ICES Journal of Marine Science: Journal du Conseil Advance Access originally published online on January 31, 2007
ICES Journal of Marine Science: Journal du Conseil 2007 64(3):425-438; doi:10.1093/icesjms/fsl042
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Environmental variability in the North Atlantic and Iberian waters and its influence on horse mackerel (Trachurus trachurus) and albacore (Thunnus alalunga) dynamics
1 Instituto Español de Oceanografía, CO Santander, Apartado 240, 39080 Santander, Spain
2 Departamento de Ciencias y Técnicas del Agua y del Medio Ambiente, E.U.I.T Minera e Ingeniería Ambiental, 39316 Tanos (Torrelavega), Spain
3 Instituto Español de Oceanografía, CO Vigo, Apartado 1552, 36280 Vigo, Spain
Correspondence to A. Lavín: tel: +34 942 291060; fax: +34 942 275072; e-mail: alicia.lavin{at}st.ieo.es
Lavín, A., Moreno-Ventas, X., Ortiz de Zárate, V., Abaunza, P., and Cabanas, J. M. 2007. Environmental variability in the North Atlantic and Iberian waters and its influence on horse mackerel (Trachurus trachurus) and albacore (Thunnus alalunga) dynamics. – ICES Journal of Marine Science, 64: 425–438.We explore the potential impact of climatic and oceanic variables on the dynamics of horse mackerel Trachurus trachurus (coastal distribution) and albacore Thunnus alalunga (oceanic distribution). Principal components analysis of a set of environmental parameters for the years 1966–2000 allowed us to characterize the system by three components. The first consisted mainly of sea surface temperature (SST; 18.5% of variability), the second was determined by the oceanic transport indices, potential energy anomaly (PEA), and the Gulf Stream Index (15.6%), and the third by the meridional wind component and Ekman transport (11.5%). Horse mackerel recruitment was negatively correlated mainly with the first thermal component, whereas albacore age 3 catches were negatively correlated with the second oceanic component and positively with the third wind component. Multiple linear regression confirmed that environmental conditions [SST, PEA, and the zonal (east–west) wind component] explained the availability of age 3 albacore to the surface fisheries for the period 1975–1999. In contrast, cross-validation analysis showed that environmental conditions did not consistently explain horse mackerel recruitment, probably because of the short time-series available (15 y).
Keywords: albacore catch, environmental conditions, horse mackerel recruitment, multivariate analysis, North Atlantic, oceanographic indices, sea surface temperature
Received 3 July 2006; accepted 30 November 2006; advance access publication 31 January 2007.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
Meteorological and oceanic parameters display variability over a range of scales, and this variability potentially influences the living marine resources at different trophic levels and in different ecosystems, particularly pelagic ecosystems, where the atmosphere/ocean interrelationship is close (Pitcher, 1995). The impact of different environmental conditions on fish populations and fisheries has been studied in depth (e.g. Beamish and McFarlane, 1989; Beamish, 1995; Cushing, 1995; Reid, 2001). Events such as depletion or increase in abundance of fish species, e.g. clupeoids or flatfish, are often synchronous over geographically broad and widely separated areas, suggesting large-scale climate forcing (Lluch-Belda et al., 1989; Cushing, 1995; Schwartzlose et al., 1999). In addition to large-scale forcing, there may be local or regional events, such as coastal upwelling or low-range thermohaline-forced currents, which can contribute extensively to recruitment variability. Most pelagic fish species inhabit areas where turbulence dominates their environment, because it acts on the advection and retention of larvae and so influences recruitment success (Lasker, 1975; Cury and Roy, 1989; Bakun, 1996; Borja et al., 1996).
Environmental conditions can be represented by indices of upwelling, turbulence, sea surface temperature (SST), wind or oceanic large or mesoscale circulation, and atmospheric factors. The distribution of these factors in time and space influences the distribution of fish spawning, larva retention areas, and food production, and hence the availability of food to fish larvae and juveniles.
The North Atlantic Oscillation (NAO) index accounts for much of the atmospheric variability in the North Atlantic (Rogers, 1984; Hurrell, 1995) and is therefore a dominant exogenous driving factor for biological systems. Interannual to interdecadal variability in the northern North Atlantic is reflected in periods of strong and weak convective mixing and in the NAO index (Dickson et al., 1996). An oceanic index, estimated similarly to the atmospheric NAO index, was presented by Curry and McCartney (2001) as a two-point baroclinic pressure difference between the subtropical and sub-polar gyre centres. This index, the potential energy anomaly (PEA), provides information on the magnitude of geostrophic velocities and mass transports by marine currents over the central and northern North Atlantic. A second index of oceanic circulation is the position of the northern wall of the Gulf Stream (Taylor, 1996).
There are two different wind regimes off the western and northern Iberian Peninsula: autumn/winter, dominated by southwesterly wind, and spring/summer, dominated by northeasterly wind. There are also two regimes in the marine current patterns of the area. In autumn, the Iberian poleward current develops and occasionally reaches the Cantabrian shelf-slope and even further north to the French shelf-slope (Sánchez and Gil, 2000; Ruiz-Villarreal et al., 2006). In summer, upwelling dominates the entire area and is stronger in the western part and diminishes through the Cantabrian Sea (Lavín et al., 2005). SST distributions show a clear picture of the different features. Warmer winter water announces the arrival of the poleward current, and low summer temperatures indicate strong upwelling. Winds are the forcing or enhancing mechanism of both events, upwelling and the arrival of the Iberian poleward current (Wooster et al., 1976; Blanton et al., 1984; Lavín et al., 1991, 2000; Cabanas et al., 2003). Occasionally, poleward current episodes develop out of season, and if this happens during fish spawning seasons, they can have a large influence on the survival of fish larvae (Villamor et al., 2004).
A synoptic description of the main circulation patterns in the Bay of Biscay and nearby North Atlantic, composed from several sources of information, is presented in Figure 1. The main North Atlantic current (NAC) and the Azores current (AC) are represented by bold arrows. Thin arrows indicate less permanent local characteristic currents (continuously detaching from the NAC), the southeastern flux from the Goban Spur to the Bay of Biscay (Pingree, 1993), the main entry of Eastern North Atlantic Central Water (ENACW) into the Bay of Biscay from the west and anticyclonic circulation inside (Colas, 2003), and the retention area west of the Iberian Peninsula and the Bay of Biscay, where ENACW is formed (Pollard et al., 1996). The Iberian poleward current and its possible entry into the Bay of Biscay as a continuation of the slope current are represented by a dashed arrow (González-Pola et al., 2005).
|
The most direct effect of climate on fish stocks is to increase or diminish the magnitude of recruitment over a period of time (Cushing, 1995). Horse mackerel (Trachurus trachurus) is a commercial species widely distributed in the Northeast Atlantic and Mediterranean Sea that supported a catch of about 275 000 t in 2000, an important source of income for local economies (FAO, 2000). Several authors have considered the influence of environmental factors on aspects related to horse mackerel recruitment, e.g. the onset of spawning (Barenbeim, 1974), the development and hatching of eggs (Pipe and Walker, 1987), and the seasonal distribution of pelagic eggs at sea (Solá et al., 1990). Recently, Santos et al. (2001) analysed the influence of upwelling along the Portuguese coast on recruitment (year-class strength) to the southern horse mackerel stock, but no predictive model was developed. One of the main problems in fish stock assessment relates to the prediction of recruitment and the subsequent forecasting of catches (Shepherd, 1997). Therefore, one of our aims here is to explore the environmental processes that influence horse mackerel recruitment, identifying the key factors, and on the basis of these factors to design a multiple linear regression model to estimate incoming recruitment (recruitment in the most recent assessment year). The cross-validation technique described by Hastie and Tibshirani (1990) and Francis (2006) was used to test the predictive power of the proposed models.
