Ecography 32: 831!839, 2009 doi: 10.1111/j.1600-0587.2009.05352.x # 2009 The Authors. Journal compilation # 2009 Ecography Subject Editor: David M. Green. Accepted 28 January 2009
Sexual patterns of prebreeding energy reserves in the common frog Rana temporaria along a latitudinal gradient K. Ingemar Jo¨nsson, Ga´bor Herczeg, Robert B. O’Hara, Fredrik So¨derman, Arnout F. H. ter Schure, Per Larsson and Juha Merila¨ K. I. Jo¨nsson (
[email protected]), Dept of Mathematics and Science, Aquatic Biology and Chemistry Group, Kristianstad Univ. College, SE-291 88 Kristianstad, Sweden. ! G. Herczeg and J. Merila¨, Ecological Genetics Research Unit, Dept of Bio- and Environmental Sciences, PO Box 65, FI-00014, Univ. of Helsinki, Finland. ! R. B. O’Hara, Dept of Mathematics and Statistics, PO Box 68, FI-00014, Univ. of Helsinki, Finland. ! F. So¨derman, Dept of Ecology and Evolution/Population Biology, Uppsala Univ., Norbyva¨gen 18D, SE-752 36 Uppsala, Sweden. ! A. F. H. ter Schure, Air Quality Health and Risk Assessment, Electric Power Research Inst. (EPRI), 3420 Hillview Avenue, P.O. Box 10412, Palo Alto, CA 94304-1338, USA. ! P. Larsson, Chemical Ecology and Ecotoxicology, Dept of Ecology, Lund Univ., Ecology Building, SE-223 62 Lund, Sweden.
The ability to store energy is an important life history trait for organisms facing long periods without energy income, and in particular for capital breeders such as temperate zone amphibians, which rely on stored energy during reproduction. However, large scale comparative studies of energy stores in populations with different environmental constraints on energy allocation are scarce. We investigated energy storage patterns in spring (after hibernation and before reproduction) in eight common frog Rana temporaria populations exposed to different environmental conditions along a 1600 km latitudinal gradient across Scandinavia (range of annual activity period is 3!7 months). Analyses of lean body weight (eviscerated body mass), weight of fat bodies, liver weight, and liver fat content, showed that 1) post-hibernation/prebreeding energy stores increased with increasing latitude in both sexes, 2) males generally had larger energy reserves than females and 3) the difference in energy stores between sexes decreased towards the north. Larger energy reserves towards the north can serve as a buffer against less predictable and/or less benign weather conditions during the short activity period, and may also represent a risk-averse tactic connected with a more pronounced iteroparous life history. In females, the continuous and overlapping vitellogenic activity in the north may also demand more reserves in early spring. The general sexual difference could be a consequence of the fact that, at the time of our sampling, females had already invested their energy into reproduction in the given year (i.e. their eggs were already ovulated), while the males’ main reproductive activities (e.g. calling, mate searching, sexual competition) occurred later in the season.
Acquisition and expenditure of energy are important factors affecting life history variation. In fact, different life history strategies largely arise by the different ways organisms allocate the available energy among maintenance, growth, defence and reproduction. An important aspect of energy acquisition and expenditure is how these are temporally related. For example, energetic expenditures of reproduction can be supported by simultaneous energy uptake (income breeders) or by using stored energy that was gathered in advance (capital breeders sensu Stearns 1992 and Jo¨nsson 1997; see also Drent and Daan 1980 for the original formulation of the capital and income breeding concepts). Long-term energy storage and continuous energy intake without storage are not mutually exclusive, and a range of different strategies, from pure capital breeding at one end, through mixed strategies, to pure income breeding at the other end is possible. Energy storage might also be energetically costly (Jo¨nsson 1997), although for ectothermic animals the effect could be minimal (Bonnet et al.
