Living to the range limit: consumer isotopic variation increases with environmental stress

Abstract

Background: Theoretically, each species' ecological niche is phylogenetically-determined and expressed spatially as the species' range. However, environmental stress gradients may directly or indirectly decrease individual performance, such that the precise process delimiting a species range may not be revealed simply by studying abundance patterns. In the intertidal habitat the vertical ranges of marine species may be constrained by their abilities to tolerate thermal and desiccation stress, which may act directly or indirectly, the latter by limiting the availability of preferred trophic resources. Therefore, we expected individuals at greater shore heights to show greater variation in diet alongside lower indices of physiological condition.

Methods: We sampled the grazing gastropod Echinolittorina peruviana from the desert coastline of northern Chile at three shore heights, across eighteen regionally-representative shores. Stable isotope values (δ13C and δ15N) were extracted from E. peruviana and its putative food resources to estimate Bayesian ellipse area, carbon and nitrogen ranges and diet. Individual physiological condition was tracked by muscle % C and % N.

Results: There was an increase in isotopic variation at high shore levels, where E. peruviana's preferred resource, tide-deposited particulate organic matter (POM), appeared to decrease in dietary contribution, and was expected to be less abundant. Both muscle % C and % N of individuals decreased with height on the shore.

Discussion: Individuals at higher stress levels appear to be less discriminating in diet, likely because of abiotic forcing, which decreases both consumer mobility and the availability of a preferred resource. Abiotic stress might be expected to increase trophic variation in other selective dietary generalist species. Where this coincides with a lower physiological condition may be a direct factor in setting their range limit.

Keywords: Diet; Intertidal; Isotopic niche; Littorinid; Physiological condition; Range limit; Stable isotopes; Stress gradient; Trophic niche width; Trophic variation.

Figure 1. Isotopic variation among Echinolittorina peruviana individuals grouped by relative height on the shore.

(A) Individual values on an δ¹⁵N-δ¹³C biplot, with standard ellipses (equivalent to a bivariate SD), corrected for small sample size (SEAc), plotted as calculated by the SIBER procedure (Jackson et al., 2011). Plot lines and symbols represent shore height (see legend). (B) E. peruviana isotopic niche width (SEA.B) at different shore heights, plotted with 95, 75 and 50% credible intervals and mode as a black point. Macroalgae includes Lessonia nigrescens, Dictyota sp. and Ulva sp. pooled.

Figure 2. Separate (A) carbon and (B) nitrogen isotopic variation of E. peruviana individuals from different shore heights.

Range estimates calculated by a Bayesian implementation of Layman et al. (2007) in the R package SIBER (Jackson et al., 2011). A black dot shows the mode, while boxes represent the 95, 75 and 50% credible intervals.

Figure 3. Comparison of isotopic niche widths (SEA.B) of E. peruviana individuals, grouped by relative height on shore, over Antofagasta Bay (black boxes) and Mejillones Peninsula (white boxes).

This geographical split follows the ecological differences identified over this region by Reddin et al. (2015). Plotted are 95, 75 and 50% credible intervals (boxes) and the mode (black or white dot). Sample sizes per height level for both geographical groups, ‘low’ n = 10, ‘mid’ n = 10, ‘high’ n = 7.

Figure 4. Trend of decreasing (A) % C and (B) % N, and (C) increasing C:N, of E. peruviana muscle tissue with increasing relative shore height.

Shown are the median (dark line), inter-quartile range (box), range (whiskers) and outliers (points).

Figure 5. Contributions of putative resources to E. peruviana diet at high (A and B), middle (C and D) and low (E and F) shore heights.

Contributions estimated by SIAR mixing models (Parnell et al., 2010) run separately for Antofagasta Bay (A, C, E) and Mejillones Peninsula (B, D, F). Plotted are contribution estimates’ 95, 75 and 50% Bayesian credible intervals (boxes), mode values (dots) and probability of differences between putative resource contributions of POM, macroalgae (L. nigrescens, Dictyota sp. and Ulva sp. pooled), and epilithic biofilm.