. .

. ....
.........: u
awnﬁf. a. . .3
1’»- v

E .

. V
1'
.9 a. win»: .

1-5 rgaﬁaﬁ
mm a. .
eiﬂhuu: . ‘

(.1

Th.
v.3: ,
3:: In .3 . V
«n :51 a. A» E
.31! 3h. .
n k H}. .
. .21: .g: 1....L‘ .
, t . .. v
{Innatﬂmvt . E. t
‘ . ‘ I £32: ,
. I .
, ‘ . . . a
. 1.... 1...... 1
....F.i....l
5...}...

o r .—’p"aan.1
. E
. V. . ’0. I" ”r
. , ﬂﬁgit 2..
:3me .. ‘ .
s
.1.

2:12. ..
3:...{3151 ,!
O - .4

‘ girl. .Wuﬂ

. vb-nuujlg .

.5 9...; hr} . h.
V‘lvi‘...‘

.1) «En... a: .
€an... .13.“.ﬁrﬁz‘uﬂ? n. 51.... .
. 0. ..r...uznxvﬁw.mﬁna ..r
c I.» . , 1 Ilhﬁuvalhuur pi). {Lars
“EA 2...... .L. ..| £34.53 1 351...}?! I:
r .. v1. :. I in.» 2.2553,. '41:! c.
. in: 3...“. if?
0.. . it a .3 .23
355,91... . id». .5 $3 , ‘ . b!
E. 2.....594. 3.1.. .. ! ahﬁha
fl. 41.xf1. 2993-1;
! ﬁz'302’ta- ’-
3.21:2.1}...
51.2.:

An...
it:
<0}...
than» C
U331
3.5:...‘1. flux- _
{3,7,3 sun” (LL
.{2..{€J~.1.lnlr :89;
1.19.2.5... 5:11....
I .
3.. 0|: ”.7.
1
q 4.29.: .3
I . q .9 .31!
VL’I...%:14\} 13:
:{1}

«4:4
3. “5‘"? 1|]ch 3.". cm”!!!
:‘uauricnxid‘humvad
ﬁrm...“ .... t

 

.li...7!.!.!. V).

J. . .
.ftiiltgii.

6. .7:
. l 334‘;
{If}? '15;

..':.d..n1lt< »_lv:5l\’)~tl65b—I.
1.. 11:12} x...ll6...5a {(50:11

3... . 1. 0
”11024. I 13".

 

 

.ﬂuuw

 

: J . a»: 5%. u; ..
85:: t .319:
. iii:
$9.5 . K. .HL. n . ‘ ‘ ‘

 

Illlllllll||||||illlmllllllllllillllUlllllllllllllllllllll

3 1293 02048 6381

LIBRARY

Mlchlgan State
Universlty

 

 

 

 

This is to certify that the

dissertation entitled

Effect of Phytic Acid on Nitrogen Retention in
Tilapia (Oreochromis niloticus)

 

presented by

Martin Alan Riche

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degmm Fish. & w11a1.

 

 

 

aJor professor

Dacia 23 , 2000

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0- 12771

 

PLACE IN RETURN BOX to remove this checkout from your record. ’
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EFFECT OF PHYTIC ACID ON NITROGEN RETENTION IN TILAPIA
(OREOCHROMS NILOTICUS)

By

Martin Alan Riche

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Fisheries and Wildlife

2000

 

ABSTRACT
EFFECT OF PHYTIC ACID ON NITROGEN RETENTION IN TILAPIA
(OREOCHROMS NILO TIC US)
By

Martin Alan Riche

Culturing tilapia in temperate regions requires the use of intensive recirculating
systems and formulated diets. Diets used typically contain ﬁsh meal (FM). The expense
associated with feeding FM based diets has resulted in efforts to ﬁnd alternative proteins
such as soybean meal (SBM). However, replacement of FM with SBM has met with
mixed success. SBM contains anti-nutritional factors (ANF) that reduce its biological
value. One such ANF is phytic acid (PA). PA has the potential to reduce availability of
amino acids. PA can be removed from SBM with the enzyme phytase at considerable
expense. A series of experiments were conducted to determine the effect of PA on
growth and efficiency in Nile tilapia, and whether phytase treatment is warranted for
increasing nitrogen (N) retention.

Two preliminary studies were conducted to determine the optimal feeding
frequency and interval between feedings to maximize growth and efﬁciency. Growth,
whole body proximate composition, and efﬁciency parameters were evaluated in ﬁsh fed
1, 2, 3, or 5 feedings day '1 to apparent satiation. Three feedings day '1 led to optimal

growth and efﬁciency.

Fish eat food at intervals determined by rate of gastric emptying. In a second
study, gastric evacuation rate (GER) in ﬁsh fed 3 or 5 meals day '1 was determined by
following movement of feed containing colored dyes over time. Equations describing
GER were VT = 67.0 e OHM), and VT = 85.0 e 01490;) for ﬁsh fed 3, and 5 meals day ’1,
respectively. The optimal interval between feedings was 4 — 5 hours.

