Quantification of niche breadth, contraction, and overlap in multispecies assemblages. A case study with allochthonous and autochthonous herbivores in Chilean Patagonia
Quantification and statistical significance of n-dimensional hypervolumes (niche breadth) and overlap among them remains a central goal in ecology. We used a novel methodology to estimate hyperdimensional niche breadth and overlap, and tested hypothesis about ecological niche behaviour depending on community configuration. We applied this method to field data of habitat use in a multispecies herbivorous assemblage in Torres del Paine National Park (with no domestic herbivores) and rangelands in the surroundings, for winter and summer seasons. We predicted that coexistence with allochthonous, potentially competing species, could modify niche parameters of autochthonous ones, provoking niche contraction and segregation. Alternatively, time lapsed since introduction could have been too short, generating then niche overlap among species. We used GDA to determine ecological gradients of habitat use. Secondly, for each species, we estimated a multidimensional non-parametric kernel density function (KDF) on the space determined by GDA axes. Species’ niche breadths (hypervolumes) were estimated as the volume occupied by 95% of each species’ KDF. Niche overlap was estimated as the sum of the probabilities associated to the smaller of the considered species’ specific KDF at each point of the GDA space. We used null-modelling approach to evaluate if observed overlap was different from random.
Results showed that autochthonous species’ niches behave as predicted under a scenario of current competition. Overlap among wild and domestic herbivorous (44.2%) was higher than that among only wild species (30.2%), as expected in a non-evolutionary assemblage. Overlap among wild species was higher in summer (34.3%) than in winter (26.0%), as expected by lower resource availability during winter. The contrary occurred for domestic species (41.2% in summer; 50.6% in winter), as expected by human influence. In particular, we examined the effect on guanaco Lama guanicoe, the most abundant wild herbivore in the study area. During winter, guanaco showed a significant niche contraction when coexisting with domestic herbivores, as compared when allopatry (53.60 vs. 26.91, GDA units; p<0.01). This result suggests that guanaco could modify its habitat use when is in sympatry with domestic herbivores, indicating an unexpected and quick effect of coexistence with competitive species on wild species niche parameters. The method used here to estimate niche hypervolumes breadth and overlap is both of friendly use and easy understanding, and it proves useful for hypothesis-based niche experiments and field works. It avoids problems related to shape of the hypervolume, and allows an usable and operational method for null hypothesis modelling.