Trophic niche changes associated with the eradication of invasive mammals in an insular lizard: an assessment using isotopes

Convergent decoupling of individual specialization and niche width during ecological release

Trophic niche has fundamental ecological importance, but many studies consider few niche metrics and most neglect critical structuring processes. Multiple processes shape trophic niches, including inter and intra-specific competition, predation and resource diversity. These processes interact and effects vary with time and taxa. The White Sands dunefield provides an ecological gradient ideal for understanding variation in niches. We measured population niche width, trophic position and individual specialization in four lizard species across habitats over 2 years. The habitats include White Sands interior, the surrounding desert scrub, and their ecotone. We used arthropod, lizard and plant stable isotopes to quantify niches. We sampled lizard competitors, predators and prey as proxies for ecological processes. We found substantial variation in niches across populations but convergence between species. Individual specialization and population niche width were surprisingly decoupled. Specialization was highest in habitats with low species diversity (White Sands) and population niche width highest at intermediate diversity (ecotone). White Sands lizards may exhibit 'ultra partitioning'; high specialization alongside low individual niche widths. Population niche width is likely constrained within White Sands by low prey diversity. High ecotonal population niche widths may be due to fewer natural enemies than desert scrub but higher resource diversity than White Sands. Trophic position and specialization were positively correlated, suggesting stronger intraspecific competition at higher trophic levels. Prey diversity, inter and intra-specific competition, and predation all interacted to shape niches. Our results highlight the need for measuring multiple components of community structure and niches, as results are likely misleading in isolation.