A series of four experiments were conducted to determine the effect of PA on
growth and N retention. An eight week grth study was conducted with phytase
treated, or untreated SBM substituted into a FM based diet at 0, 25, 50, 75, or 100 % of
the crude protein (CP). This was followed with a study utilizing the same diets to
determine CP and amino acid digestibility. Untreated SBM can replace 75 % of the CP,
and phytase treated SBM 25 % of the CP, without signiﬁcantly depressing growth and
efficiency. Growth models suggest restricting untreated SBM, and phytase treated SBM,
to 30 %, and 15 % of the dry diet, respectively. Lysine (Lys) and methionine (Met)
became limiting with increasing levels of SBM, regardless of treatment. Availability of
Lys and Met from phytase treated SBM appears to be responsible for reduction in
performance in ﬁsh fed phytase treated SBM beyond 25 % of the CP.

Tilapia were fed FM based diets supplemented with graded levels of PA during an
eight week grth study, and a digestibility study. Growth, efficiency, and digestibility
were independent of PA concentration. PA removal from SBM with phytase is not
efﬁcacious or warranted for increasing N retention in tilapia.

Measurements of pH were taken of the food and mucosa of the stomach and
intestine before feeding and at 0.5, 1, 2, 4, 6 and 8 hours postprandially. The pH values

in the gastric region were not as low as previously reported.

This body of work is dedicated to the memory of my mother, Shirley Anne
Casavant; my father, Walter Thomas Riche, Jr.; and my stepfather Jack Edward
Casavant; all of whom passed away after its inception, but before its completion
Although they were unable to see its culmination, I know they were with me in spirit,

even to the last word.

iv

ACKNOWLEDGMENTS

I wish to extend my deepest and sincerest gratitude to Dr. Donald Garling who
had the faith and patience to guide me through this program and this research. He has
been a mentor and true friend while providing me with the encouragement and
opportunities to grow and develop both personally and professionally. I would also like
to express my appreciation to the other members of my graduate committee Dr. Nathalie
Trottier, Dr. Weiming Li, and Dr. Michael Allen for their helpful advice, without which
this dissertation would not be a reality.

I give a big thank you to my dear friends Dave Haley, Betsy Haley, Jon Amberg,
and Mike Oetker for the many hours of their tireless assistance in collecting and
analyzing samples. I acknowledge Dr. Pao Ku for his technical assistance and advice
both of which, without, I could not have ﬁnished. I also acknowledge the assistance of
Patty LeChance, Tiffani Smith, and Sarah Garbrecht.

To my beautiful new bride I say thank you for the love, support, and endless
hours of work you have given me. I promise to pay you back in full for the patience,
kindness, and support you have displayed through this process. Without you I would not
have made it.

Lastly I express my appreciation to Purdue University, Illinois State University,
Zeeland Farm Services, BASF, and Zapata Proteins for their donations. Thanks also to
the CANR for providing me the ﬁnancial means to complete this dissertation. This
research was supported in part by the North Central Regional Aquaculture Center and the

Michigan Soybean Promotion Committee.

TABLE OF CONTENTS

LISTOFABBREVIATIONS

HYPOTHESIS STATEMENTS... .
Experiment 1 - Optimal Feeding Frequency
Experiment 2— Tilapia Gastric Evacuation Rate.
Experiment 3— Solvent Extracted SBM Grth Trial
Experiment 4- Phytic Acid Supplemented Fish Meal Growth TriaJIII
Experiment 5— Solvent Extracted SBM D1 gestrbrlrty Trial...

Experiment 6— -Phytic Acid Supplemented Fish Meal D1gest1brl1ty Tnal II]].

Experiment 7— Tilapia Gastrointestinal Tract pH Proﬁle...

NIETHODS AND MATERIALS...

Experiment 1 -OptimalFeed1ngFrequency ..

Experimental Design...

Proximate Analysis...

Calculations...

Statistical Analysis .
Experiment 2- Tilapia Gastric Evacuation Rate:

Experimental Diet Preparation.

Experimental Design...... I..I....

Iron Analysis...
Statistical Analysis...

Experiment 3— Solvent Extracted SEM Growth TriaIII.
Solvent Extracted SBM Diet Preparation.
ExperimentalDesign...
Phytate Analysis... .

Trypsin Inhibitor Actrvrty
Proximate Analysis...
Efﬁciency CalculationsI].
Statistical Analysis...
Experiment 4- Phytic Acid Supplemented Fish Meal Growth Trial.
Phytic Acid Supplemented Fish Meal Diets...

Experimental Design............

vi

.viii

.. .xii

.xvii

Experiment 5— Solvent Extracted SBM Digestibility Trial...
ExperimentalDesign... .
HROM and Crude ProteinII
Amino Acid Analysis...

Statistical Analysis...

Experiment 6— Phytic Acid Supplemented Fish Meal D1gest1b1lrty Trial” ..
.....41
.........41
. ......41
...44

Experimental Design..
Statistical Analysis.

Experiment 7_ Tilapia GaStl'OlnteSItlnal TraI’ICIt 'i'J'I-I‘ IIQIIOIﬁHIIe: . . . . . . . . . . . . . . . . . .

Experimental Design...
Statistical Analysis...

RESULTS... . .
Experiment I— OptIm—al Feedmg Frequency
Experiment 2— Tilapia Gastric Evacuation Rate:
Experiment 3- Solvent Extracted SBM Growth TrialI.
Experiment 4— Phytic Acid Supplemented Fish Meal Growth Trial...
Experiment 5-— Solvent Extracted SBM D1gest1brl1ty Trial...

