Habitat selection and spatial personalities of Belizean white-tailed deer

Abstract

Understanding movement ecology – broadly defined as the study of how environmental conditions shape animal movement – can provide inference into the environmental traits required to sustain viable wildlife populations. Therefore, movement models have emerged as important tools for wildlife management and conservation. However, consistent individual variation in movement behavior (e.g., habitat selection) – referred to as spatial personality – may have important ramifications for determining population size and structure yet remains understudied in the field of wildlife ecology and conservation despite being embraced by ethologists and evolutionary biologists. Conservation efforts are of critical importance in the Neotropics – one of the most threatened and biodiverse regions of the world – primarily due to habitat loss and other anthropogenic impacts. Deer (family Cervidae) in the Neotropics act as keystone species because of their importance as prey for threatened and endangered large carnivores, as well as the ecological role they play in shaping plant communities and nutrient cycling. Nevertheless, deer and other ungulates are highly understudied across most of the Neotropics. To advance understanding of the general and movement ecology of deer in the Neotropics, we used a white-tailed deer (Odocoileus virginianus nemoralis)-felid (family Felidae) system in northwestern Belize to study their habitat selection and spatial personality. We live-captured and GPS-collared 13 white-tailed deer to obtain fine-scale movement data (2-hour fix rate). We combined GPS collar data with remotely sensed landscape data and predator activity using camera traps to quantify how the environment shapes deer movement. We utilized step-selection functions (SSF), principal component analyses (PCA), linear mixed-effect models, and a metric of nestedness to assess population-level habitat selection and individual variation in habitat selection. We found that deer strongly selected for close proximity to grasslands and pastural areas, and that their space use may be influenced by puma (Puma concolor) and human predation risk. Further, we found moderately high levels of spatial personality as evidenced by repeatable habitat selection over time (R= 0.19 – 0.66). Further, we found that behavioral syndromes of deer were moderately nested (median=0.40). Our results indicate that consistent individual variation in habitat selection occurs within an ecologically valuable ungulate population, suggesting that utilizing mean values when interpreting population level SSF coefficients may overlook substantial variation and mislead conservation decision making. Although the average deer in our study area selected strongly for grassland habitat (e.g., pastural land, road clearings, etc.), indicating grasslands help sustain their abundance and population, the substantial individual variation in selection for, or avoidance of, puma risk, habitat heterogeneity, and human infrastructure (i.e., roads and buildings) obfuscates conservation action. Thus, deer conservation efforts in the Neotropics should focus on conserving and restoring grassland systems as well as further research aimed at understanding the potential impacts of spatial personality for conservation decision making.