The Effects of Stocking Density on the Growth and Survival of Nile Tilapia (Oreochromis niloticus) Fry at Son Fish Farm, Uganda

Tilapias are the world’s second most important fish species for aquaculture after the carp and this is due to their high growth rates, being prolific breeders, completing their life cycle in captivity, tolerance to environmental stress and high market demand [1]. Nile tilapia culture in Uganda traces way back in the late 1940’s just after it had proved to do better than the carp in Uganda according to experiments that were carried out at the by then Kajjansi Experimental Station [2] The culture of Tilapia expanded exponentially until a more recent overtake by the catfish making Tilapia now, Uganda’s second most cultured fish species. Nile tilapia is widely accepted on the Ugandan local market, regional markets and is also exported to the European market. Furthermore, the declining capture fisheries worldwide, (Gabriel et al., 2007 drives an increased need for aquaculture expansion in order to bridge the gap between production and demand. For Uganda, this new wave of aquaculture would be fully endowed by small scale farmers if they had adequate knowledge on the appropriate Nile tilapia fry stocking densities as it is believed to be a critical husbandry practice that can boost the production capacity and efficiency of a culture system [3] in addition to determining the economic viability of production system in intensive aquaculture [4]. Stocking densities of Nile tilapia fry normally range from 3000 to 4000 fry/m3. However a stocking density as high as 20,000 fry/m3 is also practical given good water quality [5]. SON Fish Farm is currently using a stocking density of 4000 fry/m3 in 2x1x1 meter hapas and most fish farms in Uganda don’t deviate much from this density. However, the basis of this stocking density is not justified and the definite best stocking density is not known. It is on the basis of this knowledge gap that the current study was proposed.


Introduction Background
Tilapias are the world's second most important fish species for aquaculture after the carp and this is due to their high growth rates, being prolific breeders, completing their life cycle in captivity, tolerance to environmental stress and high market demand [1]. Nile tilapia culture in Uganda traces way back in the late 1940's just after it had proved to do better than the carp in Uganda according to experiments that were carried out at the by then Kajjansi Experimental Station [2] The culture of Tilapia expanded exponentially until a more recent overtake by the catfish making Tilapia now, Uganda's second most cultured fish species. Nile tilapia is widely accepted on the Ugandan local market, regional markets and is also exported to the European market. Furthermore, the declining capture fisheries worldwide, (Gabriel et al., 2007 drives an increased need for aquaculture expansion in order to bridge the gap between production and demand. For Uganda, this new wave of aquaculture would be fully endowed by small scale farmers if they had adequate knowledge on the appropriate Nile tilapia fry stocking densities as it is believed to be a critical husbandry practice that can boost the production capacity and efficiency of a culture system [3] in addition to determining the economic viability of production system in intensive aquaculture [4]. Stocking densities of Nile tilapia fry normally range from 3000 to 4000 fry/m 3 . However a stocking density as high as 20,000 fry/m 3 is also practical given good water quality [5]. SON Fish Farm is currently using a stocking density of 4000 fry/m 3 in 2x1x1 meter hapas and most fish farms in Uganda don't deviate much from this density. However, the basis of this stocking density is not justified and the definite best stocking density is not known. It is on the basis of this knowledge gap that the current study was proposed.

Problem statement
Fish farmers in Uganda in the grow-out sector often face low yield that culminates in low profit margins given input costs. This setback, to some extent stems from lack of sufficient knowledge on the best practice of fry stocking densities in hapas or nursery ponds that would maximize production. For this reason, a study was proposed to address the concern in a local setting that could be representative of various potential fish farm sites in central Uganda.

Justification
The research was envisaged to give knowledge on appropriate stocking density of tilapia fry that would maximize production and profitability of grow-out aquaculture enterprises.

Objectives of the study
Major objective: To establish the most appropriate stocking densities for Nile tilapia fry in hapas set in earthen ponds.

Specific objectives:
To investigate the effects of stocking density on feed conversion ratios (FCR) of Nile tilapia stocked in hapas in earthen ponds.
• To determine the effect of stocking density on growth homogeneity of Nile tilapia fry in hapas in earthen ponds.
• To compare the growth (daily weight gain (DWG) and specific growth rate (SGR) rate) at different stocking densities.
• To determine the survival rate of Nile tilapia fry at various stocking densities.

