Climate controls on marine ecosystems and fish populations James E. Overland a, ⁎, Juergen Alheit b , Andrew Bakun c , James W. Hurrell d , David L. Mackas e , Arthur J. Miller f a NOAA/Pacific Marine Environmental Laboratory, 7600 Sand Point Way NE, Seattle WA 98115, USA b Institut für Ostseeforschung Warnemünde Biologie, Seestrasse 15, DE-18119, Warnemünde, Germany c Pew Institute for Ocean Science, RSMAS, University of Miami, 4600 Rickenbacker Causeway, Miami FL 33149, USA d NCAR/Climate and Global Dynamics Division, P.O. Box 3000, Boulder CO 80307, USA e Fisheries and Oceans, Institute of Ocean Sciences, 9860 West Saanich Road, Sidney, B.C., Canada V8L 4B2 f Scripps Institution of Oceanography, SIO-UCSD 0224, La Jolla CA 92093, USA abstract article info Article history: Received 1 May 2007 Received in revised form 27 January 2008 Accepted 2 December 2008 Available online 20 February 2009 Keywords: Large marine ecosystems Climate change Regime shift North Atlantic Oscillation Pacific Decadal Oscillation Red noise This paper discusses large-scale climate variability for several marine ecosystems and suggests types of ecosystem responses to climate change. Our analysis of observations and model results for the Pacific and Atlantic Oceans concludes that most climate variability is accounted for by the combination of intermittent 1–2 year duration events, e.g. the cumulative effect of monthly weather anomalies or the more organized El Niño/La Niña, plus broad-band “red noise” intrinsic variability operating at decadal and longer timescales. While ocean processes such as heat storage and lags due to ocean circulation provide some multi-year memory to the climate system, basic understanding of the mechanisms resulting in observed large decadal variability is lacking and forces the adoption of a “stochastic or red noise” conceptual model of low frequency variability at the present time. Thus we conclude that decadal events with rapid shifts and major departures from climatic means will occur, but their timing cannot be forecast. The responses to climate by biological systems are diverse in character because intervening processes introduce a variety of amplifications, time lags, feedbacks, and non-linearities. Decadal ecosystem variability can involve a variety of climate to ecosystem transfer functions. These can be expected to convert red noise of the physical system to redder (lower frequency) noise of the biological response, but can also convert climatic red noise to more abrupt and discontinuous biological shifts, transient climatic disturbance to prolonged ecosystem recovery, and perhaps transient disturbance to sustained ecosystem regimes. All of these ecosystem response characteristics are likely to be active for at least some locations and time periods, leading to a mix of slow fluctuations, prolonged trends, and step-like changes in ecosystems and fish populations in response to climate change. Climate variables such as temperatures and winds can have strong teleconnections (large spatial covariability) within individual ocean basins, but between-basin teleconnections, and potential climate-driven biological synchrony over several decades, are usually much weaker and a highly intermittent function of the conditions prevailing at the time within the adjoining basins. As noted in the recent IPCC 4th Assessment Report, a warming trend of ocean surface layers and loss of regional sea ice is likely before 2030, due to addition of greenhouse gases. Combined with large continuing natural climate variability, this will stress ecosystems in ways that they have not encountered for at least 100s of years. Published by Elsevier B.V. 1. Introduction In September 2006 a Working Group composed of the authors met at a GLOBEC Workshop on Climate Variability and Marine Ecosystems to discuss and reach consensus on characterizing the spatial and temporal scales of the physical climate forcing of ocean ecosystems, the matching of physical to biological time scales, and where, when and how climate forcing might lead to synchrony of fisheries responses among widely-separated ocean regions. The Group included Pacific and Atlantic climate scientists and those interested in lower trophic level and fisheries responses to climate. Our paper is motivated by the concern of GLOBEC scientists and fisheries managers about the importance of detecting and interpreting apparent oscillations, low frequency red noise, and multiple-state “regime shifts” of climate in controlling the year-to-year and decade- to-decade evolution of regional ecosystems and fisheries. Our goal here is to develop a conceptual model of low frequency variability in the marine climate of the Atlantic and Pacific and relate these changes to responses in large biological populations and associated ecosystems. What follows represents a concept paper rather than a review paper. The workshop was timely in that historical time series are available for a variety of indices describing ocean climate and for a variety of Journal of Marine Systems 79 (2010) 305–315 ⁎ Corresponding author. Tel.: +1 206 526 6795; fax: +1 206 526 6485. E-mail address: james.e.overland@noaa.gov (J.E. Overland). 0924-7963/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.jmarsys.2008.12.009 Contents lists available at ScienceDirect Journal of Marine Systems journal homepage: www.elsevier.com/locate/jmarsys