Biol. Bull. 211: 286-296. (December 2006)
© 2006 Marine Biological Laboratory
Stocking Density at Early Developmental Stages Affects Growth and Sex Ratio in the European Eel (Anguilla anguilla)
Mar Huertas and
Joan Cerdà*
Center of Aquaculture IRTA, 43540-Sant Carles de la Ràpita, Tarragona, Spain, and Reference Center in Aquaculture, Barcelona, Spain
* To whom correspondence should be addressed, at Lab IRTA-ICM, CMIMA-CSIC, Passeig Marítim 37–49, 08003-Barcelona, Spain. E-mail: joan.cerda{at}irta.es
 |
Abstract
|
|---|
To investigate the effect of stocking density on growth and sex ratio in European eel, four constant density conditions were tested during the transition from the glass to the elver stage for 90 days (Period 1). The test conditions combined the weight of fish per unit surface or volume (surface density or volume density) resulting in four experimental groups: low surface density (0.5 kg/m2) and low volume density (5 kg/m3) (group S0.5V5); low surface density (0.5 kg/m2) and high volume density (10 kg/m3) (group S0.5V10); high surface density (2 kg/m2) and low volume density (5 kg/m3) (group S2V5); and high surface density (2 kg/m2) and high volume density (10 Kg/m3) (group S2V10). Subsequently, fish from the S0.5V5, S2V5, and S2V10 groups were transferred to low density conditions (0.1–0.4 kg/m2 or 0.1–0.3 kg/m3) for another 21 months (630 days; Period 2). After Period 1, fish maintained at high surface density, regardless of the volume density, showed higher standard growth rates (SGRs) and RNA/DNA ratio in muscle than those cultured at low surface density. The percentage of mortality was similar in three of the groups (34.2%–41.8%), but not in the S2V10 group (83.3%). At the end of Period 2, most fish (about 95%) exhibited fully differentiated gonads, but different sex ratios were observed in each group. Thus, the S2V5 group showed a higher proportion of females (36.1%) and a lower proportion of males (56.8%) than the S0.5V5 group (11.4% and 72.5%, respectively), while all survivor fish from the S2V10 group developed into females. The gonadosomatic index and SGR were higher in females than in males. These results suggest that glass eels maintained at high surface density during the first months of growth tend to develop into females. The data also indicate that growth and sex ratio are linked processes during eel development, with growth seeming to be sex dependant rather than being influenced by the density conditions in which glass eels are maintained.
Abbreviations: GSI, gonadosomatic index • SGR, standard growth rate: in weight (SGRw) and in length (SGRL)
|
Introduction
|
|---|
The European eel (Anguilla anguilla, L. 1758) is a catadromous fish species that enters into the rivers from the sea at the so-called "glass" eel stage (i.e., transparent, mostly unpigmented fish). In the river, glass eels grow into the elver stage and pass through several stages of metamorphosis for 3 to 9 years, up to the "pubertal" yellow stage, when sex differentiation occurs (Acou et al. 2003). After about 3 to 6 more years, they reach the adult silver stage and remain in the river for a variable period of time (10 to 40 years; Walsh et al., 2004). Subsequently, adult fish return to their spawning area in the Sargasso Sea, where they reproduce and die; the resulting larvae return in a long migration to the European and North African coasts, where they metamorphose into glass eels (Tesch, 2003). Currently, the European eel is a highly appreciated species in the food market, so fishing for eels is an important activity in most European countries. But due to the fishing effort, natural populations are overexploited and this effect—together with other anthropogenic factors, such as loss of habitats, contamination, and transfer of parasites and diseases—has contributed to the decrease in the captures of both glass eels and adults during the last 20 years (ICES, 2004, 2005).
Developing aquaculture methods for the eel might reduce the fishery pressure and help to restock natural populations. This technology is well developed in European, Asian, and American countries (ICES, 2000), but several obstacles challenge its sustainability. The most important problem is that methods for controlling reproduction in captivity are lacking, so none or few viable larvae are obtained, and the culture is thus dependant on a constant input of wild glass eels. But due to the variability of these natural stocks, the growth rates of glass eels during the first months of rearing in captivity are highly heterogeneous, which produces high dispersion in weight and length of the populations and favors elevated mortalities by cannibalism (Gousset, 1990; Heinsbroek, 1991). Another important problem is that the sex ratio is skewed toward males in eel stocks under intensive culture conditions (Gousset, 1990; Colombo and Grandi, 1996; Holmgren, 1996; Beullens et al., 1997a). The reduction in the number of females in cultured stocks may hinder breeding programs, and it also decreases the ability to enhance natural populations with skewed sex ratios (Helfman, 1987; Oliveira et al., 2001).
The process of sex differentiation in teleosts is influenced by genetic, physiological, and environmental conditions (reviewed by Devlin and Nagahama, 2002; Piferrer et al., 2005). In controlled studies on the European eel, the culture density has been suggested as one important factor (Roncarati et al., 1997; Krueger and Oliveira, 1999; Davey and Jellyman, 2005), and it is believed that high density conditions lead to a higher proportion of males (Colombo and Grandi, 1996; Beullens et al., 1997a; Krueger and Oliveira, 1999; Davey and Jellyman, 2005). This hypothesis is supported by ecological studies which show that females predominate in upstream riverine habitats (low density conditions) (Ibbotson et al., 2002), while males predominate in tidal areas (high density conditions) (Oliveira and McCleave, 2000; Walsh et al., 2004). However, data on natural populations may be obscured by other factors, such as mortality (Davey and Jellyman, 2005), fishing preferences that favor one size (Vollestad and Jonsson, 1988; Naismith and Knights, 1990; Jellyman, 2001), sex-related migratory dispersion (Ibbotson et al., 2002), or a high proportion of sexually undifferentiated individuals (Krueger and Oliveira, 1999). These factors make it difficult to establish a direct relationship between sex and stocking density.
