Abstract The planktonic rotifer, Brachionus calyciflorus, displays extensive variation in the length of its anterior and posterior spines. Notably, posterolateral spines may be absent or near body length. Studies of laboratory and natural populations have identified the different factors controlling this variation and have investigated the trade-offs associated with increased spine development. Low temperature and low food availability can induce modest spine elongation that may reduce sinking rate. A kairomone released by the carnivorous rotifer Asplanchna induces pronounced spine elongation, without detectable reproductive cost, that can provide an effective defense against this predator. Endogenous mechanisms also operate: Spine development is inhibited in females hatched from fertilized resting eggs and can be promoted by increasing maternal age. Genetic variation for the length of spines in both noninduced (default) and induced phenotypes occurs among and within populations. Asplanchna in natural communities likely leads to seasonal selection for genotypes that can develop increasingly long spines. The rotifer Brachionus calyciflorus (figure 1) is a small (approximately 200 micrometers, μm) metazoan (phylum Rotifera) that is common and often very abundant in zooplankton communities of nutrient-rich freshwaters. It is a model organism for the study of morphological variation. The variation can be extensive, with very different forms once given separate names, and is controlled largely by environmental and endogenous factors, as well as by genetics. The ability of a given genotype to express very different phenotypes is a remarkable case of phenotypic plasticity in the animal kingdom. Transformations from one phenotype to another can occur rapidly and be of considerable ecological significance. In this article, I describe studies showing how nongenetic and genetic factors interact to affect the morphology and fitness of this rotifer in changing environments. Figure 1. View largeDownload slide An amictic female of Brachionus calyciflorus (Lake Littra, Australia) carrying three oviposited diploid eggs developing parthenogenetically into females. Photographed alive by author. Figure 1. View largeDownload slide An amictic female of Brachionus calyciflorus (Lake Littra, Australia) carrying three oviposited diploid eggs developing parthenogenetically into females. Photographed alive by author. General morphology, life cycle, and distribution The integument of the rotifer body contains an internal skeleton or lorica into which the anterior corona and posterior foot can be withdrawn. The ciliated corona is used for locomotion and creating feeding currents from which small (mostly 2–20 μm) particles (e.g., unicellular algae, other protists, bacteria, and detritus) are collected and transferred down the esophagus to the stomach and intestine. Although the rotifer has a foot that can potentially attach to substrata, it is almost always free swimming. B. calyciflorus and other small planktonic rotifers are eaten by the large predatory rotifer Asplanchna, some crustaceans (copepods and large cladocerans), a few insects (backswimmers and the larvae of the dipteran Chaoborus), and small fish. Therefore, these rotifers are important links in the transfer of energy to higher trophic levels in community food webs. B. calyciflorus can reproduce rapidly via female parthenogenesis to produce clonal populations. One at a time, diploid oocytes in the ovary receive yolk from the vitellarium via a cytoplasmic bridge, undergo a single mitotic maturation division, and are extruded from the mother's body cavity (pseudocoelom), although they are attached to her cloaca by a fine thread (figure 1). After oviposition, the eggs begin cleavage divisions and develop directly to hatch as juvenile females. These amictic, or female-producing, females may carry several eggs at a time, each egg in a different stage of development (figure 1), and produce a series of 10 to 20 daughters during their lifetime (1–2 weeks at 20 degrees Celsius,°C). Intermittent bisexual reproduction is initiated in expanding populations by a quorum-sensing mechanism, in which accumulation of a self-produced infochemical induces some oocytes of amictic females to develop into mictic females. These morphologically similar females produce eggs that undergo typical meiosis and are haploid. If unfertilized, these much smaller eggs develop parthenogenetically into diminutive males (figure 2a, 2b). If fertilized, they develop into dark-brown, energy-rich, diapausing embryos encased in a thick, multilayered wall (figure 2c). These so-called resting eggs sink to bottom sediments and can hatch as amictic stem females after some days, months, or years. The factors controlling the production of mictic females and the biology of resting eggs in this and other rotifers are reviewed by Gilbert (2003, 2007, 2017a, 2017b) Gilbert and Schröder (2004), and Snell (2011). Figure 2. View largeDownload slide Sexual reproduction in Brachionus calyciflorus (Lake Littra, Australia). (a) A mictic female with small, haploid eggs developing parthenogenetically into males. (b) Male. (c) A fertilized mictic female carrying two resting eggs. All photographed alive by author. Figure 2. View largeDownload slide Sexual reproduction in Brachionus calyciflorus (Lake Littra, Australia). (a) A mictic female with small, haploid eggs developing parthenogenetically into males. (b) Male. (c) A fertilized mictic female carrying two resting eggs. All photographed alive by author. B. calyciflorus is widely distributed in both the Eastern and Western Hemispheres. Genetic analyses using mitochondrial and nuclear DNA sequences of clones from different locations in several geographic areas (China, The Netherlands, the southern United States) show that, even within each area, the species is a complex of morphologically similar but genetically distinct groups of putative cryptic species (Gilbert and Walsh 2005, Xiang et al. 2011a, 2011b, Papakostas et al. 2016). The four groups from The Netherlands defined by nuclear sequence differences may hybridize (Papakostas et al. 2016) but are morphologically distinct on the basis of lorica details (Papakostas et al. 2016, Evangelia Michaloudi, Department of Zoology, Aristotle University of Thessaloniki, Thessaloniki, personal communication, 1 November 2017). Genetically distinct groups of B. calyciflorus may be reproductively isolated from one another, either because of a failure to mate or because impregnated females do not form resting eggs that hatch or develop into viable females. The mechanism of isolation between two Chinese groups (Xiang et al. 2011a) was not determined. However, males from some geographic areas clearly will not mate with females from another; they may show typical precopulatory behavior after an encounter but do not initiate copulation (Gilbert and Walsh 2005). Reciprocal cross-copulation occurred between males and females from Florida and Georgia, but not between those from Florida, Texas, and Australia (Gilbert and Walsh 2005, Gilbert 2017c). Similarly, no reciprocal cross-copulation occurred between males and females of strains from Ascarate Lake, El Paso, Texas, and a temporary pond in Bariloche, Argentina. Most currently described and morphologically distinct species of Brachionus and other rotifers probably are groups of multiple, genetically discrete groups, many of which are reproductively isolated from one another. For