al., 2011; Schöb et al., 2013; Aschehoug and Callaway, 2014; Bonanomi et al., 2016). In systems where wind-driven desiccation is an important environmental stress factor for plant growth (Lambers et al., 2008), compact and tall shrubs provided a better wind shield than smaller shrubs with loose canopies, thereby reducing evaporation (Schöb et al., 2013). However, plants growing under dense canopies could also have the cost of reduced light (Holmgren et al., 1997). Therefore, the level of shading created by nurse plants could determine a lack of a positive influence of their canopy, despite the improvements in water conditions (Reisman-Berman, 2007; Michalet et al., 2011; Holmgren et al., 2012). By contrast, below-ground competition is also expected to be intense in water limited systems, increasing the stress experienced by neighbouring species (Valladares and Pearcy, 2002; Michalet, 2007; Maestre et al., 2009; Michalet et al., 2016), and despite a canopy effect, root architecture could determine the intensity of the competition between woody and herbaceous growth forms (Callaway et al., 1991; Bucci et al., 2009; Aschehoug and Callaway, 2014; Semchenko et al., 2017). Because of sex -differences in growth, survival, reproductive output, resource allocation and physiology (Dawson and Bliss, 1989), it might be expected that the genders will perform differently in facilitative/competitive interactions with other species (Verdu et al., 2004; Montesinos et al., 2007; Graff et al., 2013). Therefore, the performance of dioecious species could increase or decrease depending on the interacting overstorey shrub species, and eventually this differential performance could result in a local spatial segregation of the sexes. We studied the interaction of adult male and female plants of a dioecious grass species with neighbouring shrub species having morphological variation in root architecture and contrasting ecological strategies (Graff and Aguiar, 2017). The shrubs selected were hemispherical, with closed and tall canopies, but one species was deep-rooted and the other was shallow-rooted. Therefore, we expected that both shrubs would be canopy facilitators, but we also expected that soil effects of shallow-rooted species would impact more negatively on the performance of grasses than those of deep-rooted species because woody and herbaceous growth forms frequently compete for resources in the upper soil layers (Davis et al., 1998; Bucci et al., 2009). Through transplanting male and female plants at different distances of shrub species and experimentally reducing root competition, we tested the potential effects of plant–plant interactions on vital rates of males and females. We studied the distribution of female and male naturally established plants at this ecological scale to confirm the importance of the interaction balance (Graff and Aguiar, 2017). MATERIALS AND METHODS Study site Field studies were conducted in the Patagonian steppe in south-western Chubut, Argentina (45°41′S, 70°16′W, 500 m a.s.l.). The study site was located within a 150-km2 area at the Río Mayo Experimental Field Station (Instituto Nacional de Tecnología Agropecuaria). This community is co-dominated by few grasses and shrubs (only three species each), which contribute more than 96 % of the total plant cover and plant biomass (Fernandez et al., 1991). Forbs contribute <1 % to plant cover. The vegetation is distributed in a two-phase mosaic, shrubs and grasses forming high cover patches within a matrix of sparse tussock grasses. The climate is arid, with an intense summer drought (Paruelo et al., 1988). Mean annual precipitation in the last 30 years (1984–2013) was 134 mm with a winter and early spring (May to September) rainy season. Average monthly temperatures range from 3 °C in winter to 16 °C in summer (1984–2013) (field meteorological station data http://anterior.inta.gov.ar/region/pas/sipas2/cmp/agromet/index.html# revised August, 2017). Strong, dry winds blow predominantly from west to east with high intensities throughout the year, averaging more than 20 km h−1 during the growing season. The scarce precipitation and its distribution during winter lead to a high summer water deficit. Therefore, stress during the growing season is mainly induced by the lack of below-ground resources (water and nutrients) and by above-ground non-resource factors (wind, temperature, etc.) (Beltrán, 1997). As in other arid ecosystems, stress factors also vary at the patch scale (Soriano and Sala, 1986; Aguiar and Sala, 1994). Wind speed is five times lower and evaporative demand two times lower near shrub patches than in bare soil areas (Soriano and Sala, 1986; Graff and Aguiar, 2017,in ecy1703-sup-0001-AppendixS1.docx), while at the same time, soil water content is significantly lower than in open spaces (Graff and Aguiar, 2017,in ecy1703-sup-0001-AppendixS1.docx). Therefore, amelioration of non-resource stress factors by shrub canopies is coupled with an increase in resource stress factors driven by the water consumption of the potential benefactor (Sala et al., 1989; Valladares and Pearcy, 2002; Maestre and Cortina, 2004). Species description Shrub species. We selected two out of the three dominant species as potential benefactor species. Shrubs were Mulinum spinosum [=Azorella prolifera (Cav.) G.M. Plunkett & A.N. Nicolas] and Senecio filaginoides, De Candolle (hereafter Mulinum and Senecio, respectively). Together the two species account for 76 % of the shrub cover in the Occidental district in the Patagonian steppe (Fernandez et al., 1991). Mulinum is a deciduous, hemispherical-cushion-form shrub, of 70 cm height. Senecio is an evergreen species small and globose shrub of 60 cm height (Fernandez et al., 1991). Both shrub species have a tap root with long lateral roots growing outward, but in Senecio, the lateral roots grow close to the soil surface and previous physiological and water use studies on adult plants