Present address. School of Environmental Sciences, University of East Anglia, Norwich NR4 7TJ, UK.
Interactive effects of Pgi genotype and temperature on larval growth and survival in the Glanville fritillary butterfly
Article first published online: 21 APR 2011
© 2011 The Authors. Functional Ecology © 2011 British Ecological Society
Volume 25, Issue 5, pages 1032–1039, October 2011
How to Cite
Kallioniemi, E. and Hanski, I. (2011), Interactive effects of Pgi genotype and temperature on larval growth and survival in the Glanville fritillary butterfly. Functional Ecology, 25: 1032–1039. doi: 10.1111/j.1365-2435.2011.01854.x
- Issue published online: 22 SEP 2011
- Article first published online: 21 APR 2011
- Received 1 November 2010; accepted 7 March 2011 Handling Editor: Sue Jackson
- larval biology;
- life-history trade-off;
- Melitaea cinxia;
- phosphoglucose isomerase;
- thermal adaptation
1. Genetic polymorphism in the gene phosphoglucose isomerase (Pgi) encoding for a glycolytic enzyme has been shown to affect many traits in adult insects, including flight metabolism, running speed, fecundity and longevity, but it is not known to what extent Pgi genotypic effects are consistent across different life stages.
2. In the Glanville fritillary butterfly, heterozygous AC adult individuals for a single nucleotide polymorphism (SNP) (AA111) in the coding region of Pgi have superior fitness to the common homozygotes (AA) in practically all life-history traits.
3. Here, we studied associations between Pgi SNP AA111 and larval and pupal weights, larval development time in three different temperatures and adult longevity. Small body size and limited mobility of larvae offer little buffer against changes in ambient temperature; hence, temperature is expected to affect greatly larval growth and development.
4. In contrast to adults, larval performance was superior in AA homozygotes in two respects. First, survival was higher in AA homozygotes under stressful conditions, represented by the low-temperature treatment in which survival was generally low. Second, the AA homozygotes had heavier pupae. In spite of the latter result, adult life span was longer in the AC heterozygotes, in support of previous studies.
5. The results on larval growth are consistent with the hypothesis of a trade-off between thermal stability and kinetic efficiency between the different isoforms of the PGI enzyme, but the results also indicate unexpected differences in the genotypic effects at different life stages.
Temperature is a key environmental factor affecting the growth and survival of organisms and the dynamics of their populations (Hazel 1995; Somero 1995; Willmer, Stone & Johnston 2000; Hochachka & Somero 2002; Sinclair et al. 2003) especially for ectothermic animals (Hazel 1995; Willmer, Stone & Johnston 2000; Hochachka & Somero 2002). Functional properties of enzymes depend strongly on temperature (Somero 1995), and many studies have demonstrated that molecular variation in metabolic enzymes and in the genes encoding for them is associated with temperature-dependent organismal performance (Riehle, Bennett & Long 2001, 2005; Podrabsky & Somero 2004). For instance, variation in glycolytic enzymes along latitudinal and altitudinal gradients has been interpreted in this context (Hoffmann 1981; Graves & Somero 1982; Powers et al. 1991; Verrelli & Eanes 2001).
The glycolytic enzyme phosphoglucose isomerase (PGI) is a good candidate for a key enzyme in thermal adaptation (Kingsolver & Watt 1983; Watt 1983). PGI has been found to be highly polymorphic in a wide range of taxa, and this variation is often correlated with variation in individual performance and fitness (Patarnello & Battaglia 1992; Katz & Harrison 1997; Dahlhoff & Rank 2000; Watt 2003). For many taxa, there is evidence for either balancing or directional selection on PGI (Corbin 1977; Watt 1977, 1992; Hoffmann 1981; Zera 1987; Patarnello & Battaglia 1992, Katz & Harrison 1997; Rank & Dahlhoff 2002, Wheat et al. 2010). Studies on Colias butterflies (reviewed by Watt 1992, 2003) and the leaf beetle Chrysomela aeneicollis (Dahlhoff & Rank 2000, Rank et al. 2007) have produced compelling evidence for temperature-dependent selection on PGI. These studies have shown that different isoforms of the PGI enzyme differ in their kinetic efficiency and thermal stability in a manner that is consistent with differences in flight performance and activity (Watt, Cassin & Swan 1983; Watt 1992, 2003) and the observed distribution of PGI genotypes in natural populations (Watt 1992, 2003; Dahlhoff & Rank 2000; Rank et al. 2007). Similarly, in the leaf beetle C. aeneicollis, the PGI allozyme that has the highest catalytic efficiency but is the least thermostable predominates in the coldest study areas, whereas the more thermostable genotype has been found to predominate in the warmer areas (Dahlhoff & Rank 2000; Rank et al. 2007). In the copper butterfly Lycaena tityrus, the PGI allozyme dominating in high-altitude populations was found to have the shortest chill-coma recovery time, suggesting a link between thermal adaptation and PGI (Karl, Schmitt & Fischer 2008).