Atlantic albacore, Thunnus alalunga (Bonnaterre, 1788), is a highly migratory species found in temperate and subtropical waters throughout the Atlantic (Bard, 1981). At present, the hypothesis is of a two-stock structure for Atlantic albacore, the northern and southern stocks being separated at 5°N for purposes of stock assessment (Anon. 2001). Immature albacore, age classes 1–4, begin to migrate in May from the central Atlantic around the Azores towards their feeding grounds in the Bay of Biscay, then following the food towards the southwest coast of Ireland, along two main migratory paths (Hue, 1980; Bard, 1981). This trophic migration is associated with the thermal structure (i.e. thermal fronts, SST) of the eastern North Atlantic temperate water masses north of 35°N, which fluctuates between years (Dao and Bard, 1971; Aloncle and Delaporte, 1982; Leroy, 1990). Migrating immature albacore are restricted mostly to areas of SST between 15°C and 20°C in the NE Atlantic (Aloncle and Delaporte, 1973; Havard-Duclos, 1973), and they concentrate in the Bay of Biscay and adjacent waters during spring and summer (Aloncle and Delaporte, 1976; Bard and Santiago, 1999). Albacore are exploited by the surface fishery in these two areas, where different fleets have been operating for several decades (Anon. 2001). Standardized catch per unit effort (cpue) of 2- and 3-year-old fish taken by the commercial surface fleets, troll, and bait boats (Goujon et al., 1996; Ortiz de Zárate et al., 2001) has been used to calibrate a fit of virtual population analysis (VPA) to catch-at-age data, to assess the state of the North Atlantic albacore stock (Anon. 2001). These commercial fishery indices, considered to be relative abundance indices by age of the stock, have shown conflicting trends over time, and one working hypothesis has been that they may represent catchability or availability of the different albacore age groups to the different fleets, rather than stock abundance per se (Anon. 2001). A significant (at the 5% level) negative cross-correlation between catches of juvenile albacore (2- and 3-group) in the North Atlantic and the Gulf Stream latitude index has been found (Ortiz de Zárate et al., 1997). Highly significant positive correlations (at the 1% level) have been demonstrated between the NAO index and average SST anomalies in summer in the Bay of Biscay fishing grounds (Bard and Santiago, 1999). Large year-on-year fluctuations of oceanic phenomena associated with the NAO index and the Gulf Stream Index (GSI) have been positively correlated with catches of 2- and 3-group albacore (Lavín et al., 1999).
Here, we explore the potential link between environmental variables, such as climate and oceanographic indices, and recruitment trends in horse mackerel, as well as the influence of the same environmental indices on catches of immature albacore taken by the surface fishery in the North Atlantic. These two species were selected to investigate our hypotheses because both are pelagic migratory species within the coastal and oceanic domain exposed to climate change in the North Atlantic Ocean and because they represent valuable species for local fishers.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
Meteorological and oceanographic data
The region of interest (Figure 1) is the northern part of the subtropical Gyre and the Atlantic temperate area around the Iberian Peninsula. Data for 1966–2000 were used in this analysis (Figure 2). The NAO index was obtained from Rogers (1990) The PEA oceanic index (Curry and McCartney, 2001) was provided by R. Curry (WHOI). Both are shown in Figure 2a. To fill two data gaps (from 1979 to 1980 and from 1982 to 1984), a cubic spline interpolation was performed (Figure 2a). The GSI was taken from Taylor (1996). Table 1 lists the data sets, sources, and acronyms used in our analysis. The average GSI from autumn of the previous year to the end of winter (October–March) was used because it coincides with part of the horse mackerel spawning season.
|
|
To take into account the spatial conditions, we selected two areas where hydrographic characteristics are rather different: west of the Iberian Peninsula, represented by the point 43°N 11°W; and the Cantabrian Sea (north of the Iberian Peninsula), represented by the point 45°N 3°W. As described previously, there are two main seasons: spring/summer, characterized by near-surface stratified waters; and autumn/winter, with mixed waters. To include this variability, the values for spring/summer (noted by s) are separated from the average annual values (noted by a) for most of the variables used. The spawning season of horse mackerel is a long one (Abaunza et al., 2003a), so it is useful to consider the annual mean values of the environmental variables. Albacore migration towards the Bay of Biscay takes place in spring and summer, so this period is also included in the analysis. SST and winds at the two points in different seasons were taken from the Comprehensive Ocean Atmosphere Data Set (COADS) (Woodruff et al., 1993), available at http://www.cdc.noaa.gov/coads/coads1a.html (locations are shown in Figure 1). Figure 2b presents the SST annual average time-series at 43°N 11°W and at 45°N 3°W. Ekman transport based on wind forcing (Bakun, 1973) was used at the western point (43°N 11°W), extending the data set of Lavín et al. (2000) (Figure 2c). The upwelling index is calculated as the offshore Ekman transport during spring and summer.
Biological data
Estimates of horse mackerel recruitment for the period 1985–1999 were taken from the ICES working group assessments (ICES, 2003) applying XSA (extended survivors analysis; Shepherd, 1999) (Figure 2b). The recruitment estimate for 2000 was not considered because estimates are generally unreliable for the most recent assessment year. Estimates of horse mackerel spawning-stock biomass (SSB) for the years 1985–1999 were also considered, to explore correlations with recruitment. A new stock definition for horse mackerel has recently been applied to the Northeast Atlantic (ICES, 2006). For exploratory purposes, the recruitment estimates of this newly defined stock have also been considered. Those estimates come from the application of an ad hoc age-structured model (ICES, 2006).
The albacore surface fishery in the eastern North Atlantic has been documented since 1920 (Bard and Santiago, 1999). However, for assessment purposes, only the period 1975–1999 is considered suitable for dynamic estimates of population parameters of the North Atlantic albacore stock (Anon, 1996, 2001).