indicate that its activity overlaps substantially with grasses (Golluscio and Oesterheld, 2007; Bucci et al., 2009). In Mulinum the lateral roots start to grow horizontally 15 cm below the soil surface (Fernández and Paruelo, 1988; Sala et al., 1989; Bucci et al., 2009) and we assume that this root architecture has a lower competition effect on herbaceous species. Sexual morphs of the grass species. We studied Poa ligularis Nees. ap. Steudel (hereafter Poa), a dominant species in non-degraded Patagonian shrub–grass steppes. Poa is a wind-pollinated and strictly dioecious bunchgrass (section Dioicopoa). The species is endemic to the Patagonian steppe and Pampa region in Argentina (www.floraargentina.edu.ar). Male (staminate) and female (pistillate) plants are dimorphic in their reproductive characters and the hairiness of the floret (www.floraargentina.edu.ar). They are also dimorphic in morpho-chemical traits related to herbivory and to the environmental variation induced by surrounding vegetation. Female morphs produce more phenolic compounds and are also less palatable to herbivores than males (see table 1 in Graff et al, 2013). Females associated with shrub patches are taller and have increased blade length and area than those in inter-patch areas (Moreno & Bertiller, 2016). On the other hand, females allocate less to the root system (i.e. less root biomass, new roots and root/shoot ratio) than males (see table 1 in Graff et al, 2013). A high investment in root biomass is necessary for drought resistance (Lambers et al., 2008) and could help also to minimize the negative impact of sharing below-ground resources with neighbours (Cahill and Casper, 2000; Wang et al., 2010; Semchenko et al., 2013). Accordingly, below-ground competitive effects of other grass neighbours have been reported to be more intense on the biomass of female adult plants than on males (Graff et al, 2013). Previous studies with Poa indicated that sex expression is constant (Bertiller et al., 2000, 2002; Graff et al., 2013), and females distribute more closely to large shrub patches in the Patagonian Monte than males (Bertiller et al., 2000, 2002). Shrub–sexual morph interactions: field experiment The effects of each shrub species on male and female Poa plants were addressed with a manipulative experiment with sex (female or male), shrub canopy effect (beneath shrub canopy or in open areas) and root effect (full or reduced root competition) arranged on a three-way factorial randomized complete block design. We selected six paddocks where sheep grazing was excluded during the experiment. Within each of these six grazing exclosures (blocks n = 6), we selected 16 Mulinum and 16 Senecio naturally established shrubs of similar height, basal diameter and far from other shrubs (>3 m), to apply the eight treatments with two subreplicates per block of each treatment combination. The eight treatments were the combination canopy × root effects for each sex. In the previous spring–summer (November–January), Poa plants growing within long-term exclosures were sexed and marked in the field to enable collection in May 2010 (when the experiment began). The 96 male and 96 female transplants comprising the entire experiment were obtained by dividing Poa plants into five-tiller individual plants. Because Poa plants are perennial, we selected stem bases with green tillers coming from the tussock periphery to form the five-tiller individuals. To isolate the shrub canopy effect (facilitation or not), transplants were placed at the edge of the shrub canopy (beneath shrub canopy), and 1 m away from the shrub edge (open areas), within the bare soil matrix and far from the influence of the shrub canopy. Transplants placed beneath shrub canopies were located east of shrubs, where wind speed is five times lower and evaporative demand is two times lower than those of exposed windward areas (Soriano and Sala, 1986; Aguiar and Sala, 1994). Therefore, facilitation via habitat amelioration is more likely to occur in eastern edges (Graff & Aguiar, 2017). The root effect (full or reduced competition) was achieved by isolating (or not) transplant roots from the influence of shrubs. Half of the transplants had a reduced root effect (or reduced competition) treatment that was achieved by digging a hole of 0.20 m in diameter by 0.30 m deep, and fitting a fine mesh cloth in the pits, which functioned as a barrier to growth of roots of neighbouring plants under beneficiary target plants (Aguiar et al., 1992; Graff et al., 2007, 2013). The holes were filled with the extracted soil after removing roots and stones by sieving. Transplants with no mesh around the sieved soil core were those that did not receive a reduced root effect treatment. Large shrub patches are frequently encircled by a dense ring of grasses that could obscure shrub effects on target species (Aguiar and Sala, 1994), so before applying the treatments, we removed the surrounding grasses to isolate the effect of the shrub on Poa plants. We monitored transplant survival during two growing seasons. Alive transplants were harvested in January 2012 by digging pits 0.20 m in diameter by 0.3 m deep where the transplants were located. Vegetative (shoot and root) and reproductive (panicles) material was separated, oven dried at 70 °C for 48 h and weighed. Roots were separated from sandy soils by flotation and sieving (0.2-mm mesh sieve). Shrub effects on transplant survival were analysed for each shrub species separately. We used the GENMOD procedure in SAS (SAS Institute, 1996). The model assumed a binomial distribution and used a logit link function because the response variable was binary (lived or not). We analysed the effect of each shrub species on Poa plants through an RCB three-way factorial ANOVA in SAS (SAS Institute, 1996). The response variable was total vegetative biomass (roots and shoots) and panicle number. Panicle biomass was not included because some seeds from female