Experiment 7— Tilapia Gastrointestinal Tract pH Proﬁle...

SUNIMARYAND CONCLUSIONS......

APPENDIX 1

RECIRCULATING SYSTEMDESIGN

APPENDIX 2
PROCESSING STEPS FOR SOLVENT EXTRACTED SOYBEAN MEAL

OBTAINED FROM ZEELAND FARM SERVICES, ZEELAND, MI... ..

APPENDIX 3

APPENDIX 4

APPENDIX 5

APPENDIX 6

vii

......37
.. ...37
......39
....39

...61
...72
......76
Experiment 6— Phytic Acid Supplemented Fish Meal D1gestrbrlrty Trial... ... ... .

....87

.143

..148

.161

.162

.163

164

.165

.166

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

LIST OF TABLES

Composition of experimental diets fed to tilapia (Oreochromis niloticus)
to determine gastric evacuation rate and intestinal motility.

Composition of solvent extracted SBM experimental diets used in grow
out and digestibility trials. Substitution of SBM into the herring meal
control diet provided 25, 50, 75, and 100 % of the crude protein. Values
shown are g/kg of the dry diet.

Mineral premix composition and dietary mineral supplementation in both
the soybean meal and phytic acid supplemented experimental diets fed to
tilapia.

Composition of phytic acid supplemented experimental diets used in grow
out and digestibility trials. Phytic acid incorporation was equivalent to
phytic acid concentrations in diets that contain SBM providing 25, 50, 75,
100, and 200 % of the crude protein. Also shown is composition of the
N-free diet fed to correct apparent digestibility coefﬁcients to true
digestibility coefﬁcients. Values shown are g/kg of the dry diet.

Phytic acid concentrations (% DM) in diets incorporating graded levels of
solvent extracted SBM substituted into a herring meal control diet on a
percent protein basis. Diets contained either untreated or phytase treated
SBM. Also shown is the Na-phytate equivalent incorporated into the
hening meal control/phytic acid supplemented diets.

Initial and ﬁnal whole body proximate composition of 0. niloticus fed a
stande commercial trout diet to apparent satiation l, 2, 3, or 5 times

day " during a four-week growth trial. All components are reported on a
dry matter basis. Values represent the mean and (SEM) of triplicate
assays on pooled samples (n=5) from three replicate treatments. Initial
values represent the mean and (SEM) of triplicate assays on pooled
samples (n=5). Different superscripts within a column indicate signiﬁcant
differences (P<0.05).

Weight gain, speciﬁc growth rate (SGR), feed efﬁciency (FE), protein
efﬁciency ratio (PER), energy retention (ER), and apparent net protein
utilization (ANPU) in 0. niloticus fed to satiation 1, 2, 3, or 5 times day "
during a four-week growth trial. Values represent the mean and (SEM) of
pooled samples (n=5) from three replicate treatments. Different
superscripts within a column indicate signiﬁcant differences (P<0.05).

viii

Table 8.

Table 9.

Table 10.

Table 11.

Table 12.

Table 13.

Partitioned ANOVA tables and parameter estimates for the multivariable
regressions examining (A) visceral weight as a function of total wet
weight, gender, and number of daily feedings; and (B) perivisceral weight
as a function of visceral weight, gender, and number of daily feedings.
Number of feedings evaluated were one, two, three, or ﬁve times a day.

Weight gain, whole body crude protein, whole body lipid, and moisture in
juvenile tilapia fed diets containing untreated, or phytase treated, solvent
extracted SBM substituted into a reference diet at 25, 50, 75, or 100 % of
the dietary protein. Values represent the mean of three replicates of
pooled samples (n=8). Different superscripts in a column, within the
phytase treatment, or within the untreated treatment, represent signiﬁcant
differences within that treatment (P<0.05). Asterisks represent signiﬁcant
differences from the control diet (P<0.05).

Speciﬁc growth rate (SGR), feed efﬁciency (FE), protein efﬁciency ratio
(PER), and apparent net protein utilization (ANPU) in juvenile tilapia fed
diets containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean of three replicates of pooled samples
(n=8). Diﬁ‘erent superscripts in a column, within the phytase treatment, or
within the untreated treatment, represent signiﬁcant differences (P<0.05).
Asterisks represent signiﬁcant differences from the control diet (P<0.05).

Weight gain, whole body crude protein, whole body lipid, and moisture in
juvenile tilapia fed a ﬁsh meal based diet supplemented with graded levels
of phytic acid as Na-phytate. Also shown is the level of SBM
incorporation (% dietary CP) providing the equivalent phytic acid
concentration. Values represent the mean of three replicates of pooled
samples (n=8). Values within a column with different superscripts are
signiﬁcantly different (P<0.01).

Speciﬁc growth rate (SGR), feed efﬁciency (FE), protein efﬁciency ratio
(PER), and apparent net protein utilization (ANPU) in juvenile tilapia fed
a ﬁsh meal based diet supplemented with graded levels of phytic acid as
Na-phytate. Also shown is the level of SBM incorporation (% dietary CP)
providing the equivalent phytic acid concentration. Values represent the
mean of three replicates of pooled samples (n=8). Values within a column
with different superscripts are signiﬁcantly different (P<0.01).