Catch-at-age for the northern albacore stock for the period 1975–1999 was estimated by Santiago and Arrizabalaga (2001) (Figure 3a) for the North Atlantic stock (Anon. 2001). Most catches are of immature albacore aged 1–4. Most of these catches are taken by the surface fishery in the study area, with longline fishery catches taking a minor share because such fleets target the adult population (>5 y old) in deeper water (
200 m) of the central Atlantic (Figure 3b). The surface fleets do not target albacore of age group 1 (Mejuto et al., 1992; Ortiz de Zárate and Cramer, 2001; Ortiz de Zárate et al., 2001). In contrast, age 2 and 3 albacore are targeted by the surface fleets, live bait boats, and midwater pelagic trawlers, mainly in the Bay of Biscay, and by driftnetters and trollers in the adjacent eastern North Atlantic (Anon. 1996, 2001). The presence of age group 4 on the fishing grounds is variable from year to year, but mainly in autumn in the Bay of Biscay fishery. Therefore, only the total annual catch for age groups 2 and 3 was used in this study. Selected biological variables are listed in Table 1.
|
Statistical analysis
Conformity of the time-series to a normal distribution was confirmed using a Kolmogorov–Smirnov test (
= 0.05).
Principal components analysis
Principal components analysis (PCA) is used to reduce the dimensionality of the set of environmental variables and to characterize variability in the data (Table 1). The number of eigenvalues considered was determined using the observed communalities and the eigenvalue screen plot; only components with an explained variance of at least 10% were retained. To seek a simple structure, i.e. a solution where causal relationships between the underlying common factors and the observed descriptors is simpler, we applied an orthogonal rotation using the varimax method (Legendre and Legendre, 1998). The rotation of the matrix of factors sorts the variance from the first eigenvalue to the last, to obtain a simpler and more significant pattern of variables in the data set (Hair et al., 1999).
Multiple linear regression analysis
The set of environmental variables for this analysis was the same as for the PCA (Table 1). For the horse mackerel recruitment time-series, the number of cases was smaller than the number of independent variables, so the number of variables was reduced. To do so, pairwise correlations between independent variables were calculated. When the simple correlation (r) was >0.6, the variable with the higher correlation with horse mackerel recruitment was selected, reducing the number of variables from 23 to 18. A similar method was used for albacore, leading to 11 explanatory variables. Final variable selection was carried out using backward elimination (Draper and Smith, 1998). For horse mackerel, the variance–covariance matrix was singular, so we finally applied a stepwise elimination method, which avoided this difficulty. Colinearity diagnostics provided by the SPSS software (Marija, 1993), such as the tolerance, the variance inflation factor, the eigenvalue, and the variance proportion explained, were also analysed at each step. Factors not significant at the 5% level were removed from the model. The residuals of the selected models were tested for autocorrelation and deviation from normality by applying the Durbin–Watson test (Draper and Smith, 1998).
To assess the performance of the horse mackerel recruitment prediction model, we applied cross-validation. The predictor variables used in the cross-validation (18 variables) were those selected in the pre-screening procedure using correlation analysis (see earlier). For cross-validation, we followed the method of Francis (2006), applying a repeated leave-one-out procedure in which we: (i) dropped all data for year i from the predictors and recruitment; (ii) applied a stepwise regression to select the best predictors; (iii) calculated a regression equation using these best predictors; and (iv) used this equation to predict the recruitment in year i. In the stepwise regression, the maximum p-value considered to add the variables to the model was 0.05, and the minimum p-value to remove a variable was 0.10. The per cent variance explained statistic (PVE; Francis 2006) was calculated to measure how much better a regression-based estimator of recruitment was than the default (arithmetic mean) estimator. PVE is defined here as:
|
|
i–ri)2 and ri and i are the true (in this case from the XSA) and estimated recruitments, respectively, in year i. The parameter i is calculated from the regression equation in MSEregression and from the arithmetic mean [meanj
i(rj)] in MSEdefault. All statistical analyses were performed using SPSS 13.0.1 and MATLAB 7.0. |
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
Principal components analysis
Environmental factors
Three principal components (PCs) from the rotated factors were selected as meaningful, accounting for 45.6% of the total variance. Correlation estimates for the three PCs are presented in the Appendix. The first component explained 18.5% of the total variability and was positively related to SST off the western Iberian Peninsula (43°N 11°W) and in the Cantabrian Sea (45°N 3°W) for the annual period and the summer season. The second component was related to the oceanic transport indices, i.e. PEA and annual, summer, and winter GSI, and explained 15.6% of the system variance. The third PC explained 11.4% of the total variation and was positively correlated with the annual "V" (north/south wind component) in the two sampling locations, 45°N 3°W (Cantabrian Sea) and 43°N 11°W (western Iberian Peninsula), and with the annual Ekman transport at the second location. Each of the remaining PCs explained <10% of the system's variability. The distribution of time-series (y) as a function of the first and second components is presented in Figure 4. Periods of extreme values of oceanic transport represented by the PEA index appear to be approximately separated.
|
Environmental factors and horse mackerel recruitment
The analysis of horse mackerel recruitment vs. the first PC, the thermal one, revealed a significant negative correlation (r = –0.74; p = 0.002) (Figure 5a). There was a weak correlation between horse mackerel recruitment and the second and third PCs, 0.37 (p = 0.18) and –0.21 (p = 0.45), respectively. Years with lower SST favoured successful recruitment, as in 1986 and 1991. There was an irregular downward trend in recruitment strength over the time-series. The lower values were during the last three years (1997–1999), coinciding with the higher values of the first PC.
|
The correlation between SSB and recruitment strength at age 0 was not significant.
The multiple regression analysis for horse mackerel recruitment selected two environmental variables: spring/summer SST at 43°N 11°W and spring/summer meridional wind component at 45°N 3°W ("V", the northerly wind). With these two variables, the model explained 65.5% (adjusted r2) of the recruitment variability in the time-series (Table 2). The collinearity statistics showed no major problems, and the value obtained for the Durbin–Watson test (D = 1.342) was inconclusive for a two-sided test about the acceptance or rejection of H0 (where H0 = no autocorrelation in the residuals). The variable with most weight in the model was SST for spring/summer at 43°N 11°W, as shown by the higher value of the standardized coefficient "beta" in comparison with the wind component V (Table 2).
|
The cross-validation method showed that the correlation of horse mackerel recruitment with environmental variables was strong by chance. The estimated PVE value was negative (–21.71), meaning that the regression estimator of recruitment is no better than the default estimator (the arithmetic mean).
During the repeated leave-one-out procedure of cross-validation, the stepwise screening method selected the variables SST4311s and V4503s in 11 of 15 y. These two variables were also selected, but in the company of others, in 1993 and 1996. In 1988, only variable SST4311s was selected, and in 1986, no variables were selected and the default recruitment (arithmetic mean) was adopted for that year. The adjusted r2 was always greater than 0.59 in all cases except 1986, where the default estimate was considered.
Environmental factors and albacore catches
The first PC representing the effect of temperature was not significantly correlated with albacore aged 2 and 3. The second PC, the oceanic one, which includes variables such as the oceanic transport indices PEA and GSI, had no significant correlation with the catch of 2-group albacore, but was significantly negatively correlated with the catch of 3-group albacore (r = –0.42; p = 0.035) (Figure 5b). Likewise, there was a positive significant correlation with catches of age 3 albacore (r = 0.42; p = 0.035) for the third meridional wind ("V", northerly wind) component. In general, the catch of age 3 albacore exhibited a decreasing trend over time, correlated with the oceanic component. In a broad sense, years with negative values of the oceanic component, seen in the PEA and GSI indices, corresponded to larger catches of age 3 albacore and were recorded during the earlier decades of the period analysed, the 1970s and the 1980s, whereas the positive values in the time-series of the oceanic component occurred together with lower catches of age 3 albacore in the 1990s. For 2 years, 1996 and 1998, the catches failed to match the general relationship.
