Decadal-scale Variability in Populations of Small Pelagic Fish

Transcription

1 1 Decadal-scale Variability in Populations of Small Pelagic Fish 2 3 Jürgen Alheit 1, Claude Roy 2 and Souad Kifani Baltic Sea Research Institute, Warnemünde, Germany; juergen.alheit@io-warnemuende.de 2 L Institut de Recherche pour le Développement (IRD), Plouzané, France ; claude.roy@ird.fr 3 Institut National de Recherche Halieutique, Casablanca, Morocco ; kifani@inrh.org.ma Abstract Decadal-scale dynamics of small pelagic fish populations from five large marine boundary currents (Kuroshio, California, Humboldt, Benguela and Canary Currents) and their possible links to climate variability are described and compared. Small pelagic clupeiform fish species such as anchovies, sardines, sardinellas, herring and sprat are characterized by decadal drastic fluctuations of biomass which are often associated with regime shifts in large marine ecosystems and their dynamics are governed by long-term, decadal-scale physical processes. Consequently, small pelagics are excellent indicators of regime shifts. When occurring in the same system, anchovies and sardines usually fluctuate out of phase. These shifts between sardine-dominated and anchovy-dominated states seem to restructure the entire ecosystem, as concomitant qualitative and quantitative changes in ecosystem components other than sardines and anchovy populations have been observed. The small pelagic fish seem thereby to respond to a bottom-up forcing of the ecosystem which itself is driven by changing ocean conditions. Evidence is emerging that these ecosystem shifts are associated with large-scale changes in subsurface processes and basin-scale circulation. In the Humboldt Current ecosystem, the shifts seem to be

2 linked to lasting periods of warm or cold water anomalies related to the approach or retreat of warm oceanic subtropical surface water (SSW) of high salinity to the coast of Peru and Chile. The famous collapse of the Peruvian anchovy around 1970 was the result of such a regime shift, not the consequence of the El Nino 1972/73 which happened after the anchovy decline was already initiated. Dynamics of Japanese sardines and anchovies and their Kuroshio Current ecosystem exhibit a surprising synchrony with processes in the Humboldt system. Although changes in basin-scale circulation must be tightly interwoven with climate variability, the direct association of shifts in the dynamics of small pelagic fish and their ecosystems and climate dynamics and mechanisms linking them, at least in the Pacific, are still largely obscure. The climate regime shift observed in 1976/77 in the North Pacific did not cause clear reactions in the dynamics of small pelagic fish and zooplankton populations of the Pacific boundary currents Contents Introduction 5.2 Decadal-scale regime shifts 5.3 California Curent ecosystem Changes in the biota Regime shifts Mechanisms linking climate to decdal-scale population dynamics 5.4 Kuroshio Current ecosystem Changes in the biota Regime shifts Mechanisms linking climate to decadal-scale population dynamics

3 Humboldt Current ecosystem Changes in the biota Regime shifts Mechanisms linking climate to decadal-scale population dynamics 5.6 Benguela Current ecosystem Changes in the biota Northern Benguela Southern Benguela Regime shifts 5.7 Canary Current ecosystem Changes in the biota Changes in geographic distribution and potential regime shifts Mechanisms linking climate to decadal-scale population dynamics 5.8 Discussion Temperature Trophodynamic aspects Regime shifts and mechanisms linking climate to decadal-scale population dynamics Synchronies and teleconnections 5.9 References Introduction 73

4 Sufficient evidence has been accumulated to show that marine ecosystems undergo decadal-scale fluctuations which seem to be driven by climate variability (e.g. Beamish 1995, Bakun 1996). Climate variability can reorganise marine communities and trophodynamic relationships and can induce regime shifts where the dominating species are replacing each other on decadal time scales. One way to predict how marine ecosystems will react to future climate variability or to climate change is to search for causal relationships of past patterns of natural variability and to draw conclusions based on retrospective studies. Long-term biological time series are essential for retrospective analysis of climate impact on marine ecosystems; however, they are scarce. Fish populations usually provide longer records than other biological components of marine ecosystems because of their economical importance. The dynamics of exploited fish populations are affected by natural environmental variability and man-made activities (fishing, habitat alteration) and retrospective studies will help to distinguish between the two. Although the potential impact of climate variability on marine ecosystems and their fisheries has been described in a number of cases (e.g. Cushing 1982, Laevastu 1993), rigorous studies on these relationships were started only in the 1990s. This was certainly stimulated by the world-wide public awareness of global changes and the predicted greenhouse effect. The initiation of global international research programmes, such as the World Climate Research Programme (WCRP) and the International Geosphere Biosphere Programme (IGBP), vastly improved co-operation across disciplinary boundaries accumulating knowledge on climate variability, particularly on the decadal scale It has been suggested that there exist teleconnections among the low-frequency fluctuations of anchovies and sardines in the Pacific which swing in synchrony (Kawasaki 1983, Chavez et al. 2003) and among Pacific and Atlantic small pelagics which fluctuate antagonistically with each

5 other (Lluch-Belda et al. 1989, Schwartzlose et al. 1999). This chapter will investigate these earlier suggestions and present an update on this debate, now that more than 20 years have passed since the classical paper of Kawasaki (1983) and almost 10 years since Schwartzlose et al. (1999). This chapter is also a contribution to the regime shift debate which has been, and still is, the focus of GLOBEC retrospective studies. We present new views on the existence and causes of regime shifts in ecosystems where small pelagic fish such as anchovies and sardines play an important role, particularly making use of recent insights about regional oceanographic processes Small pelagic fish such as anchovies and sardines are ideal targets for testing the impact of climate variability on marine ecosystems (Box 5.1). This chapter describes interdecadal variability of the large anchovy and sardine populations in the four eastern boundary currents (California, Humboldt, Benguela and Canary Currents) and the Kuroshio Current, a western boundary current. The analysis of each ecosystem is structured according to (i) changes in the biota, (ii) regime shifts and (iii) mechanisms linking climate to decadal-scale population dynamics. The final discussion focuses on tropho-dynamic aspects, climate relationships, synchronous ecosystem dynamics and possible teleconnection patterns Decadal-scale regime shifts Huge populations of sardines and anchovies are dwelling in the upwelling ecosystems of the eastern boundary currents (California, Humboldt, Canary and Benguela Currents) and in the waters around Japan. They support important fisheries, mainly for fish meal, and the well-being of the economy of the riparian countries of upwelling systems depends heavily on these fisheries. The dynamics of these anchovy and sardine populations are characterised by their inverse

6 relationships. When one species supports a large biomass and high production, the other species usually sustains a rather low biomass (Fig. 5.1). The changes in biomass are accompanied by enormous expansions and contractions of the areas of distribution (Fig. 5.2). These shifts between sardine-dominated and anchovy-dominated states seem to restructure the entire ecosystem, as concomitant qualitative and quantitative changes in ecosystem components other than sardines and anchovy populations have been observed. The small pelagic fish species seem thereby to respond to a bottom-up restructuring of the ecosystem which itself is forced by changing ocean conditions as explained below for several ecosystems. Because of their dramatic and long-lasting nature, these switches have been termed regime shifts (Lluch-Belda et al. 1989, 1992). The first use of the term regime was by Isaacs (1976) to describe distinct environmental or climatic states and regime shifts are transitions between different regimes (Lluch-Belda et al. 1989, 1992, MacCall 1996). During a Workshop on Regime Shifts in Villefranche-sur-Mer in April 2003, regime shifts in the marine realm were defined in a pragmatic way as changes in marine system functioning that are relatively abrupt, persistent, occurring at a large spatial scale, observed at different trophic levels and related to climate forcing (deyoung et al. 2004). In contrast to a climate regime shift which might happen within a very short period, an ecological regime shift cannot necessarily be pinpointed to one or two single years. The reason behind this is that marine populations often react with time lags to physical forcing due to complex recruitment processes. A difference in the timing of the shift of different populations can also be expected as different species react differently to climatic forcing depending on their particular physiological threshold values and their life history traits (Beaugrand and Reid, 2003; Beaugrand, 2004). The start and end point (turning points) of regime shifts are defined here as those brief periods where marked changes in ecosystems have been observed. For example, a shift from an anchovy to a sardine period starts when the anchovy stock begins to decrease and/or the sardine stock starts to

7 increase. At this time, at the early stages of a sardine period, the anchovy stock might still have a much larger biomass than the sardine stock, however, it is assumed that the ecosystem has begun a change from a state favourable for anchovy recruitment to a state favourable for sardine recruitment (Jacobson et al. 2001). In other words, the new sardine regime begins when the sardine stock shows the first signs of a longer-lasting increase. Often, what has been considered to be the beginning of a sardine regime is the time when the biomass of sardine surpasses that of anchovies or when negative biomass anomalies change to positive anomalies. However, at those times the sardine regime has usually already been going on for several years. Consequently, in the following, the focus will be on the turning points, i.e. those brief periods when relevant changes in the physics and several trophic levels of an ecosystem have been recorded approximately at the same time (within a period of 1-3 years) The bulk of the data on fish population dynamics stems from the fisheries. Catch data are a somewhat crude measure of fish abundance, however, they usually give an acceptable signal of the trends in population dynamics. Additional data arise from research surveys conducted by fisheries management institutions. Figures of long-term catches and biomass from the upwelling systems are in Chapter 10 (Barange et al. 2007) and are, consequently, not displayed in this chapter. Habitats are described in Chapter 3 (Checkley et al. 2007) California Current ecosystem Changes in the biota 168

8 The California sardine fishery peaked at over mt (metric tons) in 1936 and then decreased rapidly from the mid-1940s to the early 1950s to catches of mt in 1952 (Fig. 5.1; Fig. 10.3a). A fisheries moratorium was established from which was lifted in 1986 because biomass exceeded the minimum of mt. The stock recovered in the early 1980s and annual catches have exceeded mt since 1997 (Lluch-Belda et al. 1989, Schwartzlose et al. 1999, McFarlane et al. 2002). Biomass peaked in 1934 with over 4 million mt and then declined to values of under mt in the 1970s. In 1984, biomass was increasing from unmeasurably low values, rose steadily to 1.5 million mt in 1996, and subsequently declined to about 1 million mt (Fig. 10.3a) (Hill et al. 2006). In the Gulf of California, a sardine fishery started in the late 1960s and caught a peak of mt in when the fishery had fully expanded. It declined rapidly to mt in and increased again to mt in Biomass increased from 1976 to a peak in of 1.2 million mt and decreased subsequently. Recruitment of the Gulf of California population increased after 1975 until and then fell drastically (Schwartzlose et al. 1999). In British Columbia, sardines were the largest fishery from the mid-1920s to the mid-1940s when catches ranged between to mt with an annual average of mt. This fishery collapsed in 1947 and sardines disappeared completely from waters of British Columbia (Fig. 5.2). After 40 years of total absence, sardines were caught again off Vancouver Island in 1992 and have increased in an experimental fishery from 1995 to 1999 (McFarlane and Beamish 2001). Anchovy catches of the central stock increased substantially in the late 1960s, reached a peak in 1981 with mt and declined again with catches being negligible from 1990 on (Fig. 5.1; Fig. 10.2a). Catches off the Pacific coast of Baja California show similar dynamics with substantial increases in the late 1960s, a peak in 1982 with mt and negligible catches starting in Biomass started to increase in late 1960s/early 1970s and fell again substantially after In contrast, in the Gulf of

9 California, anchovy were reported for the first time in The fishery reached a peak in and then declined to zero in 1997 (Schwartzlose et al. 1999) Long-term zooplankton data stem from two time series. The CalCOFI (California Cooperative Oceanic Fisheries Investigations) programme, which was established in 1951 as a response to the decline of the California sardine, has collected zooplankton up to the present using a.505-mmmesh size net. Additional data originate from a series collected off southern Vancouver Island since 1979 (MacCall et al. 2005). Data give strong evidence that zooplankton variability at decadal time scales is intense and coherent over the full width and over alongshore distance >400 km of the California Current ecosystem (CCE) with abrupt transitions between high and low abundances (MacCall et al. 2005). Analysis of CalCOFI samples from southern California (30-35 N) showed a prolonged downward trend in total macrozooplankton biomass as measured by displacement volume from 1951 to the late 1990s by about 80% whereby it is uncertain whether the decline occurred gradually over the entire time series or more rapidly since the 1970s (Roemmich and McGowan 1995) (Fig. 5.3). McGowan et al. (2003) argue that there was a shift in zooplankton biomass around , rather than a continuous decline. The decrease of zooplankton displacement volume was driven by long-term declines of pelagic tunicates, salps in particular (Lavaniegos and Ohman 2003, 2007). There was a major decline of some (but not all) salp species after the mid-1970s and a subsequent increase since 1999 (MacCall et al. 2005). Rebstock (2001, 2002) did not find long-term trends in copepod abundance. In contrast, MacCall et al. (2005) state, based on Rebstock data, that abundance of calanoid copepods increased after 1977 and declined again around 1990 and seems to have increased again in This statement is confirmed by Lavaniegos and Ohman (2007) (using carbon biomass of copepods). However, they also state that they did not find long-term trends of copepods. The solution for

