write a summary of articles

Effect of habitat conditions on reproduction of the European anchovy (Engraulis encrasicolus) in the Strait of Sicily

G. BASILONE,1,* C. GUISANDE,2 B. PATTI,1

S. MAZZOLA,1 A. CUTTITTA,1

A. BONANNO,1 A.R. VERGARA1

AND I. MANEIRO2

1 Istituto di ricerche sulle Risorse Marine e l’Ambiente, Sezione di

Mazara del IAMC of Consiglio Nazionale delle Ricerche, via L. Vaccara 61, Mazara del Vallo (TP) Italy 2 Facultad de Ciencias del Mar, Universidad de Vigo, Campus

Universitario, 36310 Vigo, Spain

ABSTRACT

The aims of this study were to describe the reproduc- tive cycle of the European anchovy (Engraulis encra- sicolus) off the south coast of Sicily and determine whether intra- and inter-annual reproductive trait variations, if any, are adaptive responses which max- imize reproductive success under environmental fluc- tuations. Biological data were collected from purse seine and mid-water pelagic trawl commercial catches landed in Sciacca (Sicily) over 6 yr (1997–2002) at fortnightly intervals, analysing a total of 84 581 indi- viduals. No inter-annual changes in length at first reproduction were observed, with a mean pooled value of 11.26 cm for both sexes being found. Spawning intensity, indicated by gonadosomatic index, condi- tion factor and length–weight relationships, seem to be governed by food availability prior to spawning. Anchovy reproductive investment was limited by the area’s low primary production. There was a synchrony between reproductive cycle and temperature. Water warming marks the onset of a period of high water stability in the area, and its later cooling marks the onset of a period with low water stability. The rela- tionship between reproductive cycle and temperature is therefore probably a reproductive strategy having evolved to ensure that spawning takes place during the period of the year when water column stability is higher, favouring food concentration and egg and larval retention in the spawning areas.

Key words: anchovy, Engraulis encrasicolus reproductive cycle, food, temperature

INTRODUCTION

In the Mediterranean Sea there are several anchovy (Engraulis encrasicolus) spawning areas. The Strait of Sicily coastal population is the only one for which there presently is no information on its reproduction (Agostini and Bakun, 2002). Although anchovy landings in the Strait of Sicily are low compared with other Mediterranean areas (FAO, 1998), previous acoustic and Daily Egg Production Method estimates of anchovy biomass off the south coast of Sicily have varied between 7000 and 23 000 t in the period 1998– 2002 (Mazzola et al., 2000, 2002; Patti et al., 2004).

Recruitment success of pelagic fish species is strongly influenced by oceanographic factors, i.e. input of nutrients into the surface water, offshore egg and larval transport and water column stability (Peterman and Bradford, 1987; Bakun and Parrish, 1991; Borja et al., 1996; Tsai et al., 1997; Cole and McGalde, 1998; Painting et al., 1998; Guisande et al., 2001; Guisande et al., 2004a; Lloret et al., 2004). For fish, in particular the Clupeidae, inhabiting an environment with pronounced seasonal climatic variability, repro- ductive strategy must therefore be adapted to the habitat conditions in order to maximize offspring survival.

The reproductive strategy of fish species comprises a number of reproductive traits, such as timing of the reproductive cycle, length at first reproduction, spatial and temporal variations in spawning intensity, and changes in the investment in offspring quantity and quality. All of these processes may be modified in re- sponse to environmental fluctuations (Sargent et al., 1987; Guisande et al., 1998; Millán, 1999; Riveiro et al., 2000, 2003).

The onset of spawning in the European anchovy is clearly temperature dependent (Furnestin and Furn- estin, 1959; Palomera, 1992; Giráldez and Abad, 1995; Motos et al., 1996; Millán, 1999), but it is unknown whether temperature directly affects offspring success, specifically hatch success and larval survival, or is just

*Correspondence. e-mail: [email protected]

Received 21 June 2004

Revised version accepted 5 May 2005

FISHERIES OCEANOGRAPHY Fish. Oceanogr. 15:4, 271–280, 2006

� 2006 Blackwell Publishing Ltd. doi:10.1111/j.1365-2419.2005.00391.x 271

an indicator of habitat conditions favouring anchovy reproduction. Intra- and inter-annual variations in length at first reproduction, spawning intensity, and condition factor (CF) have also been observed in the European anchovy (Hay and Brett, 1988; Luo and Musick, 1991; Giráldez and Abad, 1995; Regner, 1996; Millán, 1999; Lapolla, 2001; Peebles, 2002) but it is unknown if such variations are due to changes in the environment. In addition to the lack of information about anchovy reproduction in the Strait of Sicily, it is therefore unclear what causal mechanisms, if any, af- fect its reproductive biology.

The aims of this study were to describe the repro- ductive cycle of the European anchovy near the south coast of Sicily and identify the environmental factors responsible for intra- and inter-annual variations in its reproductive traits.

MATERIAL AND METHODS

Data collection

Data were collected from purse seine and mid-water pelagic trawl commercial catches landed in Sciacca, the most representative port in the study area. Over 6 yr (1997–2002) at fortnightly intervals, 106 samples containing a total of 84 581 specimens were used to determine sex and size (length, total body weight), recording total body length (TL) to the nearest 1 mm and total body weight to the nearest 1 g. Only one sample was collected for each sampling occasion and some weeks, particularly in winter because of bad weather conditions, no fish were landed at Sciacca port. A sub-sample of 64 019 individuals was staged for maturity and used for gonadal weight (GW) measurements. Sex was determined macroscopically and maturity stages were assigned according to the five-degree maturity scale for partial spawners des- cribed in Holden and Raitt (1975). Ovary staging was used to judge seasonal development of mature gonads as an indicator of the spawning season, to support gonadosomatic index (GSI) data, and to determine the number of immature and mature specimens per size class, necessary for estimations of length at first maturity (L50).

