Estimates of escapement, exploitation rate, and number of downstream migrating European eels Anguilla anguilla in Ichkeul Lake (northern Tunisia)

Abstract

The European eel (Anguilla anguilla) stock has been declining for the last three decades and today, it is currently listed as endangered. The objective of this study was to quantify the escapement rate of silver eels to obtain an estimate of the future spawners migrating from the southern part of its distribution area in Tunisia. A mark–recapture experiment was conducted in Ichkeul Lake (Tunisia) between December 2013 and February 2014, covering the winter downstream run of eels. The size range of the downstream migrating eels was between 33 and 79 cm with 585.3 ± 156.77 g mean weight. The number of migrating silver eels was estimated to 342 221 (297 956–386 486), corresponding to a biomass of 200.2 (174.3–226.1) t, with a density of 23.55 (20.51–26.6) kg ha⁻¹. The commercial fishing rate was estimated to be 18.8% (16.6–21.5%), which translates into an escapement rate of 81.2% (78.5–83.4%). The ratio current/pristine escapement was 0.69 for the entire migration period.

Introduction

The European eel (Anguilla anguilla) stock is at a historical minimum and is outside safe biological limits (ICES, 2014). Recruitment has decreased dramatically across most of Europe since the early 1980s (Elie and Fontenelle, 1982; Dekker, 2003; ICES, 2014) and has declined by an estimated 50% in the past 10 years in northern Africa (Azeroual, 2010). Various factors have been implicated such as loss and degradation of habitat, contamination by POP and heavy metals, parasitism by Anguillicola crassus, virus (Evex), predation, barriers to migration, and overexploitation by fishing at all stages of the continental phase (from glass eels to silver eels). The European eel is currently considered as an endangered species and is consequently included in CITES Appendix II list (2007) and in IUCN Red List (2008; Jacoby and Gollock, 2014). The European Commission has adopted a regulation [(EC) No. 1100/2007], inviting member states to submit an Eel Management Plan for each river basin which constitutes a natural habitat for this species.

The increase in silver eel escapement is a primary management target within each management plan (Robinet et al., 2007). The Council Regulation urged the escapement of 40% silver eels pristine biomass (i.e. without any anthropogenic influence) as a minimum for stock recovery.

Various methods have been set up and tested to estimate silver eel escapement: directly by catching and/or counting silver eels, proxy indicators based on knowledge of yellow eel populations, and model predictions and extrapolations. ICES (2010) and Walker et al. (2011) reviewed these approaches. Total escapement estimates, where the whole run of silver eel is intercepted are rare and generally estimates have to be derived from a mark–recapture study or direct counts (ICES, 2010). Several studies on silver eels stock have been carried out across the north side of the European eel distribution area (Rosell et al., 2005; Aprahamian et al., 2007; Breteler et al., 2007; Bilotta et al., 2011; Andersson et al, 2012; Prigge et al., 2013; Marohn et al., 2014) and a few in the southern part (Bevacqua et al., 2007; Amilhat et al., 2008; Charrier et al., 2012) but none in North Africa. However, given the current fragility of the European eel populations in the Atlantic and Mediterranean, a joint effort of the countries bordering the Mediterranean appears as a necessary condition for the proper management of this resource. Rarely taken into account in the population functioning models at the international level, the eel population fractions populating the Mediterranean hydro-systems must now be integrated (Farrugio and Elie, 2011). In this context, Tunisia chose to participate in the collective effort by writing its own Eel Management Plan (DGPA, 2010a).

Before the current trade ban came into force, Tunisia was the largest exporter of live eels to the EU, after Norway (Crook, 2010). Export accounted for 80–90% of the catches, as there is little tradition of eel consumption in this country (DGPA, 2010a). This work aims to fill gaps regarding the state of the stock, the escapement rate, and the exploitation rate of silver eels from Ichkeul Lake, an important hydro-system in Tunisia and one of the main sites of eel exploitation.