1998). The function of energy stores is often to allow organisms to reproduce, or to survive over periods when feeding is constrained for one reason or another (e.g. absence of food, harsh environmental conditions). This condition often arises in temperate areas, and energy storage therefore plays an important role in the life histories of many organisms. Obviously, populations in different environments, or individuals faced with different energetic expenditures might adaptively differ in their use and pattern of energy storage. Bonnet et al. (1998) suggested that ectothermic animals are preadapted to benefit from long-term energy storage, and concluded that capital breeding is widespread among ectothermic vertebrates. This applies in particular to temperate zone anurans, most of which reproduce immediately after hibernation (Wells 1977), and thus rely on energy stores for both reproduction and survival during hibernation. However, the classification into capital breeding is somewhat complicated by the fact that the main 831
vitellogenic growth in cold temperate female anurans occurs in the summer period, and the eggs are then stored over winter until spawning in the next spring (Jørgensen 1992). This fact led Lardner and Loman (2003) to classify females of the widely distributed Rana temporaria (Linnaeus) as income breeders. However, follicular maturation followed by ovulation takes place during the hibernation period, thus relying on stored energy, and female anurans therefore use a mixed capital-income breeding tactic. Considering the wide altitudinal and latitudinal distributions of temperate zone anurans (Gasc et al. 1997), it seems obvious that the environmental constraints and challenges encountered by different species and populations within species are highly variable. This may strongly influence the needs and benefits of energy storage. Energy storage represents a general tactic under stochastic energy supply or predicted energy shortages (McNamara and Houston 1990, Jo¨ nsson 1997). For anurans, longer hibernation periods require more accumulated energy to survive the winter. Also, a higher risk of unfavourable weather conditions during spring at high altitudes or latitudes would favour larger residual energy stores at post-hibernation emergence. Furthermore, sexual differences in energy storage before reproduction might be expected. As noted above, a main part of female egg production takes place prior to hibernation, while male gametes are formed in connection with spawning (Jørgensen 1992). This may predict larger pre-breeding stores in males. Males also spend a longer time at breeding activities, and engage in more energy demanding activities during the breeding period than females do (Wells 1977). Hence, males might need more energy stores than females for use during the breeding time proper. This prediction may hold not only for amphibians, but also for many other ectothermic animals. Despite the importance of energy stores in the life histories of frogs and the expected variation among populations, relatively few studies on variation in the patterns of energy storage have been conducted (but see Elmberg 1991a, Elmberg and Lundberg 1991). Energy can be stored by amphibians as lipids, proteins and carbohydrates. Of these, lipids are perhaps the most important energy source (Fitzpatrick 1976). Lipids are primarily stored in the carcass, in the abdominal fat bodies and in the liver (Fitzpatrick 1976). Seasonal changes in energy stores have been reported from several R. temporaria populations (Smith 1950, Krawczyk 1971, Pasanen and Koskela 1974, Elmberg 1991a). Fat body size peaks in the autumn before hibernation, and declines over winter to its lowest values in spring or summer (Elmberg 1991a). For a northern Finnish population, Pasanen and Koskela (1974) reported the highest levels of liver fat in the autumn, followed by a decline over the wintering period. This pattern seems to represent the general energy storage dynamics found in amphibians in seasonal environments (Pinder et al. 1992). The aim of this study was to analyse the patterns of energy storage in breeding common frog R. temporaria populations along a latitudinal gradient from southern to northern Scandinavia. Rana temporaria is an excellent candidate for such a study, representing one of the most widespread amphibian species in the world (Gasc et al. 1997), and inhabiting many different habitats in several 832
biomes. The gradient sampled in the present study represents a shift from approximately seven to three months annual activity period (Elmberg and Lundberg 1991, Laugen et al. 2003), and thus strikingly different environmental constraints on energy storage and investments might be expected. Further, R. temporaria is an explosive breeder sensu Wells (1977), and males stay in the breeding area without feeding (J. Elmberg pers. comm.) longer than females, while they are forming a chorus and actively search and fight for females (Elmberg 1990). Females mature their eggs during hibernation (Elmberg 1991a), and come to the breeding pond only for spawning, without much possibility of mate choice (Elmberg 1987, 1991b). Here, we applied a multi-level approach by comparing 1) eviscerated body mass, 2) fat body, 3) liver mass, and 4) liver fat content relative to body length between sexes and populations along a Scandinavian latitudinal gradient. Based on the conditions discussed above, we predicted that 1) energy stores should increase towards the north and 2) males should have larger energy stores than females prior to breeding.
Methods Sampling and general measurements Adult Rana temporaria were collected in 1998 and 1999 from eight populations (one or a few ponds in close vicinity) along a ca 1600 km south-to-north gradient on the Scandinavian Peninsula (Fig. 1; latitudinal range: 55840?N! 69804?N; altitudal range: 5!485 m a.s.l.). The frogs were caught at breeding sites during the early phase of the spawning season (April!June depending on locality), just
Figure 1. Map of the locations of the Rana temporaria study populations.