In addition, the mechanisms by which stocking density may influence sex differentiation in teleosts are unclear. In captive eel stocks, under high densities, the interference competition seems to decrease the growth rate, and individuals differentiate as males (Knights, 1982, 1987; Gousset, 1990; Usui, 1991; Colombo and Grandi, 1996; Tesch, 2003), so growth rate may be involved in the mechanism of sex differentiation (Davey and Jellyman, 2005). However, other studies have shown that males can grow faster than females (Holmgren et al., 1997; Davey and Jellyman, 2005). These apparently contradictory results may be related to differences in the experimental conditions or to the specific developmental stage of the animals employed. Therefore, the stocking density conditions and the size of individuals are two factors that should be controlled in an investigation of the effects of density on the mechanism of sex differentiation in eels.
In the present study, we have combined surface and volume densities to test the effect of four stocking densities on the growth and survival of European eels during the transition from the glass to the elver stage. In addition, the effect of the density conditions—applied transiently for 3 months—on subsequent growth rates and sex differentiation was investigated.
|
Materials and Methods
|
|---|
The experimental work reported here comprised three components: the preparation, a period in which a cohort of glass eels was prepared for testing (about 25 days); and two experimental intervals (Periods 1 and 2). In Period 1, the prepared glass eels were raised through the elver or yellow stages (90 days; 3 months); and in Period 2, the elvers from Period 1 were raised to the silver stage (630 days; 21 months). Therefore, to carry out the experimental design required a continuous commitment of 745 days (>2 years)
Preparation of fish for experimentation
Glass eels were obtained from local fish farms from the Delta del Ebro (Spain). The fish were kept in 200-1 tanks and treated for 1 h with 80 ppm formaldehyde; subsequently they were exposed twice within a 2-day interval to 100 ppm hydrogen peroxide. Then, during 4 days at natural temperature, the fish were acclimated to fresh water (1–2 psu), which was followed over 5 days by an increase of temperature from 15°C to 26°C. The fish were maintained at 26°C for 2 weeks, during which they were fed with cod roe. During this interval, the glass eels progressed from a completely unpigmented stage to a stage with one rostral spot. After this 25-day preparation period, the fish were distributed into the different experimental treatments.
Experimental design
All the experimental treatments were performed in recirculated, aerated fresh water maintained at constant temperature (26°C) and photoperiod (10L:14D). The experiment comprised two periods.
Period 1.
Glass eels (0.17 g in weight) were grown through the elver stage under different stocking densities for 90 days. The effect of total biomass per unit surface or per unit volume was investigated. The densities per unit surface (surface density) were 0.5 kg/m2 and 2 kg/m2, and per unit volume (volume density) were 5 kg/m3 and 10 kg/m3. Thus, after combining the two factors, the four experimental groups were designated S 0.5V5, S 0.5V10, S2V5, and S2V10, and each treatment had six replicates (n = 150 fish per replicate; 900 fish per treatment). The fish were placed into one of two types of cylindrical containers covered with a net on one end (Fig. 1). The cylinders had a diameter of either 0.25 m for treatments with high surface density, or 0.5 m for treatments with low surface density. These containers were distributed into six tanks of 1500 1, filled with water up to 500 1, and connected to the recirculation unit. To maintain the densities constant throughout the experiment, the changes in weight of the fish were determined every 10 days by quickly weighing each container. Each of the cylinders had a device to control the surface available to the fish, and by moving up the cylinder and down perpendicular to the water level in the tank, the volume available was controlled. Thus, the conditions of surface and volume densities were adjusted every 10 days and maintained approximately constant throughout the experiment (0.5 ± 0.1 kg/m2 and 2 ± 0.26 kg/m2, and 5 ± 0.6 kg/m3 and 10 ± 1.1 kg/m3; mean ± st. error). The fish were fed daily ad libitum (4% dry weight per fresh body weight) with a commercial feed (MICROBAQ 8, DIBAQ, Spain) that was manually hydrated to reach 80% of water content. Food ration was also adjusted every 10 days.
View larger version (50K):
[in this window]
[in a new window]
|
Figure 1. Diagram of the containers used for the culture of glass eels under different density conditions during Period 1.
|
|
Period 2.