A series of studies on the Glanville fritillary butterfly (Melitaea cinxia, Fig. 1) has examined associations between life-history traits and PGI allozyme phenotypes or single nucleotide polymorphisms (SNP) in the coding region of the Pgi gene. Orsini et al. (2009); see Materials and methods) have shown that the Pgi SNP AA111 in the Glanville fritillary corresponds closely to the allozyme allele f reported by Haag et al. (2005). Heterozygous (AC) females for SNP AA111 have higher flight metabolic rate (Haag et al. 2005; Niitepõld 2010) and higher dispersal rate in the field (Niitepõld et al. 2009) than the AA homozygotes. Furthermore, heterozygous females lay larger egg clutches (Saastamoinen 2007a), at least partly because these females are able to start flying earlier in the day and have higher body temperature at low ambient temperatures than the AA homozygotes (Saastamoinen & Hanski 2008). SNP AA111 is even associated with the growth rate of natural populations under some environmental conditions (Hanski & Saccheri 2006).
Previous studies on Pgi in the Glanville fritillary and other butterflies (references above) have examined the performance and fitness components of adults, while larval performance has remained uninvestigated (with the exception of the study by Karl, Schmitt & Fischer (2008) on the copper butterfly Lycaena tityrus). It is possible that temperature-related factors play an even greater role in the development of larvae than in the performance of adults because of the more limited dispersal capacity of the former (McMillan et al. 2005). Furthermore, the small size of caterpillars offers little buffer against changes in ambient temperature (Heinrich 1993; Fordyce & Shapiro 2003). Here, we investigate the effect of temperature and Pgi genotype on larval development and survival during the last larval instar in the Glanville fritillary. We also examine the effect of population age on larval performance, contrasting individuals originating from newly established vs. old local populations. The age of the field populations is known based on a long-term survey of the entire metapopulation (Hanski 1999; Nieminen, Siljander & Hanski 2004). Previous studies have documented substantial life-history differences between females from new vs. old populations (Hanski et al. 2004; Haag et al. 2005; Saastamoinen 2007b), apparently because new populations are established by females that tend to be more dispersive than the average female in the entire metapopulation (Hanski et al. 2004; Zheng, Ovaskainen & Hanski 2009).
Materials and methods
Study species and populations
The Glanville fritillary butterfly has a univoltine life cycle in the Åland Islands in Finland (Hanski 1999; Nieminen, Siljander & Hanski 2004). Females lay large clutches of 130–160 eggs, and the larvae develop in groups of full sibs. In the fifth larval instar, larvae spin a communal web in which they diapause over winter. Larvae moult two more times after diapause and pupate after the seventh larval instar in spring (Nieminen, Siljander & Hanski 2004; Saastamoinen 2007a).
Fifth instar larvae were collected from 60 different local populations scattered across the Åland Islands in the spring 2005. The larvae were reared into butterflies that were released into a large population cage for an experiment (Saastamoinen 2007b). Another similar experiment was performed in the following summer using the offspring from the first experiment, and we used the offspring from this latter experiment in the present study. Thus, the present material represents the F2 individuals reared under common garden conditions. Larvae in each family group are full sibs, which is known based on direct observations on matings in the cage (Saastamoinen 2007b). The larvae collected in the field in 2005 originated from either newly established (1 year) or old (≥5 years) local populations. We classify individuals in the present material as coming from newly established vs. old populations based on the population type of their grand-mother; we used only butterflies from new × new and old × old crosses from the 2006 cage experiment. In the present material, both sex ratio (59% males) and population type (45% from newly established populations) were roughly equal.
