Evolutionary Ecology 14: 13±23, 2000. Ó 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Research paper
Ecology of the hybridogenetic Rana esculenta complex: dierential oxygen requirements of tadpoles SANDRINE PLENET*, FREDERIC HERVANT and PIERRE JOLY
Laboratoire HydrosysteÁmes ¯uviaux. U.M.R. CNRS. Universite Lyon 1, Villeurbanne, France (*author for correspondence, fax: 33 04 72 43 11 41; e-mails:
[email protected];
[email protected];
[email protected]) Received 4 June 1999; accepted 23 September 2000 Co-ordinating editor: N. Chr. Stenseth Abstract. Because of intrinsic demographic load induced by hybridogenesis (infertility of homotypic hybrid matings), the maintenance of hybrid lineages supposes that they present better performances (heterosis) than their host species which allows them to coexist on a long-term basis. However, this necessity of high ®tness can be relaxed if a relative niche partitioning occurs between the taxa, each of them diering in their ecological optima. In the waterfrog hybridogenetic complex (Rana esculenta complex), recent studies have revealed that hybrids show intermediate distribution between parental species across a gradient of river in¯uence (that is related to a gradient of oxygen levels), and intermediate performances of their tadpoles with regard to oxygen availability (hypoxia). In investigating oxygen consumption rates, survival time in anoxia, and metabolite contents in the three forms of the complex, the present study con®rms intermediate characteristics of hybrid tadpoles (R. esculenta) when compared to both parental lineages (R. lessonae and R. ridibunda). Whereas R. ridibunda tadpoles were the most sensitive to anoxia, R. lessonae tadpoles were the most tolerant. Because oxygen requirements of the hybrid proved to be intermediate, no heterosis was detected. These results con®rm the hypothesis of the intermediate niche hypothesis to explain the coexistence of R. lessonae and R. esculenta and the success of the hybridogens. Key words: habitat selection, heterosis, hybridogenetic Rana esculenta complex, intermediate niche, oxygen consumption, oxygen tolerance, tadpoles
Introduction Because recombinations are their only source of genetical variation, unisexual lineages are commonly considered as an evolutionary dead end for vertebrates (Muller, 1964; Vrijenhoek, 1989; Milinski, 1994). However, the long-term persistence of unisexual lineages and their potential role in speciation (Vrijenhoek, 1989) sparked interest in the mechanisms of unisexuality that lend insight into evolutionary processes. Among the dierent mechanisms that sustain unisexuality, the origin of hybridogenesis involves hybrid phenotypes. In these hybridogens, one of the parental genomes is excluded from the germinal line prior to meiosis. The other genome is endoduplicated and is the only one that contributes to the production of gametes. Because of the absence of
14 recombination between the two parental genomes, the transmission of genomes is de®ned as hemiclonal (Schultz, 1969). As a rule, hybridogenetic lineages coexist with a sexual host species (one of the original parental species) with which the backcrossing restores phenotypic hybridity (Schultz, 1977; Moore, 1984; Vrijenhoek, 1994). Hybridogenesis is known to occur in insects (Phasmoptera: Mantovani and Scali, 1992), ®sh (Poeciliopsis: Schultz, 1977; Vrijenhoeck, 1994) and frogs (waterfrog complex: Tunner, 1970; Graf and MuÈller, 1979). Several hypotheses have been proposed for explaining the success and maintenance of hemiclonal hybrids. The persistence of unisexual lineages depends on their continued competitive ability to coexist with their sexual host (Moritz et al., 1989). First, because unisexual lineages arise from hybridization processes, a heterotic advantage (heterosis hypothesis; Manwell et al., 1963; White, 1970; Schultz, 1977) has been supposed to explain the success of hybridogens when they coexist with their sexual host. According to this hypothesis, a hybrid lineage may have a broad generalist niche. Heterosis is thus assumed to compensate for the load of producing males with low ®tness (Schultz, 1971, 1977). A second hypothesis supposes that if hybridization results in intermediate characteristics that dier from those of the parental species, niche partitioning may occur. The