UBC Undergraduate Research

Can species distribution modelling improve the climate threat assessment of at-risk mosses in Canada? Menchions, Emma; Francis, Isla; Knoblauch, Victoria; Weinhagen, Caleb


British Columbia is a biodiversity “hotspot” for bryophytes, with the highest occurrence of at-risk mosses in Canada. Assessments of the impacts of future climate change on bryophytes-at-risk are based on widely used IUCN methodologies. However, the assessments are generally unsatisfactory because they are typically based on regional scale climate data and generalized qualitative predictions which have limited relevance to the organisms being assessed. There is a significant need for more quantitative and predictive threat evaluations, especially when considering complex threats associated with climate change. The geographic distributions of at-risk mosses, the factors that determine species distributions, and the way in which distributions could respond to future climate change is largely understudied in the province. Bryophytes are environmentally sensitive, often have highly specific niche requirements, and in some cases have limited dispersal capacities. As a result, changes in climate could pose a significant risk by reducing or shifting their current range. The use of species distribution modelling is a promising tool for quantitatively determining and comparing how climate threats could impact the persistence of at-risk moss distributions. Our project aims to model the distribution of the at-risk moss, Bartramia halleriana (Haller’s Apple Moss) in Canada using an ensemble of small models (ESM) species distribution modeling technique to predict areas of climatically suitable habitat at present and under future climate change scenarios of RCP 4.5 and 8.5 from projected 15EGCM climate data for 2085. In Canada, Haller’s Apple Moss is found only in the Rocky Mountains of British Columbia and Alberta. However, it also occurs in several other countries in the northern and southern hemispheres (COSEWIC 2011). The Canadian population displays high microclimatic habitat specificity and its distribution is believed to be constrained by moisture availability and regions of moderated temperatures (COSEWIC 2011). Therefore, we predict that the climatically suitable areas of the future will move 1) westward, as drier climates from the eastern Rockies in Alberta encroach on its current distribution, and 2) northward and upslope as has been observed for other plant species (Chen et al., 2011). Modelling inputs included presences of B. halleriana, pseudo-absences, seven biologically relevant climate variables and topographical data at 1 Arc Second resolution. The ANN and CTA algorithms were used to produce a final ensemble of the individual ESMs. Maps of the model projection under current and future climate scenarios were created to identify areas of potentially suitable habitat. Differences in the areas predicted to be suitable and unsuitable were compared through time. Estimates of current and future range size were used to estimate what listing category the future extent of occurrence area may be classified under within COSEWIC and BC CDC ranking systems. Contrary to some of our predictions, areas determined to be climatically suitable in the future are identified further south in the Rockies and at higher elevations (30-73 m depending on the climate change scenario)). Overall reductions in range size by 56.8% and 64.1% were estimated for the RCP 4.5 and 8.5 scenarios, respectively. The predicted classifications of the range size under COSEWIC and BC CDC systems generally increased, depending on dispersal and climate change scenarios. The models displayed high performance despite challenges, which included limited published microclimatic data, a paucity of information describing species-specific ecological requirements, limited knowledge of dispersal ability, the limited number of known occurrences, and poor coverage of high resolution environmental data. These challenges may constrain the transferability of the process to other threatened species. Finally, recommendations for future study are made for Haller’s Apple Moss and for bryophytes in general. Process documentation and R-scripts are shared in order to guide future modelling efforts for at-risk moss species distributions.

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International

Usage Statistics