example, studies on the genetics, morphology, mating behavior, and ecology of Brachionus plicatilis from brackish waters worldwide indicate a complex of 15 cryptic species (Mills et al. 2017). Spines and their variation Spiny outgrowths of the integument or body wall of B. calyciflorus are a prominent feature (figure 3). The number and especially the length of these spines can differ greatly among populations and among individuals within a population, both at a given time and from one time to another. Two pairs of anterior spines and a pair of posteromedian spines at the cloaca are always present (or constitutive), and a pair of posterolateral spines may be absent, short, or very long (figure 3). In phenotypes with the least spine development, the anterior and posteromedian spines are short, and posterolateral spines are absent (figure 3a). In phenotypes with pronounced spine development, the anterior and posteromedian spines are longer, and posterolateral spines are very long (figure 3c). Figure 3. View largeDownload slide Variation in spine development in Brachionus calyciflorus (Lake Littra, Australia). (a) An individual with short spines and no posterolateral spines. (b) An individual with longer spines and short posterolateral spines. (c) An individual with very long spines and especially long posterolateral spines. All preserved in Lugol's solution and photographed by author. Figure 3. View largeDownload slide Variation in spine development in Brachionus calyciflorus (Lake Littra, Australia). (a) An individual with short spines and no posterolateral spines. (b) An individual with longer spines and short posterolateral spines. (c) An individual with very long spines and especially long posterolateral spines. All preserved in Lugol's solution and photographed by author. Early observations on spine variation Although comparable variation in spine development occurs in many other rotifers belonging to the family Brachionidae (reviews by Gilbert 1999, 2013, 2017a), the pronounced variation in B. calyciflorus attracted very early attention and has been the focus of extensive field and experimental studies. By the late nineteenth century, the English naturalist P. H. Gosse (1889) concluded that forms with different degrees of spine development, previously given different names, belong to a single species now recognized as B. calyciflorus Pallas, 1766. Weber (1898) and Kofoid (1908) found individuals with short and long spines occurring together in plankton collected around Geneva and from the Illinois River. Dieffenbach and Sachse (1911) and Wesenberg-Lund (1930) reported seasonal progressions between these forms in German and Danish ponds, with long-spined individuals often replacing short-spined ones. Most importantly, they found that females with no posterolateral spines could produce eggs that develop parthenogenetically into daughters with long posterolateral spines and vice versa. Therefore, it became clear that there is a strong nongenetic component to the variation. Wesenberg-Lund (1930) was the first to consider factors in the environment that might affect spine development in B. calyciflorus. However, after observing populations in his Danish ponds over many years, he wrote, “All attempts to connect (spine length) with the habitat or external conditions are, as far as I can see, quite fruitless” (Wesenberg-Lund, 1930, p. 114). A breakthrough occurred in 1952, when the zoologist and rotifer biologist Paul Marais de Beauchamp published a three-page note in the Proceedings of the French Academy of Sciences about an environmental factor controlling spine development in his laboratory cultures. This factor was the large, ovoviviparous rotifer Asplanchna (figure 4). Asplanchna feeds by bumping into prey as it swims randomly through the environment. When its coronal receptors detect a suitable prey item, often a smaller rotifer, its dilatable mouth opens as its pharynx expands to suck in the prey. If the prey can enter the pharynx, it is then either transferred through the esophagus into the blind stomach or ejected from the mouth. Prey capture can be very rapid. Asplanchna girodi initiated pharyngeal expansion approximately 93 milliseconds after contacting a small rotifer (Synchaeta oblonga) and captured it in the pharynx approximately 52 milliseconds later (Gilbert and Stemberger 1985). Figure 4. View largeDownload slide The predatory rotifer Asplanchna brightwellii (Lake Littra, Australia). Adult showing brown stomach and developing embryos in the uterus. Photographed alive by author. Figure 4. View largeDownload slide The predatory rotifer Asplanchna brightwellii (Lake Littra, Australia). Adult showing brown stomach and developing embryos in the uterus. Photographed alive by author. Beauchamp's cultures of B. calyciflorus typically contained individuals with no or short posterolateral spines, even when they were initiated with long-spined individuals collected from the field. However, when his cultures also contained Asplanchna, the B. calyciflorus soon developed long posterolateral spines. He concluded that Asplanchna was releasing into the medium a factor to which B. calyciflorus responded in a very adaptive way by producing daughters with long, defensive spines. Although Asplanchna could eat individuals with short spines, it had enough difficulty ingesting those with long spines that both predator and prey populations could coexist. Beauchamp's study was the first report of a predator-induced morphological defense in any organism, a phenomenon now known to also occur in some algae, ciliated protozoa, other rotifers, crustaceans, and fish (Tollrian and Harvell 1999). However, at the time, his discovery failed to attract the attention it deserved. It may have been unnoticed, disregarded, or possibly disbelieved because of the extreme novelty of such a rapid, adaptive response. G. Evelyn Hutchinson, an eminent ecologist and limnologist at Yale University, knew of Beauchamp's paper and directed me to it when I became his graduate student in 1959. When I joined the Department of Zoology at the University of Washington in 1963 as a National Institutes of Health postdoctoral fellow to study rotifer development with W. Thomas Edmondson (who completed his doctoral research on rotifers with Hutchinson in 1942), I was able to confirm and extend Beauchamp's findings (Gilbert 1966, 1967). A few years later, Udo Halbach at the Ludwig Maximilian University of Munich reported extensive laboratory experiments and field observations that did the same (Halbach 1970, 1971). Research from 1952 to the present has shown that spine development in B. calyciflorus is controlled by some other environmental factors besides Asplanchna, by two physiological or endogenous factors, and by genotype. Although there is still much to learn about the developmental biology and evolution of spine-length variation in this rotifer, our present knowledge provides an excellent example of how multiple nongenetic and genetic factors can interact to control an ecologically important morphological trait. Environmental control of spine development Response to Asplanchna The spine-development response of B. calyciflorus to Asplanchna is very specific. All species of Asplanchna that have been tested induce the response, but no other rotifer or predator is known to do so. The response to Asplanchna requires no physical contact with the predator. It is induced by a chemical, or kairomone, leaking out of Asplanchna into the environment. The