Multiple linear regression analysis fitted to catches of age 2 albacore showed that none of the environmental variables could explain the catches. However, catches of age 3 albacore analysed in the same manner were significantly correlated with four variables: oceanic transport (PEA), the zonal wind component in summer and annually at 43°N 11°W, respectively (U4311s, U4311a), and summer SST at 45°N 3°W (SST4503s) (Table 3). This final model explained 51.7% (adjusted r2) of the variability in the 3-group albacore time-series of catches. The standardized coefficients of the predictors revealed that oceanic transport (PEA) produced the highest negative weighting in the model, followed by the negative weighting of SST in the Bay of Biscay during summer (SST4503s), and the zonal wind component in summer in the eastern Atlantic (U4311s). Although the coefficient of the annual zonal wind effect (U4311a) could not significantly explain the catch of albacore aged 3, when including this effect the model explained more variance than an otherwise identical model in which U4311a was removed. The model was furthermore meaningful in explaining the albacore hypothesis being tested. The collinearity statistical diagnostics showed no major problems, and the value for the Durbin–Watson test (D = 2.408) was again inconclusive for a two–sided equal–tailed test, about the acceptance or rejection of the null hypothesis H0 that there was no autocorrelation in the residuals.
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
The two first PCs of the environmental variables provide a useful summary of system fluctuations (Figure 4). The two periods of extreme values of oceanic transport (PEA), low values (1965–1974) and high values (1990–1997), appear to be separated. A low transport index corresponds to the negative values of the first (thermal) and the second (oceanic transport) PCs, and a high transport index (e.g. during the 1990s) to positive values of both components. As the study area (Figure 1) is located in the intergyre region between the sub-polar gyre (NAC), the subtropical gyre (AC), and the European shelf, mean circulation is weak (Maillard, 1996), although it could be influenced by changes in the strength of both main currents.
The first PC or thermal axis represents the local pattern in the Bay of Biscay. The second PC represents oceanic transport variability. A comparison of periods of low and high transport, as done by Curry and McCartney (2001) for PEA fields at 200 dbar relative to 2000 dbar, reflects the strong transport of the 1990s, uplift in the entire subtropical dome west of 40°W, and a deepening of the sub-polar bowl. In addition to the north–south dipole, changes in east–west circulation indicate an opposite sign. This coincidence of high temperature and strong transport could denote more subtropical influence in the area in years of high PEA. The mode of variability reflects changes in the circulation in the North Atlantic.
The case of horse mackerel
Rates of natural mortality of marine organisms in pelagic ecosystems are strongly size-dependent (Houde, 2002), so mortality rates are greatest early in life, then decrease at later stages of development. The causes of mortality in early life are varied and may include: starvation and nutritional deficiency, predation, physical oceanographic processes, unsuitable water quality in nursery areas, and disease (Houde, 2002). Climate factors are the origin of many of these causes, providing a basis for establishing a relationship between such factors and recruitment. Temperature is in part an environmental proxy for other physical factors, such as upwelling, wind systems, and currents, and many fluctuations in the marine ecosystem have been correlated with temperature variations (Laevastu, 1993). Temperature is the most potent environmental regulator of fish physiology, e.g. regarding spawning time via its influence on the rate of gonad maturation (Lange and Greve, 1997) or embryonic development (Fuiman, 2002).
Years with cooler SSTs near the coast during spring and summer seem to favour horse mackerel recruitment. The cooler temperatures are also in the range of best survival and development of horse mackerel eggs, which are pelagic until hatching: 12.2–15.8°C (Pipe and Walker, 1987). Trachurus trachurus has a long spawning season over the Iberian shelf, covering almost the first 8 months of the year, peaking in winter and spring on the Portuguese coast and in spring/summer off northern Spain (Solá et al., 1990; Borges and Gordo, 1991; Abaunza et al., 2003a). The cooler temperatures are a clear indication of upwelling and less stormy weather, precisely when the pelagic eggs and larvae are more abundant. Therefore, the food supply for larvae is secure, and survival of some fish species would be favoured (Cushing, 1995). There is also a strong temperature-dependent regulation of food intake through appetite in adult horse mackerel, with a doubling of the intake rate at 13°C compared with that at 10.7°C (Temming and Herrmann, 2001). This could be also related to fish condition before spawning, and hence with reproductive potential (Abaunza et al., 2003a). In this sense, the high negative correlation between the first PC, the thermal one, and horse mackerel recruitment (Figure 5a) makes biological sense.
Variable V at 45°N 3°W (spring/summer meridional wind) included in the regression model has a negative correlation, meaning that northerly winds favour horse mackerel recruitment. This is probably because in the Bay of Biscay, northerly winds bring cooler air temperatures, which favour heat transfer from the ocean to the atmosphere, so decreasing SST, itself propitious to horse mackerel recruitment. In the western part of the study area, northerly winds generate upwelling events during spring and summer, but this phenomenon should be more appropriate to the sampling station at 43°N 11°W. In addition, throughout the Cantabrian Sea (which is longitudinally orientated), northerly wind could cause young stages of fish to be retained over the continental shelf, where the environment may be more favourable for their survival. It is in spring and summer when most horse mackerel eggs and larvae are present in the Cantabrian Sea (Solá et al., 1990).
The finding that the linear regression model explained 65.5% of recruitment variability with three environmental variables could represent a significant step towards resolving one of the central problems in short-term forecasting of catch and biomass: the estimated strength of incoming year classes (Shepherd and Pope, 2002). However, one of the common errors in statistics is to use the same data to select variables for inclusion in a model as to assess their significance (Good and Hardin, 2003). Cross-validation is one of the appropriate approaches to avoid this error. The PVE statistics obtained (–20.71) was negative, so the regression model did not show a superior ability to predict horse mackerel recruitment over the application of a simple arithmetic mean. However, the stepwise screening procedure through the different years almost always selected the same two variables: SST4311s and V4503s, with an adjusted r2 always >0.59. This result accords with the high negative correlation value obtained between horse mackerel recruitment and the first PC. These arguments led us to think that one of the main issues in the reliability of the model is the length of the time-series. Francis (2006) demonstrated that an ability to detect environment–recruitment relationships depends to a great extent on the length of the time-series. With simulated data, he showed that with a time-series of 13 y, it was almost impossible to detect any environment–recruitment relationship, so the minimum number of years to detect such relationships could be close to 23 y. Our time-series spans 15 y, shorter than the ideal minimum time-series. Clearly, the predictive capacity of the model could be tested with more certainty as the time-series expands (Hair et al., 1999).