Rearing of larvae
Prediapause larvae were reared in common garden conditions (12:12 L/D, 25 and 20 °C, respectively). Diapausing larvae were kept under natural conditions out of doors. The diapause was broken at the end of February 2007, after which the larvae were reared under the intermediate-temperature conditions in Table 1 until the last moult. Larvae were fed daily with fresh leaves of the host plant Plantago lanceolata. In the wild, larvae feed gregariously until they have reached the last larval instar (Boggs & Nieminen 2004). To prevent any stress caused by rearing larvae in isolation, larvae from the same family group were kept in the same rearing box until they moulted to the last instar, after which they were reared individually.
|Treatment||Max temperature (°C)||Time (h)||Min temperature (°C)||Time (h)|
The experiment was conducted on the last instar larvae. Larvae from each family were assigned into one of three temperature treatments: low, intermediate or high (Table 1). There were mostly 6–10 larvae per family per treatment. In the low-temperature treatment, temperature alternated between 8 and 20 °C, in the intermediate treatment between 15 and 26 °C, and in the high treatment between 8 and 35 °C (Table 1). The low-temperature treatment (8/20 °C) was selected based on a pilot experiment as a treatment that increased larval mortality, though, still allowing most of the individuals to complete their development. The intermediate-temperature treatment (15/26 °C) represents the standard conditions that we have used for rearing larvae. The high-temperature treatment (8/35 °C) approximates the temperatures that larvae may maximally reach on sunny days in the field when they are able to bask in the sun (Kuussaari 1998). Lights were adjusted to correspond to the temperature, such that maximum illumination was reached simultaneously with maximum temperature, and lights were turned off for 12 h around the minimum temperature. The light conditions were the same in all treatments.
Larvae were weighed after the last moult (=larval weight) and 1 day after pupation (=pupal weight) with an analytical balance (Mettler Toledo Inc., model XS105, Columbus, OH, USA) with an accuracy of 10 μg. Pupae were kept under the intermediate-temperature conditions (above) until adults eclosed and the sex could be determined. After eclosion, butterflies were kept in cylindrical cages in a room in which temperature alternated between 20 and 28 °C, which corresponds to natural conditions in the field. Butterflies were fed with honey water solution. The cages were checked daily for dead butterflies.
Three SNPs targeting three charge-changing amino acids in the coding region of Pgi have been described and validated by Orsini et al. (2009). A combination of two of these three SNPs identifies an allozyme allele that has been correlated with individual performance and fitness in previous studies (Haag et al. 2005; Hanski & Saccheri 2006). In practice, owing to close linkage, one of the SNPs alone explains phenotypic variation as well as the combination (Orsini et al. 2009). We thus genotyped the present material for only one of the SNPs (SNP AA111). DNA extraction and SNP genotyping were carried out following the protocol described by Orsini et al. (2009). The CC homozygotes were rare in this material (four individuals), as has been observed in other materials from the Åland Islands (Orsini et al. 2009), and were hence omitted. Thus, we contrast two Pgi genotypes: the AA homozygotes and the AC heterozygotes in AA111. Genotyping was carried out with MegaBACE 1000 capillary sequencing machine (GE Healthcare, Freiburg, Germany) and called by SNP profiler. There were a few cases which were omitted because the call was not clear enough to reliably determine the genotype. We obtained reliable genotypes for 96% of the individuals, of which 167 were AA homozygotes and 126 were AC heterozygotes.
We used Linear Mixed Models with binomial error distribution (glimmix; SAS Institute Inc., Cary, NC, USA) to study the effect of Pgi genotype (AA vs. AC) on larval survival over the final instar. Population type (new vs. old), larval weight, temperature treatment and all second-order interactions were included as candidate explanatory variables. Family was added as a random factor nested within the population type. In general, we used backward elimination by removing all insignificant (P > 0·05) interactions one by one, starting with interactions whose components did not have significant main effects. Sex could not be included in the model because we were unable to sex the larvae that died.
We used linear mixed effects models (proc mixed; SAS Institute Inc., Cary, NC, USA) to analyse the effects of Pgi genotype, population type and their interaction on larval and pupal weights. In the latter case, additional explanatory variables were the temperature treatment and larval weight and their interactions. These analyses were conducted separately for females and males with family as a random factor. Model selection was performed in the same manner as above for survival. An additional analysis was run for family means of pupal weight, which avoids a difficulty arising from limited genotypic variation within families. Prior to the statistical analyses, we inspected biplots of pupal weight vs. larval weight and removed five clear outliers that had larval weight >0·1 g (see Fig. 3 below).
