© 2007 South African National Biodiversity Institute DOI: 10.1111/j.1472-4642.2007.00391.xJournal compilation © 2007 Blackwell Publishing Ltd www.blackwellpublishing.com/ddi 645 Diversity and Distributions, (Diversity Distrib.) (2007) 13 , 645–653 BIODIVERSITYRESEARCH ABSTRACT While poleward species migration in response to recent climatic warming is widelydocumented, few studies have examined entire range responses of broadly distributedsessile organisms, including changes on both the trailing (equatorward) and theleading (poleward) range edges. From a detailed population census throughoutthe entire geographical range of Aloe dichotoma Masson, a long-lived Namib Deserttree, together with data from repeat photographs, we present strong evidence that adeveloping range shift in this species is a ‘fingerprint’ of anthropogenic climatechange. This is explained at a high level of statistical significance by populationlevel impacts of observed regional warming and resulting water balance constraints.Generalized linear models suggest that greater mortalities and population declinesin equatorward populations are virtually certainly the result, due to anthropo-genic climate change, of the progressive exceedance of critical climate thresholdsthat are relatively closer to the species’ tolerance limits in equatorward sites.Equatorward population declines are also broadly consistent with bioclimaticallymodelled projections under anticipated anthropogenic climate change but, as yet,there is no evidence of poleward range expansion into the area predicted tobecome suitable in future, despite good evidence for positive populationgrowth trends in poleward populations. This study is among the first to show amarked lag between trailing edge population extinction and leading edge rangeexpansion in a species experiencing anthropogenic climate change impacts, apattern likely to apply to most sessile and poorly dispersed organisms. This providessupport for conservative assumptions of species’ migration rates when modellingclimate change impacts for such species. Aloe dichotoma ’s response to climatechange suggests that desert ecosystems may be more sensitive to climate changethan previously suspected. Keywords Bioclimatic modelling, desertification, extinction, global warming fingerprints, migration, range shift. INTRODUCTION Studies of the responses of terrestrial organisms to twentiethcentury warming have focused on range shifts of motileorganisms (e.g. Hersteinsson & Macdonald, 1992; Parmesan et al ., 1999; Warren et al ., 2001; Parmesan & Yohe, 2003),stressing poleward range boundary extensions (Parmesan et al .,1999). The studies of sessile organisms (Keeling et al ., 1996;Menzel & Fabian, 1999; Abu-Asab et al ., 2001; Fitter & Fitter,2002) have focused on non-lethal changes in the growthpatterns. Both types of studies have focused most frequently ononly a portion of a species’ range.There is concern that the adaptive responses of sessileorganisms to rapid climate change may be constrained, thuscausing population extinctions at the so-called ‘trailing edge’ ofspecies’ geographical ranges that are shifting in response to a 1 Global Change and Biodiversity Program, 2 Protea Atlas Project, South African National Biodiversity Institute, Private Bag X7, Claremont 7735, South Africa; 3 Threatened Species Programme, South African National Biodiversity Institute, Private Bag X101, Pretoria 0001, South Africa; 4 Percy FitzPatrick Institute of African Ornithology, 5 Department of Botany, 6 Leslie Hill Institute of Plant Conservation, University of Cape Town, Private Bag X3, Rondebosch 7701, South Africa; 7 Avian Demography Unit, Department of Statistical Sciences, University of Cape Town, South Africa, Private Bag X3, Rondebosch, 7701, South Africa; 8 Center for Applied Biodiversity Science, Conservation International, 1919 M St NW, Suite 600, Washington DC 20036, USA, 9 Environment Systems, ADAS, Woodthorne, Wergs Road, Wolverhampton WV6 8TQ, UK, 10 Laboratoire d’Ecologie Alpine, UMR CNRS 5553, Université J Fourier, Grenoble Cedex 9, France *Correspondence: Wendy Foden, Threatened Species Programme, South African National Biodiversity Institute, Private Bag X101, Pretoria 0001, South Africa. E-mail: [email protected] Blackwell Publishing Ltd A changing climate is eroding the geographical range of the Namib Desert tree Aloe through population declines and dispersal lags Wendy Foden 1,3,4 *, Guy F. Midgley 1,8 , Greg Hughes 9 , William J. Bond 5 , Wilfried Thuiller 1,10 , M. Timm Hoffman 6 , Prince Kaleme 4 , Les G. Underhill 7 , Anthony Rebelo 2 and Lee Hannah 8W. Foden et al. © 2007 South African National Biodiversity Institute 646 Diversity and Distributions , 13 , 645–653, Journal compilation © 2007 Blackwell Publishing Ltd changing climate (Davis & Shaw, 2001). Such trailing edgeextinctions should be first detected in terrestrial organisms thatoccupy extreme climatic environments such as deserts, whereclimate-related stresses exert direct control over populationprocesses, especially at range margins (Jordan & Nobel, 1979).It is implicitly assumed by bioclimatic modelling approaches thattrailing edge extinctions will be accompanied by simultaneousleading edge range expansions (Parmesan et al ., 1999), but lagsin population expansions due to limiting rates of dispersal andestablishment (Pitelka, 1997) undermined this assumption,especially for sessile species. Such lags have been observed, forexample, in the difference in responsiveness of alpine plant rangeresponses relative to mobile organisms such as butterflies(Walther et al ., 2002), and predicted for temperate zoneAmerican trees (Iverson et al ., 2004). Such constraints are likelyto squeeze the ranges of non-motile organisms between a zone ofactive population die-back and of constrained expansion,thereby raising their risk of extinction, at least temporarily, evenif climate change trends begin to stabilize.Here we investigate impacts of regional climate change onpopulation die-back in the long-lived, giant Namib Desert tree( Aloe dichotoma Masson), prompted by casual observations ofwidespread mortalities apparently induced by drought (i.e.water balance constraints). We test whether this pattern ofpopulation level mortality is in fact consistent with water balancechanges, and furthermore, whether a spatial pattern consistentwith the expected signal of anthropogenic climate change exists(i.e.