Without understanding the mechanistic connections between the variables being correlated, we cannot be confident that correlations will continue into the future (Gargett et al., 2001). SST or other temperature measures and wind components are factors that could be taken into account in modelling horse mackerel recruitment, as explained earlier. In any case, models based solely on environmental conditions could be inadequate. Often, there are decadal or temporary changes in oceanographic conditions, which generate new dynamics in living marine resources, associated with the complexity of ecosystem interactions. Therefore, the factors intervening in the models proposed here will require continuing revalidation outside the study period (Pepin, 2002), and if necessary an appropriate solution will need to be sought. Another complication is the possibility of spurious correlations between estimated fish recruitment and environmental factors attributable to errors in the rate of natural mortality used in the VPA (Lapointe and Peterman, 1991). This remains an unresolved problem in stock assessment, in which the most realistic and robust estimates of population parameters are required for stock management.
Recently, an EU-funded project on horse mackerel stock identification (Abaunza et al., 2003b) showed that the currently accepted boundaries of the southern horse mackerel stock are probably incorrect. Therefore, the time-series of recruitment estimates have changed (ICES, 2006). We believe that our proposed methodology for estimating incoming recruitment will nevertheless remain useful for the species. An exploratory analysis made with the preliminary recruitment estimates of the new stock unit, in this case the so-called "western stock" (ICES, 2006), and the environmental variables considered in this study showed that the best correlations were also obtained with SST variables at 45°N and 43°N (r = –0.4 and –0.36, respectively), spring–summer V at 43°N (r = –0.43), annual U at 45° (r = 0.51), and annual Qx at 43°N (r = –0.36). One must take into consideration the fact that the sampling areas used in this study are now located at the very southern limit of the newly defined western horse mackerel stock, distributed from the north of Spain to Norway. Therefore, there is not much sense in designing a multivariate regression model with predictive purposes with the data available for the newly defined Western stock. Exploratory trials showed that the best model had only one significant explanatory variable, the annual "U" (zonal wind component) at 45°N, which explained 20% (adjusted r2) of the recruitment variability. This could be because westerly winds are likely the most common environmental variable that has influence throughout the European Atlantic coast, from Norway to Portugal.
In summary, our results have confirmed the advice by Francis (2006) to use a validation technique when developing predictive recruitment models. Moreover, the horse mackerel case showed that extension of the time-series determines the ability of the regression model to detect environment–recruitment relationships.
The case of albacore
A significant association has been identified between the large-scale, low-frequency oceanic changes in the North Atlantic and their effects on catches of albacore aged 3, harvested largely in the surface fishery of the eastern North Atlantic (Figure 1). The results showed a significant negative correlation between the PEA and GSI oceanic transport indices and catches of age 3 albacore. Significant correlations in raw data time-series seem to indicate that the response of the albacore population to the PEA index probably takes place with some inertia (Ottersen et al., 2001). This relationship highlights some sort of mutual pattern between the increase in eastward oceanic transport (PEA index), limited by boundaries of the sub-polar gyre and the subtropical gyre, and the abundance of albacore aged 3 in the eastern North Atlantic.
However, no significant relationships were found when regressing catches of age 2 albacore with either the oceanic variables or the more local variables, such as SST, which is summarized by the first PC in the analyses. Therefore, the time trend in the catch of age 2 albacore could not be explained by the environmental variables studies here. Unfortunately, it is not possible to investigate this behavioural aspect further because appropriate data are lacking.
Albacore spawn in the western Atlantic during spring and summer in part of the Sargasso Sea, in tropical waters (Shiohama, 1971; Ueyanagi, 1971; Uozumi, 1996). At a juvenile stage, albacore grow rapidly, attaining 30 cm at 6 months, and soon begin to migrate east to the feeding grounds in the eastern North Atlantic at age 1, when they recruit to the surface fishery.
The hypothesized change in environmentally driven availability of the resource to the surface fishery by influencing the spatial distribution of seasonal migration was explained to some extent by the regression model (Table 3). Consequently, the decadal PEA index's higher values in the 1990s might have implied a more northward distribution of age 3 albacore, which points to changes in the annual migration routes. Only for negative PEA values do juvenile albacore seem to migrate towards the eastern North Atlantic temperate areas and reach the Bay of Biscay grounds in summer.
Concerning the possible negative effect of SST on the availability of age 3 albacore, the warm anomalies (SST4503s) during summer in the southern (45°N 3°W) Bay of Biscay could act as a limiting distribution factor given that the most suitable range of SST (17–20°C) for immature albacore in the adjacent waters of the North Atlantic precludes migration to warmer Bay of Biscay waters (Aloncle and Delaporte, 1973; Havard-Duclos, 1973; Leroy, 1990). The different age classes of immature fish that congregate on the summer fishing grounds (Figure 1) are distributed according to different SSTs of 17–21°C, and the 16°C isotherm seems to be the lower limiting temperature (Allain and Aloncle, 1968). Other authors (Bard and Santiago, 1999) found positive significant correlations (p = 0.001) between the winter NAO index and SST anomalies for the period June–August of 1986–1993 in the eastern North Atlantic, including the Bay of Biscay, that would impact albacore distribution in this area. Likewise, in the Pacific Ocean, the migration patterns of albacore were described as being wider in El Niño years, an area associated with the presence of a cold-water region in the central and southwestern North Pacific (Kimura et al., 1997).
Finally, the negative relationship with the zonal (east–west) wind component in summer (U4311s) indicated that increases in turbulence and mixing produced by westerly wind have a negative effect. In contrast, Ekman transport and upwelling events induced by easterly wind in the Bay of Biscay had a positive effect on catches of age 3 albacore. Recently, Goñi and Arrizabalaga (2005), in looking for a statistical relationship between standardized annual cpues of albacore aged 2 and 3 in the Spanish baitboat and troll fleets (Ortiz de Zárate et al., 2001; Ortiz de Zárate and Cramer, 2001) and environmental variables, using multiple linear regression and generalized least squares models, found a significant negative relationship between the 2-group albacore cpue time-series for troll and baitboat fleets and sea state and insolation in the Bay of Biscay.
A descriptive analysis of nominal catch rates (cpue) by age (1–4) of the Spanish troll and baitboat fleets targeting immature albacore during two consecutive years suggests a very high spatial and temporal variability of cpue within and between years. Higher values of cpue of albacore aged 2 and 3 have been found for the eastern North Atlantic fishing areas by Mejuto and García (1993). Nevertheless, both temperature (SST) and forage density play a key role in the distribution of albacore, which being an opportunistic feeder, concentrates in areas where prey are abundant and visibility is good. Fiedler and Bernard (1987) found for the eastern Pacific that immature albacore migrating in summer associated with thermal fronts. Also, immature albacore migrate to the eastern Pacific off California (USA) during summer (Laurs and Lynn, 1991). Those authors found that in years with strong temperature fronts, albacore were more concentrated, indicating that thermal gradients are important in the process of stimulating aggregation.