Length–weight relationships were calculated annually for males and females and combined sexes using the following equations:

TW ¼ aTLb ð1Þ

where TW is the total weight (g), TL is the total length (cm), b is the allometric growth and a is a scaling constant. The combined sex data included

males, females, and immature specimens, in order to cover the widest size range.

Reproductive cycle

The spawning period was determined by evaluating the GSI and the percentage of mature specimens. GSI was calculated monthly and seasonally, using the equation described by Bougis (1952):

GSI ¼ GW

TW ð2Þ

where GW is the gonad weight and TW is the total weight (g).

Several studies have criticized the use of the GSI because it can be influenced by fish length. As a result, some other methodologies have been proposed, al- though each of the methods has advantages and dis- advantages (DeVlaming et al., 1982; Erickson et al., 1985). The use of a relative gonadal index (RGI; Er- ickson et al., 1985) is more appropriate than GSI when the relationship between GW and total weight is non-linear. Moreover RGI is appropriate only when the slopes of the relationship GW (ln) to body weight (ln) are not different among gonadal stages (Erickson et al., 1985).

Condition factor

Condition status was investigated monthly by evalu- ating the CF with the following equation (Le Cren, 1951):

CF ¼ TW � GW

aTLb ð3Þ

where a and b are the length–weight relationship regression parameters estimated for each year (TL in cm). To avoid fluctuations due to GW change, ovary- free weight (TW ) GW) was used in the above equation rather than total body weight (TW). Gutted weights exhibit strong seasonal and interannual dif- ferences in condition because of the storage of energy (e.g. proteins, lipids) in the muscle (Lambert and Dutil, 1997a,b); as a result, indices based on gutted weights can reveal differences in condition. The Le Cren condition index is relative and independent of body size, whereas alternative indices such as Fulton’s index (1911 in Lagler, 1956) tend to increase con- siderably with size [Giráldez and Abad (1995); Millán (1999); see also Bolger and Connolly (1989) for further discussion of the characteristics of different condition indices].

Length at first maturity (L50) was calculated from the percentage of mature individuals (stages III, IV and V) occurring in the reproductive period, avoiding

272 G. Basilone et al.

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

the onset month and last month, as determined by GSI seasonal development. The length at which 50% of specimens were mature was estimated by an ana- lytical method based on a logistic non-linear regres- sion model (Hunter et al., 1992; Roa et al., 1999):

Pl ¼ a

1 þ eb0þbl1 ð4Þ

where Pl is the proportion of mature specimens at size l, and a, b0 and b1 are the asymptote, intercept and slope parameters of the logistic regression between body length and percentage of mature individuals. Confidence limits were estimated using the methodo- logy described by Roa et al. (1999).

Estimation of chlorophyll concentration

Mean monthly chlorophyll concentration in the Strait of Sicily in the period 1997–2002 derived by satellite measurement was used as an indicator of production in the area. Chlorophyll concentration is generally accepted as a primary production index (Longhurst et al., 1995), and it is related with fish biomass pro- duction (Ryther, 1969; Iverson, 1990). Although the anchovy diet in the Mediterranean sea is mainly zooplankton (Tudela and Palomera, 1997), several authors have shown that yearly Mediterranean zooplankton production is characterized by two abun- dance maxima, one in late winter–early spring and another in autumn, related to the corresponding phy- toplankton maxima (Scotto di Carlo and Ianora, 1983; Estrada et al., 1985; Siokou Frangou, 1996). Basilone et al. (2004) observed a significant relationship between chlorophyll concentration derived from satellite images and anchovy growth in the Mediterranean Sea, indi- cating that sea surface estimates of chlorophyll con- centration are good indicators of food availability for anchovy.

Monthly (September 1997 to December 2002) composite downloaded satellite images from http:// marine.jrc.cec.eu.int/frames/archive_seawifs.htm, were used to estimate the monthly average chlorophyll a concentration in the study area (Fig. 1). Chlorophyll concentration was estimated from the coast to approximately the 100-m isobath offshore (5269 km2). SeaWiFS chlorophyll estimations were not validated with in situ information. Silva et al. (2000) showed that SeaWiFS chlorophyll estimations were overestimated in the Chilean coastal upwelling, in comparison with in situ data (approximately 2.3 times higher). The real valuescanthereforebeconsiderablydifferentfromthose derived from satellite images, but the latter are still useful for comparative studies of intra- and inter-annual chlorophyll concentration variability in the area.

Sea surface temperature

Figure 1 also shows the location (37�30¢N 12�30¢E) where mean monthly 1� · 1� (spatial resolution) sea surface temperature (SST) data from 1997 to 2002 were obtained from a collection of SST analyses pre- pared at the National Centre for Environmental Pre- diction by D. Reynolds, S. Stokes and T. Smith. These analyses were determined by blending marine surface observations and satellite AVHRR data using an optimum interpolation (OI) method. A description of the OI analysis is given in Reynolds and Smith (1994). SST data were provided by the Data Support Section of the University Corporation for Atmospheric Re- search (UCAR), Boulder, Colorado, USA, from their web page http://dss.ucar.edu/datasets/ds277.0.