0.025, 0.012/16 0.028, 0.013/14 0.011, 0.0097/13 0.028, 0.020/5 0.052, 0.014/9 0.051, 0.030/4 0.023, 0.0099/22 0.023, 0.0067/22 0.023, 0.011/17 0.017, 0.0049/13 0.016, 0.0043/23 0.015, 0.0027/13 0.026, 0.0064/15 0.029, 0.0088/16 0.028, 0.0091/29 0.027, 0.0090/20 144 107 0.028, 0.059/56 0.0073, 0.0018/33 0.012, 0.016/44 0.0036, 0.0068/27 ! ! 0.039, 0.086/43 0.0026, 0.0073/12 0.022, 0.030/35 0.013, 0.025/18 0.076, 0.075/29 0.030, 0.047/13 0.034, 0.027/46 0.037, 0.039/43 0.087, 0.069/28 0.075, 0.10/19 281 165 0.48, 0.16/56 0.36, 0.13/33 0.61, 0.25/44 0.53, 0.21/27 0.78, 0.18/9 0.79, 0.11/4 0.84, 0.25/50 0.79, 0.34/22 1.12, 0.39/34 1.15, 0.25/20 1.12, 0.21/29 1.22, 0.26/13 0.71, 0.18/46 0.66, 0.15/43 1.05, 0.28/29 1.02, 0.14/21 297 183 Total N
69.1 Kilpisja¨rvi
485
67.5 Kiruna
425
65.5 Ammarna¨s
410
64.3 Umea˚
5
60.0 Uppland
45
59.9 Va¨rmland
84
56.2 Blekinge
110!120
64.5, 5.6/56 63.0, 6.3/33 65.9, 4.9/44 68.6, 7.7/27 76.1, 5.2/9 79.1, 3.2/4 72.8, 4.9/50 76.8, 6.6/22 75.9, 6.3/35 83.2, 4.9/20 79.7, 2.9/29 82.7, 4.7/13 69.2, 3.1/46 71.7, 4.3/43 72.4, 3.6/29 78.3, 3.8/21 298 183 55.4
22
! " ! " ! " ! " ! " ! " ! " ! " ! "
23.6, 6.5/55 14.6, 5.1/33 28.6, 7.3/44 22.0, 7.2/27 37.9, 8.5/9 30.3, 5.1/4 35.6, 6.3/49 27.7, 8.1/22 37.2, 9.3/35 37.3, 6.2/20 43.1, 6.2/29 39.6, 7.0/13 26.3, 4.5/46 22.2, 3.8/43 30.9, 4.1/29 30.0, 5.3/21 296 183
LIVERFAT (g/g LIVER, SD/N) FATBODY (g, SD/N) LIVER (g, SD/N) LWGT (g, SD/N) SVL (mm, SD/N) Sex
Lund
We used a hierarchical approach (Gelman and Hill 2007) to model the associations between the variables. Our data did not allow for an analysis of between-year variation, since not all populations and sites were sampled in both years, and thus data from the two years were pooled. The response variables (LWGT, FATBODY, LIVER, LIVERFAT) were all log-transformed before the analysis, to better achieve normality of the residuals. Observed values of zero (e.g. individuals that totally lacked fat bodies) were replaced in the transformed data by "5 for the analysis. If we call the vector of responses for the ith individual xi (i.e. LWGT, FATBODY, LIVER, and LIVERFAT, all log transformed)
Altitude (m)
Statistical analyses
Latitude (8N)
After thawing, the frog livers were homogenized individually in a Bligh and Dyer solution with an UltraTurrax mixer, and lipids were analysed according to Bligh and Dyer (1959). The method was modified so that dichloromethane (DCM) was used instead of chloroform. In short, for each gram of liver, 6 ml of Bligh and Dyer solution (DCM/methanol/water in the proportions: v/v 1:2:0.8) was used for homogenising/extraction of lipids. The homogenate was then treated by ultrasound for 15 min and left overnight at room temperature. The homogenate was centrifuged for 15 min at 1450 rpm (Savant) and the supernatant was transferred. The remaining homogenate was washed and mixed with 1 ml of Bligh and Dyer solution, centrifuged, and the supernatant was added to the first aliquote. This was repeated once. For each gram of homogenised liver, 2.10 ml of a DCM and water solution (v/v 1:2) was than added to the extract to obtain a twophase system with DCM/methanol/water in the proportions 2:2:1.8 (v/v). The extract was then centrifuged for 20 min at 1450 rpm (Savant). After the lower DCM phase was transferred, the extract was washed, mixed and centrifuged with 1 ml of DCM, which was also transferred. Finally, the combined DCM phase was evaporated in a vacuum centrifuge (Savant) until dry, after which lipid weight was measured gravimetrically. As liver fat was extracted from liver parts in some cases, we used the relative liver fat (the ratio of extracted liver fat weight and the weight of the liver used for extraction, LIVERFAT) in our analyses.
Population
Extraction of liver fat
Table 1. Mean values (in bold) with standard deviation (SD) and sample size (N) for the analysed variables of Rana temporaria populations. The variables are: snout-vent length (SVL), lean bodyweight (LWGT), weight of fat bodies (FATBODY), liver weight (LIVER) and relative liver fat (LIVERFAT).
after their emergence from hibernation. After capture, the animals were maintained in water at about 68C under laboratory conditions, and were anaesthetized and killed by an overdose of MS-222 (tricaine methane sulfonate) within four days. Before dissection, body length from the tip of the nose to the end of the urostyle (here referred to snout-vent length, SVL) was measured. Liver (LIVER) and fat bodies (FATBODY) were then dissected out and weighed to the nearest milligram, the former being immediately frozen, and kept at "228C until analyzed for fat content. Eviscerated body mass (lean body weight; LWGT) was also measured to the nearest 0.1 g. We note that not all measurements were taken in each individual (Table 1).
