Novel tri‐isotope ellipsoid approach reveals dietary variation in sympatric predators

Abstract

Sympatric species may partition resources to reduce competition and facilitate co‐existence. While spatial variation and specialization in feeding strategies may be prevalent among large marine predators, studies have focussed on sharks, birds, and marine mammals. We consider for the first time the isotopic niche partitioning of co‐occurring, teleost reef predators spanning multiple families. Using a novel tri‐isotope ellipsoid approach, we investigate the feeding strategies of seven of these species across an atoll seascape in the Maldives. We demonstrate substantial spatial variation in resource use of all predator populations. Furthermore, within each area, there was evidence of intraspecific variation in feeding behaviors that could not wholly be attributed to individual body size. Assessing species at the population level will mask these intraspecific differences in resource use. Knowledge of resource use is important for predicting how species will respond to environmental change and spatial variation should be considered when investigating trophic diversity.

Improving stable isotope assessments of inter‐ and intra‐species variation in coral reef fish trophic strategies

“…This suggests that using a single tracer, such as only δ 15 N, δ 13 C, or δ 34 S, may hide important dietary variation, and lead to false conclusions regarding shared resources among groups. Multivariate analyses more accurately capture the extent of the isotopic niche space, representing a more comprehensive view of a species' trophic ecology, which is likely to be inherently variable (Nalley, Donahue, & Toonen, 2022 ; Skinner, Mill, et al, 2019 ).…”

“…To extend the isotopic niche to three dimensions and include δ 34 S, Bayesian ellipsoids were fit to 95% of the data (SEV B ) (15,000 iterations, burn‐in 10,000, thin 25) and their median volume was determined (Skinner, Mill, et al, 2019 ). As with the ellipses, SEV B reflects the properties of small sample size corrected ellipsoids (SEV c ) and slightly underestimates SEV when sample sizes are smaller (Skinner, Mill, et al, 2019 ). Ellipsoid overlap among species was calculated based on Bayesian SEV B (7500 iterations, burn‐in 5000, subdivision 4).…”

“…Overlap was expressed as a median percentage where 100% indicates completely overlapping ellipsoids and 0% indicates entirely distinct ellipsoids. When ellipsoid overlap between two species was >60%, it was considered significant (Matley et al, 2017 ; Skinner, Mill, et al, 2019 ). To compare the size of the ellipse areas with the ellipsoid volumes for each species, SEA B and SEV B values and their interquartile ranges (2.5%–97.5%) were mean‐centered and scaled using the base functions in R.…”

“…For example, recent work by Besnard et al ( 2021 ; δ 15 N, δ 13 C, and δ 202 Hg isotopes) and Skinner, Mill, et al ( 2019 ; δ 15 N, δ 13 C, and δ 34 S isotopes) has shown the utility of three‐dimensional SIBER to investigate complex trophic strategies where two dimensions have failed. This three‐dimensional method successfully differentiated among pelagic shark trophic strategies that did not vary significantly by resource use ( δ 15 N and δ 13 C) but were characterized by depth gradients (additional δ 202 Hg isotope; Besnard et al, 2021 ), and revealed within‐species individual specializations in feeding strategies of teleost reef fish that were not detectable using just δ 15 N and δ 13 C (additional δ 34 S isotope; Skinner, Mill, et al, 2019 ). Incorporating additional isotopic information is relatively simple with the advancement of SIA instrumentation, yet examples of its utility across different environments are limited.…”

“…Incorporating additional isotopic information is relatively simple with the advancement of SIA instrumentation, yet examples of its utility across different environments are limited. Moreover, both two‐ and three‐dimensional isotopic niche studies have mostly focused on detecting niche variation among similar species (Espinoza et al, 2019 ; Frisch et al, 2014 , 2016 ; Matich et al, 2011 ; Shipley et al, 2019 ; Skinner, Mill, et al, 2019 ). This is likely due to the seemingly similar functional roles of many species, which raises questions as to their ability to co‐exist and share resources.…”