Causes and short‐term consequences of variation in milk composition in wild sheep

Abstract

Ecologists seek to understand the fitness consequences of variation in physiological markers, under the hypothesis that physiological state is linked to variability in individual condition and life history. Thus, ecologists are often interested in estimating correlations between entire suites of correlated traits, or biomarkers, but sample size limitations often do not allow us to do this properly when large numbers of traits or biomarkers are considered. Latent variables are a powerful tool to overcome this complexity. Recent statistical advances have enabled a new class of multivariate models—multivariate hierarchical modelling (MHM) with latent variables—which allow to statistically estimate unstructured covariances/correlations among traits with reduced constraints on the number of degrees of freedom to account in the model. It is thus possible to highlight correlated structures in potentially very large numbers of traits. Here, we apply MHM to evaluate the relative importance of individual differences and environmental effects on milk composition and identify the drivers of this variation. We ask whether variation in bighorn sheep milk affects offspring fitness. We evaluate whether mothers show repeatable individual differences in the concentrations of 11 markers of milk composition, and we investigate the relative importance of annual variability, maternal identity and morphological traits in structuring milk composition. We then use variance estimates to investigate how a subset of repeatable milk markers influence lamb summer survival. Repeatability of milk markers ranged from 0.05 to 0.64 after accounting for year‐to‐year variations. Milk composition was weakly but significantly associated with maternal mass in June and September, summer mass gain and winter mass loss. Variation explained by year‐to‐year fluctuations ranged from 0.07 to 0.91 suggesting a strong influence of environmental variability on milk composition. Milk composition did not affect lamb survival to weaning. Using joint models in ecological, physiological or behavioural contexts has the major advantage of decomposing a (co)variance/correlation matrix while being estimated with fewer parameters than in a “traditional” mixed‐effects model. The joint models presented here complement a growing list of tools to analyse correlations at different hierarchical levels separately and may thus represent a partial solution to the conundrum of physiological complexity.

Sources of variation in maternal allocation in a long‐lived mammal

“…We suggest that this increase in mass transfer efficiency is likely due to selective disappearance of poor-quality mothers from the breeding population. Individual repeatability in milk production and composition has been demonstrated in wild mammals (Lang, Iverson, & Bowen, 2009;Renaud, Blanchet, Cohen, & Pelletier, 2019) and in domestic mammals (Bergsma, Kanis, Verstegen, & Knol, 2008;VandeHaar et al, 2016); and consistent, heritable differences in milk production have been linked to differences in fitness (Gilbert et al, 2012;Spurlock, Dekkers, Fernando, Koltes, & Wolc, 2012).…”

Ontogeny shapes individual specialization

“…Alternatively, maternal effects (i.e., maternal genotype or maternal environment) alone can shape the phenotype of offspring. For example, milk quantity or quality [25] can vary among females either due to genetic differences or differences in the environments, leading to greater similarity among all offspring from the same mother (e.g. being smaller or larger in body size), which in turn could cause similarities in trophic position among siblings.…”

“…In addition, maternal effects could shape individual specialization in brown bear offspring. As a potential pathway, milk quantity or quality 32 can vary among females due to genetic differences and/or differences in their home range quality, leading to consistently larger or smaller offspring from the same mother, which in turn could cause similarities in trophic position among siblings. Last, brown bears live a solitary lifestyle except for the period of offspring rearing involving up to three years of maternal care 33 , after which female offspring often settle close to their mother's home range 34 .…”