twice daily at 0800 and 1500 h, 7 days a week, for a period of 65 days. Waste feed was minimised by observing fish during feeding and when fish lost their appetite feeding
Ž w
x .
was ceased. Fish were then harvested February late summer 1996 and survival mean Ž
. weight increment and feed conversion ratio FCR were calculated from each tank.
Proximate analyses on the whole body composition of five fish randomly selected Ž
. from each tank were also conducted and the indices for protein deposition PD , fat
Ž .
Ž .
Ž .
deposition FD , protein efficiency ratio PER , protein retention efficiency PRE , and Ž
. energy retention efficiency ERE were calculated.
2.6. Statistical analysis Ž
. The digestibility and availability experiment Experiment 1 was designed for analy-
Ž .
sis using two-factor ANOVA with meat products B, L, M or P as the first factor Ž
. Ž
. Ž
. fixed and inclusion level 15 or 30 as the second factor also fixed . Single factor
ANOVA was used to assess the difference between proximate body composition of fish fed experimental diets.
Ž .
The growth experiment Experiment 2 was designed for analysis using single-factor Ž
. ANOVA. Homogeneity of variances was assessed using Cochran’s Test Winer, 1971
and comparison between means were made using Student–Newman–Kuels multiple- range test. Differences between means were considered significant at P - 0.05. Unless
Ž .
otherwise stated, all results appear as mean standard error of mean n s 3 .
3. Results
3.1. Experiment 1, digestibility and aÕailability The analysed proximate composition of meat meals tested in this study is presented in
Table 1. The difference between the sum of the percentages of protein, fat and ash, and 100 for these meals indicates the presence of some nitrogen-free material. This may
Ž .
indicate contamination with material e.g., carbohydrate from the rumen or another extraneous source. When compared to Peruvian fish meal, the beef and lamb meals had
less protein, fat and energy; the mixed meat meal had less protein and more fat and energy; and Provine
w
had more protein and energy. Provine
w
and the mixed meat meal had less ash than the Peruvian fish meal, while the beef meal and lamb meal had more.
The amino acid content of the tested ingredients also differed and lysine was lower in all meat products than in Peruvian fish meal.
Digestibility and availability coefficients for ingredients were calculated using the Ž
. values for the reference diet and the proportion of the ingredients used Table 4 . There
Ž .
was no interaction between ingredient or inclusion level for any nutrient P 0.05 .
There were significant differences between ingredients for all nutrients except methion- Ž
. ine P - 0.05 , but significant differences between inclusion levels only occurred for six
Ž .
amino acids alanine, arginine, glycine, histidine, lysine and valine . The average
Ž .
availability coefficients n s 12 for each of these six amino acids for all ingredients included at 15 and 30, respectively, were: alanine 77.1 2.6, 79.5 2.4; arginine
Table 4 Apparent digestibility and availability coefficients of ingredients for silver perch in Experiment 1
1,2,3
Ingredient Pooled SEM
w
Beef Lamb
Mixed Provine
Nutrient
a b
c d
Dry matter 42.6
54.9 75.7
88.9 3.9
a b
b c
Energy 71.4
80.8 81.9
89.9 1.7
a b
c c
Protein 69.7
73.4 83.7
85.5 1.4
Amino acids
U a
b c
c
Ala 68.3
72.6 85.9
86.4 1.7
U a
b c
c
Arg 72.0
76.0 87.6
87.6 1.5
a b
b b
Asp 77.1
84.6 86.5
85.4 1.1
a a
a a,b
Cys 71.7
75.0 70.4
86.1 1.9
a b
c c
Glu 75.3
81.6 86.1
87.1 1.1
U a
a b
b
Gly 65.2
64.5 87.8
90.2 2.6
U a
b b
b
His 72.0
87.0 82.9
84.5 1.4
a b
b b
Ile 73.2
83.8 79.6
80.5 1.0
a b
b b
Leu 77.0
84.4 83.8
82.4 0.8
U a
b b
b
Lys 73.6
83.4 83.8
84.2 1.0
Met 82.0
83.8 83.5
83.3 0.5
a b
b b
Phe 73.4
82.2 83.3
81.6 1.0
a a
b b
Pro 68.0
67.5 86.8
89.0 2.2
a b
c c
Ser 72.6
76.8 85.9
84.4 1.3
a b
b b
Thr 74.7
83.3 84.7
84.3 1.0
a b
b b
Tyr 80.2
84.8 85.0
83.6 0.6
U a
b b
b
Val 74.6
80.5 79.9
80.1 1.0
Average 73.5
79.6 83.8
84.8
1
Ž .
Apparent digestibility or availability coefficients for test ingredients were calculated using the equation: Ž
. Ž
Apparent digestibility or availability of nutrient in test ingredients nutrient or energy concentration in test Ž
. diet=apparent digestibility
or availability of nutrient or energy in test dietyproportion of reference
Ž .
diet=nutrient or energy in reference diet=apparent digestibility or availability of nutrient or energy in the . Ž
. reference diet r proportion of test ingredient in test diet=nutrient or energy concentration in test ingredient .
2
Ž .
Values are meanspooled SEM; ns6 average of both inclusion levels . There was no interaction Ž
. P 0.05 between ingredients and inclusion level for any nutrient or amino acid. Means in the same row
with the same letter in the superscript indicate the difference between ingredients was not significant Ž
. P 0.05 .