Turbulence

The energy transferred by wind through the water col- umn creates turbulence in the surface layers. The wind- mixing index (WM) in the upper layer is usually cal- culated as the cube of wind speed (Elsberry and Gar- wood, 1978). We used this index as an indicator of turbulence in the surface layers. Data were derived from I-COADS standard data set (http://dss.ucar.edu/data- sets/ds540.1/data/msg_1deg/) (Dı́az et al., 2002) for a 1� · 1� cell centred approximately at 37�30¢N 12�30¢E.

RESULTS

Habitat oceanographic conditions

Chlorophyll concentration, SST, and the WM index showed large intra- and inter-annual variations during the period 1997–2002 (Fig. 2). Monthly mean

12.00 12.50 13.00 13.50 14.00 14.50 15.00 15.50

36.50

37.00

37.50

38.00

SICILY Mazara del Vallo

Sciacca

Licata

Cape Passero

200 m

100 m

50 m

Palermo

Catania

N

S

W E

Figure 1. Map of Sicily showing the area studied, indicating the area where chlorophyll was estimated (shadowed area) and the station from which mean monthly sea surface tem- perature and wind-mixing index from 1997 to 2002 were obtained ( ).

Anchovy reproduction and habitat conditions 273

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

chlorophyll concentrations and SST showed a signi- ficant negative correlation (Fig. 3, F1,62 ¼ 158.9, r 2 ¼ 0.72, P < 0.001, exponential fit). Sea surface temperature and the WM index were

also negatively correlated (F1,69 ¼ 38.7, r2 ¼ 0.36, P < 0.001) (Fig. 4). Water warming marks the onset

of a period where the WM is lower (Fig. 2) and, hence, the water column stability is probably higher.

Reproductive cycle

The anchovy spawning season, as defined by GSI, usually began in March or April (although in 2001 it began in June), peaked in June and July, and ended by

C on

d it

io n

f ac

to r

0.85

0.90

0.95

1.00

1.05

1.10

T em

p er

at u

re (

ºC )

14 16 18 20 22 24 26 28

G S

I

0.00

0.01

0.02

0.03

0.04

0.05

J F M A M J J A S O N D

C h

lo ro

p h

yl l

(m g

m –3

)

0.0

0.2

0.4

0.6

0.8

1.0

J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D

1997 1998 1999 2000 2001 2002

W in

d i

n d

ex m

3 s ec

–3

0

500

1000

1500

2000

2500

Figure 2. Monthly means of sea surface temperature, wind-mixing index, chlo- rophyll concentration, and condition factor and gonadosomatic index of the European anchovy over the period stud- ied in the South coast of Sicily. Females ( ) and males (().

Sea surface temperature (ºC) 12 14 16 18 20 22 24 26 28

C h

lo ro

p yl

l (m

g m

–3 )

0.0

0.2

0.4

0.6

0.8

1.0

Figure 3. Relationship between monthly sea surface tem- perature means and chlorophyll concentration from Sep- tember 1997 to December 2002 in the area studied.

Sea surface temperature (ºC) 12 14 16 18 20 22 24 26 28

W in

d m

ix in

g in

d ex

( m

3 s –3

)

0

500

1000

1500

2000

2500

Figure 4. Relationship between monthly sea surface tem- perature means and wind mixing index from January 1997 to December 2002 in the area studied.

274 G. Basilone et al.

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

mid-September or October (Fig. 2). GSI was usually higher for females than males (Fig. 2). Maturity stages data support GSI results because there was a positive linear relationship between percentages of mature fe- males and mean monthly GSI values (r

2 ¼ 0.88; F1,49 ¼ 683; P < 0.001; Fig. 5). The intra-annual variation in the reproductive cycle was clearly syn- chronized with water temperature (r2 ¼ 0.56; F1,10 ¼ 12.83 P ¼ 0.05, Fig. 6), as the onset of the reproduc-

tive cycle was associated with water warming and its end with water cooling (Fig. 2). A weak but significant negative linear relationship was detected between WM and GSI without considering any time lag (r2 ¼ 0.25, F1,55 ¼ 18.11; P < 0.001).

There were also inter-annual GSI variations during the spawning period for both female and male speci- mens. An analysis of covariance (ANCOVA) with chlorophyll concentration (log transformed) as the covariable and sex as the factor showed that GSI during the spawning period was significantly correlated with mean chlorophyll concentration from October of the previous year to May (Table 1), and that there were significant differences between males and females (Table 1, Fig. 7). The October–May period was cho- sen because it has the highest chlorophyll levels

0 20 40 60 80 100

M ea

n m

on th

ly G

S I

(f em

al e)

0.00

0.01

0.02

0.03

0.04

0.05

Figure 5. Relationship between monthly mean gonadoso- matic index (GSI) (females) and percentage of mature fe- males during spawning periods as defined by GSI and maturity stage trends.

Mean spawning period temperature °C 14 16 18 20 22 24 26 28

S p

aw n

in g

p er

io d

m ea

n G

S I

(f em

al es

)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Figure 6. Relationship between mean gonadosomatic index (females) and sea surface temperature during spawning periods.

Table 1. Analysis of covariance with chlorophyll concen- tration (log transformed) as the covariable, sex as the factor and gonadosomatic index as the dependent variable.

