The influence of natural incubation environments on the phenotypic traits of hatchling lizards

Ecology, Dec, 1997 by Richard Shine, Melanie J. Elphick, Peter S. Harlow

Both tests yielded highly significant results. The first test (one-factor MANOVA on data for hatchlings from translocated eggs, with the nest of laying as the factor) showed highly significant phenotypic variation among hatchlings depending on their nest of origin (Hotelling-Lawley trace = 4.13, [F.sub.240, 1922] = 2.20, P [less than] 0.0001). Closer inspection of one-factor ANOVAs for each trait confirmed that these nest-of-origin effects were significant (P [less than] 0.05) for most measures of hatchling size and shape, and close to significance for running speeds. However, survival rates (proportions of eggs hatching) did not depend upon their nest of origin ([[Chi].sup.2] = 23.41, 16 df, P = 0.10). The analogous MANOVA for incubation effects (i.e., phenotypes of hatchlings from translocated eggs, tested against the identity of the nest in which they incubated) also showed a highly significant effect (Hotelling-Lawley trace = 4.01, [F.sub.240, 1922] = 2.14, P [less than] 0.0001). As for nest-of-origin effects, the variables most strongly affected by incubation environment were morphological traits rather than running speeds. Survival rates of eggs to hatching also depended on the nest to which they were translocated ([[Chi].sup.2] = 68.25, 16 df, P [less than] 0.001).

In combination, these analyses show that the amongnest differences documented from the "natural" situation (i.e., hatchlings emerging from the nests in which they were originally laid) are due to a combination of effects operating prior to the time the eggs were transferred in our experiment (i.e., nest-of-origin effects plus incubation conditions in the first few days of development), and other factors that affected the eggs during incubation.

Are the thermal regimes in natural nests responsible for induction of phenotypic changes in hatchlings?

Although our analyses document an effect of incubation environment on hatchling phenotypes, we have not identified which aspects of nest environment are important in this respect. Natural nests vary in many ways, of which the two most obvious are temperature and soil moisture. However, many other variables undoubtedly differ among nests as well (e.g., trace chemicals, pesticide residues, soil characteristics such as particle size, pH) in ways that might influence hatchling phenotypes. Thus, we need to determine whether some of the incubation-induced effects are due to temperature regimes rather than to other factors.

The simplest way to see whether the thermal environment affects hatchling phenotypes is to regress phenotypic trait values (of all eggs, translocated as well as replaced) against descriptors of the thermal regime under which these eggs incubated. A significant relationship between hatchling traits and thermal variables would support the notion that nest temperatures affect hatchling phenotypes. Because laboratory experiments have shown that the phenotypes of hatchling B. duperreyi are influenced by thermal variability as well as by mean temperatures (Shine and Harlow 1996), we regressed hatchling traits against two descriptors of the thermal environment in each nest: the mean temperature, and an index of the variance relative to the mean (i.e., the residual score from the linear regression of the variance against the mean). This index provides a measure of the variance independent of mean nest temperature.