Phenotypic plasticity drives phenological changes in a Mediterranean blue tit population

Abstract

Earlier phenology induced by climate change, such as the passerines' breeding time, is observed in many natural populations. Understanding the nature of such changes is key to predict the responses of wild populations to climate change. Genetic changes have been rarely investigated for laying date, though it has been shown to be heritable and under directional selection, suggesting that the trait could evolve. In a Corsican blue tit population, the birds' laying date has significantly advanced over 40 years, and we here determine whether this response is of plastic or evolutionary origin, by comparing the predictions of the breeder's and the Robertson‐Price (STS) equations, to the observed genetic changes. We compare the results obtained for two fitness proxies (fledgling and recruitment success), using models accounting for their zero inflation. Because the trait appears heritable and under directional selection, the breeder's equation predicts that genetic changes could drive a significant part of the phenological change observed. We, however, found that fitness proxies and laying date are not genetically correlated. The STS, therefore, predicts no evolution of the breeding time, predicting correctly the absence of trend in breeding values. Our results also emphasize that when investigating selection on a plastic trait under fluctuating selection, part of the fitness‐trait phenotypic covariance can be due to within individual covariance. In the case of repeated measurements, splitting within and between individual covariance can shift our perspective on the actual intensity of selection over multiple selection episodes, shedding light on the potential for the trait to evolve.

Selection and the direction of phenotypic evolution

“…The Robertson equation has been applied to a few natural populations, those where time series of trait and fitness values are concurrently obtained with pedigrees such that trait breeding values can be estimated. Biquet et al (2022) provide an exemplary study of a blue tit population that has been followed for more than 40 years, and in which egg laying has changed to earlier and earlier spring dates, presumably because of climate change. Despite significant heritability for egg laying date and selection on it, however, modeling the expected breeding values for laying date did not reveal any temporal trend consistent with a general lack of genetic covariance between laying date and fitness during the whole period followed.…”

Conceptualizing the evolutionary quantitative genetics of phenological life-history events: Breeding time as a plastic threshold trait

“…To examine these possibilities, leading empirical studies have used the best available individual-based datasets to estimate key quantitative genetic parameters underlying observed variation in breeding time in wild vertebrate populations (e.g., Merilä et al 2001;Gienapp et al 2006;Brommer et al 2008;Charmantier et al 2008;Husby et al 2010Husby et al , 2011Germain et al 2016;Bonnet et al 2019;Ramakers et al 2019;Evans et al 2020;Biquet et al 2021). Here, breeding time is treated as a continuously distributed trait that is directly measured and analyzed on scales of ordinal dates, or days since some arbitrary seasonal start time.…”

“…Resulting analyses commonly report moderate additive genetic variance and heritability in breeding date (i.e., effectively in reaction norm elevation; Merilä et al 2001;Gienapp et al 2006;Bonnet et al 2019;Biquet et al 2021), sometimes including associative genetic effects of females' socially paired males (Germain et al 2016;Evans et al 2020). There is commonly evidence of substantial among-individual variance in reaction norm slope (i.e., I×E), with some reported evidence of additive genetic variance (i.e., G×E; Brommer et al 2008;Husby et al 2011; but see Charmantier et al 2008).…”

“…As for any examination of plasticity, such analyses will require measurements of key environmental variables that are known or hypothesized to drive trajectories of liability (and resulting transitions to breeding) across appropriate time periods. Indeed, while analyses that simply estimate additive genetic variance in observed breeding date can be accomplished without explicitly considering any form of plasticity or hence environmental covariates (e.g., Germain et al 2016;Evans et al 2020;Biquet et al 2021), analyses of environmentally driven responses of threshold traits cannot (Supporting Information S4). However, in contrast to current common practice, there is no need to identify discrete critical time windows within which values of particular environmental variables are postulated to act as "cues" (Figure 1; e.g., Brommer et al 2008;Husby et al 2010;Porlier et al 2012;Phillimore et al 2016;Bonamour et al 2019;Simmonds et al 2019b).…”

Phenotypic stasis with genetic divergence

“…Explanations for short-term phenotypic stasis have relied on showing that in many cases there were no changes in the breeding traits’ values, that is, no genetic divergence, either because of selection on unmeasured traits that are genetically correlated with observed ones or because of correlated selection due to unknown environmental covariation between observed and unobserved traits with fitness e.g., (Czorlich et al, 2022; Kruuk et al, 2002), both instances of “indirect” selection. Short-term phenotypic stasis without genetic divergence has also been explained by phenotypic plasticity allowing the tracking of environmental fluctuations e.g., (Biquet et al, 2022; de Villemereuil et al, 2020). These studies indicate that phenotypic evolution cannot be understood when considering each trait independently of others and that a multivariate description of selection and standing genetic variation is needed.…”