Source d.f. MS F P-value

Corrected model 2 0.179 880.7 <0.001 Interception 1 0.054 267.2 <0.001 Chlorophyll 1 0.354 1739.3 <0.001 Sex 1 0.006 28.509 <0.001 Error 53 856 0.000 Total 53 859

Chlorophyll (mg m–3)

0.40 0.42 0.44 0.46 0.48 0.50 0.52

G S

I

0.016

0.018

0.020

0.022

0.024

0.026

0.028

0.030

Figure 7. Relationship between mean annual gonadoso- matic index during the spawning season (May–September) and mean chlorophyll concentration from October to May. All standard deviations were lower than 0.015. Females ( ) and males (().

Anchovy reproduction and habitat conditions 275

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

(Fig. 2). Chlorophyll concentration explained 92% of the total variance observed in female GSI and 55% of that in male GSI.

Condition factor

The CF data were plotted against GSI to see if they provided any additional information. As the linear regression correlation coefficient was low (r2 ¼ 0.21), we decided to use CF data also in the environment versus biology analysis. Figure 2 shows that there were no differences in CF between males and females. However, there were clear intra-annual variations linked to timing of the phytoplankton bloom. An ANCOVA with chlorophyll concentration (log transformed) as covariable and sex as the factor showed that monthly CF had a significant correlation with monthly chlorophyll concentration, with a 4- month lag (Table 2), but there were no significant differences in CF between males and females (Table 2, Fig. 8). The existence of relationships between CF and SST or WM was also evaluated: a positive linear relationship between CF and SST was found (r2 ¼ 0.476 F1,50 ¼ 45.47, P < 0.001, 2-month lag). A weaker, negative linear relationship was found be- tween CF and WM (r2 ¼ 0.20, F1,52 ¼ 13.38, P ¼ 0.001, 3-month lag).

Length–weight relationships corroborate the find- ings obtained from analyses of variability in CF. An ANCOVA with body length (log transformed) as covariable and mean chlorophyll concentration from October to May as factor shows that total anchovy weight is related with body length (Tables 3 and 4) and changes in total anchovy weight for a given body length are mainly due to changes in the chlorophyll concentration (Table 4). A stepwise multiple regres- sion including mean temperature and chlorophyll be- tween October and May showed a significant relationship (F1,3 ¼ 18.9, r2 ¼ 0.9, P ¼ 0.049) be- tween chlorophyll concentration and the slope of the annual length–weight relationship (Fig. 9); tempera- ture was not significant (F1,3 ¼ 0.269, P ¼ 0.655).

Length at first maturity

Maturity in some specimens was evident at 9.4 cm (females and males), and all individuals longer than 13.5 cm were mature. Mean L50 values for the whole

Table 2. Analysis of covariance with chlorophyll concen- tration (log transformed) as the covariable, sex as the factor and condition factor as the dependent variable.

Source d.f. MS F P-value

Corrected model 2 0.019 14.3 <0.001 Interception 1 18.801 14346.9 <0.001 Chlorophyll 1 0.037 28.5 <0.001 Sex 1 3.410)5 0.03 0.872 Error 101 0.001 Total 104

Chlorophyll (mg m–3)

0.0 0.2 0.4 0.6 0.8 1.0 C

on d

it io

n f

ac to

r 0.85

0.90

0.95

1.00

1.05

1.10

Figure 8. Relationship between monthly anchovy condi- tion factor and monthly chlorophyll concentration with a 4- month lag. Females ( ) and males ((). The line indicates the regression considering both females and males.

Table 3. Values of regression parameters from the length– weight relationships (Eqn 1) obtained in different years for Engraulis encrasicolus.

Year

Males + females + immature

a b r 2

n

1997 0.0047 3.2203 0.95 10 212 1998 0.0030 3.3841 0.95 13 857 1999 0.0019 3.5603 0.95 12 396 2000 0.0017 3.5975 0.97 15 325 2001 0.0027 3.4050 0.95 21 721 2002 0.0021 3.5166 0.96 11 059

Table 4. Analysis of covariance with total length (log transformed) as the covariable, chlorophyll as the factor and anchovy weight as the dependent variable.

Source d.f. MS F P-value

Corrected model 5 369.151 336 902 <0.001 Interception 1 887.660 810 116 <0.001 Body length 1 1812.242 1 653 929 <0.001 Chlorophyll 4 1.365 1245 <0.001 Error 74 356 0.001 Total 74 362

276 G. Basilone et al.

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

period were 11.27 ± 0.09 cm for males and 11.24 ± 0.09 cm for females (Table 5). Although a more accurate estimation could be obtained by using microscopic and macroscopic gonad examina- tion, several authors have used only macroscopic examination for L50 anchovy estimations in the Mediterranean Sea (Hemida, 1987; Djabali et al., 1988; Pertierra, 1992; Giráldez and Abad, 1995; Millán, 1999; Sinovčić, 2000; Palomera et al., 2003).

There was little inter-annual variability in length at first maturity during the 6 yr studied for males, females and both sexes combined. An analysis of residual sum squares (Chen et al., 1992; Haddon, 2001) was applied to compare the logistic curves for years 1997 and 2002, which had the greatest difference in the estimate of length at first maturity. No significant difference in L50 between these years was found (F1,26 ¼ 0.7, P ¼ 0.49).

