The study of fish populations is essential for sustainable management of the ichthyic resources. This paper aims to assess the state and structure of freshwater fish populations in three Tunisian reservoirs in order to monitor and optimize fishing and stock enhancement for an effective management. The Technical Center of Aquaculture in cooperation with the Higher Institute of Fisheries and Aquaculture of Bizerte has implemented a sampling technique using fish nets inspired by the European standard CEN pr EN 14757. In this study, 3 reservoirs were prospected: Kasseb, Seliana (in March 2016) and Lahjar (in March 2017). The implementation of the sampling protocol allowed the capture of 5 species of freshwater fish. The yields observed are more important in terms of weight than number at the Lahjar reservoir (637.5 in d/1000m2 and 73008.3 g/1000m2) average at Seliana (240 in d/1000m2 and 45191.67 g/1000m2) and quite low at the Kasseb reservoir (25.49 in d/1000m2 and 6750 g/1000m2). The majority of the catches were obtained in the 0-3 and 3-6m strata. An alarming decline of the pike-perch stock was detected in these dams with the absence of carp and eel in all the reservoirs surveyed. It is worth mentioning that overexploitation of fish resources at the level of the explored reservoirs was also observed.
Keywords: Freshwater fish, population assessment, multi-mesh gillnets, reservoirs, management plan, Tunisia
NPUE, Number per unit effort; WPUE, Weight per unit effort; (D), Index of margalef; H, Shannon and weaver diversity index; J, Index population distribution of pielou; In d, individual; ISPAB, Higher institute of fisheries and aquaculture of bizerte; TCA, Technical centre of aquaculture; IRESA, Institution of agricultural research and higher education
Aquaculture in Tunisia is a very old activity dating back to Roman times as evidenced by some mosaics of the Bardo Museum in Tunis. The first Tunisian experience of fishing in reservoirs initiated in the 1960s by the national fisheries office through the stocking of some dams with fry of various fish species.1 Currently, 9 governorates are concerned by this fishing activity and the majority of reservoirs are located in the northern part of the country.2 The fisheries in these reservoirs involve a large number of fishermen’s and it was considered as an important economic activity, especially in the disadvantaged region of the country.1 The most commonly caught species in Tunisian reservoirs are Mullet (Mugilcephalus Linnaeus, 1758, Mugiliformes, Mugilidaeand Liza ramadaRisso, 1827, Mugiliformes, Mugilidae), Pike-perch (Sander lucioperca, Linnaeus, 1758, Perciformes, Percidae), Carp (CyprinuscarpioLinnaeus, 1758, Cypriniformes, Cyprinidae), Rudd (Scardiniuserythrophthalmus Linnaeus, 1758, Cypriniformes, Cyprinidae), Catfish (SilurusglanisLinnaeus, 1758, Siluriformes, Siluridae),Eel(Anguilla anguillaLinnaeus, 1758, Anguilliformes, Anguillidae), Tilapia (OreochromisniloticusLinnaeus, 1758, Perciformes, Cichlidae), Roach (RutilusrutilusLinnaeus, 1758, Cypriniformes, Cyprinidae) and Barbel (BarbuscallensisLinnaeus, 1758, Cypriniformes, Cyprinidae).1,3,4 Production increased from 843.5 tonnes in 2000 to exceed 1359.821 tonnes in 2017 with a decrease to 919 tonnes in 2011 and 874.407 tonnes in 2019.5 The decline of 2011 is the result of the unstable political situation of the country and the one, which occurred in 2019, can be explained by the low rate of carp and Mullet stocked for the last years in the Tunisian reservoirs. For 2019, the Carp was the most heavily fished species, with 30.28% of the total production followed by Mullet and Pike-perch with, respectively, 28.19% and 22.45%.5 This activity can be maintained and developed by monitoring the fish stocks and the status of freshwater population’s fishes in the reservoirs. Given the lack of data related to these fisheries and the necessity of management, a research and development project was initiated in collaboration between the Technical Center of Aquaculture in Tunis (CTA) and the Higher Institute of Fisheries and Aquaculture in Bizerte (ISPAB) in order to develop an adequate management plan. The goal of this study was to determinate the specific richness, spatial distribution and abundance of freshwater species. We also seek to provide details on the status of freshwater fish exploited in three reservoirs: Lahjar, Kasseb and Seliana located in different areas in Tunisia.