Apparent amino acid digestibility coefﬁcients in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean and (SEM) of four replicates of pooled
samples (n=15), except where indicated. Values with different
superscripts across a row are signiﬁcantly diﬁ‘erent (P<0.05).

ix

Table 14.

Table 15.

Table 16.

Table 17.

Table 18.

Table 19.

Table 20.

Table 21.

Orthogonal contrasts between the means of apparent amino acid
digestibility coefﬁcients in tilapia fed diets containing untreated, or
phytase treated, solvent extracted SBM substituted into a reference diet at
25, 50, 75, or 100 % of the dietary protein. Signiﬁcant differences are
shown with resultant P values.

Apparent amino acid digestibility coefﬁcients in juvenile tilapia fed ﬁsh
meal based diets supplemented with graded levels of phytic acid as Na-
phytate. Values represent the mean and (SEM) of four replicates of
pooled samples (n=15). Values with different superscripts across a row
are signiﬁcantly different (P<0.05).

Range and mean pH values from feed in the stomach and gastric mucosa
in 0. niloticus before feeding, and 0.5, 1, 2, 4, 6, and 8 hours post-
prandially. Values in parentheses represent the standard deviation of eight
values collected from ﬁve ﬁsh at each sampling period. Values within a
column with different superscripts are signiﬁcantly different (P<0.0001).

Hours between feedings, predicted gastric evacuation (%) from evacuation
curves, and consumption at the ensuing meal as a percent of satiation.

Comparison of apparent digestibility coefﬁcients for indispensable and
dispensable amino acids (AA) in tilapia (0. niloticus) fed graded levels of
SBM (25, 50, and 75 % of the crude protein) substituted into a ﬁsh meal
based diet ', or a poultry meat based diet 2.

Requirements and available dietary indispensable amino acids (1AA) for
tilapia fed graded levels of untreated, or phytase treated, SBM as a percent
of dietary crude protein

Correlations between weight gain and efﬁciency parameters, and available
individual amino acids in phytase treated SBM, and untreated SBM,
experimental diets fed to tilapia. Parameters include speciﬁc growth rate
(SGR), feed efﬁciency (FE), protein efﬁciency ratio (PER), and apparent
net protein utilization (ANPU). Experimental diets incorporated SBM at
O, 25, 50, 75, and 100 % of the dietary crude protein. Shown are IAA and
associated Pearson correlation coefﬁcients for all relationships with
(P<0.10).

Dietary methionine supplied by intact protein and L—crystalline methionine
supplementation in diets incorporating graded levels of phytase treated
SBM as a percent of dietary crude protein. Also shown are the apparent
digestibility coefﬁcients for the diets, the calculated availability of
methionine from intact protein and the supplement, and measured
available methionine.

Table 22.

Table 23.

Table 24.

Table 25.

Table 26.

Requirements an available dietary indispensable amino acids (1AA) in
tilapia fed ﬁsh meal based diets supplemented with graded levels of phytic
acid a Na-phytate.

Predictive equations and R2, (n=12) for linear regression and non-linear
regression models applied to growth and efﬁciency parameters in juvenile
tilapia fed diets containing untreated, or phytase treated, solvent extracted
SBM substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Parameters include speciﬁc growth rate (SGR), feed efﬁciency
(FE), protein efficiency ratio (PER), and apparent net protein utilization
(ANPU).

Mean pH values (1: SE) on surface of feed collected at the esophagus,
pyloric sphincter, and eight sites within excised stomachs as described in
text. Numbers within parentheses represent number of samples collected.
Samples were collected before feeding and 0.5, l, 2, 4, 6, and 8 hours
following an initial feeding. The two groups labeled 6 and 8 hours were
sampled 2 and 4 hours following a second meal, respectively.

Mean pH values (:t SE) of gastric mucosa collected at the esophagus,
pyloric sphincter and eight sites within excised stomachs as described in
text. Numbers within parentheses represent number of samples collected.
Samples were collected before feeding and 0.5, l, 2, 4, 6, and 8 hours
following an initial feeding. The two groups labeled 6 and 8 hours were
sampled 2 and 4 hours following a second meal, respectively.

Mean pH values (1 SE) of intestinal mucosa measured at three sites in the
most anterior segment, and nine other sites as described in Figure 2.
Measurements were made before feeding, and 0.5, l, 2, 4, 6, and hours
following an initial feeding. Values represent 5 measurements in each
segment.

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

LIST OF FIGURES

Top view of stomach excised from tilapia (0. niloticus). The stomach was
slit anteriorly from the pyloric sphincter to the esophagus and ventrally to
expose the contents. View indicates 10 sites where pH measurements were
made on the surface of the gastric mucosa. The sites corresponding to 1
and 2 represent the esophagus and pylorus, respectively.

Excised GI tract of tilapia (0. niloticus). Numbers indicate intestinal
segments where pH measurements were made. Readings were collected
on the contents and mucosa at three sites along the most anterior segment;
1) immediately following the pyloric sphincter and anterior to the bile
duct; 2) the middle of the segment; 3) the posterior end of the segment.
Intestinal contents and mucosa were measured at the middle of the nine
remaining segments (sites 4 - 12).

