Takuya Furukawa

Response diversity determines the resilience of ecosystems to environmental change.

A growing body of evidence highlights the importance of biodiversity for ecosystem stability and the maintenance of optimal ecosystem functionality. Conservation measures are thus essential to safeguard the ecosystem services that biodiversity provides and human society needs. Current anthropogenic threats may lead to detrimental (and perhaps irreversible) ecosystem degradation, providing strong motivation to evaluate the response of ecological communities to various anthropogenic pressures. In particular, ecosystem functions that sustain key ecosystem services should be identified and prioritized for conservation action. Traditional diversity measures (e.g. ‘species richness’) may not adequately capture the aspects of biodiversity most relevant to ecosystem stability and functionality, but several new concepts may be more appropriate. These include ‘response diversity’, describing the variation of responses to environmental change among species of a particular community. Response diversity may also be a key determinant of ecosystem resilience in the face of anthropogenic pressures and environmental uncertainty. However, current understanding of response diversity is poor, and we see an urgent need to disentangle the conceptual strands that pervade studies of the relationship between biodiversity and ecosystem functioning. Our review clarifies the links between response diversity and the maintenance of ecosystem functionality by focusing on the insurance hypothesis of biodiversity and the concept of functional redundancy. We provide a conceptual model to describe how loss of response diversity may cause ecosystem degradation through decreased ecosystem resilience. We explicitly explain how response diversity contributes to functional compensation and to spatio‐temporal complementarity among species, leading to long‐term maintenance of ecosystem multifunctionality. Recent quantitative studies suggest that traditional diversity measures may often be uncoupled from measures (such as response diversity) that may be more effective proxies for ecosystem stability and resilience. Certain conclusions and recommendations of earlier studies using these traditional measures as indicators of ecosystem resilience thus may be suspect. We believe that functional ecology perspectives incorporating the effects and responses of diversity are essential for development of management strategies to safeguard (and restore) optimal ecosystem functionality (especially multifunctionality). Our review highlights these issues and we envision our work generating debate around the relationship between biodiversity and ecosystem functionality, and leading to improved conservation priorities and biodiversity management practices that maximize ecosystem resilience in the face of uncertain environmental change.

Time-delayed response of Japanese hare distribution to landscape change along an urban gradient

Urban areas are highly dynamic landscapes and urbanization patterns are spatially complex. For effective wildlife management in urban landscapes, it is essential to consider the complex history of urbanization and its timedelayed effects on species distribution. We examined whether the history of local land use affects the time-delayed response of Japanese hares (Lepus brachyurus) to landscape change in an urban ecosystem by hierarchically reducing the factors affecting differences in the response. We evaluated the occurrence of hares in 62 forest patches along an urban gradient in the Tama Hills area, Japan. Using sites where the surrounding landscape remained relatively stable, we calculated the habitat threshold of hare occurrence based on the surrounding forest area. We extracted sites that used to be suitable for hare occurrence in the past but later on surpassed the identified habitat threshold (i.e., ‘regressive sites’) and analyzed the landscape factors explaining the extinction debt of hares. Finally, we mapped potential extinction debt throughout the study region. We detected possible extinction debt in 4 regressive sites. We also found that forest patch size and the number of surrounding forest patches were related to hare occurrence. Based on this prediction, several other sites with potential extinction debt were identified in the study region. We found that the delayed response of the Japanese hares is likely caused by differences in forest fragmentation processes at each site. Our hierarchical approach could be an effective methodology for detecting and explaining the cause of extinction debt at the local scale and would contribute to the management of urban land for wildlife conservation.

Nestedness-resultant community disassembly process of extinction debt in a highly fragmented semi-natural grassland

There are two major processes of species disassembly after landscape changes: non-random loss of species resulting in nested assemblages and species replacement resulting in spatial species turnover. Although time-lagged responses of species to landscape change have been widely recognized, few studies have empirically evaluated which of these two processes is more closely related to extinction debt (i.e., postponed species extinction following habitat loss). This study aimed to understand the underlying processes of extinction debt by partitioning β-diversity into components of species nestedness and species turnover. We measured grassland species richness at three spatial extents in a highly fragmented semi-natural grassland landscape in Japan. Dissimilarity-based β-diversity was partitioned into two components (i.e., nestedness-resultant dissimilarity [βsne] and turnover-resultant dissimilarity [βsim]), which were further analyzed using principal coordinates analyses (PCoA). The relationships between the variability of PCoA axis 1 scores and the current and past habitat proportions were evaluated. A significant positive relationship between current grassland species richness and past (i.e., the 1910s) grassland proportion was found at the largest spatial extent. The first axis of PCoA based on βsne showed significant correlation with past habitat proportions, whereas the PCoA axis based on βsim showed no significant correlation with either the current or past habitat proportions. A non-random loss of grassland species represented by nestedness underlay the extinction debt found at the landscape level. There is a chance of predicting the loss of species from the nested ranks of species which likely reflects the gradient of species vulnerability to historical landscape changes.