Reproductive seasonality in captive wild ruminants: implications for biogeographical adaptation, photoperiodic control, and life history

Abstract

Many ruminant species show seasonal patterns of reproduction. Causes for this are widely debated, and include adaptations to seasonal availability of resources (with cues either from body condition in more tropical, or from photoperiodism in higher latitude habitats) and/or defence strategies against predators. Conclusions so far are limited to datasets with less than 30 species. Here, we use a dataset on 110 wild ruminant species kept in captivity in temperate-zone zoos to describe their reproductive patterns quantitatively [determining the birth peak breadth (BPB) as the number of days in which 80% of all births occur]; then we link this pattern to various biological characteristics [latitude of origin, mother-young-relationship (hider/follower), proportion of grass in the natural diet (grazer/browser), sexual size dimorphism/mating system], and compare it with reports for free-ranging animals. When comparing taxonomic subgroups, variance in BPB is highly correlated to the minimum, but not the maximum BPB, suggesting that a high BPB (i.e. an aseasonal reproductive pattern) is the plesiomorphic character in ruminants. Globally, latitude of natural origin is highly correlated to the BPB observed in captivity, supporting an overruling impact of photoperiodism on ruminant reproduction. Feeding type has no additional influence; the hider/follower dichotomy, associated with the anti-predator strategy of 'swamping', has additional influence in the subset of African species only. Sexual size dimorphism and mating system are marginally associated with the BPB, potentially indicating a facilitation of polygamy under seasonal conditions. The difference in the calculated Julian date of conception between captive populations and that reported for free-ranging ones corresponds to the one expected if absolute day length was the main trigger in highly seasonal species: calculated day length at the time of conception between free-ranging and captive populations followed a y = x relationship. Only 11 species (all originating from lower latitudes) were considered to change their reproductive pattern distinctively between the wild and captivity, with 10 becoming less seasonal (but not aseasonal) in human care, indicating that seasonality observed in the wild was partly resource-associated. Only one species (Antidorcas marsupialis) became more seasonal in captivity, presumably because resource availability in the wild overrules the innate photoperiodic response. Reproductive seasonality explains additional variance in the body mass-gestation period relationship, with more seasonal species having shorter gestation periods for their body size. We conclude that photoperiodism, and in particular absolute day length, are genetically fixed triggers for reproduction that may be malleable to some extent by body condition, and that plasticity in gestation length is an important facilitator that may partly explain the success of ruminant radiation to high latitudes. Evidence for an anti-predator strategy involving seasonal reproduction is limited to African species. Reproductive seasonality following rainfall patterns may not be an adaptation to give birth in periods of high resource availability but an adaptation to allow conception only at times of good body condition.