DISCUSSION

This study employed a GSI time series to investigate the intensity of anchovy spawning seasonality off Si- cily. Our results show that inter-annual GSI variations

Chlorophyll (mg m–3 )

0.40 0.42 0.44 0.46 0.48 0.50 0.52

S lo

p e

of t

h e

le n

gt h

–w ei

gh t

re gr

es si

on

3.35

3.40

3.45

3.50

3.55

3.60

3.65

Figure 9. Relationship between the slope of the length– weight regression obtained each year for the European an- chovy in the South coast of Sicily (shown in Table 3) and mean chlorophyll from October to May of the previous year in the area studied. Bars indicate the confidence intervals of the length–weight regressions obtained each year.

Table 5. Annual estimates of length at first (L50) and full (L95) maturity, and maturation range (L25)75), for males, females and combined sexes over the period 1997–2002. Confidence limits (c.l.) for the analytical estimation (AE) are also listed.

Year L25)75 L95 L50 AE (c.l.) n

Males 1997 9.7–12.6 15.11 11.19 11.02–11.33 4511 1998 10.3–12.1 13.65 11.18 11.10–11.25 4485 1999 10.7–11.9 13 11.30 11.24–11.36 4061 2000 10.5–12.2 13.7 11.34 11.24–11.41 4157 2001 8.6–14.3 15.7 11.42 11.13–11.62 3103 2002 11.0–12.2 13.3 11.64 11.55–11.71 2139

1997–2002 10.1–12.4 14.4 11.27 11.22–11.31 22 456

Females 1997 9.3–12.2 14.7 10.75 10.53–10.93 5705 1998 10.3–11.9 13.26 11.09 10.99–11.17 4103 1999 10.8–11.9 12.8 11.33 11.27–11.38 3850 2000 10.7–12.3 13.6 11.52 11.43–11.59 4329 2001 9.5–12.9 15.7 11.04 10.70–11.28 2533 2002 11.3–12.3 13.1 11.8 11.71–11.86 2400 1997–2002 10.3–12.2 13.8 11.24 11.19–11.28 22 920

Total 1997 9.6–12.4 14.8 11.02 10.89–11.13 10 216 1998 10.3–12.0 13.5 11.14 11.08–11.20 8588 1999 10.7–11.9 12.9 11.31 11.27–11.35 7911 2000 10.6–12.2 13.6 11.41 11.35–11.47 8486 2001 8.9–13.6 17.5 11.27 11.07–11.41 5636 2002 11.1–12.3 13.2 11.71 11.65–11.76 4539 1997–2002 10.2–12.3 14.1 11.26 11.23–11.29 45 376

Anchovy reproduction and habitat conditions 277

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

for both males and females were mainly governed by changes in the amount of chlorophyll in the period preceding spawning, and so both the intensity and seasonality of spawning were governed by oceanogra- phic processes in the months prior to spawning.

These processes are probably associated with prey availability, increase in anchovy energy reserves and gonadal development. Inter-annual differences in length at first reproduction were small and not signi- ficant.

The length at first maturity of a species is important life history information for fish resource management. It has been suggested that inter-annual L50 changes occur in the European anchovy (Giráldez and Abad, 1995; Millán, 1999), although significant inter-annual differences have not been demonstrated in these studies. Inter-annual variability in L50 has been attributed to spawning tactics to ensure survival of early life history stages (Alheit, 1989). However, we observed no significant differences in L50 from one year to another.

We observed a clear relationship between spawning intensity, CF, and parameters of length–weight regression models with proxy variables believed to represent the productivity in the study area. In gen- eral, spawning intensity increased several months after peak chlorophyll concentrations. Positive correlations between reproductive investment and prior phyto- plankton and zooplankton abundance (food concen- tration) have been previously observed in clupeid species (Hay and Brett, 1988; Luo and Musick, 1991; Regner, 1996; Millán, 1999; Lapolla, 2001; Peebles, 2002). Our results show that anchovy spawning intensity and CF were affected by chlorophyll con- centration. Basilone et al. (2004) also observed that anchovy body growth is limited by chlorophyll con- centration in the Strait of Sicily, where it is at the lower end of the range observed in different anchovy spawning areas (Basilone et al., 2004). This low chlorophyll concentration may be a consequence of low primary production, and probably explains why the pooled mean GSI for both sexes during the spawning period in the Strait of Sicily was only 0.025. This value is similar to the 0.029 obtained in the Middle-North Adriatic (Croatia) (Sinovčić, 2000), an area with a slightly higher chlorophyll concentration than the Strait of Sicily (Basilone et al., 2004). However, in other areas such as the Bay of Cádiz and along the coast of Malaga, Spain, where the primary production is much higher than in the Strait of Sicily or Middle-North Adriatic (Basilone et al., 2004), the mean GSI for European anchovy during the spawning season was 0.037 (Millán, 1999) and 0.033 (Giráldez

and Abad, 1995), respectively. Our findings thus sug- gest that anchovy reproductive investment may be limited by primary production in some spawning areas.

However, we cannot explain why the relationship between GSI and CF apparently reaches an asymptote at high chlorophyll concentrations in our study area (Figs 7 and 8). As mentioned above, higher GSI val- ues are obtained for anchovy in other areas and, therefore, the asymptotic value of approximately 0.03 for GSI or 1.1 for CF is clearly not the maximum for this species. A possible explanation is that other fac- tors, which have not been taken into account, may be also limiting anchovy reproduction in the Strait of Sicily.

