• PNAS Teaching Resources Portal
  • Science Sessions: The PNAS Podcast Program

Environmental and geographic variables are effective surrogates for genetic variation in conservation planning

  1. Richard A. Fullera
  1. aSchool of Biological Sciences, The University of Queensland, Brisbane, QLD 4072, Australia;
  2. bSchool of Earth and Environmental Sciences, The University of Queensland, Brisbane, QLD 4072, Australia
  1. Edited by Richard M. Cowling, Nelson Mandela Metropolitan University, Port Elizabeth, South Africa, and approved September 29, 2017 (received for review June 22, 2017)


To protect biodiversity for the long term, nature reserves and other protected areas need to represent a broad range of different genetic types. However, genetic data are expensive and time-consuming to obtain. Here we show that freely available environmental and geographic variables can be used as effective surrogates for genetic data in conservation planning. This means that conservation planners can, with some confidence, design protected area systems to represent intraspecific genetic diversity without investing in expensive programs to obtain and analyze genetic data.


Protected areas buffer species from anthropogenic threats and provide places for the processes that generate and maintain biodiversity to continue. However, genetic variation, the raw material for evolution, is difficult to capture in conservation planning, not least because genetic data require considerable resources to obtain and analyze. Here we show that freely available environmental and geographic distance variables can be highly effective surrogates in conservation planning for representing adaptive and neutral intraspecific genetic variation. We obtained occurrence and genetic data from the IntraBioDiv project for 27 plant species collected over the European Alps using a gridded sampling scheme. For each species, we identified loci that were potentially under selection using outlier loci methods, and mapped their main gradients of adaptive and neutral genetic variation across the grid cells. We then used the cells as planning units to prioritize protected area acquisitions. First, we verified that the spatial patterns of environmental and geographic variation were correlated, respectively, with adaptive and neutral genetic variation. Second, we showed that these surrogates can predict the proportion of genetic variation secured in randomly generated solutions. Finally, we discovered that solutions based only on surrogate information secured substantial amounts of adaptive and neutral genetic variation. Our work paves the way for widespread integration of surrogates for genetic variation into conservation planning.


  • ?1To whom correspondence should be addressed. Email: jeffrey.hanson{at}uqconnect.edu.au.
  • Author contributions: J.O.H., J.R.R., C.R., and R.A.F. designed research; J.O.H., J.R.R., C.R., and R.A.F. performed research; J.O.H. analyzed data; and J.O.H., J.R.R., C.R., and R.A.F. wrote the paper.

  • The authors declare no conflict of interest.

  • This article is a PNAS Direct Submission.

  • Data deposition: The code, data, and results reported in this paper have been deposited in GitHub (10.5281/zenodo.843625).

  • See Commentary on page 12638.

  • This article contains supporting information online at www.danielhellerman.com/lookup/suppl/doi:10.1073/pnas.1711009114/-/DCSupplemental.

Published under the PNAS license.

Online Impact