and alternatives proposed. Occasional sexuality in the cyclical parthenogenetic life cycle of monogononts permits application of the biological species concept to ...
Hydrobiologia 186/187: 299-310, 1989. C. Ricci, T. W. Snell and C. E. King (eds), Rotifer Symposium V. © 1989 Kluwer Academic Publishers. Printed in Belgium.
299
Systematics, reproductive isolation and species boundaries in monogonont rotifers Terry W. Snell Division of Science, University of Tampa, Tampa, FL 33606, USA
Key words: mating, reproductive isolation, Rotifera, sexuality, systematics
Abstract The typological concept of rotifer species and the morphological basis of rotifer systematics is reviewed and alternatives proposed. Occasional sexuality in the cyclical parthenogenetic life cycle of monogononts permits application of the biological species concept to this group. Data from cross-mating experiments with Asplanchna, Brachionusand Epiphanesillustrate the usefulness of reproductive isolation as a criterion for species boundaries. Populations from different geographic regions are often interfertile indicating that rotifer species are genetically integrated over wide areas. The main types of isolating mechanisms operating in monogononts are reviewed. The role of behavioral reproductive isolation in maintaining species boundaries is examined. The use of a mate recognition bioassay which estimates the probability of copulation and quantifies the degree of isolation is described. Recent work of the mechanism of mate recognition is reviewed. It is concluded that the biological species concept is applicable to rotifers and that a more experimental approach to determining species boundaries is both feasible and desirable.
Introduction Although the concept of species is essential in systematics and evolution, it lacks a universal definition. The most widely accepted species concept was developed by Dobzhansky and Mayr beginning in the 1930's (see Dobzhansky, 1970 and Mayr, 1970 for reviews). The hallmark of the biological species concept is that species are distinguished by their reproductive isolation from other species. Because of its prominent role, reproductive isolation has been the subject of intense study (Littlejohn, 1981). Despite its widespread acceptance for several years, the biological species concept recently has been vigorously attacked by a number of authors. Hauser (1987) reviews the objections to the biological species concept and examines 15 recently published alter-
native views. He concludes that the biological species concept provides an objective means for separating species and represents the causal factor which produces and maintains discrete evolutionary units. Hauser refutes the major criticisms of the biological species concept which are based on its lack of universality, impracticality and inapplicability to species separated in time. It is concluded that the biological species concept does not need to be changed or rejected on the basis of the prevailing criticisms. Perhaps the strongest stimulus for re-examining the species concept has come from Paterson (1982, 1985). He argues that the prevailing biological species concept is actually an 'isolation concept'. Paterson reasons cogently that speciation occurs as a side-effect of adaptations that promote fertilization and is therefore an evolu-
300 tionary effect rather than a cause. Selection acts to enhance fertilization systems in populations, a consequence of which is speciation. Defined in this way, species are the most inclusive population of biparental organisms which share a common fertilization system. Paterson's 'recognition concept' of species more clearly separates evolutionary causes from effects and does not rely on relational characteristics to define species. The recognition concept of species is gaining wide attention, but it is not without critics (Templeton, 1987). Against this background of controversy, the classification of rotifers proceeds, as it does in all organisms, without a unanimous notion of species. Yet the question of how to best organize rotifer diversity into meaningful taxonomic and evolutionary units persists. There is a basic impression that some populations are similar due to genetic integration, probably resulting from gene flow, stabilizing selection, developmental homeostasis or some combination of these factors. Other populations are decidedly different, probably due to evolutionary divergence in isolation. But how can these ideas be forged into a systematics of rotifers consistent with modem evolutionary theory? As presently conceived, rotifer species are groupings based on morphological similarity of easily measured characters. These are for the most part visually distinguished at the light or electron microscope level. Data are available for only a few species on biochemical traits or on reproductive isolation from cross-mating experiments. Discontinuities among rotifer species are identified by the judgement of systematists, so rotifer taxonomy is based on the classical typological concept of species. Species boundaries occur where phenotypic variation between groups is too large to fit comfortably into one group. In theory, rotifer taxonomy is based on evolutionary systematics, rather than numerical phenetics or cladistics (Mayr, 1981). But in practice it is based almost entirely on morphology. Rotifer systematics is therefore more phenetic than evolutionary or cladistic, but has not utilized the modern numerical taxonomic methods of this
approach. A first attempt to apply cladistic methodology to rotifers appears in this volume (Wallace & Colburn, 1989). For the most part, rotifer genera are morphologically well defined as taxa, but establishing species boundaries has been a very difficult problem for rotifer systematists (Ruttner-Kolisko, 1963). The use of morphology to define rotifer species is confounded by extensive morphological variation, like cyclomorphosis, that is typical of the group (Hutchinson, 1967). In spite of these difficulties with morphology, rotifer systematics has not employed approaches that have been helpful in establishing species boundaries in other animals. Reproductive isolation, perhaps the most natural marker of a species boundary, has seldom been utilized in rotifers. While theoretically attractive, there is some question whether the criterion of reproductive isolation is applicable to rotifers. The problem is rooted in the question asked by Pejler (1977a): is parthenogenesis essentially obligatory in monogonont rotifers? Obligatory parthenogenesis is well known in the class Bdelloidea where males have never been observed and the cytological evidence supports the idea that bdelloids are agamospecies. If most monogononts are also obligatorily parthenogenetic, reproductively definable boundaries between species do not exist and rotifer systematics can only be based on subjective decisions. Whether sexual reproduction occurs occasionally, rarely, or never in monogononts therefore will determine if reproductive isolation is useful as a criterion for rotifer species boundaries. Holman (1987) has compared the recognizability of species in bdelloids and monogononts. My objective in this paper is to explore the feasibility of establishing a more experimental definition of rotifer species. The use of crossmating experiments and mate recognition bioassays as reproductive isolation tests is demonstrated. It is argued that these tools make it possible to define rotifer species boundaries empirically. The small data base from cross-mating experiments between species and among strains within species is reviewed. The most common
301 types of reproductive isolating mechanisms operating in rotifers is characterized and the role of behavioral isolation in defining and maintaining species boundaries discussed. The use of the probability of copulation for quantifying the degree of isolation is described as is recent work on mechanisms of mate recognition.