10 understanding these somewhat confusing, contradicting statements is probably that over the long period of 56 years, there was no overall change, but short term fluctuations lasting several years. A comparison of zooplankton phenology (occurrence of seasonal peak of zooplankton biomass) between the periods and revealed that the biomass peak occurred two months earlier in the second period (McGowan et al. 2003) Regime shifts Recent decadal-scale regime shifts in the North Pacific have been described for 1925, 1947, 1977, 1989 and 1998/99 (King 2005). Paleo-ecological studies indicate that such regime shifts might have occurred already for centuries (e.g. Baumgartner et al. 1992, Field et al. 2007). Regime shifts reported since 1960s have been studied relatively well, as data series of ocean sampling are available. Around 1977, dramatic changes in zooplankton, invertebrate and fish populations from around the northern Pacific rim have been reported (Ebbesmeyer et al. 1991, Hare and Mantua 2000, King 2005). It has been argued that these populations responded to a dramatic change of the spatial pattern of atmospheric forcing over the North Pacific basin (e.g. McGowan et al. 2003, King 2005). In the mid-1970s, the Aleutian Low pressure system intensified and shifted southwards in association with substantial changes in the ocean system causing unusually warm, upper-ocean temperatures throughout the northeastern Pacific (Schwing et al. 2005). This pattern is now recognized as the positive phase of the Pacific Decadal Oscillation (PDO) (Hare and Mantua 2000). The PDO is an index which is based on the dominant spatial pattern of SST variation across the North Pacific. The positive phase is characterized by an overall cooling of the central subarctic Pacific and a warming along the coastal northeastern Pacific. This east-west constellation is known as the classic PDO mode. Around 1988, the atmospheric forcing over

11 the North Pacific changed again, when the Aleutian Low pressure system switched to a weakerthan-normal state and a north-south pattern of SST has been observed which was named Victoria pattern (King 2005). However, SSTs remained warm along the west coast of North America. Another North Pacific regime shift has been postulated for 1998 when the Aleutian Low pressure system intensified again. This affected mainly the most southerly regions, i.e. the Central North Pacific and the CC. In the CCE, cooling of coastal waters and enhanced southward flow of water and organisms were recorded (King 2005). The associated deepening of the thermocline and decreased stratification resulted in increasing phytoplankton biomass, both in amount and seaward extent, and higher zooplankton biomass throughout the CCE, whereby the species composition returned to patterns similar to those during the mid-1980s Recently, a debate was initiated about the exact beginning of the 1970s regime shift in the North Pacific, particularly in the CCE. Schwing et al. (2004) challenged the dogma of a Pacific-wide regime shift in and provided evidence that ocean temperatures below the mixed layer and in the northern extremes of the North Pacific began warming around 1970 (Schwing et al. 2005) and did not show a clear shift at the surface in Also, many fishery time series suggest population shifts near They interprete a regime shift as an evolving phenomenon whose signals propagate into different regions, depths and fields having different response times, depending in part on the process that is directly supplying the climate signal. In this context, it is interesting to re-evaluate the assumed rapid change in CCE zooplankton biomass and phenology observed in the 1970s (Roemmich and McGowan 1995, McGowan et al. 2003) which has been presented by many authors as one of the most convincing proofs for the regime shift in the CCE (e.g. Hayward 1997). A closer study of the respective data reveals that CalCOFI zooplankton sampling during the 1970s was only intermittent (Fig. 5.3) (Roemmich and

12 McGowan 1995 Fig. 2, MacCall et al Fig. A2.5), and was too sparse to resolve the timing of changes in zooplankton abundance during the period from 1970 to Also, the more objective analysis of the CalCOFI zooplankton time series done by Lavaniegos and Ohman (2007) does not support the interpretation of an abrupt shift in the mid-1970s. All the above described changes of zooplankton might very well have occurred already around 1970, in association with the decline of sub-surface temperatures as recorded by Schwing et al. (2004), preceding the climatic shift There was a clear sardine regime from the 1920s to the 1940s, very likely associated with the climatic regime shifts in 1925 and However, for the second half of the 20 th century, not much correspondence between the climatic regime shifts in the North Pacific and California sardine dynamics can be found. Catches are not a valid representation of sardine dynamics because of the moratorium up to However, biomass has steadily increased since the early to mid-1980s to 1996 (Fig. 10.3a) (Hill et al. 2006). Anchovy catches started to increase in the late 1960s, began to decrease in 1982 and reached very low levels after 1989 (Fig. 10.2a). Interestingly, during the period of increasing anchovy catches from about 1965 to the early 1980s, sardine biomass was very low. When anchovy catches declined, sardine biomass began to increase. Since the time when anchovy catches were very low, sardine biomass reached rather high levels. Maybe the 1988 climatic regime shift had an impact on both species as anchovy catches declined thereafter, sardine biomass increased strongly after 1990 and sardines occurred again off British Columbia in Also, the increase in anchovy catches in the late 1960s might be related to the subsurface warming around 1970 reported Schwing et al However, the anchovy fishery was more dependent on the price of fish meal than on the availability of fish (Thomson et al. 1985). Thus, the timing of the changes in landings may be misleading (A.

13 MacCall, pers. comm.). Also, the increase in anchovy catches in the late 1960s was strongly influenced by a change in regulations (A. MacCall, pers. comm.) In conclusion, it is doubtful whether drastic changes were exerted on zooplankton (CalCOFI data) and small pelagic fishes in the CCE by the climate shift as characterized by the change of the Aleutian Low. The zooplankton data do not exhibit a corresponding step-like change and the time pattern of sardine dynamics is unclear, as it was not observable due to extremely low abundance during the mid-1970s. Catch data of anchovies and biomass information of sardines in the CCE point to an anchovy regime from the late 1960s to about the early 1980s and a sardine regime with an unclear beginning, maybe from the early 1980s to the present Mechanisms linking climate to decadal-scale population dynamics McGowan et al. (2003) list three explanatory hypotheses for the decline of zooplankton biomass and populations in the CCE in the 1970s: 1) Variation in coastal upwelling intensity, varying the input of deeper, nutrient-rich water to the lighted zone (Bakun 1990) 2) Variations in horizontal input of cooler, fresher, nutrient-rich water from the north (Chelton et al. 1982) 3) Long-term warming and deepening of mixed layer, post 1977, leading to an increase in stratification and resulting in change in nutrient content of waters from below (Roemmich and McGowan 1995). McGowan et al. (2003) suggest that the combined effects of a deeper thermocline (nutricline) and increased stability (increased stratification) - both processes were thought to be a result of

14 the regime shift scenario in association with warming waters - have led to a reduced supply of nutrients to the euphotic zone. This, in turn, was coincident with the zooplankton biomass decrease, as there seemed to be a correlation between thermocline depth (deepening by 17%) and zooplankton biomass (74% decline) due largely to less gelatinous zooplankton. Ocean warming occurred down to a depth of at least 200 m and may be a result of more frequent incursions of water from the Subtropical Gyre (Bograd and Lynn 2003). More recently, Palacios et al. (2004) agree with the mechanisms proposed by McGowan et al. (2003). However, they point out that the thermocline deepened several years earlier. The thermocline depth had a shallow period from 1950 to 1966 and deepened from 1969 to Also, the thermocline was relatively weak between 1950 and 1969 and strengthened afterwards No trophodynamic relationships can be deduced from the long-term CalCOFI zooplankton data and catch records of sardines and anchovies. Both, anchovies and sardines increased during the period of plankton sampling, whereas zooplankton biomass declined steadily. However, the lack of any trophodynamic relation is not surprising considering that (i) the plankton net mesh size was.505 mm (which is too large for major food items of sardines and anchovies) and (ii) the major contribution to the zooplankton decline was due to salps which do not seem to be a relevant food item of sardines and anchovies. Because of their low energy content, the long-term decline in salp biomass is not likely to be of much consequence to planktovorous fish, however, in the Benguela Current, sardine stomachs were found which were packed with salps (H. Verheye, pers. comm.). Unfortunately, there is no long-term information on zooplankton smaller than.505 mm Kuroshio Current ecosystem 336

15 Changes in the biota The sardine population of Japan is distributed in the Sea of Japan and in the waters east of Japan, namely coastal waters, the Kuroshio Current Ecosystem (KCE) and the Kuroshio Current Extension (KC Extension) and the Oyashio Current. Catch records for the Japanese sardine have been kept since at least Yields have been between 0.1 and 0.6 million mt until 1927 (Schwartzlose et al. 1999). Then, mainly based on the Sea of Japan subpopulation, they increased sharply until 1942 with a peak of 1.6 million mt in 1936 (Fig. 5.1). Thereafter, they decreased again to former levels until the late 1950s when they dropped dramatically to 0.02 million mt and less during the 1960s. Sardine catches, mainly from the Pacific subpopulation, started to increase in 1971 and those from the Sea of Japan in Catches reached levels well above 1 million mt from 1976 to 1994 with a peak of 4.5 million mt in 1988 when they started to decrease steadily again. The transition from an anchovy to a sardine regime in the KCE ecosystem in was further indicated by (i) the instantaneous surplus production rate (ISPR) turning positive (Jacobson et al. 2001; Fig. 5a), (ii) the decrease of the economically important post-larval anchovy (shirasu) fishery around 1969 (Fig. 5.4) (Nakata et al. 2000), (iii) the sardine mass spawning in Tosa Bay in 1970 (Kawasaki 1993), (iv) the first appearance of sardine eggs in the Enshu-nada Sea in 1971 (Nakata et al. 2000) and (v) the decrease of summer zooplankton biomass in the Oyashio region, which serves as an important feeding ground for sardines and started to decline around 1972 (Chiba et al. 2006) (Fig. 5.3). Also, new results of fish scale accumulation rates in anaerobic sediments demonstrate that the Japanese anchovy population decreased around 1970 (Poster, M. Kuwae, ASLO Summer Meeting 2006, Victoria). The commercially important squid, Todarodes pacificus, also started to decrease around 1970 (Sakurai et al. 2000). After a decline since the late 1950s, catches of the Pacific saury, Cololabis

16 saira, started to increase again in 1971 (Watanabe et al. 2003). In the mid-1980s, Japanese waters switched back to anchovy dominance. ISPR turned negative (Jacobson et al. 2001). In 1986, anchovy shirasu (post-larvae) catches started to increase again (Fig. 5.4), whereas sardine recruitment began to decline and a dramatic decrease of sardine juveniles was observed. In 1987, anchovy catches began their increase and, in 1988, sardine catches reached their peak and began then to decrease (Fig. 5.1). At the same time, sardine shirasu catches started to decline (Fig. 5.4), whereas anchovy shirasu appeared in 1988 in Suruga bay and Enshu-nada to Ise and Mikawa bays (Kondo 1991). Anchovy increase and sardine decrease in the late 1980s are reflected well by catches from the post-larval fish fishery (Nakata et al. 2000). Also, squid increased again since the late 1980s (Sakurai et al. 2000). Zooplankton biomass in the Oyashio as collected with a 333 micron net increased from the early 1950s to 1970, with a short decline in the early 1960s (Fig. 5.3) (Odate 1994, Chiba et al. 2006). From around 1970 on, it decreased again up to the mid- 1980s and increased again until Biomass of KC Extension copepods > 1mm decreased from 1970, when sampling began, to the early 1990s and then increased again (Nakata and Hidaka 2003) (Fig. 5.3). All these changes are summarized in Table Regime shifts A number of physical processes indicate that substantial changes occurred in the waters east of Japan in the late 1960s which reversed around the mid- to late 1980s, for example changes in the winter mixed-layer depth (MLD) in the KC Extension and in the movement of the Oyashio Current (Yasuda 2003). The MLD in the KC Extension area shifted around 1967 (Yasuda et al. 2000) (Fig. 5.5). The MLD was deeper than the long-term mean between 1967 and 1985 (Yatsu and Kaeriyama 2005). The dynamics of SST in the southern recirculation area of the Kuroshio

17 Extension followed the MLD by a lag of 2-3 years, as they decreased in 1969 and increased again in 1988 (Yasuda et al. 2000, Noto and Yasuda 1999, Yatsu and Kaeriyama 2005). Data on dynamics of sardine and anchovy (biomass, catches, mass appearances of spawn and post-larvae, anchovy scale accumulation rates), squid and zooplankton show turning points at the same time. In addition, sardine recruitment was negative during most years of the 1960s and in most years after 1988 (Yatsu and Kaeriyama 2005). Noto and Yasuda (1999, 2003) report a significant positive correlation between the mortality coefficient of sardine post-larvae to age 1 and the SST of the KC Extension and its southern recirculation area (Fig. 5.6). The high SSTs in this area from and the abrupt increase in SSTs since 1988 (Fig.5.5) were accompanied by low sardine production. Wada and Jacobson (1998) and Yatsu et al. (2005), also, considered conditions in the KCE during the period from as a regime favourable for sardines whereas it was unfavourable during and after The Oyashio front near the island of Honshu has shifted north- and southwards on decadal time scales (Oyashio intrusions) (Yasuda 2003). In the late 1960s, there was a change from northwards shifts to southward shifts until the mid-1980s after which northward shifts dominated again (Yasuda 2003). The Pacific saury fishing grounds are known to be related to this meridional shift of the Oyashio Front (Yasuda and Watanabe 1994). The turning points of Oyashio zooplankton biomass exhibit similar dynamics with a high around 1970 and a low in the late 1980s (Odate 1994, Chiba et al. 2006) (Fig. 5.3). All these data indicate that the ecosystem of the Japanese anchovies and sardines east of Japan experienced a regime shift around 1969 to 1971, when the system changed to sardine dominance, and in the mid- to late 1980s, when it reversed to a state more favourable for anchovies. Clearly, the dynamics of major pelagic populations are associated with these shifts, not with the 1977 climatic shift of the Northeastern Pacific (Yasunaka and Hanawa 2002, Yatsu and Kaeriyama 2005).

