Sex, size and isotopes: cryptic trophic ecology of an apex predator, the white shark Carcharodon carcharias

Abstract

Demographic differences in resource use are key components of population and species ecology across the animal kingdom. White sharks (Carcharodon carcharias) are migratory, apex predators, which have undergone significant population declines across their range. Understanding their ecology is key to ensuring that management strategies are effective. Here, we carry out the first stable isotope analyses of free-swimming white sharks in South Africa. Biopsies were collected in Gansbaai (34.5805°S, 19.3518°E) between February and July 2015. We used Stable Isotope Bayesian Ellipsis in R and traditional statistical analyses to quantify and compare isotopic niches of male and female sharks of two size classes, and analyse relationships between isotopic values and shark length. Our results reveal cryptic trophic differences between the sexes and life stages. Males, but not females, were inferred to feed in more offshore or westerly habitats as they grow larger, and only males exhibited evidence of an ontogenetic niche shift. Lack of relationship between δ¹³C, δ¹⁵N and female shark length may be caused by females exhibiting multiple migration and foraging strategies, and a greater propensity to travel further north. Sharks < 3 m had much wider, and more diverse niches than sharks > 3 m, drivers of which may include individual dietary specialisation and temporal factors. The differences in migratory and foraging behaviour between sexes, life stages, and individuals will affect their exposure to anthropogenic threats, and should be considered in management strategies.

Fig. 1

a K-means cluster analysis of averaged δ¹³C and δ¹⁵N data for white sharks (Carcharodon carcharias) from the Gansbaai aggregation separated by sex and size category: female sharks < 3 m (closed black circles n = 6), female sharks > 3 m (open grey circles n = 10), male sharks < 3 m (closed black triangles n = 5), and male sharks > 3 m (open grey triangles n = 5). Three clusters were indicated in the analysis (1, 2, 3 demarcated by a dashed line). b Linear regression (y = 0.35x × 19.17, R² = 0.15, P = 0.043) and k-means cluster analysis results of averaged and outlier-removed δ¹³C and δ¹⁵N data; female sharks < 3 m (closed black circles n = 4), female sharks > 3 m (open grey circles n = 10), male sharks < 3 m (closed black triangles n = 3), and male sharks > 3 m (open grey squares n = 5); two clusters were indicated by the analysis (1 and 2, demarcated by a dashed line)

Fig. 2

Relationships between a female length and δ¹³C, b male length and δ¹³C, c female length and δ¹⁵N, and d male length and δ¹⁵N, for white sharks sampled at the Gansbaai aggregation

Fig. 3

Isotopic niches of 22 white sharks sampled at the Gansbaai aggregation a SIBER generated biplots of averaged δ¹³C and δ¹⁵N values with small sample size corrected standard ellipse areas (SEAc) for female sharks < 3 m (closed black circles, solid black line n = 6), female sharks > 3 m (open grey circles, solid grey line n = 10), male sharks < 3 m (closed black triangles, dashed black line n = 5), and male sharks > 3 m (open grey triangles, dashed grey line n = 5). b Credible intervals (95, 75, 50%) of Bayesian estimates of SEAc for averaged δ¹³C and δ¹⁵N values for female sharks < 3 m, female sharks > 3 m, male sharks < 3 m, male sharks > 3 m. c Averaged and outlier-removed δ¹³C and δ¹⁵N values with small sample size corrected standard ellipse areas (SEAc), for female sharks < 3 m (closed black circles, solid black line n = 4), female sharks > 3 m (open grey circles, solid grey line n = 10), male sharks < 3 m (closed black triangles, dashed black line n = 3), and male sharks > 3 m (open grey triangles, dashed grey line n = 5). d Credible intervals (95%, 75%, 50%) of Bayesian estimates of SEAc for averaged and outlier-removed δ¹³C and δ¹⁵N values for female sharks < 3 m, female sharks > 3 m, male sharks < 3 m, male sharks > 3 m