hybridogenetic taxon might be best suited for intermediate niches in which the parental species are inferior competitors (intermediate niche hypothesis; Moore, 1984; Vrijenhoek, 1989). This last hypothesis supposes that each parental species is adapted to a dierent zone along an ecological gradient. Hybrid individuals would occupy the intermediate habitats across such a gradient. These alternatives were investigated in the water frog Rana esculenta complex that is interesting in several respects. In the Rana esculenta complex, because of the inviability of homospeci®c matings (Berger and Uzzell, 1977) and the low success of hybrid males in heterospeci®c matings (e.g. Blankenhorn, 1977; Abt and Reyer, 1993), the male sex represents a heavy cost that has to be compensated by high female ®tness. This is assumed to be achieved by the higher fecundity of hybrid females (which can be threefold that of R. lessonae females, Berger and Uzzell, 1980) and by better performances of hybrid ospring when facing stressful conditions such as hypoxia, predation, competition, and pond drying (Tunner and Nopp, 1979; Semlitsch and Reyer, 1992a; Semlitsch, 1993a; Gutman et al., 1994; Berger and Rybacki, 1997; Hotz et al., 1999). These results support the heterosis hypothesis to explain the ecological success of the unisexual R. esculenta. In contrast, several studies suggest that habitat utilization varies among the dierent taxa of the hybridization complex (Blankenhorn, 1977; Rybacki and Berger, 1994; Lada et al., 1995), and that, among other factors, oxygen availability may in¯uence this variation (A. Pagano, S. PleÂnet, P. Joly, unpublished observations; PleÂnet et al., 2000).
15 PleÂnet et al. (2000) even found intermediate responses of the hybrids when compared to their parental taxa with respect to environmental oxygen availability. These results suggest that hybrids might be best suited for intermediate habitat conditions, thus supporting the intermediate niche hypothesis. Knowledge of oxygen requirements and the tolerance to variation in oxygen of tadpoles from each taxon of the R. esculenta complex appears crucial for testing between these alternative hypotheses. In this respect, the present paper aims at comparing both oxygen consumption and survival in tadpoles of three genotypes (R. lessonae, designated LL; R. esculenta, designated RL, and R. ridibunda, designated RR). Moreover, as the major storage form of carbohydrates (i.e. glycogen) is assumed to represent the main source of energy sustaining metabolism during hypoxic/anoxic transitions (Smith, 1950; Hochachka et al., 1973; Pasanen and Koskela, 1974; Hochachka, 1980), metabolite levels were also measured to estimate the energetic potentialities of tadpoles of each lineage when facing hypoxic or anoxic environments.
Materials and methods Animals All tadpoles were obtained from isolated clutches in natural populations from the RhoÃne ¯oodplain (Southeast of France). For a given lineage, all the experiments were carried out on tadpoles from several clutches (3 for LL, 4 for RL and 3 for RR). Taxonomic identi®cations were ensured on subsamples of 10 embryos (2 days after hatching) from each clutch by enzyme electrophoresis on the whole body after the digestive tract had been removed (enzymes PGM2, MPI and LDH-B; see Uzzell and Berger, 1975; Pagano et al., 1997). After identi®cation, tadpoles of each genotype were randomized and kept in large tanks ®lled with aerated tap water until they reached developmental stage 25 (Gosner, 1960). Tadpoles necessary for all the experiments were then sampled and reared individually in small boxes ®lled with aerated tap water (normoxic conditions) until they reached developmental stage 28 (Gosner, 1960). They were fed with rabbit food ad libitum. The experiments were conducted in June 1998, in a room with a controlled temperature (20.1 0.05 °C) under LD-10:14 photoperiod. In order to obviate subsequent in¯uence of body size and development on the metabolic rate, all experimental individuals were `a priori' selected to be similar in length, body mass, and developmental stage (Gosner, 1960). Tadpoles used in the measurement of the rate of O2 consumption (VO2) were starved for at least 1 day before the test to minimize production of faeces and associated microbial VO2 (Feder, 1981).
