North American Journal of Aquaculture 68:192–197, 2006 [Article]

Ó Copyright by the American Fisheries Society 2006

DOI: 10.1577/A05-012.1

Alternate-Day Feeding Strategy for Nile Tilapia Grow Out in the

Philippines: Marginal Cost–Revenue Analyses
REMEDIOS B. BOLIVAR AND EDDIE BOY T. JIMENEZ
Freshwater Aquaculture Center and College of Fisheries, Central Luzon State University,
Science City of Muñoz, Nueva Ecija 3120, Philippines

CHRISTOPHER L. BROWN*
Marine Biology Program, Florida International University, 3000 Northeast 151st Street, AC1 378,
North Miami, Florida 33181, USA

Abstract.—Nile tilapia Oreochromis niloticus were produced in an experimental, on-farm grow-out cycle in
which economic and production efficiencies were partially examined. Milled feeds were provided either daily
(control) or on alternate days (experimental) at a given percentage of body weight, ranging from 20% initially
to 2% at the end of the experiment. Growth, survival, feed conversion efficiency, cost, yield, and income were
considered under the two strategies. Although the cost of feeds was cut in half by the experimental feeding
treatment, feeding on alternate days did not reduce Nile tilapia growth or production performance variables,
and yields were not significantly different. A trend favoring larger fish among the controls was not significant,
and profit margins were higher for the alternate-day feeding strategy than for the control strategy. Although
feed conversion ratios varied considerably among the nine participating farms, the improved efficiency in the
experimental groups was consistently observed and was statistically significant. It is possible that the
improved performance attained by alternate-day feeding is a result of reduced feed waste, either through more
complete consumption of or improved nutrient absorption from available feeds.

Feed contributes substantially to the cost of
commercial fish production, and this expense may
make the difference between profitable and unprofit-
able culture enterprises. This challenge to fish farmers
includes those that cultivate tilapias Oreochromis spp.
in the Philippines (Brown et al. 2004). The de-
termination of aquaculture feeding strategies based on
mathematical, economic, and nutritional models can be
rather complex (Cacho 1993). Partly because of the
opportunistic nature of tilapias, models based on
feeding standardization for these species can be
somewhat unpredictable. Some feeding restriction
protocols lower feed costs without a net reduction in
crop yields, but it is not known whether these result in
more efficient feed consumption, better feed utilization
(increased feed conversion ratio [FCR]), or both. It has
also been argued on the basis of water quality concerns
that optimal feeding levels should be below the level
that supports maximal growth; regular feeding of fish
to the point of satiation increases the risk of waste, feed
decomposition, and compromised fish health (Hatch
and Kinnucan 1993). Often, decisions about the
amount of feed to use are based primarily on the
observations and assumptions of the farmer as opposed
to a detailed understanding of fish nutrition.
* Corresponding author: brownch@fiu.edu
Received February 2, 2005; accepted December 2, 2005
Published online April 18, 2006
Feed management strategies have been investigated
for several species of fish (Singh and Srivastava 1984;
Srikanth et al. 1989; Diana 1997). De Silva (1985,
1989) presented a number of mixed feeding schedules
that involved feeding Nile tilapia O. niloticus alter-
nately with high- and low-protein feeds. De Silva
(1985, 1989) found that Nile tilapia responded well to
some of the mixed feeding schedules (i.e., high-protein
feed alternated with low-protein feed) and that fish on
mixed schedules exhibited growth performance that
was comparable to the performance of fish reared
entirely on the high-protein feed. A substantial cost
saving on feeds was realized, and the nitrogen loading
of the system was reduced by some of these feeding
schedules (De Silva 1985; De Silva et al. 1993).
Tilapia culture is expanding in importance world-
wide, particularly in Southeast Asia, and the generation
of new strategies for tilapia culture could have an
economic impact throughout the region. Based on our
most recent experience, we anticipate that these studies
on feeding strategies will provide insight into the
suitability of a variety of cost-cutting options available
to farmers—not so much as a strict paradigm to be
followed, but rather a series of choices and expected
benefits to use depending on the circumstances of each
farm. Many farmers consider fish culture to be a mix of
art and science, and the ideal advice from the research
community does not always come in the form of