A sample of fish from each treatment cylinder was collected for biochemical analysis. (see below). Following Period 1, the remaining fish were removed from the cylinders and released—for an additional growth period of 630 days—into the tanks of the recirculation unit, which were filled with water up to 1000 1. The replicates of the same density treatment from Period 1 (except the S0.5V10 group; see Results) were replaced into their original tank (n = 324 fish in S 0.5V5; n = 281 fish in S2V5; n = 60 fish in S2V10). At the start of Period 2, the fish had already reached the elver or yellow stage, and they were therefore fed daily ad libitum (4% dry weight per fresh body weight) with commercial extruded pellets for European eel (DIBAQ, Spain). Food ration was adjusted monthly to the total biomass, which was estimated in each tank successively by removing all its fish and weighing them. At the start of Period 2, the initial density in the S0.5V5 and S2V5 groups was 0.3–0.4 kg/m2 (0.2–0.3 kg/m3); but in the S2V10 group it was 0.1 kg/m2 (0.1 kg/m3). After 630 days, the density of the S0.5V5 and S2V5 groups was 18.3 and 20.1 kg/m2 (15.5 and 17.0 kg/m3), respectively, while the density of the S2V10 group was 3.6 kg/m2 (3.1 kg/m3).
Growth rates, and determination of DNA and RNA content
The standard growth rate (SGR) and the mortality in each experimental group were determined as follows. The fish in each group were counted and individually weighed each month during Period 1, and every 7 months during Period 2. At the end of both periods, the length of each fish was also measured. The SGR in weight (SGRw) was calculated as (lnWf– lnW0/T) x 100, where Wf is the mean weight of fish from each treatment in a given time, W0 is the mean weight in the previous time, and T is the time interval (days) (Appelbaum et al., 1998). The SGR in length (SGRL) was calculated using the analogous equation. The mortality was calculated as the percentage of individuals that disappeared within a time interval. The gonadosomatic index (GSI) was calculated as gonad weight/body weight x 100.
At the end of Period 1, a sample of fish (32%–44% per treatment) was sacrificed, frozen at –80°C, and used to estimate by biochemical analysis the growth of each group. A sample of liver and muscle was rapidly dissected from each fish, and the DNA and RNA content (µg/mg dry weight) was determinated by a fluorometric procedure following the method of Caldarone et al. (2001). Since the quantity of DNA is believed to be stable normally while the quantity of RNA, primarily associated with ribosomes, is closely related to the rate of protein synthesis, the ratio RNA/DNA was calculated as an index of fish growth (Kawakami et al., 1999; Gwak and Tanaka, 2001).
Histology
At the end of Period 2, a piece of the gonad from each fish was fixed in Bouin’s fixative for 2 h at room temperature, dehydrated, and embedded in paraffin wax. Sections (3 µm thick) were cut and stained with hematoxylin and eosin and examined with a Zeiss Axioskop 2 plus light microscope.
Statistics
Data were presented as means ± st. error of the mean (SEM). Data on SGRw (Period 1 and 2) were analyzed by three-way nested ANOVA (surface and volume densities nested by time) with a randomized complete block design (RCB) with unbalanced data, in which each treatment had six blocks or replicates (Cobb, 1998). After identifying interacting factors (surface and volume densities), statistical differences in SGRw between treatments were analyzed by two-way ANOVA with RCB for Period 1, and by one-way ANOVA at each time point for Period 2 (Cobb, 1998). The rest of the data, except those on sex ratio, were analyzed by one- or two-way ANOVA. In all cases, when the interaction between factors was statistically significant post hoc comparisons were performed following the Dunn or Holm-Sidack methods (Zar, 1996; Cobb, 1998) to detect differences between treatments. Data on sex ratio were statistically analyzed by 
2 goodness-of-fit test (Wilson and Hardy, 2002). Differences were considered significant at P < 0.05.
|
Results
|
|---|
Effects of stocking density on growth and mortality
During Periods 1 and 2, the activity and behavior of the fish depended markedly on the density conditions under which they were maintained. In general, fish cultured at higher surface density were more dynamic, with higher swimming and feeding activity, while those cultured in low surface density remained mainly grouped at the bottom of the containers.
Statistical analysis (three-way nested ANOVA) of data on SGRW indicated a significant (P < 0.02) interaction surface x volume, while the standard growth rate in weight (SGRW) in all groups significantly (P < 0.001) changed during the culture period. Thus, at day 30, fish at low surface and volume densities (S0.5V5) showed higher SGRW than fish at higher densities (S2V5 and S2V10) (Fig. 2). However, at day 60 and day 90, the growth pattern under each treatment was reversed, since the SGRW of the S2V5 and S2V10 groups was higher than that of the S0.5V5 and S0.5V10. Thus, the mean weight of fish at the end of Period 1 was 1.2 ± 0.1 g for the S0.5V5 and S0.5V10 groups, and 2.0 ± 0.1 g and 2.6 ± 0.2 g for S2V5 and S2V10, respectively, which indicated that surface density rather than volume density was the main factor affecting growth.
View larger version (42K):
[in this window]
[in a new window]
|
Figure 2. Standard growth rate in weight (SGRw) of fish maintained under different density conditions during Periods 1 and 2. Data are expressed as the mean ± SEM (n = 6 replicates per treatment). Values with different letters indicate statistically significant (P < 0.05) differences at a given sampling time, while the numbers indicate significant differences among sampling times (P < 0.001) (see Materials and Methods).