Do monogononts reproduce sexually? Males have not been observed in many monogonont species, but are they really absent or are sampling programs so inadequate as to make male detection improbable? Long term studies by Carlin (1943) provide some insight into the intermittent nature of male production by Polyarthra and Notholca in the Motala River in Sweden. Male production did not occur in every population in every year. In P. vulgaris, for example, male production did not occur in 1935 and occurred during only one week in 1936, despite the near continuous presence of this species. Similar patterns were observed for P. major and N. caudata. Of the 23 monogonont species closely monitored by Carlin, 52% produced males at some time during the five years of sampling. Although many species eventually produced males, these were only detectable with a long-term sampling program with frequent observations. The difficulties of confirming sexual reproduction in natural rotifer populations were further elucidated by King & Snell (1980) who quantitatively sampled Asplanchna girodi in Golf Course Pond (Florida) daily or bidaily from April through July of 1977. In that study, hundreds of liters were sampled each time, providing one of the most finely detailed views of population dynamics for any natural rotifer population. Males were present (> 0.001 per liter) for 15 days in April, 35 days in May-June and 12 days in July. Male density exceeded 1 per liter for only 13 days in April and 5 days in June. If the sampling program had not been so intensive, it is unlikely that males would have been detected in this population. The presence of males only suggests sexual reproduction. Resting egg (cyst) bearing females also were
detected to verify the completion of sexual reproduction. Cyst bearing female density exceeded 1 per liter for 8 days in April and 5 days in June. These observations clearly illustrate the ephemeral nature of sexual reproduction in natural rotifer populations and the rigorous sampling program required for detection. It is therefore not surprising that males and cyst-bearing females are not often observed in natural rotifer populations. Although sexual reproduction has not been observed in many natural rotifer populations, sex is common in lab populations, especially when rotifers are first isolated from the field. Males and cyst production have been observed for many species of Asplanchna and Brachionus. More species will have to be cultured, however, before it is known whether these genera typify monogononts. At present, it seems safe to conclude that at least many Asplanchna and Brachionus species are not obligate parthenogens and that sexual reproduction is an important feature of their life cycle.
Defining species boundaries with mating experiments The classical approach to defining species boundaries as prescribed by the biological species concept is the mating experiment. Several mating experiments using monogonont rotifers have been completed, verifying the feasibility of this approach with species amenable to laboratory culture. Early attempts to cross different strains of a single rotifer species suggested that strains from distant geographical locations could be mated. Shull (1911) crossed an Epiphanes senta strain from New York with one from Maryland and later (Shull, 1915) crossed a strain from England with one from Lincoln, Nebraska. Also working in the same area of Nebraska, Hertel (1942) successfully crossed four E. senta strains collected from surrounding ponds. Gilbert (1963) attempted to mate three brachionid species, all collected from a small, slightly alkaline pond near New Haven, Con-
302 necticut. Brachionus calyciflorus and B. angularis were planktonic whereas B. quadridentatus was sometimes planktonic, but usually attached to submerged objects. All three species commonly co-occur in alkaline ponds in temperate regions throughout the world. Gilbert found that the sexual periods of B. calyciflorus and B. angularis overlapped, but all attempts to mate these species failed. Interspecific matings with B. quadridentatus also failed indicating that the reproductive isolation of these species is complete. Mating experiments between geographically separatedAsplanchna brightwelli populations were carried out by Birky (1967). Populations collected from Bloomington, Indiana were mated with populations from Paris, Tennessee and produced viable F, F2 and backcross progeny. These populations clearly share a common gene pool, but intra- and interpopulational crosses were not equally fertile. Intrapopulational crosses (self fertilization)
