Cumulative reproductive costs on current reproduction in a wild polytocous mammal.

Svenja B Kroeger,Jane M Reid,Kenneth B Armitage,Julien G A Martin,Daniel T Blumstein

doi:10.1002/ece3.4597

Abstract

The cumulative cost of reproduction hypothesis predicts that reproductive costs accumulate over an individual's reproductive life span. While short‐term costs have been extensively explored, the prevalence of cumulative long‐term costs and the circumstances under which such costs occur alongside or instead of short‐term costs, are far from clear. Indeed, few studies have simultaneously tested for both short‐term and cumulative long‐term reproductive costs in natural populations. Even in mammals, comparatively little is known about cumulative effects of previous reproduction, especially in species with high variation in offspring numbers, where costs could vary among successful reproductive events. Here, we quantify effects of previous short‐term and cumulative long‐term reproduction on current reproduction probability and litter size in wild female yellow‐bellied marmots (Marmota flaviventer) and test how these effects vary with age and between two contrasting environments. We provide evidence for cumulative long‐term effects: females that had both reproduced frequently and weaned large litters on average in previous years had decreased current reproduction probability. We found no evidence for short‐term reproductive costs between reproductive bouts. However, females weaned larger litters when they had weaned larger litters on average in previous years and had lower current reproduction probability when their previous reproductive success was low. Together these results suggest that, alongside persistent among‐individual variation, long‐term reproductive history affects current reproductive success.

Highlights

A central assumption of life‐history theory is that resources are limited, and trade‐offs are expected, where allocation to cur‐ rent reproduction reduces future reproduction and survival (Cody, 1966; Williams, 1966)
To control for potential effects of mother‐daughter competi‐ tion on female current reproduction probability and litter size, we included a fixed effect of the number of sexually mature daughters living in the same colony as the mother in each year (“Mat_daugh‐ ters”)
TA B L E 1 Generalized linear mixed‐effects model estimating effects of previous short‐term reproduction (RLY, reproduced last year) and cumulative long‐term reproduction (PRF, previous reproductive frequency; and previous average litter sizes (PALS), previous average litter size), age, valley, age at first reproduction (AFR), and number of sexually mature daughters living in the same colony (Mat_daughters) on current reproduction probability of female yellow‐bellied marmots

Summary

| INTRODUCTION

A central assumption of life‐history theory is that resources are limited, and trade‐offs are expected, where allocation to cur‐ rent reproduction reduces future reproduction and survival (Cody, 1966; Williams, 1966). Most previous empirical studies that estimated long‐term repro‐ ductive costs were carried out in the context of early‐late life trade‐ offs and the disposable soma theory of senescence, related measures of previous reproductive performance were restricted to a defined time period (“early life,” e.g., up to age 5, willow tits, P. mon‐ tanus, Orell & Belda, 2002; between 3 and 6 years, mountain goats, O. americanus, Panagakis et al, 2017). These studies were testing for a decrease in performance after a defined point in “late life” (i.e., after the defined “early life” period). As reproductive costs may be more likely detected under unfavorable conditions (e.g., Tavecchia et al, 2005), we hy‐ pothesized that reproductive costs are more likely to be observed in the higher elevation environment than the lower elevation environment

| METHODS

Findings

| DISCUSSION