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

Do hatchling phenotypes differ between eggs from different nests?

The first prediction from the hypothesis of incubation-induced phenotypic effects is that there will be significant among-nest variation in the phenotypes of hatchling lizards. To test this prediction, we used nest identification number as the factor in a one-factor MANOVA on the data for eggs that had been returned to their original nests for incubation. These represent the "natural" condition for this population. Our test revealed significant differences among nests in the phenotypes of hatchling lizards within this group (Hotelling-Lawley trace = 8.48, [F.sub.240, 1847] = 4.35, P [less than] 0.0001). More detailed examination using a series of one-factor ANOVAs showed that this significant result was due to strong differences among nests in traits such as incubation period ([F.sub.15, 140] = 37.53, P [less than] 0.0001), and hatchling morphology (mass: F.sub.15, 140] = 5.62, P [less than] 0.0001; SVL: [F.sub.15, 140] = 4.78, P [less than] 0.0001; tail length: [F.sub.15, 140] = 2.40, P [less than] 0.005). These effects persisted through the 1st wk of the lizard's life (at 1 wk old, mass: [F.sub.15, 140] = 6.42, P [less than] 0.0001; SVL: [F.sub.15, 140] = 3.72, P [less than] 0.0001; tail length: [F.sub.15, 140] = 3.23, P [less than] 0.005). Bodily proportions also differed among lizards emerging from different nests, at hatching, and at 1 wk of age (P [less than] 0.05 for all tests on residual scores for body shape). Egg survival rates also differed among nests ([[Chi].sup.2] = 59.8, 15 dr, P [less than] 0.0001). However, running speeds of hatchlings were not affected (over 1 m: [F.sub.15, 140] = 0.72, P = 0.76; over 25 cm: [F.sub.15, 140] = 0.97, P = 0.49).

Because hatchlings from different nests were exposed to different durations of laboratory incubation, it is possible that some of these among-nest differences in hatchling phenotypes were generated during exposure to laboratory rather than field conditions. However, we found no significant correlations between the proportion of total incubation spent in the laboratory, vs. any other trait except incubation period. Hence, these analyses confirm that naturally incubated eggs from different nests produce hatchlings with significantly different phenotypes.

We now proceed to disentangle some of the factors contributing to this among-nest variation.

Are among-nest differences in hatchling phenotypes due to nest-of-origin effects, to incubation conditions, or to both?

We can answer this question by analyzing data on the translocated eggs only. We used one-factor MANOVA, with the factor being either (1) the identification number of the nest in which the eggs were originally laid; or (2) the identification number of the nest in which they incubated. If the former MANOVA provides a significant result, it means that part of the phenotypic variation among hatchling skinks is due to some factor operating prior to the time we removed the eggs from their original nests. This factor could be a "nest-of-origin" effect (e.g., due to genes, maternal investment patterns, or to some environmental effect that operated prior to oviposition or in the first few days after the eggs were laid). If the second MANOVA yields a significant result, it means that incubation conditions influence hatchling phenotypes.