were
77%
successful,
whereas
crosses between Indiana and Tennessee populations had only a 48% success rate, suggesting partial reproductive isolation. Birky also attempted to cross A. brightwelli from Tennessee with A. girodi from the same pond. No interspecific matings were ever observed although both species self fertilized with > 6 0 % success. Asplanchna brightwelli and A. girodi were therefore completely reproductively isolated despite the fact that both species occupied the same region of the pond and both reproduced sexually at the same time. No hybrids were observed verifying the effectiveness of the isolating mechanisms. Mating barriers between rotifer species are not always complete. Ruttner-Kolisko (1969) successfully crossed B. uroeolariswith B. quadridentatus. Males of both species copulated with conspecific as well as heterospecific females, although the probability of copulation was not quantified so male preferences may have gone undetected. However, while pre-mating barriers were absent, post-mating barriers were observed. Crosses between B. quadridentatus females and B. uroeolaris males produced cysts but none hatched. The reciprocal cross, in contrast, was fertile but the small number of mictic females
utilized precluded quantification of post-mating barriers. Differences in the outcome of reciprocal crosses illustrates the asymmetry common in post-mating reproductive isolation (Giddings & Templeton, 1983). The species boundary between B. uroeolaris and B. quadridentatus therefore is not clearly defined. Even though these species appear to be morphologically distinct, further analysis of their reproductive isolation is warranted. Similar experiments have been conducted with A. brightwelli, A. intermedia and A. sieboldi (Gilbert et al., 1979). Crosses between A. brightwelli females and A. intermedia or A. sieboldi males yielded a 39% fertilization rate in 74 attempts indicating little if any pre-mating isolation among these species. However, since none of the cysts hatched, zygote inviability was the likely cause of the complete post-mating isolation. The reciprocal cross with A. intermedia or A. sieboldi females and A. brightwellimales yielded a fertilization rate of 47% in 47 crosses, but again none of the cysts hatched. Crosses between A. intermedia and A. sieboldi indicated no pre-mating isolation, but too few crosses were completed to conclude anything about post-mating barriers. The general conclusions from this study were that A. brightwelli is reproductively isolated from A. intermedia and A. sieboldi by real, quantifiable post-mating barriers. Additional crosses are necessary to determine if A. intermedia and A. sieboldi are likewise completely reproductively isolated. Only a few strains were used in this study so more populations need to be examined to assess the range of intraspecific variation for reproductive isolation. Results of interspecific crosses among rotifer species are summarized in Table 1. Two unpublished works provide additional information on reproductive isolation among geographically separated Asplanchna strains. Birky (unpublished, reported in Birky & Gilbert, 1971) attempted to cross his Tennessee and Indiana strains of A. brightwelli with a German strain which was morphologically identical and physiologically similar. No cysts were produced from any of these crosses. Snell (unpublished, reported in King, 1977) attempted to cross
303 Table 1. A summary of interspecific crosses attempted with monogonont rotifers. Brachionus species: caly - calyciflorus. ang - angularis. quad - quadridentatus.plic -plicatilis, rubens - rubens, urcea - uroeolaris. Asplanchna species: bright - brightwelli, girodi - girodi, inter - intermedia, sieboldi - sieboldi. Cross
Pre-mating barrier
d V caly X ang caly X quad caly X Synchaeta caly X Euchlanis ang X caly plic X Synchaeta plic X rubens urceo X quad quad X urceo
no mating no mating no mating no mating no mating no mating no mating none* none*
bright X girodi bright X inter bright X sieboldi sieboldi X inter
? none* none* none*
Post-mating barrier Brachionus no cyst hatching none? Asplanchna no cysts produced no cysts hatching no cysts hatching ?
Reference
Gilbert 1963 Gilbert 1963 Gilbert 1963 Gilbert 1963 Gilbert 1963 Snell & Hawkinson 1983 Snell unpublished Ruttner-Kolisko 1969 Ruttner-Kolisko 1969 Birky 1967 Gilbert et al. 1979 Gilbert et al. 1979 Gilbert et al. 1979
* Probability of copulation not quantified, subtle differences in male mating preferences could exist.