192
 NILE TILAPIA ALTERNATE-DAY FEEDING STRATEGY
a single best management practice, but rather a set of
options that farmers can invoke at their own discretion.
We previously conducted a series of farm trials on
the costs and benefits of feeding strategies available to
tilapia farmers (Brown et al. 2000, 2004; Bolivar et al.
2004). Tests of production efficiency of variable
feeding paradigms are not always conclusive, but
some studies—including those of Diana et al. (1996)—
have raised doubts about the efficacy of pushing the
feeding schedule to maximize growth. Our studies have
shown that semi-intensive tilapia culture can be
improved by reducing feed costs without compromis-
ing fish production.
The first study compared the production of Nile
tilapia in fertilized ponds with the initiation of
supplemental feeding at either 45 or 75 d poststocking
(Brown et al. 2000). Seven commercial farms were
involved in this study in which ponds were assigned to
each of the two treatments. Tilapia fingerlings (mean
weight ¼ 0.11 g) were stocked at a density of 4 fish/m2,
and a farm-made feed of rice bran and fish meal was
used beginning on either day 45 or day 75 and
continuing through the 150-d grow-out cycle. Yield,
daily weight gain, and survival rate were determined,
and cost analyses of the two feeding strategies were
carried out by use of a paired t-test. There were no
significant differences observed in growth (mean final
weight or mean daily weight gain) and survival (85–
87%) for the two treatments. The extrapolated fish
yields were 5,104 kg/ha after the 45-d onset versus
4,926 kg/ha after the 75-d onset, but this difference was
not statistically significant (P . 0.05). The earlier
feeding strategy required 37% more prepared feed than
the delayed feeding strategy, sharply increasing pro-
duction cost and reducing profit. The relatively better
FCR associated with delayed feeding may have
resulted from reduced waste, compensatory growth,
or both.
Our second study (Brown et al. 2000) compared the
growth, survival, and yield of Nile tilapia fed daily at
either 100% or 67% of experimentally determined
satiation. Analysis of growth performance demonstrat-
ed that the reduction of ration to 67% of satiation had
no effect on growth or yield, and again the more
moderate feeding strategy was more profitable. A third
study combined the initiation of feeding at day 75 and
100% or 67% satiation feeding. The results of this
study again demonstrated that the feeding of fish at
a 67% satiation level reduced the production cost in
tilapia culture relative to the 100% satiation feeding
(Brown et al. 2004), although the overall profitability
of the combined reduction of ration and delayed onset
of feeding was compromised in comparison with the
individual feed reduction methods tested previously.
193
A fourth feeding strategy was tested in the present
study; feeding on alternate days was compared with
daily. Growth, production, and economic factors were
monitored in this study, and we present the results here.
Methods
Nine commercial tilapia farmers in Luzon, Republic
of the Philippines, participated in this trial, which took
place from November 2002 to April 2003 and
corresponded to the dry season. At each participating
farm site, two ponds of similar sizes were assigned one
of two feeding schedules: daily feeding or alternate-day
feeding. One goal of this study was to test contrasting
feeding strategies under actual operating conditions at
commercial farms. Inherent in meeting that goal is the
somewhat less than complete control over small
variables, but this approach does provide insight into
the real-world applicability of certain techniques.
Because of the active participation of farmers and their
readiness to adopt profitable methods, the usual
extension process was unnecessary.
The ponds were stocked with androgen-sex-reversed
Nile tilapia of the GenoMar supreme tilapia, or GST,
strain. This strain is genetically selected for faster
growth and was acquired from GIFT Foundation
International, Inc. Fingerlings with a mean (6 SE)
weight of 0.19 6 0.03 g were stocked into the ponds at
a rate of 4 fish/m2. All ponds were fertilized weekly
with inorganic fertilizers at 28 kg N/ha and 5.6 kg P/ha.
Water depth of the ponds was maintained at 1 m.
Feeding was begun at the outset of the study, initially
with fry mash and progressing to finisher feeds in the
last month of the study. The fish were provided with
commercial feeds following the daily or alternate-day
feeding schedule based on a feeding guide for the
adjustment of feed amount and type. The highest
percentage of body weight (20%) was fed initially, and
the feed percentage decreased to 2% in fish over 200 g
in weight. Varying types of commercial feed were used
in this study, starting with fry mash (36% protein and
8% fat), then starter pellets (34% protein and 6% fat),
grower pellets (31% protein and 5% fat), and finisher
feed (29% protein and 5% fat). The rations used in the
control group (fed daily) were intended to be
representative of a feeding schedule used under typical
commercial production conditions as opposed to
deliberate overfeeding.
Two ponds of approximately 500 m2 were selected
at each participating farm; one pond was used for each
treatment. Ponds were not aerated mechanically.
Additional water was occasionally provided to main-
tain the water level at a 1-m depth.
A sample of 50 fish was obtained from each pond
every month to monitor weight gain and to provide the
194 BOLIVAR ET AL.

