Keystone Species 444 from those in nearby areas where sea otters had been locally extirpated by hunting In waters surrounding islands with sea otters, kelp forests at shallow depths dominated nearshore subtidal communities, and within these kelp forests sea urchins, an important consumer of kelp, were relatively small and urchin biomass was low (Figure 2) On neighboring islands without sea otters, however, sea urchins were large, urchin biomass was high, and kelps were absent On the basis of these data and the observation that sea otters preferentially feed on sea urchins, Estes and coworkers postulated that sea otters are a keystone species in this community They argued that predation by sea otters reduced sea urchin grazing, thereby maintaining kelp forests and associated species By documenting the ‘‘keystone predator’’ phenomenon in a different ecosystem, this research bolstered the view that such dynamics were potentially widespread Subsequent work supported this hypothesis and expanded our understanding of the conditions facilitating a keystone role for sea otters In a large-scale study published in 1995, Estes and his colleague David Duggins determined the spatial generality of sea otter effects and tested the prediction that after sea otters colonize new areas, sea urchin-dominated subtidal communities would become kelp-dominated (Estes and Duggins, 1995) Surveys in the Aleutians and in southeast Alaska showed that the differences associated with sea otter presence or absence in the earlier studies were general in space – kelps were abundant in the presence of sea otters and scarce in the absence of sea otters Further, considering all surveyed sites, sea urchin and kelp abundance were strongly inversely correlated (Figure 3) Finally, sea urchin abundance declined sharply at sites invaded by sea otters but did not change at sites where sea otter abundance remained constant An interesting regional difference was that rates of increase in kelp abundance in the presence of invading sea otters were higher in southeast Alaska Estes and Duggins suggested that higher rates of kelp increase resulted from differences in the recruitment of sea urchins, which was much greater in the Aleutians than in southeast Alaska (1995) Coupled with sea otter preference for larger, adult urchins, higher recruitment rates of Aleutian Islands Otters absent 20 40 60 Sea urchins per 80 100 20 40 60 80 100 Depth (m) 12 15 18 Kelp density (inds./0.25 m2) Otters present 0.252 20 Amchitka Is Alaid Is Adak Is Nizki Is Shemya Is 10 21 27 Amchitka−Kirilof point 20 40 Amchitka−Bat Island 60 80 100 20 40 60 Percent total vegetation cover 20 40 Sea urchins per 0.25 60 80 100 20 40 60 80 100 Depth (m) Southeast Alaska 80 100 12 15 Kelp density (inds./0.25 m2) 24 Otters absent Otters present Surge bay 20 Sitka sound Torch bay 10 18 21 24 27 Amchitka−Kirilof point 20 40 Shemya 60 80 100 20 40 60 Percent total vegetation cover 80 100 Figure Total cover of macroalgae and sea urchin density by depth at three locations on Amchitka Island – Kirilof Point, Bat Island, and Kirilof Rocks (all with sea otters) – and one at Shemya Island (without sea otters) m, vegetation cover; , sea urchin density Reproduced from Estes JA, Smith NS, and Palmisano JF (1978) Sea otter predation and community organization in western Aleutian Islands, Alaska Ecology 59: 822–833, with permission from ESA 100 300 500 700 900 Sea urchin biomass (g/0.25 m2) Figure Density of kelp (individuals per 0.25 m2) vs estimated sea urchin biomass (gram per 0.25 m2) for the Aleutian Islands and southeast Alaska Points show averages for sites within locations Sea urchin biomass was estimated from samples of population density, size–frequency distribution, and the functional relation between test diameter and wet mass Reproduced from Estes JA and Duggins DO (1995) Sea otters and kelp forests in Alaska – Generality and variation in a community ecological paradigm Ecological Monographs 65: 75–100, with permission from ESA