response involves elongation of the anterior and posteromedian spines, de novo induction or elongation of posterolateral spines that are already present or constitutive, and an increase in body length (lorica without spines). In addition, an increase in lorica thickness and hardness has been reported (Yin et al. 2017). The Asplanchna-controlled spine-length variation is continuous, with the extent of spine elongation depending on the exposure to Asplanchna. In one experiment (Gilbert 2011), posterolateral spine length gradually increased from about 5 μm (for the basic or default phenotype) to 50 μm as Asplanchna density increased from 0 to 8 individuals per liter (figure 5). The maximal effect of Asplanchna on spine development in a clone from Lake Littra, Australia, with constitutive posterolateral spines is shown in table 1. Figure 5. View largeDownload slide The effect of Asplanchna brightwellii density on spine development in Brachionus calyciflorus. Values are means ± 1 SE. From Gilbert (2011). Figure 5. View largeDownload slide The effect of Asplanchna brightwellii density on spine development in Brachionus calyciflorus. Values are means ± 1 SE. From Gilbert (2011). Table 1. The effect of Asplanchna girodi on body and spine lengths (μm) of adult females from a clone of Brachionus calyciflorus (Lake Littra, Australia). Spines Body PL PM AL AM Treatment Mean (M) 95% confidence interval (CI) M 95% CI M 95% CI M 95% CI M 95% CI Total (body + PL + AM) Control 193.2 4.9 73.9 3.3 29.6 3.1 41.1 2.4 61.4 2.9 328.5 Asplanchna 229.9 6.3 206.7 7.6 95.9 4.3 71.0 3.7 84.1 3.5 520.7 Spines Body PL PM AL AM Treatment Mean (M) 95% confidence interval (CI) M 95% CI M 95% CI M 95% CI M 95% CI Total (body + PL + AM) Control 193.2 4.9 73.9 3.3 29.6 3.1 41.1 2.4 61.4 2.9 328.5 Asplanchna 229.9 6.3 206.7 7.6 95.9 4.3 71.0 3.7 84.1 3.5 520.7 Abbreviations: The spines are PL for posterolateral, PM for posteromedian, AL for anterolateral, AM for anteromedian. Note: From Gilbert (2017c). View Large The Asplanchna kairomone has not been identified but is proteinaceous and unstable under natural conditions. Medium conditioned by live Asplanchna lost all activity when incubated with 0.01% Pronase at 37°C for 2 hours (Gilbert 1966, 1967) and after several days at 20°C–25°C (Gilbert 1967, Halbach 1970). At 20°C, it had a half-life of approximately 1.5 days (Halbach 1970). Therefore, the concentration of Asplanchna kairomone at any time depends on the rate of its production and degradation. Hopefully, future work with modern techniques will purify and identify the kairomone. The critical or labile period for Asplanchna-induced spine development occurs during oogenesis, when individuals are oocytes in the maternal body cavity (Gilbert 1966, 1967, 1999, 2012). The kairomone may act directly on the oocyte or indirectly by triggering a change in maternal physiology. Therefore, the phenotype of an individual is fixed for life, or irreversible, after the egg from which it develops is oviposited. Even so, the developmental response to Asplanchna can occur very rapidly. When adult females without posterolateral spines were exposed to Asplanchna after oviposition of their first egg, their second or third and all subsequently oviposited eggs developed into females with long posterolateral spines (Gilbert 2012). In the absence of Asplanchna kairomone, Asplanchna-induced spine development is reversed after one or two generations. Newborn individuals with long, induced spines produced daughters with greatly reduced spines (Gilbert 2012, Yin et al. 2015a). Incomplete reversal after one generation likely is due to the transfer of residual inducer from the mother to her oocytes. During postnatal growth of individuals with Asplanchna-induced spines, posterolateral spine length increases more slowly than body length. Therefore, young individuals have disproportionately long spines. In some results extracted from Gilbert (1967) and later analyzed, six neonates with body lengths of 140 μm were measured again as adults with a mean body length of 191 μm. Mean ratios of spine to body length decreased 18% from 0.62 (standard error (SE) = 0.06) to 0.51 (SE = 0.05) [paired t-test; t5 = 10.44; p = .0001]. The negative allometric growth of the posterolateral spines was the same with and without Asplanchna kairomone. A newborn female from an Australian clone with especially pronounced Asplanchna-induced spines is shown in figure 6; here, the ratio of posterolateral-spine length to body length is 0.95. Figure 6. View largeDownload slide A newborn Brachionus calyciflorus (Lake Littra, Australia) with long Asplanchna-induced spines. (a) Normal position of posterolateral spines during swimming with corona extended. (b) Lateral extension of articulating posterolateral spines in same individual with corona retracted (and foot extended). Photographed alive by author. Figure 6. View largeDownload slide A newborn Brachionus calyciflorus (Lake Littra, Australia) with long Asplanchna-induced spines. (a) Normal position of posterolateral spines during swimming with corona extended. (b) Lateral extension of articulating posterolateral spines in same individual with corona retracted (and foot extended). Photographed alive by author. Ecological significance of response to Asplanchna In natural ecosystems, carnivorous species of Asplanchna commonly occur with B. calyciflorus. When they can eat this rotifer in laboratory cultures, they can reproduce very rapidly (Gilbert 1977, 2016, Sarma et al. 1998, 2002, 2003, Nandini et al. 2003, Santos-Medrano et al. 2017). In one experiment at 26°C, amictic females of Asplanchna sieboldii fed B. calyciflorus produced an average of 18.8 offspring during their lifetime, resulting in a 5.5-fold increase in population size per day (Gilbert 1977). The rate at which Asplanchna can ingest B. calyciflorus is highly variable and depends on many factors: the species of Asplanchna, the body sizes of Asplanchna and B. calyciflorus (both of which increase greatly from neonate to adult), the temperature, the hunger level of Asplanchna, the density of B. calyciflorus, and the spine development of B. calyciflorus. B. calyciflorus of the basic morph with minimal spine development can be readily eaten. Satiated adult A. brightwellii at 20°C ate approximately 12 juveniles or approximately 6 adults in 30 minutes (Halbach 1971). Juvenile A. sieboldii at 25°C, and starved for 8 hours, ate approximately 10 or approximately 20 individuals in 30 minutes when these were at densities of 4 and 16 individuals per milliliters, respectively (Nandini and Sarma 1999). B. calyciflorus individuals with Asplanchna-induced spines are more difficult for Asplanchna to capture and ingest than those of the noninduced basic morph (table 2; Gilbert 1967, 1980, Halbach 1971, Iyer and Rao 1996). Their somewhat larger bodies and much longer spines greatly increase their total length (table 1) and thus can mechanically inhibit passage into Asplanchna’s pharynx or esophagus. In addition, their articulating posterolateral spines fling out laterally after an encounter with Asplanchna, making them even more difficult to capture and ingest (figures 6 and 7). This spine movement is due to contact-induced coronal retraction and the resulting increase in internal hydrostatic pressure. Figure 7. View largeDownload slide The Asplanchna-induced spines of Brachionus calyciflorus can prevent capture by Asplanchna brightwellii. Photographed alive by author in Gilbert (1966). Figure 