|
|
By the end of Period 1, all of the fish had developed to the elver stage, and some of them had reached the yellow stage. At this point, the fish from the S0.5V10 group were discarded since they showed the same SGRW after 90 days as the S0.5V5 group. All of the fish from the remaining groups (S0.5V5, S2V5 and S2V10)—with their six replicate subgroups combined—were transferred to separate tanks where they were maintained for 630 days (Period 2). In contrast to the differences in growth among treatments during Period 1, all groups showed a gradual decrease in SGRW during Period 2 (Fig. 2). However, at day 300, the SGRW of the S2V5 group decreased more than that of the S0.5V5 and S2V10 groups, while at 720 days the SGRW of group S0.5V5 was the lowest and that of S2V10 was the highest. At the end of the experiment, the mean weight of fish from S0.5V5, S2V5 and S2V10 was, respectively, 114 ± 7 g, 163 ± 16 g, and 609 ± 75 g, the differences among the three groups being statistically significant (P < 0.05). During Period 2, fish from the three groups reached the silver stage (100% in the S2V10, and 60%–70% in the S0.5V5 and S2V5 groups), as judged by the increase of eye size, the change in skin color, and the silvering of the swim bladder (Oliveira and McCleave, 2000; Durif et al., 2005). Mortality was the same in all groups during the first 510 days (27%–36%), but during the following 210 days, no fish died at all.
The differences in SGRW between treatments detected after Period 1 correlated with the RNA/DNA ratio in liver and muscle of fish (Fig. 3). In muscle, and independently of the volume conditions, the RNA/DNA ratio was 2–3 times higher in high surface density fish than in lower surface density fish (two-way ANOVA, P < 0.05). In the liver, however, the interaction surface x volume was statistically significant (P < 0.01), only the S2V10 group showing a higher (P < 0.05) RNA/DNA ratio.
View larger version (37K):
[in this window]
[in a new window]
|
Figure 3. RNA/DNA ratio in liver and muscle from fish maintained under different density conditions. The ratios were determined at the end of Period 1 (90 days). Data are means ± SEM (n = 20–50 fish). In a given tissue, different letters at the top of each bar indicate statistically different values (P < 0.05).
|
|
Independent statistical analysis at each sampling time during Period 1 (days 30, 60, and 90) indicated that the cumulative percentage of mortality was dependant on the density conditions (two-way ANOVA; significant interaction surface x volume, P < 0.03) (Table 1). At day 90, the percentage of mortality was similar among the S0.5V5, S0.5V10, and S2V5 groups (ranging from 34.2% ± 3.5% to 41.8% ± 4.1%), while in the S2V10 group this percentage was significantly (P < 0.05) higher (83.3% ± 5.3%).
Progress of metamorphic stages
The effects of the four density treatments on the progress of glass eels, as they advanced through metamorphosis, was assessed at the end of Period 1. In brief, the fish (n = 100–350 per treatment) were anaesthetized with phenoxyethanol (5 ppm), and the percentage of eels at each metamorphic stage was determined with reference to the classification of such stages by Elie et al. (1982). The classification is based upon a clearly described series of patterns of rostral and caudal pigmentation in developing eels.
The fish were initially obtained as transparent, unpigmented glass eels (stages IV–VA). By the end of their 25-day preparation for these experiments (i.e., the beginning of Period 1), the glass eels in every treatment had progressed from lacking pigment to a stage with one rostral spot (stage VB). The spot is due to one or a few melanophores that are visible on the surface of the brain; the rest of the body remains transparent.
By the end of Period 1, the surviving fish had reached the elver or yellow stage. Those from the S0.5V5, S0.5V10, and S2V5 groups were at stages VIA4, VIB, and VII, so we could detect no statistical differences between the treatments (two-way ANOVA; Table 2). Although fish from the S2V10 group seemed to have advanced through metamorphosis somewhat further than the other treatment groups, the differences were not statistically significant.
Sex ratio and growth rates
At the end of Period 2, all fish were sacrificed, and the sex and stage of gonad development was evaluated. Undifferentiated and ambisexual (Syrski) gonads, as well as testis and early ovaries, were very difficult to differentiate macroscopically since all appeared as a slightly lobed rope located in the dorsal part of the abdominal cavity. Therefore, they were classified by histological examination (Table 3). Fully developed ovaries were, however, easier to identify macroscopically, since they are larger than undifferentiated and Syrski gonads and testis and appear characteristically as paired undulate ribbons that cover the abdominal cavity laterally from the liver to the anus.
Macrocroscopic and histological analysis revealed that the sex ratio in each experimental group was different (Table 4). Both the S0.5V5 and S2V5 groups showed a similar proportion of fish with either undifferentiated or Syrski gonads (Fig. 4A), but the S0.5V5 group showed a higher proportion of males and a lower ratio of females than the S2V5 group. In contrast, all fish from the S2V10 group were females.
View larger version (92K):
[in this window]
[in a new window]
|
Figure 4. Photomicrographs of histological sections of eel undifferentiated gonad and testis. (A) Undifferentiated gonad showing Sertoli cells (Sc) surrounding primordial germ cells (PGC). (B) Testis at stage I showing Sc and Leydig cells (Lc), spermatogonia type A (SgA), and spermatogonia type B (SgB). (C) Testis at stage II showing Sc and Lc, SgA, SgB, and spermatocytes (Spc). Ct, connective tissue; St, seminiferous tubule. Scale bars, 20 µm.
|
|
Detailed histological analysis indicated that, in all treatments, the testis of males developed up to stage I, which showed SgA and SgB, or to stage II, where the appearance of Spc suggested that meiosis had started (Fig. 4B, C). In females, the ovary developed to the early ovary stage or up to the oil droplet stage, in which yolk incorporation (vitellogenesis) is initiated (Fig. 5). More advanced stages of gonadogenesis were not found under any treatment. However, the density conditions did appear to affect the development of the gonads (Table 4). In the S2V5 group, males showed a more advanced stage of spermatogenesis than the S0.5V5 males, while the females from this group had less advanced oogenesis in comparison with the S0.5V5 females. In the S2V10 group, all females were at the oil droplet stage, the most advanced stage of oogenesis observed during the experiment.
