Targeting climate diversity in conservation planning to build resilience to climate change

Nicole E Heller,Stuart B Weiss,Ryan Branciforte,Elisabeth Micheli,Alan L Flint,Amanda Recinos,Jason Kreitler,Lorraine E Flint,David D Ackerly

doi:10.1890/es14-00313.1

Abstract

Climate change is raising challenging concerns for systematic conservation planning. Are methods based on the current spatial patterns of biodiversity effective given long‐term climate change? Some conservation scientists argue that planning should focus on protecting the abiotic diversity in the landscape, which drives patterns of biological diversity, rather than focusing on the distribution of focal species, which shift in response to climate change. Climate is one important abiotic driver of biodiversity patterns, as different climates host different biological communities and genetic pools. We propose conservation networks that capture the full range of climatic diversity in a region will improve the resilience of biotic communities to climate change compared to networks that do not. In this study we used historical and future hydro‐climate projections from the high resolution Basin Characterization Model to explore the utility of directly targeting climatic diversity in planning. Using the spatial planning tool, Marxan, we designed conservation networks to capture the diversity of climate types, at the regional and sub‐regional scale, and compared them to networks we designed to capture the diversity of vegetation types. By focusing on the Conservation Lands Network (CLN) of the San Francisco Bay Area as a real‐world case study, we compared the potential resilience of networks by examining two factors: the range of climate space captured, and climatic stability to 18 future climates, reflecting different emission scenarios and global climate models. We found that the climate‐based network planned at the sub‐regional scale captured a greater range of climate space and showed higher climatic stability than the vegetation and regional based‐networks. At the same time, differences among network scenarios are small relative to the variance in climate stability across global climate models. Across different projected futures, topographically heterogeneous areas consistently show greater climate stability than homogenous areas. The analysis suggests that utilizing high‐resolution climate and hydrological data in conservation planning improves the likely resilience of biodiversity to climate change. We used these analyses to suggest new conservation priorities for the San Francisco Bay Area.

Highlights

Biological communities are already changing in response to climate change
We examined the climate space captured in each network by plotting the distribution of climatic water deficit (CWD), summer Tmax, and winter Tmin using frequency counts and density plots of occurrences of each value
Comparing the two LU based network scenarios, we found that individual landscape units in the iNetLU were, on average, two and half times more likely to show a greater range in winter Tmin and CWD than in vNetLU, but did not differ for summer Tmax

Summary

Introduction

Biological communities are already changing in response to climate change. Climate is a primary driver of species distributions; as climate warms, populations are shifting in abundance, demographic traits, and distribution (Parmesan 2006, Moritz et al 2008, Bellard et al 2012, Rapacciuolo et al 2014). The protection of diverse physical conditions on the landscape, taking into consideration the effects of topography and geologic structure on the cycling of energy and water, is one approach to climate adaptation planning (Ackerly et al 2010, Anderson and Ferree 2010, Beier and Brost 2010, Schloss et al 2011) This is an alternative or complementary approach to species distribution modeling (SDM), which is commonly used in climate adaptation planning (e.g., Hannah et al 2002, Araujo et al 2005, Hannah et al 2007), yet has some widely acknowledged limitations. SDM is sensitive to variation in algorithms, future scenarios, the spatial resolutions of data inputs, and life history traits (e.g., Araujo et al 2004, Kueppers et al 2005, Elith and Graham 2009, Franklin et al 2013)

Methods

Results

Discussion

Conclusion