ARCHIVAL WEBSITE
You are viewing the archived website of Pond Dynamics / Aquaculture CRSP. When using this website, please understand that links may be broken and content may be out of date. You can view more information on the continuation of PD/A CRSP research archived at AquaFish Innovation Lab.
8KR3 - Relative Contribution of Supplemental Feed and Inorganic Fertilizers in Semi-Intensive Tilapia Production
PD/A CRSP Seventeenth Annual Technical Report
Previous Section
Table of Contents
Next Section


Cite as: [Author(s), 2000. Title.] In: K. McElwee, D. Burke, M. Niles, X. Cummings, and H. Egna (Editors), Seventeenth Annual Technical Report. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, [pp. ___.]

Relative Contribution of Supplemental Feed and Inorganic Fertilizers in Semi-Intensive Tilapia Production

Eighth Work Plan, Kenya Research 3 (8KR3)
Final Report

Karen Veverica
Department of Fisheries and Allied Aquacultures
Auburn University, Alabama, USA

Jim Bowman
Department of Fisheries and Wildlife
Oregon State University
Corvallis, Oregon, USA

Wilson Gichuri, Paul Izaru, and Patricia Mwau
Department of Hydrobiology
University of Nairobi
Nairobi, Kenya

Thomas Popma
Department of Fisheries and Allied Aquacultures
Auburn University, Alabama, USA

Abstract

A 20-week experiment was conducted at Sagana Fish Farm, Kenya, to characterize the productive capacity of ponds at this new CRSP research site and to determine least-cost combinations of rice bran and inorganic fertilizer. Twelve 800-m2 ponds were stocked with juvenile (32 g each) Oreochromis niloticus at 20,000 ha-1 and Clarias gariepinus fingerlings (average weight 4.6 g) at 2,400 ha-1. Ponds contained about half sex-reversed and half mixed-sex tilapia, with an estimated ratio of approximately 75% males to 25% females at stocking. Four treatments were applied in triplicate as follows: 1) Urea and DAP to provide 16 kg N ha1 wk-1 and 4 kg P ha-1 wk-1; 2) Urea and DAP applied to give 8 kg N and 2 kg P ha-1 wk-1, plus rice bran fed at 60 kg ha-1 d1; 3) Rice bran fed at 120 kg ha-1 d-1; and 4) Rice bran as in Treatment 3 and fertilizer as in Treatment 2. Net fish yield averaged 1,127, 1,582, 1,607, and 2,098 kg ha-1 for Treatments 1 through 4 respectively. Fish in ponds receiving rice bran (Treatments 2, 3, and 4) were still growing rapidly at harvest time, but the growth rate of fish in Treatment 1 was beginning to decrease near the end of the experiment. Treatment 1 was the most cost-effective, but Treatments 1, 2, and 4 all resulted in fairly similar net profits. Input costs for Treatments 1 and 2 will be of interest to fish farmers, although it is possible that fish raised using only fertilizer at the rates used in Treatment 1 may never reach market size at this stocking density. Fish had reduced growth towards the end of the culture period and resulting low final average weights, which were less than 100 g. If rice bran had cost 3.5 ksh or less per kilogram, profit for Treatment 3 would have surpassed that of Treatment 1. If rice bran had cost less than 5.8 ksh per kg, Treatment 2 would have been more profitable than Treatment 1.

Introduction

Aquaculture development in Kenya has been hampered by a lack of nutritionally complete feeds. The application of chemical fertilizers can enhance natural food production and indirectly provide protein to complement energy-rich rice bran (> 16 kcal digestible energy (DE) g-1 protein; National Research Council, 1993). A 20-week experiment was conducted at Sagana Fish Farm, Kenya, to characterize the productive capacity of ponds at this new CRSP research site and to determine least-cost combinations of rice bran and inorganic fertilizer. In addition to the work reported here, samples of fish tissue, fish feed, soil, and plankton from the experiment were provided to researchers at the University of Arkansas at Pine Bluff for a companion study entitled "Nutritional Contribution of Natural and Supplemental Foods for Nile Tilapia: Stable Carbon Isotope Analysis." (See pp. 29–31 of this report)