A. brightwelli from Florida with Birky's strains from Tennessee and Indiana and the German strain. None of these crosses produced cysts, yet all strains were successfully self fertilized. Reproductive isolation among geographically separated Brachionusplicatilisstrains was investigated by Ruttner-Kolisko (1983, 1985). A strain cultured in Ruttner-Kolisko's lab in Austria for 20 years was crossed with strains from Scotland and Colorado. Austria, Scotland and Colorado males did not attempt to mate with Austria females. Austria females apparently have lost the ability to elicit male mating responses after many years of parthenogenetic reproduction in the laboratory. Austria males, in contrast, were able to fertilize Scotland and Colorado females and produce viable cysts. Crosses between Scotland males and Colorado females produced viable cysts, but the reciprocal cross with Scotland females and Colorado males was infertile. Self fertilization in Scotland and Colorado strains was highly successful. Gene flow therefore is possible among all three strains, but all are at least partially reproductively isolated from one another. These experiments illustrate that reproductive isolation is often asymmetrical between strains and reproductive
incompatibilities can be due to either males or females. Results of intraspecific crosses among strains of a single rotifer species are summarized in Table 2. These studies permit several conclusions about reproductive isolation in rotifers. (1) It is possible to identify absolute species boundaries, consistent with the biological species concept, among closely related rotifer species. Cross-mating experiments have accomplished this for B. calyciflorus, B. angularis and B. quadridentatus, A. brightwelli,A. sieboldi and A. intermedia. These taxa are good biological species which are not exchanging genes and are pursuing independent evolutionary courses. (2) Absence of hybrids suggests that reproductive isolation is usually complete even in species with overlapping sexual periods (e.g. B. calyciflorus and B. angularis, A. brightwelli and A. girodi). (3) Populations within a species often are genetically integrated over a wide range and retain the ability to exchange genes. This is demonstrated with A. brightwelli from Indiana and Tennessee and B. plicatilisfrom Austria, Scotland and Colorado. (4) Postmating barriers effectively separate closely related species as in A. brightwelli, A. sieboldi
304 Table 2. A summary ofintraspecific crosses attampted with monogonont rotifers. Asplanchna abbreviations: Ind - Indiana, Tenn - Tennessee, Fla - Florida, 4 temporal pops - all from Lake Thonotosassa, Florida. Brachionus abbreviations: 3 temporal pops - all from McKay Bay, Florida; geographic pops - from different countries worldwide. Cross Epiphanes senta New York X Maryland England X Nebraska 4 Nebraska pops Asplanchna brightwelli Ind X Tenn Ind X Germany Tenn X Germany Fla X Ind Fla X Tenn Fla X Germany 4 temporal pops Brachionusplicatilis Austria X Scotland Austria X Colorado Scotland X Colorado 3 temporal pops 8 geographic pops
Pre-mating barrier
Post-mating barrier
Reference
none* none* none*
none none none
Shull 1911 Shull 1911 Hertl 1942
none* complete complete complete complete complete none*
isolation, isolation, isolation, isolation, isolation,
barrier barrier barrier barrier barrier
none* none* none* none
partial unknown unknown unknown unknown unknown partial
Birky 1967 Birky unpublished Birky unpublished Snell unpublished Snell unpublished Snell unpublished King 1977
partial partial partial
partial
Ruttner-Kolisko 1983 Ruttner-Kolisko 1983 Ruttner-Kolisko 1983 Snell & Hawkinson 1983 Snell & Hawkinson 1983
* probability of copulation not quantified, subtle differences in male mating preferences could exist.
and A. intermedia. (5) Sibling species may be common in monogont rotifers as suggested by the reproductive isolation of Tennessee and Indiana A. brightwelli from morphologically identical strains from Florida and Germany.
Reproductive isolating mechanisms Mating experiments illustrate that many monogonont rotifer species can be defined as reproductively isolated evolutionary units. An important consideration is how reproductive isolation among species arises and how it is maintained. Several examples of reproductive isolating mechanisms in rotifers will provide insight into this question. Reproductive isolating mechanisms are broadly classified as pre-mating or post-mating (Mayr, 1963). Pre-mating barriers include seasonal, habitat, behavioral and mechanical mechanisms; post-mating barriers include gametic incompatability, hybrid inviability and hybrid
sterility. Examples of seasonal reproductive isolation in rotifers are found in Carlin (1943) who quantitatively sampled the Motala River for five consecutive years. Polyarthra vulgaris and P. remata produced males from September through November in each of the five years sampled. Polyarthramajor consistently produced males July through October and P. dolichoptera produced males April through June. Even though females of these species sometimes co-occurred in the plankton, the sexual periods of P. major and P. dolichoptera did not overlap one another or those of P. vulgaris and P. remata. Non-overlapping sexual periods enforce reproductive isolation by preventing males and mictic females from meeting. In some types of environments seasonal reproductive isolation can breakdown. Pejler (1956) investigated the reproductive isolation between P. vulgaris and P. dolichoptera in Lake Luossajarvi in Sweden. Polyarthra vulgaris was eurythermal, preferred oxygen rich surface waters and had a sexual period stretching from summer
305 to early winter. Polyarthra dolichoptera, in contrast, preferred deep, cold water that was oxygen poor and reproduced sexually in early spring. This seasonal and ecological isolation was very effective in maintaining species boundaries, as no hybrids were ever detected in the lake. These species also co-occurred in tarns, small ponds 1-3 meters deep that are abundant in the region. Here, P. vulgaris dominated, but intermediate forms were common. Pejler argued that the environment in the shallow tarns broke down the reproductive isolation between these two species leading to hybridization. This example illustrates the importance of the environment in maintaining species separation. Examples of habitat isolation can be found in various Brachionus species. Ruttner-Kolisko (1974) described the ecological characteristics of several brachionid species in the uroeolarisgroup. Brachionus uroeolaris is found primarily in freshwater where it is mainly benthic, often attaching to submerged vegetation; B. plicatilis is found in brackish waters and soda lakes where it is usually planktonic; and B. rubens is also found in freshwater, often epizooic on Daphnia. The habitat differences of these three species promotes their reproductive isolation. Additional examples of habitat isolation among six species of Brachionus are provided by Miracle et al. (1987). They analyzed the distribution of these species using discriminant analysis of 17 physical and chemical parameters. Separation was achieved on the basis of sulphate, temperature, chloride, alkalinity, conductivity and 02 concentration, suggesting considerable habitat specialization by these brachionids. Studies like this and that of Bogdan & Gilbert (1987) suggest that rotifers finely partition the environment providing ample opportunity for habitat isolation. Behavioral reproductive isolating mechanisms are among the most important in animals (Dobzhansky, 1970). Mate recognition in monogononts and its role in the development of reproductive isolation is an area of active research. This work will be described in detail in the next section. Examples of mechanical isolation have been
