To learn more about our current work, please visit AquaFish Innovation Lab.
PD/A CRSP Aquanews-Summer 2002

Contents Previous Page
CRSP Homepage

Graduates’ Corner

Congratulations to CRSP-sponsored graduate students Enos Mac’Were, Bethuel Omolo, and Robert Carpenter on completion of their degrees. In March 2002, Enos Mac’Were earned a Master of Science degree from the Fisheries Department of the Faculty of Forest Resources and Wildlife Management at Moi University, Kenya. His major professor was Charles Ngugi. Bethuel Omolo earned a Master of Science degree from the Department of Fisheries and Allied Aquacultures at Auburn University. A CRSP-sponsored graduate student profiled in the Summer 2000 issue of Aquanews, he defended his thesis in November 2001. Ron Phelps was his major professor. Rob Carpenter, also advised by Ron Phelps, earned a Master of Science degree from the Department of Fisheries and Allied Aquacultures at Auburn University in August 2002.

(abstract of Enos Mac’Were’s M.S. thesis)

Image of Clarias gariepinus courtesy of FishBase (<>), reproduced with permission from Dr. Guy Teugels, Royal Museum for Central Africa, Tervuren, Belgium.

Oreochromis niloticus and Clarias gariepinus were reared for 180 days in fertilized ponds receiving different feeds to compare growth and financial analysis. Supplemental feeds tested were a readily available agricultural by-product (rice bran), commercially available animal diet (pig finisher pellet), and a pelleted test diet. Test diet was formulated to contain 20% crude protein (CP). The two pellets were compared to rice bran.
Twelve 800-m2 ponds, with four replicates for each of three feeds and fertilizer regimes were limed at 2.5 ton ha-1 just prior to filling with water. Each pond was stocked with 1550 tilapia and 50 Clarias averaging 89 g and 330 g respectively. Fish were fed twice a day at 10 A.M. and 4 P.M. at estimated 2% body weight. Each pond was fertilized at 5 kg P ha-1 wk-1 and 20 kg N ha-1 wk-1 from diammonium phosphate (DAP) and urea.
Water quality parameters were not significantly different (p>0.05) among treatments except total alkalinity in which TDP was higher than RB, and both dawn and afternoon pH and dissolved oxygen where ponds receiving RB and PFP had highest and lowest (p<0.05) values respectively.
All three feeds contained much lower protein than expected. Rice bran (RB) had 5% CP content against expected 10%. Pig finisher pellet (PFP) advertised to contain 14% had 10% while test diet pellet (TDP) had only 11%.
RB had significantly lower values (P<0.05) in fish growth rate, gross and net fish yields, net annualized production and with higher values on apparent feed conversion ratio compared to both PFP and TDP. However, there were no significant differences (p>0.05) in survival rate and stocking size of fish among treatments.
Relative profitability analysis using partial and full enterprise budgets revealed that PFP was the best feed followed by RB at normal selling price of Kshs 90 kg-1 fish. At higher prices of Kshs 120 kg-1 and if price varies by size in favour of bigger fish PFP would still be a better choice followed by TDP while RB would be least profitable. RB and PFP had significantly lower (P<0.05) break-even prices on total cost than TDP. RB recorded least operational cost.
This study assessed relative profitability of the three nutrient regimes and gave full enterprise budget for the most profitable one. It should therefore be a valuable contribution to development of management practices involving nutritionally incomplete feeds. In addition, the development of cost-effective tilapia feed may increase the profitability of fish farming in the region and stimulate commercial aquaculture enterprises.

(abstract of Bethuel Oduor Omolo’s M.S. thesis)

Research was conducted in 0.04-ha ponds at Auburn University Fisheries Research Station to study changes in feed conversion efficiency by channel catfish after reaching a minimum market weight of 250 g and assess the implications for economically optimum harvest weight. Fish from five nominal weight classes (250, 500, 1000, 1500, and 2000 g) were stocked in two randomly selected ponds per weight class and fed a 32% protein, floating feed once daily to apparent satiety. In order to maintain similar water quality in all ponds, the stocking densities were such that, based on existing bioenergetic models, predicted average daily feed consumption at water temperatures of 29°C would be near 50 kg/ha in all ponds. All observed mortalities were counted and weighed. Ponds were harvested when cumulative feed consumption exceeded 150% of initial fish biomass or 1600 kg/ha (average rearing time 86 d).
After harvest, fish were restocked to produce two additional replications of the 500-, 1000-, and 2000-g nominal weight classes, following the same feeding criteria. Four of the 16 ponds failed one or both inclusion criteria for further analysis (survival > 70% and fish recovery > 95% of initial stock). In the remaining 12 ponds, average water temperature was 27 C, fish weights were not related (each P >0.10; r2 <0.20) to dissolved oxygen concentration at dawn (5.9 mg/L), survival (95%), recovery of stocked fish (99%), feeding rate (31 kg/ha-d), and standing crop of fish at harvest (3341 kg/ha). Feed conversion ratios (FCR, feed/gain) ranged from 1.47 to 3.54 and were related to fish weight (P<0.01, r2=0.63). The relationship between FCR and average weight was described by the equation FCR=1.41 + 0.54Wkg. This relationship suggested that at current feed and fingerling prices ($215/metric ton and $60/1000, respectively), feed plus fingerling costs (others factors held constant) per metric ton of net fish yield is lowest if fish are harvested at a weight of 1000 g. If fingerling prices remain at current rates ($60/1000) but feed prices increased by 85% to $401/metric ton, the smaller weight class of catfish (250–500 g) became the economically superior size class due to the least cost of input requirement to obtain a net fish yield of one ton. When feed prices held at current rates but fingerling prices raised, the next larger size class (1000-1500 g) became economically superior only after fingerling prices are increased by 133% to $140/1000.

Abstract of Robert Harper Carpenter’s M.S. thesis

Inheritance of sex ratio was studied using three strains of Oreochromis niloticus progeny sets from nine individual pair spawns, designated as families, from April to October of 2000. Five males were placed in hapas with half the females from their family and were allowed to spawn during the summer months. Ten sets of progeny from each family, eight from Family VI, were collected after at least sixty days of grow-out. Sex ratios of all 88 progeny groups were determined by microscopic examination of gonads. Sex ratio of each spawn was compared to that of the parent family, family mean, and population mean using Chi-square tests. Of the nine families, four produced progeny where 100% of the ten sets were significantly different from the parent ratio (P < 0.05). Of the other five families, four had 40 to 90% of the progeny sets different from the parent while Family VIII only reported 10% of the sets to be significantly different. Results of progeny sets when compared to the family mean, mean proportion of males for all that families spawns, revealed three families where 10% of the progeny sets were significantly different. The remaining families revealed 20 to 50% of progeny sets different from the family mean. Similar differences were found when comparing each spawn to the population mean. Calculations for heritability of sex ratio revealed an h2 value of 0.09 with an R2 value of 0.04. Results suggest sex ratio is not an inherited trait.
A temperature study was also performed on spawns from three families to determine the effect of environment on sex determination. Spawns were separated into equal groups at swim-up and one group was reared in temperature controlled aquaria at 35±1 °C for 30 days and the other at ambient temperature averaging 29±1 °C. High temperature skewed the sex ratio to male in two of the three families tested giving evidence that there was an environmental component in sex determination. This effect, however, is not always expressed.

Contents Previous Page
CRSP Homepage