Hybrid projections improve prediction of distributional shifts of invasive and native seaweeds under climate change

Abstract

Aim Using correlative species distribution models (SDMs) to predict species' range shifts may have limited predictive power when extrapolating into climatic conditions outside those used to train the models. The inclusion of physiological responses to add mechanistic knowledge can increase the reliability of predictions, but it is seldom applied. The objective of this study was to demonstrate this approach with two coexisting and potentially competing macroalgae in Europe and North America-the native Chondrus crispus and the invasive Grateloupia turuturu-and discuss the implications in broader biological and ecological contexts. Location Worldwide. Time Period 2000-2014 and 2090-2100. Major Taxa Studied Marine macroalgae. Methods We (i) developed SDMs from distributional records aiming to represent the realised niche of C. crispus and G. turuturu; (ii) provided an experimental proxy of thermal fundamental niche by determining growth and survival thermal physiological thresholds (PTs), using it to project the worldwide distribution of the selected macroalgae; and (iii) combined both approaches into a hybrid one derived from the SDM habitat suitability projections post-filtered with the PTs excluding thermally unsuitable areas. Results Hybrid projections detected and corrected extrapolations of SDMs in areas outside the fundamental niche of thermal tolerance of species, with greater restrictions for the native C. crispus due to its lower upper thermal PT compared to the invasive G. turuturu. Both species are expected to experience future distribution changes and move towards the poles. However, because the invasive species presents a wider thermal niche unfilling state, its potential northward range shift at southern latitudes in the Northern Hemisphere will be smaller. Main Conclusions Habitat suitability projections were more realistic and complete with the correction of the SDMs with the PTs into hybrid projections than those obtained from the two approaches independently and could provide an early warning of the potential replacement of the native species by the invasive one at the contracting limits of distribution.