16 Rates of oxygen consumption (VO2) VO2 was assessed using a closed respirometry system. Individuals were introduced into a 300 ml glass metabolic chamber ®lled with normoxic dechlorinated tap water. Oxygen depletions (less than 15%) inside the system were monitored during 10 h with a WTW Oxi 538 oxygen meter/recorder. Control ¯asks (with no tadpole) served for adjusting O2 uptake calculations. Because of the relative inactivity of the tadpoles in the metabolic chambers, rates of O2 consumption were considered as approximating basic requirements (standard metabolic rate). For all taxa, minimum resting rates of VO2 under standardized conditions were measured at the same period of the day to avoid the eects of a possible circadian respiration rhythm. When removed from the metabolic chambers, the tadpoles were measured (total length), staged according to Gosner (1960), and weighed (wet body mass) after blotting water with paper towels. Thirty larvae of each genotype were individually tested following this procedure. Tolerance to anoxia For each genotype, three groups of ten tadpoles of similar body sizes and similar developmental stages were used to assess the eect of anoxia on survival. Each group was introduced into a glass incubation ¯ask (vol. = 1000 ml) ®lled with anoxic water. In each ¯ask, air was then displaced using pure nitrogen (containing 0.1 ppm O2) before ®lling with anoxic water (PO2 = 0.0 0.03 kPa). Anoxic water was generated by bubbling nitrogen for 1 h; water oxygen concentration was checked using an O2 meter (WTW Oxi 538). Each ¯ask containing tadpoles was then sealed for a speci®c period of time (from 50 to 180 min). Survival tests were conducted at 20.2 0.4 °C. Then the animals were removed from the ¯asks and placed in oxygenated water. In order to avoid recording quiescence as death, a 1-day recovery period was allowed before tabulations. Each animal was recorded as physiologically dead when it had not recovered within 24 h. All animals were measured, weighed, and staged as described above. The lethal time for 50% of the population (LT50) in anoxia was estimated by the `Trimmed Spearman±Karber' method (Hamilton et al., 1977). Metabolites assays In order to measure key metabolite contents (glycogen, proteins, and triglycerids), ten tadpoles of each genotype were selected. Each individual was quickly frozen before being lyophilized (VIRTIS lyophilizator, Trivac D4B). Individuals were then weighed, and their digestive tubes removed to eliminate the bolus. Individuals were homogenized and stored in capped vials at )75 °C until
17 ready for key metabolite assays. Glycogen content was measured following Hervant et al. (1995). Total protein and triglycerid contents were extracted according to Elendt (1989) and measured using speci®c test-combinations (Búhringer, Mannheim, Germany). All assays were performed using a recording spectrophotometer (BECKMANN DU-6) at 25 °C. Enzymes, coenzymes and substrates used for enzyme assays were purchased from Búhringer (Mannheim, Germany) and Sigma (St. Louis, USA). Statistical analyses Dierences in mean morphological measures (body size and mass) and mean metabolite contents among genotypes were tested for signi®cance by one-way analysis of variance (ANOVA). Dierences in developmental stage among genotypes were tested by a Kruskal±Wallis test. Data were log-transformed when they lacked homoscedasticity (mass; DagneÂlie, 1986). Dierences in mean rates of VO2 among genotypes were also evaluated by a one-way ANOVA. Schee's test was used for comparisons between pairwise VO2 means. Comparison of LT50 among taxa relied on 95% con®dence interval.
Results Tadpoles from the three genotypes used in VO2 assessments were of similar mass (ANOVA, df = 85; F = 1.99; P = 0.144) and developmental stage (Kruskal±Wallis, df = 86; H = 0.80; P = 0.671) (Table 1). Only R. ridibunda tadpoles diered slightly in body length from the other taxa (Schee's tests, P < 0.05) (Table 1). Tadpoles used in survival experiments were also similar in size and mass among taxa (ANOVA, df = 179; F = 1.31; P = 0.273; F = 2.25; P = 0.108, respectively), with only dierences in developmental stages (R. esculenta were slightly more advanced than the others; Kruskal± Wallis, df = 180; H = 13.38; P = 0.001) (Table 2). The mean individual metabolic rates diered signi®cantly among genotypes (ANOVA; df = 85; F = 19.41; P = 0.0001; Schee's tests were signi®cant Table 1. Oxygen consumption rates of the water frog tadpoles in normoxia
R. ridibunda R. esculenta R. lessonae
n
Stage
Mass (g)
Length (mm)
VO2 (mg O2/h/g wet mass)
27 29 30
27.9 (0.2) 27.9 (0.2) 27.9 (0.2)
0.111 (0.005) 0.104 (0.003) 0.114 (0.004)
2.43 (0.03) 2.23 (0.03) 2.25 (0.03)
0.288 (0.012) 0.245 (0.008) 0.214 (0.004)
Values are in mean (SE).