Study site
The research was carried out at Source of the Nile (SON) fish farm, a mono-sex tilapia culture fish farm located in Bugungu, Njeru sub county in Buikwe district. One of the primary nursing ponds at this farm (G ponds) were used for the experiment (Figure 1).

Research design
The case study involved stocking Nile tilapia fry at different stocking densities and maintaining uniform management practices for a period of four weeks. Data was collected weekly to assess impact of stocking density on survival and growth of fry.

Experimental design
Two replicates were used for each treatment and these were fed on 20%, 18% and 15% body weight in week 1,2 and three respectively. This was done using a 45% crude protein feed.

Specific methods
Harvested fry of mean length 12.4 ± 0.16 mm and mean weight of 0.03 ± 0.0016g originating from the same brood were randomly selected and stocked in hapas in various stocking densities that is;1000fry/m 3 , 1330 fry/m 3 , 2000fry/m 3 , 2670fry/m 3 , 4000fry/m 3 and 5330 fry/m 3 , respectively . Fish were fed manually with 20%, 18%, and 15% of body weight daily for the first week, second week and for the rest of the experiment duration, respectively. Aquatro feeds (45% crude protein) obtained from Livestock Feeds Limited (LFL) in Mauritius were used for three weeks with a feeding frequency of eight rounds per day. Every after one week, measurements were carried out on fish fry between 09:00 -11:30 am depending on the weather of the day and finally at the time of grading. Standard length and body weight were taken to the nearest millimeter and nearest 0.01g, respectively using calibrated graph paper and a mini digital pocket gram scale. The fish were removed from the hapas between the 24 th and 25 th day and graded.
To determine survival at grading time, number of fish in the hapa was established and computed against the number at stocking.

Statistical analysis
Data collected from the experiment was analyzed using one-way ANOVA and when ANOVA indicated that there was a statistical difference between the stockings densities means, Tukeys Tests were used to compare these means. The software of statistics calculations and model was SPSS statistics version 20 and all statistical tests considered significant at (P < 0.05).

Length and weight
The mean length at the beginning of the experiment and at week one sampling were not significant for all the stocking densities (P =1.0 and 0.192, respectively). At sampling two (week two) there was a significance difference in the mean length of all the three samples (P < 0.0001). Tukeys test indicated a significance between the highest density of 5330fry/m 3  The relationships between stocking density and mean length at week one, two and three samplings are shown in Figure 2.
All the stocking densities showed a gradual increase in weight with density of 1000 fry/m 3 having the greatest gain from 0.03 g average body weight at stocking to 0.43 ± 0.057 whilst 5330 fry/m 3 registered the lowest from 0.03 average body weight at stocking to 0.24 ± 0.042g by grading time (Figure 3).

Survival rate
From the experiment, the results were divided into two groups depending on the percentage survival. The first group had stocking densities 2670,1330,1000, and 2000 fry/m 3 in order of decreasing survival but all having high survivals of above 90%. The second group had densities 4000 and 5330 fry/m 3 and these had less than than 90% as far as survival is concerned (Table 1).
There existed a high negative correlation between survival and stocking density (P= -0.835) ( Figure 4).   Food conversion ratio, specific growth rate and daily weight gain Results of food conversion ratio (FCR) specific growth rate (SGR) and daily weight gain (DWG) are shown in Table 2. There was a great significant difference in the food conversion ratios and specific growth rates of the six stocking densities (F = …..; P<0.0001). FCR was lowest at stocking density of 1000 fry/m 3 (1.63) while 5330 fry/m 3 had the highest FCR (2.22). The daily weight gain decreased with increasing stocking density with an anomaly existing between densities of 2670 and 4000 where by 2670 had a lower daily weight gain (0.0116 g/day) than 4000 (0.0121g/day) which was higher in density. Daily weight gain and specific growth rate both decreased with increasing stocking density with 5330 fry/m 3 having almost just half the daily weight gain as 1000fry/m 3 , however in both cases there was an anomaly at a stocking density of 2670 fry/m 3 which had lower daily weight gain and specific growth rate than expected.
Linear regression showed a great correlation between food conversion rate and the growth rate using mean length. (R 2 =0.878) with food conversion able to explain up to 77% of the length (R 2 =0.77). There existed a very good correlation as well between stocking density and growth rate using mean length (R=0.928 and R 2 =0.86). This also happened between stocking density and feed utilization or conversion level. (R 2 =0.844 and R 2 = 0.712) However the effect of stocking density on growth could not be explained by the regression (ANOVA P=0.52) while the effect of stocking density on feed utilization level and that of feed utilization level on growth were significantly explained by the regressions ( P=0.035 and 0.022 respectively).