18 Mechanisms linking climate to decadal-scale populations dynamics Japanese sardines and anchovies spawn in inshore waters of central and southern Japan and in the KCE. The KCE and the KC Extension are important for recruitment success as they transport larvae to nursery areas in the northeast of Japan into the KC Extension area (Nakata et al. 1994, Noto and Yasuda 1999, Yatsu et al. 2005). The Kuroshio-Oyashio Transition Zone is a nursery and feeding ground for juveniles (Yatsu et al. 2005). Consequently, hydrographical processes in all these areas are potentially of importance to sardines and anchovies Changes in the MLD, which affect the depth and rate of vertical mixing, appear to be an important mechanism through which climatic variations can induce biological changes (Limsakul et al. 2001) as shown for several locations in the central and western North Pacific. Polovina et al. (1995), working with data from the northwest Hawaiian islands, argued that, when the mixed layer deepens, nutrients from subthermocline depths are brought into the mixed layer, thereby increasing phytoplankton production. Yasuda et al. (2000) demonstrated that the winter (Jan-Mar) mixed layer depth (MLD) in the Kuroshio Extension area was particularly deep from the mid- 1960s to the mid-1980s and that there was a drastic change from a deep to a shallow phase between (Fig. 5.5). Limsakul et al. (2001) showed that MLD in the south of the Kuroshio Current, about 450 km south of Shikoku Island, Japan, increased from around 1970 to early/mid-1980s and then decreased again. In the KC Extension, the decadal-scale variation pattern of the January and February MLD is quite similar to winter SST dynamics (Yasuda 2003), both of which started to change in the mid-/late 1960s and reversed in the mid-/late 1980s (Noto and Yasuda 1999, Yasuda et al. 2000), slightly preceding the shifts from anchovy to sardine

19 ( ) and back to anchovy ( ) in the KC Extension. Spring and summer MLD (April - June) in the Oyashio also deepened in 1970 and started shoaling again in 1985 (Chiba et al. 2006). Shortly thereafter, zooplankton biomass began to decline and started to increase again in the late 1980s, a short time after the shallowing of the MLD (Odate1994, Chiba et al. 2006) (Fig. 5.3). Chiba et al. (2006) argue that, in contrast to the impact of a deepening MLD on productivity as reported by Polovina et al. (1995), a deepening MLD in the more northern latitudes of the Oyashio might decrease productivity as this region is not nutrient, but light limited. Interestingly, in the KC Extension, the shoaling of the MLD preceded the abrupt increase in SSTs in which, in turn, preceded the climatic shift (Yasuda et al. 2000, Yasuda 2003). According to Yasuda (2003), variation in the KC could induce the climate regime shift. The increase in the heat transport by the KC Extension may be related to the spin-up of the subtropical gyre (Yasuda et al. 2000, Yasuda and Hanawa 1997). Consequently, Yasuda et al. (2000) suggest that variations in the KC system are a major driving force in the control of the long-term climate variability in the North Pacific. Further, they propose to use the change in MLD to predict dynamics of the Japanese sardine stock Anchovies and sardines in the waters around Japan have been alternating over the last 100 years on a decadal-scale pattern (Fig. 5.1). The mechanisms causing this alternation are largely unknown, however, the timing of the changes (turning points) from sardine to anchovy periods and back to sardines can be determined now rather exactly. At the time of these ecosystem shifts substantial changes in physical and biological variables in the waters off the east coast of Japan have been observed. Evidence is emerging that these ecosystem shifts are associated with largescale changes in subsurface processes and basin-scale circulation. Sardines seem to thrive at periods of reduced biomass of meso-zooplankton in the KC (Nakata and Hidaka 2003) and in the

20 Oyashio Current ecosystems (Odate 1994, Chiba et al. 2006) as collected with 300 micron nets which consists mainly of larger copepods whereas anchovies appear to suffer during these periods Humboldt Current ecosystem Changes in the biota Peruvian anchovy catches of the northern/central stock (Fig. 5.1; Fig. 10.2a) peaked in 1970 at 11 million mt, fell dramatically from 1970 to 1972, remained between 0.5 and 3 million mt until 1982 and decreased to an extremely low level during the early 1980s, also in response to the El Niño of In 1984, the stock recovered and catches rose to 3 million mt in Catches increased steadily thereafter to a peak of 9.8 million mt in 1994, dropped to 1 million mt in 1998 because of another strong El Niño and have maintained at 6-9 million mt since (Alheit and Niquen 2004). Sardine spawning (Zuta et al. 1983) and catches (Serra 1983) were insignificant during the 1950s and 1960s. From 1964 to 1971, the only distinct spawning areas were in northern Peru and northern Chile (Bernal et al. 1983). After 1971, the sardine expanded to the northern and southern extremities of both refuge areas. Sardine spawning off Peru from 1966 to 1968 was poor and limited to the region between 6 and 10 (Zuta et al. 1983). After 1969, an increase of spawning was observed and, after the El Niño event of , sardine spawning increased strongly and the spawning area expanded considerably (Fig. 5.2). From 1973 on, the area of distribution and the abundance of sardines increased notably in Ecuador, Peru and Chile (Zuzunaga 1985). Between 1976 and 1980, spawning increased further (Zuta et al. 1983). From 1964 to 1973, no eggs or larvae were observed in Chile south of 25. However, a new spawning

21 area off Talcahuano was established subsequently (Fig. 5.2) (Bernal et al. 1983, Serra 1983). Sardine spawning increased and the geographic distribution of spawning expanded during warm years. Sardine catches of Peru and Chile increased steadily from less than mt in 1970 to more than 3 million mt in 1979, peaked 1985 with 5.5 million mt and decreased thereafter to almost zero at present (Fig. 5.1; Fig. 10.3a). Zooplankton volumes declined in 1969 and preceded the anchovy crash in 1970 (Ayon et al. 2004) (Fig.5.3), whereas the recovering anchovy stock in 1984 preceded the increase of zooplankton biomass by about two years. However, up to 2001, zooplankton biomass had by far not reached the high values of the 1960s. As zooplankton volumes, phytoplankton volumes started to increase again around 1987, after the recovery of the anchovy (Sánchez 2000) Regime shifts All seven HCE anchovy and sardine stocks (Alheit and Niquen 2004) show clear decadal variability in abundance and, in spite of the wide geographical distances between their habitats, they seem to swing in synchrony. When E. ringens supports high biomasses, S. sagax exhibits low population levels and vice versa. Consequently, the HCE has passed through alternating anchovy and sardine periods on the decadal time scale. Critical periods of transition, turning points, were when the famous Peruvian anchovy stock started to collapse, and , when the HCE switched back from a sardine to an anchovy system. The fact that all seven anchovy and sardine populations showed dramatic changes around these periods indicates the regime shift character of these processes. Major changes in zooplankton and fish populations were observed between 1969 and 1974, well before

22 Over the last forty years, two regime shifts of the HCE have been recorded (Alheit and Niquen 2004) was the turning point when the HCE changed from an anchovy- to a sardinedominated system and dramatic changes were observed during this short period. Anchovy recruitment collapsed in 1971 (Gulland 1982, Mendelssohn and Mendo 1987). Zooplankton biomass decreased drastically along the entire Peruvian coast (Alheit and Niquen 2004, Ayon et al. 2004, Carrasco and Lozano 1989). Off northern Chile zooplankton biomass also diminished (Bernal et al. 1983) and the composition of the ichthyoplankton community changed markedly (Loeb and Rojas 1988). Subsequently, the once largest fish population in the world, the Peruvian anchovy collapsed and increasing sardine spawning was recorded. Whereas the anchovy went down to very low biomass levels, the sardine biomass increased steadily and surpassed that of the anchovy in the mid-1970s. Zooplankton biomass stayed at very low levels, just as the anchovy. In the mid-1980s, the HCE shifted back to an anchovy system, but zooplankton biomass increased only several years later. However, phytoplankton biomass also started to increase in the mid- 1980s (Sanchez 2000). Thus, when the HCE shifted in the late 1960s changes were recorded in zooplankton and fish, whereas, when it changed back, signals were observed in phytoplankton and fish (Alheit and Niquen 2004). All these changes are summarized in Table Mechanisms linking climate to decadal-scale population dynamics In the HCE, the regime shifts seem to be linked to lasting periods of warm or cold water anomalies related to the approach or retreat of warm subtropical oceanic waters to the coast of Peru and Chile (Santander and Flores 1983, Tsukayama 1983, Alheit and Bernal 1993, Alheit and Niquen 2004). When the SST anomalies of the El Niño 4 region are subtracted from those of the El Niño 1&2 region (Trenberth and Stepaniak 2001) and the results plotted cumulatively, the

23 resulting cusums indicate changes in 1970 and 1986 (Fig. 5.7). This is interpreted as oceanic warm water masses moving to the coast between 1970 and 1986 (E. Hagen, pers. comm.). Phases during the descending part of the curve parallel anchovy regimes (1950s to about 1970; 1985 up to the present) and the ascending phase from about was characterised by sardine dominance (Alheit and Niquen 2004). These results were recently confirmed by analysis of longterm salinity data from waters off Peru, up to 60 nm offshore. (Poster, O. Morón, Humboldt Current Symposium, Lima, Peru, December 2006; Checkley et al. 2007). These data demonstrate the approach of oceanic subtropical high salinity water masses toward the Peruvian coast. From 1960 to the late 1960s, coastal waters off Peru were dominated by cold coastal water (CCW; C; PSU). Thereafter, subtropical surface water (SSW; C; PSU) approached the coastal realm from the late 1960s to the mid-1980s. When the SSW retreated again offshore, it was replaced by CCW. Decadal SST variability along the western coast of South America has been studied by Montecinos et al. (2003). The first principal component (PC1) of the normalized SST time series (at 23.5 S) shows a temperature increase from about 1970 up to 1983 when SSTs decrease again. Associated with this decadal-scale variability of SSTs is a decadal-scale oscillation of the (modeled) thermocline (Pizarro and Montecinos 2004). During the late 1960s, the thermocline was very shallow. It then deepened continuously up to the early 1980s, when it became shallower again (Fig. 5.7). At the end of the 1960s, the thermocline was on average 10 m shallower than during the beginning of the 1990s (Pizarro and Montecinos 2004). Interestingly, the turning points of the thermocline are both a few years before the turning points of the cusums curve of the tem anomalies of the El Niño regions (Fig. 5.7). These processes described for the Humboldt are very similar to the suggestions by Chavez et al. (2003) that the California Current weakened and moved shoreward during the sardine period, whereas a stronger and broader California Current during the anchovy period was associated with a

24 shallower coastal thermocline from California to British Columbia, leading to enhanced primary production From the timing of all these physical processes, it emerges that the waters off Peru, and probably Chile, were dominated by CCW from the beginning of measurements in the early 1960s up to the late 1960s, a period of cooler SSTs and a shallow thermocline. This environment seems to have favoured a high biomass of meso-zooplankton > 300 micron and provided excellent recruitment conditions for the anchovy (Alheit and Niquen 2004). The intrusion of SSW beginning in the late 1960s resulted in warmer SSTs and a deeper thermocline. This new subtropical environment led to the decline of meso-zooplankton biomass and the crash of the Peruvian anchovy stock in which provided about 20% of the world fisheries yield at that time. Clearly, the anchovy stock collapsed before the El Niño as a combined result of a changing environment and, probably, also of heavy fishing pressure. The advance of SSW caused an entire suite of adverse conditions for the anchovy (Alheit and Niquen 2004). Predation on all life stages increased whereas the abundance of their most important food source, large calanoid copepods, decreased enhancing recruitment failure of the anchovy population. The new environment seems to have provided feeding conditions favourable for sardine and less favourable for anchovy which need larger particles than sardines (Alheit and Niquen 2004, van der Lingen et al. 2007). As zooplankton was collected with 300 micron nets, there is no information on long-term dynamics of zooplankton which passes through the meshes of these nets. In the mid-1980s, the SSW moved again offshore. Consequently, anchovy biomass increased again and the sardine populations started to decline. 575