Mean daily intake of tilapia 0. niloticus fed to satiation 1, 2, 3, or 5 times
day’l. The curve, as drawn, approaches an asymptote representing
maximum daily intake at 3 feedings clay'1 .Points represent means of 3
replicates of pooled samples (n= 5) overl a 29 day period. Mean daily
intakes of ﬁsh fed 2, 3, or 5 times day were signiﬁcantly greater than
ﬁsh fed 1 time day “ (P<0. 05), but were not signiﬁcantly different from
each other.

Cumulative feed consumed by 0. niloticus fed to satiation 1,2, 3, or 5
times day Regression analysis indicated the equations best describing
the relationships were; y=2 97x +2. 00, R2=O 997 (0); y=4. 09x +0 93,
R2=O. 993 (a); y=4. 61x +3 97, R2=O. 997 ( a); and y=4 80+5 06, R2=0 997
(O). Slope ratio analysis indicated cumulative consumption by ﬁsh fed 1
meal day was signiﬁcantly less than those fed 3, or 5 meals clay'1
(P<0.05). No other differences were detected.

Mean daily intake of 0. niloticus fed 1, 2, 3, or 5 meals day '1 partitioned
into mean consumption at each feeding. Values represent the mean
consumption of three replicate groups fed during a four week trial.

Weight gain in 0. niloticus fed to satiation 1, 2, 3, or 5 times day. 1 Fish
fed 1 meal day exhibited a signiﬁcantly lower Increase in weight than
ﬁsh fed multiple meals day, which were not signiﬁcantly different from
each other (P<0. 05).

xii

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Rate of gastric evacuation in 0. niloticus fed an experimental diet
containing Fe203 as an external marker at time 0. Fish were fed either 3
feedings day " to satiation (3:00, 12:00, and 17:00), or 5 feedings day “ to
satiation (8:00, 10:00, 12:00, 15:00, and 17:00). Cumulative iron is
relative to total iron collected 24 hrs following feeding. The gastric
evacuation rates for the two treatments are described as 66.5 e '0‘1520‘ +0001)
R2=01.90 (3 feedings day “ ); and 85.1 e 41‘5“" +000", R2=O.97 (5 feedings
day ' ).

Rate of iron appearance at terminus of 0. niloticus intestinal tract in ﬁsh
fed an experimental diet containing Fe203 as an external marker at time 0.
Fish were fed either 3 feedings day " to satiation (8:00, 12:00, and 17:00),
or 5 feedings day “‘ to satiation (8:00, 10:00, 12:00, 15:00, and 17:00).
Following the initial feeding, subsequent meals consisted of a similar diet
without F8203.

Linear increase in iron appearing at the terminal segment of 0. niloticus
intestinal tract during the ﬁrst eight hours following the feeding of an
experimental diet containing F e203 as an external marker. Cumulative
iron is relative to total iron collected over 24 hrs.

Weight gain of juvenile tilapia fed experimental diets relative to a control
diet Experimental diets contained untreated, or phytase treated, solvent
extracted SBM substituted into the control diet at 25, 50, 75, and 100 % of
the dietary protein. Asterisks indicate experimental diets resulting in
signiﬁcantly different growth rates relative to the control diet (P<0.05).

Orthogonal contrasts of weight gain in juvenile tilapia fed diets containing
untreated, or phytase treated, solvent extracted SBM substituted into a
reference diet at 25, 50, 75, or 100 % of the dietary protein. Values
represent the mean of three replicates of pooled samples (n=8). Error bars
represent SBM. Different labels within a level of substitution represent
signiﬁcant differences (P<0.05).

Orthogonal contrasts of speciﬁc growth rate (SGR) in juvenile tilapia fed
diets containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean of three replicates of pooled samples
(n=8). No signiﬁcant differences were detected between treatments at any
level of substitution.

xiii

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Orthogonal contrasts of feed efﬁciency in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean of three replicates of pooled samples
(n=8). Error bars represent SEM. Different labels within a level of
substitution represent signiﬁcant differences (P<0.01).

Orthogonal contrasts of protein efﬁciency ratio in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean of three replicates of pooled samples
(n=8). Error bars represent SEM. Different labels within a level of
substitution represent signiﬁcant differences (P<0.05).

Orthogonal contrasts of apparent net protein utilization in juvenile tilapia
fed diets containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean of three replicates of pooled samples
(n=8). Error bars represent SEM. Different labels within a level of
substitution represent signiﬁcant differences(P<0.05).

Dietary phytic acid concentrations in ﬁsh meal based experimental diets
supplemented with phytic acid as Na-phytate.

Apparent crude protein digestibility (ACPD) in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary
protein. Values represent the mean of four replicates of pooled samples
(n=15), except the two diets with 100 % substitution, where insufﬁcient
material limited the analysis to two and three replicates for the treated, and
untreated diets, respectively. Error bars represent SEM. No signiﬁcant
differences were detected among the treatments.

Apparent crude protein digestibility (ACPD) in juvenile tilapia fed ﬁsh
meal based diets supplemented with graded levels of phytic acid as Na-
phytate. Values and error bars represent the mean and SEM of four
replicates of pooled samples (n=15).

