Spatial overlap of sea ice-associated predators and prey in western Hudson Bay
Description
Our objective was to identify distribution hotspots and spatial relationships among polar bears, Arctic foxes, ringed seals, and bearded seals in the sea ice environment. We hypothesized that both polar bears and Arctic foxes, which face limited interspecific competition, maximize their spatial overlap with resources. We further hypothesized that persistence in space use is low between years, due to the dynamic nature of sea ice. We predicted high spatial overlaps between polar bears and all three other species, a large overlap between Arctic foxes and ringed seals, and a low overlap between Arctic foxes and bearded seals, and between ringed seals and bearded seals reflecting their distinct habitat preferences. Finally, we did not expect high overlap between the yearly hotspots of polar bears and ringed seals based on the temporal dynamics of sea ice. Data: Observation type, lat/long coordinates, and Year necessary to create hotspots. Note, Polar Bear hunting attempt was used to create hotspot "Structure".
Files
Steps to reproduce
We conducted hotspot analyses in R v. 4.2 [70] using RStudio v.2024.12.0.467 [71] for each species pooling years using package sfhotspot v.0.8.0[72], and the Getis-Ord Gi* (Gi*) statistic [73]. The hotspot method detects spatial clusters of high (hotspots) and low (coldspots) values by comparing each observation to its surrounding neighbors. We defined neighbors using Queen’s case contiguity (i.e., grid cells that share at least one corner or edge). The Gi* statistic assigns a z-score, which indicates whether the observed clustering significantly differs from a random spatial distribution, with positive z-scores representing hotspots and negative z-scores indicating cold spots. We considered hotspots at statistical significance level of α ≤ 0.05. To examine spatial relationships among species, we calculated mutual and directional overlaps between hotspots. We imported the hotspots into ArcGIS Pro v. 3.3.0 (ESRI ArcGIS Pro©, Environmental Systems Research Institute Inc., Redland, CA), converted them to polygons, and extracted the contour of significant hotspots. Mutual overlap was calculated for all species pairs, while directional overlap quantified the proportion of Arctic fox hotspots overlapped by polar bear hotspots, of polar bear and seal hotspots overlapped by seal-kill or structure hotspots, and the proportion of seal kill, and structure hotspots overlapped by Arctic fox hotspots. For polar bear tracks (n=3 years) and ringed seals (n=4 years), we obtained enough observations to assess temporal hotspot persistence and examine temporal changes in their spatial relationship (Table 3) by measuring distances between yearly hotspot centroids within each species and between species within each year. To ensure consistency of between-year comparisons, we constrained the analysis to hotspot areas within the minimum convex polygon encompassing the region common to all surveys. We assessed interannual variation in the overlap between polar bear and ringed seal hotspots. Descriptive statistics are provided as mean ± SE. To assess relationships with sea ice, we calculated the centroid coordinates for each species hotspot and measured centroid distances to shore, and quantified hotspot proportion comprising landfast vs. pack ice. We obtained weekly regional ice data for Hudson Bay from the Canadian Ice Service Digital Archives (https://iceweb1.cis.ec.gc.ca/Archive/page1.xhtml; 2024). To delineate landfast ice, we selected continuous polygons with a 10/10th ice concentration that were attached to the shore [74]. We calculated the maximum extent of landfast ice for each considered period (years pooled or yearly) and quantified the overlap between landfast ice and species' hotspots. For context, we calculated in ArcGIS Pro v. 3.3.0 the mean percent of the study area covered by landfast ice, and calculated the maximum distance between landfast edge and the coast.
Institutions
- University of Alberta
- University of Manitoba
- Environment and Climate Change Canada
- San Diego Zoo Institute for Conservation Research
Categories
Funders
- Polar Bears International
- MITACS
- Banrock Station Environmental Trust
- Canadian Association of Zoos and Aquariums
- Hauser Bears
- Zuest Family Foundation
- Manitoba Agriculture and Resource Development
- Polar Continental Shelf Program
- Quark Expeditions
- San Diego Zoo Wildlife Alliance
- World Wildlife Fund Canada