View larger version (99K):
[in this window]
[in a new window]
|
Figure 5. Photomicrographs of histological sections of eel previtellogenic ovary at early stage (A), growing oocyte stage (B), and oil droplet stage (C). Ct, connective tissue; Sg, spermatogonia; Lc, Leydig cell; Yv, yolk vesicle (also called Balbiani body); Od, oil droplet; O1, O3, and O4, oocytes type 1, 3, and 4, respectively. Scale bars, 200 µm.
|
|
Once the sex and stage of gonad development were determined, the gonadosomatic index (GSI) and standard growth rates in width (SGRW) and length (SGRL) of males and females at each gonad stage were further analyzed. The results indicated that these indices were dependent on the sex and stage of gonad development, and were not significantly affected by the density conditions under which the glass eels were maintained (three-way ANOVA: P < 0.26, P < 0.08, and P < 0.07, for GSI, SGRW, and SGRL, respectively; data not shown). Thus, data from the density treatments were grouped by gonad type and stage and were analyzed by one-way ANOVA (Table 5). Females with early ovaries and males with testis at stage I or II had similar GSI, whereas the GSI of females, but not that of males, progressively increased with advanced stage of oogenesis (up to 12 times in females with ovaries at the oil droplet stage). The GSI of fish carrying undifferentiated or Syrski gonads was not determined because the small size of these gonads made extraction and accurate weight measurements difficult. The SGRW and SGRL for fish at each gonad stage showed the same pattern: they were the lowest in fish with Syrski and undifferentiated gonads, followed by males and then females. Within each sexual stage, the SGRW and SGRL increased with advanced stages of gonad development. However, females with early ovaries had growth rates similar to those of males with testis at stage I, and they grew less than males with testis at stage II.
View this table:
[in this window]
[in a new window]
|
Table 5 Gonadosomatic index (GSI), standard growth rates in weight (SGRw) and length (SGRL) of fish at each gonad stage after Period 2
|
|
|
Discussion
|
|---|
The effect of stocking density on the growth of cultured European glass eels is poorly documented. Some authors have suggested that low densities are associated with higher growth rates, but some of these studies were limited to volume density or were biased by increases in mortality or density during the experiments (Degani, 1983, 1988; Degani et al., 1985; Roncaratti et al., 1997). In the present work, an elevated surface density (S2V5 and S2V10 groups) applied for 90 days resulted in the highest growth rates (Fig. 2), although high density conditions in terms of both surface and volume (S2V10) dramatically increased mortality (80%; Table 1).
Consistent with these observations, the RNA/DNA ratio in muscle was higher in the S2V5 and S2V10 groups than in the S0.5V5 and S0.5V10 groups (Fig. 3). However, the stocking densities tested apparently did not affect the metamorphic processes occurring during the transition to the elver stage (Table 2), which agrees with the fact that the pigmentation processes of developing eels are independent of their growth rates (Jegstrup and Rosenkilde, 2003).
The differences in standard growth rate in weight (SGRW) among the S0.5V5, S0.5V10, and S2V5 groups were attributable only to the density conditions, since the level of mortality in these groups was similar (Table 1), and furthermore, was comparable to that observed in other works (Degani, 1983; Gousset, 1992; Roncaratti et al., 1997). In the S2V10 group, however, a much higher mortality was recorded, which may be related to the poor water quality caused by the high density.
The different conditions of surface density tested here through the elver stage (Period 1) were associated with different sex ratios after the second period of growth up to the silver stage (Period 2). Thus, the groups maintained at high surface density (S2V5) had a higher proportion of females and a lower proportion of males than those under low surface density (S0.5V5) (Table 4). In both groups, the percentage of fish with undifferentiated and Syrski (ambisexual) gonads was low, and therefore their number did not affect the final sex ratio. The S2V10 group was 100% females, which may support our conclusion that high densities result in a higher proportion of females. But the level of mortality of this group was higher than in other groups during Period 1, and this high mortality may have influenced the sex ratio. In any case, our observations are apparently in contrast with those reported by Roncarati et al. (1997), who found that high density treatments produced a high ratio of males. In their study, however, the density treatments were calculated per unit volume, while the surface densities (0.03 to 0.12 kg/m2) were, in fact, close to or lower than the low surface density conditions used in our study. Therefore, our results may agree with those of Roncarati et al. (1997). In eel farms, fish are cultured at high densities and the majority of individuals are males (Gousset, 1990). However, the actual initial densities at which glass eels are maintained are usually not higher than 0.2–0.3 kg/m2 (Usui, 1991; Gousset, 1992; Liao et al., 2002), which also supports our data. In wild populations of eels, several authors have observed a high proportion of females in upper riverine areas where density is low, while males are abundant in lower river areas and streams where density is high (De Leo and Gatto, 1996; Holmgren et al., 1997; Walsh et al., 2004). In these studies, however, the effect of density is difficult to separate from other factors, such as fishing, diseases, food availability, and migration patterns along the river (Tzeng, 2000; Ibbotson et al., 2002; Davey and Jellyman, 2005). Thus, the relationship between density and sex ratio in wild stocks is unclear.