plants were dispersed before we harvested the plants. We had three crossed factors: sex (S) (males and females), distance to neighbour (D) (close and far) and barrier to root competition (B) (with or without). Transformed data of total vegetative biomass (ln) and panicle number (sqrt) were used to meet the assumption of homoscedasticity in ANOVA. We averaged the two subreplicates per block for statistical analyses. This resulted in six true replicates (n = 6). We also evaluated the predicted positive, negative and net effects of shrubs on sexes by calculating the Relative Interaction Index (RII;Armas et al., 2004) per block. Response variables were survival, total vegetative biomass at harvest and panicle number at harvest. RIInet effects: (CB−− FB+)/(CB−+ FB+), where C is close to shrub canopy, F is far from shrubs, and B is the barrier to root competition treatment (+ and − superscripts indicate with or without the mesh cloth, respectively). RII is symmetrical around 0 and ranges from +1 to − 1. We constructed the 95 % confidence intervals from the RII values obtained per block. Positive values indicate facilitation, negative values indicate competition and values not significantly different from zero indicate neutral/no effects (Armas et al., 2004). We tested whether RII values differ from zero using a t-test for single means. Shrub–sexual morph distribution: descriptive study To complement the field experiment, we included an assessment of the cumulative frequency distribution of males and females in relation to the distance from shrub canopies measured during the flowering season. We used four long-term exclosures to sheep grazing and other large herbivores (settled in 1956, 1972, 1983 and 1994, respectively) to minimize the influence of large herbivores on the spatial distribution of plants. In each of the four exclosures, we selected shrubs of modal size along an imaginary line forming ten pairs of single plants of Mulinum–Mulinum and ten Senecio–Senecio shrubs. Selections were based on the criteria that the distance between the edge of each shrub was 2 m (which is approximately the maximum distance between shrub pairs found in exclosures; Cipriotti and Aguiar, 2010). Transects were placed from the edge of one shrub to the half distance of the other shrub edge to fully characterize the zone of influence of only one shrub on the distribution of Poa plants. At each transect, male and female distance to a focal shrub was recorded. We merged the data points from the ten transects per shrub per exclosure and then determined cumulative frequency distributions of the number of Poa plants at distances from shrubs. The distributions of cumulative frequencies of sexual morphs occurrence at varying distances from Mulinum or Senecio edges were compared using Kolmogorov–Smirnov tests. We also calculated the sex-ratio in the long-term exclosures. In each exclosure we established 50 × 1-m linear transect and we counted the total number of female and male plants. RESULTS Shrub–sexual morph interactions A sex-biased survival of Poa plants was only detected in the interaction with Mulinum shrubs (Fig. 1). Females were more favoured by Mulinum canopies than males (Sex × Distance P = 0.0076). Almost all female plants survived when grown beneath Mulinum canopies, whereas survival of females was 37 % lower in open areas. Male survival was not affected by the distance to Mulinum canopies (Supplementary Data Table S1). Fig. 1. View largeDownload slide Sex-biased survival response of sexual morphs of Poa ligularis to Mulinum spinosum shrubs. Proportion of female (white bars) and male (grey bars) plants that survived growing beneath Mulinum canopies (Close) and in open areas (Far) (Sex × Distance, P = 0.0076). Fig. 1. View largeDownload slide Sex-biased survival response of sexual morphs of Poa ligularis to Mulinum spinosum shrubs. Proportion of female (white bars) and male (grey bars) plants that survived growing beneath Mulinum canopies (Close) and in open areas (Far) (Sex × Distance, P = 0.0076). We found strong sex-related differences in total biomass at harvest in response to shrub species interactions (Fig. 2). While Mulinum canopies had positive effects on female total biomass, they had negative effects on male biomass (Sex × Distance, P<0.0001, Fig. 2A). Although plants tended to accumulate more biomass when shrub roots were excluded (Barrier, P = 0.08), this was a weak effect and no interaction with sexual morphs was detected (Sex × Barrier, P = 0.64) (Table S2). Fig. 2. View largeDownload slide Sex-biased growth response of sexual morphs of Poa ligularis to the interaction with shrubs. Total biomass at harvest of female (white bars) and male (grey bars) plants interacting with (A) Mulinum spinosum and (B) Senecio filaginoides shrubs. Plants grew in the close periphery to or far from potential nurses and with (B+) or without (B-) the barrier that reduces root competition on them. S, sex; D, distance from shrub; B, barrier factors. Fig. 2. View largeDownload slide Sex-biased growth response of sexual morphs of Poa ligularis to the interaction with shrubs. Total biomass at harvest of female (white bars) and male (grey bars) plants interacting with (A) Mulinum spinosum and (B) Senecio filaginoides shrubs. Plants grew in the close periphery to or far from potential nurses and with (B+) or without (B-) the barrier that reduces root competition on them. S, sex; D, distance from shrub; B, barrier factors. Senecio shrubs did not affect the survival of female and male plants, but interacted with sexual morphs in an opposite way to that found for Mulinum (Fig. 2B). On average, females accumulated less biomass growing close to than far from Senecio, whereas males did not show differences in their biomass (Sex × Distance, P = 0.0149). Females and males were also affected differently by Senecio root competition (Fig. 2A) growing twice as large with the barrier (B+) than without it (B-) (Sex × Barrier, P = 0.0302). By contrast, males did not show differences in total biomass between the barrier treatments (Table S2). A sex-biased response in panicle production was also detected in the interaction with shrubs (Fig. 3). In general, the patterns of response for reproduction were similar to those for plant biomass, but with different intensity depending on the shrub species. Females produced three times more reproductive structures close to Mulinum than far from them. By contrast, male panicle number was 40 % lower close to than far from the shrub (Sex × Distance, P = 0.0002), independently of the competition treatment. Similar but less significant results were found with Senecio (Sex × Distance, P = 0.04) (Table S3). Fig. 3. View largeDownload slide Sex-biased panicle production of sexual morphs in response of Poa ligularis to the interaction with shrubs. Panicle number produced at the end of the second growing season of female (white bars) and male (grey bars) plants interacting with (A) Mulinum spinosum and (B) Senecio filaginoides shrubs. Plants grew in the close periphery to or far from potential nurses and with (B+) or without (B-) the barrier that reduces root competition on them. S, sex; D, distance from shrub; B, barrier factors. Fig. 3. View largeDownload slide Sex-biased panicle production of sexual morphs in response of Poa ligularis to the interaction with shrubs. Panicle number produced at the end of the second growing season of female (white bars) and male (grey bars) plants interacting with (A) Mulinum spinosum and (B) Senecio filaginoides shrubs. Plants grew in the close periphery to or far from potential nurses and with (B+) or without (B-) the barrier that reduces root competition on them. S, sex; D, distance from shrub; B, barrier factors. Net effects of shrubs were significant for female performance but not for male performance (P < 0.05). Whereas in the interaction with Mulinum the net effect on females was positive for survival, growth and panicle production, in the interaction with Senecio the balance was negative for growth (Fig. 4; Table S4). Fig. 4. View largeDownload slide Net interaction effects of shrubs on Poa ligularis females and males. Bars represent the Relative Interaction Index (RII; Armas et al., 2004) of Mulinum and Senecio shrubs on each sexual morph. White bars for females and grey bars for males. Values are RII means and 95 % confidence intervals. Positive values indicate facilitation, negative values indicate competition and values not significantly different from zero indicate neutral/no effects (Cristina Armas et al., 2004). F, females; M, males. Asterisks indicate significant deviation from zero value (one-sample t-tests): *P < 0.05, **P < 0.01. Calculations of the effects are explained in detail in the Methods. Fig. 4. View largeDownload slide Net interaction effects of shrubs on Poa ligularis females and males. Bars represent the Relative Interaction Index (RII; Armas et al., 2004) of Mulinum and Senecio shrubs on each sexual morph. White bars for females and grey bars for males. Values are RII means and 95 % confidence intervals. Positive values indicate facilitation, negative values indicate competition and values not significantly different from zero indicate neutral/no effects (Cristina Armas et al., 2004). F, females; M, males. Asterisks indicate significant deviation from zero value (one-sample t-tests): *P < 0.05, **P < 0.01. Calculations of the effects are explained in detail in the Methods. Shrub–sexual morph distribution A total of 217 Poa plants were recorded in the 80 transects placed from the edge of one shrub to the half distance of the other shrub edge, 111 in the Mulinum–Mulinum transects and 106 in the Senecio–Senecio transects. Females were distributed more closely to Mulinum shrub patches than males (Distance, 0.4122, P = 0.002), with a median of 40 and 65 cm, respectively (Fig. 5A). Females and males were distributed similarly from Senecio shrubs (P = 0.677), but no female individuals were found in close proximity of Senecio (up to 14 cm). The 5–10 % percentiles of female numbers were at 18.2–21 cm, whereas those for males were at 0–2.4 cm (Fig. 5B). Fig. 5. View largeDownload slide Distribution of shrub–sexual morphs. Cumulative frequency of female (dotted curve) and male (solid line) plants at increasing distances from Mulinum spinosum (deciduous, deep-rooted) and Senecio filaginoides (evergreen and shallow-rooted) shrubs. Inset figures are box-and-whisker plots representing the 5–95 % percentile distribution and median values. Point outside whiskers are outliers. Fig. 5. View largeDownload slide Distribution of shrub–sexual morphs. Cumulative frequency of female (dotted curve) and male (solid line) plants at increasing distances from Mulinum spinosum (deciduous, deep-rooted) and Senecio filaginoides (evergreen and shallow-rooted) shrubs. Inset figures are box-and-whisker plots representing the 5–95 % percentile distribution and median values. Point outside whiskers are outliers. Sex ratio was female-biased and increased with exclosure age. Sex ratio was 2.2: 1, 1.7: 1, 1.3: 1 and 1.1: 1 in the exclosures settled in 1956, 1972, 1983 and 1994, respectively. DISCUSSION In dioecious plant species, population performance and viability can be projected in terms of reproductive females. Male-biased sex ratios then serve to lower the effective population size (Ne), which can lead to extinction (Hultine et al., 2008, 2016). It has been proposed that positive interactions with other plant species may counteract the direct effects of the abiotic environment, allowing plants to thrive and maintain a viable population in suboptimal habitats (Hacker and Gaines, 1997; Bruno et al., 2003), but this concept has not been explored for dioecious species. Our results confirmed that facilitation, mediated by shrub cover, could help females of dioecious plants to cope with environmental severity, but also that