833
then we assumed that this vector is multivariately normally distributed: xi !MVN(mi ; V)
(1)
with a vector of expected values mi ; and a covariance matrix V: We then model mi as: mi #as(i) $bs(i) (Lp(i) "L)$fs(i) (Si "S)$gp(i)
(2)
i.e. as a linear function of a population effect (/gp(i) ; where p(i) is a factor representing the population of origin of the ith individual), a sex effect (/as(i) ; where s(i) is the sex of the individual), and of bs(i) and fs(i) ; which are sex-specific effects of population latitude Lp (with mean L) and logtransformed body size (Si, with mean S; we used SVL as a body size proxy), respectively. The mean-centering was done to improve the estimation: gp(i) and as(i) are thus the expected values at the mean latitude and body size. We assume that the population effect, gp(i) ; is also multivariately normally distributed: gp(i) !MVN(0; Vp )
(3)
where 0 is a vector of zeroes and Vp is a covariance matrix. The model was fitted using a Bayesian approach (Gelman et al. 2004), as this allows greater flexibility in the modelling. For this, we need to give prior distributions. These were chosen to be non-informative: V"1 ; V"1 p !Wish (R; 5)
(4)
as(i) ; bs(i) !N(0; 104 )
(5)
where Wish (A, x) is a Wishart distribution with parameters A (a square matrix of the same dimension as V) and x (a scalar). Here R is a diagonal matrix, with 10 "4 on the diagonals. Using x #5 gives a uniform prior for the correlations. N(a, b) denotes a normal distribution with mean a and variance b. Because not all response variables were measured for all individuals, and removing observations with missing data would reduce the power of the analysis, the missing responses were estimated by multiple imputation (Gelman et al. 2004, Nakagawa and Freckleton 2008), in effect they were treated as extra parameters and estimated as such. The uncertainty in their values is therefore incorporated into the analysis, affecting the uncertainty in the parameters of interest. The model was fitted using MCMC in OpenBUGS (Thomas et al. 2006). Five chains were run, a burn-in of 40 000 iterations was removed after which 20 000 iterations were run, and the chains thinned by ten iterations to give a total of 10 000 iterations. Convergence was checked by eye. The results are presented as estimates of parameters and other statistics derived from the model. This is preferable to just presenting p-values which only show whether an estimate is different from zero, and not how important variables are in relation to each other and to total model fit (Cohen 1994, Johnson 1999). Instead, we focused on the parameters and their biological interpretation, as well as on estimating the relative sizes of different sources of variation. The reported posterior densities represent the likely values of the parameters in the light of the data analysed. 834
Results The descriptive statistics for the analysed variables are given in Table 1. Figure 2 shows the expected population means for LWGT, FATBODY, LIVER and LIVERFAT along the latitudinal gradient (all corrected for SVL), and the estimates of the slopes of the effect of latitude are presented in Table 2. FATBODY and LIVER increased towards north in both males and females (Table 2, Fig. 2). There was some evidence that LWGT increased with latitude in females, but decreased with latitude in males (Fig. 2). Although the effect of latitude on LWGT was uncertain, the difference in the slopes was clearly negative, i.e. the slope was smaller in males than in females (Table 2). We did not detect any latitudinal pattern with regard to LIVERFAT (Table 2, Fig. 2), and the slopes had wide confidence intervals. For LIVER, FATBODY and LWGT, males had higher values (Fig. 2), but the sex differences decreased with increasing latitude for LWGT (Fig. 3). Although there was a tendency for a decrease with latitude also for FATBODY and LIVER, lack of effect can not be ruled out (Fig. 3). In contrast, for LIVERFAT females had higher values than males, but the sex difference became less pronounced with increasing latitude, being uncertain in the northernmost populations (Fig. 3). The estimates of the posterior mode and 95% highest posterior density interval of the residual correlation matrix are given in Table 3. The main effects are a positive correlation between LWGT and LIVER, and a smaller negative correlation between LIVER and LIVERFAT. Apart from the size effect, most of the variation in the corrected traits data is due to latitude (Fig. 4), but there is also a large amount of variation between individuals in FATBODY.
Discussion Our results suggest that both the different environmental constraints and the sexual differences in the timing of energy investment for reproduction determine the energy storage patterns in pre-breeding R. temporaria. Comparable large-scale studies are scarce (but see Elmberg 1991a for a mainly altitudinal comparison), and the present study provides new insights into the energy storage patterns of capital breeder amphibians facing different patterns of energy demands and acquisition. We found that both liver weight and fat body weight increased towards the north in both sexes, while lean body weight in females also tended to increase with latitude. Pasanen and Koskela (1974) reported a similar latitudinal increase of liver glycogen content in R. temporaria. Glycogen is known as an important alternative to lipids as an energy source in semi-aquatic anurans (Smith 1950, Pasanen and Koskela 1974). Elmberg (1991a) showed that at a higher altitude (but similar latitude) R. temporaria had a higher rate of fat body growth in summer than frogs at a lower altitude. It seems obvious that frogs facing longer hibernation need larger pre-hibernation energy reserves, but why would the trend be the same just after hibernation? One possible explanation is that reproduction needs more energy in the north than in the south. However,
Figure 2. Effect of latitude on posterior population means of traits. Variables are: lean bodyweight (LWGT, g), weight of fat bodies (FATBODY, g), liver weight (LIVER, g) and relative liver fat (LIVERFAT, g/g LIVER). Points: posterior mode, boxes: 50% highest posterior density intervals and whiskers: 95% highest posterior density intervals. Female effects have been shifted slightly for clarity. The figure shows variables on the natural scale. However, in the statistical analyses all variables were log-transformed.