reported for rotifers, but there is some doubt about their significance. Ruttner-Kolisko (1983) described a series of mating experiments with B. plicatilis where some interstrain crosses failed. She attributed this failure to differences in lorica thickness of newborn females and the males ability to penetrate the lorica with hyperdermic insertion of the penis. If this accurately represents copulation, it would be an example of mechanical reproductive isolation based on lorica thickness. There is some doubt, however, whether males penetrate the lorica since most copulations occur in the coronal region which is not covered by the lorica (Gilbert, 1963; Snell & Hoff, 1987). Examples of post-mating reproductive isolation in rotifers are not as abundant as premating barriers. There is some evidence of hybrid inviability in crosses among A. brightwelli, A. sieboldi and A. intermedia (Gilbert et al., 1979). King (1977) reported partial post-mating isolation among A. brightwelli populations collected at different times from Lake Thonotosassa, Florida. Four discrete populations were collected in 1974: April, June, September and November. The hatchability of cysts produced from intrapopulational crosses of was much higher than that of interpopulational crosses.
Mate recognition and species boundaries Mate recognition is a crucial behavior with wide ranging evolutionary consequences. Mate recognition systems restrict gene exchange between species and promote divergence among populations. Divergence in mate recognition systems can be an important component of speciation (Paterson, 1982; Thornhill & Alcock, 1983; Giddings & Templeton, 1983; Nevo & Capranica, 1985; Ryan & Wilczynski, 1988). Incorrect recognition leads to gamete wastage which is a serious error considering the small number of gametes produced by male rotifers (Snell & Hoff, 1987). Inviable or reduced fertility hybrids lower reproductive output and overall fitness. As a result, strong selection for precise mate recognition systems is expected.
306 Mate recognition in rotifers is based on coronal contact chemoreception by the male of a chemical on the body surface of females (Gilbert, 1963). Mating begins when a male makes head-on contact with a conspecific female. Chemosensory receptors in the male's coronal region (Clement et al., 1983) sense a glycoprotein present on females (Snell et al., 1988). If a male senses the correct glycoprotein, he begins copulatory behavior by circling the female, skimming over her body while maintaining coronal contact. After several seconds of circling, the male moves toward the female's corona where he then inserts his penis into her pseudocoelom. The entire copulatory sequence takes an average of 1.2 minutes (Snell & Hoff, 1987). Males are smaller, faster swimming than females and actively discriminate conspecific females. Females swim about randomly, taking no apparent role in mate recognition or copulation. Mate recognition behavior of male rotifers can be a useful tool for establishing species boundaries. Males discriminate conspecifics using a precise biochemical mechanism that has evolved over thousands of generations. This recognition system can be used as a simple bioassay to probe the dimensions of a species' gene pool and its boundaries with adjacent species. Mate recognition occurs at the instant of malefemale contact. A conspecific contact usually results in circling of the female by the male. Circling behavior is easily detected as it differs markedly from normal male swimming behavior. A heterospecific contact results in males changing direction and swimming away. Observation of 50-100 encounters provides enough data to calculate a probability of copulation (Pc), the liklihood that males will initiate mating behavior upon contacting a female. If no mating attempts are observed after a suitable number of encounters (Pc = 0), reproductive isolation is assumed to be complete between the populations tested. If P is greater than 0, the species are potentially able to exchange genes, but further tests are necessary to determine if mating is consumated or if postmating barriers exist. Calculation of Pc allows detection of partial reproductive isolation by
quantitatively comparing homogamic and heterogamic pairings (Snell & Hawkinson, 1983). The absence of mating behavior is therefore sufficient evidence for reproductive isolation, but if mating occurs additional experiments are necessary to establish species boundaries. This approach to determining species boundaries is appealing because it relies on rotifer males rather than human systematists to discriminate species. Mate recognition is a natural bioassay for species boundaries and can be carried out quickly with high reliability. The protocol is simple and requires minimal equipment. The application of this bioassay to laboratory populations should not be difficult because male production is common in many cultured species. The mating bioassay will be most useful in sorting out sibling species where morphological criteria are not discriminating. Sibling species are a common problem in cladocera (Hebert, 1987) which are sexually similar to rotifers. Another application will be in cases where cyclomorphosis confuses phenotypic variation. Although, the mate recognition bioassay is not a panacea, it is a useful tool that, when properly applied, can uncover biologically meaningful discontinuities in rotifer diversity. Using mate recognition to discriminate rotifer species is appealing for other reasons. Mate recognition is a central element of the recognition concept of species (Paterson, 1985) and is gaining considerable attention. Specific mate-recognition systems (SMRS) (Paterson 1978, 1980) are involved in signaling between mating partners and are therefore a subset of the fertilization system. In Paterson's view, species boundaries are set by the limits of shared fertilization systems, a primary component of which is the SMRS. Species are genetically cohesive units that are resistent to change (homeostatic) because of stabilizing selection of the SMRS. Speciation occurs as an incidental effect when the SMRS breaks down in allopatry. In the case of rotifers, the limits of the SMRS can be defined experimentally by the mate recognition bioassay. The recognition concept of species makes several predictions (Lambert & Paterson, 1983), many of which are testable with a rotifer model.