18 Table 2. Lethal time for 50% of the population of the water frog tadpoles in anoxia (n = 60)
R. ridibunda R. esculenta R. lessonae
Stage
Mass (g)
Length (mm)
LT50 (min)
27.2 (0.1) 27.6 (0.1) 27.1 (0.1)
0.093 (0.002) 0.091 (0.002) 0.089 (0.003)
2.15 (0.02) 2.11 (0.02) 2.15 (0.02)
62.0 (59.7/64.0) 125.5 (120.2/129.3) 147.0 (140.6/152.5)
Values are in mean (SE).
between each ones, P < 0.05) (Table 1). R. ridibunda showed the highest rates of O2 consumption, whereas R. lessonae showed the lowest, and R. esculenta were found to be intermediate. In anoxia, R. ridibunda LT50 was signi®cantly lower than those of the other genotypes (Table 2). R. ridibunda tadpoles never survived after 62 min. Although intermediate, R. esculenta LT50 was closer to that of R. lessonae than that of to R. ridibunda. Under normoxic conditions, glycogen and protein levels diered signi®cantly among genotypes (ANOVA, df = 29; P < 0.001; Table 3). R. ridibunda showed the lowest glycogen and triglycerid contents, and the highest protein contents. Conversely, R. lessonae showed the highest glycogen and triglycerid contents, and the lowest protein content. R. esculenta showed intermediate metabolic characteristics.
Discussion Rates of oxygen consumption and tolerance to anoxia In normoxia, the oxygen consumption rates of the studied tadpoles were higher than those of other Rana species reared and tested under approximately similar conditions (20±25 °C; developmental stages 25±30). Whereas they ranged from 0.214 to 0.288 mg O2/h/g wet mass in our experiments, they were estimated approximately at 0.075, 0.100, and 0.175 mg O2/h/g wet mass in R. castesbiana,
Table 3. Metabolite contents of the water frog tadpoles (n = 10)
R. ridibunda R. esculenta R. lessonae
Glycogen (lmol glycosyl/g dw)*
Triglycerids (lmol/g dw)
Proteins (g/g dw)
153.1 (7.5) 183.0 (5.4) 216.6 (12.3)
7.2 (0.7) 7.9 (0.7) 8.8 (0.5)
319.7 (23.2) 315.0 (14.1) 309.9 (13.9)
* dw = dry weight; Values are in mean (SE).
19 R. berlandieri, and R. pipiens, respectively (Noland and Ultsch, 1981; Feder, 1983; Crowder et al., 1998). Despite the probable in¯uence of slight variation in experimental procedure, these dierences are large enough to consider the tadpoles of the European water frogs as ecologically more oxygen-dependent than the other species studied. An overview of the relationship between oxygen requirement and habitat use would shed a new light on these dierences, as suggested by some former studies (e.g. Noland and Ultsch, 1981). Adaptation to low-oxygen availability involves the ability to reduce metabolite rate and energy requirement thus inducing a conservation of metabolic reserves (Hervant et al., 1998). Therefore, the higher survival times shown by R. esculenta and R. lessonae probably resulted from two mechanisms: (1) a higher storage of fermentable fuels (mainly glycogen) to sustain the anaerobic metabolism for a longer time, and (2) a lower metabolic rate in normoxia (Hochachka et al., 1973; Hochachka, 1980). Because of their higher metabolic rate, lower LT50, and lower glycogen content, R. ridibunda tadpoles proved less able to withstand hypoxic conditions than did R. esculenta and R. lessonae. The higher metabolic rate of R. ridibunda tadpoles combined with the lower oxygen anity of the blood in adult individuals (Nopp and Tunner, 1985; Lutschinger, 1988), the higher critical pressure Pc50 in adults (18 mmHg, Wolvekamp and Lodewijks, 1934; Nopp and Tunner, 1985), the lower survival of adults in hypoxic conditions (Tunner and Nopp, 1979), and the lower performance (growth and development) of tadpoles during oxygen depletion (PleÂnet et al., 2000) lead us to consider this genotype as the less tolerant one to oxygen de®cit. Conversely, the lower metabolic rate of R. lessonae tadpoles, the higher blood O2 capacity in adult individuals (Nopp and Tunner, 1985; Lutschinger, 1988), the lower Pc50 in adults (12 mmHg; Wolvekamp and Lodewijks, 1934; Nopp and Tunner, 1985), and the indierence in growth and development responses of tadpoles during oxygen depletion allow us to consider this genotype as the more resistant one to hypoxia. With respect to these physiological and developmental responses, the hybrid R. esculenta showed intermediate characteristics. In addition, the higher lipid stores (triglycerids) in R. esculenta and R. lessonae demonstrated the tadpoles of these taxa to be better adapted to food depletion than R. ridibunda tadpoles. In ®nding intermediate performances of the hybrid, our results diverge from those of Tunner and Nopp (1979) who showed that R. esculenta froglets resist better to hypoxia than parental species. However, the experimental conditions strongly diered from ours in being conducted on froglets at very low oxygen pressure (0.6 0.06 kPa) and very low temperature (4.1 0.7 °C) so as to simulate hibernation conditions. Nopp and Tunner (1985) also found that blood anity for oxygen among the three taxa of the complex varied according to the length of the acclimation of the animals to hypoxia. This question of the in¯uence of acclimation remains to be investigated in tadpoles.
