Tit wit: environmental and genetic drivers of cognitive variation along an urbanization gradient

Abstract

Cognitive abilities can promote acclimation to life in cities. However, the genetic versus environmental drivers of cognition have rarely been studied in the wild and there exists a major knowledge gap concerning the role of cognition in adaptation to urban contexts. We evaluate cognitive variation in wild great tits (Parus major; N = 393) along an urban gradient, and estimate the genetic basis of this variation using a combination of a common garden experiment, quantitative genetic analysis, and genome-wide association study. Specifically, we measure inhibitory control abilities which affect how animals respond to novel challenges. We find that wild urban and forest tits do not clearly differ in inhibitory control performance (number of errors or the latency to escape) during a motor detour task; a result that was consistent in birds from urban and forest origins reared in a common garden (N = 73) despite average performance differing between wild and captive birds. Cognitive performance was repeatable (R = 0.35-0.38) and showed low to moderate heritability in the wild (h² = 0.16-0.28, but both estimates had high uncertainty). We identified five SNPs that were associated with the number of errors during the task, with two of these SNPs linked to genes related to serotonergic and dopaminergic systems that are known to play important roles in cognition. Altogether, our study finds limited evidence that inhibitory control abilities have evolved under novel urban contexts, yet reveals some evidence for a genetic basis of this cognitive trait in great tits.

Keywords: Cognition; Common garden experiment; GWAS; Great tit; Heritability; Inhibitory control.

Fig. 1

Cognitive assay used to measure performance related to inhibitory control in wild and common garden contexts. a side view of the cognitive assay cage in the field, b diagram showing assay components from above including a possible route to the exit in red, and c a video clip showing how the cognitive task was analyzed including the measurements of interest (note bird detouring barrier on the right)

Fig. 2

The effect of habitat type (forest vs. urban) and associated 95% credible intervals on a the number of errors and b the latency to escape in wild and common garden (CG) contexts. Estimates have been back transformed from Table 1 to show effects on the observed data scale. Forest and urban estimates are shown in green and grey, respectively. See Fig. S8 for these plots shown with raw data

Fig. 3

Visualizations of the variance explained by each random effect in the animal model (see Table 2) for top panel a,b the number of errors and bottom panel c,d the latency to escape in wild birds. Panels show a,c the relative percentage of each variance component and b,d 95% Highest Density Interval (HDI; in grey) of the posterior distributions of the variance ratios. From top to bottom, the posterior distributions are shown for heritability (h²), site (V_SITE), permanent environment (V_PE), and residual (V_R) ratio variance. The solid line corresponds to zero and dashed bold line corresponds to the median value

Fig. 4

Manhattan plot (on the left) with –log10 p-values for the association between marker genotype and a number of errors, b Latency of escaping for all 228444 SNPs and 253 individuals and their associated Q-Qplot (on the right). On the left: The X-axis is the genomic position of the SNPs in the genome, and the Y-axis is the negative log base 10 of the P-values. The green solid line indicates the genome-wide Bonferroni-corrected significant threshold (here log10(p) = 6.69) and the green dotted line the threshold for suggestive significant associations. Each chromosome is colored differently. SNPs with stronger associations with the trait will have a larger Y-coordinate value. On the right: Quantile–quantile (QQ) –plot of P-values. The Y-axis is the observed negative base 10 logarithm of the P-values, and the X-axis is the expected observed negative base 10 logarithm of the P-values under the assumption that the P-values follow a uniform [0,1] distribution. The grey surface show the 95% confidence interval for the QQ-plot under the null hypothesis of no association between the SNP and the trait