307 Mate recognition bioassays have been used to examine species boundaries in several brachionid rotifers. Gilbert (1963) mated Brachionus calyciflorus and B. angularismales with B. calyciflorus, B. angularis,B. quadridentatus, Synchaeta sp. and Euchlanis sp. females. Males of these two species attempted to mate only with conspecific females demonstrating the strict species specificity of the mating reaction. Brachionus plicatilis males are also strictly species-specific in their mating activity when exposed to Synchaeta bicornis females (Snell & Hawkinson, 1983) or B. rubens females (Snell, unpublished). Snell and Hawkinson also examined mating reactions among temporally isolated B. plicatilis populations from the same bay as well as spatially and geographically separated populations. No differences in Pc were detected among the temporally separated populations. However, spatially and geographically separated populations showed mating preferences. Then strains from around the world were characterized for probabilities of copulation. PC values ranged from 78% to 6.2% in all possible pairings. In no case was P = 0, indicating that all strains have retained their mating compatability despite wide geographical separation. This suggests that a globally distributed species like B. plicatilis (Pejler, 1977b) can maintain an integrated gene pool. Post-mating barriers need to be investigated, but an intriguing question is raised. How is mate recognition maintained in geographically isolated populations separated for thousands of generations? Theoretical models of the evolution of reproductive isolation suggest that this is sufficient time for the establishment of reproductive barriers (Nei et al., 1983). Several possible causes for the maintenance of species integrity in the absence of gene flow have been postulated (e.g. Hutchinson, 1968; Van Valen, 1982). Strict species specificity of the mating reaction may not be characteristic of all rotifers. In Asplanchna, Gilbert et al. (1979) report that male A. brightwelli, A. intermedia and A. sieboldi attempt to mate with females of all three species. Only when post-mating barriers were examined, along with six morphological traits, could these
species be discriminated. Asplanchna males also often attempt to copulate with conspecific males, a phenomenon that is not observed in brachionids. This further suggests that the mating reaction in Asplanchna may not be as specific as in brachionids.
Molecular aspects of mate recognition The female's signal and the male's reception are the basis of mate recognition on rotifers. More knowledge of the chemical nature of the signal and how it varies among rotifer species will provide insight into how reproductive isolation develops. Gilbert (1963) showed that mate recognition in rotifers is based on contact chemoreception and provided initial biochemical observations of the mate recognition factor on female B. calyciflorus. Using biochemical techniques considered crude by today's standards, Gilbert concluded that the mate recognition factor is a small, amphoteric molecule less than 4000 daltons with aromatic characteristics. More recently, Snell et al. (1988) examined the biochemical basis of mate recognition in B. plicatilis. They showed that females heated to 100 ° C for 5 minutes lost their ability to elicit a male mating reaction. In contrast, freezekilled females retained their ability to elicit male mating activity at a level indistinguishable from live females. Snell et al. (1988) also showed that the mate recognition factor was destroyed by the proteases proteinase K, pronase E and chymotrypsin and the glycohydrolase fl-amylase. These observations clearly demonstrated that the mate recognition factor is a heat labile glycoprotein and that the carbohydrate portion of the molecule is necessary for mate recognition. The mate recognition glycoprotein (MRG) of B. plicatilisfemales has been further characterized by Snell & Nacionales (1989a). They showed that male mating activity is blocked by exposing females to the lectin concanavalin A (con A) and a lectin isolated from the lentil Lens culinaris. These lectins selectively bind to glycoproteins containing terminal -D-mannosyl, -D-glucosyl or
308 -N-acetylglucosamine residues. Their binding to the MRG suggests the presence of these residues and their blocking of mating activity reinforces the conclusion that the carbohydrate portion of the MRG is important in mate recognition. When males are exposed to these lectins, mate recognition also is blocked, suggesting that lectins bind to the male receptor of the MRG, causing disruption of mate recognition. Using fluoresence microscopy, localization of the MRG on females and its receptor on males has been accomplished (Snell & Nacionales, 1989b). Con A conjugated to the fluorochrome FITC (fluoroisothiocyanate) produced fluoresence at sites where Con A bound to the MRG. On females the MRG was highly localized in the buccal field and corona, dorsal and lateral antennae, pedal sphincter and tip of the foot. The MRG receptor on males is also highly localized in the coronal region, dorsal and lateral antennae, tip of the foot, tip of the penis and the accessory gland. The sites of highest MRG concentration on females correspond to the sites where the most copulations occur (Snell & Hoff 1987). It appears that the MRG guides males to a site on the female's body surface that is favorable for penetration. The goal of research on the biochemical features of mate recognition in rotifers is to better understand the molecular basis of reproductive isolation. Since mate recognition is the cornerstone of reproductive isolation, it is important to know how barriers to cross-mating develop. What kind of molecular changes cause reproductive isolation? Are the mutations responsible simple or complex? This could indicate how difficult it is for behavioral reproductive isolation to evolve. Isolation and characterization of the MRG will provide opportunities to develop biochemically based, rapid screening assays of reproductive isolation. Information like this will make it easier to sort out discontinuities in rotifer diversity and provide better tools for reconstructing rotifer evolution.