correct amount of feed (ending with 2% of body weight
per fish). After 120 d, the fish were harvested by
seining, followed by the complete draining of the
ponds. The total number of fish was counted, and the
fish were bulk weighed. Final mean weight, daily
weight gain, gross yield, and survival rate were
calculated. The total amount of feed given in each
treatment was also estimated at the end of the study.
The data were analyzed statistically by use of t-tests
and a mixed-model analysis (discussed below).
TABLE 2.—Marginal cost–revenue analysis of daily and
alternate-day feeding schedules for pond-reared Nile tilapia in
Luzon, Philippines. The analysis takes into account the
variable costs and revenues associated with the two feeding
strategies as well as two categories of fixed costs, namely,
labor (estimated as US$2.69/d) and other fixed costs
(construction, fuel, fertilizer, and acquisition of fingerlings;
estimated as $588 per pond per grow-out cycle).
Feeding schedule
Cost and profit Daily Alternate

Fish were harvested after 120 d of culture at an Variable costs

average size of approximately 150 g. This is a preferred
size in the Philippines market; tilapia are frequently
sold whole for pan-frying, and larger sizes or fillets are
not typically available. Farmers recorded the amount of
fish harvested and sold at the end of the experiment.
Additional data were recorded on the amount and cost
of feeds used and the time spent feeding fish in each of
the two treatments. Growth and performance data are
presented in Table 1, which includes as a footnote the
formula for calculation of the specific growth rate. Cost
data were used as the basis for a partial budget analysis,
in which the variable costs of daily and alternate-day
feeding schedules could be contrasted; these results are
shown in Table 2.
A mixed-model statistical analysis was carried out
on the growth and survival data by use of Statistical
Analysis System (SAS) software (Littell et al. 1996).
The experimental design was subjected to analysis with
nine random farms as blocks and two fixed feeding
treatments (daily and alternate-day feeding methods),
each applied to one pond per farm. Whole-pond
responses (total harvest, survivorship, and FCR) were
TABLE 1.—Mean 6 SE performance parameters of pond-
reared Nile tilapia subjected to daily and alternate-day feeding
schedules after the 120-d culture period at commercial farms
in Luzon, Philippines.
Feeding schedule
Feeds $1,987.86 $865.31
Labor $323.17 $161.58
Profit per hectarea $447.50 $556.01
a
Includes deduction of approximate fixed costs.
thus analyzed in a randomized complete block design
for a situation involving no within-block replication.
Survival fractions were arcsine-square-root trans-
formed to improve normality. The individual fish
masses at harvest were treated as measurements from
random fish within fixed pond treatments within
random farms. Individual fish mass was log trans-
formed prior to analysis to improve within-pond
normality. Analyses were performed in SAS, version
9.1.3 by use of the MIXED procedure with the
CONTAIN denominator degrees of freedom used as
the default (Littell et al. 1996).
Results and Discussion
Table 1 presents growth data, survival rates, yields,
FCRs, and other performance data from the nine
participating commercial farms. The growth trends of
Nile tilapia in the daily and alternate-day feeding
schedules over 120 d of culture are shown in Figure 1.