25 Obviously, dynamics of anchovies and sardines in the HCE are governed by long-term, decadalscale physical processes. Whereas single El Niño events cause short-term perturbations of the environment which are unfavourable for the anchovy, the anchovy has the ability to recover very fast during periods when CCW is preponderant as observed, for example after the strong El Niño events in and (Alheit and Niquen 2004) Benguela Current Ecosystem Changes in the biota The Benguela Current Ecosystem consists of two subsystems: the northern part, off Namibia, from the Angola-Benguela front (14 S-16 S) to the permanent upwelling cell off Lüderitz (26 S) and the southern part, off South Africa, from the Lüderitz cell to East London (28 E). The southern part includes the upwelling region along the southwestern coast of southern Africa and extends over the Agulhas Bank along the south coast (Cury and Shannon 2004). Each of the two sub-systems has an independent sardine and anchovy stock. Spawning and fishing areas of sardines expand and contract with periods of high and low abundance (Lluch-Belda et al. 1989). Decadal-scale changes in the Benguela Current ecosystem have been described and discussed recently by Crawford et al. (2001), Cury and Shannon (2004) and van der Lingen et al. (2006a) Northern Benguela Catches of northern Benguela sardines increased throughout the 1950s and 1960s up to a maximum of 1.4 million mt in 1968 (Fig. 5.1; Fig. 10.3b) (Schwartzlose et al. 1999). They collapsed in , were between 0.3 and 0.7 million mt from and finally fell to

26 very low levels since. There might have been a slight recovery in the early 1990s. Virtual population analysis of spawner biomass demonstrates similar dynamics with a peak in Anchovy fisheries started in 1964, but they never reached the high yields as for sardine. In 1978, when catches peaked at 0.4 million mt, anchovy catch for the first time surpassed that of sardine. However, in 1984 catches fell to very low levels and have stayed there since, with the exception of At present, both species are at extremely low levels. It seems that anchovy replaced the sardine after the collapse in the late 1960s. Now that both species are very low they have apparently been replaced by a suite of different pelagic species such as horse mackerel (Trachurus capensis) and bearded goby (Sufflogobius bibarbatus) which were recorded in the purse seine fishery for the first time in 1971 (Crawford et al. 2001). Anomalies in the condition factor of sardine were negative from and then positive from , probably because of a density-dependent response to high and low biomass, respectively (Crawford et al. 2001). Also, after 1967, the proportion of spawners to non-spawners in the sardine population switched from 10-50% to 70-95%, indicating a density-dependent response whereby the population put more effort into reproduction as biomass declined (Crawford et al. 2001). Sardine egg concentrations have decreased and egg distribution has been contracted to the north since the collapse. Also, anchovy egg distribution has shifted northwards (van der Lingen et al. 2006b). At higher population levels, anchovy and sardine used to migrate between the northern spawning and the southern feeding grounds in the Walvis Bay area. Intense fishing may have interrupted this migration pattern (Boyer et al. 2001). The substantial decrease of the Walvis Bay spawning is suggested to be the result of removal of older, migratory specimen (Daskalov et al. 2003) or of a selective change in migratory behaviour as a reaction to heavy fishing activities in the Walvis Bay area (Bakun 2001). The change in pelagic fish community structure was reflected in predator diets. The proportion of sardines in the diet of Cape gannets, Cape cormorants and African

27 penguins declined whereas the pelagic goby gained importance (Cury and Shannon 2004). Also, sardine lost and the pelagic goby gained importance in seal diets after the sardine collapse There appears to be a decade-scale decline in zooplankton biomass and copepod abundance off Walvis Bay during the 1960s and 1970s until about the mid-1980s when the trend reversed and a subsequent increasing trend over the next two decades became evident. This cyclic pattern, which closely follows that of sea surface temperature in the same area, declining since the mid-1960s to a minimum in 1984 and rising subsequently until the present, is thought to be related to advective loss and retention processes associated with coastal upwelling (H.M. Verheye, pers. comm.) Southern Benguela Catches of southern Benguela sardine increased from the 1950s to a peak of 0.4 million mt in 1962 and then collapsed from 1963 throughout the mid-1960s to 0.07 million mt in 1967 (Fig. 5.1; Fig. 10.3b). Sardine increased again from 1985 onwards and has surpassed now the levels of the 1960s (van der Lingen et al. 2006b). Southern Benguela anchovies were caught in larger quantities in the 1960s, particularly as a reaction to the sardine collapse. For 30 years, from 1966 to 1996 anchovy catches were larger than sardine catches with a peak of 0.6 million mt in 1987 and 1988 (Fig. 5.1; Fig. 10.2b). Catches decreased thereafter, but increased again since As a density-dependent reaction, conditon factor of sardine started to increase from 1969 and stayed high until 1985, when it declined steadily (Crawford et al. 2001). At present, both species are on rather high levels. Considerable shifts in spawning grounds of anchovies and sardines have been reported (van der Lingen et al. 2006b). Anchovy shifted its major spawning activities since 1996 from the western to the central and eastern Agulhas Bank. The sardine has two major spawning grounds, the west coast from north of Cape Columbine to Cape Point and the south coast, the

28 Agulhas Bank. Relative importance of the two areas kept on changing periodically since the 1960s. As with the anchovy, most sardine spawning has been recorded at the south coast since The dynamics of anchovies and sardines as inferred from catch data are confirmed by other sources. The seabirds producing guano eat mainly small pelagics. Fluctuations in the quantity of guano produced are based on the availability of these fish to birds. Thus, the sardine collapse in the mid-1960s resulted in substantially reduced guano yields (Crawford and Shelton 1978). The long-term dynamics of anchovies and sardines are reflected in their predator s diets (Crawford et al. 2001). The frequency of occurrence of sardine in snoek (Thyrsites atun) stomachs fell from more than 50% in 1962 to less than 10% in 1964, just at the time when the sardine stock collapsed (van der Lingen et al. 2006b, Fig. 7a). At the same time, the percentage of anchovy increased to over 70%. Studies on the diet of Cape gannets show that there was a good sardine year class in 1983 which triggered the subsequent increase in sardine biomass (Crawford et al. 2001). Sardine increased steadily in the diet of Cape gannets from Biomass of sardine increased considerably after 1985 and at the same time the condition factor began to decrease simultaneously (Crawford et al. 2001). Around , there was a change in the relative abundance of larvae of non-harvested mesopelagic fish species (Loeb 1988). Total copepod abundance in a single location in the southern Benguela increased by two orders of magnitude from (Verheye et al. 1998). During the period of the preferentially filter-feeding sardine dominance, smaller-sized cyclopoid copepods made up 41% of the crustacean zooplankton. In contrast, when the preferentially particle-feeding anchovy was dominant, the proportion of cyclopoid crustacean zooplankton was significantly higher (56%) (Fig. 8.16) (Verheye and Richardson 1998, Verheye 2000, van der Lingen et al. 2007). This differential

29 predation impact of sardine and anchovy focusing on different size spectra of crustacean zooplankton might be a mechanism leading to the alternating dominance of sardines and anchovies (Alheit and Niquen 2004, van der Lingen et al. 2006a) (see below) Regime shifts Northern and southern Benguela ecosystems which are separated by the strong Lüderitz upwelling cell provide quite distinct habitats for anchovies and sardines. In the northern Benguela, two kinds of environmental anomalies have been recorded periodically which lead to shifts in distribution and decline in recruitment and catches of pelagic resources, as was observed particularly during the 1990s. The Benguela Niño events, which occur about once every decade, are associated with warm waters penetrating onto the Namibian shelf. Frequently occurring lowoxygen or anoxic events (Weeks et al. 2002) also lead to high mortalities in the pelagic realm of the northern Benguela. The low-oxygen event in and the immediately following Benguela Niño in early 1995 probably inhibited the potential recovery of northern Benguela sardines in the mid-1990s, beside the heavy fishing pressure. Another substantial difference between the two Benguela sub-systems is that anchovies and sardines in the southern Benguela make use of the non-upwelling Agulhas Bank environment for completing their life cycle. Consequently, their dynamics are not directly comparable with those from pure upwelling ecosystems. This might also explain why the alternation between anchovy and sardine periods, which is typical for upwelling systems, has not been observed in the southern Benguela recently. Whereas there seems to have been a shift in the southern Benguela from a sardine to an anchovy regime in the early/mid-1960s with concomitant changes in the zooplankton (Verheye et al. 1998) and predator communities (Cury and Shannon 2004) and large changes in the structure and

30 functioning of the pelagic component of the southern Benguela ecosystem (Crawford et al.), both species exhibit recently very high population sizes at the same time since the late 1990s in the southern Benguela. Dynamics of small pelagics in the northern Benguela took a different turn. After the collapse of the sardine in the late 1960s, catches of anchovies increased. However, the increase in biomass of other pelagic species such as horse mackerel, pelagic goby and even jelly fish was much more conspicuous and led Cury and Shannon (2004) to the assumption that there was a regime shift from sardines to other non-anchovy pelagic species. There was probably a substantial transfer of production to the mid-water which implies that the trophic functioning of the system changed considerably. Changes in the trophic regime of the northern Benguela ecosystem continued into the 1970s, whereas the regime shift in the southern Benguela was completed in the 1960s (Crawford et al. 2001). In addition, both, anchovies and sardines in the northern Benguela have been at extremely low biomass levels since the early 1990s, partly caused by the devastating warm water and low-oxygen events of 1993 to The major changes in the entire pelagic Benguela system took place in the 1960s and the early 1980s, when, respectively, the two sardine populations decreased and that of the southern system recovered again (Crawford et al. 2001). Apparently, these changes in the trophic structure and functioning of the northern Benguela ecosystem were associated with intrusions of warm, saline surface waters onto the Namibian shelf, the Benguela Niños (Crawford et al. 2001). Also, the major changes in the southern Benguela were approximately coincident with Benguela Niños (Crawford et al. 2001). The causes of the regime shifts in both Benguela subsystems remain largely obscure Canary Current ecosystem 718

31 Changes in the biota In contrast to the other ecosystems analyzed in this article, the Canary Current ecosystem (CanCE) is not dominated by a Sardinops/Engraulis species pair. Instead, it is characterized by Sardina pilchardus and Sardinella aurita and S. maderensis. Anchovies (E. encrasicolus), although present in this ecosystem, do not play such an important role as in the other systems. Biogeographically, the CanCE can be divided into two regions with a transition between Cape Barbas (23 N) and Cape Blanc (21 N), the northern one of which is characterized by North Atlantic Central Water (NACW) whereas the southern one is dominated by South Atlantic Central Water (SACW) (Mittelstaedt, 1983). S. pilchardus and E. encrasicolus are predominant in the NACW area, off the Iberian and Moroccan coasts, whereas Sardinella aurita and S. maderensis are dwelling in the SACW area, off Mauritania and Senegal (Fréon, 1988, Marchal, 1991) Off NW Africa, between 20 N-36 N, there are three important fishing areas for S. pilchardus: in the North (36 N-33 N), the Centre (32 30 N to 27 N), and the South (26 N to 20 N) (Belvèze and Erzini 1983). It is assumed that they correspond to separate stocks, but some seasonally occurring partial mixing is likely. The biomass in the central and southern areas has widely fluctuated during the last century. Sardine has largely dominated the total landings of pelagic fish north of Cape Bojador (26 N) since the beginning of fishing activities in the 1920s. The northern and central stocks provided the bulk of the catches until the 1970s. After 1966, an increase of sardine spawning was observed south of Cape Juby (28 N) and, from 1970 onward, the area of distribution and the abundance of sardine increased notably toward the south after several years of good spawning (Holzlöhner 1975, Barkova and Domanevsky 1976). In contrast, in the

