Tue, Aug 03, 2021:On Demand
Background/Question/Methods
Niche conservatism (NC) is the process by which species retain aspects of their ancestral fundamental niche and related traits. Species distributions in space and time are determined not only by local conditions but a tendency of related species to share environmental tolerances. As a result, species’ phylogenetic relationships have the potential to inform our estimates of their ecological niches and their future in a changing world. The aim of this study is to i) explore scenarios where phylogenetic information elevates model performance in predicting a species niche and, ii) quantify the degree of NC required to detect a significant phylogenetic signal. For a given phylogeny, we simulated niches for synthetic species under 3 continuous trait evolution models with low to high degrees of NC; a white noise (WN) model where simulated niches are independent, a Brownian motion (BM) model where niches of sister species are weakly related and finally an Ornstein-Uhlenbeck (OU) model where niches stabilize at a central tendency across the phylogeny. Using these niche parameters, we generated distribution data and fit them with a baseline single species distribution model (SSDM) and a joint species distribution model (JSDM) that can incorporate a phylogenetic distance matrix.
Results/Conclusions We found the JSDM performed up to 60% better than the SSDM, especially in the case of data-deficient species. This improvement can be attributed partly to the inclusion of a co-occurrence matrix that captures overlap in environmental niche space. Removing the phylogenetic structure from the JSDM did not worsen performance, indicating in most runs, co-occurrence was sufficient in explaining inter-species dependence. However, preliminary results show models predict data-deficient species best under the WN trait model. In the WN case, the phylogenetic distance matrix was able to supplement a weak co-occurrence signal. These first results indicate a complex interplay between the co-occurrence and phylogenetic matrices in borrowing strength across species. Utilizing phylogenetic dependence in species niche and distribution predictions is an area of active development with currently limited understanding of its potential and utility. Through our simulation framework, we demonstrate phylogenetic information can be used in cases where available data limits the use of traditional modelling approaches, e.g. in the cases of species lacking occurrence records or realized niches are artificially constrained by anthropogenic forces.
Results/Conclusions We found the JSDM performed up to 60% better than the SSDM, especially in the case of data-deficient species. This improvement can be attributed partly to the inclusion of a co-occurrence matrix that captures overlap in environmental niche space. Removing the phylogenetic structure from the JSDM did not worsen performance, indicating in most runs, co-occurrence was sufficient in explaining inter-species dependence. However, preliminary results show models predict data-deficient species best under the WN trait model. In the WN case, the phylogenetic distance matrix was able to supplement a weak co-occurrence signal. These first results indicate a complex interplay between the co-occurrence and phylogenetic matrices in borrowing strength across species. Utilizing phylogenetic dependence in species niche and distribution predictions is an area of active development with currently limited understanding of its potential and utility. Through our simulation framework, we demonstrate phylogenetic information can be used in cases where available data limits the use of traditional modelling approaches, e.g. in the cases of species lacking occurrence records or realized niches are artificially constrained by anthropogenic forces.