water depletion by neighbours may prevent females from benefitting from seemingly favourable conditions above ground. The current organization of sexes in the steppe environment confirmed that our experimental findings are important (sensuTirado et al., 2015) in the context of spatial and temporal variability controlling vital rates of plants in arid ecosystems. Poa ligularis females interact with shrubs more strongly than males. Positive canopy effects as well as negative below-ground effects were more intense for female vital rates than for male vital rates, and were also dependent on the shrub species. Canopies of Mulinum shrubs facilitated the survival, growth and reproduction of females, whereas there was a lack of a positive influence of Senecio canopies. The variation in the level of shade below both shrubs could explain the difference in their effect. Both shrubs have been described as hemispherical cushion-like shrubs, but with contrasting phenologies (Golluscio et al., 2005). Mulinum is a deciduous species and Senecio is an evergreen species that intercepts more radiation than Mulinum in summer, when dry conditions are more marked (Graff and Aguiar, 2017). Previous studies have demonstrated that Poa ligularis increases chlorophyll content and photoprotective carotenoid concentration under drought (Cenzano et al., 2013). Protection against high irradiance could be a cost in the shade rather than a benefit (Valladares, 2004; Valladares et al., 2008). Therefore, the greater shade below Senecio canopies could partially explain the lack of a positive influence on Poa plants. Additionally, Senecio’s root competition was also strong on female growth. Negative root effects were more intense in the interaction with the shrub with the shallowest rooting depth. Overall net positive effects of Mulinum, the deep-rooted shrub, on females were also supported by the spatial organization in the field: naturally established female Poa plants tended to be distributed more closely to shrub edges than males. Therefore, based on the probably higher costs of reproduction (Graff et al, 2013), females of a dioecious grass in this arid environment were more strongly favoured by shrubs than males, but only when positive effects through habitat amelioration outweighed the negative effects of resource competition. In anemophilous dioecious plants, it is frequent to find that sexes segregate by habitat, with the females being located in habitats of high quality and in the shelter of vegetation patches. By contrast, the males frequently grow in the more marginal zones or in more exposed sites with faster air currents (i.e. distant from a vegetation patch). The latter would enhance male function, by distributing their pollen and increasing opportunities for mating (Bierzychudek and Eckhart, 1988; Bertiller et al., 2000). For Poa ligularis, pollen production and dispersal occur in spring, after a humid winter season, and when soil water is not a constraint (Bertiller et al., 2000). However, females complete fruiting at the beginning of the dry season (late spring to early summer). Thus, the reproductive effort of males can be maintained even in relatively more exposed sites (i.e. distant from a patch) than females. Confirming these ideas, we have found a consistent gender differential response on panicle production related to shrub distance. Females produce more panicles when close to rather than far from shrubs, whereas males showed the opposite pattern, suggesting a reproductive disadvantage in sheltered microsites for males (Dawson and Bliss, 1989; Delph, 1999). Despite the prediction of a male preponderance in arid environments, previous studies in harsh environments did not confirm this pattern because of unexpected female-biased sex ratios (Graff et al., 2013; Moreno and Bertiller, 2016). Furthermore, female numbers sometimes increase along regional environmental stress gradients (Gehring and Linhart, 1993; Moreno and Bertiller, 2016). The mechanisms responsible for the female-biased sex ratios observed in mature plant populations are not well understood and could be caused by both genetic and ecological factors (Field et al., 2013). Our study sheds light on an unexplored factor that could drive the sex ratio of dioecious species in arid environments: facilitation. In general terms, facilitation enhances plant growth, reproduction and survival of understorey species, eventually expanding their realized niche (Bruno et al., 2003). Facilitation can also explain some of the patterns in plant species interactions occurring in stressed ecosystems (Michalet, 2006). However, certain species are more likely to act as facilitators than others, depending on the ways the species influence their environment and resources (Callaway, 2007; Macek et al., 2016). In our current experiment, performed in grazing exclosures, we found that only Mulinum shrubs exerted positive net effects on females. Furthermore, females generally grew closely associated with this species in the field. Mulinum is a palatable species and increases in numbers in exclosures to sheep grazing (Fernández et al., 1992; Cipriotti and Aguiar, 2012). Current density is 0.31, 0.24 and 0.14 individuals m–2 in the exclosures settled in 1972, 1983 and 1994, respectively (M. R. Aguiar et al., unpubl. data). We propose that this indirect effect of grazers on benefactor density may explain the female-biased pattern previously found in exclosures (Graff et al., 2013). Moreover, it could help to understand why female ratio increases with exclosure age in this study. Our study and findings may have significance in predicting the potential response of dioecious species to global environmental changes. It has been proposed that populations of dioecious species will be directly threatened by an increase in aridity due to the higher susceptibility of female plants to abiotic stress (Hultine et al., 2008, 2016). Our study with two different shrub