Figure 5 : Reaction norms for each family separated by source island; the first and second columns of graphs are for individuals from the open and shaded islands, respectively (i.e., reaction norms reflect the trait values for all offspring from a single mother. The third column represents the reaction norms averaged for each family group (red curve = open island, blue curve = shaded island, black curve = average of all family groups). (a) Reaction norms for body mass; (b) Reaction norms for snout-vent length; (c) Reaction norms for spring speed; (d) Reaction norms for water loss. All reaction norm slopes reported are measured directionally from the open to the shaded incubation treatment. That is, a positive slope conveys a larger trait value in the shaded treatment and a negative slope conveys a larger trait value in the open treatment. The y -axis has been changed for the plots showing the average reaction norms to better see the patterns.
Table 1. Effect of incubation treatment, island, and covariates on hatchling phenotypes. Statistically significant interaction terms were retained in the final models. Descriptive statistics (means and standard deviations) for each island and incubation treatment are reported in supplemental Table 2). Estimates for treatment and source island effects were calculated with the shaded incubation treatment and shaded island as references. In the random effects columns σ2 = the residual variance, τ00 = the random intercept variance (maternal ID was the random effect in models for mass and SVL), ICC = interclass correlation.