32 historical Moroccan sardine fishery grounds (the northern part of the central area), a constant decrease of the catches, spreading from north to south, began around 1968 and in the early 1990s the fishery, that operated between 32 N and 30 N, collapsed (Belvèze and Erzini 1984, Kifani 1998) (Fig. 5.1). During the 1970s, south of Cape Bojador (26 N) to Cape Blanc (20 N), the rather tropical Sardinella and Scomber japonicus were replaced in the catch composition by Sardina which had expanded southwards. During this period of high abundance of sardine off the Saharan area (south of 25 N), significant catches have been recorded as far south as Cape Verde (14 N) (Fréon 1988) Other remarkable events that were recorded during the period of sardine outburst off the Sahara are the dramatic increase in the early 1970s of snipefish (Macrorhamphosus scolopax and M. gracilis), both in Moroccan and Iberian waters (Brêthes 1979), as well as the sudden explosion of the triggerfish (Balistes carolinensis) that spread geographically from Ghana to Mauritania, occupying the pelagial during the first two years of its life cycle (Caverivière 1991, Fréon and Misund 1999). Triggerfish and snipefish collapsed simultaneously in NW African waters after they had expanded their biomass respectively to more than 1 million mt in the late 1970s/early 1980s (Sætersdal et al. 1999, Belvèze 1984). Outburst and collapse of snipefish remain largely unexplained, while the triggerfish expansion was linked to South Sahelian rivers runoff deficit during the 1970s and 1980s (Gulland and Garcia 1984, Caverivière 1991) In the mid-1990s, the biomass of the Saharan sardine stock between 20 N and 26 N crashed drastically from more than 5 million mt in 1995 to about 1 million mt in This collapse does not seem to be linked to fishing pressure, as the USSR fishing fleet had left the area already in the early 1990s. Thereafter, the sardine recovered again gradually. In the early 1980s, most of the total biomass of sardinellas was located between Cape Verde (14 40'N) and Cape Roxo

33 (12 20'N) (Sætersdal et al. 1999), while, after the mid-90s, more than 50% of the total biomass of these species was found between 20 N and 25 N, indicating a clear northward shift in the distribution of sardinellas towards the Saharan region. Recently, over the last two years, some scattered concentrations of sardinella have been caught further north, off Cape Juby. Zooplankton data collected over the NW African continental shelf and slope between 16 N and N revealed over the Mauritanian shelf a pronounced increase up to 1998 of zooplankton species which are usually prevailing in the waters off Senegal and Guinea (Sirota et al. 2004) Changes in geographic distribution and potential regime shifts Although a large amount of decadal-scale variability has been described for small pelagic fish populations off NW Africa, neither regime shifts nor alternating population fluctuations between anchovies and sardines have been observed, in contrast to other eastern boundary systems. Possible reasons, as stated above, are that a different sardine species is dwelling in the CanCE and that anchovy biomass is rather negligible (Fig. 5.1). In addition, there are no decadal-scale long-term time series on other biological components than fish available from the waters off NW Africa which would indicate regime shifts. Instead of regime shifts, this region seems to be governed by large-scale shifts of distributional boundaries of small pelagic fish populations moving the centres of gravities of the populations and the transition zones between S. pilchardus and Sardinella spp. northwards and southwards along the NW African coast (Kifani 1998, Binet et al. 1998) Over the period in which there have been major fisheries operating, two transformations of habitat geography of S. pilchardus off NW Africa have been observed and, over a multi-decadal

34 time scale, gradual 2000 km southward expansions along the coast have occurred (Binet et al. 1998, Kifani 1998, Bakun 2005). In the 1920s, the southward limit of sardines has been reported to be off northern Morocco. By the 1950s, the species had become abundant as far south as Mauritania (Belvéze 1984), and, around the mid-1970s, sardines were fairly common even off Senegal (Fréon 1988, Bakun 2005). Thereafter, the distribution area contracted again northwards and sardine abundance was drasticaly reduced south of 25 N in (Bakun 2005). Subsequently, sardines expanded their range far southward and were caught again in Senegalese waters during the 1990s (Binet et al. 1998, Bakun 2005). The southward migration in the late 1960s was followed by the emergence of a new sardine fishery off the Sahara which increased quickly from to mt (Binet et al. 1998). At the same time, in the early 1970s, the small pelagic community in Saharan waters switched from a community dominated by horse mackerel, mackerel and sardinellas to a sardine dominated one (Gulland and Garcia, 1984). When the sardine range expanded to the south, the distribution area of sardinella shifted equatorward (Kifani 1998). In 1974, some sardines were fished off Senegal. A few years later the southern boundary of the sardine population started moving back to the north again and by sardine had almost disappeared from Mauritanian waters (Binet et al. 1998). A second southward extension of the sardine population occurred in the late 1980s all the way to Senegal leading again to high catches off the Sahara). Thus, twice in twenty years a southward extension of the geographical range of the sardine was recorded, followed by high catches. Again, in the mid- 1990s, when the abundance of sardine off Saharan declined, sardinellas extended northwards (Fréon et al. 2006). It is not clear whether these alternate shifts in biogeographic boundaries of sardines and sardinellas might indicate regime shifts. 812

35 Mechanisms linking climate to decadal-scale population dynamics The reason for the collapse in the early 1990s of the traditional Moroccan sardine fisheries in the area surrounding Cap Sim and Agadir Bay (32 N-30 N) is still an unresolved issue. Several hypotheses have been proposed (Belvèze and Erzini 1984, Do-Chi and Kiefer 1996, Kifani 1998, Bakun 2005). It seems that this collapse could result from a reduced sardine migration from the primary reproductive zone in the south (near Tantan, between 28 N and 29 N) to the adult feeding grounds located further north in the region of active upwelling between Cap Sim and Agadir Bay where the major fishing ports are located. A decrease of the wind induced upwelling and the associated increasing trend in sea surface temperature has been proposed as an explanation for the cessation of the migration to the feeding grounds (Kifani 1998, Bakun 2005) The increase in sardine biomass that occurred during the 1970s in the southern part of the Canary current (26 N to 20 N) and the related southward extension of the sardine population was associated with the strengthening of upwelling off the Sahara region and further south off Mauritania and Senegal (Sedykh 1978, Holzlöhner 1975, Binet 1988, Binet et al. 1998). There is also indication that the second southward extension of the sardine population that occurred in the late 1980s was related to colder than usual temperatures and increased upwelling (Binet et al. 1998, Roy and Reason, 2001). The mechanisms involved are still unclear. Binet et al. (1998) hypothesized that intensification of the trade winds along the NW African coast enhanced upwelling activity and southward transport whereby phytoplankton production was most probably boosted by enhanced upwelling, but not matched by zooplankton grazing, due to the brevity of the water residence time over the shelf. This led to new distribution patterns of primary and secondary production that are in favour of sardines which can feed on phytoplankton in contrast to the zooplankton-feeding sardinellas (Binet et al. 1998). A recent study of the impact of

36 the continental shelf geometry on the structure of a wind-driven upwelling explains the mechanism which separates the upwelling area from the coast, that is observed off Sahara (Estrade 2006, Estrade et al., subm.). It has been shown, that enhanced upwelling-favourable wind over this wide and shallow shelf tends to move the core of the uwelling from the nearshore domain to the mid-shelf area. This creates an inner front that allows retention of biological material such as larvae in the highly productive nearshore environment. This separation of the upwelling core from the near-coastal area that is observed and modelled under enhanced wind forcing is another potential mechanism to explain the postive link between upwelling intensity and sardine abundance off the Sahara region. The increased occurrences of tropical species north of Cap Blanc after the mid-1990s seem to be associated with a northward shift of the boundary between the NACW and the SACW toward N latitudes and warmer SSTs off NW Africa and changes in the main current pattern (Ostrowski and Strømme 2004, Sirota et al. 2004, Ostrowski 2005). The Atlantic Multidecadal Oscillation, AMO (Kerr 2000), a distinct signal of multidecadal variability of North Atlantic SST was in a positive phase between 1925 and Another positive period is in progress since the mid-1990s. In contrast, cold periods were recorded from the end of the 19th century until 1925 and from the mid-1960s to the mid-1990s. Interestingly, the timing of the turning points of AMO anomalies seems to be consistent with the timing of the shifts recorded in the CanCE, such as trade wind intensity, current and species distribution and catches (Fig. 5.9). Also, Moses et al. (2006), using a century length proxy record from corals (Siderastrea radians) at Pedra de Lume ( N, W) on the island of Sal, Cape Verde Islands, in the eastern tropical North Atlantic, show that long-term salinity data correlate positively with the two major North Atlantic SST indices, the Tropical North Atlantic Index (TNA) and Atlantic Multidecadal Oscillation (AMO).

37 Discussion Temperature For many authors (e.g. Chavez et al. 2003), it has been a puzzle that sardine biomass increased in the eastern boundary currents, when SSTs increased, whereas the Japanese sardine exploded when temperatures decreased. However, the argument that Japanese sardines increase in abundance when temperatures decrease is not valid as a closer look at their ambient temperatures and the respective literature reveals. Already Tomosada (1988) stated that sardine catch increased in the 1930s and the 1970s when the temperature in the coastal sardine spawning grounds increased. He also reported that, at the same time, temperature in the fishing grounds decreased. Also, when the KC adopts its large meandering phase (Kasai et al. 1993, 1996), the frequency of warm water Kuroshio intrusions into the coastal zone such as the Enshu-nada Sea, which is one of the most important larval nursery areas, is increased. Consequently, during the high abundance phase of the Pacific sardine population off Japan, shelf temperatures were mainly elevated (Nakata et al. 2000, Yasuda et al. 1999). After 1988, shelf temperatures started to decrease again. In addition, it has to be considered, that the Pacific sardine of Japan expands during its high abundance phase from cooler coastal waters into a warm water current, the KC. Consequently, the effect is the same in eastern boundary currents (CC, HC) as well as in the western boundary current (KC): sardine populations in their expansion phase extend their area of distribution into warmer waters. 882

38 Trophodynamic aspects The alternation between anchovy and sardine periods may be trophodynamically mediated (van der Lingen et al. 2006a, 2007). Intermittent mixing such as upwelling leads to relatively cool temperatures in the upper layers and favours food chains dominated by diatoms and large calanoid copepods. This is the favourite feeding environment for the preferentially particulatefeeding anchovy as demonstrated for the Humboldt (Alheit and Niquen 2004) and the Benguela Current ecosystems (van der Lingen et al. 2006a). More stable water column situations caused by e.g. relaxed upwelling and/or El Niño leading to warmer temperatures in the upper layers lead to flagellate-dominated food chains and a shift in the size spectrum of the crustacean zooplankton towards small-sized copepods such as cyclopoid copepods (Alheit and Niquen 2004, van der Lingen et al. 2006a). For example, abundances of small cyclopoid copepods (Oncaea spp. and Oithona spp.) in the Humboldt Current ecosystem increased from 3-fold to 1 order of magnitude (Oithona) as the El Niño event developed from January 1997 to January 1998 (González et al. 2000). This is the feeding environment more favourable for the non-selective filter-feeding sardine (Alheit and Niquen 2004, van der Lingen et al. 2006a). Physical forcing leading to different size spectra of phyto- and zooplankton provides by trophodynamical mediation feeding environments favourable for anchovies or for sardines, respectively (van der Lingen et al. 2006a, b) (Fig. 8.6.Beng-4; van der Lingen 2007) Regime shifts and mechanisms linking climate to decadal-scale populations dynamics Major hypotheses about potential mechanisms for the observed low frequency population swings of anchovies and sardines and their alternations are discussed by MacCall (2007). A climatic

39 regime shift has been reported for the North Pacific region for (e.g. Hare and Mantua 2000, King et al. 2005) and numerous publications from Pacific rim ecosystems in the western and eastern North Pacific and the eastern South Pacific have described ecosystem regime shifts occurring in reaction to this climate shift. However, more exact investigation of biological data series reveals that in several cases dramatic changes in the abundance of key populations did not occur in the mid-1970s, but some years earlier, between the late 1960s and the early 1970s. Often, the occurrence of a regime shift was concluded because biomass anomaly values of time series were crossing the zero line. However, the crossing of the zero line has no biological meaning. Consequently, the concept of a North Pacific or even a pan-pancific ecosystem regime shift in the mid-1970s has to be seriously questioned. The scrutiny of zooplankton and pelagic fish time series of three major Pacific ecosystems such as the waters east of Japan, the CCE and the HCE reveals that many of these populations did not exhibit abrupt changes in response to the climate shift of the North Pacific. However, a critical period for anchovies, sardines and zooplankton in waters east of Japan and the HCE was around when anchovy populations (Fig. 5.1) and meso-zooplankton in the Oyashio, off Peru and off Chile declined (Fig. 5.3) and sardine populations started to increase. For the same period, a number of major changes in physical variables such as SST and subsurface processes (depth changes in MLD and thermocline) have been described. Around the mid- to late 1980s, all these changes reversed. Obviously, the waters east of Japan and the HCE experienced regime shifts and from the mid- to late 1980s. Studies of major changes in physical sub-surface processes which were observed at the same time of changes in major fish and zooplankton populations give rise to speculations that they were caused by changes in basin-scale gyre circulation, particularly by more frequent incursions of water from the subtropical gyre in the CCE (Bograd and Lynn 2003, McGowan et al. 2003) or a spin-up of the subtropical gyre in the Kuroshio Extension