species supports the idea that the indirect effects of climate change may be as important as the direct effects. In general, climate change has been projected to increase aridity in temperate drylands worldwide during this century (Feng and Fu, 2013), but it will be coupled with a substantial increase in deep soil drought conditions during the growing season (Schlaepfer et al., 2017), which may differentially affect woody plants according to their root architecture. In particular, stress-tolerant dominant species with shallow roots (Schlaepfer et al., 2017), and species with thick evergreen leaves, low specific leaf area and high leaf dry matter content will be favoured (Gross et al., 2013). This specialization of the species pool, expected under extreme conditions, will probably change the facilitation/competition balance towards an increase in the importance of competition as an assembly driver (McCluney et al., 2012; Liancourt et al., 2013; Siefert et al., 2015). Climate change will thus decrease the reproductive success of many dioecious species by directly altering their individual fitness (Hultine et al., 2008, 2016), but also indirectly, by limiting the role of shrubs as benefactors of females, amplifying dioecious species loss. CONCLUSIONS Interactions with other plants may reduce or enhance the effect of abiotic stress on the seemingly maladapted sex to arid environments. We found that these biased interactions are evident in the current organization of sexes in the field, confirming our experimental findings. Therefore, indirect effects of climate change on population sex ratios may be expected if benefactor species abundances are differentially affected. SUPPLEMENTARY DATA Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Table S1: Analysis of the survival response of Poa ligularis. Table S2: Analysis of the growth response of Poa ligularis. Table S3: Analysis of the panicle production of Poa ligularis. Table S4: Analysis of the net interaction outcomes among shrubs and Poa ligularis sexual morphs. ACKNOWLEDGEMENTS We thank Ignacio P. Rodriguez and Gastón Oñatibia for assistance in the field. This work was supported by [ANPCyT- PICT 00462] and [UBA - G0032] to M.R.A. P.G. was sponsored by a post-doctoral fellowship from CONICET. M.R.A. and P.G. are members of CONICET. Instituto Nacional de Tecnología Agropecuaria (INTA) granted access to the facilities and experimental field. LITERATURE CITED Aguiar MR, Sala OE. 1994. Competition, facilitation, seed distribution and the origin of patches in a Patagonian steppe. Oikos 70: 26– 34. Google Scholar CrossRef Search ADS Aguiar MR, Sala OE. 1999. Patch structure, dynamics and implications for the functioning of arid ecosystems. Trends in Ecology and Evolution 14: 273– 277. Google Scholar CrossRef Search ADS PubMed Aguiar MR, Soriano A, Sala OE. 1992. Competition and facilitation in the recruitment of grass seedlings in Patagonia. Functional Ecology 6: 66– 70. Google Scholar CrossRef Search ADS Armas C, Ordiales R, Pugnaire FI. 2004. Measuring plant interactions: a new comparative index. Ecology 85: 2682– 2686. Google Scholar CrossRef Search ADS Aschehoug ET, Callaway RM. 2014. Morphological variability in tree root architecture indirectly affects coexistence among competitors in the understory. Ecology 95: 1731– 1736. Google Scholar CrossRef Search ADS PubMed Beltrán A. 1997. Caracterización microclimatica del distrito occidental de la estepa Patagónica . Magister thesis, University of Buenos Aires. Bertiller MB, Ares JO, Graff P, Baldi R. 2000. Sex-related spatial patterns of Poa ligularis in relation to shrub patch occurrence in northern Patagonia. Journal of Vegetation Science 11: 9– 14. Google Scholar CrossRef Search ADS Bertiller MB, Sain CL, Bisigato AJ, et al. 2002. Spatial sex segregation in the dioecious grass Poa ligularis in northern Patagonia: the role of environmental patchiness. Biodiversity and Conservation 11. Bierzychudek P, Eckhart V. 1988. Spatial segregation of the sexes of dioecious plants. American Naturalist 132: 34– 43. Google Scholar CrossRef Search ADS Bonanomi G, Stinca A, Chirico GB, et al. 2016. Cushion plant morphology controls biogenic capability and facilitation effects of Silene acaulis along an elevation gradient. Functional Ecology 30: 1216– 1226. Google Scholar CrossRef Search ADS Bruno JF, Stachowicz JJ, Bertness MD. 2003. Inclusion of facilitation into ecological theory. Trends in Ecology and Evolution 18: 119– 125. Google Scholar CrossRef Search ADS Bucci SJ, Scholz FG, Goldstein G, et al. 2009. Soil water availability and rooting depth as determinants of hydraulic architecture of Patagonian woody species. Oecologia 160: 631– 641. Google Scholar CrossRef Search ADS PubMed Cahill JF, Casper BB. 2000. Investigating the relationship between neighbor root biomass and belowground competition: field evidence for symmetric competition belowground. Oikos 90: 311– 320. Google Scholar CrossRef Search ADS Callaway RM. 2007. Positive Interactions and Interdependence in Plant Communities . Berlin: Springer Science & Business Media. Callaway RM, Nadkarni NM, Mahall BE. 1991. Facilitation and interference of Quercus douglasii on understory productivity in central California. Ecology 72: 1484– 1499. Google Scholar CrossRef Search ADS Cenzano AM, Varela MC, Bertiller MB, Luna MV. 2013. Effect of drought on morphological and functional traits of Poa ligularis and Pappostipa speciosa, native perennial grasses with wide distribution in Patagonian rangelands, Argentina. Australian Journal of Botany 61: 383–393. Cipriotti PA, Aguiar MR. 2010. Resource partitioning and interactions enable coexistence in a grass–shrub steppe. Journal of Arid Environments 74: 1111– 1120. Google Scholar CrossRef Search ADS Cipriotti PA, Aguiar MR. 2012. Direct and indirect effects of grazing constrain shrub encroachment in semi-arid Patagonian steppes. Applied Vegetation Science 15: 35– 47. Google Scholar CrossRef Search