Larval survival during the final instar averaged 80%, but it was greatly affected by temperature, declining from 100% in the highest temperature to only 55% in the lowest temperature (Fig. 2; F = 13·1, d.f. = 2, P < 0·0001). Larval weight in the beginning of the final instar and population type did not affect survival, but Pgi genotype had a significant effect (F = 4·7, d.f. = 1, P = 0·032). Although the interaction between temperature and Pgi genotype was not formally significant, it is apparent that the genotypes differed in the low-temperature treatment, in which 63% of the AA individuals but only 48% of the AC individuals survived until pupation (Fig. 2).
Larval and pupal weights
None of the explanatory variables, including population type, genotype and their interactions, had a significant effect on larval weight in the beginning of the final instar. However, although not representing a significant difference, we note in relation to the results presented below that the AA homozygotes tended to be heavier than the AC heterozygotes especially in females (2·3% and 9·6% heavier, P = 0·74 and 0·094 in males and females, respectively).
Pupal weight was strongly affected by temperature (Table 2), essentially because in both sexes pupae were much heavier in the high-temperature treatment than in the intermediate and low-temperature treatments (Fig. 3; no significant difference between the latter two, Tukey’s pairwise comparison). Secondly, pupal weight was affected by the larval weight, as individuals that were heavy in the beginning of the final larval instar tended to have heavier pupae (Table 2).
|Factor||Type of variable||d.f.||Males||Females|
|F or z||P value||F or z||P value|
|Pop type × genotype||Fixed||1||1·62||0·206||3·95||0·051|
Genotype had a significant effect on pupal weight in both males and females (Table 2), with the AA homozygotes being heavier than the AC heterozygotes, except in the low-temperature treatment, where there was no difference (Fig. 3). In both sexes, the difference between the genotypes was more apparent in new populations, and the interaction between population type and genotype was significant at 5% level in females. The AA homozygotes from new populations were heavier than the AA individuals from old populations, although the main effect of population type was not significant (Table 2).
In the above analyses, family was included as a random factor. This involves a potential problem in analysing the genotypic effect, because there was relatively little genotypic variation within families. We therefore conducted a complementary analysis on family means separately for the different temperature treatments. In the low and high-temperature treatments, there was no significant genotypic effect in either males or females (see also Fig. 3). In the intermediate-temperature treatment, families with high frequency of AA individuals had significantly heavier pupae than families with low frequency of AA individuals (females: F = 7·5, d.f. = 1, P = 0·021, males: F = 6·83, d.f. = 1, P = 0·034). Thus, the analysis on family means is consistent with the genotypic effect in Table 2. It is noteworthy that in all six analyses on family means, families from new populations had heavier pupae than families from old populations (P = 0·016); though, the effect was significant only in males in the intermediate temperature.
Adult life span
Pupal weight may have various fitness consequences for adult butterflies (Discussion). In the present experiment, we measured adult life span. Treatment had no significant effect on adult life span (females: P = 0·15 and males: P = 0·66); hence, we focus below on the effects of pupal weight and genotype. We removed any effect of treatment by taking residuals from an anova. In females, the AC heterozygotes had significantly longer life span than the AA homozygotes (F = 9·86, d.f. = 1, P = 0·01), and the interaction between genotype and residual pupal weight was marginally significant (F = 4·62, d.f. = 1, P = 0·05), because of stronger effect of pupal weight in the AC heterozygotes than in the AA homozygotes (Fig. 4a). The pattern was the same in males (Fig. 4b) but the effects of genotype and pupal weight on life span were not significant.