20 Evolutionary and ecological implications The evolution of the R. esculenta complex and the adaptive advantage of hybrid lineages remain partially resolved because of the lack of quantitative studies on the ecological determinants of waterfrog distributions in natural habitats. Several studies converge in demonstrating higher ®tness of R. esculenta hybrids that is expected to compensate for the inviability of homotypic R. esculenta matings when hybrids coexist with R. lessonae. Such studies dealing with dierential performances among waterfrog genotypes have used food level, density, competition, drying regime, and predation as constraining factors (e.g. Semlitsch and Reyer, 1992a, b; Semlitsch, 1993a±c). It is commonly considered that these factors act as main selective pressures that in¯uence the evolution of tadpole growth and developmental traits (Wilbur and Collins, 1973; Newman, 1988; Denver, 1997). However, we lack knowledge about the variation of such factors among the natural habitats occupied by each genotype of the complex to predict their potential role in the spatial dynamics of the hybridization system. Among the factors that are decisive to explain habitat partitioning among the genotypes of the complex, Pagano et al. (Unpublished observations) identi®ed river in¯uence (¯ooding rate) and oxygen availability. Oxygen availability has also been shown to dierentially in¯uence tadpole growth and development, with hybrid tadpoles showing intermediate responses to constant hypoxia (PleÂnet et al., 2000). The present study con®rms these previous results in highlighting the intermediate oxygen requirements of R. esculenta tadpoles. All these studies converge toward the absence of a heterosis eect with regard to oxygen requirements and rather support the intermediate niche hypothesis (Moore, 1984) to partly explain the coexistence of R. lessonae and R. esculenta. Previous investigators have shown that physiological requirements of tadpoles may be related to breeding site selection (e.g. R. pipiens and Bufo terrestris: Noland and Ultsch, 1981). Because of the intolerance of R. ridibunda tadpoles to hypoxia, ridibunda adults are expected to breed in habitats with higher oxygen levels than R. esculenta or R. lessonae habitats, as evidenced by several studies (Pagano, 1999; Pagano et al., Unpublished observations). However, the lower tolerance of hybrid tadpoles compared to R. lessonae tadpoles suggests con¯icting interests in breeding habitat selection by R. esculenta. R. esculenta that reproduces with R. lessonae, must breed in the habitats occupied by R. lessonae, even in the most hypoxic ones, as observed in ®eld studies (Pagano, 1999; Pagano et al., Unpublished observations). According to our results, this would constrain R. esculenta tadpoles to experience physiological stress which could reduce their performance. A possibility for hybrids to avoid very low oxygen levels may be microhabitat selection by hybrid tadpoles. Another possibility would be for R. esculenta
21 females to select intermediate habitats according to oxygen availability for laying eggs. In demonstrating that hybrid tadpoles perform better in hypoxia than parental ridibunda tadpoles, our results suggest that the ridibunda haplogenome bene®ts from hybridization in extending its ecological range. However, the intermediate performances of the esculenta genotype in hypoxia suggest that the ridibunda genome carries some load concerning the ability of the hybrid to fully invade the ecological niche of R. lessonae. Further investigations in the ®eld or ®eld experiments are needed to provide responses concerning habitat selection by females and performances of R. esculenta and R. lessonae tadpoles in natural habitats diering in oxygen availability. Indeed it has not yet been demonstrated whether oxygen availability really aects the ®tness ratio in R. lessonae±R. esculenta natural populations. On the other hand, recent studies have found that R. esculenta populations may be composed of several ridibunda hemiclones that show differential performances (Semlitsch et al., 1996, 1997). Hence, variation of tadpole performances among hemiclones need to be taken into account in future investigations dealing with any studies of niche partitioning.