The synchrony between seasonal reproductive cycle and temperature observed in our study is well known in anchovy species (Furnestin and Furnestin, 1959; Palomera, 1992; Motos et al., 1996; Regner, 1996; Millán, 1999; Lapolla, 2001). In fact, spawning in temperate fish species is often cued by temperature (Moyle and Cech, 1988). However, we do not believe that this synchrony is due to a threshold temperature below which anchovy reproduction cannot take place: in fact, European anchovy can spawn within a tem- perature range of 11.6–27.5�C (Regner, 1996) and would therefore have the potential to breed all year round in the Strait of Sicily.

Water warming marks the onset of a period with low WM (probable high water stability), and cooling, later in the year, marks the onset of a period with high WM (probable low water stability; Figs 2 and 4). Water column stability may favour spawning success because (1) turbulent conditions limit primary pro- duction, hence decreasing food availability for larvae (Lasker, 1975; Huntsman and Barber, 1977) and (2) a stratified water column allows food particles to aggre- gate in sufficient concentration for larval feeding (Lasker, 1981). The effect of water column stability on primary production is not the factor marking the onset of the spawning period (usually in March–April), be- cause the phytoplankton bloom in this area occurs between November and April (Fig. 2).

The synchrony between reproductive cycle and temperature is probably a reproductive strategy which has evolved to enable spawning during the period of the year when water column stability is higher. Stable sea conditions favour the aggregation of prey, higher larval feeding success and hence larval survival. The low WM may also favour retention of eggs and larvae in the spawning areas (Penven et al., 2000), as wind- induced currents will be weaker. Offshore transport during the spawning season can have a negative effect on larval survival of pelagic fish species, by carrying

278 G. Basilone et al.

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

eggs and larvae to areas where food concentrations are too low to ensure high larval survival rates (Bailey, 1981; Parrish et al., 1981; Penven et al., 2000).

The low biomass of anchovy in the Strait of Sicily compared with other Mediterranean areas (Agostini and Bakun, 2002) may therefore be partly due to food limitation of reproduction and growth. Anchovy in this area have evolved to spawn when seasonal wind speeds are lowest, even though temperatures and chlorophyll a concentrations are higher at other times of the year.

ACKNOWLEDGEMENTS

The authors would like to thank: the referees and Associate editor for their comments; the SeaWiFS Project (Code 970.2) and the Distributed Active Ar- chive Center (Code 902) at the Goddard Space Flight Center, Greenbelt, MD20771, for the production and distribution of the SeaWiFS raw data, respectively; the Inland and Marine Waters Unit of the Joint Research Centre (IMW-JRC) for the processing and the distri- bution of the final products. This paper was based on data collected during the following research projects: DGXIV MED 96/052 and DGXIV MED 98/070.

REFERENCES

Agostini, V. and Bakun, A. (2002) ‘‘Ocean triads’’ in the Mediterranean Sea: physical mechanisms potentially struc- turing reproductive habitat suitability (with example appli- cation to European anchovy, Engraulis encrasicolus). Fish Oceanogr. 11:129–142.

Alheit, J. (1989) Comparative spawning biology of anchovies, sardines, and sprats. Rapp.p-v. reun. Cons. int. Explor. Mer. 191:7–14.

Bailey, K.M. (1981) Larval transport and recruitment of Pacific hake, Merluccius productus. Mar. Ecol. Prog. Ser. 6:1–9.

Bakun, A. and Parrish, R.H. (1991) Comparative studies of coastal pelagic fish reproductive habitats: the anchovy (Engraulis anchoita) of the southwestern Atlantic. ICES J. Mar. Sci. 48:342–361.

Basilone, G., Guisande, C., Patti, B. et al. (2004) Linking habitat conditions and growth in the European anchovy (Engraulis encrasicolus). Fish. Res. 68:9–19.

Bolger, T. and Connolly, P.L. (1989) The selection of suitable indices for the measurement and analysis of fish condition. J. Fish. Biol. 34:171–182.

Borja, A., Uriarte, A., Valencia, V., Motos, L. and Uriarte, A. (1996) Relationships between anchovy (Engraulis encrasico- lus L.) recruitment and the environment in the Bay of Bis- cay. Sci. Mar. 60:179–192.

Bougis, P. (1952) Rapport hepatosomatique et rapport gonado- somatique chez Mullus barbatus L. Bull. Soc. Zool. 74:326–330.

Chen, Y., Jackson, D.A. and Harvey, H.H. (1992) A comparison of von bertalanffy and polynomial functions in modelling fish growth data. Can. J. Fish. Aquat. Sci. 49:1228–1235.

Cole, J. and McGalde, J. (1998) Clupeid population variability, the environment and satellite imagery in coastal upwelling systems. Rev. Fish. Biol. Fish. 8:445–471.

DeVlaming, V., Grossman, G. and Chapman, F. (1982) On the use of Gonosomatic index. Comp. Biochem. Physiol 73:31– 39.

Dı́az, H., Folland, C., Manabe, T., Parker, D., Reynolds, R. and Woodruff, S. (2002) Workshop on advances in the use of historical marine climate data. WMO Bulletin 51:377–380.

Djabali, I., Mouhoubet, R. and Hemida, F. (1988) Résultats des travaux réalisés sur les stocks de sardines et des anchois des côtes algéroises. FAO, Rap. Pêches 395:112–120.

Elsberry, R.L. and Garwood, R.W. (1978) Sea-surface tem- perature anomaly generation in relation to atmospheric storms. Bull. Am. Meteor. Soc. 59:786–789.

Erickson, D.L., Hightower, J.E. and Grossman, G. (1985) The relative gonadal index: an alternative index for quantifica- tion of reproductive condition Comp. Biochem. Physiol. 81:117–120.