40 (Yasuda et al. 2000, Yasuda and Hanawa 1997). Also, the approach and retreat of the warm SSW to and from the Peruvian coast indicates that changes in basin-scale circulation might be involved in decadal-scale dynamics of anchovy and sardine populations and respective ecosystem regime shifts. The same argument with respect to circulation is brought forward by MacCall (2007) who suggests that flow patterns in the Japanese system and eastern boundary currents are strongly associated not only with characteristic shifts in ocean temperature, but also with shifts in the physical and biological structure of entire ecosystems. In conclusion, alternations between anchovy and sardine populations and regime shifts in respective ecosystems seem to be brought about by changes in basin-scale circulations which alter the feeding (and maybe the reproductive) environment such that they are either favourable to anchovies or to sardines The possible association between Pacific ecosystem regime shifts and basin-scale circulation changes seems to be compatible with the flow hypothesis of MacCall (2007). Although changes in basin-scale circulation are tightly interwoven with climate variability, the direct association of shifts in the dynamics of small pelagic fish and their ecosystems and climate dynamics, at least in the Pacific, is still obscure. The major change in the PDO was observed when the anomalies jumped from negative to positive values thereby crossing the zero line. However, at the time of the observed regime shifts and mid- to late 1980s there are also clear signals in the PDO which exhibits turning points during these brief periods. From the late 1950s to 1970, PDO anomalies became more and more negative. Then, the trend turned around and anomalies became more and more positive until the mid-/late 1980s. Thus, the timing of the turning points of PDO anomalies corresponds to the shifts recorded in the ecosystems. Also, there is an interesting link between historic sardine fishing periods of Japanese sardines and the dynamics of the Aleutian Low (Yasuda et al. 1999). Spring temperatures over the NW coast of N America are

41 correlated with the dynamics of the Aleutian Low and Kuroshio temperatures. Using tree rings, a NW coast temperature curve could be reconstructed for several centuries indicating that high sardine yields are in synchrony with positive temperature anomalies over the NW coast. Consequently, it is concluded that the dynamics of the Japanese sardine is governed by Aleutian Low dynamics Synchronies and teleconnections The long-term dynamics of the HCE are characterized by alternating sardine and anchovy regimes and an associated restructuring of the entire ecosystem from phytoplankton to the top predators (Alheit and Niquen 2004). These regime shifts seem to be linked to lasting periods of warm or cold water anomalies related to the approach or retreat of warm subtropical oceanic waters (SSW) to the coast of Peru and Chile. Phases with mainly negative temperature anomalies parallel anchovy regimes (1950s- about 1970; 1985-up to now) and the rather warm period from about was characterised by sardine dominance. The transition periods (turning points) from one regime to the other were and which are much earlier than those suggested by Chavez et al. (2003) who suggested mid-1970s and mid- to late 1990s, respectively. The KCE is similarly characterized by alternating periods of sardines and anchovies. The most recent transition periods between the two Kuroshio species were strikingly synchronous to those of their Humboldt cogeners (Fig. 5.8). The striking synchronies between the HCE and the KCE extend to the timing of (i) changes between temperature regimes, (ii) sub-surface processes (changes in depth of thermocline depth, MLD position), (iii) anchovy and sardine periods, (iv) dynamics of zooplankton and (v) other nektonic populations. The question remains: how were the changes in temperature regimes and sub-surface processes in both ecosystems synchronized.

42 A very likely mechanism are synchronized flow patterns of basin-scale circulation of the gyres in the North and South Pacific. Coastal sea level is a well-established indicator of integrated boundary current flow (Chelton et al. 1992, MacCall 2007). Using sea level data, MacCall (2007) demonstrates a Pacific-wide tendency toward synchrony in fluctuations in sea levels and associated boundary current strenghts. The claim of Kawasaki (1983) and Chavez et al. (2003) that Pacific anchovy and sardine populations fluctuate in synchrony cannot be confirmed. Whereas fluctuations of sardines off Japan and California might have been in synchrony during the 1930s and 1940s, they are clearly out of phase since the 1970s (Fig. 5.1)

43 1003 Box 5.1: Small Pelagic Fish Small pelagic fishes such as sardine, anchovy, sardinella, sprat and others represent about % of the total annual world fisheries catch. They are widespread and occur in all oceans. They support important fisheries all over the world and the economies of many countries depend on those fisheries. They do respond dramatically and quickly to changes in ocean climate. Most are highly mobile; have short, plankton-based food chains and some even feed directly on phytoplankton. They are short-lived, highly fecund and some can spawn all year-round. These biological characteristics make them highly sensitive to environmental forcing and extremely variable in their abundance (Hunter and Alheit 1995). Thousandfold changes in abundance over a few decades are characteristic for small pelagics and well-known examples include the Japanese sardine, sardines in the California Current, anchovies in the Humboldt Current, sardines in the Benguela Current or herring in European waters. Their drastic stock fluctuations often caused dramatic consequences for fishing communities, entire regions and even whole countries. Their dynamics have important economic consequences as well as ecological ones. They are the forage for larger fish, seabirds and marine mammals. The collapse of small pelagic fish populations is often accompanied by sharp declines in marine bird and mammal populations that depend on them for food (Hunter and Alheit 1995). Major changes in abundance of small pelagic fishes may be accompanied by marked changes in ecosystem structure. The great plasticity in the growth, survival and other life-history characteristics of small pelagic fishes is the key to their dynamics and makes them ideal targets for testing the impact of climate variability on marine ecosystems References 1026

44 Alheit, J. and Bakun, A. (in press). Population synchronies within and between ocean basins: apparent teleconnections and implications as to physical-biological linkage mechanisms. J. mar. Syst Alheit, J. and Bernal, P. (1993). Effects of physical and biological changes on the biomass yield of the Humboldt Current ecosystem. In Large Marine Ecosystems - Stress, Mitigation, and Sustainability, ed. K. Sherman, L. M. Alexander, and B. Gold, American Association for the Advancement of Science, pp Alheit, J. and Niquen, M. (2004). Regime shifts in the Humboldt Current ecosystem. Prog. Oceanogr., 60, Ayón, P., Purca, S. and Guevara-Carrasco, R. (2004). Zooplankton volume trends off Peru between 1964 and ICES J. mar. Sci., 61, Bakun, A. (1987). Monthly variability in the ocean habitat off Peru as deduced from maritime observations, In The Peruvian Anchoveta and Its Upwelling Ecosystem: Three Decades of Change, ed. D. Pauly and I. Tsukayama. Manila: International Center for Living Aquatic Resources Management (ICLARM), pp Bakun, A. (1990). Global climate change and intensification of coastal ocean upwelling. Science, 247,

45 Bakun, A. (1996). Patterns in the Ocean: Ocean Processes and Marine Population Dynamics, San Diego: University of California Sea Grant Bakun, A. (2001). School-mix feedback : a different way to think about low frequency variability in large mobile fish populations. Prog. Oceanogr., 49, Bakun, A. (2005). Regime Shifts. In The Sea, vol.13, ed. A. R. Robinson and K. Brink. Cambridge, Massachusetts: Harvard University Press, pp Barange, M., Bernal, M., Cercole, M. C., Cubillos, L., Cunningham, C. L., Daskalov, G. M., de Oliveira, J. A. A., Dickey-Collas, M., Hill, K., Jacobson, L., Köster, F. W., Masse, J., Nishida, H., Niquen, M., Oozeki, Y., Palomera, I., Saccardo, S. A., Santojanni, A., Serra, R., Somarakis, S., Stratoudakis, Y., van der Lingen, C. D., Uriarte, A. and Yatsu, A. (2007). Current trends in the assessment and management of small pelagic fish stocks (this volume) Barkova N.A. and Domanevsky L.N. (1976). Some peculiarities of sardine (Sardina pilchardus) distribution and spawning along Northwest Africa. ICES. C.M. 1976/J: 6 Pelagic fish (Southern) Committee, 15p Baumgartner, T.R., Soutar, A. and Ferreira-Bartrina, V. (1992). Reconstruction of the history of Pacific sardine and northern anchovy populations over the past two millenia from sediments of the Santa Barbara Basin, California. CalCOFI Rep., 33,

46 Beamish, R.J. (ed.). (1995). Climate Change and Northern Fish Populations. Can. Spec. Publ. Fish. Aqat. Sci., 121, Beaugrand, G. (2004). The North Sea regime shift: evidence, causes, mechanisms and consequences. Prog. Oceanogr., 60, Beaugrand, G. and Reid, P.C. (2003). Long-term changes in phytoplankton, zooplankton and salmon related to climate. Glob. Change Biol., 9, Belvèze, H. (1984). Biologie et dynamique des populations de sardine peuplant les côtes atlantiques marocaines et propositions pour un aménagement des pêcheries. Unpublished Ph.D. Thesis, Université de Bretagne Occidentale, Brest Belvèze, H. and Erzini, K. (1983). The influence of hydroclimatic factors on the availability of the sardine (Sardina pilchardus, Walbaum) in the Moroccan Atlantic fishery. FAO Fish. Rep., 291, Bernal, P., Robles, F.L. and Rojas, O. (1983). Variabilidad fisica y biologica en la region meridional del sistema de corrientes Chile-Peru. FAO Fish. Rep., 291, Binet, D. (1988). Rôle possible d'une intensification des alizés sur le changement de répartition des sardines et sardinelles le long de la côte ouest africaine. Aquat. Living Resour., 1,

47 Binet, D., Samb, B., Sidi, M. T., Levenez, J. J. and Servain, J. (1998). Sardine and other pelagic fisheries changes associated with multi-year trade wind increases in the southern Canary current. In Global versus Local Changes in Upwelling Systems, ed. M. H. Durand, P. Cury, R. Mendelssohn, C. Roy, A. Bakun, and D. Pauly. Paris: ORSTOM Editions, pp Bograd, S. J. and Lynn, R.J. (2003). Long-term variability in the Southern California Current System. Deep-Sea Res. II, 50, Boyer, D. C., Boyer, H. J., Fossen, I. and Kreiner, A. (2001). Changes in abundance of the northern Benguela sardine stock during the decade 1990 to 2000 with comments on the relative importance of fishing and the environment. S. Afr. J. mar. Sci., 23, Brêthes, J. C.(1979). Contribution à l étude des populations de Macrorhamphosus scolopax (L. 1759) et Macrorhamphosus gracilis (Lowe, 1839) des côtes atlantiques marocaines. Bull. de l Inst. Pêches Marit.. N 24 (juillet 1979) Carrasco, S. and Lozano, O. (1989). Seasonal and long-term variations of zooplankton volumes in the Peruvian sea, In The Peruvian Upwelling Ecosystem: Dynamics and Interactions. Ed. D. Pauly, P. Muck, J. Mendo and 1. Tsukayama. Manila: International Center for Living Aquatic Resources Management (ICLARM), pp Caverivière, A. (1991). Quelques considérations méthodologiques sur les campagnes scientifiques de chalutage de fond en Afrique Occidentale. ICES C.M. 1991/D:

48 Chavez, F. P., Ryan, J., Lluch-Cota, S. E. and Niquen, M. (2002). From anchovies to sardines and back : multidecadal change in the Pacific Ocean. Science, 299, Checkley, D., Ayon, P., Baumgartner, T., Bernal, M., Coetzee, J. C., Emmett, R., Guevara, R., Hutchings, L., Ibaibarriaga, L., Nakata, H., Oozeki, Y., Planque, B., Schweigert, J., Stratoudakis, Y. and van der Lingen, C. (2007). Habitats of small, pelagic fish (this volume) Chelton, D. B., Bernal, P. A. and McGowan, J. A. (1982). Large-scale interannual physical and biological interaction in the California Current. J. mar. Res., 40, Chiba, S., Tadokoro, K., Sugisaki, H. and Saino, T. (2006). Effects of decadal climate change on zooplankton over the last 50 years in the western subarctic North Pacific. Glob. Change Biol., 12, Crawford, R. J. M. and Shelton, P. A. (1978). Pelagic fish and seabird interrelationships off the coasts of South West and South Africa. Biol. Conserv., 14, Crawford, R. J. M. and 23 co-authors. (2001). Periods of major change in the structure and functioning of the pelagic component of the Benguela ecosystem, Unpublished proceedings of a GLOBEC/SPACC workshop in Cape Town, Marine and Coastal Management, Department of Environmental Affairs and Tourism, Private Bag X2, Rogge Bay, 8012, South Africa. 1142

49 Csirke, J., Guevara-Carrasco, R., Cardenas, G., Niquen, M. and Cipollini, A. (1996). Situación de los recursos anchoveta (Engraulis ringens) y sardina (Sardinops sagax) a principios de 1994 y perspectivas para la pesca en el Perú, con particular referencia a las regiones norte y centro de la costa Peruana. Bol. Inst. del Mar del Perú-Callao, 15, pp Cury, P. and Shannon, L. (2004). Regime shifts in upwelling ecosystems: observed changes and possible mechanisms in the northern and southern Benguela. Prog. Oceanogr. 60, pp Cushing, D. (1982). Climate and Fisheries. London: Academic Press Daskalov, G. M., Boyer, D. C. and Roux, J.-P. (2003). Relating sardine Sardinops sagax abundance to environmental variables in the northern Benguela. Prog. Oceanogr., 59, DeYoung, B., Harris, R., Alheit, J., Beaugrand, G., Mantua, N. and Shannon, L. (2004). Detecting regime shifts in the ocean: data considerations. Prog. Oceanogr., 60, Do Chi, T. and Kiefer, D. A. (ed.) (1996). Report of the Workshop on the Coastal Pelagic Resources of the Upwelling System off Northwest Africa: Research and Predictions. FI:TCP/MOR/4556(A) Field Document 1. Rome: FAO Ebbesmeyer, C., Cayan, D. R., McLain, D. R., Nichols, F. H., Peterson, D. H., Redmond and K. T. (1991) step in the Pacific climate: forty environmental changes between ( and ). In Proc. Seventh Ann. Pacific Climate (PACLIM) Wkshp, April (1990), ed. J. L.