Previous studies on the Glanville fritillary have shown that the AC heterozygotes in the Pgi SNP AA111 are superior to the AA homozygotes in several adult life-history traits. The AC heterozygotes have higher flight metabolic rate (Haag et al. 2005; Niitepõld 2010), they are able to fly at lower ambient temperatures (Saastamoinen & Hanski 2008, Niitepõld et al. 2009), have higher dispersal rate in the field (Niitepõld et al. 2009) and they lay larger egg clutches (Saastamoinen & Hanski 2008). The AC heterozygotes have also longer adult life span than the AA homozygotes (Klemme & Hanski 2009; Saastamoinen, Ikonen & Hanski 2009); though, in the latter study the difference was significant in only females. Similarly, in our results the AC females had longer life span than the AA females, especially among large-bodied individuals (Fig. 4a). So far, there is only one exception to the superior performance of the AC heterozygotes in adult butterflies: under high ambient air temperatures, the AA homozygotes have higher flight metabolic rate and dispersal rate in the field than the AC heterozygotes (Niitepõld et al. 2009). This may reflect a trade-off between molecular kinetic efficiency vs. thermal stability of the PGI molecule as suggested for the leaf beetle C. aeneicollis (Dahlhoff & Rank 2000; Rank et al. 2007) and Colias butterflies (Watt 1992, 2003). However, not all associations between molecular variation in Pgi and life-history traits relate to temperature, and there may be other factors apart from possible molecular trade-offs that maintain polymorphism in this locus, such as spatial and temporal variation in selection (Reznick 1985; Roncé & Olivieri 2004), for which there is some evidence for the Glanville fritillary (Hanski et al. 2004; Zheng, Ovaskainen & Hanski 2009).
One very strong mechanism that contributes to polymorphism in Pgi SNP AA111 is low fitness of the CC homozygotes, reflected in their lower than expected frequency in random population samples (including the present material) as well as among the progeny of AC × AC crosses in the laboratory (Orsini et al. 2009). The reason for the clear deficiency of the CC homozygotes is not known, but it is possible that the C allele in SNP AA111 is linked with a strongly deleterious or lethal mutation in a common haplotype. It is noteworthy that the CC homozygotes in this SNP occur at the expected frequency in some other parts of the geographical range of the Glanville fritillary than in Finland (I.H. unpublished).
To return to the relative performances of the AA and AC genotypes, yet another factor that may maintain polymorphism is antagonistic pleiotropy, in which a genotype that is advantageous for some traits incurs a cost in terms of other traits. Antagonistic pleiotropy may also involve trade-offs between different life stages (Bradshaw & Holzapfel 1996; Feder et al. 1997; Sorensen & Loeschcke 2004). For instance, in the leaf beetle C. aeneicollis, the PGI genotype affects thermal tolerance in a dissimilar manner in different life-history stages (Neargarder, Dahlhoff & Rank 2003; Rank et al. 2007). Thus, adults that possessed the PGI-1 allele run faster than individuals with the PGI-4 allele following an exposure to stressful temperatures, but in larvae the result was the opposite (Rank et al. 2007; see also McMillan et al. 2005).
The present results support the notion of dissimilar genotypic effects in different life-history stages. In contrast to the previous results on adult butterflies (references above), our results on larvae revealed two traits in which the AA homozygotes performed better than the AC heterozygotes. First, larval mortality during the final instar was significantly lower in AA homozygotes under stressful conditions, represented by the low-temperature treatment. Second, the AA homozygotes had significantly heavier pupae than the AC heterozygotes.
Lower larval survival of AC heterozygotes in low temperature is seemingly contradictory to previous findings indicating that, at the adult stage, the AC heterozygotes do better than AA homozygotes in low ambient temperatures (Saastamoinen & Hanski 2008; Niitepõld et al. 2009). However, in the case of larval survival, low temperature was so low that it greatly increased larval mortality, and the important consideration here is likely to be stressful conditions for development rather than low temperature per se. Adult butterflies encounter a diverse thermal environment and are able to regulate their body temperature by behavioural means, whereas larvae in the present experiment were forced to experience the experimental temperature regime continuously. In any case, results on thermal tolerance for both C. aeneicollis and the Glanville fritillary indicate that the Pgi genotypic effects are complex and may differ between the different life stages.
Larval weight in the beginning of the final instar was not affected by population type nor by Pgi genotype, but pupal weight was affected by both factors. Therefore, the genotypic effect became evident during the last instar, which is the stage during which larvae gain most of their ultimate weight in absolute terms. At the pupal stage, the AA homozygotes were heavier in high and intermediate-temperature treatments than the AC heterozygotes, whereas in the low-temperature treatment AC males were heavier and in females there was no difference between the genotypes. These results can be interpreted in terms of a trade-off between thermal stability and kinetic performance, consistent with previous studies on adult butterflies showing superior performance of AC individuals in lower ambient temperatures (Saastamoinen & Hanski 2008; Niitepõld 2010).
