Ecology and Management of Mesocarnivores in New Hampshire and Beyond

Date of Award

Spring 2026

Abstract

Mesocarnivores play critical roles in ecosystems by maintaining structure, controlling rodents, and serving as sentinels of global change, yet effective monitoring and ecological understanding of many species remains lacking. My dissertation addresses this gap through four interconnected chapters spanning local, regional, and national scales. In Chapter 1, I surveyed furbearer biologists across the United States to document how monitoring programs have evolved over the past four decades. Camera traps were the most widely adopted contemporary method (80.6% of states), yet biologists rated them among the least influential for management decisions—a "Methods Paradox" in which harvest-based methods retain primacy despite declining use. Funding and staffing constraints emerged as the dominant barriers to program improvement. Better integrating camera trap data into management frameworks, potentially by calibrating relative abundance indices against independent density estimates, could enhance the utility of these increasingly common datasets. In Chapter 2, I evaluated four common metrics—occupancy, local abundance, relative abundance, and density—using camera trap data from 75 New Hampshire sites. Density and relative abundance were strongly correlated, while occupancy and density showed the weakest relationship. Habitat associations varied across metrics such that a significant finding in one rarely held in another, undermining the common practice of using occupancy as a density proxy. In Chapter 3, I used dynamic occupancy modeling to examine top-down, bottom-up, and abiotic influences on red fox winter distribution across 197 camera trap sites spanning a ~60,000 km² landscape in Maine. Contrary to expectation, coyotes associated positively with red fox occurrence rather than suppressing it, and snow depth and forest cover emerged as the dominant drivers—suggesting habitat filtering and abiotic limitation outweigh interference competition in winter. In Chapter 4, I integrated abundance-based multispecies modeling with simple slope analysis across 250 New Hampshire camera trap sites and 45 species pairs. Over one-third of pair-context cases were context dependent, with dampening the dominant outcome. Context dependency was strongly pair-specific and unpredictable from body mass ratio or dietary overlap, and my slope classification framework detected twice as many context-dependent associations as conventional significance testing. Together, these chapters expose a recurring gap between the frameworks most commonly applied to mesocarnivores and the complexity those frameworks fail to capture. The methods developed here are transferable across ecosystems and taxa, with direct applications to harvest regulation and conservation monitoring, and offer a more empirically grounded foundation for mesocarnivore management in a rapidly changing world.

Document Type

Dissertation

First Advisor

Remington J Moll

Department or Program

Natural Resources and Environmental Studies

Degree Name

Doctor of Philosophy

This document is currently not available here.

Share

COinS