Estrada, M., Vives, F. and Alcaraz, M. (1985) Life and the productivity of the open sea. In: Western Mediterranean. R. Margalef (ed.) Oxford: Pergamon Press, pp. 148–197.

FAO (1998) FAO Yearbook of Fishery Statistics: Capture Pro- duction, Vol. 82, 1996. Rome: FAO.

Furnestin, J. and Furnestin, M.L. (1959) La reproduction de la sardine et de l’anchois des cotes atlantiques du Maroc. Rev. Trav. Inst. Peches Marit. 23:79–102.

Giráldez, A. and Abad, R. (1995) Aspects on the reproductive biology of the Western Mediterranean anchovy from the coasts of Malaga (Alboran Sea). Sci. Mar. 59:15–23.

Guisande, C., Riveiro, I., Solá, A. and Valdés, L. (1998) Effect of biotic and abiotic factors in the biochemical composition of wild eggs and larvae of several fish species. Mar Ecol. Prog. Ser. 163:53–61.

Guisande, C., Cabanas, J.M., Vergara, A.R. and Riveiro, I. (2001) Effect of climate on recruitment success of Atlantic Iberian sardine (Sardina pilchardus). Mar. Ecol. Prog. Ser. 223:243–250.

Guisande, C., Vergara, A.R., Riveiro, I. and Cabanas, J.M. (2004a) Climate change and abundance of the Atlantic Iberian sardine (Sardina pilchardus). Fish. Oceanogr. 13:91– 101.

Haddon, M. (2001) Modelling and Quantitative Methods in Fish- eries. Boca Ratón, Florida: Chapman and Hall/CRC, 424 pp.

Hay, D.E. and Brett, J.R. (1988) Maturation and fecundity of Pacific herring (Clupea harengus pallasi). Can. J. Fish. Aquat. Sci. 45:399–406.

Hemida, F. (1987) Contribution à l’étude de l’anchois Engraulis encrasicolus (Linné, 1758) dans la région d’Alger: biologie et exploitation. Thèse de Magister, USTHB, Alger.

Holden, M.J. and Raitt, D.F.S. (1975) Manual of Fisheries sci- ence. Part 2 – Methods of resources investigation and their application. FAO fish. tech. pap. 115:211 pp.

Hunter, J.R., Macewicz, B.J., Lo, N.C. and Kimbrell, C.A. (1992) Fecundity, spawning and maturity of female Dover sole Microstomus pacificus, with an evaluation of assumption and precision. Fish. Bull. 90:101–128.

Huntsman, S.A. and Barber, R.T. (1977) Primary production off northwest Africa: the relationship to wind and nutrient conditions. Deep-Sea Res. 24:25–33.

Iverson, R.L. (1990) Control of marine fish production. Limnol. Oceanogr. 35:1593–1604.

Anchovy reproduction and habitat conditions 279

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.

Lagler, K.F. (1956) Freshwater Fishery Biology, 2nd edn. Dubu- que, Iowa: W.C. Brown Publishers.

Lambert, Y. and Dutil, J.D. (1997a) Condition and energy reserves of Atlantic cod (Gadus morhua) during the collapse of the northern Gulf of St. Lawrence stock. Can. J. Fish. Acquat. Sci. 54:104–112.

Lambert, Y. and Dutil, J.D. (1997b) Can simple condition indices be used to monitor and quantify seasonal changes in the energy reserves of Atlantic cod (Gadus morhua)? Can. J. Fish. Acquat. Sci. 54:2388–2400.

Lapolla, A.E. (2001) Bay anchovy Anchoa mitchilli in Narra- gansett Bay, Rhode Island. II. Sapwing season, hatch-date distribution and young-of-the-year growth. Mar. Ecol. Prog. Ser. 217:103–109.

Lasker, R. (1975) Fish criteria for survival of anchovy larvae: the relation between inshore chlorophyll maximum layers and successful first feeding. Fish. Bull. 73:453–462.

Lasker, R. (1981) Marine Fish Larvae. Morphology, Ecology, and Relation to Fisheries. Seattle: University of Washington Press, 131 pp.

Le Cren, E.D. (1951) The length–weight relationship and sea- sonal cycle in gonad weight and condition in the perch (Perca fluviatilis). J. Anim. Ecol. 20:201–219.

Lloret, J., Palomera, I., Salat, J. and Sole, I. (2004) Impact of freshwater input and wind on landings of anchovy (Engraulis encrasicolus) and sardine (Sardina pilchardus) in shelf waters surrounding the Ebre (Ebro) River delta (north-western Mediterranean). Fish. Oceanogr. 13:102–110.

Longhurst, A., Sathyendranath, S., Platt, T. and Caverhill, C. (1995) An estimate of global primary production in the ocean from satellite radiometer data. J. Plankton Res. 17:1245–1271.

Luo, J. and Musick, J.A. (1991) Reproductive biology of the bay anchovy in Chesapeake Bay. Trans. Am. Fish. Soc. 120:701– 710.

Mazzola, S., Garcı́a, A. and Garcı́a Lafuente, J. (2000) Dis- tribution, Biology and Biomass Estimates of the Sicilian Channel Anchovy. DGXIV MED 96/052 Final Report.

Mazzola, S., Garcia, A. and Garcia Lafuente, J. (2002) The Sicilian Channel Anchovy Fishery and the Underlying Oceano- graphic and Biological Processes Conditioning their Interannual Fluctuations. DGXIV MED 98/070 Final Report.

