FORUM Extinction debt in fragmented grasslands: paid or not? Cousins, Sara A.O. Department of Botany and Department of Physical Geography and Quaternary Geology, Stockholm University, Sweden; Fax 146 8164818; E-mail sara.cousins@natgeo.su.se Abstract Fragmentation of grasslands and forests is considered a major threat to biodiversity. In the case of plants, the effect of fragmentation or landscape context is still unclear and published results are divergent. One explanation for this divergence is the slow response of long-lived plants, creating an extinction debt. However, this has not been empirically confirmed. In this study, data were compiled from broad-scale studies of grasslands from throughout the world that relate plant diversity to fragmentation effects. Only seven studies from northern Europe, out of a total 61, gave any information on actual habitat frag- mentation in time and space. In landscapes with 410% grassland remaining, present-day species richness was related to past landscape or habitat pattern. In landscapes with o10% grassland remaining, in contrast, plant spe- cies richness was more related to contemporary landscape or habitat pattern. Studies from landscapes with 410% grassland remaining supported the concept of an extinc- tion debt, while studies from more fragmented landscapes did not provide any evidence of an extinction debt. In order to make generalisations about historical legacies on species diversity in grasslands it is important to consider a range of highly transformed landscapes, and not only landscapes with a high amount of grassland remaining. Keywords: History; Landscape; Literature review; Plants; Species diversity. Fragmentation Effects and the Concept of Extinction Debt Recently, there has been an increasing aware- ness of the importance of considering larger scales, i.e. the landscape scale, when analysing the status and the threats to species diversity. Habitat frag- mentation (i.e. the tendency of habitats to become smaller and more isolated) has been identified as one of the greatest threats to biodiversity. There are ex- tensive reviews available on fragmentation effects (Harrison & Bruna 1999; Debinski & Holt 2000; Fahrig 2003; Ewers & Didham 2006), mostly on animals. These reviews have made it clear that it is difficult to draw general conclusions because study sites may vary in size and heterogeneity, different types of organisms react differently to fragmenta- tion effects, and many studies were not conducted in real landscapes (Harrison & Bruna 1999). For grassland plants, the results are even more con- founding. No relationship has been found between species richness and habitat area in grasslands in Sweden and Estonia (Eriksson et al. 1995; Pa¨rtel & Zobel 1999; Kiviniemi & Eriksson 2002), thus defy- ing spatial ecological theories (MacArthur & Wilson 1967; Hanski 1999). Studies based on old maps and aerial photographs have documented relationships between past landscape patterns and present-day species diversity in grasslands (Pa¨rtel et al. 1999; Norderhaug et al. 2000; Cousins & Eriksson 2002; Lindborg & Eriksson 2004; Gustavsson et al. 2007), and other types of fragmented habitat (Piessens & Hermy 2006; Vellend et al. 2006). The possibility of an extinction debt (Tilman et al. 1994) and an inertia to environmental change (i.e. habitat decline) could explain why past landscapes reflect present species patterns better than contemporary landscape pat- terns. Particularly long-lived plants decline slowly and can form remnant populations that may survive for decades after environmental change (Eriksson 1996; Dahlstro¨ m et al. 2006; Helm et al. 2006; Lind- borg 2007). In contrast, other studies have found no effects of past landscape patterns on present-day plant species patterns in grasslands (Bruun 2000; Adriaens et al. 2006; Cousins et al. 2007; O ¨ ster et al. 2007), providing no evidence of an extinction debt. Furthermore, several authors have argued that there is a threshold for species extinctions in landscapes: when there is between 10–30% of the target habitat remaining (Andre´ n 1994; McIntyre & Hobbs 1999; McIntyre et al. 2000; Cousins et al. 2003; Betts et al. 2007). It might also be possible to apply such a threshold for the time since change. This might weaken the direct spatial threshold effect, particu- larly for long-lived plants with no or slow extinction rates at the beginning of the fragmentation event that Journal of Vegetation Science 20: 3–7, & 2009 International Association for Vegetation Science 3