RESUMO
The relationship between growth and age-specific telomere length, as a proxy of somatic state, is increasingly investigated, but observed patterns vary and a predictive framework is lacking. We outline expectations based on the assumption that telomere maintenance is costly and argue that individual heterogeneity in resource acquisition is predicted to lead to positive covariance between growth and telomere length. However, canalization of resource allocation to the trait with a larger effect on fitness, rendering that trait relatively invariant, can cause the absence of covariance. In a case study of common tern (Sterna hirundo) chicks, in which hatching order is the main determinant of variation in resource acquisition within broods, we find that body mass, but not telomere length or attrition, varies with hatching order. Moreover, body mass and growth positively predict survival to fledging, whereas telomere length and attrition do not. Using a novel statistical method to quantify standardized variance in plasticity, we estimate between-individual variation in telomere attrition to be only 12% of that of growth. Consistent with the relative invariance of telomere attrition, we find no correlation between age-specific body mass or growth and telomere attrition. We suggest that common tern chicks prioritize investment in long-term somatic state (as indicated by canalization of telomere maintenance) over immediate survival benefits of growth as part of an efficient brood reduction strategy that benefits the parents. As such, interspecific variation in the growth-telomere length relationship may be explained by the extent to which parents benefit from rapid mortality of excess offspring.