7. View largeDownload slide The Asplanchna-induced spines of Brachionus calyciflorus can prevent capture by Asplanchna brightwellii. Photographed alive by author in Gilbert (1966). Table 2. The posterolateral-spine lengths of juvenile and adult Brachionus calyciflorus affect the probability of capture and ingestion by Asplanchna brightwellii. Brachionus Proportion captured in pharynx after contact Proportion ingested after capture Juveniles (approximately 150 μm) No spines 0.95 1.00 Short spines 0.90 1.00 Long spines 0.20 0.29 Adults (approximately 225 μm) No spines 0.80 1.00 Short spines 0.40 0.63 Long spines 0.20 0 Brachionus Proportion captured in pharynx after contact Proportion ingested after capture Juveniles (approximately 150 μm) No spines 0.95 1.00 Short spines 0.90 1.00 Long spines 0.20 0.29 Adults (approximately 225 μm) No spines 0.80 1.00 Short spines 0.40 0.63 Long spines 0.20 0 Note: Short and long spines are defined as their proportion of body length: 0.1–0.3 and 0.5–1.0, respectively. The Asplanchna are young females starved for 6 hours after being well fed for 12 hours. Probabilities are means of 20 trials conducted at 20°C. From Halbach (1971). View Large Asplanchna-induced spines can protect even the small and most vulnerable juveniles of B. calyciflorus. In some direct observations made by Halbach (1971), also shown in table 2, the probability of Asplanchna ingesting them after an encounter was 0.95 for those with no posterolateral spines and only 0.06 for those with long ones. The disproportionally longer spines of small juveniles (see above) are therefore especially valuable. The highly adaptive morphological response of B. calyciflorus to Asplanchna demonstrated in laboratory experiments certainly also occurs in natural ecosystems. Two lines of indirect evidence support this relationship. First, published records of zooplankton communities from many geographic areas show that long-spined individuals of B. calyciflorus commonly occur when Asplanchna is present (Gilbert 1967, Green 1974, Bertani et al. 2013). In addition, multiple observations of zooplankton in many small ponds over several months showed that the posterolateral-spine length of B. calyciflorus was directly related to the population density of Asplanchna (Halbach 1970, Halbach and Jacobs 1971). A highly significant relationship (R = +0.91) in one pond between Asplanchna density and the ratio of posterolateral spine length to body length of concurrently developed, neonate B. calyciflorus is shown in figure 8b and 8c. Figure 8. View largeDownload slide Seasonal changes in a German pond with Brachionus calyciflorus and Asplanchna. (a) Temperature. (b) Population densities of Brachionus and Asplanchna. (c) The ratio of posterolateral spine length to body length of neonate Brachionus in the pond and the maximum for this ratio in the offspring of Brachionus collected from the field and exposed to Asplanchna in a laboratory bioassay. The means ± 1 SE are within symbols. (d) The coefficient of variation for maximum spine length ratio. From Halbach and Jacobs (1971). Figure 8. View largeDownload slide Seasonal changes in a German pond with Brachionus calyciflorus and Asplanchna. (a) Temperature. (b) Population densities of Brachionus and Asplanchna. (c) The ratio of posterolateral spine length to body length of neonate Brachionus in the pond and the maximum for this ratio in the offspring of Brachionus collected from the field and exposed to Asplanchna in a laboratory bioassay. The means ± 1 SE are within symbols. (d) The coefficient of variation for maximum spine length ratio. From Halbach and Jacobs (1971). Second, Asplanchna densities that induce spine development in laboratory populations of B. calyciflorus are much lower than those frequently occurring in natural communities. In one experiment, 1.2 and 8 Asplanchna brightwellii per liter induced 50% maximal and maximal length of the posterolateral spines, respectively (figure 5; Gilbert 2011). Asplanchna densities associated with pronounced spine development in Halbach's ponds were 10–100 or more individuals per liter. Most convincingly, there is strong direct evidence for the control of B. calyciflorus spine length by Asplanchna kairomone in natural systems. Laboratory bioassays of filtered water from several ponds containing Asplanchna and long-spined B. calyciflorus showed that the water alone induced females without posterolateral spines to produce daughters with long ones (Gilbert 1967, Gilbert and Waage 1967, Halbach 1970). Furthermore, studies in German ponds demonstrated that the posterolateral spine length of B. calyciflorus was nicely correlated with both the density of Asplanchna and the ability of bioassayed water to induce spine development (Halbach 1970). The tendency of posterolateral spine length to closely track both increases and decreases in Asplanchna density can be attributed to the rapid response of B. calyciflorus to the presence of Asplanchna, the rapid decay of the Asplanchna kairomone, and the reversibility of the response after one or two generations. Although Asplanchna-induced spine development must effectively reduce Asplanchna predation on B. calyciflorus in natural communities, it still may not prevent high mortality from this predator. For example, Bertani and colleagues (2013) found that Asplanchna strongly depressed B. calyciflorus abundance even though the B. calyciflorus had developed long spines. Green and Lan (1974) noted that the proportion of long-spined B. calyciflorus in the stomachs of adult A. brightwellii was just slightly lower than that in the rotifer plankton. However, they did not compare lorica and spine lengths of ingested individuals with those in the plankton, and they did not analyze the stomach contents of smaller A. brightwellii less able to eat large and long-spined Brachionus. Asplanchna-induced spine development is not known to affect the susceptibility of B. calyciflorus to any other predator. Certainly it had no effect against adult females of the copepod Mesocyclops edax; all B. calyciflorus with either very short or long posterolateral spines (spine length to body length ratios of 0.09 and 0.54, respectively) were captured and ingested after attacks (Gilbert 1980). Responses to temperature and food concentration Moderate spine development in B. calyciflorus can be induced both by low temperature (Gilbert 1967, Halbach 1970, Pourriot 1973) and low food availability (Halbach 1970, Stemberger 1990). Halbach nicely demonstrated the independent and additive effects of temperature, food level, and Asplanchna kairomone on the ratio of posterolateral spine length to body length in adults of two German strains of B. calyciflorus (figure 9). As temperature decreased from 25°C to 10°C at the high food level without Asplanchna kairomone, this ratio increased gradually from negligible to about 0.2. Also, without Asplanchna kairomone, spine lengths were appreciably higher at the lower food concentration at 25°C and 20°C. However, even with the additive effects of low temperature and low food availability, the spine-to-body-length ratio never exceeded 0.2. Asplanchna kairomone induced much longer spines in all treatments, especially at the lower temperatures and food level, with the highest spine-to-body-length ratios (approximately 0.7) occurring between 20°C and 10°C. Figure 9. View largeDownload slide The additive effects of temperature, food level, and Asplanchna on spine development in Brachionus calyciflorus. From