The underlying mechanisms linking climate and SST as a "proxy" to migrations of immature albacore and the distribution of these mechanisms in the eastern North Atlantic are yet to be investigated, ideally with independent data sets. The use of an oceanic transport index, such as the PEA, which reflects the cumulative nature of the ocean and integrates atmospheric (NAO index) and thermal (SST) factors, seems likely to be more useful than using single variables when searching for environmental indices related to fisheries, especially for species with broadly oceanic behaviour, such as North Atlantic albacore.
|
Conclusions |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
Taking into account the effect of SST on both horse mackerel recruitment and the albacore fishery, long-term variability in SST in the area could be playing the dominant role. The upper waters of the Bay of Biscay have experienced progressive warming during the past and the present century. Mean surface water temperatures increased by 1.4°C in the south-eastern Bay of Biscay over the period 1972–1993 (0.6°C per decade) and by 1.03°C over the past century (Koutsikopoulos et al., 1998; Planque et al., 2003). The increase in heat content stored in the water column appears to be greatest in the 200–300 m layer (González-Pola and Lavín, 2003), and it is in this layer that Eastern North Atlantic Central Waters (ENACW) respond quickly to climate forcing in areas of water mass formation located in the northern Bay of Biscay and adjacent areas. In the southern Bay of Biscay (around our location, at 45°N 3°W), temperature has increased during the last decade in the ENACW by 0.032°C y–1 (González-Pola et al., 2005).
|
Appendix |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
Contribution of each environmental variable to the principal components.
|
|
Acknowledgements |
|---|
We thank R. Curry (WHOI) for providing the PEA data, J. Hurrell and R. Dickson for the NAO index, A. Taylor for the GSI and the COADS, and J. Carranza, I. Huskin, and E. Gallo for their collaboration. Additionally, we are very grateful to two anonymous referees for their thoughtful suggestions on how best to present the data, to R. I. C. C. Francis for his valuable advice on the cross-validation analyses, and to editor V. Trenkel for her extensive editorial assistance.
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionConclusionsAppendixReferences |
|---|
-
Abaunza P., Gordo L., Karlou-Riga C., Murta A., Eltink A. T. G. W., García Santamaría M. T., Zimmermann C.,, et al. (2003a) Growth and reproduction of horse mackerel, Trachurus trachurus (Carangidae). Reviews in Fish Biology and Fisheries 13:27–61.[CrossRef][Web of Science]
Abaunza P., Murta A., Molloy J., Nascetti G., Mattiucci S., Cimmaruta R., Magoulas A.,, et al. (2003b) Final Report of the EU funded project HOMSIR: "A multidisciplinary approach using genetic markers and biological tags in horse mackerel (Trachurus trachurus) stock structure analysis". , Brussels295 Code QLK5-Ct1999-01438. European Commission.
Allain C. and Aloncle H. (1968) Migrations du germon en fonction des variations thermiques du milieu entre le secteur portugais et le SO de 
Irlande. Campagne de "La Pelagia" et de la "Thalassa" du 18 mai au 30 septembre 1968. Science et Pêche. Bulletin Institut des Pêches Maritime 176:1–15.
Aloncle H. and Delaporte F. (1973) Start of albacore fishing campaign along the European coast. A "zone test", radial Cap St. Vincent/Azores. 1:371–377 ICCAT Collective Volume of Scientific Papers.
Aloncle H. and Delaporte F. (1976) Marquages de germons par l'ISTPM, 1967–74. 5:216–220 ICCAT Collective Volume of Scientific Papers.
Aloncle H. and Delaporte F. (1982) Le germon: évolution d'une pêche et d'une recherche. La Pêche Maritime 1249:206–212.
Anon. (1996) Informe Reunión final PSG de atún blanco (junio 1994). 43:395 ICCAT Colección de documentos científicos.
Anon. (2001) Report of ICCAT SCRS Albacore Stock Assessment Session (Madrid, Spain, October 9 to 15, 2000). 52:1283–1390 ICCAT Collective Volume of Scientific Papers.
Bakun A. (1973) Coastal upwelling indices, west coast of North America, 1946–1971. 671:103 NOAA Technical Reports, NMFS SSRF–.
Bakun A. (1996) Patterns in the Ocean: Ocean Processes and Marine Population Dynamics. California Sea Grant(La Jolla, CA) pp. 323.
Bard F. X. (1981) Le thon germon (Thunnus alalunga. Bonaterre 1788) de Ocean Atlantique. pp. 333 Thése Doctorat. État. Universitat. París VI.
Bard F. X. and Santiago J. (1999) Review of albacore (T. alalunga) historical surface fisheries data (1920–1975) for possible relationships with the North Atlantic Oscillation. 49: pp. 311–323 ICCAT Collective Volume of Scientific Papers.
Barenbeim D. (1974) On regularities in the variability of the spawning terms of horse mackerel (Trachurus trachurus Linne) within the spawning area. 14: pp. 9 ICES Document CM 1974/J.
Beamish R. J. (1995) Climatic change and northern fish populations. Canadian Special Publication of Fisheries and Aquatic Sciences 121:739.
Beamish R. J. and McFarlane G. A. (1989) Effects of ocean variability on recruitment and an evaluation of parameters used in stock assessment models. Canadian Special Publication of Fisheries and Aquatic Sciences 108:379.
Blanton J. O., Atkinson L. P., Castillejo F. F., Lavín A. (1984) Coastal upwelling off the Rias Bajas, Galicia, Northwest Spain. 1. Hydrographic studies. 183:79–90 Rapports et Procès-Verbaux des Réunions du Conseil International pour l'Exploration de la Mer.
Borges M. F. and Gordo L. S. (1991) Spatial distribution by season and some biological parameters of horse mackerel (Trachurus trachurus L.) in the Portuguese continental waters (Division IXa). 54:16 ICES Document CM 1991/ H.
Borja A., Uriarte A., Valencia V., Motos L., Uriarte A. (1996) Relationships between anchovy (Engraulis encrasicolus L.) recruitment and the environment in the Bay of Biscay. Scientia Marina 60:Suppl. 2, 179–192.
Cabanas J. M., Lavín A., García M. J., González-Pola C., Tel E. (2003) Oceanographic variability in the northern shelf of the Iberian Peninsula (southern Bay of Biscay), 1990–1999. 219:71–79 ICES Marine Science Symposia.
Colas F. (2003) Circulation et dispersion lagrangiennes en Atlantique Nord-Est, PhD thesis, Université Bretagne Occidentale, France. Numéro de ordre 943:253.
Curry R. G. and McCartney M. S. (2001) Ocean gyre circulation changes associated with the North Atlantic Oscillation. Journal of Physical Oceanography 31:3374–3400.[CrossRef][Web of Science]
Cury P. and Roy C. (1989) Optimal environmental window and pelagic fish recruitment success in upwelling areas. Canadian Journal of Fisheries and Aquatic Sciences 46:670–680.
Cushing D. (1995) Population Production and the Regulation in the Sea. A Fisheries Perspective(Cambridge University Press, Cambridge) pp. 354.
Dao J. C. and Bard F. X. (1971) Les migrations estivales des germons immatures. 4: pp. 35–50 CNEXO Rapports Scientifiques et Techniques.