In the present experiment, different density treatments were applied for 90 days during the transition from the glass to the elver stage and were effective at altering the sex ratio (Table 4). These results suggested that during the selected 90-day period of early eel development, the eels might have been sensitive to the effects of density on sex differentiation. After 90 days, the gonads were not yet morphologically differentiated, since gonadal differentiation usually occurs at the yellow stage (i.e., eels of 20–22 cm in length, which appeared at about 300 days in our experiment) (Colombo and Grandi, 1996). Altogether, our observations could indicate that in the European eel, as in other fish species (Devlin and Nagahama, 2002), the physiological mechanisms underlying sex differentiation may precede its morphological expression. Earlier reports, however, have proposed that effective sex reversal in the eel must cover both physiological and morphological differentiation (Chiba et al., 1993; Davey and Jellyman, 2005), so exposure to different densities for longer than 90 days might have been more effective at increasing the differences in sex ratio. But glass eels are more sensitive than later stages to hormonal sex reversal (Colombo and Grandi, 1995; Grandi et al., 2000); therefore, the effect of density on sex differentiation might indeed have occurred early during the 90-day period of transition from the glass to the elver stage.
As with other eel species (Holmgren et al., 1997; Oliveira, 2002; Walsh et al., 2003), we observed a relationship between growth rate and sex ratio (Table 5). Our data confirm that females have higher SGRs than males (Holmgren and Mosegaard, 1996; Oliveira, 2002), but they are in conflict with reports that males grow faster than females (Voellestad and Jonsson, 1988; Holmgren et al., 1997). In these latter studies, however, fish were grown from 10 g to 80 g (i.e., from 300 days until 500 days in our experiment) — a period during which we also observed a higher SGRW in the group with more males (S0.5V5). But females differentiate during this time, and thus the energetic cost of developing the ovary may have compromised body growth. Nevertheless, it is difficult to determine which of the two factors, growth rate or sex differentiation, precedes the other. Some authors have suggested that growth rate is one of the key determinants of subsequent sex differentiation, which implies that fish with the same SGR have the same sex (Helfman, 1987; Davey and Jellyman, 2005). However, in a given stock, slow-growing fish have been identified as females (Chiba et al., 1993). In our study, females with early and developing ovaries had SGRW and SGRL higher than those of males with early testis, but similar to those of males with testis at advanced stages (stage I and II). These observations thus support our view that, in the eel, sex differentiation occurs before changes in growth rates.
The mechanism by which density may affect sex differentiation and growth remains uncertain, although several authors have suggested that social interactions play a role (Devlin and Nagahama, 2002; Davey and Jellyman, 2005). During the present work, we also observed differences in behavior that depended on the density treatment and might lead to different life strategies. Thus, fish maintained at low surface density became territorial, gregarious, and less active, whereas fish at high density lacked the space for a proper habitat and became active swimmers, which could enhance aggressive behavior and stimulate the intake of food or the food conversion ratio. The role of odorant signals may also be involved in the interaction between density and sex differentiation (Devlin and Nagahama, 2002) since eels have a highly developed olfactory epithelium sensitive to the odorants of conspecifics (Pesaro et al., 1981; Saglio, 1982; Huertas et al., 2006). In fact, the walls and bottom of the containers in which our fish lived were covered with mucus (although the rate of water circulation was at standard levels), and mucus has been described as a highly active conspecific odorant in the eel (Huertas et al., 2005).
In summary, the present work shows that surface density is the most critical factor affecting the growth and sex ratio of glass eels. Regardless of volume density, treatments with high surface density resulted in higher standard growth rates and a higher proportion of females, suggesting that sex differentiation and growth rate are linked processes in European eel. Our data suggest that the labile period during which density may affect sex differentiation might occur within the first 3 months of growth, indicating that physiological sex differentiation occurs earlier than its morphological expression. Further studies will be necessary to elucidate the specific physiological processes that mediate the effect of stocking density on sex differentiation in the eel.
|
Acknowledgments
|
|---|
We thank Dr. M. Blázquez for critical reading of the manuscript and helpful discussions. We also thank J.L. Celades, G. Maciá, and M. Monllaó for their assistance during fish maintenance and sampling. This work was supported by grants from the Reference Center in Aquaculture (Generalitat de Catalunya, Spain) to J.C. M.H. was the recipient of a fellowship from the Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA, Spain).
|
Footnotes
|
|---|
Received 6 June 2006; accepted 28 August 2006.
|
Literature Cited
|
|---|
Acou, A., F. Lefebre, P. Contournet, G. Poizat, J. Panfili, and A. J. Crivelli. 2003. Silvering of female eels (Anguilla anguilla) in two sub-populations of the Rhône delta. Bull. Fr. Pêche Piscic. 368:55–68.
Appelbaum, S., A. Chernitsky, and V. Birkan. 1998. Growth observations on European eel (Anguilla anguilla L.) and American (Anguilla rostrata Le Seur) glass eels. Bull. Fr. Pêche Piscic. 349:187–193.