Variable Daily Alternate day

Initial mean weight (g) 0.19 6 0.03 0.19 6 0.03

Final mean weight (g) 167.35 6 53 137.79 6 72
Total weight gain (g) 167.16 6 53 137.60 6 72
Mean daily weight gain (g/d) 1.39 6 0.44 1.15 6 0.60
Specific growth (%)a 5.58 6 0.22 5.35 6 0.31
Survival (%) 55.34 6 20 63.42 6 26
Feed conversion ratiob 2.24 6 0.73 1.00 6 0.34
Fish yield (kg) 222 6 102 200 6 88
Quantity of feeds (kg) 475 6 185 208 6 145
Extrapolated fish yield (kg/ha) 2,994 6 808 2,807 6 857
Extrapolated quantity of feeds (kg/ha) 6,331 6 1,088 2,690 6 601
a
Specific growth rates (SGRs) were calculated as SGR ¼ 100
(logeWf
- logeWo)/(tf - to), where Wo and Wf are fish weights at the beginning
(to) and end (tf) of the feeding trials. FIGURE 1.—Growth trends of pond-reared Nile tilapia
b
The difference between the two feeding treatments was significant (P subjected to daily and alternate-day feeding schedules at
, 0.05). commercial farms in Luzon, Philippines.
 NILE TILAPIA ALTERNATE-DAY FEEDING STRATEGY
FIGURE 2.—Growth of Nile tilapia subjected to daily (left-
hand boxes in each pair) and alternate-day (right-hand boxes)
feeding schedules on nine participating commercial farms in
Luzon, Philippines. Error bars indicate SDs. Dots indicate
outlying points that have been factored into the calculations.
While growth performance varied between farms, as
some farms appeared to have better fish growth than
others, the mean growth performance of fish in the
daily and alternate-day feeding groups lacked any
significant differences (P . 0.05). Although a trend
favoring a faster growth rate in the daily-fed group was
observed, the apparent difference was not statistically
significant (P . 0.05). Since diets were restricted in the
alternate-day feeding schedule, this did not translate to
differences in growth rate in a linear fashion, suggest-
ing differences in food utilization efficiency. Similarly,
the specific growth rates were comparable in the two
feeding schedules, and in some farms the specific
growth rates were even higher in the alternate-day-fed
group than in the daily-fed group.
The mean survival rates obtained were 55.35% for
the fish that were fed daily and 63.42% for those that
were fed on alternate days. Higher survival values were
reported for the experimental groups (fed on alternate
days) than for the controls at seven of the nine
participating farms. However, this apparent difference
in survival rate for the two feeding schedules was not
statistically significant (P . 0.05). We saw no
evidence to suggest that the moderate survival rates
observed in this experiment were treatment related or
a result of nutritional deficiency, since there were no
significant differences between treatments and since
the fish fed on alternate days had, if anything, a trend
favoring improved survival.
The administered amount of prepared feeds averaged
475 kg in the group that was fed daily and 208 kg in
the group fed on alternate days. Despite the use of more
than twice as much feed for daily feeding, as compared
with alternate-day feeding, the mixed-model analysis
confirmed similarities in growth and survivorship
195
FIGURE 3.—Feed conversion ratios (FCRs) of Nile tilapia,
depicted in pairs from the mixed-model analysis. Ponds
subjected to daily (dots) and alternate-day (triangles) feeding
schedules are depicted for each of the nine participating
commercial farms in Luzon, Philippines.
between the two treatments. Log-transformed final
weights of individual fish did not differ significantly by
treatments within farms (F ¼ 2.12; df ¼ 1, 16; P ¼
0.165). Similarly, neither arcsine–square-root-trans-
formed survivorship (F ¼ 0.78; df ¼ 1, 16; P ¼
0.390) nor total harvest (F ¼ 0.23; df ¼ 1, 16; P ¼
0.641) differed significantly between feeding treat-
ments within farms. Total harvest did not correlate with
feed input either within or between feeding treatments
(Figure 2), but FCRs differed significantly between the
animals fed daily and those fed on alternate days (F ¼
21.6; df ¼ 1, 16; P ¼ 0.003; Table 1, Figure 3).
We do not have a means of quantifying the
consumption of plankton by Nile tilapia, although
planktonic grazing can be an important component of
the diet among tilapia species. Grazing undoubtedly
contributed to some of the very low FCRs seen in this
study. Past studies have shown that tilapia grown in
fertilized ponds can be produced at a reduced cost by
initiating the addition of prepared feeds relatively later
