Improving species distribution models to predict climate change impacts: accounting for species traits, weather extremes, landscape dynamics and demography

Natalie J. Briscoe (1), Michael R. Kearney (2), Brendan A. Wintle (3)

1 School of BioSciences, University of Me lbourne, Vi ctoria, Australia, 3010, n b riscoe@unimelb .edu.au, @n jbriscoe

2 School of BioSciences, University of Me lbourne, Vi ctoria, Australia, 3010, mrke @u nimelb.ed u.au, @ecophys

3 School of BioSciences, University of Me lbourne, Vi ctoria, Australia, 3010, b .wintle@unimelb.edu.au

How  climate  change  affects  species  will  depe nd on physiology, landscape, population and dispersal dynamics, and existing stressors such as habitat loss. To generate robust predictions that can form the basis for conservation  actions  we  therefore  need an approach that integrates physiological  constra ints on survival  and  reproduction  with  other  demographic  processes  and  threats.  We  outline  a modelling framework that addresses this gap by integrating biophysical models with population dynamics models – and demonstrate its application using the koala ( Phascolarctos cinerus) as a case study. We first developed a biophysical model that calculates the survival, energy and water requirements of koalas based on daily weather, individual traits and local habitat features. By coupling this information with data on potential energy and water intake, we estimated how climate change will influence vital rates. Biophysical model output was then integrated with a stage -specific population dynamics model to predict how climate change will  interact  with  other  demographi c  processes  and  threats  to impact  population  persistence.  Our innovative  approach  overcomes  the  widely -recognized  short-comings  of  current  static,  correlative approaches to predicting the impacts of climate change and provides insight into the mechanisms t hrough which climate change will influence species persistence. This modelling framework can be used to identify key traits and quantify how trait variation influences survival and reproduction – a strong foundation for considering evolutionary responses. We illustrate how this approach can be used to identify climate refugia that provide suitable habitat and buffer populations from extreme events, as well as to evaluate how future management actions are likely to influence populations.