TABLE 1. Experimental design for fry production; 2 replicates per treatment.


 Treatment 1 2 3

 No. of brood fish/hapa 10 20 30
 Brood fish density
 (#/a hapa) 2.7 5.4 8.1
 No. of females/hapa 8.5(8) 14.5(16) 23.5(24)
 (desired no.)
 No. of males/hapa 1.5(2) 5.5(4) 6.5(6)
 (desired no.)
 Sex ratio ( feaale:male) 5.7 2.6 3.6
 Initial mean wt. of females(g) 133 138 158
 Initial mean wt. of males(g) 170 187 204





 TABLE 2. Stocking and harvest dates of replications of 6 stocking densities of
 tilapia fry fed for 63 days.


 Stocking Density Replication Stocking Harvest
 (#/m ) Date Date

 26 1 11 Jan 17 Mar
 2 11 Jan 17 Mar
 3 25 Jan 31 Mar

 52 1 28 Dec 2 Mar
 2 11 Jan 17 Mar
 3 25 Jan 1 Apr

 78 1 28 Dec 2 Mar
 2 11 Jan 18 Mar
 3 8 Feb 13 Apr

 104 1 28 Dec 2 Mar
 2 25 Jan 1 Apr
 3 8 Feb 13 Apr

 130 1 11 Jan 16 Mar
 2 25 Jan 30 Mar
 3 25 Jan 31 Mar

 155 1 11 Jan 17 Mar
 2 25 Jan 30 Mar
 3 25 Jan 31 Mar






The daily feeding rate, which was based on the modification of a
rate used by Snow et al. (1983), was 15% of initial body weight
for three weeks, followed by weekly adjustments to 5% of body
weight based on biweekly sampling. After 63 days, the finger-
lings were harvested, sorted into centimeter-length groups,
weighed and counted.
 All of the tanks were filled with well water. There was no fer-
tilization, aeration or water exchange during the 9-week experi-
mental period. Several water quality variables were measured
weekly using standard methods (APHA et al., 1980). Samples
were collected early in the morning. Dissolved oxygen (DO) and
temperature were measured with a YSI Model 51-A
polarographic oxygen meter. A glass electrode was used to
measure pH. Total alkalinity was measured by titration with stan-
dard acid to the methyl red-bromcresol green end point. Levels of
total ammonia-nitrogen and nitrite-nitrogen were determined by
the phenate method and diazotization method, respectively.
Chlorophyll a was measured by using the acetone extraction
method (Vollenweider, 1969).


RESULTS AND DISCUSSION


Fry Production
 Fry production was extremely variable (Table 3). For example,
 harvests from individual replicates of Treatment 3 (high density)
 ranged from 0 to 4,947 fry. Hughes and Behrends (1983) ob-
 tained similar variability with T. nilotica and attribute this to
 variability in fecundity, differences in spawning frequency, and
 the relative asynchrony of spawning cycles of individual females.
 Fry production increased with an increase in density of brood
fish (Table 4). Production in Treatment 3 was 23.8 fry/m2 of
hapa/day, or nearly twice the production of Treatment 1 (low
density). It appears that brood fish density of at least 8.1/m2
would be necessary to maximize fry production of T. area.
Hughes and Benrends (1983) reported that fry production of T.
nilotica was greater at a brood fish density of 5/m2 than at 10/m2.
Uchida and King (1962) found that fry production of T. mossam-
bica peaked at a brood fish density of 11.2/m2.
 As the density of brood fish increased, fry production per
 female decreased (Table 4). Production of fry/female/day was
 5.3 for Treatment 1 compared to 3.8 for Treatment 3. A similar



 TABLE 3. Mean number ( standard error) of tilapia fry collected from hapas
 during 7 harvests at 2-week intervals. Each harvest is the mean of 2 replications.


 Harvest 1 2 3


 1 498 87 590
 (+493) (+43) (+590)


 2 960 1040 357
 (+544) (+.01) (+357)


 3 766 292 2618
 (+766) (+222) (+2330)


 4 565 1421 908
 (+556) (+345) (+ 136)


 5 1476 2248 2020
 (+354) (+886) (+ 284)


 6 1050 660 3426
 (+ 432) (+660) (+1248)


 7 0 810 445
 (+0) (+310) (+ 268)

 Mean 759 937 1480
 (+184) (+228) (+ 425)

 Total 5316 6557 10,362






TABLE 4. Tilapia fry production in hapas during a 117-day period. Each value is
the mean of 2 replicates.

Treatment 1 2 3

Fry/m2 hapa/day 12.2 15.1 23.8

IFry/ female/day 5.3 3.9 3.8

)ry/ g female/day1 0.03 0.02 0.02

 Based on the average of initial and final weights of females.


VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


237