Mean pH values recorded at 10 locations in the stomach (gastric region),
and 12 locations in the intestine (intestinal region) two hours following an
early morning feeding. Measurements were made on the surface of the
mucosa and on the surface of the recovered digesta (feed). Error bars
represent the standard deviation of 5 samples, except gastric segment 1
and 2 which represent 2, and 3 samples, respectively.

xiv

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Mean pH values measured on the gastric mucosa, surface of digesta
collected form the stomach, and homogenized digesta collected ﬁ'om the
stomach before feeding, and 0.5, l, 2, 4, 6, and 8 hours postprandially.
Error bars represent the SEM of ﬁve ﬁsh collected at each sampling
period

(A) Speciﬁc growth rate (SGR), and (B) feed efﬁciency (FE) as a function
of feeding frequency in 0. niloticus fed to satiation 1, 2, 3, or 5 times

day ". The relationship for SGR is described as SGR = — 0.2209x2 +
1.5187x - 1.0476 (R2 = 0.999), and for FE as FE = - 0.0832x2 + 0.5595x
- 0.3027 (R2 = 0.979).

Semilogarithmic plot of cumulative iron collected from 0. niloticus fed an
experimental diet containing Fe203 as an external marker at time 0, against
postprandial time. Fish were fed either 3 feedings day '1 to satiation (8:00,
12:00, and 17:00), or 5 feedings day " to satiation (8:00, 10:00, 12:00,
15:00, and 17:00). Cumulative iron is relative to total iron collected 24
hrs following feeding. The lines describing the relationships are

y = -0.153x + 4.205, R2=0.90 (3 feedings day "); and y = —0.149x + 4.443,
R2=0.97 (5 feedings day '1). The slopes 0.153 and 0.149 represent the
instantaneous rates of evacuation.

Growth data (% increase) of juvenile tilapia fed graded levels of (A)
untreated solvent extracted SBM, and (B) phytase treated solvent extracted
SBM, as a percent of the crude protein. Data were ﬁtted with a two-slope
broken line model utilizing a non-linear procedure as described in
methods. Dotted lines indicate model parameter estimates. Diminishing
returns and lower growth are predicted with greater than 30.6 % of the CP
as untreated solvent extracted SBM (A), and 13.7 % of the CP as phytase
treated solvent extracted SBM (B).

Growth data (% increase) of juvenile tilapia fed graded levels of (A)
untreated solvent extracted SBM, and (B) phytase treated solvent extracted
SBM, as a percent of the crude protein. Data were ﬁtted with a quadratic
model. Dotted lines indicate model parameter estimates. Diminishing
returns and lower growth are predicted with greater than 38 % of the CP as
untreated solvent extracted SBM (A), and 17 % of the CP as phytase
treated solvent extracted SBM (B).

Dietary Met and available Met in relation to dietary CP. Diets contained
graded levels of phytase treated SBM (A), or untreated SBM (B) as a
percent of CP in the diet.

Dietary Lys and available Lys in relation to dietary CP. Diets contained
graded levels of phytase treated SBM (A), or untreated SBM (B) as a
percent of CP in the diet.

XV

Figure 27.

Figure 28.

Figure 29.

Available methionine (% dietary requirement) as a function of L-
crystalline methionine dietary supplementation. Methionine was
supplemented into phytase treated SBM, and untreated SBM diets
incorporating SBM at 50, 75, and 100 % of the dietary crude protein to
meet the dietary methionine requirement for tilapia. Curves were ﬁtted by
quadratic regression.

Weight gain in relation to dietary crude protein (CP), and as a function of
dietary phytic acid supplementation as Na-phytate (g/kg dry diet).

Conceptual ﬂow diagram of the recirculating system used in the
experiments

xvi

LIST OF ABBREVIATIONS

AA — Amino acids

Ala - Alanine

ACPD — Apparent crude protein digestibility
ANF — Anti-nutritional factor

ANPU -— Apparent net protein utilization
Arg - Arginine

Asp — Asaprtic acid

BAPNA - Nor-benzoyl-DL—arginine-p-nitroanilide hydrochloride
BOD - Biological oxygen demand

BW — Body weight

C02 -— Carbon dioxide

CP - Crude protein

Cys — Cysteine

DE — Digestible energy

DDI — Distilled deionized water

DM — Dry matter

DO - Dissolved oxygen

DMSO - Dimethyl sulfoxide

FCR - Feed conversion ratio

FE — Feed efﬁciency

FM - Fish meal

GE -- Gross energy

xvii

GER — Gastric evacuation rate

GET — Gastric evacuation time

GI - Gastrointestinal

Glu - Glutamic acid

Gly — Glycine

His - Histidine

HROM - Hydrolysis resistant organic matter
IAA - Indispensable amino acid

Ile —- Isoleucine

IU- Internaional Unit of enzyme activity
Leu — Leucine

Lys — Lysine

MDI — Mean daily intake

Met - Methionine

MS—222 - Tricainemethane sulfonate

N — Nitrogen

NH3/NH4+ - Ammonia/ammonium

N02” - Nitrite

NO3' - Nitrate

P:DE - Crude proteinzdigestible energy ratio (g/MJ)
PER —- Protein efﬁciency ratio.