Beullens, K., E. H. Eding, P. Gilson, F. Ollevier, J. Komen, and C. J. J. Richter. 1997a. Gonadal differentiation, intersexuality and sex ratios of European eel (Anguilla anguilla L.) maintained in captivity. Aquaculture 153:1–2.
Beullens, K., E. H. Eding, F. Ollevier, J. Komen, and C. J. J. Richter. 1997b. Sex differentiation, changes in length, weight and eye size before and after metamorphosis of European eel (Anguilla anguilla L.) maintained in captivity. Aquaculture 153:1–2.
Caldarone, E. M., M. Wagner, J. S. Onge-Burns, and L. J. Buckley. 2001. Protocol and guide for estimating nucleic acids in larval fish using a fluorescence microplate reader. P. 21, Northeast Fisheries Science Center Reference Document 01-11, National Marine Fisheries Service, Woods Hole, MA.
Chiba, H., K. Iwatsuki, K. Hayami, and K. Yamauchi. 1993. Effects of dietary estradiol-17 beta on feminization, growth and body composition in the Japanese eel (Anguilla japonica). Comp. Biochem. Physiol. A Comp. Physiol. 106A:367–371.
Cobb, G. W. 1998. Design and Analysis of Experiments. Springer-Verlag, New York.
Colombo, G., and G. Grandi. 1995. Sex differentiation in the European eel: histological analysis of the effects of sex steroids on the gonad. J. Fish Biol. 47:394–413.
Colombo, G., and G. Grandi. 1996. Histological study of the development and sex differentiation of the gonad in the European eel. J. Fish Biol. 48:493–512.
Davey, A., and D. Jellyman. 2005. Sex determination in freshwater eels and management options for manipulation of sex. Rev. Fish Biol. Fish. 15:37–52.
Degani, G. 1983. The influence of low density on food adaptation, cannibalism and growth of eels (Anguilla anguilla (L.)). Bamidgeh 35:53–60.
Degani, G. 1988. Influence of high loading density on growth and survival of European glass eels. Prog. Fish-Cult. 50:178–181.
Degani, G., D. Levanon, and C. Dosoretz. 1985. Growth of Anguilla anguilla densities in out-door containers with Tilapia aurea. Prog. Fish-Cult. 47:114–117.
De Leo, G. A., and M. Gatto. 1996. Trends in vital rates of the European eel: evidence for density dependence? Ecol. Appl. 6:1281–1294.
Devlin, R. H., and Y. Nagahama. 2002. Sex determination and sex differentiation in fish: an overview of genetic, physiological, and environmental influences. Aquaculture 208:191–364.
Durif, C., S. Dufour, and P. Elie. 2005. The silvering process of Anguilla anguilla: a new classification from the yellow resident to the silver migrating stage. J. Fish. Biol. 66:1025–1043.
Elie, P., R. Lecomte-Finiger, I. Cantrelle, and N. Charlon. 1982. Définition des limites des différents stades pigmentaires durant la phase civelle d’Anguilla anguilla L. (Poisson téléostéen anguilliforme). Vie Milieu 32:149–157.
Gousset, B. 1990. European Eel (Anguilla anguilla L.) farming technologies in Europe and in Japan: application of a comparative analysis. Aquaculture 87:209–235.
Gousset, B. 1992. Eel Culture in Japan. Musée Océanographique, Monaco.
Grandi, G., F. Poerio, G. Colombo, and M. Chicca. 2000. Effects of diet supplementation with carp ovary on gonad differentiation and growth of the European eel. J. Fish Biol. 57:1505–1525.
Gwak, W.S., and M. Tanaka. 2001. Developmental change in RNA:DNA ratios of fed and starved laboratory-reared Japanese flounder larvae and juveniles, and its application to assessment of nutritional condition for wild fish. J. Fish Biol. 59:902–915.
Heinsbroek, L. 1991. A review of eel culture in Japan and Europe. Aquacult. Fish Manag. 22:57–72.
Helfman, G., J. 1987. Reproductive ecology of the American eel. Pp. 42–56 in American Fisheries Society Symposium: 1. M. K. Dadswell, C. M. Moffitt, R. L. Saunders, R. A. Rulifson, and J. E. Cooper, eds. Int. Symp. on Common Strategies of Anadromous and Catadromous Fishes. 9–13 March, 1987. Boston, MA.
Holmgren, K. 1996. Effect of water temperature and growth variation on the sex ratio of experimentally reared eels. Ecol. Freshw. Fish 5:203–212.
Holmgren, K., and H. Mosegaard, 1996. Implications of individual growth status on the future sex of the European eel. J. Fish Biol. 49:910–925.
Holmgren, K., H. Wickstrom, and P. Clevestam. 1997. Sex-related growth of European eel, Anguilla anguilla, with focus on median silver eel age. Can. J. Fish. Aquat. Sci. 54:2775–2781.
Huertas, M., P. C. Hubbard, A. P. Scott, A. V. M. Canário, and J. Cerdá. 2005. Evidence for involvement of chemical communication in reproduction of the eel Anguilla anguilla. Comp. Biochem. Physiol. A 141:S90 (Abstract).
Huertas, M., A. P. Scott, P. C. Hubbard, A. V. M. Canário, and J. Cerdá. 2006. Sexually mature European eels (Anguilla anguilla L.) stimulate gonadal development of neighbouring males: possible involvement of chemical communication. Gen. Comp. Endocrinol. 147:304–313.[ISI][Medline]
Ibbotson, A., J. Smith, P. Scarlett, and M. Aprhamian. 2002. Colonisation of freshwater habitats by the European eel Anguilla anguilla. Freshw. Biol. 47:1696–1706.