in the grow-out cycle (Diana et al. 1994, 1996).
Fertilizer alone promotes algae and plankton blooms,
which support growth that is equivalent to feeding the
tilapia for about the first 30 d of production (Green
1992). In the present study, the amount of prepared
feeds consumed was significantly higher in the daily
feeding schedule (P , 0.05), although this was not
reflected in a corresponding increase in growth in the
daily-fed fish. The FCRs differed greatly between the
two feeding strategies. A significantly better FCR (P ,
0.05) was obtained for fish fed on alternate days than
for fish fed daily (Table 1; Figure 3). These results are
distinctly different from reports of the effects of
alternate-day feeding practices and their consequences
in channel catfish Ictalurus punctatus. Experimental
feeding on alternate days has reportedly increased food
consumption in this species; channel catfish fed on
alternate days actually consumed more food than those
fed once or more per day (Lovell 1980). It was
196 BOLIVAR ET AL.
consequently recommended that channel catfish should
be fed once or twice daily (Lovell 1980). Other studies
have indicated that channel catfish on restricted or
semi-restricted diets, including alternate-day feeding,
grow more efficiently in some cases. Feed restriction in
this species has also resulted in reduced growth, but
seasonal changes are also probably a factor in this
report (Bosworth and Wolters 2005). Additional
research on alternate-day feeding in channel catfish
attributes some differences in fingerling yield and
survival—but not growth—to feeding restrictions; this
approach may have utility to some farmers (K. Engle,
University of Arkansas, personal communication).
Some other species, such as red drum Sciaenops
ocellatus, are commonly fed on alternate days (Worts
2004).
Our partial cost–benefit analysis considered reve-
nues from gross sales and variable production costs,
including feeds and labor, under the two feeding
regimes. Taking only variable expenses and revenues
into account, the value of the alternate-day feeding
approach becomes evident (Table 2). Considering also
the estimated fixed costs of fuel and fingerlings, among
other things, we project that the daily feeding schedule
would be slightly unprofitable (a loss of US$448/ha),1
while the alternate-day feeding schedule would yield
a positive net return of $556/ha (Table 2). Cutting the
cost of feed in half without significantly diminishing
production has obvious and profound potential for
application. The results also suggest that the fish fed on
alternate days were healthy and that neither growth nor
survival was in any way compromised. Obviously,
farmers have no use for a feeding paradigm that
weakens fish or makes them immunocompromised.
Certainly, the Nile tilapia evolved with lower feed
requirements than even the most minimalistic of the
farm-scale production feeding paradigms that we have
tested.
Philippine farmers that have expressed an interest in
the alternate-day schedule also expressed hesitance
about the possible consequences because of the lack of
an appropriate study of this feeding practice. In the
current study, feeding on alternate days led to a cost
savings of more than 50% and reduction in fish
nutrition by half; however, the lower feed bill was not
associated with diminished production. We do not have
a means of ascertaining the physiological basis for the
similar growth rates and production values of the
control and experimental fish, but differences in FCR
lead us to suspect that reduced feeding led to less waste
1 Brown, C. L., R. B. Bolivar, and E. B. T. Jimenez. 2004.
Values were calculated in terms of Philippine pesos and
converted to U.S. dollars via the exchange rate at the time of
this study, 55.699 pesos/dollar.
and perhaps to improved nutrient absorption. It is also
possible that reduced feed waste led to improved water
quality and reduced exposure of fish to pathogens.
Our investigations to date have provided three clear
options for reducing feed bills without negatively
impacting production. Delaying the onset of feeding in
fertilized ponds (Brown et al. 2000), feeding daily but
at a reduced rate (Bolivar et al. 2004), or feeding on
alternate days (present study) can be done without
diminishing the harvest. Experimentation to date does
suggest that there are limits to the degree that feeding
can be reduced without compromising crop value, as
a combination of the first two feed reduction strategies
