Trophic specialization on unique resources despite limited niche divergence in a celebrated example of sympatric speciation

Abstract

Trophic niche partitioning is observed in many adaptive radiations and is hypothesized to be a central process underlying species divergence. However, patterns of dietary niche partitioning are inconsistent across radiations and there are few studies of niche partitioning in putative examples of sympatric speciation. Here, we conducted the first quantitative study of dietary niche partitioning using stomach contents and stable isotope analyses in one of the most celebrated examples of sympatric speciation: the cichlid radiation from crater lake Barombi Mbo, Cameroon. We found little evidence for trophic niche partitioning among cichlids, including the nine species coexisting in the narrow littoral zone. Stable isotope analyses supported these conclusions of substantial dietary overlap. Our data, however, did reveal that five of eleven species consume rare dietary items, including freshwater sponge, terrestrial ants, and nocturnal foraging on shrimp. Stomach contents of the spongivore (Pungu maclareni) were 20% freshwater sponge, notable considering that only 0.04% of all fishes consume sponges. Overall, we conclude that cichlid species in lake Barombi Mbo overlap considerably in broad dietary niches-in part due to the large proportion of detritus in the stomach contents of all species-but there is evidence for divergence among species in their diet specializations on unique resources. We speculate that these species may utilize these additional specialized resources during periods of low resource abundance in support of Liem's paradox.

Keywords: Liem’s paradox; adaptive radiation; dietary niche partitioning; dietary specialization; ecological divergence; specialization; sympatric speciation; trophic niche partitioning.

Figure 1:

Dietary profiles of cichlid species from Barombi Mbo by prey item proportion. Each color represents a different dietary component. Bar length is based on the average proportion of each prey item. Representative live photographs of most species reared in a common lab environment are also shown, excluding K. dikume which was never successfully recovered alive.

Figure 2:

Box and whisker plots displaying the proportions of A) shrimp B) ants, C) Trichopteran larvae, D) Corvospongilla sponge, E) plant tissue, and F) fish found in each species’ stomachs. Total sample size was 203 individuals collected in 2010 and 2016 from multiple sites around the lake.

Figure 3:

Non-metric multidimensional scaling (NMDS) ordination of dietary item proportions for nine out of the eleven cichlid species from Barombi Mbo. Ordination is based on Bray-Curtis similarity index (stress = 0.103).

Figure 4:

The diet of large Konia eisentrauti and Pungu maclareni individuals contains a larger proportion of plant tissue and Corvospongilla sponge respectively, while the diets of small K. eisentrauti and P. maclareni individuals contain a larger proportion of detritus. Graphs display the linear relationship between the log(standard length(mm)) of individual fish and the proportion (arcsine-transformed) of A) detritus and B) plant tissue for K. eisentrauti, and C) detritus and D) Corvospongilla sponge for P. maclareni. Points represent individuals, lines represent the predicted results of a LOESS model with proportion of a diet item as the response variable and log(standard length(mm)) as the predictor variable.

Figure 5:

Scatterplots of δ¹³C and δ¹⁵N isotopic values for A) all eleven cichlid species from Barombi Mbo and B) only Stomatepia species. δ¹³C offers insight into ultimate carbon source (littoral vs pelagic) while δ¹⁵N values describe relative trophic position.

Figure 6:

Individuals that are part of specialist groups exhibit a wider range of detritus in their diets and more individuals consume secondary or specialized diet items. Density plots displaying the number of individuals per species consuming a given proportion (arcsine-transformed) of 1) detritus and 2) the second most-consumed item (generalist species) or their specialized food item (specialist species).