Halbach (1970). Figure 9. View largeDownload slide The additive effects of temperature, food level, and Asplanchna on spine development in Brachionus calyciflorus. From Halbach (1970). The ecological significance of the responses to temperature and food level is not clear. Increased form resistance due to somewhat greater spine lengths at low temperatures and food concentrations may reduce sinking rate and thus the energy needed to maintain position in the water column (Hutchinson 1967, Stemberger 1990). Responses of other rotifers to environmental factors Some other species of Brachionus, and some species of Keratella, exhibit similar spine-development responses to Asplanchna (reviews by Gilbert 1999, 2013, 2017a). Although Asplanchna is the only predator known to induce spine development in B. calyciflorus and other congeners, crustaceans (cladocerans, such as Daphnia, and copepods), as well as Asplanchna, can induce spine development in several species of Keratella. In Keratella tropica, the response to Asplanchna is different from and much more pronounced than that to crustaceans (Gilbert 2011a). In addition, some rotifers besides B. calyciflorus develop longer spines at low temperatures. This has been observed in seasonal studies of natural populations of Keratella cochlearis (Conde-Porcuna et al. 1993, Green 2005) and demonstrated in laboratory experiments with both K. cochlearis (Lindstrom and Pejler 1975) and K. tropica (Gilbert 2011b). In K. tropica, the spine-development responses to Daphnia and low temperature are additive (Gilbert 2011b). Endogenous control of spine development Two endogenous or physiological factors can affect spine development in B. calyciflorus. First, as was noted by early and recent investigators, individuals hatched from resting eggs typically have the basic or default phenotype. They have short spines, with no or extremely short posterolateral spines, even if their mictic-female mothers had long posterolateral spines. Fertilization and development of the resting-egg embryo may somehow inhibit spine development. Second, increasing maternal age alone can promote spine development in some clones. When populations of some B. calyciflorus clones from Florida and Georgia were cultured under environmental conditions that induced no spine development (no Asplanchna kairomone, 20°C and a high concentration of the alga Cryptomonas erosa), some individuals had no posterolateral spines, some had short ones, and others had ones as long as those typically induced by Asplanchna. This variation was associated with maternal age (Schröder and Gilbert 2009). Ratios of posterolateral spine length to body length gradually increased from first-born to twelfth-born females: from approximately 0.15 to 0.5 for Florida Clone 1 adults and from 0 to approximately 0.4 for Florida Clone 2 adults (figure 10). Figure 10. View largeDownload slide The effect of maternal age, or birth order, on spine development in two clones of Brachionus calyciflorus (Florida, United States). Data from Schröder and Gilbert (2009). Figure 10. View largeDownload slide The effect of maternal age, or birth order, on spine development in two clones of Brachionus calyciflorus (Florida, United States). Data from Schröder and Gilbert (2009). This maternal-age effect is a striking example among animals of how the age of a female can have a major effect on the morphology of her offspring. It may have evolved as a bet-hedging strategy to protect late-born individuals from Asplanchna in case Asplanchna occurs at densities too low to induce spine development (Schröder and Gilbert 2009, Gilbert and McPeek 2013). Also, it should favor these individuals whenever Asplanchna does induce spine development, because the effects of birth order and Asplanchna can be additive (Gilbert 2012). A model using life-table and spine-length data demonstrated the impact of such a pronounced maternal-age effect on the frequency of posterolateral-spine lengths in B. calyciflorus populations (Gilbert and McPeek 2013). In the absence of Asplanchna, and therefore with no selection for long spines, individuals of early birth orders with no or short spines soon dominate the population. However, if Asplanchna occurs and preys selectively on individuals with no or short spines, the frequency of late-born individuals with longer defensive spines greatly increases. Genetic control of spine development Constitutive posterolateral spines B. calyciflorus has some spines that always are present, or constitutive, but whose length can be affected by environmental and endogenous factors: two pairs of anterior spines and a pair of posteromedian spines. Posterolateral spines may or may not be constitutive. Females of some well-studied clones had no constitutive posterolateral spines and developed these spines only when they were induced by low temperature, low food availability, or Asplanchna kairomone (Gilbert 1966, 1967, Halbach 1970). In different clonal populations from Lake Littra, Australia, females never, sometimes, or always had short or moderately long constitutive posterolateral spines (Gilbert 2017c). Responses to maternal age There can be considerable genetic variation for the maternal-age effect (Schröder and Gilbert 2009). Florida Clone 1 had longer posterolateral spines than Florida Clone 2 at all birth orders (figure 10). Among seven clones from a Georgia strain, four showed spine-length increases with birth order in most or all generations to different extents, and three never developed posterolateral spines over six generations. A clone from Lake Littra, Australia, with constitutive posterolateral spines (figure 3b) exhibited a very modest but significant maternal-age effect (Gilbert 2017c). The mean length of these spines in adults gradually increased from 62 μm for first-born females to 78 μm for fourteenth-born females. Responses to environmental factors All populations of B. calyciflorus appear to have the potential to develop long spines in the presence of Asplanchna. The association between long-spined individuals and Asplanchna in natural plankton communities is worldwide (see above). Furthermore, all strains of B. calyciflorus so far cultured with Asplanchna in the laboratory showed strong spine-development responses. Tested strains were from Australia (Gilbert 2017c); Chile (Aránguiz-Acuña et al. 2010); China (Yin et al. 2015); France (Beauchamp 1952); Germany (Halbach 1970); Mexico (Gama-Flores et al. 2011, Sarma et al. 2011); and Florida, Texas, and Washington in the United States (Gilbert 1967, 2012). However, there is genetic variation for the degree of spine development that can be induced. Beauchamp (1952) compared the effect of Asplanchna on populations from two localities in Paris and found that one developed much longer spines than the other. Similarly, Halbach (1970) found that clone Cn from Wolfsee near Scheinfeld produced longer spines in response to Asplanchna, low temperature, and low food concentration than clone T-3 from a fishpond near Winterhausen (figure 9). Genetic variation for the maximal spine-development response of B. calyciflorus to Asplanchna also likely occurs within populations. This variation could be introduced when fertilized resting eggs from a sediment egg bank hatch into stem females that then develop clones within the planktonic population (Gilbert 2017b). Halbach and Jacobs (1971) found strong evidence for such clonal variation in a pond near Würzburg, Germany (see below). Genetic variation