The study was conducted in three Tunisian reservoirs: Lahjar, Kasseb and Seliana (Figure 1) *Lahjar reservoir: (36°504’18’’N, 11°01’14’’E): It is located in the northeast of Tunisia6 near the city of Kelibia with a maximum depth ranging from 13 to 14m and surface area of 254 ha. Lahjar reservoir is used exclusively for irrigation of a large surrounding area.7 *Kasseb reservoir: (36°45’32”N, 9°0’20”E): It was built in 1968 and it belongs to the north hydrological watershed of Madjerda, situated at some kilometers (20Km) of Beja city. The reservoir has a surface area of 430ha and contains 70Mm3 of water. It is used mainly for drinking water supply of Tunis city with 40Mm3/year.7 *Seliana reservoir: (36°09’26’’N, 9°20’55’’E): It was built in 1978 and it has a surface area of 600ha. It is situated near the North of Seliana city and it belongs to the south hydrological watershed of Medjerda. Seliana reservoir is used for irrigation at the perimeter of Gaafour/Laaroussia city with 23Mm3/year.7
Fish sampling operations were carried out with multi-mesh monofilament gillnets inspired from the European standard CEN pr EN 14757.8 These nets, adapted to Tunisian reservoirs, were used in Tunisia since 2013. There are formed by 8 different meshes (18, 24, 28, 35, 40, 55, 70 and 80 mm, knot-to-knot). Each net has 20 m length and 1.5 m (benthic nets) and 6 m (pelagic nets) depth. A stratified random sampling was used to take into account the irregular spatial distribution of fish in reservoirs. Pelagic nets were put by pairs in the deepest part at depths multiple of 6 (0-6 m and 6-12 m, etc.).3,9–11 A total of 11, 48 and 20 benthic nets were implemented, respectively, for Lahjar, Kasseb and Seliana reservoirs. Additionally, one and 6 pelagic nets were implemented, respectively, in Lahjar and kasseb reservoir; for Seliana reservoir the low depth (<6m) makes the use of pelagic nets useless (Table 1). Nets were set in late afternoon and left for approximately 12 hours, according to the European Standard and they were removed the following morning. For Kasseb and Seliana reservoirs, sampling was conducted during spring (March 2016) and for Lahjar in April 2017. Captured fishes were identified, referred to the datasheets of kraiem12 counted by species, measured to the nearest millimeter total lengths and weighted to the nearest gram wet mass. Catches (expressed as Number Per Unit Effort "NPUE") and biomass (Weight Per Unit Effort "WPUE") were calculated for each reservoir. NPUE and WPUE for gill nets were expressed per 1000 m2 net.
Reservoirs |
Sampling period |
Benthic nets per strata |
Total Benthic nets |
Total Pelagic nets |
||||
0 – 2.9 |
3 – 5.9 |
6 – 11.9 |
12 – 19.9 |
19 – 34.9 |
||||
(m) |
(m) |
(m) |
(m) |
(m) |
||||
Lahjar |
Apr-17 |
5 |
5 |
1 |
0 |
0 |
11 |
1 |
Kasseb |
Mar-16 |
10 |
10 |
10 |
10 |
8 |
48 |
6 |
Seliana |
Mar-16 |
10 |
10 |
0 |
0 |
0 |
20 |
0 |
Table 1 Number of benthic and pelagic nets used in Lahjar, Kasseb and Seliana reservoirs
The diversity, of the freshwater fish population in the studied reservoirs, was expressed by specific richness determined by the index of Margalef (d)13 and calculated according to the formula:
The Shannon and Weaver diversity index (H’)43 and the index population distribution of Pielou (J’) were calculated according to the formulas of Barbault1:
Specific composition
The implementation of the standardized sampling protocol allowed us to evaluate the species richness. The latter is variable at the level of the studied reservoirs. The most important species richness was recorded in Lahjar and Seliana dams with the presence of 4 different species for each, i.e. 40% of the global richness; followed by the Kasseb reservoir where, only, two species were reported representing 20% of the global specific richness (Figure 2). A total of 5 species were captured in this investigated dams; Mullet (Liza ramada), Rudd (Scardinuserythrophtalmus), Roach (Rutilusrubilio), Pike-perch and Barbel (Barbuscallensis) (Table 2). In this study, the Margale index (d) varied from 3.84 at Lahjar to 3.82 at Seliana and to 1.69 at Kasseb reservoir. The Shannon index shows values varying between 1.35 in Seliana, 0.29 in Lahjar and 0.24 in Kasseb dam. The application of the Shannon index to the collected data showed that the fish populations in the different studied reservoirs are not very diversified: H' is often lower than 1.5.
Reservoirs |
Roach |
Pike-perch |
Mullet |
Rudd |
Barbel |
Rutilusrubilio |
Sander lucioperca |
Liza ramada |
Scardinuserythrophtalmus |
Barbuscallensis |
|
Lahjar |
× |
× |
× |
× |
|
Kasseb |
× |
× |
|||
Seliana |
× |
× |
× |
× |
Table 2 Captured species in the investigated reservoirs
Numerical and weighted yields
The highest yields were recorded in Lahjar dam with 637.5 ind/1000m2 and 73008 g/1000m2, representing a population of 612 in d and a biomass of 70088g, followed by Seliana dam which showed an average yields: 45191g/1000m2 and 240 in d/1000m2. The lowest numerical and weight yields were recorded at Kasseb dam with 25.49 in d/1000m2 and 6750 g/1000m2 (Table 3).
Reservoirs |
Species |
Captures |
Yields |
|||||||
Numerical (ind) |
Numerical (%) |
Biomass |
Biomass |
Numerical (ind/1000m2) |
Weighted (g/1000m2) |
|||||
(g) |
(%) |
|||||||||
Lahjar |
Roach |
588 |
96.08 |
57735 |
82.38 |
612. 50 |
60140.62 |
|||
Mullet |
16 |
2.61 |
9084 |
12.96 |
16.67 |
9462.5 |
||||
Rudd |
3 |
0.49 |
349 |
0.5 |
3.13 |
363.54 |
||||
Pike-perch |
5 |
0.82 |
2920 |
4.17 |
5.21 |
3041.66 |
||||
Total |
612 |
100 |
70088 |
100 |
637.5 |
73008.3 |
||||
Kasseb |
Barbel |
25 |
96.15 |
6660 |
96.73 |
24.51 |
6529.41 |
|||
Mullet |
1 |
3.85 |
225 |
3.27 |
0.98 |
220.59 |
||||
Total |
26 |
100 |
6885 |
100 |
25.49 |
6750 |
||||
Seliana |
Mullet |
85 |
59.03 |
20165 |
74.37 |
141.67 |
33608.33 |
|||
Barbel |
2 |
1.39 |
100 |
0.37 |
3.33 |
166.67 |
||||
Rudd |
45 |
31.25 |
4775 |
17.61 |
75 |
7958.33 |
||||
Pike-perch |
12 |
8.33 |
2075 |
7.65 |
20 |
3458.33 |
||||
Total |
144 |
100 |
27115 |
100 |
240 |
45191.67 |
Table 3 Numerical and weighted abundances of ichthyic species in Lahjar, Kasseb and Seliana reservoirs
In Lahjar Dam, 612 individuals belonging to 4 different species, were captured in 11 benthic nets (total area 720m²) and one in the pelagic net (area 240m²). The species captured are dominated by the Roachwith 505 specimens sampled in the benthic nets, against 83 for the pelagic net. This species shows a numerical percentage of 96.08% followed by Mullets with 2.61% (16 individuals). The numerical yields are respectively 612.5 in d/1000m2 for Roach and 16.67 in d/1000m2 for Mullet. Pike-perch and Rudd show very low numbers of individuals respectively (5 and 3 individuals) with respective percentages of 0.82% and 0.49%. Their numerical yields are as follows: 6.94ind/1000m2 for Pike-perch and 4.17ind/1000m2 for Rudd (Table 3). Regarding the weighted yields, the lowest values are observed for Rudd (0.48kg/1000m2) and Pike-perch (4.06 kg/1000m2). Mullet and Roach have higher yields: they are respectively around 12.62 kg/1000m2 and 69.04 kg/1000m2 In Kasseb reservoir the catch yields vary between 0.98ind/1000m2 for Mullet and 24.51 in d/1000m2 for Barbel. The weight yields recorded for Mullet and Barbel are about 220.59 g/1000m2 and 6529.41g/1000m2 for a biomass of 6660g and 225g, respectively (Table 3). In Seliana reservoir, a total of 144 specimens were captured which is corresponding to a numerical yield of 240 in d/1000m2 and a weight yield of 45191.67kg/1000m2. The weight and number catches were dominated by Mullet followed by a forage fish (Rudd). The yields recorded are 33608.33g/1000m2, 141.67 in d/1000m2 for Mullet and 7958.33g/1000m2; 75 in d/1000m2 for Rudd. Barbel and Pike-perch showed low numbers, respectively 2 and 12 individuals. Their numerical yields oscillate respectively, between 3.33 in d/1000m2 and 20 in d/1000m2 and their weight yields were 166.67g/1000m2 for Barbel against 3458.33g/1000m2 for Pike-perch(Table 3).
Spatial distribution of catches
The majority of the catches, collected in Lahjar reservoir, were obtained at the level of the 0-3m strata. These catches represent 69.4% of the total number of fish caught, followed by the stratum 3-6m (30.6%). In this reservoir, Roach dominate in all depth strata. Pike-perch and Rudd are found only in the 3-6m stratum. Mullet occupy the 0-3m stratum. Pelagic nets caught only Roach in the 6-12 m stratum (Figure 3). In Kasseb reservoir, Barbel was captured by benthic nets, as well as by pelagic nets and in all strata, mainly that of 0-3m. The only Mullet specimen was caught in the 0-3m stratum (Figure 4). The species populating Seliana reservoir colonize the two sampled strata (0-3m and 3-6m) with a quantitative predominance in the superficial stratum (0-3m). Rudd and Mullet are present in the whole reservoir, while the Pike-perch and the specimens of the Barbel are present only in the 0-3m stratum (Figure 5).
Population structure Roach
The size of fish caught only in Lahjar reservoir ranged between 15 and 27cm for Roach. Besides the size frequencies histogram shows the presence of two age classes (Figure 6).
Mullet
The size classes of the captured specimens of Mullet in Lahjar reservoir is ranged from 35 to 42cm, the length frequency distribution show the presence of two age classes. For Seliana reservoir, the size structures of the catches of Mullet are mainly composed by three age classes. The average size is about 28.98±84cm with an average weight of 237.24±33.54g. The Kasseb reservoir shows a deficit of Mullet (Figure 7).
Rudd
The three specimens of Rudd captured in Lahjar reservoir show an average size of 21.33cm with an average weight of 116,33g; but for the Kasseb reservoir, the average length of the catches is about 20.6±2cm for an average weight equal to 106.1±21g. The majority of the catches in Lahjar and Kasseb reservoirs belong to adult sizes. The fry and juvenile stages are not included in our catches because the minimum mesh length used for fishing is 18mm (Figure 8).
Pike-perch
The size classes of the catches of pike-perch in Lahjar reservoir is ranged between 39.7 and 43.8cm. In Kasseb reservoir, the size classes of this latter are ranged from 18 to 22cm with an average length of 23.5±9.67cm and a weight of 172.92±364g. Two age classes were caught in the two reservoirs (Figure 9).
Barbel
Barbel is an endemic species, which is quite widespread in Kasseb reservoir. This species presents class sizes varying between 26 and 33.5cm. We identify two age classes at the level of the captured individuals (Figure 10).
In this study, species richness and fishing yields are low, as is the case for most of the reservoirs in Tunisia due to the recent introduction of species and the annual seeding of Mullet fry. In addition, the spatial distribution and survival of fish seems to be strongly influenced by oxygen content and temperature. A similar study was carried out by Holubová14 at the Rímovdam (Czech Republic) and showed a positive correlation between temperature and the density of freshwater fish present in the epipelagic stratum. These authors noted that the increase in temperature in late spring increases the rate of encounters with predators, which leads to the probability of group constitution. Other research, such as Society15 and Weetman,16 confirm that high temperature stimulates fish activity. Similarly, the small changes in temperature can lead to changes in distribution.17 Furthermore, in Lake Balaton in Hungary, György18 showed that the variability of catch per unit effort is mainly related to the degree of water transparency which can influence the efficiency of the nets. On the other hand, Macy19 and Figuerd.20 noted that fish activity increases with light intensity, are very elevated in high light but low in darkness as well as with water transparency, the fish is more likely to be entangled in the net in turbid water than in clear water. Previous studies showed that prey species escape periods of high light and seek protection in deep layers or shaded objects.21–23 In contrary, Kerry and Bellwood24 showed that this behavior is related to light avoidance phenomenon more than predators. According to the literature, the percentage of piscivorous fish is considered as an indicator of water quality.25 Therefore, the potential piscivorous biomass, including all size classes, shows a total imbalance in all sampled reservoirs. In our study the fish communities are composed mainly by cyprinids, (588 Roach) in Lahjar reservoir with a percentage of 75% of the total catches within the 3 dams, as is the case for most lakes in France6. In Tunisia, Sidi Salem reservoir proves this statement; this water reservoir is characterized by a high abundance of Roach and Rudd. These two species, with a phytophilic character, seem to have a good reproductive success in Tunisian reservoirs.10
Recent studies revealed that fish populations are influenced mainly by the depth of the dam as well as the amount of chlorophyll "a" it contains.26,27 The majority of cyprinids (Freshwater bream, Roach and bleak) colonize the epipelagic part of the reservoir which represents a rich feeding habitat especially for the forage species.28 This result is in agreement with our results: the majority of Roach as well as Rudd captured, respectively, at Lahjar and Seliana reservoirs were recorded at the upper strata 0-3m and 3-6m. The same result was mentioned by Mili10 following a study in Ghezela dam. Barbel occupies mainly the stratum 0-3m in Kasseb, but it is also foundin deeper strata (6-12m and 12-20m). These results are comparable to those of Vašek29 who found that the majority of fishes are located in the superficial layers. The work of Djemali30 and Mili10 indicate that going from the dike to upstream, the distribution of fish biomass is influenced by water depth. These authors state that the highest concentrations were recorded at the surface layers of the deepest areas of the reservoir. Hydro acoustic studies conducted in Tunisian dams showed that fish densities vary considerably between the surface and the bottom due to the mobility of ichthyic resources and the lack of predictable behavior in these animals.30 ,31 The study of Vašek31 which focused on the analysis of fish yields and distribution in European reservoirs, showed that pike-perch harbor the lower epilimnion and upper metalimnion with nyctimeral migrations from the pelagic layer to the coastal areas.
Inter-organism interactions are major factors affecting the composition and location of fishes at the trophic level, particularly predator-prey interactions, that can alter the spatial distribution of fishes.26,33–35 This is concordance with the results found by Mili36 stating that the decline of Barbel stocks in Tunisian reservoirs is due to the introduction of carnivorous predatory species. The present study shows that the presence of Pike-perchis linked to that of Roach for Lahjar and Rudd for Seliana reservoirs. The low numbers sampled during the month of March within these two reservoirs, is probably related to the behavior of this species during the breeding season. This lack of Pike-perchand Mulletin Lahjar and Kasseb reservoirs can be explained by the fact that these species are strongly targeted by fishermen in these dams. The study conducted by Mili10 in Sidi Salem reservoir (Tunisia) showed that the Mullets seem to be overexploited in this dam. Furthermore, referring to the study of Winfield,37 the analysis of size structures of fish caught by multi-mesh nets showed that these gears catch more medium-sized individuals compared to other sampling methods. In similar studies, larger individuals were caught near the dam, but smaller individuals and juveniles were concentrated near the shore. These areas have better trophic conditions but may shelter carnivores come to lay eggs.38,29 Additionally, Carol35 found that freshwater ecosystem productivity and nutrient concentrations increase from the dam to the river.
In this study, all Pike-perch captured were adults since sexual maturity of this species is reached at the age of one year for a size of 24 cm and a weight of 90g.39 Roach caught in Lahjar have sizes varying between 15 and 27cm and belonging to adult specimens. Indeed, the sexual maturity of this species is reached at the age of one year, which corresponds to a size between 11.92cm for males and 12.04cm for females.29 The majority of specimens of Rudd caught are adult given that the size at first sexual maturity for males is about 11.99cm and around 12.48cm for females.40 Barbel size frequencies in Kasseb reservoir, illustrates the presence of adult individuals. Sexual maturity of this species is reached for males at 19.6cm and for females at 27.7cm.41 Deceliere-Vergès and Guillard42 indicate that the abundance of species in the pelagic stratum would be better assessed by hydro acoustic surveys than nets because sampling biases can be due to the passive nature of the multi-mesh nets and the escapement of small individuals.
Reservoirs are essential for the sustainable management of aquatic resources and for fisheries production in areas far from the coast in Tunisia.43 In our study, the application of a sampling by multi-mesh nets inspired by the European standard CEN14575 and adapted to Tunisian dams has allowed the capture of 5 species of fish with important values of numerical and weight yield at Lahjar reservoir against average values recorded in Seliana demand very low ones at Kasseb. Therefore, a rapid intervention is necessary for the preservation of the specific richness of the ichthyofauna by a reinforcement of the Pike-perch and Mullet population through seeding the dams with fry.
We are most thankful for the Institution of Agricultural Research and Higher Education (IRESA) that has provided financial support for the research project with impact (AMBISEPT) from which this paper emerged. This work is part of a collaborative project involving the TCA and ISPAB, Tunisia. We would like to thank TCA technical staff and students from ISPA Bizerte for their efforts on the practical part of this work.
The author declares no conflict of interest.
© . This is an open access article distributed under the terms of the, which permits unrestricted use, distribution, and build upon your work non-commercially.