Future research In my view, there are several areas that would benefit from a focused research effort. These include the following: 1) There is more information to be extracted from cross-mating experiments. Although these are tedious, they yield a large amount of information. A better understanding of intraspecific variation in reproductive barriers is needed. Do intraspecific barriers to gene flow exist and how do they compare to interspecific barriers? Enough strains within a single species should be tested to characterize these barriers. We need to clearly establish species boundaries for at least one model rotifer species using the criterion of reproductive isolation. Brachionids are good candidates for this. 2) The mate recognition bioassay should be used as a rapid screening technique to examine species boundaries in as many rotifer species as possible. Closely related species in the families Brachionidae (genera like Epiphanes, Brachionus, Euchlanis) or Synchaetidae (Synchaeta, Polyarthra) are good candidates because several species can be cultured, males are common and many species have been described. Several predictions of the recognition concept of species could be tested. 3) More work is needed on mate recognition in rotifers. Are glycoproteins the basis of recognition phenomena in all species? What molecular changes in the carbohydrate moieties lead to reproductive isolation? Is it possible to construct a phylogeny of monogononts using molecular differences in mate recognition glycoproteins? Acknowledgments I thank Mike Childress and Lisa Nacionales for expert technical assistance. Bob Wallace made several suggestions that improved the manuscript. This material is based upon work supported by the National Science Foundation under Grant
309 OCE-8600305 and by the National Institutes of Health under grant ES-04749.
References Birky, C. W., Jr. & J. J. Gilbert, 1971. Parthenogenesis in rotifers: The control of sexual and asexual reproduction. Amer. Zool. 11: 245-266. Birky, C. W., Jr., 1967. Studies on the physiology and genetics of the rotifer Asplanchna. III. Results of outcrossing, selfing and selection. J. exp. Zool. 165: 104-116. Bogdan, K. G. & J. J. Gilbert, 1987. Quantitative comparison of food niches in some freshwater zooplankton. Oecologia 72: 331-340. Carlin, B., 1943. Die Planktonrotatorien des Motalastrom. Medd. Lunds Univ. Limnol. Institut., 258 pp. Clement, P., E. Wurdak & J. Amsellem, 1983. Behavior and ultrastructure of sensory organs in rotifers. Hydrobiologia 104: 89-130. Dobzhansky, T., 1970. Genetics of the Evolutionary Process. Columbia University Press, New York, 505 pp. Giddings, L. V. & A. R. Templeton, 1983. Behavioral phylogenies and the direction of evolution. Science 220: 372-378. Gilbert, J. J., 1963. Contact chemoreceptors, mating behavior and reproductive isolation in the rotifer genus Brachionus. J. Exp. Biol. 40: 625-641. Gilbert, J. J, C. W. Birky, Jr. & E. S. Wurdak, 1979. Taxonomic relationships of Asplanchna brightwelli,A. intermedia and A. sieboldi. Arch. Hydrobiol. 87: 224-242. Hauser, C. L., 1987. The debate about the biological species concept - a review. Z. zool. Syst. Evolut.-forsch. 25: 241-257. Hebert, P. D. N., 1987. Genotypic characteristics of the cladocera. Hydrobiologia 145: 183-193. Hertel, E. W., 1942. Studies on vigor in the rotifer Hydatina senta. Physiol. Zool. 15: 304-324. Holman, E. W., 1987. Recognizability of sexual and asexual species of rotifers. Systematic Zoology 36: 381-386. Hutchinson, G. E., 1967. A Treatise on Limnology. Volume 2. Introduction to Lake Biology and Limnoplankton. John Wiley & Sons, New York, 1115 pp. Hutchinson, G. E., 1981. When are species necessary? In: Population Biology and Evolution, R. C. Lewontin (ed.), Syracuse University Press, Syracuse, NY: 177-186. King, C. E. & T. W. Snell, 1980. Density-dependent sexual reproduction in natural populations of the rotifer Asplanchna girodi. Hydrobiologia 73: 149-152. King, C. E., 1977. Genetics of reproduction, variation, and adaptation in rotifers. Arch. Hydrobiol. Beih. 8: 187-201. Lambert, D. M. & H. E. H. Paterson, 1983. On bridging the gap between race and species: The isolation concept and an alternative. Proc. linn. Soc. N.S.W. 107: 501-514. Littlejohn, M., 1981. Reproductive isolation: a critical