Pro -— Proline

Phe — Phenyalanine

xviii

SD - Standard deviation

SEM — Stanard error of the mean
SGR — Speciﬁc growth rate

Ser — Serine

Thr — Threonine

TIA — Trypsin inhibitor activity
TIU —- Trypsin inhibitor units
Tris- Tris(hydroxymethyl) aminomethane
Trp - Tryptophan

TSAA - Total sulfur amino acids
Tyr — Tyrosine

Val — Valine

xix

INTRODUCTION

Tilapia is a generic term applied to the collection of Cichlidae species within the
genera Oreochromis, T ilapia, and Sarotherodon. They are warmwater species
originating on the African continent, and are found in a diverse range of habitats. Many
of the Tilapine species and their hybrids are cultured worldwide as a food source.

Tilapia are well suited for culturing. They exhibit rapid growth, high resistance to
stress and disease, endure submarginal water quality, readily utilize relatively low quality
feedstuffs, and are highly prolific. Current conventional wisdom dictates tilapia culture
requires recirculating systems for economic viability. Recirculating systems allow for
intensive culture of organisms while maintaining temperature and removing waste
products from the aquatic environment (Appendix 1). In much of the temperate regions,
such systems are imperative for maintaining the warm water temperatures required by
tilapia. The Nile tilapia (Oreochromis niloticus) were chosen for this study because it is
the foremost species of tilapia cultured domestically, as well as world-wide.

A major limiting constraint associated with recirculating systems is the
accumulation of nutrients, particularly nitrogenous products, and solids. Nitrogenous
wastes associated with feeds and feeding are divided into a solid fecal fraction, and a
soluble fraction associated with gill and urinary excretions. A number of mechanical
methods for the separation and removal of the solid fraction exist, and this area continues
to be a rich topic for aquacultural engineers. While numerous bioﬁltration devices exist
for handling the soluble fraction, they all utilize nitrifying bacteria for the oxidation of
NH3/NI'14 + to the end product N03 '. However, the efﬁcacy of nitrifying bacteria is

variable and contingent upon relatively unpredictable and often hard to manage

parameters such as dissolved oxygen (DO), pH, temperature, C02, biological oxygen
demand (BOD), and competitive heterotrophic bacteria, as well as circulating
NHg/NHr *, N02 ', and N03 ' levels (Wheaton 1993).

The difﬁculty and expense associated with nitrogen and solids removal makes
reducing wastes entering the system the most viable and desirable alternative. It is
estimated 10 % of ingested nitrogen is excreted in the feces with as much as 66 %
excreted as ammonia (Cho 1993 ). These values are based on feeding high quality ﬁsh
meal proteins. The values are higher with lower quality plant protein substituted diets.
Despite these statistics the industry is moving towards replacing expensive, and
sometimes unavailable, ﬁsh meal products with less expensive plant protein feedstuffs.
Solids reduction and increased nitrogen retention are obvious starting points for reducing
nitrogen inputs into the system.

The most promising plant protein alternatives identiﬁed to date are soybean
products. However, partial or complete replacement of ﬁshmeal (FM) with soy products
has met with mixed success in tilapia species (Davis and Stickney 1978; Jackson et al.
1982; Viola and Arieli 1983; Shiau et al. 1987; 1989; 1990; Davies et al. 1989; De Silva
and Gunasekera 1989; El-Dahhar and El-Shazly 1993). Generally, poor performance is
in direct relationship to the level of FM replacement (Shiau et al. 1990). Reasons cited
for decreased performance (reduced growth, protein efﬁciency ratio, apparent net protein
utilization, and increased feed conversion ratio) were residual trypsin inhibitors,
unbalanced amino acids, methionine deﬁciency, or undigestible polysaccharide anti-

nutritional factors. However, evidence to support these conclusions is lacking.

Soybean meal (SBM) based diets supplemented with methionine (Shiau et al.
1987; Shiau et al. 1989) or methionine and lysine (El-Dahhar and El-Shazly 1993) did
not increase performance to the level of FM control diets. Methionine was found not
limiting in diets with 24 % CP, but limiting in diets with 32 % CP, and methionine
supplementation only had a signiﬁcant effect on performance parameters when diets
contained SBM as the sole source of protein (Shiau et al. 1989). However, methionine
levels in these studies were above reported requirements determined for tilapia (Jackson
and Capper 1982). Substitution of SBM for PM at 30 % of the CP had no signiﬁcant
effect on total protein or dry matter digestibility, but led to a signiﬁcant reduction at
greater than 33 % substitution (Shiau et al. 1987; Shiau et al. 1989). These ﬁndings
would indicate bioavailability of methionine is less than expected, or the available amino
acids are unbalanced. However, increasing the ratio of soy ﬂour in a blend from 25:75 to
75:25 (soyflourzfeather meal) had no effect on apparent availability of methionine
(Sadiku and Jauncey 1995). The currently accepted maximum replacement of ﬁshmeal
with SBM in tilapia diets is 30 % of the crude protein (Shiau et al. 1990).

