HyDiaD: A hybrid species distribution model combining dispersal, multi-habitat suitability, and population dynamics for diadromous species under climate change scenarios

Abstract

Diadromous species are particularly vulnerable to climate change because they utilize both marine and freshwater habitat to complete their life cycles. Dispersal plays an important role in restraining the distribution of plant and animal species, and is a key mechanism to allow diadromous species to adapt to changes in habitat suitability, but it is often not included in species distribution models that explore population trends under climate scenarios. The objective of this study was to develop a model to estimate potential shifts in diadromous populations in the Atlantic area of Europe under two climate change scenarios and multiple global climate models. To address the question of range-shift responses, a hybrid approach for diadromous species distribution (HyDiaD) was developed that incorporated two components: i) statistical static models of habitat suitability describing the influence of environmental factors on species occurrence, and ii) biological processes relevant for the distribution of the species, such as population demography and dispersal dynamics. Hybrid models were developed using a novel approach that incorporated both population and between-catchment dispersal dynamics specific to each species. Occupancy data for diadromous species in a subset of Atlantic Area catchments were first validated by regional experts, and boosted regression trees were applied to estimate habitat suitability within each catchment based on historical physical and climatic environmental predictors from the continental and marine domains. Habitat suitability was then used in a population dynamics model that incorporated between-catchment dispersal and local population growth. Results for different-sized catchments were compared using time series of spawner density and saturation rate, which estimated how much of the available habitat was being utilized. Many of the species-specific values used in HyDiaD were estimated through a survey of diadromous species experts, and group consensus was reached by calculating weighted averages. The HyDiaD model was applied to two shad species (Alosa alosa and A. fallax) to explore population trends projected annually from 1951 to 2100. Projected trends indicated that under XXIst century climate scenarios, habitat suitability is expected to increase for A. fallax, but decrease for A. alosa. Projected trends also indicated an increase in the rate of annual variability for A. alosa, particularly in the southern part of its range. Future studies can utilize the HyDiaD model to explore distribution trends for other diadromous species under climate change scenarios.