Methods and Materials

This experiment was conducted in twelve newly renovated CRSP research ponds at Sagana Fish Farm, Kenya, beginning on 31 October 1997 and ending on 25 March 1998 (144 days). Lime was applied to all ponds at a rate of 5 t ha-1 prior to the experiment. The newest ponds received the lime treatment just prior to filling. Twelve 800-m2 ponds were stocked with juvenile (32 g each) Oreochromis niloticus at 20,000 ha-1 and Clarias gariepinus fingerlings (average weight 4.6 g) at 2,400 ha1. Ponds contained half sex-reversed and half mixed-sex fish, with an estimated ratio of approximately 75% males to 25% females at stocking.

Four treatments were applied in triplicate as follows:

Due to the relative newness of some ponds and a suspected high P adsorption capacity of newly exposed pond bottoms, the ponds were blocked according to the following criteria:

Ponds were assigned randomly in a split block design, with one replicate of each treatment in each block.

Dissolved oxygen concentration, temperature, and pH were measured weekly in the morning and afternoon. Total alkalinity, chlorophyll a, Secchi disk visibility, and total ammonia nitrogen concentration were measured every two weeks. Total nitrogen, mineral nitrogen, total phosphorus, and soluble reactive phosphorus were analyzed monthly. Samples for water chemistry were taken on Mondays, fertilizing was done on Tuesdays, and dissolved oxygen and temperature readings were done on Thursdays.

Ponds were sampled monthly for fish growth and drained completely after 20 weeks. Tilapia were separated by sex, counted, and weighed at draining. Tilapia reproduction was weighed and subsamples were counted. Clarias were counted and weighed. Two fish of each species were taken from each pond for proximate analysis at the beginning, in the middle (day 65), and at the end of the experiment. The whole fish, including viscera, was used for proximate analysis.

Rice bran was sampled as each lot arrived and combined into three batches. Each of the batches was analyzed for protein, fiber, lipids, ash, and phosphorus. Proximate analyses of fish and feedstuff was done using the Kjeldahl method for protein, Soxhlet extraction for lipids, and muffle furnace for ash and phosphorus.

Results and Discussion

Although the Treatments 1, 2, and 3 were intended to be iso-nitrogenous, the rice bran contained less protein than expected. Total nitrogen and phosphorus inputs are summarized in Table 1. The first few lots of rice bran (all in Batch 1) contained excessively high quantities of fiber, and large amounts of husks were observed in these lots. Proximate analyses of the batches are given in Table 2.

Some of the ponds in Blocks 1 and 2 still had residual lime on their bottoms after draining; however, there were no significant differences in fish production between blocks. Block assignments were therefore not taken into account in the statistical analyses of other results. At harvest, the average weights of tilapia were 89, 106, 106, and 131 g, and Clarias weights were 110, 217, 236, and 296 g for Treatments 1, 2, 3, and 4, respectively (Table 3). Male tilapia and Clarias showed significantly different average weights among treatments but differences among female tilapia were significant only for Treatments 1 and 4. Survival ranged from 67 to 88%; there were no significant differences by treatment. Males made up 65 to 71% of total tilapia numbers at draining.

Fish began spawning during the first month of the experiment. However, due to the presence of Clarias, few fingerlings survived to harvest.

Net fish yield averaged 1,127, 1,582, 1,607, and 2,098 kg ha1 for Treatments 1, 2, 3, and 4, respectively (Table 3). Fish in Treatments 2, 3, and 4 were still growing rapidly at harvest time, but the growth rate of fish in Treatment 1 was beginning to decline near the end of the experiment (Figures 1 and 2).

Supply waters at Sagana typically have total alkalinity (TA) levels between 10 and 20 mg l-1 as CaCO3. TA levels in all ponds were higher than this throughout the experiment. The average TAs of ponds receiving rice bran rose to 70 mg l-1 or above by day 46 and remained relatively steady throughout the remainder of the experiment, whereas those of ponds receiving only chemical fertilizer descended to levels between about 35 and 55 mg l-1 after the first month (Figure 3).

