Abstract

Topic: Movements and habitat use

Seasonal variation in moose habitat selection in a Northern Ontario population

Alexander Hamish Bezzina¹, Joseph Northrup², Brent Patterson², Dani Freund³

Trent University, 3024 Rosslyn Road, Thunder Bay Ontario, CA
Ontario Ministry of Natural Resources, Wildlife Research & Monitoring Section Adjunct Professor, Trent University, Environmental and Life Sciences Graduate Program Trent University, DNA Building, 2140 East Bank Drive, Peterborough ON K9L 1Z8
PhD Student - Environmental and Life Sciences Trent University - Ontario, CA

Abstract
Background: In Ontario, Canada, moose (Alces alces) populations have declined substantially over the last 20 years and the underlying causes remain largely unquantified. A better understanding of moose movement and habitat selection are key to uncovering the mechanisms underlying this decline, and for informing management. Further, understanding seasonal variation in habitat selection can help to inform the optimum timing of aerial abundance surveys which assume changes in habitat use occur at a fixed time every winter in Ontario. Step-selection functions (SSFs) are a common framework for understanding habitat selection but typically assume stationarity over time. Objective: The primary objective of this research is to assess whether habitat selection changes across and within seasons for moose and whether these changes vary by sex. Methods: This study site is 2800 km2 of boreal forest surrounding Dog Lake, located 50 km North of Thunder Bay, Ontario. We captured 83 adult moose (21 male and 62 female) via aerial darting during February-March 2024 and December-March 2024-2025. Each moose was fitted with a GPS collar programmed to record hourly fixes. We combined these data with LiDAR data and land cover classifications to create SSFs using the amt package in R with all coefficients varying by individual. We fit models with functional responses on the land cover classifications to assess, in continuous fashion, how habitat selection varied over time. Results: Moose habitat selection was influenced by a combination of terrain features, forest structure, and proximity to roads, with seasonal variation and sex differences observed across coefficients. Functional responses illustrated potential non-linearities in these effects. This research provides valuable insights into the spatiotemporal dynamics of moose habitat selection, which will subsequently be used to better understand seasonal variation in the tradeoffs between predation risk and forage acquisition. It also helps to refine understanding of the timing of population surveys. The findings will inform wildlife management and monitoring strategies and support efforts to address the decline in moose populations in Ontario.