Behavioural heterogeneity across killer whale social units in their response to feeding opportunities from fisheries

Abstract

Intra-population heterogeneity in the behavioural response of predators to changes in prey availability caused by human activities can have major evolutionary implications. Among these activities, fisheries, while extracting resources, also provide new feeding opportunities for marine top predators. However, heterogeneity in the extent to which individuals have responded to these opportunities within populations is poorly understood. Here, we used 18 years of photo-identification data paired with statistical models to assess variation in the way killer whale social units within a subantarctic population (Crozet Islands) interact with fisheries to feed on fish caught on fishing gear (i.e., depredation behaviour). Our results indicate large heterogeneity in both the spatial and temporal extents of depredation across social units. While some frequently depredated on fishery catches over large areas, others sporadically did so and in small areas consistently over the years. These findings suggest that killer whale social units are exposed to varying levels of impacts of depredation, both negative (potential retaliation from fishers) and positive (food provisioning), on their life history traits, and may explain the contrasted demographic patterns observed within the declining population at Crozet but also potentially within the many other killer whale populations documented depredating on fisheries catches worldwide.

Keywords: Orcinus orca; depredation; human‐wildlife conflicts; intra‐population variation; marine top predator; social network.

FIGURE 1

Network graph showing the associations among the 79 killer whales of the Crozet Islands used for the study between 2005 and 2022. Individuals are represented by nodes (coloured rectangles) and associations by edges (lines) between nodes. Colours represent social units and edges are weighted by the simple ratio index (SRI). The alpha‐numeric codes used to identify individuals and social units are from the photo‐identification catalogue of the population (Tixier, Gasco, et al., 2021). The graph was laid out using the ForceAtlas2 algorithm in R (Jacomy et al., 2014).

FIGURE 2

The spatial range of the depredation events during which killer whale social units were sighted (red points) estimated by minimum convex polygons (MCPs), around the Crozet Islands between 2005 and 2022.

FIGURE 3

Proportion of the fishing area as estimated from minimum convex polygons (MCPs) in which social units were sighted during depredation events around the Crozet Islands between 2005 and 2022. (a) Annual mean proportion (% ± Standard Error (SE)) of the fishing area over which each social unit was sighted during depredation events; (b) cumulative proportion of the fishing area over which each social unit was sighted during depredation events over the years following the year social units were first sighted during a depredation event.

FIGURE 4

Proportion of the fishing area as estimated from the number of spatial cells in which social units were sighted during depredation events around the Crozet Islands between 2005 and 2022. Sightings were gridded in 0.1° × 0.1° spatial cells. (a) Annual mean proportion (% ± standard error (SE)) of the fishing area over which each social unit was sighted during depredation events; (b) cumulative proportion of the fishing area over which each social unit was sighted during depredation events over the years following the year social units were first sighted during a depredation event.

FIGURE 5

The relative probability of each of the 17 killer whale social units of the Crozet Islands to be present during depredation events between 2005 and 2022, as estimated by a GLMM fitted to the occurrence of killer whale social units during depredation events. Error bars are 95% confidence intervals.