Material and methods

Study area

The Ichkeul Lake is located in the southeast of Bizerta in northern Tunisia on the central south Mediterranean coasts (Figure 1). Covering an area of 85 km², this lake is a part of the Ichkeul National Park which includes a mountain “the Jebel” and marshes. It is registered with three international conventions: Biosphere Reserve in 1977 (UNESCO), World Heritage component in 1979 (UNESCO), and Wetland of International Importance in 1980 (Ramsar Convention). The lake is indirectly connected to the sea through the lagoon of Bizerta via Wadi Tinja. It has an average depth of 1.5 m and receives freshwater from six wadies, three of which are equipped with dams. During winter, those wadies supply the lake with freshwater which once mixed with the rainfall, causes a rise in water level, and flows towards the lagoon of Bizerta. On the other hand, the combination of drought and low water drop causes a reversal of water flow during summer and, therefore, an increase in salinity. The salinity fluctuates thus between 16 and 34 g l⁻¹ and temperatures are between 11.8 and 29.1°C (Ramdani et al., 2001).

Ichkeul Lake is the subject of a traditional fishing based on fish migration (Romdhane, 1998; Chaouachi and Ben Hassine, 2001). Since 1998, this activity has been granted to a private operating company (Tunisia Lagoons Company SLT). The exploitation is carried out on a stock of marine origin, which depends on the arrival of juveniles from the sea, basically mullet fry and glass eels (Lemoalle and Vidy, 1983; Kraïem et al., 2003). Eels represented 26.1% of the total catches in 2009 (DGPA, 2010b). Only silver eels are exploited during the migration period. Yellow eels are not particularly targeted but are subject to bycatch and glass eel fishing is prohibited by a ministerial order (JORT, 1995).

The downstream migration season of A. anguilla occurs in Tunisia between late October and early February of each year (Hizem Habbechi, 2014). In addition to the traditional weir, and after obtaining an authorization, fishers set up a net barrier to catch the migrating eels (Figure 1). This net barrier is not completely closed since it is divided into two halves to provide a passage for small boats. It consists of a set of fykenets “capétchades” which are passive gears with 10 mm mesh traditionally used in the Mediterranean lagoons. Traps are emptied daily by fishers and contain both silver and yellow eels.

Eel samples

Only the silver eels were considered for the colour-marking experiment, but the yellows were also counted to know their percentage in total catch. To verify silver eel stage, we based on qualitative criteria: a contrast in colour between a dark dorsal surface, and a silvery ventral surface; the presence of a well-differentiated lateral line on the flanks and large eyes (Acou et al., 2005).

Colour-marking and recapture experiment

The colour-marking experiment took place twice: a first campaign from 9 to 12 December 2013 in which 701 silver eels were tagged with green and a second campaign in 6 January 2014 in which 284 silver eels were tagged with red. To facilitate eel handling, they were anaesthetized using Eugenol (diluted in ethanol) at a concentration of 6 ml l⁻¹. Each silver eel was weighed to the nearest gramme and sized to the nearest millimetre. Sex was assigned according to the size of individuals (i.e. eels with a total length longer than 45 cm were considered female; De Leo and Gatto, 1995; Tesch, 2003; Durif et al., 2005). Eels were then tagged with acrylic paint (“Van Gogh” permanent green light 618) or ink (“Magic Color” process magenta MC 620) inside the base of the dorsal fin with a syringe. This kind of paint was previously used by Amilhat et al. (2008) and Acou et al. (2010). Cotton soaked in iodine-based antiseptic (Betadine^®) was applied on the sting's spot to prevent infection. Eels were subsequently placed in aerated tanks until their total recovery. The release operations were made after each campaign (13 December and 7 January, respectively) in the lake upstream the net barrier (Table 1). Recaptured eels were recorded and checked for marks by fishers every day until the end of the migration period in February when they removed the net barrier.

Data analysis

The number of migrating silver eels was estimated by the Stratified Population Analysis System (SPAS software; Arnason et al., 1996) using the Pooled Petersen estimator (Seber, 1982). This method was used in similar studies to estimate silver eels stock (Caron et al., 2003; Amilhat et al., 2008; Charrier et al., 2012). The confidence intervals (CI) were calculated at α = 0.05. To have the total silver eels catch, we eliminated yellow eels from the total catch obtained from fishers. This total silver eel weight was converted into numbers of eels using the mean weight. The exploitation rate, escapement, and escapement rate were calculated according to:

• Exploitation (%) = silver eels total catch in number/pooled Petersen estimate

• Escapement = pooled Petersen estimate − silver eels total catch in number

• Escapement (%) = escapement/pooled Petersen estimate.

Pristine biomass estimate

The pristine biomass was estimated using the historical data of Ichkeul Lake before the decline in 1980 and considered by the scientific community as the maximum historical recruitment (ICES, 2010). Only the years between 1973 and 1979 were taken into account, since the fishing method used before that date did not reflect the real potential of the lagoon (Figure 2).