review. In: Evolution and Speciation, W. R. Atchley & D. S. Woodruff (eds.), Cambridge University Press, Cambridge: 298-334. Mayr, E., 1963. Animal Species and Evolution. Belknap Press, Harvard University, Cambridge, MA. Mayr, E., 1970. Populations, Species and Evolution. Belknap Press, Harvard University, Cambridge, MA. Mayr, E. 1981. Biological classification: Toward a synthesis of opposing methodologies. Science 214: 510-516. Miracle, M., M. Serra, E. Vicente & C. Blanco, 1987. Distribution of Brachionus species in Spanish mediterranean wetlands. Hydrobiologia 147: 75-81. Nei, M., T. Maruyama & C. Wu, 1983. Models of evolution of reproductive isolation. Genetics 103: 557-579. Nevo, E. & R. R. Capranica, 1985. Evolutionary origin of ethological isolation in cricket frogs, Acris. Evol. Biol. 19: 147-214. Paterson, H. E. H., 1978. More evidence against speciation by reinforcement. S. African J. Science 74: 369-371. Paterson, H. E. H., 1980. A comment on 'mate recognition systems'. Evolution 34: 330-331. Paterson, H. E. H., 1982. Perspectives on speciation by reinforcement. S. African J. Science 78: 53-57. Paterson, H. E. H., 1985. The recognition concept of species. In: Species and Speciation, E. S. Vrba (ed.), Transvaal Museum Monograph No. 4, Transvaal Museum, Pretoria, Rep. S. Africa: 21-29. Pejler, B., 1956. Introgression in planktonic Rotatoria with some points of view on its causes and conceiveable results. Evolution 10: 246-261. Pejler, B., 1977a. General problems on rotifer taxonomy and global distribution. Arch. Hydrobiol. Beih. 8: 212-220. Pejler, B., 1977b. On the global distribution of the family Brachionidae (Rotatoria). Arch. Hydrobiol. Beih. Supplement 2: 255-306. Ruttner-Kolisko, A., 1963. The interrelationships of the Rotatoria. In: The Lower Metazoa, E. C. Dougherty, (ed.), Univ. Calif. Press, Berkeley, Calif: 263-272. Ruttner-Kolisko, A., 1969. Kreuzungexperimente zwischen Brachionus urceolaris and Brachionus quadridentatus, ein Beitrag zur Fortpflanzungbiologie der heterogonen Rotatoria. Arch. Hydrobiologie 65: 397-412. Ruttner-Kolisko, A., 1974. Plankton Rotifers. Binnengewasser 26 suppl.: 1-146. Ruttner-Kolisko, A., 1983. The significance of mating processes for the genetics and for the formation of resting eggs in monogonont rotifers. Hydrobiologia 104: 181-190. Ruttner-Kolisko, A., 1985. Results of individual cross-mating experiments in three distinct strains of Brachionusplicatilis (Rotatoria). Verh. int. Ver. Limnol. 22: 2979-2982. Ryan, M. J. & W. Wilczynski, 1988. Coevolution of sender and receiver: Effect on local mate preference in cricket frogs. Science 240: 1786-1788. Shull, A. F., 1911. Studies in the life cycle of Hydatinasenta. II. The role of temperature, of chemical composition of the medium, and of internal factors on the ratio of parthenogenetic to sexual forms. J. exp. Zool. 10: 117-166.
310 Shull, A. F., 1915. Inheritance in Hydatinasenta. IV. Characters of females and their parthenogenetic eggs. J. exp. Zool. 18: 145-186. Snell, T. W. & C. A. Hawkinson, 1983. Behavioral reproductive isolation among populations of the rotifer Brachionus plicatilis. Evolution 37: 1294-1305. Snell, T. W. & F. H. Hoff, 1987. Fertilization and male fertility in the rotifer Brachionus plicatilis. Hydrobiologia 147: 329-334. Snell, T. W., M. J. Childress & B. C. Winkler, 1988. Characteristics of the mate recognition factor in the rotifer Brachionusplicatilis.Comp. Biochem. Phys. 89A: 481-485. Snell, T. W. & M. A. Nacionales, 19 8 9a. Sex pheromone communication in Brachious plicatilis (Rotifera). Submitted.
Snell, T. W. & M. A. Nacionales, 1989b. Localization of the mate recognition glycoprotein on the rotifer Brachionus plicatilis. Submitted. Templeton, A. R., 1987. Species and speciation. Evolution 41: 233-235. Thornhill, R. & J. Alcock, 1983. The Evolution of Insect Mating Systems. Harvard University Press, Cambridge, MA. Van Valen, L., 1982. Integration of species: Stasis and biogeography. Evol. Theory 6: 99-112. Wallace, R. L. & R. A. Colburn, 1989. Phylogenetic relationships within the phylum Rotifera: orders and genus Notholca. Hydrobiologia 186/187: 311-318.
