Effects of telomere length on reproductive success
Finally, the effect of fledgling TL on the total number of recruits produced per individual (LRS, lifetime reproductive success) was analyzed. We included only individuals that survived until breeding and that were not removed during the artificial selection (and thus allowed to reproduce). In the high population, 10 out of 22 recruiting individuals produced at least one recruit and 39 out of 80 individuals did so in the low population. While reproduction per semay accelerate telomere loss (Sudyka, 2019), we test here the predictive value of early-life TL and/or tarsus length on subsequent reproductive output (Eastwood et al., 2019). Since LRS and lifespan (measured in years from first to last observation) were highly correlated (high population; Pearson’s r =0.83,p <0.0001, low population; Pearson’sr =0.70, p <0.0001), we tested whether TL or tarsus length predict LRS while controlling for lifespan (which is equivalent to the individual average annual reproductive success, ARS). We fitted a set of generalized linear mixed models with a Poisson error distribution separately for each population to facilitate model convergence, using the package glmmTMB (Brooks et al. 2017). Sex was included as fixed factor and brood identity and year were included as random intercepts in all models. We compared the same 9 candidate models for each population using AICc and models were validated using the DHARMa package (Hartig, 2019).