Growth homogeneity
At stocking the fry sampled fell in three homogenous groups according to length in millimeters that is, 10-11mm with 13 fry, 12-13mm with 29 fry and 14-15mm with 8 fry hence there was some degree of homogeneity at stocking. This homogeneity was however not realized by the first sampling after one week (Table 3) as length ranged in up to 5 groups in all the six stocking densities. In the second and third week samplings, there was an observed trend of increasing homogeneity with increasing stocking density. Homogenous growth was observed in the high stocking densities with all the samples in the highest density (5330 fry/m 3 ) falling in just three groups and 36 out of the 50 sampled being below 24 mm. 21 18 24   The number of fry in lower higher classes (longer fry) also decreases as the stocking density increases hence shorter fry produced (Tables 2,4,5,6).

Stocking density and growth
The effect of population density is usually seen to be either density dependent or density independent. Wiener et al [6] suggested that stocking density that negatively affect fish growth are density dependent. In the current study, the first sampling showed no significant effect of stocking density on growth while the second and third showed a great significant effect of stocking density on growth. This can be explained by the fact that at appropriate stocking density before attainment of carrying capacity, the fish grow properly thus at this stage stocking density did not affect fish growth by sampling one. This can also be explained by the slow growth after immediate stocking of fry due to stress from the harvesting process thus no extreme utilization of resources in the hapa.
By the second sampling the highest stocking density (5330 fry/ m 3 ) was significantly different from the rest of the stocking densities. This indicated that as early as two weeks it had reached its carrying capacity which is one of the reasons to explain reduced growth rate. In sampling three, there was no significance between density 4000 and 5330 fry/m 3 implying that in a time period of one week 4000 fry/m 3 had also increased in biomass to carrying capacity thus affecting growth. 2670 fry/m 3 density reported lower growth rates than 4000 fry/m 3 even though they did not differ significantly, this can be explained by the high survival rates of 96.9% as compared with 82% in 4000 fry/m 3 making them have insignificantly different number of fry. These results are in agreement with Yousif [7] and Breine et al. [8] who reported that it's a generally accepted principle that increasing stocking density adversely affects fish growth. High stocking densities are associated with stress, competition for food and living space, voluntary appetite suppression and more energy expenditure in antagonistic interactions [9,10]. According to Bacellos et al [11], stress leads to increased cortisol production by even resting plasma cortisol and these increases with increasing stocking density, he continues that higher cortisol concentrations are considered as chronicle response to social stress due to high stocking density and this impairs fish growth due to mobilization of dietary energy by physiological alterations caused by stress.

Survival rate
In this study, the high survival rates of tilapia fry at high stocking density (82.9% at 5330 fry/m 3 ) indicate amenability of tilapia to intensive culture [3]. However, this can also be attributed to favorable environmental conditions during the experiment. This is in agreement with El Sherif [11] who found out that survival of up to 100% could be linked to good environmental conditions. Lower survivals were realized when stocking density was higher than 2670 fry/m 3 . This means that tilapia fry can survive at high densities of up to around 2670 fry/m 3 but at extremely high densities the survival rate is significantly affected, this can be due to stress, decreasing water quality due to increased biomass and other density dependent factors at this point not only affecting growth rate, but also cause increase mortality rates. Thus 4000 fry/m 3 and 5330 fry/m 3 are in agreement with Osofero [12] who found an inverse relationship between stocking density and survival while 1000,1330,2000 and 2670 fry/m 3 are in line with Alhassan [3] who realized no effect of stocking density on survival. The high stocking densities (4000 fry/m 3 and 5330 fry/m 3 ) are not in agreement with Alhassan [3] because he considered very low stocking densities (8, 10 and 12 fry/m 3 ) thus could not find the limit to the insignificant effect stocking density on survival as the current study did ( Figure 5).