reproductive investment is lower in northern populations of both males (relative testis mass, Hettyey et al. 2005) and females (relative ovary mass, Miaud et al. 1999) of R. temporaria. Also, despite the generally lower air temperature at high latitudes, thermal conditions in breeding ponds are more favourable (i.e. warmer) in the north, due to higher insolation and more shallow breeding ponds (Olsson and Uller 2003). On the other hand, in northern R. temporaria populations, the subsequent ovarian cycle (i.e. vitellogenesis of next years’ eggs) starts already at the time of spawning (Elmberg 1991a), while more southern European populations have a ‘‘resting period’’ of 1!4 months (Jørgensen 1992). Hence, females in populations without a resting period might need larger energy reserves at the start of the activity period to support the early vitellogenic growth, which has been reported to be sensitive to interruptions by adverse conditions (e.g. starvation; Jørgensen 1992). This difference between southern and northern Scandinavian populations can also be formulated within a capital-income breeding context. While female R. temporaria in southern populations may rely on a more
income breeding tactic to support egg development (Lardner and Loman 2003), northern populations without a resting period must support early egg development by stored energy. Thus, although all females of R. temporaria rely on a mixed tactic of capital-income breeding, the capital component should be larger in northern populations. Male R. temporaria, on the other hand, are more strictly capital breeders, regardless of latitude. This difference between the sexes is consistent with the observed higher levels of energy stores in male R. temporaria compared to females. In male R. temporaria, explanations for increasing energy reserves towards north are harder to understand from the reproductive viewpoint. The operational sex ratio seems to change towards a higher female/male ratio in the north (Hettyey et al. 2005 and references therein, see also Alho et al. 2008 for a detailed study of subarctic populations), which translates to decreased male-male competition (Wells 1977). This would predict a lower energy demand in northern males. Further, the length of the breeding period seems to be unrelated to latitude, or a slight decrease
Table 2. Posterior densities for slopes of latitudinal effect (posterior mode and 95% highest posterior density interval). Variables are: lean body weight (LWGT), weight of fat bodies (FATBODY), liver weight (LIVER) and relative liver fat (LIVERFAT). All variables were logtransformed, and corrected for body size (logSVL) in the analyses. All parameters multiplied by 100. Male LWGT FATBODY LIVER LIVERFAT
"0.72 8.12 2.56 0.83
("1.65, 0.16) (5.07, 11.0) (1.07, 4.00) ("4.53, 6.98)
Female 0.55 ("0.40, 1.43) 9.7 (6.45, 13.4) 2.72 (1.38, 4.46) "0.82 ("6.48, 5.23)
Difference (male-female) "1.29 "1.60 "0.42 1.60
("1.74, ("5.60, ("1.44, ("0.72,
"0.80) 1.92) 0.73) 4.54)
835
Figure 3. Predicted male/female ratios of trait population means (solid line) across latitude and 95% posterior predictive confidence interval (shaded region). Variables are: lean bodyweight (LWGT), weight of fat bodies (FATBODY), liver weight (LIVER) and relative liver fat (LIVERFAT). The figure shows variables on the natural scale. However, in the statistical analyses all variables were logtransformed.
towards north can be detected (Table 6 in Elmberg 1990). On the other hand, in northern Scandinavia (e.g. Kiruna, Kilpisja¨rvi) spawning activities take place around the clock, whereas in southern Scandinavia (e.g. Lund), very little spawning occurs during day time (unpubl.). A possible explanation for larger energy stores at high latitudes could be that northern frogs need permanently high energy reserves as a buffer against more uncertain environmental conditions. Energy stores are often needed to buffer against predicted or unpredicted shortages in energy supply (McNamara and Houston 1990, Jo¨ nsson 1997). Data on variability in spawning time in southern and northern Scandinavia do not suggest that northern frogs face more unpredictable conditions (in terms of hibernation length) than southern frogs (Terhivuo 1988, Loman 2003). However, the risk of encountering ‘‘bad conditions’’, making foraging and energy gathering problematic (e.g. in spring), is higher at high than at low latitudes (Laugen et al. 2003). Further, northern R. temporaria have more pronounced iteroparous life histories, with longer life spans and lower yearly reproductive effort (Miaud et al. 1999, Hettyey
et al. 2005). Hence, northern frogs are expected to be more risk-averse (and therefore have wider energy margins) than southern populations, since their life-time fitness relies more on a longer life span. Such need for permanently high energy reserves in the north might be traded off with other energy demanding life history traits in southern populations. Besides the decreasing yearly reproductive effort and higher age at maturity in the north (Miaud et al. 1999, Hettyey et al. 2005), the S-shaped adult body size distribution of R. temporaria along the same latitudinal gradient used in the present study (Laugen et al. 2005) might be a result of the latitudinal differences in the optimal energy allocation strategy. Such links between latitudinal body size patterns and adaptive life histories have also been discussed in other ectotherms such as the eastern fence lizard Sceloporus undulatus (Angilletta et al. 2004). Similarly, the pattern of energy storage along our latitudinal gradient may not only reflect the energetic need determined by the abiotic (e.g. thermal) environment, but may also be a consequence of an adaptive general life history. In accordance with this, Niewiarowski (2001) reported that
Table 3. Posterior densities for residual correlation matrices (posterior mode and 95% highest posterior density interval). Variables are: lean body weight (LWGT), weight of fat bodies (FATBODY), liver weight (LIVER) and relative liver fat (LIVERFAT). All variables were logtransformed, and corrected for body size (logSVL) in the analyses.
