Positive Sire Effects and Adaptive Genotype by Environment Interaction Occur despite Pattern of Local Maladaptation in Roadside Populations of an Amphibian
The global road network causes many negative ecological effects. Contrasting our knowledge of these effects, insights into evolutionary consequences of roads remain undeveloped. Here, we study a suite of populations of the Wood Frog that appear to be evolving maladaptively in response to road-adjacency. Specifically, when raised together in roadside pools, roadside populations survive at lower rates compared to populations away from roads. To begin to understand the cause of this survival disadvantage, we investigated potential parental and genetic sources of maladaptation. First, to assess whether parental effects might induce maladaptation, we measured adult body weight to length ratio (‘relative weight') and its influence on offspring survival in a reciprocal transplant experiment across 12 populations. Next, to assess whether genetic effects might limit adaptive responses in offspring, we estimated genetic correlations between environments for survival and fitness-related traits. We found that relative weight was higher in roadside populations and, for males, had a positive influence on offspring survival. This demonstrates a novel transgenerational effect of Wood Frog adult males but suggests that this effect is not causing maladaptive survival. Genetic correlations indicated that a subset of roadside genotypes respond adaptively to road-adjacency despite population level maladaptive survival. This suggests that metapopulation dynamics and/or high levels of nonadditive genetic variance may be limiting adaptation or that insufficient time has elapsed for adaptation to occur. Together, these results highlight the complexity and scale of responses to a pervasive feature of landscape alteration revealed by evolutionary approaches.
Study region and reciprocal transplant design. Location of 12 pools comprising reciprocal transplant are shown on a map of the region. Gray symbols indicate roadside pools; white symbols indicate woodland pools. Each roadside-woodland pool pair comprising a population level transplant shares a common symbol shape. Interstate highway (I-84) is marked with an interstate highway symbol and can be seen running in a southwest-northeast direction. Primary roads are heavily shaded, while secondary roads are lightly shaded. Bar graph inset shows mean (±1 SE) specific conductance (μS) for roadside and woodland pools. For roadside pools, conductivity is shown as the average of surface and bottom values. This is because specific conductance in the bottom waters of roadside pools is nearly twice that of the surface water. No such vertical gradient exists in woodland pools.
Adult relative weight. Mean adult relative weight (±1 SE) is shown for breeding adult females and males from the roadside (dark bars) and woodland (light bars) populations. Relative weight was measured following capture on the inbound migration to breed. Measurements were made just prior to pairing males and females together to breed. Relative weight is shown as weight (g) divided by snout–vent length (SVL [mm]).
Adult male body condition influence on offspring survival. Offspring survival is shown as the proportion of embryos surviving to larval feeding stage in relation to sire body condition (i.e., body weight [g] divided by SVL [mm]). Proportion survival was measured at the enclosure level. Fitted lines representing the fixed effects from the mixed model are shown. Points and lines are color and symbol coded across each level of the Pt x Et interaction. Open symbols and black lines correspond to the roadside population in the roadside environment (open circles, solid black line) and woodland environment (open triangles, dashed black line). Filled (gray) symbols correspond to the woodland population in the woodland environment (filled circles, solid gray line) and roadside environment (filled triangles, dashed gray line).
Family level variation in larval survival, size (SVL), and developmental rate (DR) for roadside populations (black lines) and woodland populations (gray lines). Both population types presented similar patterns of genotypic expression across environments. Pattern of expression in DR suggests the absence of a genotype by environment interaction, as the phenotype is consistent between woodland and roadside environment. However, the pattern of expression in larval survival and SVL indicates a genotype by environment interaction is present, as genotypes present different phenotypes between woodland and roadside environments.
Genetic correlations in roadside and woodland environments. Genetic correlations of survival, size (SVL), and developmental rate (DR) are shown separately for the roadside (top panel) and woodland (bottom panel) populations. Genetic correlation modes of the posterior density distributions are indicated by the dot, while means are indicated by the triangle. Solid lines represent the 95% credible interval around the mean. Genetic correlations that depart from 1 indicate a genotype by environment interaction.
Contributor Notes
Associate Editor: A. Hendry.
From “Eco-Evolutionary Dynamics in Cold Blood,” an ASIH-sponsored symposium at the 2016 Joint Meeting of Ichthyologists and Herpetologists in New Orleans, Louisiana.