Acknowledgements We thank Raymond D. Semlitsch for constructive comments on the manuscript. We also thank A. Pagano for his assistance in the laboratory taxonomic identi®cations. A preliminary English version of the manuscript was revised by E. Pattee.
References Abt, G. and Reyer, H.-U. (1993) Mate choice and ®tness in a hybrid frog: Rana esculenta females prefer Rana lessonae males over their own. Behav. Ecol. Sociobiol. 32, 221±228. Berger, L. and Rybacki, M. (1997) Competition between tadpoles of water frogs of the Rana esculenta complex. Zool. Pol. 42, 141±154. Berger, L. and Uzzell, T. (1977) Vitality and growth of progeny from dierent egg size classes of Rana esculenta L. (Amphibia, Salientia). Zool. Pol. 26, 291±317. Berger, L and Uzzell, T. (1980) The eggs of European Water frogs (Rana esculenta complex) and their hybrids. Folia biologica (Krakow) 28, 3±25. Blankenhorm, H.J. (1977) Reproduction and mating behaviour in Rana lessonae-Rana esculenta mixed populations. In D.H. Taylor and S.J. Guttman (eds) The Reproductive Biology of Amphibians. Plenum Press, New York, pp. 389±410. Crowder, W.C., Nie, M. and Ultsch, R. (1998) Oxygen uptake in bullfrog tadpoles (Rana catesbeiana). J. Exp. Zool. 280, 121±134. DagneÂlie, P. (1986) TheÂories et MeÂthodes Statistiques. Applications Agronomiques. II ± Les meÂthodes de l'infeÂrence statistique. Les Presses agronomiques de Gembloux, Belgique, pp. 109±240.
22 Denver, R.J. (1997) Proximate mechanisms of phenotypic plasticity in amphibian metamorphosis. Amer. Zool. 37, 172±184. Elendt, B.P. (1989) Eects of starvation on growth, reproduction, survival and biochemical composition of Daphnia magna. Arch. Hydrobiol. 116, 415±433. Feder, M.E. (1981) Eect of body size, trophic state, time of day, and experimental stress on O2 consumption of anuran larvae: an experimental assessment and evaluation of the literature. Comp. Biochem. Physiol. 70, 497±508. Feder, M.E. (1983) Response to acute aquatic hypoxia in larvae of the frog Rana berliandieri. J. Exp. Biol. 104, 79±95. Gosner, K.L. (1960) A simpli®ed table for staging anuran embryos and larvae with notes on identi®cation. Herpetologica 16, 183±190. Graf, J.-D. and MuÈller, W.P. (1979) Experimental gynogenesis provides evidence of hybridogenetic reproduction in the Rana esculenta complex. Experientia 35, 1574±1576. Gutman, E., Hotz, H., Semlitsch, R.D., Guex, G.-D., Beerli, P., Berger, L. and Uzzell, T. (1994) Spontaneous heterosis in larval life-history traits of hemiclonal water frog hybrids. Zool. Pol. 39, 527±528. Hamilton, M.A., Russo, R.C. and Thurston, R.V. (1977) Trimmed Spearman-Karber method for estimating median lethal concentrations in toxicity bioassays. Environ. Sci. Technol. 