3225 Reports Ecology, 86(12), 2005, pp. 3225–3231  2005 by the Ecological Society of America MARINE-DERIVED NUTRIENTS, COMMERCIAL FISHERIES, AND PRODUCTION OF SALMON AND LAKE ALGAE IN ALASKA DANIEL E. SCHINDLER, 1,4 PETER R. LEAVITT, 2 CURTIS S. BROCK, 2 SUSAN P. JOHNSON, 1 AND PAUL D. QUAY 3 1 School of Aquatic and Fishery Sciences, University of Washington, Box 355020, Seattle, Washington 98195 USA 2 Limnology Laboratory, University of Regina, Regina, Saskatchewan 242 0A2 Canada 3 School of Oceanography, University of Washington, Seattle, Washington 98195 USA Abstract. For decades ecologists have recognized the potential importance of marine- derived nutrients (MDN) deposited in freshwater ecosystems by spawning anadromous salmon. Previous studies have shown that some MDN are retained in freshwater ecosystems. A popular hypothesis linking MDN to salmon population productivity posits that MDN provided by post-spawning mortality of salmon are critical for salmon population dynamics because they enhance prey populations in the freshwater ecosystems used as nursery hab- itats. We tested this hypothesis by reconstructing historical sockeye salmon populations for the last 300 years in Bristol Bay, Alaska. Stable nitrogen isotope chronologies in lake sediments and sockeye catch and escapement histories show that commercial fisheries intercepted about two-thirds of MDN bound for freshwater spawning grounds since about 1900. Reconstruction of lake algal production using fossil pigments shows that this loss of MDN has reduced lake algal productivity to about one-third of its level before commercial fishing. However, contrary to expectation, recent sockeye population sizes (sum of spawning escapement and fishery catch) in the last century were equivalent to those before the advent of commercial fishing. These data demonstrate that the MDN subsidy is important for the productivity of coastal lakes but that some sockeye salmon populations are limited by other features of ecosystems such as the amount of suitable spawning habitat. Key words: biogeochemistry; marine-derived nutrients; mixing model; 15 N; paleolimnology; pri- mary production; salmon; salmon enhancement; sediments; stable isotopes; sustainability. INTRODUCTION The importance of wide-ranging organisms for con- necting spatially isolated ecosystems has been increas- ingly recognized in nearly all of the world’s ecosystems (Polis et al. 1997, 2004). In many cases, cross-system fluxes of nutrients and energy are critical for main- taining the productivity, diversity, and community structure of recipient ecosystems. Pacific salmon pro- vide one of the most impressive and widely cited ex- amples of mobile animals linking spatially isolated eco- systems because their migrations move substantial quantities of nutrients and energy into coastal fresh- water ecosystems where salmon spawn and then die (Gende et al. 2002, Naiman et al. 2002, Schindler et al. 2003). Because 95% of salmon growth is accu- mulated during the marine phase of their life cycle, most of the nutrients and energy deposited from spawn- ing populations is derived from marine ecosystems. Thus, populations of spawning salmon produce a sub- sidy of nutrients and energy to the freshwater ecosys- tems used as spawning and nursery habitats. The importance of marine-derived nutrients (MDN) for the productivity of freshwater ecosystems and of Manuscript received 19 November 2004; revised 13 April 2005; accepted 16 June 2005; final version received 29 July 2005. Corresponding Editor: C. R. Hupp. 4 E-mail: deschind@u.washington.edu salmon populations themselves has been speculated about for decades. Mass-balance studies show that MDN can represent a major nutrient source to nursery ecosystems with dense salmon populations (Donaldson 1967, Krohkin 1967, Moore and Schindler 2004). More recently, stable-isotope analyses of the biota in eco- systems used by spawning salmon demonstrate that substantial amounts of salmon-derived nutrients are re- tained in recipient freshwater and riparian ecosystems (Gende et al. 2002, Naiman et al. 2002). These lines of evidence have bolstered the hypothesis that the pro- ductivity of freshwater coastal ecosystems is intimately dependent on the seasonal pulse of MDN deposited by anadromous salmon populations. Because many sal- mon spend a large fraction of their lives in freshwaters, this hypothesis has been extended to suggest that the productivity of salmon populations is ultimately de- pendent on the MDN deposited in nursery habitats by previous generations of salmon (Stockner and Mac- Isaac 1996, Schmidt et al. 1998, Finney et al. 2000, Stockner 2003). Despite having only circumstantial support for this hypothesis, it has become a major focus for salmon management and recovery efforts of im- pacted populations in the Pacific Northwest (Stockner and MacIsaac 1996, Stockner 2003). Here we use paleolimnological time series and long- term fisheries records from the 20th century to test the hypothesis that fishery interceptions of sockeye salmon