50 Betancourt and V. L. Sharp. California Department of Water Resources Interagency Ecological Studies Program Tech. Rep. 26, Estrade, P. (2006). Mécanisme de décollement de l'upwelling sur les plateaux larges et peu profond d'afrique du Nord-Ouest. Unpublished Ph.D. thesis, Université de Bretagne Occidentale Estrade P., Marchesiello, P., Colin de Verdière, A. and Roy, C. (in press). Cross shelf structure of coastal upwelling: a two dimensional extension of Ekman's theory and a mechanism for inner shelf upwelling shut down Field, D.B., Baumgartner, T. R., Ferreira, V., Gutierrez, D., Lozano, H., Salvatteci, R. and Soutar, A. (2007). Fish scales in marine sediments as indicators of variability in small pelagic fishes (this volume) Fréon, P. (1988). Réponses et adaptations des stocks de clupéidés d'afrique de l'ouest à la variabilité du milieu et de l'explotation: Analyse et réflexion à partir de l'exemple du Sénégal. Etudes et Thèses, ORSTOM Fréon, P. and Misund, O.A. (1999). Dynamics of pelagic fish distribution and behaviour: effect on fisheries and stock assessment. London: Blackwell Fishing News Books Fréon, P., Alheit, J., Barton, E. D., Kifani, S. and Marchesiello, P. (2006). Modelling, forcasting and scenarios in comparable upwelling ecosystems: California, Canary and Humboldt. In

51 Benguela: Predicting a Large Marine Ecosystem, ed. V.Shannon, G. Hempel, P. Melanotte- Rizzoli, C. Moloney and J. Woods. Amsterdam: Elsevier, pp Gulland, J. (1982). Management policies and models of pelagic stocks: The Peruvian experience Faculdad de Ciencias Biológicas, Pontifica Universidad Católica de Chile, Monografías Biológicas No. 2, Gulland, J. A. and Garcia, S. (1984). Observed patterns in multispecies fisheries. In Exploitation of marine communities, ed. R.M. May. Berlin: Springer-Verlag, pp Hare, S. R. and Mantua, N. J. (2000). Empirical evidence for North Pacific regime shifts in 1977 and Prog. Oceanogr. 47, Hayward, T. L. (1997). Pacific ocean climate change: atmospheric forcing, ocean circulation and ecosystem response. TREE, 12, Hill, K. T., Lo, N. C. H., Macewicz, B. J. and Felix-Uraga, R. (2006). Assessment of the Pacific sardine (Sardinops sagax caerulea) population for U.S. management in NOAA Tech. Memo. NMFS-SWFSC, 386, Holzlöhner, S. (1975). On the recent stock developement of Sardina pilchardus Walb. off Spanish Sahara. ICES CM 1975/J:

52 Hunter, J. R. and Alheit, J. (1995). International GLOBEC Small Pelagic Fishes and Climate Change program. GLOBEC Report No Isaacs, J. D. (1976). Some ideas and frustrations about fishery science. Cal. Coop. Oceanic Fish. Invest. Rep., 18, Jacobson, L. D., de Oliveira, J. A. A., Barange, M., Cisneros-Mata, M. A., Felix-Uraga, R., Hunter, J. R., Kim, J. Y., Matsuura, Y., Niquen, M., Porteiro, C., Rothschild, B., Sanchez, R. P., Serra, R., Uriarte, A. And Wada, T. (2001). Surplus production, variability, and climate change in the great sardine and anchovy fisheries. Can. J. Fish. Aqua. Sci., 58, Kasai, A., Kimura, S. and Sugimoto, T. (1993). Warm water intrusion from the Kuroshio into the coastal areas south of Japan. J. Oceanogr., 49, Kasai, A., Koizumi, N. and Sugimoto, T. (1996). Transport and survival of Japanese sardine larvae (Sardinops melanostictus) in the Kuroshio-Oyashio region: A numerical approach. In Survival strategies in early life stages of marine resources, ed. Y. Watanabe,Y. Yamashita and Y. Oozeki. Rotterdam: A.A. Balkema, pp Kawasaki, T. (1983). Why do some pelagic fishes have wide fluctuations in their numbers? Biological basis of fluctuation from the viewpoint of evolutionary ecology. FAO Fish. Rep., 291,

53 Kawasaki, T. (1993). Recovery and collapse of the Far Eastern sardine. Fish. Oceanogr., 2, Kerr, R. A. (2000). A North Atlantic climate pacemaker for the centuries. Science, 288, Kifani, S. (1998). Climate dependant fluctuatiosn of the Moroccan sardine and their impact on fisheries. In Global versus local changes in upwelling systems, ed. M. H. Durand, P. Cury, R. Mendelssohn, C. Roy, A. Bakun and D. Pauly. Paris: Editions ORSTOM, pp King, J. R. (2005). Report of the Study Group on Fisheries and Ecosystem Responses to Recent Regime Shifts. PICES Scientific Rep Kondo, K. (1991). Interspecific relation between Japanese sardine and anchovy populations that reflects the essential mutual relation between fluctuation mechanisms of the two species based on organism-environment coupling. In Long-term variability of pelagic fish populations and their environment, ed. T. Kawasaki, S. Tanaka, Y. Toba and A. Taniguchi. Pergamon Press, pp Laevastu, T. (1993). Marine Climate, Weather and Fisheries. Oxford: Blackwell Scientific Publications Lavaniegos, B. E. and Ohman, M.D. (2003). Long term changes in pelagic tunicates of the California Current. Deep-Sea Res. II, 50,

54 Lavaniegos, B. E. and Ohman, M. D. (2007). Coherence of long-term variations of zooplankton in two sectors of the California Current system. Prog. Oceanogr., doi: /j.pocean Limsakul, A., Saino, T., Midorikawa, T. and Goes, J. (2001). Temporal variations in lower trophic level biological environments in the northwestern North Pacific Subtropical Gyre from 1950 to Prog. Oceanogr., 49, Lluch-Belda, D., Crawford, R. J. M., Kawasaki, T., MacCall, A. D., Parrish, R. H., Schwartzlose, R. A. and Smith, P. E. (1989). World-wide fluctuations of sardine and anchovy stocks: the regime problem. S. Afr. J. mar. Sci., 8, Lluch-Belda, D., Schwartzlose, R. A., Serra, R., Parrish, R., Kawasaki, T., Hedgecock, D. and Crawford, R. J. M. (1992). Sardine and anchovy regime fluctuations of abundance in four regions of the world oceans: a workshop report. Fish. Oceanogr., 1, Loeb, V. J. (1988). Report on the analysis of the Cape area historical ichthyoplankton data base. Unpubl. Rep., Sea Fisheries Research Institute, Cape Town Loeb, V.J. and Rojas, O. (1988). Interannual variation of ichthyoplankton composition and abundance relations off northern Chile, Fish. Bull., 86,

55 MacCall, A.D. (1996). Patterns of low-frequency variability in fish populations of the California Current. CalCOFI Rep., 37, MacCall, A. D. (2007). Mechanisms of low frequency fluctuations in sardine and anchovy populations (this volume) MacCall, A. D., Batchelder, H., King, J., Mackas, D., Mantua, N., McFarlane, G. A., Perry, I., Schweigert, J. and Schwing, F. (2005). Appendix 2: Recent ecosystem changes in the California Current system. In Report of the Study Group on Fisheries and Ecosystem responses to Recent Regime Shifts, ed. J. R. King. PICES Scientific Rep. 28, pp Marchal E. (1991). Location of the main West African pelagic stocks. In Variabilité, instabilité et changement dans les pêcheries ouest africaines, ed. Ph. Cury and C. Roy. Paris: ORSTOM, pp McFarlane, G. A. and Beamish, R. J. (2001). The re-occurrence of sardines off British Columbia characterises the dynamic nature of regimes. Prog. Oceanogr., 49, McFarlane, G. A., Smith, P. E., Baumgartner, T. R. and Hunter, J. R. (2002). Climate variability and Pacific sardine populations and fisheries. Am. Fish. Soc. Symp., 32, McGowan, J., Bograd, S. J., Lynn, R. J. and Miller, A. J. (2003). The biological response to the 1977 regime shift in the California Current. Deep-Sea Res. II, 50,

56 Mendelsohn, R. and Mendo, J. (1987). Exploratory analysis of anchoveta recruitment off Peru and related environmental series. In The Peruvian Anchoveta and its Upwelling Ecosystem: Three Decades of Change, ed. D. Pauly and I. Tsukayama. ICLARM Studies and Reviews 15, Mittelstaedt, E. (1983). The upwelling area off Northwest Africa. A description of phenomena related to coastal upwelling. Prog. Oceanog., 12, Montecinos, A., Purca, S. and Pizarro, O. (2003). Interannual-to-interdecadal sea surface temperature variability along the western coast of South America. Geophys. Res. Lett., 30 (11) 1570, ( ) Moses C. S., Swart, P. K. and Rosenheim, B. E. (2006). Evidence of multidecadal salinity variability in the eastern tropical North Atlantic. Paleoceanography, 21 (3), doi: /2005PA Nakata, H., Funakoshi, S. and Nakamura, M. (2000). Alternating dominance of postlarval sardine and anchovy caught by coastal fishery in relation to the Kuroshio meander in the Enshu-nada Sea. Fish. Oceanogr., 9, Nakata, K. and Hidaka, K. (2003). Decadal-scale variability in the Kuroshio marine ecosystem in winter. Fish. Oceanogr., 12, Nakata, K., Hada, A. and Matsukawa, Y. (1994). Variations in food abundance for Japanese sardine larvae related to the Kuroshio meander. Fish. Oceanogr., 3,

57 Noto, M. and Yasuda, I. (1999). Population decline of the Japanese sardine, Sardinops melanostictus, in relation to sea surface temperature in the Kuroshio Extension. Can. J. Fish. Aquat. Sci., 56, Noto, M. and Yasuda, I. (2003). Empirical biomass model for the Japanese sardine, Sardinops melanostictus, with sea surface temperature in the Kuroshio Extension. Fish. Oceanogr., 12, Odate, K. (1994). Zooplankton biomass and its long-term variation in the western North Pacific Ocean, Tohoku Sea area. Japan. Bull. Tohoku natl. Fish. Res. Inst., 56, (in Japanese, Engl. Abstr.) Ostrowski, M. and Strømme, T. (2004). Evolution of coastal SST in Northeast subtropical Atlantic and distribution patterns of small pelagic fish from Mauritania to Morocco. ICES Symposium on The Influence of Climate Change on North Atlantic Fish Stocks May 2004, Bergen (Poster B4) Palacios, D. M., Bograd, S. J., Mendelssohn, R. and Schwing, F. B Long-term and seasonal trends in stratification in the California Current. J. geophys. Res., 109, doi: /2004JC Pizarro, O. and Montecinos, A. (2004). Interdecadal variability of the thermocline along the west coast of South America. Geophys. Res. Lett., 31, L20307 (1-5). 1353

