Candidate gene-environment interactions and their relationships with timing of breeding in a wild bird population

Abstract

Monitoring and predicting evolutionary changes underlying current environmental modifications are complex challenges. Recent approaches to achieve these objectives include assessing the genetic variation and effects of candidate genes on traits indicating adaptive potential. In birds, for example, short tandem repeat polymorphism at four candidate genes (CLOCK, NPAS2, ADCYAP1, and CREB1) has been linked to variation in phenological traits such as laying date and timing of migration. However, our understanding of their importance as evolutionary predictors is still limited, mainly because the extent of genotype-environment interactions (GxE) related to these genes has yet to be assessed. Here, we studied a population of Tree swallow (Tachycineta bicolor) over 4 years in southern Québec (Canada) to assess the relationships between those four candidate genes and two phenological traits related to reproduction (laying date and incubation duration) and also determine the importance of GxE in this system. Our results showed that NPAS2 female genotypes were nonrandomly distributed across the study system and formed a longitudinal cline with longer genotypes located to the east. We observed relationships between length polymorphism at all candidate genes and laying date and/or incubation duration, and most of these relationships were affected by environmental variables (breeding density, latitude, or temperature). In particular, the positive relationships detected between laying date and both CLOCK and NPAS2 female genotypes were variable depending on breeding density. Our results suggest that all four candidate genes potentially affect timing of breeding in birds and that GxE are more prevalent and important than previously reported in this context.

Keywords: ADCYAP1; CLOCK; CREB1; GxE; NPAS2; candidate gene; incubation duration; laying date.

Figure 1

Mean female NPAS2 genotypes observed on the 40 farms (colored circles, see legend) in the study system in southern Québec, Canada. Number of females observed between 2010 and 2013 is represented by different circle sizes (range: 2 – 45). Forest patches (green), rivers and lakes (blue), other land uses (mostly agriculture; white), elevation (100-m gray isolines), latitude and longitude (in decimal degrees; thin black lines) are also represented. This figure was produced with QGIS 2.0 (QGIS Team Development, 2013).

Figure 2

Predictions from the linear mixed models of tree swallow laying date (Julian days; A-D) and incubation duration (days; E-F) correlate with candidate gene genotypes (CLOCK: A, B; NPAS2: C, E; ADCYAP1: D; CREB1: F). Interactions with breeding density (A, C), latitude (D) and May temperature (F) are presented for the first (gray) and third (black) quartile of environmental values. Genotype frequency histograms for male (gray) or female (white) and 95% confidence intervals of predictions (from models with year included as a fixed effect) are also presented on each panel.