for Asplanchna-induced spine development also occurs in the Lake Littra population of B. calyciflorus, because the effects of genotype and Asplanchna on posterolateral spine length are additive (Gilbert 2017c). For example, with Asplanchna, a clone with constitutive posterolateral spines developed much longer posterolateral spines than one without them: on average, from 74 to 207 μm in the former (table 1) and from 0 to 159 μm in the latter. Ratios of posterolateral spine length to lorica length for these Asplanchna-induced adults averaged 0.90 and 0.76, respectively. Clonal variation for the presence of constitutive posterolateral spines in Lake Littra could be maintained by varying directional selection (Gilbert 2017c). Clones with no constitutive posterolateral spines may have an advantage when Asplanchna is rare or absent if there is a cost associated with greater spine development (see below). On the other hand, clones with constitutive posterolateral spines, and therefore an ability to develop especially long spines in the presence of Asplanchna, may be more fit when Asplanchna predation is intense. Microevolution of spine length in natural populations Because natural populations of B. calyciflorus may have genetic variation for the magnitude of spine development induced by Asplanchna, increases in the abundance of Asplanchna in plankton communities may cause rapid, seasonal selection for clones developing the longest spines. Halbach and Jacobs (1971) provided evidence for such microevolution in the Würzburg pond. They showed that the strong temporal relationship between the abundance of Asplanchna and the posterolateral spine length of newborn Brachionus (figure 8b, 8c) could be explained in part by an increase in the frequency of clones able to develop the longest spines. When 10 amictic females were isolated on each of eight dates between January and mid-September and then were exposed to Asplanchna in a laboratory bioassay to determine the spine lengths of their offspring, the mean maximum ratio of spine to body length of the offspring increased from approximately 0.25 in January to approximately 0.5 in September (figure 8c). Also, the coefficient of variation for the mean ratio of spine length to body length of these offspring decreased (figure 8d), indicating that clonal diversity in the Brachionus population for Asplanchna-induced spine-length development was high in January and February and then decreased as Asplanchna selected for the most responsive clones. The fitness cost of spine development The induction of longer spines in B. calyciflorus by low temperature, low food availability, and especially Asplanchna suggests that it increases fitness only during those conditions and is otherwise disadvantageous (Gilbert 1980, 2013). Therefore, induced spine elongation likely involves some cost and occurs only when the fitness benefit exceeds this cost. There could be several types of cost (Gilbert 2013). Evolution of the potential to respond to the Asplanchna kairomone through genetic, sensory, and developmental mechanisms may involve a cost such that genotypes capable of developing longer spines may be less fit when Asplanchna is absent. This possibility has not been investigated but is consistent with the above-noted seasonal selection by Asplanchna for B. calyciflorus clones that develop increasingly long posterolateral spines. Another type of cost could occur within a clone. Individuals with long, induced spines could have a lower reproductive potential if energy was diverted from reproduction to spine development, or required to compensate for effects of altered hydrodynamics. However, there is no convincing evidence for such an allocation cost in B. calyciflorus (see Gilbert 2013 for a review; Yin et al. 2015b, 2017). For example, in each of three experiments conducted by Gilbert (2012), populations of individuals with or without Asplanchna-induced spines had population growth rates that were not significantly different from one another (figure 11). Figure 11. View largeDownload slide Three experiments showing posterolateral-spine lengths and instantaneous population growth rates of Brachionus calyciflorus cultured with and without Asplanchna girodi. The values are means +1 SE. From Gilbert (2012). Figure 11. View largeDownload slide Three experiments showing posterolateral-spine lengths and instantaneous population growth rates of Brachionus calyciflorus cultured with and without Asplanchna girodi. The values are means +1 SE. From Gilbert (2012). Although long spines in B. calyciflorus may not involve a cost in simple laboratory experiments, they may in certain environments or interactions with other organisms. For example, individuals with long, Asplanchna-induced spines may be more susceptible to other predators. There is some evidence for such a trade-off in Keratella tropica: Individuals with long, copepod-induced spines are protected from predation by this predator but are more susceptible to predation by an aquatic insect, the backswimmer Buenoa fuscipennis (Marinone and Zagarese 1991, Zagarese and Marinone 1992). When pronounced spine development is induced by maternal age in the absence of Asplanchna, any cost associated with longer spines would be unlikely to appreciably reduce the population growth rate. That is because the population is primarily composed of individuals born early in their mothers’ lives and therefore with short spines. Yin and colleagues (2015b) reported that the induction of spine development in B. calyciflorus by Asplanchna at a low food concentration was associated with a decrease in the level of sexual reproduction, as was indicated by a decrease in the proportion of mictic females and the number of resting eggs produced. They suggested that this effect is a defense cost. However, sexual reproduction generally is considered to be more costly than parthenogenetic reproduction (Williams 1975, Maynard-Smith 1978). The potential for population growth is reduced because of the production of males, and females pass only one half of their genes to their offspring (meiosis cost). A factor that decreases the proportion of mictic females in a B. calyciflorus population should increase population growth via female parthenogenesis and thus possibly the production of more resting eggs by a larger population in the future (Gilbert 2002, Serra et al. 2004, 2005). Conclusions Spine lengths in B. calyciflorus are highly variable, with posterolateral spines absent or almost as long as the body. This variation is controlled by several environmental factors (temperature, food concentration, and a kairomone from the predator Asplanchna), two endogenous factors (development of stem females from fertilized resting eggs and maternal age), and the genotype. The individual and combined effects of these factors should explain spine-length distributions occurring in natural populations in different ecosystems and at different times within ecosystems. The pronounced spine-development response to Asplanchna is exhibited by all tested strains of B. calyciflorus, clearly occurs in natural ecosystems, and can provide an effective morphological defense against Asplanchna predation. The less dramatic responses to temperature and food concentration demonstrated in laboratory experiments likely contribute to spine-length variation in natural systems and may increase fitness by decreasing sinking rate. There is genetic variation in spine length among and within populations, both