Results

Silver eel characteristics

During the two colour-marking campaigns, 985 silver eels with a mean weight of 585.3 ± 156.8 g were caught and marked. Among them, 97% were females (Table 2). Sizes ranged from 33 to 79 cm and dominant sizes were 63 and 69.5 cm (Figure 3).

Estimates of silver eel number, exploitation, and escapement rates

One hundred and thirty-three (19%) silver eels were recaptured during the first campaign (green marks) and 51 (18%) of the second one (red marks; Table 1). Furthermore, 56% of eels marked in green and 71% for those in red were recovered within the same month of their release. However, a few eels lingered and left the lake during the following months (29% in January and 5% in February for eels with green marks and 29% in February for those with red marks). This indicates that the crossing time in the lake is not the same for all eels and may vary between 1 and 3 months.

Total catch was obtained from fishers, but it included both silver and yellow eels. By removing the proportion of yellow eels, we obtained 37 578 kg silver eels captured in total corresponding to 64 209 individual. Most of the catch (51%) was made in December.

The number of migrating silver eels was estimated at 342 221 (297 956–386 486) eels by the pooled Petersen estimator, corresponding to a biomass of 200.2 (174.3–226.1) t and a density of 23.55 (20.51–26.6) kg ha⁻¹. The rate of exploitation by the commercial fishery was estimated at 18.8% (16.6–21.5%), ∼4.42 kg ha⁻¹, which corresponds to an escapement rate of 81.2% (78.5–83.4%) or 19.13 kg ha⁻¹ (Table 3).

Pristine biomass estimation

The average catch between 1973 and 1979 in Ichkeul Lake was 71.6 T. The current catch in 2014 was 39.6 T, which corresponds to 55% of the historical one. As the fishing effort has not undergone any major changes and remained relatively constant in this area since 1973, and with the current biomass of 23.6 kg ha⁻¹, this corresponds to an average biomass of 42.6 kg ha⁻¹ for the period 1973–1979.

Discussion

Silver eels characteristics

Our results show that in Ichkeul Lake, the sex ratio is largely in favour of females (over 97%) during the sampling period. These observations are in agreement with BHH (unpublished data) who has studied Ichkeul Lake's eel population during 4 years (2004–2007). According to this author, the migratory fraction is composed of 86% females in winter and the lake produces 70% females over the whole migration period. The predominance of females was also recorded in other North African and Mediterranean sites such as the Sebou wadi and the mouth of the Moulouya in Morocco (Yahyaoui, 1991) as well as in the lagoons of Porto Pino (Rossi and Cannas, 1984) and Comacchio (De Leo and Gatto, 1995) in Italy. According to Capoccioni et al. (2014), only silver females from the westernmost and central part of the Mediterranean basin would be able to reach the Sargasso Sea in time for spawning, while just a small number of silver males could, which highlights the importance of females in this area.

Estimates of silver eel number, exploitation, and escapement rates

The migration peak in Ichkeul Lake occurred in December the year of our study. According to the fishing data, 51% of the silver eel catch was recorded in that month. The migration peak in Tunisia is shifted relatively to what can be found elsewhere in the distribution area of the species. The timing of migration is related to the geographical location of the continental life phase and thus to the distance that migrating eels have to travel to get the Sargasso Sea. Indeed, silver eels mostly migrated in August in northern Norway (Bergersen and Klemetsen, 1988), in September and October in the River Imsa, southern Norway (Vøllestad et al., 1986) as well as in the Burrishoole River, Ireland (Poole et al., 1990), while it happens in November in Bages-Sigean lagoon, France (Amilhat et al., 2008). By leaving earlier, eels from northern latitudes which have more distance to cover will likely reach the spawning grounds at the same time as other subpopulations (Bruijs and Durif, 2009).

Silver eel biomass was 23.55 kg ha⁻¹, according to our estimates. This result is close to 30 kg ha⁻¹ mentioned by Amilhat et al. (2008) in Bages-Sigean lagoon, France. However, the number of migrating eels from Bages-Sigean was three times higher than that of Ichkeul (respectively, 1 120 112 and 342 221 eels). This can be explained by the difference in sex ratio between the two subpopulations, and consequently, the median weight. (Silver eels from Ichkeul are mainly females with a median weight of 0.6 kg, while Bages-Sigean is mostly composed of males with 0.1 kg.) Our estimate is also close to Rossi's (1979) in Comacchio lagoon, Italy (20 kg ha⁻¹) and higher than 4.8–6.9 kg ha⁻¹ mentioned by Acou et al. (2009) in the Oir River, France, which are also primarily composed of females and whose area is close to that of Ichkeul (Comacchio, 100 km²; Oir, 87 km²; and Ichkeul 85 km²). Indeed, the lagoons are significantly more productive than freshwater systems since growth rates are greater in habitat close to the sea (Amilhat et al., 2008; Daverat et al., 2012).

In Ichkeul Lake, silver eel yield was estimated at 4.42 kg ha⁻¹. It is higher than that reported in Imsa River (Norway) by Vøllestad and Jonsson (1988) (2.27 kg ha⁻¹).

In addition, the escapement rate from Ichkeul Lake (81.2%) is relatively high compared with the range of escapement estimates reported in the literature (23–87%) and mentioned by Amilhat et al. (2008).

Our estimates are restricted to the winter run and exclude the autumn one. According to earlier studies in this area, silver male eels begin to leave the lake in late October with a peak in November, but continue their migration until January. On the other hand, the female migration peak occurs between December and January (BHH, unpublished data). Fishers targeting females for their large size set up the net barrier at this moment. Therefore, 30% of silver eels had left Ichkeul Lake when the colour-marking experiment has begun. To rectify this issue, we have included the estimated autumn run based on its percentage to have an overview of the whole migration in the Lake. Thus, the number of migrating silver eels increased to 488 887 eels corresponding to 33.65 kg ha⁻¹. Similarly, the annual escapement rate rose to 86.9% (29.23 kg ha⁻¹), since no eel has been captured during the autumn run (Table 4).

Handling due to tagging may cause stress to the fish and, therefore, modify the eel behaviour which might delay its migration. The colour-marking method was previously tested by Amilhat et al. (2008) in a controlled environment over a 6 months period. These authors reported that neither mark visibility loss nor induced mortality was observed. Most of the eels in our study were recaptured within the same month in which they were marked and released. Incomplete reporting of marked eel by fishers could have happened, especially during the days of high catches.

Some eels may have migrated after the fishing closure, in late February or March, or even the next year. Breteler et al. (2007) reported that a few eels marked in 2004 in the Rhine River were detected in summer 2005. Likewise, Feunteun et al. (2000) indicate that among the eels candidate to emigration, 12% remains in the watershed including 3.4% which regress to the yellow stage.

In addition, a false declaration of the total catch as well as the poaching by illegal fishers, which increases during the eel migration season according to the STL fishery manager, could lead to a lower biomass, a higher exploitation rate, and therefore to a lower escapement rate.

Current vs. pristine biomass

The European Council Regulation (EC) No. 1100/2007 for the European eel stock recovery imposes as a reference point the escapement of at least 40% of the silver eel biomass relative to the best possible escapement estimation that would have existed if the stock had not had any anthropogenic influence (i.e. pristine situation). As this objective is currently difficult to achieve, the regulation proposes the use of approximations to skirt this problem, such as the historical data before the decline in 1980, the production extrapolation by habitat type, or the analogy with similar systems.

In Tunisia, as well as in many European countries, we unfortunately have no study concerning the eel stock status or its biomass estimation. However, historical catch series are available from 1962, and until 1972 only 5 T year⁻¹ were registered on average because fishing was done in an archaic way with a small number of artisanal traps. Since 1973, an Italian company has been exploiting eels in Tunisia and the production has increased (Hizem Habbechi, 2014). For this reason, only the years between 1973 and 1979 were taken into account to get an estimate of the pristine biomass in Ichkeul Lake (Figure 2).

To meet the Eel Regulation reference point of a 40% escapement of potential spawners, 17.04 kg ha⁻¹ silver eels should leave the system every year. In the light of our results, 19.13 kg ha⁻¹ corresponding to 45% (38–52%) of the pristine biomass, mostly females, escaped from Ichkeul Lake during the winter run. The escapement increases to 29.2 kg ha⁻¹ (69% of the pristine biomass) if we extrapolate to the entire migration season of 2013/2014.

Using a population dynamic model, Aalto et al. (2016) suggested that current escapement is 35% of the pristine biomass levels across the Mediterranean basin. Other studies, with different approaches for calculating the pristine biomass, show current/pristine ratios lower or equal to the reference point (Table 5). Therefore, the ratio obtained in our study appears superior to those available in the literature.