LWGT FATBODY LIVER
836
FATBODY
LIVER
LIVERFAT
0.06 ("0.02, 0.17)
0.33 (0.24, 0.41) "0.03 ( "0.12, 0.07)
"0.13 ("0.25, 0.01) 0.14 ("0.02, 0.27) "0.33 ("0.46, "0.20)
Figure 4. Proportion of variance explained by latitude, population and individual. Points: posterior mode, boxes: 50% highest posterior density intervals and whiskers: 95% highest posterior density intervals. Variables are: lean bodyweight (LWGT), weight of fat bodies (FATBODY), liver weight (LIVER) and relative liver fat (LIVERFAT). All variables were log-transformed in the analyses.
genetically based energy allocation patterns (growth vs storage) differed between S. undulatus lizard populations: individuals from a higher latitude and altitude population invested relatively less in growth and more into storage and reproduction than conspecifics from a lower latitude and altitude population. Schultz and Conover’s (1997) study on Atlantic silversides Menidia menidia revealed a similar pattern; higher latitude fish had genetically based higher rate of lipid accumulation than low latitude fish. Although latitudinal variation in life history traits is well documented in some other ectotherms, such as the brown trout Salmo trutta (Jonsson and L’Abe´e-Lund 1993) and the Eurasian perch Perca fluviatilis (Heibo et al. 2005), records of energy storage are rarely included in the analyses. Several studies have shown that the annual cycles of the different energy stores are associated with the reproductive cycles and the hibernation period in anurans (e.g. Rana esculenta: Schlaghecke and Blu¨ m 1978, Bufo canorus: Morton 1981, Acris crepitans and Bufo woodhousei: Long 1987, Rana ridibunda: Loumbourdis and KyriakopoulouSklavounou 1991, Rana hexadactyla: Das 1996), as well as in many other ectotherms such as lizards (S. undulatus: McKinney and Marion 1985, Liolaemus bitaeniatus: Ramı´res-Pinilla 1995, Tupinambis teguixin: Herrera and Robinson 2000), and fish (Gambusia affinis: Reznick and Braun 1987, M. menidia: Schultz and Conover 1997). These stores are therefore the primary sources for maintenance and reproductive activities (e.g. follicular maturation, spermatogenesis, behavioural activities) during the time when energy intake is absent or minimal. Further, experimental studies in ranid anurans have also supported a
functional link between lipid stores and vitellogenesis (Rana cyanophlyctis: Prasadmurthy and Saidapur 1987, Rana tigrina: Saidapur and Hoque 1996, Girish and Saidapur 2000) and testicular activity (R. hexadactyla: Kasianthan et al. 1978). Considering the difference between sexes in the timing of the main energetic investments into reproduction in most temperate zone anurans, i.e. main female energy investment is prior to the mating period while main male energy investment is during the mating period, one might also expect sexual differences in energy stores after hibernation. In fact, we found considerable sexual differences in relative lean body weight, fat body weight and liver weight, and a less clear difference in liver fat content between male and female R. temporaria. In line with the expectations, males had higher lean body weight, fat body weight and liver weight than females. However, females had higher liver fat content than males in the southern populations. Thus, in general, males were in better body condition and had more energy reserves prior to breeding than females. Sexual differences in fat body size around spawning time have been reported from several other populations of R. temporaria in Europe, with larger fat bodies in males as the general pattern (Jørgensen 1992). We also found, more unexpectedly, that the sexual differences were less pronounced at the northern margin of the species’ distribution. The sexual difference in lean body weight decreased with latitude, but was still present in the northernmost site. Furthermore, fat body weight increased with latitude while the sexual difference tended to decrease, and for liver fat content the difference between sexes disappeared towards north. We are not aware of any previous reports indicating similar changes 837
over latitude in the sexual pattern of energy storage in other amphibians. In summary, we found that despite having a hibernation period more than twice as long, northern R. temporaria had considerably more energy reserves after hibernation ! prior breeding than their southern conspecifics. This can be explained by the continuous and overlapping vitellogenetic cycles of northern females, and possibly as a buffer against more uncertain environmental conditions in the northern populations. Males had more energy reserves than females, probably because females had already invested most of the energy needed for the reproduction in the given year (producing egg clutches), while males were still to face their main energy investments (behavioural activities). The sexual differences became less pronounced towards north. Although the evolutionary background of the latter result is far from understood, it clearly emphasizes the importance of incorporating not only environmental conditions but also reproductive roles in analyses of energy allocation and storage patterns in breeding ectotherms. Acknowledgements ! We are grateful to S. Andersson, J. Elmberg, A. Ja¨rvinen, B. Lardner, J. Loman, G. Sahle´n and R. Tramontano for help with locating suitable populations and collecting frogs. We also thank J. Elmberg and the anonymous reviewers for valuable comments on the manuscript. The study was performed with the permission (C21/98) of the Ethical Committee of Uppsala Univ. Our study was financially supported by The Swedish Research Council (KIJ, JM) and the Academy of Finland (GH, JM, RBO’H).