58 Polovina, J. J., Mitchum, G. T. and Evans, G. T. (1995). Decadal and basin-scale variation in mixed layer depth and the impact on biological production in the central and North Pacific, Deep-Sea Res. I, 42, Rebstock, G. A. (2001). Long-term stability of species composition in calanoid copepods off southern California. Mar. Ecol. Prog. Ser. 215, Rebstock, G. A. (2002). Climatic regime shifts and decadal-scale variability in calanoid copepod populations off southern California. Glob. Change Biol., 8, Roemmich, D. and McGowan, J. (1995). Climatic warming and the decline of zooplankton in the California Current. Science, 267, Roy, C. and Reason, C. (2001). ENSO related modulation of coastal upwelling in the eastern Atlantic. Prog. Oceanog., 49, Sætersdal, G., Bianchi, G., Strømme, T. and Venema, S. C. (1999). The DR. FRIDTJOF NANSEN Programme Investigations of fishery resources in developing countries. History of the programme and review of results. FAO Fish. Techn. Pap., Sakurai, Y., Kiyofuji, H., Saitoh, S., Goto, T. and Hiyama, Y. (2000). Changes in inferred spawning areas of Todarodes pacificus (Cephalopoda: Ommastrephidae) due to changing environmental conditions. ICES J. mar. Sci., 57,

59 Sánchez, S. (2000). Variación estacional e interannual de la biomasa fitoplanktonica y concentraciones de chlorofila a, frente a la costa Peruana durante Bol. Inst. Mar Perú-Callao, 19, Santander, H. and Flores, R. (1983). Los desoves y distribucion larval de quatro especies pelagicas y sus relaciones con las variaciones del ambiente marino frente al Peru. FAO Fish. Rep., 291, Schwartzlose, R. A., Alheit, J., Bakun, A., Baumgartner, T. R., Cloete, R., Crawford, R. J. M., Fletcher, W. J., Green-Ruiz, Y., Hagen, E., Kawasaki, T., Lluch-Belda, D., Lluch-Cota, S. E., MacCall, A. D., Matsuura, Y., Nevarez-Martinez, M.O., Parrish, R. H., Roy, C., Serra, R., Shust, K. V., Ward, M. N. And Zuzunaga, J. Z. (1999). Worldwide large-scale fluctuations of sardine and anchovy populations. S. Afr. J. mar. Sci., 21, Schwing, F. B., Mendelssohn, R. and Bograd, S. J. (2004). When did the 1976 regime shift occur? North Pacific Marine Science Organization, 13 th Annual Meeting, Honolulu, Hawaii, USA, abstract Schwing, F., Batchelder, H., Crawford, W., Mantua, N., Overland, J., Polovina, J. and Zhao, J.-P. (2005). Appendix 1. Decadal-scale climate events. In Report of the Study Group on Fisheries and Ecosystem Responses to Recent Regime Shifts, ed. J. R. King. PICES Scientific Rep. 28, pp Sedykh, K. A. (1978). The coastal upwelling off Northwest Africa. ICES. C.M. 1978/C:

60 Serra, J. R. (1983). Changes in the abundance of pelagic resources along the Chilean coast. FAO Fish. Rep., 291, Sirota, A., Chernyshkov, P. and Zhigalova, N. (2004). Water masses distribution, currents intensity and zooplankton assemblage off Northwest African coast. ICES CM 2004/N: Thomson, C. A., Grover, A. and Craig, W. L. (1985). Status of the California coastal pelagic fisheries. NMFS Southwest Region Admin. Rept. SWR p Tomosada, A. (1988). Long term variation of sardine catch and temperature. Bull. Tokai Reg. Fish. Res. Lab., 126, 1-9 (in Japanese, Engl. abstr.) Trenberth, K. E. and Stepaniak, D. P. (2001). Indices of El Niño evolution. J. Clim., 14, Tsukayama, I. (1983). Recursos pelagicos y sus pesquerias en el Peru. Rev. Com. Perm. Pac. Sur, 13, van der Lingen, C. D., Hutchings, L. and Field, J. G. (2006a). Comparative trophodynamics of anchovy Engraulis encrasicolus and sardine Sardinops sagax in the southern Benguela: are species alternations between anchovy and sardine in the southern Benguela trophodynamically mediated? Afr. J. mar. Sci., 28,

61 van der Lingen, C. D., Shannon, L. J., Cury, P., Kreiner, A., Moloney, C. L., Roux, J.-P. and Vaz-Velho, F. (2006b). Resource and ecosystem variability, including regime shifts, in the Benguela Current ecosystem. In Benguela: Predicting a Large Marine Ecosystem, ed. V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods. Amsterdam: Elsevier, pp van der Lingen, C. D., Bertrand, A., Bode, A., Brodeur, R., Cubillos, L., Espinoza, P., Friedland, K., Garrido, S., Irigoien, X., Möllmann, C., Rodriguez-Sanchez, R., Tanaka, H.and Temming, A. (2007). Trophic dynamics of small pelagic fish (this volume) Verheye, H.M Decadal-scale trends across several marine trophic levels in the southern Benguela upwelling system off South Africa. Ambio, 29(1): Verheye, H. M. and Richardson, A. J. (1998). Long-term increase in crustacean zooplankton abundance in the southern Benguela upwelling region ( ): bottom-up or top-down control? ICES J. mar. Sci., 55, Verheye, H. M., Richardson, A. J., Hutchings, L., Marska, G. and Gianakouras, D. (1998). Longterm trends in the abundance and community structure of coastal zooplankton in the southern Benguela system, S. Afr. J. mar. Sci., 19, Wada, T. and Jacobson, L.D. (1998). Regimes and stock recruitment relationships in Japanese sardine (Sardinops melanostictus), Can. J. Fish. Aquat. Sci., 55,

62 Watanabe, Y., Kurita, Y., Noto, M., Oozeki, Y. and Kitagawa, D. (2003). Growth and survival of Pacific saury (Cololabis saira) in the Kuroshio-Oyashio transitional waters. J. Oceanogr., 59, Weeks, S.J., Currie, B. and Bakun, A. (2002). Massive emissions of toxic gas in the Atlantic. Nature, 415, Yasuda, I. (2003). Hydrographic structure and variability in the Kuroshio-Oyashio Transition Area. J. Oceanogr., 59: Yasuda, I. and Watanabe, Y. (1994). On the relationship between the Oyashio front and saury fishing grounds in the north-western Pacific: a forecasting method for fishing ground locations. Fish. Oceanogr., 3, Yasuda, I., Tozuka, T., Noto, M. and Kouketsu, S. (2000). Heat balance and regime shifts of the mixed layer in the Kuroshio Extension. Prog. Oceanogr., 47, Yasuda, I., Sugisaki, H., Watanabe, Y., Minobe, S.-S. and Oozeki, Y. (1999). Interdecadal variations in Japanese sardine and ocean/climate. Fish. Oceanogr., 8, Yasuda, T. and Hanawa, K. (1997). Decadal changes in the mode waters in the midlatitude North Pacific. J. Phys. Oceanogr., 27,

63 Yasunaka, S. and Hanawa, K. (2002). Regime shifts found in the northern hemisphere SST field. J. meteor. Soc. Japan, 80, Yatsu, A. and Kaeriyama, M Linkages between coastal and open-ocean habitats and dynamics of Japanese stocks of chum salmon and Japanese sardine. Deep-Sea Res. II, 52, Yatsu, A., Watanabe, T., Ishida, M., Sugisaki, H. and Jacobson, L.D. (2005). Environmental effects on recruitment and productivity of Japanese sardine Sardinops melanostictus and chub mackerel Scomber japonicus with recommendations for management. Fish. Oceanogr., 14, Zuta, S., Tsukayama, I. and Villanueva, R. (1983). El ambiente marino y las fluctuaciones de las principales poblaciones pelagicas de la costa peruana. FAO Fish. Rep., 291, Zuzunaga, J. (1985). Cambios del equilibrio poblacional entre la anchoveta (Engraulis ringens) y la sardina (Sardinops sagax), en el sistema de afloramiento frente al Perú. In El Nin., Su Impacto en la Fauna Marina, ed. W. Arntz, A. Landa and J. Tarazona. Boletin Instituto del Mar del Perú-Callao, Volumen Extraordinario, pp

64

65 Table 5.1: Timing of events in Humboldt and Kuroshio Current ecosystems Year Biological changes HCE Biological changes KCE Physical changes HCE Physical changes KCE 1969 Per. zooplankt. 1 anch. shirasu 13 salinity 21 Kuroshio Ext. SST 25 Chil. zooplankt. 2 subtrop. surface water shift rel. abund. mesopel. larval fish 3 approaches coast 10 anch. surplus prod rel. abund. Chil. Sard. 5 sard. spawn. Tosa Bay 14 thermocline (model) 22 rel. abund. Chil. horse mackerel 5 anch. scale accumulation in anoxic SST anomalies 23 Chil. hake catches 6 sediments 15 turbulence mixing index 24 bonito catches 7 sqid catches 16 sard. spawn anch. biomass 9 sard. catches in Pac. 17 PDO anch. catches 10 sard. eggs Enshu-nada Sea 13 anch. recruitm. 11 saury catches 18 Oyashio summ. zoopl El Nino 1973 El Nino El Nino 1983 lowest anch. biomass ever 10 El Nino thermocline (model) anch. recruitment 9 SST anomalies 23 anch. biomass 9 salinity sard. catches 10

66 1986 offshore phyto- and micro- juvenile sard. 20 zooplankton 12 anch. shirau catches 13 sard. recruitment 20 squid catches coastal phytoplankton 12 anch. catches 17 PDO 1988 sard. shirasu 13 Kuroshio Ext. SST 25 sard. catches 17 1 Carrasco and Lozano 1989, 2 Bernal et al. 1983, 3 Loeb and Rojas 1988, 4 Jacobson et al. 2001, 8 Zuta et al. 1983, 9 Csirke et al. 1996, 10 Alheit and Niquen 2004, 11 Mendelsohn and Mendo 1987, 12 Sanchez et al. 2000, 13 Nakata et al. 2000, 14 Kawasaki 1993, 15 M. Kuwae, poster at ASLO Summer Meeting 2006, Victoria, Canada 16 Sakurai et al. 2000, 17 Schwartzlose et al. 1999, 18 Watanabe et al. 2003, 19 Chiba et al. 2006, 20 Yatsu and Kaeriyama 2005, 21 O. Moron, poster at Humboldt Current Symposium 2006, Lima, 22 Pizarro and Montecinos 2004, 23 Montecinos et al. 2003, 24 Bakun 1987, 25 Noto and Yasuda 2003

67 Figure legends Fig. 5.1: Normalized catches of anchovies (black line) and sardines (stipled line) from California Current, Japan, Humboldt Current, Northern and Southern Benguela Current and from NW Africa. Respective peak catches were set at 100%. California Current: anchovy peak-316*10 3 mt; sardine peak-718*10 3 mt. Japan: anchovy peak-430*10 3 mt; sardine peak-4 488*10 3 mt. Humboldt Current: anchovy peak *10 3 mt; sardine peak-5 621*10 3 mt. Northern Benguela: anchovy peak-376*10 3 mt; sardine peak-1400*10 3 mt. Southern Benguela: anchovy peak-597*10 3 mt; sardine peak-410*10 3 mt. Morocco: sardine peak-1087*10 3 mt; anchovy peak- 264*10 3 mt. Fig. 5.2: Minimum (left) and maximum (right) areas of distribution of Japanese, Humboldt Current and California Current sardines (redrawn after Hunter and Alheit 1995). Fig.5.3: Zooplankton time series. A: Copepods > 1mm in Kuroshio in winter (reprinted from Nakata and Hidaka 2003 with permission from ). B: Total zooplankton wet weight (mg m -3 ; May-July mean) in Oyashio region (reprinted from Chiba et al with permission from ). C: Meso-zooplankton volumes, excluding jelly fish, from Peru (redrawn after Ayón et al with permission from ). D: Zooplankton volumes from California Current (reprinted from Roemmich and McGowan 1995 with permission from AAAS).

68 Fig. 5.4: Annual catches of larval anchovy and sardine (shirasu fishery) in western Enshu-nada Sea (reprinted from Nakata et al with permission from ). Fig. 5.5: Dynamics of the Mixed Layer Depth in the Kuroshio Current Extension (reprinted from Yasuda et al with permission from ). Fig. 5.6: February SST anomalies in Kuroshio Extension Southern Area (broken line) and natural mortality coefficient anomalies of sardines, from postlarva to age 1, (solid line) (reprinted from: Noto, M. and Yasuda, I Population decline of the Japanese sardine, Sardinops melanostictus, in relation to sea surface temperature in the Kuroshio Extension. Can. J. Fish. Aquat. Sci., 56, , with permission from NRC Research Press). Fig. 5.7: Cumulative plot of seven years running mean of standardized annually averaged SST anomalies. The SST anomaly values of El Nino region 4 were subtracted from those of El Nino regions 1 and 2 and plotted cumulatively (E. Hagen, unpubl.). Fig. 5.8: Atlantic Multidecadal Oscillation and normalized catches of sardine and sardinella off NW Africa. Fig. 5.9: Normalized anchovy and sardine catches from Humboldt and Kuroshio Currents.

69 Figure 1.

70 Figure 2.

71 Figure 3.

72 Figure 4. Figure 5.

73 Figure 6. Figure 7.

74 Figure 8.