ADS Davis MA, Wrage KJ, Reich PB. 1998. Competition between tree seedlings and herbaceous vegetation: support for a theory of resource supply and demand. Journal of Ecology 86: 652– 661. Google Scholar CrossRef Search ADS Dawson TE, Bliss LC. 1989. Patterns of water use and the tissue water relations in the dioecious shrub, Salix arctica: the physiological basis for habitat partitioning between the sexes. Oecologia 79: 332– 343. Google Scholar CrossRef Search ADS PubMed Dawson TE, Ehleringer JR. 1993. Gender-specific physiology, carbon isotope discrimination, and habitat distribution in boxelder, Acer negundo. Ecology 74: 798– 815. Delph LF. 1999. Sexual dimorphism in life history. In: MA Geber, TE Dawson LD, eds. Gender and Sexual Dimorphism in Flowering Plants . Berlin: Springer-Verlag, 149– 173. Google Scholar CrossRef Search ADS Feng S, Fu Q. 2013. Expansion of global drylands under a warming climate. Atmospheric Chemistry and Physics 13: 10081–10094. Fernández RJ, Paruelo JM. 1988. Root systems of two Patagonian shrubs: a quantitative description using a geometrical method. Journal of Range Management 41: 220– 223. Google Scholar CrossRef Search ADS Fernandez RJ, Sala OE, Golluscio RA. 1991. Woody and herbaceous aboveground production of a Patagonian steppe. Journal of Range Management 44: 434– 437. Google Scholar CrossRef Search ADS Fernández RJ, Nuñez AH, Soriano A. 1992. Contrasting demography of two Patagonian shrubs under different conditions of sheep grazing and resource supply. Oecologia 91: 39– 46. Google Scholar CrossRef Search ADS PubMed Field DL, Pickup M, Barrett SCH. 2013. Ecological context and metapopulation dynamics affect sex-ratio variation among dioecious plant populations. Annals of Botany 111: 917– 923. Google Scholar CrossRef Search ADS PubMed Gehring JL, Linhart YB. 1993. Sexual dimorphisms and response to low resources in the dioecious plant Silene latifolia (Caryophyllaceae). International Journal of Plant Sciences 154: 152– 162. Google Scholar CrossRef Search ADS Goldberg DE. 1990. Components of resource competition in plant communities. In: Grace JB, Tilman D, ed. Perspectives on Plant Competition . San Diego: Academic Press, 27– 49. Google Scholar CrossRef Search ADS Golluscio RA, Oesterheld M. 2007. Water use efficiency of twenty-five co-existing Patagonian species growing under different soil water availability. Oecologia 154: 207– 17. Google Scholar CrossRef Search ADS PubMed Golluscio RA, Oesterheld M, Aguiar MR. 2005. Relationship between phenology and life form: a test with 25 Patagonian species. Ecography 28: 273– 282. Google Scholar CrossRef Search ADS Graff P, Aguiar MR. 2017. Do species’ strategies and type of stress predict net positive effects in an arid ecosystem? Ecology 98: 794– 806. Google Scholar CrossRef Search ADS PubMed Graff P, Aguiar MR, Chaneton EJ. 2007. Shifts in positive and negative plant interactions along a grazing intensity gradient. Ecology 88: 188– 199. Google Scholar CrossRef Search ADS PubMed 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 Gross N, Börger L, Soriano-Morales SI, et al. 2013. Uncovering multiscale effects of aridity and biotic interactions on the functional structure of Mediterranean shrublands. Journal of Ecology 101: 637– 649. Google Scholar CrossRef Search ADS Hacker SD, Gaines SD. 1997. Some implications of direct positive interactions for community species diversity. Ecology 78: 1990– 2003. Google Scholar CrossRef Search ADS Holmgren M, Scheffer M, Huston MA. 1997. The interplay of facilitation and competition in plant communities. Ecology 78: 1966– 1975. Google Scholar CrossRef Search ADS Holmgren M, Gómez-Aparicio L, Quero JL, Valladares F. 2012. Non-linear effects of drought under shade: reconciling physiological and ecological models in plant communities. Oecologia 169: 293– 305. Google Scholar CrossRef Search ADS PubMed Hultine KR, Bush SE, West AG, et al. 2008. Gender-specific patterns of aboveground allocation, canopy conductance and water use in a dominant riparian tree species: Acer negundo. Tree Physiology 28: 1383– 1394. Google Scholar CrossRef Search ADS Hultine KR, Grady KC, Wood TE, et al. 2016. Climate change perils for dioecious plant species. Nature Plants 2: 16109. Google Scholar CrossRef Search ADS PubMed Juvany M, Munné-Bosch S. 2015. Sex-related differences in stress tolerance in dioecious plants: a critical appraisal in a physiological context. Journal of Experimental Botany 66: 6083– 6092. Google Scholar CrossRef Search ADS PubMed Lambers H, Chapin FS, Pons TL. 2008. Plant water relations. In Plant Physiological Ecology . New York, NY: Springer, 163– 223. Liancourt P, Spence LA, Song DS, et al. 2013. Plant response to climate change varies with topography, interactions with neighbors, and ecotype. Ecology 94: 444– 453. Google Scholar CrossRef Search ADS PubMed López RP, Squeo FA, Armas C, Kelt DA, Gutiérrez JR. 2016. Enhanced facilitation at the extreme end of the aridity gradient in the Atacama Desert: a community-level approach. Ecology 97: 1593– 1604. Google Scholar CrossRef Search ADS PubMed Lortie CJ, Reid AM. 2012. Reciprocal gender effects of a keystone alpine plant species on other plants, pollinators, and arthropods. Botany 90: 273– 282. Google Scholar CrossRef Search ADS Macek P, Prieto I, Macková J, Pistón N, Pugnaire FI. 2016. Functional plant types drive plant interactions in a Mediterranean mountain range. Frontiers in Plant Science 7: 662. Google Scholar CrossRef Search ADS PubMed Maestre FT, Cortina J. 2004. Do positive interactions increase with abiotic stress? A test from a semi-arid steppe. Proceedings of the Royal Society B: Biological Sciences 271: S 331– 333. Google Scholar CrossRef Search ADS Maestre FT, Callaway RM, Valladares F, Lortie CJ. 2009. Refining the stress-gradient hypothesis for competition and facilitation in plant communities. Journal of Ecology 97: 199– 205. Google Scholar CrossRef Search ADS McCluney KE, Belnap