proved less profitable than either strategy alone
(Bolivar et al., unpublished).
Acknowledgments
We gratefully acknowledge the support of the
farmer–collaborators who willingly made their pond
facilities available for this study. We also thank Raul
Moncarz for his assistance with the economic analysis.
This work was a component of the Aquaculture
Collaborative Research Support Program (ACRSP),
supported by U.S. Agency for International Develop-
ment (USAID) grant number LAG-G-00–96–90015–
00 and by contributions from the participating
institutions. The ACRSP accession number is 1288.
The opinions expressed herein are those of the authors
and do not necessarily reflect the views of USAID.
References
Bolivar, R. B., E. B. T. Jimenez, J. R. Sugue, and C. L.
Brown. 2004. Effect of stocking size on the yield and
survival of Nile tilapia (Oreochromis niloticus L.) on-
grown in ponds. Pages 574–583 in R. Bolivar, G. Mair,
and K. Fitzsimmons editors. New dimensions in farmed
tilapia: proceedings from the Sixth International Sym-
posium on Tilapia in Aquaculture. Philippines Bureau
of Fisheries and Aquaculture Research and American
Tilapia Association, Manila.
Bosworth, B. G., and W. Wolters. 2005. Effects of short-term
feed restriction on production, processing, and body
shape traits in market-weight channel catfish, Ictalurus
punctatus (Rafinesque). Aquaculture Research 36:344–
351.
Brown, C. L., R. B. Bolivar, E. B. T. Jimenez, and J. P.
Szyper. 2000. Timing of the onset of supplemental
feeding of Nile tilapia (Oreochromis niloticus) in ponds.
Pages 237–240 in K. Fitzsimmons and J. C. Filho editors.
Tilapia aquaculture in the 21st century: proceedings from
the Fifth International Symposium on Tilapia in
Aquaculture. Philippines Bureau of Fisheries and Aqua-
culture Research and American Tilapia Association,
Manila.
Philippine studies support moderate feeding in tilapia.
Global Aquaculture Alliance Advocate 7:70.
 NILE TILAPIA ALTERNATE-DAY FEEDING STRATEGY
Cacho, O. J. 1993. Development and implementation of a fish
farm bioeconomic model: a three-stage approach. Pages
55–74 in U. Hatch and H. Kinnucan editors. Aquaculture
economics. Westview Press, Boulder, Colorado.
De Silva, S. S. 1985. Performance of Oreochromis niloticus
(L.) fry maintained on mixed feeding schedules of
differing protein content. Aquaculture Fisheries Manage-
ment 16:331–340.
De Silva, S. S. 1989. Reducing feed costs in semi-intensive
aquaculture systems in the tropics. Naga, the ICLARM
Quarterly 12(4):6–7.
De Silva, S. S., B. D. Mitchell, and M. C. Nandeesha. 1993.
Reducing feed costs in semi-intensive finfish culture: an
update on mixed feeding schedules and an idea for
enhancing endogenous food supply in ponds. Naga, the
ICLARM Quarterly 16(1):24–26.
Diana, J. S., C. K. Lin, and K. Jaiyen. 1994. Supplemental
feeding of tilapia in fertilized ponds. Journal of the World
Aquaculture Society 25:497–506.
Diana, J. S., C. K. Lin, and Y. Yin. 1996. Timing of
supplemental feeding for tilapia production. Journal of
the World Aquaculture Society 27(4):410–419.
Diana, J. S. 1997. Feeding strategies. Pages 245–262 in H. S.
Egna and C. E. Boyd editors. Dynamics of pond
aquaculture. CRC Press, Boca Raton, Florida.
Green, B. W. 1992. Substitution of organic manure for
197
pelleted feed in tilapia production. Aquaculture 101:213–
222.
Hatch, U., and K. Kinnucan. 1993. Introduction. Pages 1–16
in U. Hatch and H. Kinnucan editors. Aquaculture
economics. Westview Press, Boulder, Colorado.
Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D.
Wolfinger. 1996. SAS system for mixed models. SAS
Institute, Inc., Cary, North Carolina.
Lovell, R. T. 1980. Practical fish diets. Food and Agriculture
Organization of the United Nations, ADCP/REP/80/11,
Rome.
Singh, R. P., and A. K. Srivastava. 1984. Effect of feeding
frequency on the growth, consumption, and gross
conversion efficiency in the siluroid catfish, Hetero-
pneustes fossilis (Bloch). Bamidgeh 36:80–91.
Srikanth, G. K., M. S. Nandeesha, P. Keshavanath, T. J.
Varghese, H. P. C. Shetty, and N. Basavaraja. 1989. On
the applicability of mixed feeding schedules for common
carp, Cyprinus carpio var. communis. Pages 254–260 in
E. A. Huisman, P. Zonneveld, and A. H. M. Boumans
editors. Aquaculture research in Asia: management
techniques and nutrition. Center for Agricultural Pub-
lishing, Pudoc, Wageningen, The Netherlands.
Worts, W. A. 2004. Organic fertilization in culture ponds.
World Aquaculture 35(2):64–65.