11, 714±719. Hervant, F., Mathieu, J., Garin, D. and FreÂminet, A. (1995) Behavioral, ventilatory and metabolic responses to severe hypoxia and subsequent recovery of the hypogean Niphargus rhenorhodanensis and the epigean Gammarus fossarum (Crustacea: Amphipoda). Physiol. Zool. 68, 223±244. Hervant, F., Mathieu, J. and Messana, G. (1998) Oxygen consumption and ventilation in declining oxygen tension and posthypoxic recovery in epigean and hypogean aquatic crustaceans. J. Crust. Biol. 18, 717±727. Hochachka, P.W. (1980) Living without Oxygen. Closed and Open Systems in Hypoxia Tolerance. Harvard University Press, New York. Hochachka, P.W., Fields, J. and Mustafa, T. (1973) Animal life without oxygen: basic biochemical mechanisms. Am. Zool. 13, 543±555. Hotz, H., Semlitsch, R.D., Gutmann, E., Guex, G.-D. and Beerli, P. (1999) Spontaneous heterosis in larval life-history traits of hemiclonal frog hybrids. Proc. Natl. Acad. Sci. 96, 2171±2176. Lada, G.A., Borkin, L.J. and Vinogradov, A.E. (1995) Distribution, population systems and reproductive behavior of green frogs (hybridogenetic Rana esculenta complex) in the central Chernozem territory of Russia. Russian J. Herpetol. 2, 46±57. Lutschinger, G. (1988) Zur uÈberwinterung der wasserfroÈsche in den donau- und marchauen (OstoÈsterreich). Jb. Feldherp. 1, 153±160. Mantovani, B. and Scali, V. (1992) Hybridogenesis and androgenesis in the stick insect Bacillus rossius-grandii benazzii (Insecta, Phasmatodea). Evolution 46, 783±796. Manwell, C., Baker, C.M.A. and Childers, W. (1963) The genetics of hemoglobin in hybrids-I. A molecular basis for hybrid vigor. Comp. Biochem. Physiol. 10, 103±120. Milinski, M. (1994) Hybridogenetic frogs on an evolutionary dead end road. TREE 9: 62. Moore, W.S. (1984) Evolutionary ecology of unisexual ®shes. In B.J. Turner (ed) Evolutionary Genetics of Fishes. Plenum Press, New York, pp. 329±398. Moritz, C., Brown, W.M., Densmore, L.D., Wright, J.W., Vyas, D., Donnellan, S., Adams, M. and Baverstock, P. (1989) Genetic diversity and the dynamics of hybrid parthenogenesis in Cnemidophorus (Teiidae) and Heteronotia (Gekkonidae). In R.B. Dawley and J.P. Bogart (eds) Evolution and Ecology of Unisexual Vertebrates. Mus. Bull. 466, 87±112. Muller, H.J. (1964) The relation of mutation to mutational advance. Mutat. Res. 1, 2±9. Newman, R.A. (1988) Genetic variation for larval anuran (Scaphiopus couchii) development time in an uncertain environment. Evolution 42, 763±773. Noland, R. and Ultsch, G.R. (1981) The roles of temperature and dissolved oxygen in microhabitat selection by the tadpoles of a frog (Rana pipiens) and a toad (Bufo terrestris). Copeia 3, 645±652. Nopp, H. and Tunner, H. (1985) Zur sauerstobindung des blutes von Rana ridibunda Pall., Rana lessonae Cam. und Rana esculenta L. (Ranidae, Anura) bei normaler und erniedrigter sauerstospannung. Sitzungsb. OÈsterr. Akad. Wiss. 194, 125±130.
