Repository logo
 

LIFE HISTORY AND POPULATION DYNAMICS OF SABLE ISLAND HORSES

Date

2015-07-31

Journal Title

Journal ISSN

Volume Title

Publisher

ORCID

Type

Degree Level

Masters

Abstract

Individual-level life-history strategies are the rails that guide population dynamics. Due to the difficulty of conducting long-term, individual-based studies, current management practices often focus on estimating and controlling demographic rates with little consideration for the individual-level responses that guide them. This approach cannot account for important factors such as age-specific responses to changes in population density or long-term impacts of conspecific density and resource limitation. As such, population-level approaches may fail to predict age structure or the rate of population growth. Recent studies of mammals and birds have shown that short-term changes in factors such as population density can have lasting impacts on vital rates of individuals. These results highlight the importance of long-term individual-based analyses in understanding population dynamics. However, very few researchers have thus far been able to isolate and study interacting effects of density and resources on life histories apart from processes such as predation, interspecific competition, and management of anthropogenic disturbance. The feral horses (Equus ferus caballus) of Sable Island, Nova Scotia, Canada, exist in a natural though simplified system without predation, human interference, or interspecific competition (they are the island’s only terrestrial mammal, numbering approximately 500 individuals). Here I determined the roles of local conspecific density and an interacting resource gradient in guiding the reproduction and survival of adult female Sable Island horses (2008–2012). I used body condition (estimates of subcutaneous fat) as an indication of resource allocation towards the often conflicting purposes of reproduction and maintenance. Reproduction was best predicted by body condition (reproducing females were in relatively poorer condition) but there was also evidence of density-dependence in reproductive success. Survival was predicted by and positively related to body condition. Survival was also predicted by an interaction between conspecific density and location on the island consistent with expectations of a known east-west resource gradient that occurs on Sable Island (in available water and forage). Greater variability in fitness estimates in resource-poor, eastern Sable Island suggests that regions of low density and resources may be high risk/high reward habitats. Such habitats may be disproportionately avoided by young animals and exploited by senescent animals. All feral horses are descended from domesticated animals and recent work has found evidence of artificially selected life-history traits in unmanaged populations of domestic mammals like cattle, sheep, and horses (e.g., reproducing even at high densities and earlier in life than expected). I therefore attempted to determine if effects of artificial selection existed in the Sable Island population by examining age-based contributions to population growth and the relationship between reproduction (foaling) and female mortality. Perhaps due to the population’s long history of low management (>250 years), I failed to find any strong evidence of artificially selected life-history traits in Sable Island horses. That is, life history trade-offs in survival and reproduction in Sable Island horses were more similar to wild species of large herbivores inhabiting natural environments, than other populations of feral ungulates. My research suggests a rarely documented but fascinating instance of reversal of artificial selection by natural selection for a domesticated species like the horse.

Description

Keywords

ecology, equid, life history, habitat selection, density dependence

Citation

Degree

Master of Science (M.Sc.)

Department

Biology

Program

Biology

Part Of

item.page.relation.ispartofseries

DOI

item.page.identifier.pmid

item.page.identifier.pmcid