J, Collins SL, et al. 2012. Shifting species interactions in terrestrial dryland ecosystems under altered water availability and climate change. Biological Reviews of the Cambridge Philosophical Society 87: 563– 582. Google Scholar CrossRef Search ADS PubMed Michalet R. 2006. Is facilitation in arid environments the result of direct or complex interactions? New Phytologist 169: 3– 6. Google Scholar CrossRef Search ADS PubMed Michalet R. 2007. Highlighting the multiple drivers of change in interactions along stress gradients. New Phytologist 173: 3– 6. Google Scholar CrossRef Search ADS PubMed Michalet R, Xiao S, Touzard B, et al. 2011. Phenotypic variation in nurse traits and community feedbacks define an alpine community. Ecology Letters 14: 433– 443. Google Scholar CrossRef Search ADS PubMed Michalet R, Schöb C, Xiao S, et al. 2016. Beneficiary feedback effects on alpine cushion benefactors become more negative with increasing cover of graminoids and in dry conditions. Functional Ecology 30: 79– 87. Google Scholar CrossRef Search ADS Montesinos D, Verdú M, García-Fayos P. 2007. Moms are better nurses than dads: gender biased self-facilitation in a dioecious Juniperus tree. Journal of Vegetation Science 18: 271– 280. Google Scholar CrossRef Search ADS Moreno L, Bertiller MB. 2016. Variation of morphological and chemical traits in sexes of the dioecious perennial grass Poa ligularis in relation to shrub cover and aridity in Patagonian ecosystems. Population Ecology 58: 189– 197. Google Scholar CrossRef Search ADS Noy-Meir I. 1973. Desert ecosystems: environment and producers. Annual Review of Ecology and Systematics 4: 25– 51. Google Scholar CrossRef Search ADS Obeso JR. 2002. The costs of reproduction in plants. New Phytologist 155: 321– 348. Google Scholar CrossRef Search ADS Palmer MW, Dixon PM. 1990. Small-scale environmental heterogeneity and the analysis of species distributions along gradients. Journal of Vegetation Science 1: 57– 65. Google Scholar CrossRef Search ADS Paruelo JM, Aguiar MR, Golluscio RA. 1988. Soil water availability in the Patagonian Arid Steppe: gravel content effect. Arid Soil Research and Rehabilitation 2: 67– 74. Google Scholar CrossRef Search ADS Reisman-Berman O. 2007. Age-related change in canopy traits shifts conspecific facilitation to interference in a semi-arid shrubland. Ecography 30: 459– 470. Google Scholar CrossRef Search ADS Sala OE, Golluscio RA, Lauenroth WK, Soriano A. 1989. Resource partitioning between shrubs and grasses in the Patagonian steppe. Oecologia 81: 501– 505. Google Scholar CrossRef Search ADS PubMed SAS. 1996. SAS System for Windows, Release 6.12. User’s Guide. Cary, NC: SAS Institute, Inc. Schlaepfer DR, Bradford JB, Lauenroth WK, et al. 2017. Climate change reduces extent of temperate drylands and intensifies drought in deep soils. Nature Communications 8: 14196. Google Scholar CrossRef Search ADS PubMed Schöb C, Armas C, Guler M, Prieto I, Pugnaire FI. 2013. Variability in functional traits mediates plant interactions along stress gradients. Journal of Ecology 101: 753– 762. Google Scholar CrossRef Search ADS Semchenko M, Abakumova M, Lepik A, Zobel K. 2013. Plants are least suppressed by their frequent neighbours: the relationship between competitive ability and spatial aggregation patterns. Journal of Ecology 101: 1313– 1321. Google Scholar CrossRef Search ADS Semchenko M, Lepik A, Abakumova M, Zobel K. 2017. Different sets of belowground traits predict the ability of plant species to suppress and tolerate their competitors. Plant and Soil 424: 157– 169. Google Scholar CrossRef Search ADS Siefert A, Violle C, Chalmandrier L, et al. 2015. A global meta-analysis of the relative extent of intraspecific trait variation in plant communities. Ecology Letters 18: 1406– 1419. Google Scholar CrossRef Search ADS PubMed Soriano A, Sala OE. 1986. Emergence and survival of Bromus setifolius seedlings in different microsites of a Patagonian arid steppe. Israel Journal of Botany 35: 91– 100. Tirado R, Bråthen KA, Pugnaire FI. 2015. Mutual positive effects between shrubs in an arid ecosystem. Scientific Reports 5: 14710. Google Scholar CrossRef Search ADS PubMed Valladares F. 2004. Photoinhibition and drought in Mediterranean woody saplings: scaling effects and interactions in sun and shade phenotypes. Journal of Experimental Botany 56: 483– 494. Google Scholar CrossRef Search ADS PubMed Valladares F, Pearcy RW. 2002. Drought can be more critical in the shade than in the sun: a field study of carbon gain and photo-inhibition in a Californian shrub during a dry El Nino year. Plant, Cell and Environment 25: 749– 759. Google Scholar CrossRef Search ADS Valladares F, Zaragoza-Castells J, Sánchez-Gómez D, et al. 2008. Is shade beneficial for Mediterranean shrubs experiencing periods of extreme drought and late-winter frosts? Annals of Botany 102: 923– 933. Google Scholar CrossRef Search ADS PubMed Verdú M, García-Fayos P. 1998. Female biased sex ratios in Pistacia lentiscus L. (Anacardiaceae). Plant Ecology 135: 95– 101. Google Scholar CrossRef Search ADS Verdu M, Villar-Salvador P, Garcia-Fayos P. 2004. Gender effects on the post-facilitation performance of two dioecious Juniperus species. Functional Ecology 18: 87– 93. Google Scholar CrossRef Search ADS Wang P, Stieglitz T, Zhou DW, CahillJr JF. 2010. Are competitive effect and response two sides of the same coin, or fundamentally different? Functional Ecology 24: 196– 207. Google Scholar CrossRef Search ADS Yang Z, Liu X, Zhou M, et al. 2015. The effect of environmental heterogeneity on species richness depends on community position along the environmental gradient. Scientific Reports 5: 15723. Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For permissions, please e-mail: firstname.lastname@example.org. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
Annals of Botany – Oxford University Press
Published: May 22, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera