Retrospective studies of marine ecosystem variability in the North Pacific Ocean, conducted largely under the auspices of the North Pacific Marine Science Organization (PICES) and its Climate Change & Carrying Capacity Program, have identified rapid, coincident, and persistent changes in both climate and marine ecosystems. Although the initial studies in this field focused on the coincident and dramatic changes in both fisheries and climate that occurred during the winter of 1976/77, a growing body of evidence suggests that similar kinds of shifts also occurred in the 1920s, 1940s and the late 1980s. The current debate among scientists is whether these observations are generated by low frequency-high amplitude climate cycles, natural interventions (step functions) in the mean state of nature, stochastic processes, or some combination thereof.

To address these questions, a one day symposium on The Nature and Impacts of North Pacific Climate Regime Shifts was sponsored by the PICES Science Board on October 15, 1999 at the Eighth Annual Meeting of PICES in Vladivostok, Russia. The symposium was devoted to: (1) a detailed analysis of events just preceding and the impacts resulting from the 1976/77 event, (2) to the physical mechanisms behind regime shifts, and (3) observational evidence for other regime shifts this century. The symposium attracted 27 abstracts considering topics from ocean physics and climate to plankton and fisheries. Eleven of these contributions included authors from the host country of the Annual Meeting. The descriptions that follow highlight the significant contributions of each paper to the symposium and to our understanding of this aspect of the natural world. Papers focusing on evidence of regime shifts are placed earlier in the volume than those dealing with mechanisms.

The 1976-77 climatic regime shift was widely accepted among biologists in the North Pacific long before the general community of climatologists agreed upon its lasting significance. Hare and Mantua explore this apparent paradox by examining 100 different environmental time series, 31 climatic and 69 biological, for the period 1965-1997. Features of the 1976/77 change are reproduced in both physical and biological data sets, and evidence is found for a 1989 shift as well. The latter is better expressed in the biological records, and the authors conclude that monitoring North Pacific and Bering Sea ecosystems may allow for an earlier identification of regime shifts than is possible from climate data alone. This is largely due to the striking expression of regime shifts on the biology of the North Pacific. McFarlane et al. make use of this biological magnification to question whether there has been a change in climate since 1976-1977. On the basis of observed changes in salmon, hake and groundfish in British Columbia, they suggest that another regime shift occurred in 1989.

While much of the interest in the 1976/77 regime shift has been in the eastern North Pacific, the effects were basin-wide. Zhang and his colleagues show what transpired in Korean waters. Both physical and biological changes were apparent after both 1976 and 1988. Biomass and production of saury decreased after 1976, while those of sardine and filefish increased. After 1988, recruitment, biomass, and production of sardine collapsed while those of mackerel substantially increased. The paper discusses plausible hypotheses, linking oceanographic and biological processes, for these changes.

Park and Oh have confirmed that a significant warming occurred in the East Asian Marginal Seas in the 1960s. They further suggest that this warming is related to SST changes in the western equatorial Pacific rather than the SST changes in the central Equatorial Pacific (Nino3,4 region). Their results suggest that the teleconnection between the topics and mid-latitudes should be considered in more detail especially in the western parts of the Pacific basin.

Kang et al. examined physical and biological time series for evidence of climatic impacts on the ecosystems of the Korean peninsula. Many of the biological series, including chlorophyll a, zooplankton biomass, tree ring width and fish catches showed a response the climate regime shift of 1976/77. Tree ring growth was also highly correlated with precipitation with extreme growth occurring in El Niño years between 1969 and 1987. The mixed layer depth off the east coast of Korea was deeper and more variable after the 1976-77 regime shift resulting in a large change in the composition of fish catches.

Suga et al. conducted a detailed analysis of the salinity changes of North Pacific Tropical Water (NPTW) associated with the 1970s regime shift, in addition to an improved documentation of the climatological mean state of NPTW. NPTW, which is characterized by the subsurface salinity maximum, is formed roughly in the region 20-30°N and 140°E-140°W, then advects toward the west. NPTW occupies 10-23°N, 100-200 m depths in the southwestern North Pacific. Repeated observation along 137°E successfully captured salinity increases of NPTW. Suga et al. suggested that the salinification is caused by increased formation rate, and not by precipitation and evaporation in the formation region. Their conclusion indicates that ocean dynamics play an important role in salinity changes.

Yasuda and Noto analyzed SST changes and their mechanisms in the Kuroshio Extension region in the 1980s. They found that a significant mixed layer shoaling occurred in the mid 1980s which preceded the 1988/89 basin scale regime shift by a few years. Furthermore, they showed that the mid-1980s change was caused by the horizontal transport divergence of the Kuroshio Extension, indicating the importance of the ocean dynamics. It is suggested that the Kuroshio Current system plays major roles in forming changes in mixed layer depth and SST in the mid- to late-1980s in the Kuroshio Extension region and the subsequent basin-scale climate regime shift in the late 1980s.

Savelieva et al. examined variability in the Siberian High in addition to the Aleutian Low and considered the impact on precipitation and river discharge over Siberia. They found evidence of a 1970 shift in the activity of the Siberian High preceding the 1977 shift in the Aleutian Low. Following the 1970 shift, winter precipitation and, hence, river discharge, increased over West Siberia while both decreased over East Siberia.

Rogachev documented dramatic changes of water characteristics in the Kuril Islands region in the northwestern North Pacific from 1990-1996. He showed a series of changes in temperature, salinity, Oyashio transport and eddy strength near Boussole Strait. These changes may have been a consequence of large scale atmosphere-ocean change. In particular, Rogachev showed that precipitation in the Okhotsk Sea increased abruptly in 1994, and that this anomalous precipitation is associated with the meridional migration of the precipitation pattern.

Overland et al. brought a new perspective to the debate on mechanisms behind decadal climate variability, suggesting that chaos may play a substantive role. Their conceptual model uses the notion of preferred semi-stable states that never quite repeat, a notion conceptually similar to Lorenz' butterfly strange attractor. They make the argument that the North Pacific climate, as indexed by the PDO, PNA and NPI, show properties that are more characteristics of chaotic behaviour than an oscillator or steady state model. While chaos implies non-predictability, Overland et al. note that the preferred modes of the North Pacific tend to be stable with decadal scale residence times punctuated by rapid transitions between modes.

Observational evidence for decadal scale climate variability in the North Pacific is widespread but complicated to interpret. There is now a general consensus that there are at least two components to the observed decadal variability. Miller and Schneider review the leading theories on forcing mechanisms which include: stochastic atmospheric forcing, atmospheric teleconnections, midlatitude ocean-atmosphere interactions, tropical-extratropical interactions, oceanic teleconnections and intrinsic ocean variability. It is clear that the Aleutian Low plays a large role in upper ocean dynamics, in generating temperature anomalies, in turbulent mixing and Ekman pumping and advection. They conclude with a review of the biological impacts of decadal variability and possible mechanisms linking the physics and biology.

In a study of low frequency variability in the North Pacific, Minobe found evidence for both bidecadal and pentadecadal variations. Major regime shifts in the 1920s, 1940s, and 1970s involved simultaneous phase reversals of these two scales of variation that arise from different physical processes but interact with each other. In Minobe's view, the change in 1988/89 was a minor regime shift, and it is too early to tell whether observed changes in 1998/99 are the start of a major event. If one believes that the next major regime shift will involve simultaneous phase reversals between the bidecadal and pentadecadal variations, prediction with an ambiguity of about ten years should be feasible.

In summary, many authors supported earlier work by finding additional evidence for the 1976/77 change, or finding new ways to look at it. There was an increasing congruence of views about a change in 1989, particularly on the North American side, but again, the mechanism remains obscur. Others found similar persistent changes at other locations and times. Of particular interest at the Vladivostok meeting was some tantalizing new evidence of a major change in 1999, with only its persistence remaining uncertain. This and related topics will be explored further at a major PICES-cosponsored symposium on climate variation and marine ecosystems that was held in March 2000 in La Jolla, California.