Year of Award

2019

Document Type

Thesis

Degree Type

Master of Science (MS)

Degree Name

Fish and Wildlife Biology

Department or School/College

W.A. Franke College of Forestry and Conservation

Committee Chair

Hugh Robinson

Committee Co-chair

Michael Mitchell

Commitee Members

Paul Lukacs, Dave Ausband

Keywords

unmarked, camera trapping, density, cougar, low density

Subject Categories

Population Biology

Abstract

Abundance estimates can inform management policies and are used to address a variety of wildlife research questions, but reliable estimates of abundance can be difficult and expensive to obtain. For low-density, difficult to detect species, such as cougars (Puma concolor), the costs and intensive field effort required to estimate abundance can make working at broad spatial and temporal scales impractical. Remote cameras have proven effective in detecting these species, but the widely applied methods of estimating abundance from remote cameras rely on some portion of the population being marked or uniquely identifiable, limiting their utility to populations with naturally occurring marks and populations that have been collared or tagged. Methods to estimate the abundance of unmarked populations with remote cameras have been proposed, but none have been widely adopted due, in part, to difficulties meeting the model assumptions. I examined the robustness of one model for estimating abundance of unmarked populations, the time-to-event model, to violating assumptions using walk simulations. I also tested the robustness of the time-to-event model to the low sample sizes of species that live at low densities by applying it alongside genetic spatial capture recapture on two populations of cougars (Puma concolor) in Idaho, USA. The time-to-event model is robust to many potential violations of assumptions but biased by incorrectly estimating movement speed and non-random sampling. The time-to-event model can effectively estimate the density of species living at low density and was more precise than and as reliable as genetic spatial capture recapture. Camera based abundance estimates that do not require individual identification, such as the time-to-event model, solve many of the challenges of monitoring low-density, difficult to detect species and make broad scale, multi-species monitoring more feasible.

Download

Share

COinS
 

© Copyright 2019 Kenneth E. Loonam