Dickson R., Lazier J., Meincke J., Rhines P., Swift J. (1996) Long-term coordinated changes in the convective activity of the North Atlantic. Progress in Oceanography 38:241–295.[CrossRef]
Draper N. R. and Smith H. (1998) Applied Regression Analysis. 3rd edn (John Wiley, New York)706.
FAO. (2000) FAO Yearbook of Fishery Statistics, 86/1. (FAO, Rome)713.
Fiedler P. C. and Bernard H. J. (1987) Tuna aggregation and feeding near fronts observed in satellite imagery. Continental Shelf Research 7:871–881.[CrossRef][Web of Science]
Francis R. I. C. C. (2006) Measuring the strength of environment-recruitment relationships: the importance of including predictor screening within cross-validations. ICES Journal of Marine Science 63:594–599.
Fuiman L. A. (2002) Special considerations of fish eggs and larvae. In Fuiman L. A. and Werner R. G. (Eds.). Fishery Science(Blackwell Science, Oxford) The Unique Contributions of Early Life Stages.
Gargett A. E., Li M., Brown R. (2001) Testing mechanistic explanations between environmental factors and marine fisheries. Canadian Journal of Fisheries and Aquatic Sciences 58:208–219.
González-Pola C. and Lavín A. (2003) Seasonal cycle and interanual variability of the heat content on a hydrographic section off Santander (southern Bay of Biscay), 1991–2000. ICES Marine Science Symposia 219:343–345.
González-Pola C., Lavín A., Vargas-Yánez M. (2005) Intense warming and salinity modification of intermediate water masses in the southeastern corner of the Bay of Biscay for the period 1992–2003. Journal of Geophysical Research 110:14 (C05020, doi:10.1029/2004JC002367).
Goñi N. and Arrizabalaga H. (2005) Analysis of juvenile North Atlantic albacore (Thunnus alalunga) catch per unit effort by surface gears in relation to environmental variables. ICES Journal of Marine Science 62:1475–1482.
Good P. I. and Hardin J. W. (2003) Common Errors in Statistics (and How to Avoid Them). (John Wiley, New York)222.
Goujon M., Antoine L., Gascuel D. (1996) Trends of abundance indices for albacore tuna (Thunnus alalunga) obtained by GLM fitting of the French troll and baitboat catch per unit of effort data for the period 1967 to 1986. 43:295–300 ICCAT Collective Volume of Scientific Papers.
Hair J. F., Anderson R. E., Tatham R. L., Black W. C. (1999) Análisis Multivariante. 5th edn (Prentice-Hall, Iberia, Madrid)799.
Hastie T. J. and Tibshirani R. J. (1990) Generalized Additive Models. (Chapman and Hall, London)335.
Havard-Duclos F. (1973) Albacore fishery in the Bay of Biscay: influence of temperature on school movement. 1:341–370 ICCAT Collective Volume of Scientific Papers.
Houde E. D. (2002) Mortality. In Fuiman L. A. and Werner R. G. (Eds.). Fishery Science(Blackwell Science, Oxford) The Unique Contributions of Early Life Stages.
Hue S. B. (1980) New knowledge on the migration of albacore (T. alalunga) in the Northeast Atlantic. 9: pp. 344–352 ICCAT Collective Volume of Scientific Papers.
Hurrell J. W. (1995) Decadal trends in the North Atlantic Oscillation: regional temperatures and precipitation. Science 269:676–679.
ICES. (2003) Report of the Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy. 07:572 ICES Document CM 2003/ACFM.
ICES. (2006) Report of the Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy. 08:631 ICES Document CM 2006/ACFM.
Kimura S., Nakai M., Sugimoto T. (1997) Migration of albacore, Thunnus alalunga, in the North Pacific Ocean in relation to large oceanic phenomena. Fisheries Oceanography 6:51–57.[CrossRef][Web of Science]
Koutsikopoulos C., Beillois P., Leroy C., Taillefer F. (1998) Temporal trends and spatial structures of the sea surface temperature in the Bay of Biscay. Oceanologica Acta 21:335–344.[CrossRef][Web of Science]
Lange U. and Greve W. (1997) Does temperature influence the spawning time, recruitment and distribution of flatfish via its influence on the rate of gonadal maturation? German Journal of Hydrography 49:251–263.
Laevastu T. (1993) Marine Climate, Weather and Fisheries. (Fishing News Books, Cornwall)204.
Lapointe M. F. and Peterman R. F. (1991) Spurious correlations between fish recruitment and environmental factors due to errors in the natural mortality rate used in virtual population analysis (VPA). ICES Journal of Marine Science 48:219–228.
Lasker R. (1975) Field criteria for survival of anchovy larvae: the relation between inshore chlorophyll maximum layers and successful first feeding. Fishery Bulletin US 73:453–462.
Laurs R. M. and Lynn R. J. (1991) North Pacific albacore ecology and oceanography. 105:69–87 NOAA Technical Report, NMFS Circular.
Lavín A., Cabanas J. M., Moreno-Ventas X., Ortiz de Zárate V., Porteiro C., Abaunza P. ( November 1999) Variability of the environmental conditions and fish communities in the Atlantic Iberian waters and its relationship to the NAO. 1999 Poster to the NAO Chapman Conference on NAO, Orense (Spain).
Lavín A., Díaz del Río G., Cabanas J. M., Casas G. (1991) Afloramiento en el noroeste de la península Ibérica. Índices de afloramiento para el punto 43°N 11°W. Periodo 1966–1989. 91:40 Informes Técnicos del Instituto Español de Oceanografía.
Lavín A., Díaz del Río G., Cabanas J. M., Casas G. (2000) Afloramiento en el noroeste de la península Ibérica. Índices de afloramiento para el punto 43°N 11°W. Periodo 1990–1999. 15:25 Datos y Resúmenes del Instituto Español de Oceanografía.
Lavín A., Valdés L., Sánchez F., Abaunza P., Forest A., Boucher J., Lazure P., et al. (2005) The Bay of Biscay: the encountering of the ocean and the shelf. In. In Robinson A. and Brink K. (Eds.). The Seas, 14(Harvard University Press, Cambridge, MA, USA).
Legendre P. and Legendre L. (1998) Numerical Ecology. Developments in Environmental Modelling, 20 2nd edn (Elsevier, Amsterdam) pp. 853.
Leroy C. (1990) Télédétection satellitaire et pêche du germon Thunnus alalunga dans le NE atlantique. ICCAT Recueil de Documents Scientifiques 32:147–157.
Lluch-Belda D., Crawford R. J. M., Kawasaki T., MacCall A. D., Parrish R. H., Schwartzlose R. A., Smith P. E. (1989) World-wide fluctuations of sardine and anchovy stocks: the regime problem. South African Journal of Marine Science 8:195–205.
Maillard C. (1996) Atlas Hydrologyque de l'Atlantique Nord-Est. (IFREMER, Brest)260.
Marija J. (1993) Norusis/SPSS Inc. SPSS for Windows. , Chicago, USA Base System User's Guide Release 6.0. SPSS Inc.
Mejuto J., Conser R., García B. (1992) Age specific standardized indices of abundance for albacore (Thunnus alalunga) from the Spanish surface fleets in the North Atlantic, 1983–1990. 39:269–283 ICCAT Collective Volume of Scientific Papers.