3
Inclusion content did not affect digestibility or availability coefficients of any nutrient or those amino acids not marked with an asterisk. For those amino acids marked with an asterisk, availability coefficients
Ž .
were lower when ingredients were included at 15 compared with P - 0.05 .
77.6 2.2, 82.0 2.1; glycine 75.6 3.8, 78.2 3.6; histidine 79.6 2.3, 83.5 1.6; lysine 80.0 1.7, 82.5 1.2; and valine 77.8 1.5, 80.6 1.2.
The differences in amino acid availability between the 15 and 30 inclusion levels Ž
. were small average 3.1, range 2.4–4.4 . In every case, the standard error was larger
for the amino acid availability at the 15 inclusion level compared to the 30 inclusion level. The differences may be attributed to the more variable results obtained from using
the lower inclusion level and may be biologically insignificant.
For dry matter, energy and protein digestibility coefficients increased significantly Ž
.
w
with protein content P - 0.05 ; Provine was the highest, followed by the mixed meal,
D.A.J. Stone
et al.
r Aquaculture
186 2000
311 –
326
319 Table 5
Ž .
a
Growth performance, survival, feed utilisation, and carcass composition of silver perch after a 65-day feeding trial Experiment 2 Ž
. Diet fishmeal
b
Ž .
Ž .
Ž . Ž .
Ž . 1 27
2 13 3 6
4 0 5 0
Ž . Initial weight g
11.900.16 12.200.02
12.100.09 12.200.08
12.100.10
a a
a,b b
b
Ž . Weight increment g
60.402.29 60.201.01
53.902.30 52.301.75
50.001.47 Ž .
Survival 99.200.40
100.000.00 99.600.40
99.600.40 100.000.00
y1 b
b a
a a
Ž .
Total feed intake kg tank 7.610.26
7.640.19 6.430.41
6.660.14 6.230.13
c
Feed conversion ratio 1.480.01
1.440.02 1.460.02
1.500.02 1.470.02
d a
a,b a,b
b a,b
Protein efficiency ratio 2.100.02
2.040.02 2.010.03
1.960.03 2.010.03
y1 e
a a
b b
b
Ž .
Protein deposition g fish dry weight
10.540.30 10.370.38
9.370.19 8.800.20
8.980.33
f
Ž . Protein retention efficiency
36.700.26 35.180.71
34.961.18 32.930.72
36.020.63
y1 g
Ž .
Fat deposition g fish dry weight
11.020.76 11.370.38
10.211.20 10.140.86
9.420.23
h
Ž . Energy retention efficiency
65.320.79 65.801.82
65.972.05 64.942.38
65.990.76
i d
b c
b a
Ž .
Feed cost AUSrkg fish 1.120.01
1.010.01 1.060.02
0.980.01 0.890.01
Carcass composition dry weight basis of whole fish Initial
Moisture 67.37
59.670.28 60.000.33
59.610.48 59.990.44
59.580.04 Crude protein
42.63 41.880.52
41.650.91 41.561.34
40.711.42 42.440.40
Crude fat 31.37
41.911.24 43.540.88
42.712.46 43.951.66
42.440.40 Ash
15.40 10.610.39
10.760.03 11.080.52
10.900.39 10.790.13
a
Ž .
Values are meansSEM for three replicate tanks. Means in rows which share the same superscript were not significantly different P 0.05; ANOVA; SNK .
b
No added crystalline amino acids.
c
w x
FCR s weight of food, adjusted to 92 dry matterrwet weight fish gain .
d
w Ž
. Ž
. .
Ž Ž
. Ž
. .x
PDs final weight dry basis =final protein content dry basis r100 y initial weight dry basis =initial protein content dry basis r100 .
e
w Ž
. Ž
. .
Ž Ž
. Ž
. .x
FDs final weight dry basis =final fat content dry basis r100 y initial weight dry basis =initial fat content dry basis r100 .
f
w Ž .
Ž .x
PER s individual weight gain g rindividual protein intake by fish g dry weight .
g
wŽŽ .
Ž ..
x PRE s
final dry weight=final dry weight body protein y initial dry weight=initial dry weight body protein rdry weight protein intake=100 .
h
wŽŽ .
Ž ..
x ERE s
final dry weight=final body energy y initial dry weight=initial body energy rdry weight energy intake=100 .
i
Ž . Ž
. Feed cost s the cost of ingredients to produce 1 kg fish FCR=cost of ingredients
excludes processing, handling, and transport costs; see Table 3 .
lamb meal then beef meal. Significant differences are indicated in Table 4. The availability of all amino acids were averaged and the values were 84.8 for Provine
w
, 83.8 for the mixed meat meal, 79.6 for lamb meal, and 73.5 for beef meal.
Availability coefficients for non-essential amino acids, alanine, glycine, proline and serine were significantly higher for the mixed meat meal and Provine
w
compared with other products.
3.2. Experiment 2, growth Ž
. Diet had no significant effect on survival, FCR, PRE, FD, or ERE
P 0.05 . Ž
. However, there was a significant diet effect P - 0.05 on weight increment, PER and
Ž .
PD, and all of these indices were lower for diets without fish meal Table 5 . Ž
. There were no significant differences P 0.05 in the moisture, crude protein, crude
fat, and ash composition of silver perch carcasses fed different experimental diets for 65 Ž
. Ž
. days Table 5 . However, there was a difference P 0.05 between the initial and final
moisture, crude fat, and ash carcass composition for fish from each of the experimental diets. Carcass protein composition remained unchanged.
4. Discussion
