While recent experimental work on a variety of reptile species has demonstrated that incubation temperature influences hatchling phenotypes, the biological significance of such phenotypic variation remains unclear. Incubation temperature may exert significant long-term phenotypic effects. Alternatively, such influences may be temporary, or negligible relative to effects induced by genetic factors, or by the environmental conditions experienced after hatching. Even if incubation temperature exerts long-term effects on phenotype, this might occur indirectly (by influencing hatching dates) rather than by direct modifications of developmental processes. We quantified the influences of the source population, incubation temperature and rearing environment, on the phenotype of the Australian garden skink (Lampropholis guichenoti) from populations that differ in nest temperature and phenotype. Intcrpopulation differences in the phenotypes of young lizards were found to be a product of all three factors. However, the long-term effects of both population and incubation temperature operated indirectly (through variation in the date of hatching) rather than directly (through genetic or developmental factors). That is, once all temporal effects were removed, the only discernible influence on juvenile phenotypes was their rearing environment. Thus, some of the most important influences on lizard phenotypes may operate via modifications of hatching date.

Rapid embryonic development is a potential adaptation to cold climates. We tested this hypothesis for nine species of Sceloporus lizards from four species groups and one species of Urosaurus, an outgroup, using observations corrected for incubation temperature and hatchling size. Phylogenetically based comparisons indicated that relatively rapid development is characteristic of the scalaris species group and relatively slow development is characteristic of the undulatus species group. Comparisons within these lineages were therefore used to determine whether develop mental rate was related to climate, as judged by elevation and latitude. Within the scalaris species group, developmental rates of cool climate populations of S. aeneus and S, scalaris were not faster than that of a warm climate population of S. scalaris, Within the undulatus species-group, the developmental rate of cool climate populations or species were not faster than those of warm climate populations or species. In general, developmental rates of Sceloporus are lineage specific and do not appear to be adapted to local climates.

Geographic variation in phenotypes can result from proximate environmental effects as well as from underlying genetic factors. Reciprocal transplant experiments, in which organisms are moved from one area to another, offer a powerful technique to partition the effects of these two factors. However, many studies that have utilized this technique have focused on the post-hatching organism only and ignored potential effects of environmental influences acting during embryonic development. We examined the phenotypic responses of hatchling scincid lizards (Lampropholis guichenoti) incubated in the laboratory under thermal regimes characteristic of natural nests in two study areas in southeastern Australia. Although the sites were less than 120 km apart, lizards from these two areas differed in thermal regimes of natural nests, and in hatchling phenotypes (morphology, locomotor performance). We incubated eggs from each area under the thermal regimes typical of both sites. Some of the traits we measured (e.g. hatchling mass and snout-vent length) showed little or no phenotypic plasticity in response to differences in incubation conditions, whereas other traits (e.g. incubation period, tail length, inter-limb length, body shape, locomotor performance) were strongly influenced by the thermal regime experienced by the embryo. Thus, a significant proportion of the geographic variation in morphology and locomotor performance of hatchling lizards may be directly induced by differences in nest temperatures rather than by genetic divergence. We suggest that future studies using the reciprocal transplant design should consider environmental influences on all stages of the life-history, including embryonic development as well as post-hatching life.

1. A major focus of life-history research has been the analysis of reproductive effort (RE). However, while clearly defined in theory, RE has proved very difficult to measure. Consequently, researchers have looked for indices that estimate the components of RE. Uncritical use of indices of reproductive investment, such as measures of costs of reproduction, without verification of the underlying assumptions of the link between these traits, may lead to spurious conclusions. In this paper, the common assumption that the physical burden of the clutch impairs locomotor ability in gravid reptiles is examined.2. Two neighbouring populations of Australian scincid lizards (Lampropholis guichenoti) are similar in adult body sizes, body shapes and reproductive output (egg sizes, clutch sizes, relative clutch masses).3. Despite morphological and reproductive similarities, the effects of pregnancy on maternal locomotor ability (running speeds, as measured in a laboratory raceway) differed dramatically between the two populations.4. Lizards from the two populations ran at similar speeds when nongravid (i.e. after egg-laying), but pregnancy significantly reduced running speeds in one population and increased them in the other. Thus, superficial similarities in body size and reproductive output masked a strong divergence in the locomotor 'costs' of reproduction.5. Caution is advised if using simple measures of reproductive output (e.g. relative clutch mass) as indices of reproductive effort, or to generalize results even among conspecific populations.