Ponds in Treatment 1 had the highest average chlorophyll a concentrations (Tables 4 and 5). After December (month 2) the ponds in Treatment 3 (rice bran only) developed good algal blooms; however, prior to the second month they had little phytoplankton, and dissolved oxygen concentrations were frequently less than 1 mg l-1 in the morning.

Ponds that received only inorganic fertilizer (Treatment 1) had significantly more suspended silt (inorganic matter) than other ponds (tss - vss in tables 4 and 5), and suspended inorganic matter was linearly and negatively correlated with bran input rate (Figure 4). Nesting activities in monosex tilapia ponds can be a source of pond levee erosion. The high clay content of the pond soils at sagana means they would be subject to erosion from activities such as nest building by male tilapia and from wind. Additions of organic matter such as chicken litter have been reported to reduce levels of suspended silt in ponds (Boyd, 1982; Teichert-Coddington et al., 1992).

Soluble reactive phosphorus concentrations were never higher than 0.05 mg l-1 in any pond and were not greater than 0.03 mg l-1 in ponds receiving rice bran (Figure 5). Total ammonia nitrogen never surpassed 0.5 mg l-1 except on the final sampling date in Treatment 1 (Figure 6). Nitrate and nitrite were low throughout the experiment as well; average nitrite concentrations never surpassed 0.1 mg l-1 NO2-N, and average nitrate concentrations always remained below 0.25 mg l-1 NO3-N (Table 5). ";Black cotton"; soils, such as the ones in which the Sagana ponds are built have notoriously high phosphorus-adsorption rates. Assuming a clay content of 80% (ponds at Sagana are reported to have as much as 90% clay soils), a phosphorus adsorption capacity of 350 kg ha-1 was estimated using the formula proposed by Shrestha and Lin (1996). Only 80 kg ha-1 was added to most ponds over the 20 weeks, so the phosphorus adsorption capacity was far from satisfied, and little difference was observed among blocks of ponds.

Nitrogen efficiency (kg net fish yield per kg N applied) averaged 3.52, 5.42, 6.09, and 4.95 for Treatments 1, 2, 3, and 4, respectively. The N:P ratio of the inputs was lowest for Treatment 3.

At Rwasave Fish Station, Rwanda, an experiment was conducted in which fish were fed rice bran at 5 g fish-1 d-1 in ponds stocked at a density of 2 male tilapia m-2. A mean net yield of 2,620 kg ha-1 and a feed conversion ratio (FCR) of 7.6 were obtained after 192 days. No chemical fertilizers were applied, but small additions of chicken manure and grass were used as fertilizer (Verheust et al., 1992). The FCRs obtained in Treatments 3 and 4 of this experiment, in which rice bran was applied at 6 g tilapia-1 d-1, are much higher (10.5 and 8 for Treatments 3 and 4, respectively) because the fish started out at a smaller size (32 g in this experiment versus 80 g in the Rwanda experiment) and they could not consume all the bran. Also, the ponds were harvested before market size was reached, thereby not allowing recovery from the overfeeding at the beginning of the experiment. A study that combined inorganic fertilization and feeding rice bran to tilapia (reaching a maximum application of 46 kg ha-1 d-1) obtained a mean net yield of 1,160 kg ha-1 after 159 days (Perschbacher and Lochmann, 1995). This is somewhat less than the net yield in Treatment 2 of this experiment (1,582 kg ha-1).

Fish of both species contained significantly higher lipid levels when fed rice bran (Treatments 2, 3, and 4) than when in the fertilizer-only Treatment 1 (Tables 6a and 6b).

Using the results of Treatment 1 (fertilizer only) as the baseline, partial enterprise budget analysis demonstrates that none of the other treatments resulted in greater returns to operating costs. However, in comparing Treatment 4 with Treatment 3, applying fertilizer in addition to the high bran application was definitely beneficial in the economic sense (Table 7).

