Range expansion through fragmented landscapes under a variable climate.

Jonathan Bennie,Crispin T R Holloway,Jenny A Hodgson,Chris D Thomas,David B Roy,Tom Brereton,Robert J Wilson,Callum R Lawson

doi:10.1111/ele.12129

Abstract

Ecological responses to climate change may depend on complex patterns of variability in weather and local microclimate that overlay global increases in mean temperature. Here, we show that high-resolution temporal and spatial variability in temperature drives the dynamics of range expansion for an exemplar species, the butterfly Hesperia comma. Using fine-resolution (5 m) models of vegetation surface microclimate, we estimate the thermal suitability of 906 habitat patches at the species' range margin for 27 years. Population and metapopulation models that incorporate this dynamic microclimate surface improve predictions of observed annual changes to population density and patch occupancy dynamics during the species' range expansion from 1982 to 2009. Our findings reveal how fine-scale, short-term environmental variability drives rates and patterns of range expansion through spatially localised, intermittent episodes of expansion and contraction. Incorporating dynamic microclimates can thus improve models of species range shifts at spatial and temporal scales relevant to conservation interventions.

Highlights

Climate change is causing shifts in species distributions (Parmesan & Yohe 2003; Chen et al 2011), but limitations remain in our understanding of how climate-driven changes in demographic rates translate into range shifts under realistic conditions of limited dispersal, landscape heterogeneity and environmental stochasticity
We calibrate the metapopulation model using patch occupancy observations from 1982 to 1991, and empirically test the model using independent data for the subsequent 18 years (1991–2009). We show that both population and metapopulation models including dynamic microclimate variation: (1) quantitatively out-perform models in which spatial and temporal variability in temperature are not included; and (2) capture important features of the episodic and spatially localised patterns of range expansion, that are important for understanding, predicting and managing the responses of species distributions to climate change
The best performing models incorporated the effects of regional climate only, or regional climate interacting with random transect-specific effects, reinforcing the importance of interactions between interannual climatic variability and local habitat characteristics in driving population dynamics in this system

Summary

Introduction

Climate change is causing shifts in species distributions (Parmesan & Yohe 2003; Chen et al 2011), but limitations remain in our understanding of how climate-driven changes in demographic rates translate into range shifts under realistic conditions of limited dispersal, landscape heterogeneity and environmental stochasticity. The environmental conditions that affect the survival and fecundity of individuals vary greatly over fine spatial scales, because of variation in topography and habitat structure (Ashcroft et al 2009; Daly et al 2010; Sears et al 2011), and considerable interannual and decadal fluctuations are superimposed upon long-term natural and anthropogenic climate trends (Karl et al 1995; Canning-Clode et al 2011). The combined effects of short-term, fine-scale variation in climate on population survival and colonisation in fragmented landscapes remain to be incorporated in realistic, empirically tested models of species’ range shifts

Methods

Results

Conclusion