for the length of constitutive spines and also for the magnitude of spine development induced by environmental factors and maternal age. Clonal diversity within populations should be high because of extensive hatching of resting eggs (Gilbert 2017b) and suggests the possibility of rapid microevolution caused by strong clonal selection. For example, the seasonal increase in posterolateral spine length in one pond could be explained by an increase in both the abundance of Asplanchna and the frequency of clones able to develop the longest spines. The plasticity of spine length in B. calyciflorus suggests fitness trade-offs associated with spine development, with longer spines being beneficial only during inducing conditions. Pronounced spine development induced by Asplanchna does not reduce population growth rate in simple laboratory environments but may prove to reduce fitness in some conditions or biological interactions in natural systems. The well-characterized morphological effect of the Asplanchna kairomone on spine development in B. calyciflorus is an unusually clear case of the transduction of an environmental signal (either directly or through maternal influence) into the major change of an easily measurable offspring trait. As such, this system presents an interesting future substrate for developmental genetic analysis using modern molecular and bioinformatic tool sets. Acknowledgments I thank Evangelia Michaloudi for permission to cite unpublished observations, Carey Nadell and the two anonymous reviewers for improving the manuscript, Chris Maute and Reed Detar for their assistance with the preparation of figures, and Dartmouth College for support. References cited Aránguiz-Acuña A, Ramos-Jiliberto R, Sarma N, Sarma SSS, Bustamante RO, Toledo V. 2010. Benefits, costs and reactivity of inducible defences: An experimental test with rotifers. Freshwater Biology 55: 2114– 2122. Google Scholar CrossRef Search ADS Bertani I, Leonardi S, Rossetti G. 2013. Antipredator-induced trait changes in Brachionus and prey selectivity by Asplanchna in a large river under low-discharge conditions: Evidence from a field study. Hydrobiologia 702: 227– 239. Google Scholar CrossRef Search ADS Conde-Porcuna JM, Morales-Baquera R, Cruz-Pizarro L. 1993. Effectiveness of the caudal spine as a defense mechanism in Keratella cochlearis. Hydrobiologia 255–256: 283– 287. Google Scholar CrossRef Search ADS De Beauchamp P. 1952. Un facteur de la variabilité chez les Rotifères du genre Brachionus. Comptes Rendus de l’Académie des Sciences 234: 573– 575. Dieffenbach H, Sachse R. 1911. Biologisches Untersuchungen an Rädertieren in Teichgewassern. Internationale Revue der Gesamten Hydrobiologie und Hydrographie, Biol. Suppl. 3: 1– 93. Gama-Flores JL, Huidobro-Salas ME, Sarma SSS, Nandini S. 2011. Effects of predator (Asplanchna) type and density on morphometric responses of Brachionus calyciflorus (Rotifera). Allelopathy Journal 27: 289– 300. Gilbert JJ. 1966. Rotifer ecology and embryological induction. Science 151: 1234– 1237. Google Scholar CrossRef Search ADS PubMed Gilbert JJ. 1967. Asplanchna and postero-lateral spine production in Brachionus calyciflorus. Archiv für Hydrobiologie 64: 1– 62. Gilbert JJ. 1977. Effect of the non-tocopherol component of the diet on polymorphism, sexuality, biomass, and reproductive rate of the rotifer Asplanchna sieboldi. Archiv für Hydrobiologie 80: 375– 397. Gilbert JJ. 1980. Further observations on developmental polymorphism and its evolution in the rotifer Brachionus calyciflorus. Freshwater Biology 10: 281– 294. Google Scholar CrossRef Search ADS Gilbert JJ. 1999. Kairomone-induced morphological defenses in rotifers. Pages 127– 141 in Tollrian R, Harvell CD, eds. The Ecology and Evolution of Inducible Defenses . Princeton University Press. Gilbert JJ. 2002. Endogenous regulation of environmentally-induced sexuality in a rotifer: A multigenerational parental effect induced by fertilization. Freshwater Biology 47: 1633– 1641. Google Scholar CrossRef Search ADS Gilbert JJ. 2003. Environmental and endogenous control of sexuality in a rotifer life cycle: Developmental and population biology. Evolution and Development 5: 19– 24. Google Scholar CrossRef Search ADS PubMed Gilbert JJ. 2007. Timing of diapause in monogonont rotifers. Pages 11– 27 in Alekseev V, Gilbert JJ, De Stasio B, eds. Diapause in Aquatic Invertebrates: Theory and Human Use . Springer. Gilbert JJ. 2009. Predator-specific inducible defenses in the rotifer Keratella tropica. Freshwater Biology 54: 1933– 1946. Google Scholar CrossRef Search ADS Gilbert JJ. 2011a. Induction of different defences by two enemies in the rotifer Keratella tropica: Response priority and sensitivity to enemy density. Freshwater Biology 56: 926– 938. Google Scholar CrossRef Search ADS Gilbert JJ. 2011b. Temperature, kairomones, and phenotypic plasticity in the rotifer Keratella tropica (Apstein, 1907). Hydrobiologia 678: 179– 190. Google Scholar CrossRef Search ADS Gilbert JJ. 2012. Predator-induced defense in rotifers: Developmental lags for morph transformations, and effect on population growth. Aquatic Ecology 46: 475– 486. Google Scholar CrossRef Search ADS Gilbert JJ. 2013. The cost of predator-induced morphological defense in rotifers: Experimental studies and synthesis. Journal of Plankton Research 35: 461– 472. Google Scholar CrossRef Search ADS Gilbert JJ. 2017a. Non-genetic polymorphisms in rotifers: Environmental and endogenous controls, development, and features for predictable or unpredictable environments. Biological Reviews 92: 964– 992. Google Scholar CrossRef Search ADS Gilbert JJ. 2017b. Resting-egg hatching and early population development in rotifers: A review and a hypothesis for differences between shallow and deep waters. Hydrobiologia 796: 235– 243. Google Scholar CrossRef Search ADS Gilbert JJ. 2017c. Spine development in two taxa of Brachionus calyciflorus from Lake Littra, Australia: Constitutive and induced defenses against Asplanchna. Journal of Plankton Research 39: 962– 971. Google Scholar CrossRef Search ADS Gilbert JJ, McPeek MA. 2013. Maternal age and spine development in a rotifer: Ecological implications and evolution. Ecology 94: 2166– 2172. Google Scholar CrossRef Search ADS PubMed Gilbert JJ, Schröder T. 2004. Rotifers from diapausing, fertilized eggs: Unique features and emergence. Limnology and Oceanography 49: 1341– 1354. Google Scholar CrossRef Search ADS Gilbert JJ, Stemberger RS. 1985. Prey capture in the rotifer Asplanchna girodi. Verhandlungen Internationale Vereinigung für theoretische und angewandte Limnologie 22: 2997– 3000. Gilbert JJ, Waage JK. 1967. Asplanchna, Asplanchna-substance, and posterolateral spine length variation of the rotifer Brachionus calyciflorus in a natural environment. Ecology 48: 1027– 1031. Google Scholar CrossRef Search ADS Gilbert JJ, Walsh EJ. 2005. Brachionus calyciflorus is a species complex: Mating behavior and genetic differentiation among four geographically isolated strains. Hydrobiologia 546: 257– 265. Google Scholar CrossRef Search ADS Gosse PH. 1889. B. pala, Ehrenberg. Pages 117– 118 in Hudson CT, Gosse PH. The Rotifera or Wheel-Aninalcules, Volume II, Family XVIII . Brachionidæ. Longmans, Green. Green J. 2005. Morphological variation of Keratella cochlearis (Gosse) in a backwater of the River Thames. Hydrobiologia 546: 189– 196. Google Scholar CrossRef Search ADS Green J, Lan OB. 1974. Asplanchna and the spines of Brachionus calyciflorus in two Javanese