In insects, heavier pupae are often correlated with greater lifetime fecundity in females (Honek 1993), essentially because larger individuals have more resources for reproduction. This applies especially to species that accumulate all or most of the resources for adult life during the larval stage, whereas in species that feed as adults the reproductive performance is more dependent on adult foraging success (Tammaru & Haukioja 1996). Previous studies on the Glanville fritillary have found only limited correlation or no correlation at all between pupal weight and lifetime fecundity (Saastamoinen 2007a; Klemme & Hanski 2009), and we found no relationship or even an inverse relationship between pupal weight and adult longevity. Clearly, resources accumulated during the larval stage do not necessarily suffice to guarantee high adult fitness. Nonetheless, it is possible that large pupal size is advantageous under some conditions, for instance when the availability of adult resources is limited, which would give a relative advantage to the AA homozygotes with heavier pupae. This conjecture is supported by our re-analysis of the results of Saastamoinen, Ikonen & Hanski (2009), who maintained adult butterflies under two conditions, in warm temperature with food and in cyclic warm-cool temperature without food. In the former case, pupal weight had no significant effect on life span (P = 0·15 for the effect of pupal weight, n = 310 butterflies of both sexes), but in the no-food treatment, individuals with heavier pupae lived longer (P = 0·02 for pupal weight, P = 0·0001 for sex, no significant interaction, n = 124 butterflies). The results of Saastamoinen, Ikonen & Hanski (2009) demonstrate that adult butterflies use body resources in the thorax and the abdomen for both maintenance and reproduction; though, which resources (body part) was used depended on the environmental conditions.
Apart from differences between the Pgi genotypes, our results revealed differences between the two population types – newly established vs. old local populations. The general significance of population age stems from the population ecological context: the Glanville fritillary occurs as a large metapopulation in the Åland Islands with a high rate of population turnover (Hanski 1999), and new populations tend to be established by more dispersive females with correlated life-history differences to old-population females (Hanski et al. 2002, 2004). Based on previous studies, it is apparent that the life-history differences between the new and old populations (Hanski et al. 2002, 2004) are associated with the Pgi genotype (Haag et al. 2005; Saastamoinen 2007b; Zheng, Ovaskainen & Hanski 2009), but apparently other genes are involved as well. This conjecture is supported by a gene expression study (Wheat et al. 2011), which found many other genes with significant expression differences between new and old populations. In particular, females originating from newly established populations had higher expression of genes involved in egg provisioning and maintenance of flight muscle proteins than females originating from old populations (Wheat et al. 2011).
In conclusion, the present results show significant Pgi genotypic effects on larval survival and performance, and the results indicate that the genotypic effects are dependent on temperature, sex and population type (new vs. old populations). Importantly, and contrary to previous results on adults, the AA homozygotes in SNP AA111 performed better than the AC heterozygotes in terms of larval survival and growth. Molecular-level mechanisms of the strong Pgi genotypic effects reported here and in the previous studies remain unclear, but the present results underscore the importance of studying all life-history stages while examining the association between molecular variation and variation in individual performance and fitness.
We thank Suvi Ikonen for help in rearing the larvae, Luisa Orsini ja Toshka Nyman for help with genotyping, Marjo Saastamoinen and three anonymous referees for comments on the manuscript, and the Lammi Biological Station, Academy of Finland (grant numbers 131155, 38604 and 44887, Finnish Centre of Excellence Programmes 2000–2005, 2006–2011) and the European Research Council (AdG number 232826) for funding.