Mejuto J. and García B. (1993) A simple descriptive and comparative analysis of the activity of the peninsular Spanish fleets targeting albacore (Thunnus alalunga) in the Atlantic ocean from 1990 to 1991. 40:306–315 ICCAT Collective Volume of Scientific Papers.
Ortiz de Zárate V. and Cramer J. (2001) Standardized age specific catch rates for albacore, Thunnus alalunga, from the Spanish bait boat fishery in the northeast Atlantic, 1981–1999. 52:1418–1428 ICCAT Collective Volume of Scientific Papers.
Ortiz de Zárate V., Lavín A., Moreno–Ventas X. (1997) Is there a relationship between environmental variables and the surface catch of albacore Thunnus alalunga (Bonnaterre, 1788) in the North Atlantic? 48:250–252 ICCAT Collective Volume of Scientific Papers.
Ortiz de Zárate V., Ortiz M., Cramer J. (2001) Standardized age specific catch rates for albacore, Thunnus alalunga, from the Spanish troll fishery in the northeast Atlantic, 1981–1999. 52: ICCAT Collective Volume of Scientific Papers.
Ottersen G., Planque B., Belgrano A., Post E., Reid P. C., Stenseth N. C. (2001) Ecological effects of the North Atlantic Oscillation. Oecologia 128:1–14.[CrossRef][Web of Science]
Pepin P. (2002) Population analysis. In Fuiman L. A. and Werner R. G. (Eds.). Fishery Science(Blackwell Science, Oxford) The Unique Contributions of Early Life Stages.
Pingree R. D. (1993) Flow of surface waters to the west of the British Isles and in the Bay of Biscay. Deep Sea Research, Part II 40:369–388.[CrossRef]
Pipe R. K. and Walker P. (1987) The effect of temperature on development and hatching of scad, Trachurus trachurus L, eggs. Journal of Fish Biology 31:675–682.[CrossRef][Web of Science]
Pitcher T. J. (1995) The impact of pelagic fish behaviour on fisheries. Scientia Marina 59:295–306.[Web of Science]
Planque B., Beillois P., Jégou A-M., Lazure P., Petitgas P., Puillat I. (2003) Large-scale hydroclimatic variability in the Bay of Biscay: the 1990s in the context of interdecadal changes. ICES Marine Science Symposia 219:61–70.
Pollard R. T., Griffiths M. J., Cunningham S. A., Read J. F., Perez F. F., Rios A. F. (1996) Vivaldi 1991 – A study of the formation, circulation and ventilation of eastern North Atlantic Central Water. Progress in Oceanography 37:167–192.[CrossRef]
Reid D. G. (2001) Shelf edge fisheries and oceanography studies (SEFOS). Fisheries Research 50:204.
Rogers J. C. (1984) The association between the North Atlantic Oscillation and the Southern Oscillation in the northern hemisphere. Monthly Weather Review 112:1999–2015.[CrossRef][Web of Science]
Rogers J. C. (1990) Patterns of low frequency monthly sea level pressure variability (1899–1986) and associated wave cyclone frequencies. Journal of Climate 3:1364–1379.[CrossRef][Web of Science]
Ruiz Villarreal M., González-Pola C., Díaz del Río G., Lavín A., Otero P., Piedracoba S., Cabanas J. M. (2006) Oceanographic conditions in North and Northwest Iberia and their influence on the Prestige oil spill. Marine Pollution Bulletin 53:220–238.[CrossRef][Web of Science][Medline]
Sánchez F. and Gil J. (2000) Hydrographic mesoscale structures and Poleward Current as a determinant of hake (Merluccius merluccius) recruitment in the southern Bay of Biscay. ICES Journal of Marine Science 57:152–170.
Santiago J. and Arrizabalaga H. (2001) North Atlantic albacore catch-at-age estimates for the period 1975–1999 (updated). 52:1475–1480 ICCAT Collective Volume of Scientific Papers.
Santos A. M. P., Borges M. F., Groom S. (2001) Sardine and horse mackerel recruitment off Portugal. ICES Journal of Marine Science 58:589–596.
Schwartzlose R. A., Alheit J., Bakun A., Baumgartner T. R., Cloete R., Crawford R. J. M., Fletcher W. J.,, et al. (1999) Worldwide large-scale fluctuations of sardine and anchovy populations. South African Journal of Marine Science 21:289–347.
Shepherd J. G. (1997) Prediction of year-class strength by calibration regression analysis of multiple recruit index series. ICES Journal of Marine Science 54:741–752.
Shepherd J. G. (1999) Extended survivors analysis: an improved method for the analysis of catch-at-age data and abundance indices. ICES Journal of Marine Science 56:584–591.
Shepherd J. G. and Pope J. G. (2002) Dynamic pool models II: short-term and long-term forecasts of catch and biomass. In Hart P. J. B. and Reynolds J. D. (Eds.). Handbook of Fish Biology and Fisheries. 2(Blackwell Publishing, Oxford) Fisheries.
Shiohama T. (1971) Studies on measuring changes in the character of the fishing effort of the tuna longline fishery. Concentrations of the fishing effort to particular areas and species in the Japanese Atlantic fishery. Bulletin of the Far Seas Fisheries Laboratory 5:107–130.
Solá A., Motos L., Franco C., Lago de Lanzós A. (1990) Seasonal occurence of pelagic fish eggs and larvae in the Cantabrian Sea (VIIIc) and Galicia (IXa) from 1987 to 1989. 26:38 ICES Document CM 1990/H.
Taylor A. (1996) North-South shifts of the Gulf Stream: ocean-atmosphere interactions in the North Atlantic. International Journal of Climatology 16:559–583.[CrossRef][Web of Science]
Temming A. and Herrmann J.-P. (2001) Gastric evacuation of horse mackerel. 1. The effects of meal size, temperature and predator weight. Journal of Fish Biology 58:1230–1245.[CrossRef][Web of Science]
Ueyanagi S. (1971) Larval distribution of tunas and billfishes in the Atlantic Ocean. 71:297–305 FAO Fisheries Report.
Uozumi Y. (1996) An historical review of Japanese longline fishery and albacore catch in the Atlantic Ocean. 43:163–170 ICCAT Collective Volume of Scientific Papers.
Villamor B., Gonzalez-Pola C., Lavín A., Valdés L., Lago de Lanzós A., Franco C., Cabanas J. M., et al. (2004) Distribution and survival of larvae of mackerel (Scomber scombrus) in the North and North-west of the Iberian Peninsula, in relation to environmental conditions during spring 2000. 07:34 ICES Document CM 2004/J.
Woodruff S. D., Lubker S. J., Wolter K., Worley S. J., Elms J. D. (1993) Comprehensive Ocean-Atmosphere Data Set (COADS), Release 1a: 1980–92. Earth System Monitoring 4:1–8.
Wooster W. S., Bakun A., McLain D. R. (1976) The seasonal upwelling cycle along the eastern border of the North Atlantic. Journal of Marine Research 34:131–141.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||