Millán, M. (1999) Reproductive characteristics and condition status of anchovy Engraulis enchrasicolus L. from the Bay of Cadiz (SW Spain). Fish. Res. 41:73–86.

Motos, L., Uriarte, A. and Valencia, V. (1996) The spawning environment of the Bay of Biscay anchovy (Engraulis encra- sicolus L.). Sci. Mar. 60:117–140.

Moyle, P.B. and Cech, J.J. (1988) Fishes: and Introduction to Ichthyology. Englewood Cliffs, NY: Prentice-Hall.

Painting, S.J., Hutchings, L., Hugger, J.A., Korrubel, J.L., Richarson, A.J. and Verheye, H.M. (1998) Environmental and biological monitoring for forecasting anchovy recruit- ment in the southern Benguela upwelling region. Fish. Oceanogr. 7:364–374.

Palomera, I. (1992) Spawning of anchovy, Engraulis encrasicolus L., in the Northwestern Mediterranean relative to hydrogra- phic features in the region. Mar. Ecol. Prog. Ser. 79:215–223.

Palomera, I., Tejeiro, B. and Alemany, X. (2003) Size at first maturity of the NW Mediterranean anchovy. In: Document Presented to the GFCM-SAC Sub-Committee on Stock Assessment Working Group on Small Pelagic Species, Tanger, Morocco, March 2003.

Parrish, R.C., Nelson, C.S. and Bakun, A. (1981) Transport mechanisms and reproductive success of fishes in the Cali- fornia Current. Biol. Oceanog. 1:175–203.

Patti, B., Bonanno, A., Basilone, G. et al. (2004) Interannual fluctuations in acoustic biomass estimates and in landings of small pelagic fish populations in relation to hydrology in the Strait of Sicily. Chem. and Ecol. 20:365–375.

Peebles, E.R. (2002) Temporal resolution of biological and physical influences on bay anchovy Anchoa mitchilli egg abundance near a river-plume frontal zone. Mar. Ecol. Prog. Ser. 237:257–269.

Penven, P., Roy, C., Colin de Verdièr, A. and Largier, J. (2000) Simulation of a coastal jet retention process using a baro- tropic model. Oceanologica Acta 23:615–634.

Pertierra, J.P. (1992) Biologı́a pesquera de la anchoa, Engraulis encrasicolus, del litoral Catalán. Thesis, Barcelona, University of Barcelona, 281 pp.

Peterman, M.R. and Bradford, M.J. (1987) Wind speed and mortality rate of a marine fish, the northern anchovy (Eng- raulis mordax). Science 235:354–356.

Regner, S. (1996) Effects of environmental changes on early stages and reproduction of anchovy in the Adriatic Sea. Sci. Mar. 60:167–177.

Reynolds, R.W. and Smith, T.M. (1994) Improved global sea surface temperature analyses. J. Clim. 7:929–948.

Riveiro, I., Guisande, C., Lloves, M., Maneiro, I. and Cabanas, J.M. (2000) Importance of parental effects on larval survival in Sardina pilchardus. Mar. Ecol. Prog. Ser. 205:249–258.

Riveiro, I., Guisande, C., Franco, C., Lago de Lanzos, A., Ma- neiro, I. and Vergara, A.R. (2003) Egg and larval amino acid composition as indicators of niche resource partitioning in pelagic fish species. Mar. Ecol. Prog. Ser. 260:255–262.

Roa, R., Ernst, B. and Tapia, F. (1999) Estimation of size at sexual maturity: an evaluation of analytical and resampling procedures. Fish. Bull. 97:570–580.

Ryther, J.H. (1969) Relationship of photosynthesis to fish pro- duction in the sea. Science 166:72–76.

Sargent, R.C., Taylor, P.D. and Gross, M.R. (1987) Parental care and the evolution of egg size in fishes. Am. Nat. 129:32–46.

Scotto di Carlo, B. and Ianora, A. (1983) Standing stocks and species composition of Mediterranean zooplankton. UN- ESCO Rep. Mar. Sci. 20:59–69.

Silva, C., Yánez, E., Barbieri, M.A. et al. (2000) Exploring the Association between Small Pelagic Fisheries and SeaWiFS Chlorophyll and AVHRR Sea Surface Temperature in the North of Chile. Sixth International Conference on Remote Sensing for Marine and Coastal Environments, Charleston, South Carolina.

Sinovčić, G. (2000) Anchovy, Engraulis encrasicolus (LIN- NAEUS, 1758): biology, population dynamics and fisheries case study. Acta Adriat. 41:3–53.

Siokou Frangou, I. (1996) Zooplankton annual cycle in a Mediterranean coastal area. J. Plankton Res. 18:203–223.

Tsai, C.F., Chen, P.Y., Chen, C.P., Lee, M.A., Shiah, G.Y. and Lee, K.T. (1997) Fluctuation in abundance of larval anchovy environmental conditions in coastal waters off south-western Taiwan as associated with the El Niño-Southern Oscillation. Fish. Oceanogr. 6:238–249.

Tudela, S. and Palomera, S. (1997) Trophic ecology of the European anchovy Engraulis encrasicolus in the Catalan Sea (Northwest Mediterranean). Mar. Ecol. Prog. Ser. 160:121– 134.

280 G. Basilone et al.

� 2006 Blackwell Publishing Ltd, Fish. Oceanogr., 15:4, 271–280.