- 2004) Checkerspot reproductive biology. On the Wings of Checkerspots: A Model System for Population Biology (eds P. Ehrlich & I. Hanski). pp. 92–112, Oxford University Press, New York. & (
- 1996) Genetic constraints to life-history evolution in the pitcher-plant mosquito, Wyeomyia smithii. Evolution, 50, 1176–1181. & (
- 1977) Phosphoglucose isomerase polymorphism and natural-selection in sand crab, Emerita talpoida. Evolution, 31, 331–340. (
- 2000) Functional and physiological consequences of genetic variation at phosphoglucose isomerase: heat shock protein expression is related to enzyme genotype in a montane beetle. Proceedings of the National Academy of Sciences of the United States of America, 97, 10056–10061. & (
- 1997) Selective maintenance of allozyme differences among sympatric host races of the apple maggot fly. Proceedings of the National Academy of Sciences of the United States of America, 94, 11417–11421. , , & (
- 2003) Another perspective on the slow-growth/high-mortality hypothesis: chilling effects on swallowtail larvae. Ecology, 84, 263–268. & (
- 1982) Electrophoretic and functional enzymic evolution in 4 species of Eastern Pacific barracudas from different thermal environments. Evolution, 36, 97–106. & (
- 2005) A candidate locus for variation in dispersal rate in a butterfly metapopulation. Proceedings of the Royal Society B-Biological Sciences, 272, 2449–2456. , , & (
- 1999) Metapopulation Ecology. Oxford University Press, Oxford, UK. (
- 2006) Molecular-level variation affects population growth in a butterfly metapopulation. Plos Biology, 4, 719–726. & (
- 2002) Population history and life history influence the migration rate of female Glanville fritillary butterflies. Oikos, 98, 87–97. , , , & (
- 2004) Variation in migration propensity among individuals maintained by landscape structure. Ecology Letters, 7, 958–966. , , , & (
- 1995) Thermal adation in biological-membranes – is homeoviscous adaptation the explanation. Annual Review of Physiology, 57, 19–42. (
- 1993) Warm Caterpillars and Hot Maggots. The Hot-Blooded Insects: Strategies and Mechanisms of Thermoregulation (ed. B. Heinrich). pp. 382–410, Harvard Univ. Press, Cambridge. (
- 2002) Temperature. Biochemical adaptation: Mechanism and Process in Physiological Evolution, pp. 290–438. Oxford University press, Oxford. & (
- 1981) Evolutionary genetics of Metridium senile .2. Geographic patterns of allozyme variation. Biochemical Genetics, 19, 145–154. (
- 1993) Intraspecific variation in body size and fecundity in insects – a general relationship. Oikos, 66, 483–492. (
- 2008) Phosphoglucose isomerase genotype affects life-history traits and cold stress resistance in a Copper butterfly. Functional Ecology, 22, 887–894. , & (
- 1997) Balancing selection on electrophoretic variation of phosphoglucose isomerase in two species of field cricket: Gryllus veletis and G. pennsylvanicus. Genetics, 147, 609–621. & (
- 1983) Thermoregulatory strategies in Colias butterflies – thermal-stress and the limits to adaption in temporally varying environments. American Naturalist, 121, 32–55. & (
- 2009) Heritability of and strong single gene (Pgi) effects on life-history traits in the Glanville fritillary butterfly. Journal of Evolutionary Biology, 22, 1944–1953. & (
- 1998) Biology of the Glanville Fritillary Butterfly (Melitaea cinxia). Ph.D Thesis, University of Helsinki, Helsinki. (
- 2005) Natural temperature variation affects larval survival, development and Hsp70 expression in a leaf beetle. Functional Ecology, 19, 844–852. , , & (
- 2003) Variation in thermal tolerance is linked to phosphoglucose isomerase genotype in a montane leaf beetle. Functional Ecology, 17, 213–221. , & (
- 2004) Structure and dynamics of Melitaea cinxia metapopulations. On the Wigs of Checkerspots: A Model System for Population Biology (eds P. Ehrlich & I. Hanski). pp. 63–91, Oxford University Press, Oxford. , & (
- 2010) Genotype by temperature interactions in the metabolic rate of the Glanville fritillary butterfly. Journal of Experimental Biology, 213, 1042–1048. (
- 2009) Flight metabolic rate and Pgi genotype influence butterfly dispersal rate in the field. Ecology, 90, 2223–2232. , , , , , , , & (
- 2009) Fitness differences associated with Pgi SNP genotypes in the Glanville fritillary butterfly (Melitaea cinxia). Journal of Evolutionary Biology, 22, 367–375. , , , , & (