Conclusions

Treatment 1 (fertilizer only) was the most cost-effective, but the net profits from Treatments 1, 2, and 4 were similar (Table 8). Input costs for Treatments 1 and 2 will definitely be of interest to fish farmers. Treatment 1 resulted in the highest profit (Table 8); however, it is possible that fish raised using only fertilizer at the rates in Treatment 1 may never reach market size at this stocking density, as evidenced by their reduced growth towards the end of the experiment (Figures 1 and 2).

In order to further increase production over that obtained using Treatment 2, applying additional fertilizer may be a better solution than increasing bran inputs. Diana et al. (1996) found that adding supplemental feed (floating pellets, 30% crude protein) after fish reached 150 g resulted in greater annual profit than either fertilization only or feeding right from the start. In this experiment, the bran can be considered to function partly as a feed and partly as an organic fertilizer.

Bran prices vary in Kenya. Rice bran can be purchased for as little as 3 KSh kg-1 (by special arrangement with government-owned rice mills), but 6 KSh kg-1 is more common for farmers buying retail. At the price of 6 KSh kg-1, rice bran should be applied sparingly and not as a fertilizer. It is rumored that some unscrupulous retailers mix husks (obtained free) with bran. Proximate analyses indicate that some of this bran may have obtained early in the experiment. Wheat bran is available in greater quantities than rice bran and retails for 5 to 7 KSh kg1. One batch was tested and resulted in 14.64% protein, 2.9% lipids, and 12.52% fiber. Wheat bran may present a better alternative for farmers and should be tested in future PD/A CRSP experiments. It is now being used at Sagana Fish Farm. Maize bran (8% protein) and maize germ (16% protein) are also available in even greater quantities than wheat bran, but prices are higher per unit protein. If rice bran had cost 3.5 ksh kg-1 or less, then profit for Treatment 3 would have surpassed the profit of Treatment 1. If rice bran had cost less than 5.8 ksh kg-1, then Treatment 2 would have been more profitable than Treatment 1. Since this experiment was conducted, rice bran prices in the Sagana area have dropped to just over 2 KSh kg-1, making it much more attractive as a component in pond management regimes.

Anticipated Benefits

This experiment has provided data on the comparative value of inorganic fertilizers and low-cost supplemental feeds as pond inputs for semi-intensive tilapia production at Sagana, thus providing the basis for the development of more efficient production strategies for pond systems in Kenya and similar areas of Africa. In addition, determining pond productivity using high nutrient input levels at Sagana provides data for comparison of the Africa site with other CRSP sites in Southeast Asia, Central America, and South America.

Literature Cited

Boyd, C.E., 1982. Water quality management for pond fish culture. Developments of Aquaculture and Fisheries Science, Vol. 9. Elsevier, Amsterdam.

Diana, J.S., C.K. Lin, and Y. Yi, 1996. Timing of supplemental feeding for tilapia production. J. World Aquacult. Soc., 27(4):410–419.

National Research Council, 1993. Nutrient Requirements of Fish. National Academy Press, Washington, D.C., 114 pp.

Perschbacher, P.W. and R. Lochmann, 1995. Effects of form of defatted rice bran offered on Nile tilapia production in ponds. In: H. Egna, J. Bowman, B. Goetze, and N. Weidner (Editors), Twelfth Annual Technical Report, Pond Dynamics/Aquaculture CRSP, Office of International Research and Development, Oregon State University, Corvallis, Oregon, pp. 79–83.

Shrestha, M.K. and C.K. Lin, 1996. Determination of phosphorus saturation level in relation to clay content in formulated pond muds. Aquacult. Eng., 15(6):441–459.

Teichert-Coddington, D.R., B.W. Green, and R.P. Phelps, 1992. Influence of site and season on water quality and tilapia production in Panama and Honduras. Aquaculture, 105:297–314.

Verheust, L., E. Rurangwa, and K.L. Veverica, 1992. Production and growth of supplementally fed Oreochromis niloticus males stocked at three densities in fertilized ponds. In: H. Egna, M. McNamara, and N. Weidner (Editors), Ninth Annual Administrative Report. Pond Dynamics/Aquaculture CRSP, Oregon State University, Corvallis, Oregon, pp. 34–35.

Previous Section
Table of Contents
Next Section