Year of Award

2022

Document Type

Thesis

Degree Type

Master of Science (MS)

Degree Name

Wildlife Biology

Department or School/College

Wildlife Biology Program

Committee Chair

Mark Hebblewhite

Committee Co-chair

Jedediah Brodie

Commitee Members

Hugh Robinson, Dale Miquelle

Keywords

Camera traps, density estimation, unmarked ungulates, Panthera tigris, snow tracking, Formozov-Malyushev-Pereleshin (FMP)

Subject Categories

Population Biology

Abstract

Efforts to recover endangered carnivore populations are often limited by insufficient populations of prey. When recovering prey populations, estimates of population density are invaluable metrics to monitor recovery efforts. In Russia, wildlife managers use the Formozov-Malyushev-Pereleshin (FMP) snow tracking method to estimate densities of ungulate prey of the Amur tiger (Panthera tigris). Yet, increasing variability in snow conditions and other challenges have limited its reliability. Camera traps offer a promising alternative approach since managers already use cameras to monitor tigers. However, the assumptions and study design necessary to implement capture-recapture models for tigers are different from those needed to implement models for unmarked populations of prey. In Chapter 1, I estimated densities of wild boar (Sus scrofa), red deer (Cervus canadensis ssp. xanthopygus), roe deer (Capreolus pygargus), and sika deer (Cervus nippon) using Random Encounter models (REM), Space-To-Event models (STE), and Time-To-Event models (TTE), then compared these with FMP estimates within Sikhote-Alin Biosphere Zapovednik. Estimates from the STE and FMP were the most similar, though there were challenges implementing the STE to data from motion-trigger cameras. All models detected a >90% decline in wild boar density due to African Swine Fever. Simulations indicated that greater survey effort for all camera-based methods would be required to achieve a coefficient of variation of 20% (an objective set for this study area in 2006). This is likely cost-prohibitive for many conservation programs due to the high costs of randomly deploying many cameras. To examine the influence of study design on detections of ungulate prey, in Chapter 2 I compared relative abundance indices (RAIs) of prey using: (1) cameras placed on roads to monitor tigers; (2) cameras placed using systematic random sampling; and (3) “off-road” cameras placed 150 meters away from road cameras. Both road and off-road RAIs were greater than random RAIs, and our attempt to approximate representative sampling with off-road cameras ultimately did not work. These results highlight the importance of random sampling to meet the assumptions of unmarked estimators. Detection data of prey species from cameras placed for tiger monitoring should not be used to estimate true abundance of prey species using these models.

Share

COinS
 

© Copyright 2022 Scott Johnston Waller