Linkages between stream and forest food webs : Shigeru Nakano ’ s legacy for ecology in Japan

Masashi Murakami Tomakomai Experimental Forest, Hokkaido University Forests, Tomakomai 053-0035, Japan. Food webs, describing the links among consumers and their prey in ecological communities, have a long history of study in ecology dating back to Charles Elton’s research in the 1920s [1,2]. Subsequently, however, experimental work on food webs has focused within habitats on the effects of species interactions (e.g. competition and predation) and physical factors (e.g. disturbance) on community composition and dynamics through time [3]. By contrast, food-web theory largely addressed architectural features of published food webs (connectance, species richness and food-chain length) that influenced system stability in models [4–7]. In the 1990s, a renaissance occurred in research on food webs as better empirical data describing highly resolved food webs emerged from a variety of systems [8,9]. In particular, as ecologists began viewing the world more broadly, they began to consider trophic processes occurring across habitats at the landscape scale, rather than focusing primarily on food-web links within habitats. For example, Gary Polis became intrigued by the role of marine-derived allochthonous organic matter (i.e. sea wrack washed up on beaches) in supporting unusually high densities of spiders on desert islands off the coast of Baja California [10]. Subsequent work showed that these predators caused a trophic cascade, suppressing terrestrial herbivorous insects and thereby indirectly defending island plants [11]. This led Polis et al. [12] to champion the view that food webs are often ‘subsidized’by the movement of materials, energy or organisms across habitat boundaries from adjacent ecosystems. They reasoned that such resource subsidies might strongly affect recipient communities, for example by maintaining high densities of predators that initiate trophic cascades. Recently published research on stream ecosystems by Shigeru Nakano and his colleagues in Japan, who had followed Polis’work closely, built on these ideas and has made major advances at the frontiers of food-web ecology. During the past five years, other ecologists began demonstrating the importance of spatial dynamics of prey movements across the aquatic–terrestrial interface [13–16], but Nakano and his colleagues also became interested in how these fluxes varied through time. Their most recent study demonstrated that reciprocal complementary seasonal shifts in fluxes of prey between a stream and its forested watershed sustained higher densities of consumers in both habitats than would otherwise be supported in either habitat alone [17]. This research broke new conceptual ground, exploring temporal shifts in fluxes and food-web controls across landscape scales, and presenting some of the best data ever amassed on the dynamics and community impacts of cross-habitat subsidies [3]. Tragically, both Gary Polis and Shigeru Nakano, along with another American and two Japanese ecologists, died when their boat capsized in a violent storm in the Sea of Cortez while returning from one of Polis’ island study sites [18–20]. Although Nakano started from relatively humble academic beginnings in Japan, he had risen rapidly to reach a pinnacle in both Japanese and international ecology, which was demonstrated by the interest that world leaders in food-web research such as Polis took in his work. Here, we show the origins of Nakano’s research on spatial and temporal subsidies that link stream and terrestrial food webs, summarize the most recent findings, place them in context with other studies, and consider his legacy for future research in this field.