Hierarchical Bayesian model reveals the distributional shifts of Arctic marine mammals

Abstract

Aim Our aim involved developing a method to analyse spatiotemporal distributions of Arctic marine mammals (AMMs) using heterogeneous open source data, such as scientific papers and open repositories. Another aim was to quantitatively estimate the effects of environmental covariates on AMMs’ distributions and to analyse whether their distributions have shifted along with environmental changes. Location Arctic shelf area. The Kara Sea. Methods Our literature search focused on survey data regarding polar bears (Ursus maritimus), Atlantic walruses (Odobenus rosmarus rosmarus) and ringed seals (Phoca hispida). We mapped the data on a grid and built a hierarchical Poisson point process model to analyse species’ densities. The heterogeneous data lacked information on survey intensity and we could model only the relative density of each species. We explained relative densities with environmental covariates and random effects reflecting excess spatiotemporal variation and the unknown, varying sampling effort. The relative density of polar bears was explained also by the relative density of seals. Results The most important covariates explaining AMMs’ relative densities were ice concentration and distance to the coast, and regarding polar bears, also the relative density of seals. The results suggest that due to the decrease in the average ice concentration, the relative densities of polar bears and walruses slightly decreased or stayed constant during the 17‐year‐long study period, whereas seals shifted their distribution from the Eastern to the Western Kara Sea. Main conclusions Point process modelling is a robust methodology to estimate distributions from heterogeneous observations, providing spatially explicit information about ecosystems and thus serves advances for conservation efforts in the Arctic. In a simple trophic system, a distribution model of a top predator benefits from utilizing prey species’ distributions compared to a solely environmental model. The decreasing ice cover seems to have led to changes in AMMs’ distributions in the marginal Arctic region.

Integrating experimental and distribution data to predict future species patterns

“…Moreover, an important assumption behind most of them is that relationships between the observed patterns of environment and species distribution will remain unchanged over the study region and time 6 . Due to non-stationarity of ecosystem processes 11–16 such an assumption seems unrealistic and will likely be violated under future climate conditions, when statistical patterns between current species distributions and the environment are expected to become uncoupled 17,18 . Moreover, under future climate scenarios statistical SDMs are often applied outside the environmental gradient where they have been initially trained 6,10,19,20 in which case the results may become unreliable 21 .…”

“…Although it is widely acknowledged that communities are more than the sum of the parts, most SDMs neglect biological mechanisms despite the fact that species interactions often explain unexpected responses to climate change 13,20,30,31 , and most extinctions attributed to climate change to date have involved altered species interactions 32 . Thus, informing statistical SDMs with the species-specific tolerance limits of locally adapted populations, and including the main species interactions into these models, will significantly increase model realism and improve climate change projections 16,20,33,34 .…”

Bayesian model based spatiotemporal survey designs and partially observed log Gaussian Cox process

“…The water areas are colored according to the inclusion probability at day 165/171 with white corresponding to less than 0.1% / 5% inclusion probability for pike perch and herring, respectively. 2017; Mäkinen and Vanhatalo, 2018). In these applications, collecting data is typically time consuming and expensive.…”

“…The model and motivating example. The main motivation for our work comes from species distribution modeling where LGCPs have received increasing interest in recent years (e.g., Warton and Shepherd, 2010;Chakraborty et al, 2011;Renner and Warton, 2013;Yuan et al, 2017;Mäkinen and Vanhatalo, 2018). We use species distribution modeling as a running example in this work even though the model arises in other applications as well.…”

“…. , n might correspond to, for example, surveyed lattice grid cells (Chakraborty et al, 2011;Mäkinen and Vanhatalo, 2018) or survey sites (Vanhatalo, Hosack and Sweatman, 2017).…”