References Alho, J. S. et al. 2008. Female-biased sex ratios in subarctic common frogs. ! J. Zool. 275: 57!63. Angilletta, M. J., Jr et al. 2004. Bergmann’s clines in ectotherms: illustrating a life history perspective with sceloporine lizards. ! Am. Nat. 164: 169!183. Bligh, E. G. and Dyer, W. J. 1959. A rapid method of total lipid extraction and purification. ! Can. J. Biochem. Physiol. 37: 911!917. Bonnet, X. et al. 1998. Capital versus income breeding: an ectothermic perspective. ! Oikos 83: 333!343. Cohen, J. 1994. The earth is round (p B.05). ! Am. Psychol. 49: 997!1003. Das, I. 1996. Reproductive strategy and lipid mobilization in Rana hexadactyla at a seasonal locality. ! Ethol. Ecol. Evol. 8: 125!133. Drent, R. H. and Daan, S. 1980. The prudent parent: energetic adjustments in avian breeding. ! Ardea 68: 225!252. Elmberg, J. 1987. Random mating in a boreal population of European common frogs Rana temporaria. ! Holarct. Ecol. 10: 193!195. Elmberg, J. 1990. Long-term survival, length of breeding season, and operational sex ratio in a boreal population of common frogs, Rana temporaria L. ! Can. J. Zool. 68: 121!127. Elmberg, J. 1991a. Ovarian cyclicity and fecundity in boreal common frogs Rana temporaria L. along a climatic gradient. ! Funct. Ecol. 5: 340!350. Elmberg, J. 1991b. Factors affecting male yearly mating success in the common frog, Rana temporaria. ! Behav. Ecol. Sociobiol. 28: 125!131. Elmberg, J. and Lundberg, P. 1991. Intraspecific variation in calling, time allocation and energy reserves in breeding male
838
common frogs Rana temporaria. ! Ann. Zool. Fenn. 28: 23!29. Fitzpatrick, L. C. 1976. Life history patterns of storage and utilization of lipids for energy in amphibians. ! Am. Zool. 16: 725!732. Gasc, J.-P. et al. 1997. Atlas of amphibians and reptiles in Europe. ! Societas Europea Herpetologica and Museum National d’Histoire Naturelle (IEGB/SPN), Paris. Gelman, A. and Hill, J. 2007. Data analysis using regression and multilevel/hierarchical models. ! Cambridge Univ. Press. Gelman, A. et al. 2004. Bayesian data analysis, 2nd ed. ! Chapman and Hall. Girish, S. and Saidapur, S. K. 2000. Interrelationship between food availability, fat body, and ovarian cycles in the frog, Rana tigrina, with a discussion on the role of fat body in anuran reproduction. ! J. Exp. Zool. 286: 487!493. Heibo, E. et al. 2005. Latitudinal variation in life-history traits in Eurasian perch. ! Ecology 86: 3377!3386. Herrera, E. A. and Robinson, M. D. 2000. Reproductive and fat body cycles of the Tegu lizard, Tupinambis teguixin, in the Llanos of Venezuela. ! J. Herpetol. 34: 598!601. Hettyey, A. et al. 2005. Does testis weight decline towards the Subarctic? A case study on the common frog, Rana temporaria. ! Naturwissenschaften 92: 188!192. Johnson, D. H. 1999. The insignificance of statistical testing. ! J. Wildl. Manage. 63: 763!772. Jonsson, B. and L’Abe´e-Lund, J. H. 1993. Latitudinal clines in life-history variables of anadromous brown trout in Europe. ! J. Fish Biol 43 (Suppl. A): 1!16. Jo¨ nsson, K. I. 1997. Capital and income breeding as alternative tactics of resource use in reproduction. ! Oikos 78: 57!66. Jørgensen, C. B. 1992. Growth and reproduction. ! In: Feder, M. E. and Burggren, W. W. (eds), Environmental physiology of the amphibians. Univ. of Chicago Press, pp. 439!466. Kasianthan, S. et al. 1978. Fat bodies and spermatogenesis in south Indian green frog Rana hexadactyla Lesson. ! Boll. Zool. 45: 15!22. Krawczyk, S. 1971. Changes in the lipid and water content in some organs of the common frog (Rana temporaria L.) in the annual cycle. ! Acta Biol. Cracov. Ser. Zool. 14: 211!237. Lardner, B. and Loman, J. 2003. Growth or reproduction? Resource allocation by female frogs Rana temporaria. ! Oecologia 137: 