ICES. 2000. International Council for the Exploration of the Sea. Report of the ICES/EIFAC Working Group on Eels. ICES CM 2000/ACFM:6. 87.
ICES. 2004. International Council for the Exploration of the Sea. Report of the ICES/EIFAC Working Group on Eels. ICES CM 2004/ACFM:09. 87.
ICES. 2005. International Council for the Exploration of the Sea. Report of the ICES/EIFAC Working Group on Eels. ICES CM 2005/ACFM:01. Ireland. 71.
Jegstrup, I. M., and P. Rosenkilde. 2003. Regulation of post-larval development in the European eel: thyroid hormone level, progress of pigmentation and changes in behaviour. J. Fish Biol. 63:168–175.
Jellyman, D. J. 2001. The influence of growth rate on the size of migrating female eels in Lake Ellesmere, New Zealand. J. Fish Biol. 58:725–736.
Kawakami, Y., N. Mochioka, R. Kimura, and A. Nakazono. 1999. Seasonal changes of the RNA/DNA ratio, size and lipid contents and immigration adaptability of Japanese glass-eels, Anguilla japonica, collected in northern Kyushu, Japan. J. Exp. Mar. Biol. Ecol. 238:1–19.
Knights, B. 1982. Body dimensions of farmed eels (Anguilla anguilla L.) in relation to condition factor grading sex and feeding. Aquacult. Eng. 1:297–310.
Knights, B. 1987. Agonistic behaviour and growth in the European eel, Anguilla anguilla L., in relation to warm-water aquaculture. J. Fish Biol. 31:265–276.
Krueger, W. H., and K. Oliveira. 1999. Evidence for environmental sex determination in the American eel. Anguilla rostrata. Environ. Biol. Fishes 55:381–389.
Liao, I. C., Y.-K. Hsu, and W. C. Lee. 2002. Technical innovations in eel culture systems. Rev. Fish. Sci. 10:433–450.
Naismith, I. A., and B. Knights, 1990. Modelling of unexploited and exploited populations of eels, Anguilla anguilla (L.), in the Thames estuary. J. Fish Biol. 37:975–986.
Oliveira, K. 2002. Sexually different growth histories of the American eel in four rivers in Maine. Trans. Am. Fish. Soc. 131:203–211.
Oliveira, K., and J. D. McCleave. 2000. Variation in population and life history traits of the American eel, Anguilla rostrata, in four rivers in Maine. Environ. Biol. Fishes 59:141–151.
Oliveira, K., J. D. McCleave, and G. S. Wippelhauser. 2001. Regional variation and the effect of lake:river area on sex distribution of American eels. J. Fish Biol. 58:943–952.
Pesaro, M., M. Balsamo, G. Gandolfi, and P. Tongiorgi. 1981. Discrimination among different kinds of water in juvenile eels, Anguilla anguilla (L.). Monit. Zool. Ital. 15:183–191.
Piferrer, F., M. Blázquez, L. Navarro, and A. González. 2005. Genetic, endocrine, and environmental components of sex determination and differentiation in the European sea bass (Dicentrarchus labrax L.). Gen. Comp. Endocrinol. 142:102–110.[ISI][Medline]
Roncarati, A., P. Melotti, O. Mordenti, and L. Gennari. 1997. Influence of stocking density of European eel (Anguilla anguilla, L.) elvers on sex differentiation and zootechnical performances. J. Appl. Ichthyol. 13:131–136.
Saglio, P. 1982. Piégeage d’anguilles (Anguilla anguilla L.) dans le milieu naturel au moyen d’extraits biologiques d’origine intraspécifique. Mise en évidence de I’attractivité phéromonale du mucus épidemique. Acta Ecol. 3:223–231.
Tesch, F.-W. 2003. The Eel. Blackwell Science, Oxford, United Kingdom.
Tzeng, W. 2000. Differences in size and growth rates of male and female migrating Japanese eels in Pearl River, China. J. Fish Biol. 57:1245–1253.
Usui, A. 1991. Eel Culture. 2nd ed. Blackwell Science, London.
Vollestad, L., and B. Jonsson. 1988. A 13-year study of the population dynamics and growth of the European eel Anguilla anguilla in a Norwegian river: evidence for density-dependent mortality, and development of a model for predicting yield. J. Anim. Ecol. 57:983–997.
Walsh, C. T., B. C. Pease, and D. J. Booth. 2003. Sexual dimorphism and gonadal development of the Australian longfinned river eel. J. Fish Biol. 63:137–152.
Walsh, C. T., B. C. Pease, and D. J. Booth. 2004. Variation in the sex ratio, size and age of longfinned eels within and among coastal catchments of south eastern Australia. J. Fish Biol. 64:1297–1312.
Wilson, K., and I. C. W. Hardy. 2002. Statistical analysis of sex ratios: an introduction. Pp. 48–92 in Sex Ratios: Concepts and Research Methods, I. C. Hardy, ed. Cambridge University Press, Cambridge.
Zar, J. H. 1996. Biostatistical Analysis. Prentice-Hall, Upper Saddle River, NJ.
TOP
Abstract
Introduction