sewage ponds. Freshwater Biology 4: 223– 226. Google Scholar CrossRef Search ADS Halbach U. 1970. Die Ursachen der Temporalvariation von Brachionus calyciflorus Pallas (Rotatoria). Oecologia 4: 262– 318. Google Scholar CrossRef Search ADS PubMed Halbach U. 1971. Zum Adaptivwert der zyklomorphen Dornenbildung von Brachionus calyciflorus Pallas (Rotatoria). I. Räuber-Beute-Beziehung in Kurzzeit-Versuchen. Oecologia 6: 267– 288. Google Scholar CrossRef Search ADS PubMed Halbach U, Jacobs J. 1971. Seasonal selection as a factor in rotifer cyclomorphosis. Die Naturwissenschaften 57: 326. Google Scholar CrossRef Search ADS Hutchinson GE. 1967. Introduction to Lake Biology and the Limnoplankton. A Treatise on Limnology , vol. 2. Wiley. Iyer N, Rao TR. 1996. Responses of the predatory rotifer Asplanchna intermedia to prey species differing in vulnerability: Laboratory and field studies. Freshwater Biology 36: 521– 533. Google Scholar CrossRef Search ADS Kofoid CA. 1908. Plankton studies V: The plankton of the Illinois River, 1894–1899, Part II: Constituent organisms and their seasonal distribution. Bulletin of the Illinois State Laboratory of Natural History 8: 3– 361. Lindstrom K, Pejler B. 1975. Experimental studies on the seasonal variation of the rotifer Keratella cochlearis (Gosse). Hydrobiologia 46: 191– 197. Google Scholar CrossRef Search ADS Marinone MC, Zagarese HE. 1991. A field and laboratory study on factors affecting polymorphism in the rotifer Keratella tropica. Oecologia 86: 372– 377. Google Scholar CrossRef Search ADS PubMed Maynard Smith J. 1978. The Evolution of Sex . Cambridge University Press. Mills S et al. 2017. Fifteen species in one: Deciphering the Brachionus plicatilis species complex (Rotifera, Monogononta) through DNA taxonomy. Hydrobiologia 796: 39– 58. Google Scholar CrossRef Search ADS Nandini S, Sarma SSS. 1999. Effect of starvation time on the prey capture behaviour, functional response and population growth of Asplanchna sieboldi. Freshwater Biology 42: 121– 130. Google Scholar CrossRef Search ADS Nandini S, Pérez-Chávez R, Sarma SSS. 2003. The effect of prey morphology on feeding behavior and population growth of the predatory rotifer Asplanchna sieboldi: A case study using five species of Brachionus (Rotifera). Freshwater Biology 48: 2131– 2140. Google Scholar CrossRef Search ADS Papakostas S et al. 2016. Integrative taxonomy recognizes evolutionary units despite widespread mitonuclear discordance: Evidence from a rotifer cryptic species complex. Systematic Biology 65: 508– 524. Google Scholar CrossRef Search ADS PubMed Pourriot R. 1973. Rapports entre la température, la taille des adultes, la longueur des œufs et le taux de dévelopment embryonnaire chez Brachionus calyciflorus Pallas (Rotifère). Annales d’Hydrobiologie 4: 103– 115. Santos-Medrano GE, Robles-Vargas D, Hernández-Flores S, Rico-Martínez R. 2017. Life table demography of Asplanchna brightwellii Gosse, 1850 fed with five different prey items. Hydrobiologia 796: 169– 179. Google Scholar CrossRef Search ADS Sarma SSS, Nandini S, Dumont HJ. 1998. Feeding preference and population growth of Asplanchna brightwelli (Rotifera) offered two non-evasive prey rotifers. Hydrobiologia 361: 77– 87. Google Scholar CrossRef Search ADS Sarma SSS, Larios-Jurado PS, Nandini S. 2002. Population growth of Asplanchna sieboldi fed two Brachionus spp. (Rotifera) raised on green alga and baker's yeast. Hydrobiologia 467: 63– 69. Google Scholar CrossRef Search ADS Sarma SSS, Pavón-Meza EL, Nandini S. 2003. Comparative population growth and life table demography of the rotifer Asplanchna girodi at different prey (Brachionus calyciflorus and Brachionus havanaensis) (Rotifera) densities. Hydrobiologia 491: 309– 320. Google Scholar CrossRef Search ADS Sarma SSS, Resendiz RA, Nandini S. 2011. Morphometric and demographic responses of brachionid prey (Brachionus calyciflorus Pallas and Plationus macracanthus (Daday)) in the presence of different densities of the predator Asplanchna brightwellii (Rotifera: Asplanchnidae). Hydrobiologia 662: 179– 187. Google Scholar CrossRef Search ADS Schröder T, Gilbert JJ. 2009. Maternal age and spine development in the rotifer Brachionus calyciflorus: Increase of spine length with birth orders. Freshwater Biology 54: 1054– 1065. Google Scholar CrossRef Search ADS Serra M, Snell TW, King CE. 2004. The timing of sex in cyclically parthenogenetic rotifers. Pages 135– 146 in Moya A, Font E, eds. Evolution: From Molecules to Ecosystems . Oxford University Press. Serra M, Snell TW, Gilbert JJ. 2005. Delayed mixis in rotifers: An adaptive response to the effects of density-dependent sex on population growth. Journal of Plankton Research 27: 37– 45. Google Scholar CrossRef Search ADS Snell TW. 2011. A review of the molecular mechanisms of monogonont rotifer reproduction. Hydrobiologia 662: 89– 97. Google Scholar CrossRef Search ADS Stemberger RS. 1990. Food limitation, spination, and reproduction in Brachionus calyciflorus. Limnology and Oceanography 35: 33– 44. Google Scholar CrossRef Search ADS Tollrian R, Harvell CD, eds. 1999. The Ecology and Evolution of Inducible Defenses . Princeton University Press. Weber E-F. 1898. Faune rotatorienne du basin du Léman. Revue Suisse de Zoologie et Annales du Musée d’Histoire Naturelle de Genève 5: 263– 785. Wesenberg-Lund C. 1930. Contributions to the biology of the Rotifera, Part II: The periodicity and sexual periods. Det Kongelige Danske Videnskabernes Selskabs Skrifter, Naturvidenskabelig og Mathematisk 9: 1– 230. Williams GC. 1975. Sex and Evolution. Princeton University Press. Xiang X-L, Xi Y-L, Wen X-L, Zhang G, Wang J-X, Hu K. 2011a. Genetic differentiation and phylogeographical structure of the Brachionus calyciflorus complex in eastern China. Molecular Ecology 20: 3027– 3044. Google Scholar CrossRef Search ADS Xiang X-L, Xi Y-L, Wen X-L, Zhang G, Wang J-X, Hu K. 2011b. Patterns and process in the genetic differentiation of the Brachionus calyciflorus complex, a passsively dispersing freshwater zooplankton. Molecular Phylogenetics and Evolution 59: 386– 398. Google Scholar CrossRef Search ADS Yin XW, Niu CJ. 2008. Polymorphism in stem females and successive parthenogenetic generations in Brachionus calyciflorus Pallas. Aquatic Ecology 42: 415– 420. Google Scholar CrossRef Search ADS Yin XW, Zhao NX, Wang BH, Li WJ, Zhang ZN. 2015a. Transgenerational and within-generational induction of defensive morphology in Brachionus calyciflorus (Rotifera): Importance of maternal effect. Hydrobiologia 742: 313– 325. Google Scholar CrossRef Search ADS Yin XW, Zhou YC, Li XC, Li WX. 2015b. Reduced investment in sex as a cost of inducible defence in Brachionus calyciflorus (Rotifera). Freshwater Biology 60: 89– 100. Google Scholar CrossRef Search ADS Yin XW, Jin W, Zhou Y, Wang P, Zhao W. 2017. Hidden defensive morphology in rotifers: Benefits, costs, and fitness consequences. Scientific Reports 7 ( art. 4488). Zagarese HE, Marinone MC. 1992. Induction and inhibition of spine development in the rotifer Keratella tropica: Evidence from field observations and laboratory experiments. Freshwater Biology 28: 289– 300. Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the American Institute of Biological Sciences. All rights reserved. For permissions, please e-mail: firstname.lastname@example.org
BioScience – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
It’s easy to organize your research with our built-in tools.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud