Observations on the biology of Nile tilapia , Oreochromis niloticus L . , in Tekeze Reservoir , Northern Ethiopia

The study was done to examine some aspects of the reproductive biology of Nile tilapia (Oreochromis niloticus) in Tekeze Reservoir a newly man made hydropower reservoir. A total number of 1826 specimens of Nile Tilapia (O. niloticus) were collected from the reservoir from July 2015 to June 2016. Size at first maturity, sex ratio, gonado-somatic index (GSI), breeding season and fecundity were studied. The overall sex-ratio female: Male (F: M) was 1.6:1 which is deviated from the expected 1:1 sex ratio (χ2 = 10.13; p<0.05). Length at first maturity was 14 and 15 cm for females and males respectively. The breeding period of females extended from January to September. Two annual breeding seasons were noticed. A minor breeding season which extended from January to March and a major breeding season from July to September and monthly Condition Factor (CF) (mean ± SD) value of O. niloticus in the reservoir ranged from 1.73±0.03 in July to 2.05±0.02 in November for males and 1.61±0.02 in July to 1.99±0.05 in April for females. Fecundity ranged between 399 to 2129 g. Fecundity correlated well with total weight (r = 0.086), total length (r = 0.77) and ovary weight (r = 0.72) than with gonadal weight (r = 0.86). Nile tilapia at Tekeze Reservoir showed isometric growth pattern (b=2.92), indicating that the reservoir was favorable for the fish growth. For proper management of the fish species in the reservoir, it is better for the fishermen not to capture the fish during the breeding season of the year in the reservoir.


INTRODUCTION
Nile tilapia (Oreochromis niloticus) is an important fish in the ecology of tropical and sub-tropical region including Ethiopia and of great commercial importance in the fisheries in many African lakes (Britton and Harper, 2008).It is also the most popular species of the bony fish for aquaculture in Africa (Abdel et al., 2007).This is attributed to many positive qualities including tolerance to poor water quality, wide range of food, and plasticity in growth, firm flesh and good taste (Fryer and Iles, 1972).Other advantages are its herbivorous nature and its mouthbrooding habits (Pena-Mendoza et al., 2005); extended breeding seasons and their short generation *Corresponding author.E-mail: tsegayteam331@yahoo.com.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License time (Ibrahim et al., 2008).Tsegay et al. (2016a) reported that, O. niloticus is the most dominant and commercially important species in the newly formed Tekeze Reservoir, and accounts about 82.4% of the total production of fish in the reservoir.Nile tilapia is the most popular fishes in Ethiopia as well as around the reservoir due to its value as a commercial and subsistence fishes for most of the inhabitants living around the reservoir.This is because it has fewer bones in its flesh compared to the Barbus species.
Descriptions of reproductive strategies and the assessment of fecundity are fundamental topics in the study of the biology and population dynamics of fish species and also for evaluation of the reproductive potential of individual fish species.This will increase our knowledge about the state of a stock and improves standard assessments of many commercially valuable fish species (Murua et al., 2003).Moreover, the availability of data based on reproductive parameters and environmental variation leads to a better understanding of observed fluctuations in reproductive output and enhances our ability to estimate recruitment (Kraus et al., 2002).
Previous work carried out on O. niloticus in Tekeze Reservoir focused mainly on proximate composition and mineral content, and types of diet of O. niloticus (Tsegay et al., 2016b).This investigation provides information on sex ratio, maturity stages, gonado somatic index, breeding season and fecundity of O. niloticus obtained from the Tekeze Reservoir to form a base-line data which can be used as fish species for sustainable utilization and better management of Tekeze Reservoir fisheries.

Study area
The study was conducted at Tekeze reservoir.The reservoir is constructed for hydropower generation which was built in 2009 over Tekeze River, the major river in Ethiopia and is a tributary of Nile River.Tekeze Reservoir is located at the border of Amhara and Tigray regions.The reservoir has a catchment area of about 30,390 km 2 while a maximum total water storage capacity is about 9.293 billion m 3 .Tekeze reservoir has maximum length of 75 km and maximum width of 6 km, and covering an area of about 16,040 hectares with mean depth of 58 m.The reservoir extends from 13°20'N and 38°44' and 12°49'N and 38°41'E with an altitude of 3,750 ft.The catchment area is characterized by annual dry and rainy seasons.The rainy seasons extends from June to August while the dry season extends from September to May.The substratum of the reservoir is mainly rocky.Shoreline is without vegetation.The rock which is submerged is eventually decomposed by weather change and there is formation of sand in the reservoir.

Water quality parameters
Physico-chemical parameters of the reservoir water were tested according to standard methods described by American Public Health Association (APHA) (1995,2005).The physical parameters, including color, odor and transparency were measured.The pH and Teame et al. 87 water temperature were detected using a waterproof digital pH meter and thermometer (CP-411, ELMETRON).Dissolved oxygen concentrations and conductivity were measured using a waterproof digital, dissolved oxygen meter (conductivity/oxygen meter CCO-401, ELMETRON).Transparency was measured using secchi disc attached to graduated plastic rope.

Fish sampling and measurements
A total of 1826 individuals (981 female and 845 male) specimens were caught during the study.Fish samples were collected on monthly basis from July 2015 to June 2016 at three sampling stations.Fish were captured by gill nets (stretched mesh sizes of 6, 8, and 10, 12, and 14 cm).Soon after collection, total length (TL) and total weight (TW) were measured for each fish using measuring board and digital sensitive balance, to the nearest centimeter and gram, respectively.Each fish was dissected and the sex of the fish was identified through macroscopic examination of gonad stages using keys (Holden and Raitt, 1974;Babiker and Ibrahim, 1979).
After dissection, the sex and maturity stage of each fish were determined following standard methods (Siddiqui, 1997;Babiker and Ibrahim, 1979).The maturity level of each gonad was classified into six groups and graded as immature: (1) Recovering spent (2) Developing virgin (3) Ripening (4) Ripen and (5) Spent Samples of ovaries in maturity stage IV were removed and preserved in a labeled plastic can containing Gilson's fluid for fecundity estimation (Begeanal, 1978).The preserved ripe gonads were taken to Abergelle Agricultural Research Center Laboratory, for further investigation.The percentage of male and female O. niloticus having gonad stages III, IV and V in different length groups were plotted against length for each sex.The length at first maturity, L50, was determined from the relationship between the percentages of mature fish at different size classes using the logistic function (SPSS Program).

Length-weight relationship and condition factor
Length-weight relationship of O. niloticus was calculated using least squares regression analysis as described in Bagenal and Tesch (1978) as follows: TW=a TLb Where, TW= total weight in grams, TL= total length in centimeters, a and b= are intercept and slope of the regression line, respectively.
The well-being or condition factor of each fish was determined by computing Fulton condition factor as described by Bageanal and Tesch (1978).The condition factor of individual fish was calculated and then monthly mean values were determined for each sex separately.Condition factor of each fish was calculated as:

The Gonado Somatic index (GSI)
The Gonado Somatic index (GSI) used as an indicator parameter for reproduction; was calculated for each specimen as the percentage of gonad weight to that of the fish total weight (Khallaf and Authman, 2010) using the following equation: Where: GW is the weight of gonads (ovary and testes) in gram and TW is total weight of fish in gram.

Statistical analysis methods
First data were tested for its homogeneity of variances using Levene's test.A chi-square test was applied to decide if the sexratio of captured fish were varied between months (Sokal and Rohlf, 1981).A two tailed t test was used to compare the estimates of L50, fecundity, condition factor, and Gonado Somatic Index (GSI).Analysis of variance (ANOVA) was also used to define variations between the monthly GSI and condition factor (CF) values.To test relationship between total length and total weight of the fish, fecundity and total length; fecundity and total weight linear regression method was applied.All statistical analyses were performed using statistical package for social science (SPSS) Version 18 and significance of differences was judged at p < 0.05.

Physico-chemical parameters of the reservoir
The mean depth of the reservoir was 58 m, and the maximum depth of 183 m.Data on water temperature, dissolved oxygen, pH, conductivity, maximum depth and chlorophyll a are presented in Table 1.During the present study, the area was characterized by two seasons: a dry season (October to May) and a rainy season (June to September).The reservoir water level increased from June to September as a result of heavy rains.During the study, the mean water temperature of Tekeze reservoir was 26.9°C, and the mean dissolved oxygen and pH were 5.82±1.74g/Land 8.01±0.88,respectively.The mean chlorophyll a content was 81±5.25µg/Lwhereas ammonia and alkalinity were 0.39±0.11mg/L and 138.02±6.55mg/L,respectively (Table 1).

Sex ratio and length at maturity
A total of 1826 O. niloticus were studied of which 845 (46.28 %) were males and 981 (53.72%) were females.The overall sex ratio was 1:1.16 (males: females), with significant deviations from one on chi-square analysis (χ2 = 10.13;p<0.05) (Table 2).The fish specimens ranged from 6 to 37 cm in total length and the total weight ranged from 5 and 795 g.During the study, higher proportion of the sampled fishes for both sexes ranged in size between 20 and 34 cm.The peak was between 22 and 30 cm for both sexes.This length group alone was about 36% for females and 29% for males.Fish over 34 cm and below 10 cm TL were least represented in the sample.
Length at first maturity (L 50 ) has a great importance in the determination of optimum mesh size (Mehanna, 2007).The smallest sexually mature male was 14 cm TL whereas the same for female was 12.5 cm TL.Based on graphical methods, male O. niloticus reached 50% sexual maturity (L 50% ) at 15 cm TL and females at 14 cm TL.Although the sizes of L 50% were not significantly different from each other (p > 0.05), females appeared to reach sexual maturity at a relatively smaller size than males.The smallest length recorded in the catch of sample was 11.0 cm, which is slightly less than L 50 .This means that O. niloticus in this reservoir was slightly exploited specially at spawning period, so it is recommended to increase the mesh size used to catch fish for lengths greater than 14.0 cm.Humason (1972) pointed out that length at maturity was 11.7 cm (males) and 12.0 cm (females) for O. niloticus in Coatetelco Lake, Mexico.
While, Pena-Mendoza et al. (2005) found that these lengths for O. niloticus at Emiliano Zapata dam, Morelos, Mexico were 15.1 and 15.2 cm for females and males, respectively.The size at first maturity in the Nile at Khartoum for males and females Tilapia nilotica is 11.4 and 14.3 cm, respectively (Babiker and Ibrahim, 1979).While, Morales (1991) mentioned that the tilapias attain their sexual maturity at three months old with a total length of 8 to 16 cm.It was cited (de Graaf et al., 1999) that the first maturation size for reared Nile tilapia is 30 to 50 g.These differences arise because the sexual maturity is a function of the size and may be influenced by the abundance and seasonal availability of food, temperature, photoperiod and other environmental factors and different localities.Siddiqu et al. (1997) reported that fish size at maturity was influenced by the feeding level, which affects the growth (Table 2).

Distribution of matured stages
Monthly variations of the gonadal developmental stages for female O. niloticus from Tekeze reservoir are presented in Figure 1.All the different stages appeared in all months.The immature stages (I and II) represented the dominant maturity stages throughout the year with a peak in June, November, April and May.Mature stage (III), spawning stage (IV) and spent stage (V) of O. niloticus reached maximum percentages in July to September.The active reproductive period could be expected by the sum of the percentage of stages III and IV.Active reproductive period was recorded during January to February in minor peak and July to September with a major peak for both sexes.The number of each gonadal development stage in females is illustrated in Figure 2. According to gonadic maturation stages (Figure 2), 38.370% of the total fishes were maturing (II), 42.33% were mature (III) and 18.85% were ripe (IV).Therefore, 61.18% of the fishes were in the reproductive process.The highest proportion of average mature (III) and ripe (IV) gonadal stage in females was found in July (Figure 2).

Breeding season
Nile tilapia individuals with ripe gonads were caught throughout the year.However, the frequency of ripe male and female varied considerably between months (Figure 3).The main reproductive period for females, was July, August and September which was followed by a reproductively quiescent period between January and February.In Tekeze Reservoir, O. niloticus spawned all year round.However, the main breeding season occurred between July and September followed by a minor one between January and March.Similar studies conducted on O. niloticus in the rift valley Lakes Ziway and Hawassa (Zenebe, 1998;Demeke, 1996) also revealed bi-modal breeding patters.Njiru et al. (2006) reported that most tilapia species breed continuously throughout the year with increased breeding during periods of intense sunshine or rainfall.Several studies indicated that the peak breeding season of O. niloticus could be triggered by increase in temperature, solar radiation or rainy season and rise in water level (Trewavas, 1983;Stewart, 1988).Thus the main breeding season of O. niloticus in Tekeze Reservoir might be triggered by the main rainy season which increases the availability of food (plankton) and the minor breeding period of the fish in the reservoir could be associated with the high temperature and solar radiation (Figure 3).

Condition factor (CF)
The mean monthly CF value of O. niloticus significantly varied between months (p<0.05).However, no significant difference was observed between sexes (p>0.05).Monthly CF (mean ± SD) values ranged between 1.73±0.03 in July and 2.05±0.02 in November for males and from 1.61±0.02July to 1.99±0.05 in April for females.Lowest monthly CF was recorded in July in both sexes (Table 3).The average CF values of O. niloticus for male and female were 1.91 and 1.82, respectively.Similar results were reported by Gashaw Zenebe (2008) in Lake Koka (1.87), Lake Ziway (1.81) and Lake Langano   , 2002).The difference in CF value in different lakes might be attributed to the differences in food availability between water bodies.There was no significant difference in CF between females and males in Tekeze Reservoir (p>0.05).However, there was significant difference in CF values between months (p<0.05)indicating differences in the condition of the fish with seasons.Relatively, the lowest value was recorded in July and coincided with the main spawning period of the fish (Table 3).

Lengthweight relationship
The relation between TL and TW of combined sexes were curvilinear.The slope of the equation calculated for both sexes was very close to the theoretical value of b=3, that was 2.92 (for male b=3.03 and for female b=2.91) (Figure 3).The poor condition of O. niloticus in Tekeze reservoir in July might be attributed to the stress resulting from intense breeding activities which cost high energy for the fish.The length-weight relationship of female and male of O. niloticus in Tekeze reservoir was described best by the following regression equations: Male: TW=0.0253TLThis has also been confirmed in earlier studies from Lake Ziway, Tana and Turkana (Stewart, 1988;Zenebe, 1988).
Other environmental factors such as availability of food and food quality, feeding rate, degree of parasitism (Getachew, 1987;Teshima et al., 1987;stewart, 1988) and pollution have been reported to affect the condition factors of O. niloticus (Bakhoum, 1994;Khallaf et al., 2003).Maternal mouth brooders like O. niloticus fast or take less food during the early stages and probably throughout the brooding period (Fryer and Iles, 1972) (Figure 1).

Gonado Somatic index (GSI)
Monthly variations in GSI of O. niloticus revealed that both sexes followed nearly the same trend; however, males had higher mean values in July whereas females have higher value in August (Figure 4).In females, two peaks of GSI values were observed during February and August (Figure 4).This means that females could breed more than once in a year.

Fecundity
Fecundity was estimated for 30 females ranging from 14 to 37 and 78.8 g to 711 g TW.The total number of ripe eggs ranged between 399 to 2129.The fecundity ranged between 104 to 709 eggs corresponding to fish length of 12.5 to 20.9 cm with an egg diameter of 1.0 to 3.0 mm (Gomez-Márquez et al., 2003).Pena -Mendoza et al. (2005) found that fecundity ranged from 243 to 847 eggs per fish (O.niloticus) in Zapata dam, Mexico with egg diameter from 3.00 to 3.70 mm.Within a given species, fecundity may vary as a result of different adaptations to environmental habitats (Witthames et al., 1995).Even within a stock, fecundity is known to vary annually, undergo long-term changes (Kjesbu et al., 1989) and has been shown to be proportional to fish size (and hence, age) and condition.Thus, fish size and condition are key parameters to properly assess fecundity at the population level (Murua et al., 2003).In addition, the variation in fecundity may be attributed to differential abundance of food within the members of the population.Siddiqui et al. (1997) pointed out that fecundity increased with increased feeding levels.Moyle and Cech (2000) mentioned that in mouth brooding cichlids, the fecundity is considerably low because the parents assure the survival of the offspring, Fecundity was positively correlated with total weight (r=0.86),total length (r=0.77) and ovary weight (r=0.72)(Figure 5).

Conclusion
O. niloticus breed throughout the year however; August and February are the major breeding seasons.The length -weight relationship of female and male of O. niloticus in Tekeze Reservoir shows isometric growth pattern, indicating the well-being of the fish.However, the condition of the fish was found to be vary between months but not between sexes.Fecundity was more correlated with total weight than ovary weight.The physico-chemical parameters of Tekeze reservoir are generally optimal for the survival, growth and reproduction of O. niloticus.The fishing activities occurred during the spawning period the major peak of breeding (July to September) should be regulated since that is the period of reproductive activities for Nile tilapia species (O.niloticus) at the reservoir.Therefore, the management option for the species should aim at protecting immature males and females so that they can reach breeding size and produce eggs to replenish the stock.

Figure 1 .Figure 2 .
Figure 1.Monthly variations of all maturity stages of female O. niloticus from Tekeze reservoir.

Figure 3 .
Figure 3.Total length and total weight relationships for both sexes.

Figure 5 .
Figure 5. Relation between (a) fecundity and total weight (b) fecundity and total length (c) fecundity and ovary weight of O. niloticus from Tekze Reservoir

Table 1 .
Morphological, physical and chemical characteristics of Tekeze reservoir.

Table 3 .
Monthly (Mean ±SE) of the Fulton Condition Factor of O. niloticus captured from Tekeze reservoir.