Models based on management scenarios have been developed to maintain a sustainable fishery (Bevacqua et al., 2007; Charrier et al., 2010). The current escapement of 36.2% in the Or lagoon relative to the pristine biomass would reach 40% from a 10 mm mesh, 45.4% from a minimum size capture of 23 cm and 45.8 and 50.6%, respectively, if a fishery closure was applied, respectively, during (April, May, June) and (July, August, September) (Charrier et al., 2010). In Tunisia, the law allows the use of fykenets of 10 and 15 mm mesh for eel fishing and 30 cm is the minimum catch size (JORT, 1995). In addition, the fishery in Ichkeul Lake focuses on the migration period of female eels which starts in December, allowing the escapement of 10.1 kg ha⁻¹ of silver eels that left the lake with the first flood, before the start of the fishing season. During the rest of the year, the eel is a bycatch and fishers catch other fish like mullet or sea bass. Our study has demonstrated that the practiced law in Tunisia as well as the management policy in Ichkeul Lake are effective to meet the reference point of the European regulation and preserve the species.

Nevertheless, the estimated number of escaping eels from Ichkeul Lake will not necessarily reach the Sargasso Sea because various factors can hinder their migration. Moreover, low muscle fat content (Svedang and Wickström, 1997; Belpaire et al., 2009), parasitosis by A. crassus (Palstra et al., 2007; Sjöberg et al., 2009), chemical contaminations (Palstra et al., 2006; Pierron et al., 2008; Elie and Girard, 2009, 2014; Geeraerts and Belpaire, 2010), viruses such as EVEX (van Ginneken et al., 2005) could also have a harmful impact on the eel migration. In Tunisia, some information is available on the eel health status. Anguillicola crassus was reported for the first time in 1995 (Maamouri et al., 1999). The prevalence of this parasite in Ichkeul Lake reaches 32 and 56.3% for silver males and females, respectively (Hizem Habbechi et al., 2012).

Conclusion

Our work has allowed establishing an initial assessment on the escapement rate and the number of silver eels in Ichkeul Lake. However, it is based on an estimate of the current state and cannot be extrapolated to future years. In the Loire River (France), the escapement estimated by mark–recapture was different over the years ranging from 352 000 potential spawners in 2001/2002 to 150 000 in 2008/2009 (Acou et al., 2010). Similarly, the results cannot be extrapolated to the other main eel production sites in Tunisia (Ghar El Melh lagoon and Tunis Lake) because each has its own peculiarities. Although they are all in the north of the country, and silver eels are mostly females during the fishing season, the fishing effort is not the same. Tunis Lake is subject to massive poaching since 2011, while fishing is controlled in Ghar El Melh lagoon where independent fishers renew their permits from the local authorities every year. Besides, unlike Ichkeul Lake, these two lagoons have direct access to the sea.

The Ichkeul Lake contributes significantly to the renewal of the eel stock with some 342 221 individuals of silver eels leaving the watershed during the winter run, and 488 887 for the entire migration season in 2013/2014.

Furthermore, to get a better idea on the migratory stock, researches must combine quantitative and qualitative aspects. In this line, an European Eel Quality Database started to collect recent data of contaminants and diseases over the distribution area of the eel since 2007 (Belpaire et al., 2011). In addition to the information available about muscle fat content and the contamination by A. crassus, heavy metals assessment is being analysed, and investigations concerning the other pollutants (PCB…) and viruses have to be initiated for a global idea about the Tunisian eel population capacity to undertake the seaward migration in favourable conditions.

Acknowledgements

The authors would like to thank the fishers of Ichkeul Lake to have actively participated in this study as well as Messrs. Riadh Ben Nsir and Fathi Lazzez the managers of the SLT Company for providing free eels. All our gratitude to Dr Elsa Amilhat from the Center of Education and Research on Mediterranean Environments (University of Perpignan Via Domitia, France) who explained to us the eels tagging methodology and the use of SPAS software, and to the Irstea (Bordeaux, France) staff for their material and scientific assistance. We are very grateful to the INSTM technical staff and to the voluntary students Hela Jaziri, Wafa Cherif, and Boutheina Ziadi for their help during the tagging operations. Finally, we gratefully acknowledge the editor Dr Caroline Durif and two anonymous reviewers who provided useful and constructive comments that greatly improved the quality of our manuscript. The research was financed by the PRO-EEL Project (FP7).

References

Author notes