23 Pagano, A., Joly, P. and Hotz, H. (1997) Taxon composition and genetic variation of water frogs in the mid-RhoÃne ¯oodplain. C. R. Acad. Sci. Paris 320, 759±766. Pagano, A. (1999) Ecologie du complexe d'hybridation de grenouilles vertes: cas d'une plaine alluviale. PhD, Univ. Lyon 1, France, 128 pp. Pasanen, S. and Koskela, P. (1974) Seasonal and age variation in the metabolism of the common frog Rana temporaria L. in northern Finland. Comp. Biochem. Physiol. 47, 635±654. PleÂnet, S., Pagano, A., Joly, P. and Fouillet, P. (2000) Variation of plastic responses to oxygen availability within the hybridogenetic Rana esculenta complex. J. Evol. Biol. 13, 20±28. Rybacki, M. and Berger, L. (1994) Distribution and ecology of water frogs in Poland. Zool. Pol. 39, 293±303. Schultz, R.J. (1969) Hybridization, unisexuality, and polyploidy in the teleost Poeciliopsis (Poeciliidae) and other vertebrates. Amer. Natur. 103, 606±619. Schultz, R.J. (1971) Special adaptive problems associated with unisexual ®shes. Amer. Zool. 11, 351±360. Schultz, R.J. (1977) Evolution and ecology of unisexual ®shes. Evol. Biol. 10, 277±331. Semlitsch, R.D. and Reyer, H.-U. (1992a) Performance of tadpoles from the hybridogenetic Rana esculenta complex; interactions with pond drying and interspeci®c competition. Evolution 46, 665±676. Semlitsch, R.D. and Reyer, H.-U. (1992b) Modi®cation of anti-predator behaviour in tadpoles by environmental conditioning. J. Anim. Ecol. 61, 353±360. Semlitsch, R.D. (1993a) Asymmetric competition in mixed populations of tadpoles of the hybridogenetic Rana esculenta complex. Evolution 47, 510±519. Semlitsch, R.D. (1993b) Eects of dierent predators on the survival and development of tadpoles from the hybridogenetic Rana esculenta complex. Oikos 67, 40±46. Semlitsch, R.D. (1993c) Adaptive genetic variation in growth and development of tadpoles of the hybridogenetic Rana esculenta complex. Evolution 47, 1805±1818. Semlitsch, R.D., Schmiedehausen, S., Hotz, H. and Beerli, P. (1996) Genetic compatibility between sexual and clonal genomes in local populations of the hybridogenetic Rana esculenta complex. Evol. Ecol. 10, 531±543. Semlitsch, R.D., Hotz, H. and Guex, G.-D. (1997) Competition among tadpoles of coexisting hemiclones of hybridogenetic Rana esculenta: support for the frozen niche variation model. Evolution 51, 1249±1261. Smith, C.L. (1950) Seasonal changes in blood sugar, fat body, live glycogen and gonads in the common frog Rana temporaria. J. Exp. Biol. 26, 412±429. Tunner, H.G. and Nopp, H. (1979) Heterosis in the common European water frog. Naturwissenschaften 66, 268±269. Tunner, H.G. (1970) Das serumeiweibbild einheimischer wasserfosche und der hybridcharakter von Rana esculenta (serum protein patterns of European green frogs and the hybrid character of Rana esculenta). Son. Verhand. Deutsch. Zool. 64, 352±358. Uzzell, T. and Berger, L. (1975) Electrophoretic phenotypes of Rana ridibunda, Rana lessonae, and their hybridogenetic associate, Rana esculenta. Proc. Acad. Nat. Sci. Phila. 127, 13±24. Vrijenhoek, R.C. (1989) Genetic and ecological constraints on the origins and establishment of unisexual invertebrates. In R.B. Dawley and J.P. Bogart (eds) Evolution and Ecology of Unisexual Vertebrates. Mus. Bull. 466, 24±31. Vrijenhoek, R.C. (1994) Unisexual ®sh: model systems for studying ecology and evolution. Ann. Rev. Ecol. Syst. 25, 71±96. Wilbur, H.M. and Collins, J.P. (1973) Ecological aspects of amphibian metamorphosis. Science 182, 1305±1314. Wolvekamp, H.P. and Lodewijks, J.M. (1934) UÈber die sauerstobindung durch hamoglobin vom frosch (Rana esculenta und Rana temporaria). Zeits. Wissens. Biol., Abt. C. Zeits vergl. Physiol. 20, 382±387.