Growth homogeneity
Yousif [7] realized that as initial size was homogenous for all densities and ample daily feedings done, one would expect fish in each density to be slightly but not significantly different in final body sizes. In this study however although the initial size was homogenous stocking density had an effect on final size among individuals of initially the same size. The little or no homogeneity in first sampling occurring at high stocking density can be due to tilapia's aggressive behavior [4] where some fish maintain a domain position, grow faster while others in low dominance hierarchy end up being chronically stressed leading to appetite suppression, reduced food intake and low growth rates thus decreasing homogeneity in growth of fish from the same brood stock and feed on the same food under the same environment. In the second and third sampling growth homogeneity however increased with increased stocking density with 5330 fry/m 3 being the most homogeneous. This can be related to studies on Artic charr. According to Jobling 1985 [13], change in interspecies relationships that defend territories is reduced as stocking density increases to extreme in Artic charr, this could also explain findings that a positive relationship exists between stocking density and homogeneity in Artic Charr [14] and thus highest homogeneity at high tilapia stocking densities in this study. This is still in agreement with Kanawabe [15] who worked on Salmoids and found out that rearing them at high densities reduces territorial behavior and breaks the dominance hierarchy. This reduces the chronic stress by fish in low dominance hierarchy thus improving their growth; however the improved growth doesn't indicate that every individual is growing fast but rather increased growth of slow growers and reduced growth of fast growers due to reduced dominance over food.
Since Nile tilapias raised in tropics are generally fast growing with little slow growers, there is a general significant decrease in growth rate at high densities and high growth homogeneity. This finding is however not in agreement with the negative relationships found between  stocking density and growth homogeneity in red tilapia (4,16). Joblings [16] realized that proper dominance hierarchies in Artic charr were established in three months so it can also be assumed that three weeks was not enough time for establishment of proper social interactions and dominance hierarchies in this study.

Food conversion ratio (FCR), specific growth rate (SGR) and daily weight gain (DWG)
Food conversion ratio increased with increasing stocking density thus as stocking densities increased the fry became less efficient in utilizing the food for somatic growth. This is attributed to increased stress due to decreasing water quality as stocking density is increased far beyond the carrying capacity [17]. This is in agreement with Guimareaes et al. [18], who found out that efficient utilization of diets may vary within a single species because of not only the particular strain of fish but also environmental factors. On the contrary, Osofero [12] reported no effect of stocking density on food conversion ratio which he attributed to using the same feed in the same environment. Hence in the current study the environment played a very great role in influencing the food conversion ratio through water quality.
Daily weight gain and specific growth rate generally decreased with increasing stocking density indicating the decreasing feed utilization ability as explained above. However the anomaly at 2670 fry/m 3 is attributed to the very high survival rate (Figure 4) which made the carrying capacity to be realized faster. This affected growth negatively thus lowering both the specific growth rate and daily weight gain.   • Feed conversion ratios, specific growth rates and daily weight gain are as well affected by stocking density in this study.

Conclusions and Recommendations
• Many individuals grow at less than their maximum potential rate at high stocking densities thus growth homogeneity • Fish growth is plastic hence the proper growth rate lacking at some stocking densities can be regained in secondary nursing hapas by proper stocking and growth homogeneity at grading, but at the same time larval fish are very delicate hence need proper management and this is can be achieved when they are in one unit which calls for higher stocking density. Thus survival of fish is the most important factor in the primary nursing hapas since lost fish can't be regained in the following rearing stages and this rules of stocking densities of 4000 fry/ m 3 and 5330 fry/m 3 • Further research is needed to understand how stocking density affects growth and survival in secondary nursing happas and other grow out stages that follow afterwards [19][20][21][22][23][24][25].

Recommendations
According to this study, it is recommended to have stocking densities not exceeding 2670 fry/m 3 in hapas placed in earthen but depending on one's required output and resources available to manage carrying capacity this can vary.