- 1992) Glucosephosphate isomerase and fitness-effects of temperature on genotype dependent mortality and enzyme-activity in 2 species of the genus Gammarus (Crustacea, Amphipoda). Evolution, 46, 1568–1573. & (
- 2004) Changes in gene expression associated with acclimation to constant temperatures and fluctuating daily temperatures in an annual killifish Austrofundulus limnaeus. Journal of Experimental Biology, 207, 2237–2254. & (
- 1991) Genetic mechanism for adapting to a changing environment. Annual Review of Genetics, 25, 629–659. , , & (
- 2002) Allele frequency shifts in response to climate change and physiological consequences of allozyme variation in a montane insect. Evolution, 56, 2278–2289. & (
- 2007) Phosphoglucose isomerase genotype affects running speed and heat shock protein expression after exposure to extreme temperatures in a montane willow beetle. Journal of Experimental Biology, 210, 750–764. , , , & (
- 1985) Costs of reproduction – an evaluation of the emperical-evidence. Oikos, 44, 257–267. (
- 2001) Genetic architecture of thermal adaptation in Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America, 98, 525–530. , & (
- 2005) Changes in gene expression following high-temperature adaptation in experimentally evolved populations of E. coli. Physiological and Biochemical Zoology, 78, 299–315. , & (
- 2004) Life history evolution in metapopulations. Ecology, Genetics and Evolution of Metapopulations (eds I. Hanski & G. O). pp. 227–258, Elsvier Academic Press, San Diego CA. & (
- 2007a) Life-history, genotypic, and environmental correlates of clutch size in the Glanville fritillary butterfly. Ecological Entomology, 32, 235–242. (
- 2007b) Mobility and lifetime fecundity in new versus old populations of the Glanville fritillary butterfly. Oecologia, 153, 569–578. (
- 2008) Genotypic and environmental effects on flight activity and oviposition in the Glanville fritillary butterfly. American Naturalist, 171, E701–E712. & (
- 2009) Significant effects of Pgi genotype and body reserves on lifespan in the Glanville fritillary butterfly. Proceedings of the Royal Society B-Biological Sciences, 276, 1313–1322. , & (
- 2003) Insects at low temperatures: an ecological perspective. Trends in Ecology & Evolution, 18, 257–262. , , & (
- 1995) Proteins and temperature. Annual Review of Physiology, 57, 43–68. (
- 2004) Effects of relative emergence time on heat stress resistance traits, longevity and hsp70 expression level in Drosophila melanogaster. Journal of Thermal Biology, 29, 195–203. & (
- 1996) Capital breeders and income breeders among Lepidoptera – consequences to population dynamics. Oikos, 77, 561–564. & (
- 2001) Clinal variation for amino acid polymorphisms at the Pgm locus in Drosophila melanogaster. Genetics, 157, 1649–1663. & (
- 1977) Adaption at specific loci. 1. Natural-selection on phosphoglucose isomerase of Colias butterflies – biochemical and population aspects. Genetics, 87, 177–194. (
- 1983) Adaptation at specific loci. 2. Demographic and biochemical-elements in the maintenance of the Colias Pgi polymorphism. Genetics, 103, 691–724. (
- 1992) Eggs, enzymes, and evolution – natural genetic-variants change insect fecundity. Proceedings of the National Academy of Sciences of the United States of America, 89, 10608–10612. (
- 2003) Mechanistic Studies of Butterfly Adaptations. Butterflies – Ecology and Evolution Taking Flight (eds C. Boggs, W. Watt & P. Ehrlich). pp. 319–352, University of Chicago Press, Chicago. (
- 1983) Adaptation at specific loci. 3. Field behavior and survivorship differences among Colias Pgi genotypes are predictable from invitro biochemistry. Genetics, 103, 725–739. , & (
- 2010) Nucleotide polymorphism at a gene (Pgi) under balancing selection in a butterfly metapopulation. Molecular Biology and Evolution, 27, 267–281. , , , & (
- 2011) Functional genomics of life history variation in a butterfly metapopulation. Molecular Ecology, in press. , , , , , , & (
- 2000) Environmental Physiology of Animals. Blackwell Science Ltd., Oxford. , & (
- 1987) Temperature-dependent kinetic variation among phosphoglucose isomerase allozymes from the wing-polymorphic water strider, Limnophorus canaliculatus. Molecular Biology and Evolution, 4, 266–285. (
- 2009) Modelling single nucleotide effects in phosphoglucose isomerase on dispersal in the Glanville fritillary butterfly: coupling of ecological and evolutionary dynamics. Philosophical Transactions of the Royal Society B-Biological Sciences, 364, 1519–1532. , & (