541!546. Laugen, A. T. et al. 2003. Latitudinal countergradient variation in the common frog (Rana temporaria) developmental rates ! evidence for local adaptation. ! J. Evol. Biol. 16: 996!1005. Laugen, A. T. et al. 2005. Do common frogs (Rana temporaria) follow Bergman’s rule? ! Evol. Ecol. Res. 7: 717!731. Loman, J. 2003. Inventering av vanlig groda och a˚kergroda i Ska˚ne 2002. Rapport fo¨ r fa¨ltarbete samt ja¨mfo¨ relse med tidigare inventeringar. ! Miljo¨ enheten, Ska˚ne i Utveckling 2003:19, La¨nsstyrelsen i Ska˚ne La¨n, in Swedish. Long, D. R. 1987. A comparison of energy substrates and reproductive patterns of two anurans, Acris crepitans and Bufo woodhousei. ! Comp. Biochem. Physiol. A 87: 81!89. Loumbourdis, N. S. and Kyriakopoulou-Sklavounou, P. 1991. Reproductive and lipid cycles in the male frog Rana ridibunda in northern Greece. ! Comp. Biochem. Physiol. A 99: 577!583. McKinney, R. B. and Marion, K. R. 1985. Reproductive and fat body cycles in the male lizard, Sceloporus undulatus, from Alabama, with comparisons of geographic variation. ! J. Herpetol. 19: 208!217. McNamara, J. M. and Houston, A. I. 1990. The value of fat reserves in terms of avoiding starvation. ! Acta Biotheor. 38: 37!61. Miaud, C. et al. 1999. Variations in life-history traits in the common frog Rana temporaria (Amphibia: Anura): a literature
review and new data from the French Alps. ! J. Zool. 249: 61!73. Morton, M. L. 1981. Seasonal changes in total body lipid and liver weight in the Yosemite toad. ! Copeia 1981: 234!238. Nakagawa, S. and Freckleton, R. P. 2008. Missing in action: the dangers of ignoring missing data. ! Trends Ecol. Evol. 23: 592!596. Niewiarowski, P. H. 2001. Energy budgets, growth rates, and thermal constraints: toward an integrative approach to the study of life-history variation. ! Am. Nat. 157: 421!433. Olsson, M. and Uller, T. 2003. Thermal environment, survival and local adaptation in the common frog, Rana temporaria. ! Evol. Ecol. Res. 5: 431!437. 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. A 47: 635!654. Pinder, A. W. et al. 1992. Estivation and hibernation. ! In: Feder, M. E. and Burggren, W. W. (eds), Environmental physiology of the amphibians. Univ. of Chicago Press, pp. 250!274. Prasadmurthy, Y. S. and Saidapur, S. K. 1987. Role of fat bodies in oocyte growth and recruitment in the frog Rana cyanophlyctis (Schn.). ! J. Exp. Biol. 243: 153!162. Ramı´res-Pinilla, M. P. 1995. Reproductive and fat body cycles of the oviparous lizard Liolaemus bitaeniatus (Sauria: Tropiduridae). ! J. Herpetol. 29: 256!260.
Reznick, D. N. and Braun, B. 1987. Fat cycling in the mosquitofish (Gambusia affinis): fat storage as a reproductive adaptation. ! Oecologia 73: 401!413. Saidapur, S. K. and Hoque, B. 1996. Long-term effects of ovariectomy on abdominal fat body and body masses in the frog Rana tigrina during the recrudescent phase. ! J. Herpetol. 30: 70!73. Schlaghecke, R. and Blu¨ m, V. 1978. Seasonal variations in liver metabolism of the green frog Rana esculenta (L.). ! Experientia 34: 456!458. Schultz, E. T. and Conover, D. O. 1997. Latitudinal differences in somatic energy storage: adaptive responses to seasonality in an estuarine fish (Atherinidae: Menidia menidia). ! Oecologia 109: 516!529. Smith, C. L. 1950. Seasonal changes in blood sugar, fat body, liver glycogen, and gonads in the common frog, Rana temporaria. ! J. Exp. Biol. 26: 412!429. Stearns, S. C. 1992. The evolution of life histories. ! Oxford Univ. Press. Terhivuo, J. 1988. Phenology of spawning for the common frog (Rana temporaria L.) in Finland from 1846 to 1986. ! Ann. Zool. Fenn. 25: 165!175. Thomas, A. et al. 2006. Making BUGS Open. ! R News 6: 12!17. Wells, K. D. 1977. The social behavior of anuran amphibians. ! Anim. Behav. 25: 666!693.
839