Soybeans have anti-nutritional factors (ANF) with the potential to reduce their
biological value and result in pathological states in monogastric animals (Rackis 1974;
Liener 1994). Phytic acid (1,2,3,5/4,6-hexakis dihydrogen phosphate), one such ANF,
exists as a salt of mono- and divalent cations in legumes and cereals. Phytate, the salt of
phytic acid, binds divalent cations making them unavailable, resulting in mineral
deﬁciencies. Reduced mineral bioavailability has been demonstrated in rainbow trout

(Spinelli et al. 1983; Cain and Garling 1995; Riche and Brown 1996), channel catﬁsh

(Satoh et al. 1989), carp (Hossain and Jauncey 1993), and tilapia (McClain and Gatlin
1988).

Phytate also complexes with proteins. These complexes occur via direct bonding
with phytic acid forming a binary complex and through mineral-phytate—protein ternary
complexes (Cheryan 1980; Reddy et al. 1989). In acidic environments, such as Tilapine
stomachs (pH l.0-2.0), half of the phosphorus moieties of phytic acid are negatively
charged creating an environment favorable for binding proteins with e-amino groups on
lysine, imidazole groups on histidine, and guanidyl groups on arginine. In alkaline
environments, such as Tilapine intestine (pH 8.5-8.8) protein-cation-phytate complexes
are favored. These protein-phytate and protein-mineral-phytate complexes are more
resistant to proteolytic digestion in vitro and in viva (Singh and Krikorian 1982; Satterlee
and Abdul-Kadir 1983; Grabner and Hofer 1985; Knuckles et al. 1985; Carnovale et al.
1988; Vaintraub and Bulmaga 1991; Caldwell 1992).

In vitro studies indicate phytic acid inhibits pepsin and trypsin activity.
Decreased pepsin activity is linearly related to phytate level and independent of digestion
time (Knuckles et al. 1985). Inhibition is inversely correlated to the degree of phytate
hydrolysis, and is strongly affected by pH. Maximal inhibition occurs near pH 2.0
(Camus and Laporte 1976; Vaintraub and Bulmaga 1991) suggesting the potential for
decreased proteolytic digestion efﬁciency in tilapia. Decreased activity is likely due to
complex formation making sites on the protein less susceptible to enzymatic attack.

As with pepsin, the inhibition of trypsin is dependent on phytate concentration.
The inhibition is a function of temperature, calcium concentration, and contact time

(Singh and Krikorian 1982). The mechanism by which this inhibition occurs is ill

deﬁned Possible explanations for the inhibition are decreased activation of the zymogen
form, increased autolysis of trypsin, formation of the ternary complex, or competitive
sequestration of Ca“ ions between phytate and trypsin (Singh and Krikorian 1982;
Vaintraub and Bulmaga 1991; Caldwell 1992).

Decreased protein digestibility of diets supplemented with salts of phytic acid led
to depressed growth and poor performance in rainbow trout (Spinelli et al. 1983),
Chinook salmon (Richardson et al. 1985), and carp (Hossain and Jauncey 1993). Carp
are stomachless ﬁsh and do not produce pepsin, providing further evidence that trypsin
activity may be altered.

The addition of microbial phytase to swine diets signiﬁcantly increased total tract
digestibility of CP, and all amino acids except cystine and proline. Ileal digestibilty of
methionine and arginine were also signiﬁcantly increased (Mroz et al. 1994). Nitrogen
retention was increased and daily nitrogen excretion was reduced 20-25 % suggesting
improved amino acid balance. Nitrogen balance studies indicated signiﬁcantly higher
fecal and urinary nitrogen losses in rats fed a high phytate bran ﬂour diet (Satterlee and
Abdul-Kadir 1983). These authors also demonstrated increased protein digestibility,
higher biological value, and better protein efﬁciency ratios with diets containing lower
phytate levels using both in vitro and in viva techniques.

Monogastric animals, including ﬁsh, are unable to hydrolyze the complexing
phosphate groups on phytate due to a lack of intestinal secretions of the enzyme phytase.
However, diets prepared with phytase either as an enzymatic pretreatment, or as an
additive, result in better performance. Cain and Garling (1995) found rainbow trout fed

diets pretreated with phytase exhibited superior growth compared to ﬁsh fed the same

diets without pretreatment. The authors suggested the increased performance may be
attributable to improved protein quality.

In early work, it was suggested tilapia did not possess a functional stomach. The
investigators suggested tilapia contain an intestinal bulb, an enlarged region of the
anterior intestine found in stomachless ﬁshes (Bowen 1982). However, recent evidence
suggests tilapia contain the histological, morphological and physiological characteristics
required for gastric digestion. Tilapia exhibit tremendous plasticity of the GI tract, which
lends itself to adaptability and a high degree of variability (Smith 1989; Boujard and
Leatherland 1992). It is this characteristic which was likely responsible for the
confusion.

Morphologically, tilapia exhibit distinct muscular cardiac and pyloric sphincters
(Moriarity 1973) as well as endodermal epithelium (Smith 1989). Histologically, chief
cells, oxyntic cells (Kapoor et al. 1975 ), and gastric glands (Al-Hussaini and Kholy
1953) have been identiﬁed Physiologically, acid secretion is under neuronal control
(Fish 1960) and tilapia exhibit peristaltic movement (Moriarity 1973).

Acidic proteases have been identiﬁed in tilapia (Fish 1960; Moriarity 1973). One
such protease has been isolated and characterized from 0. niloticus stomach mucosa
(Yamada et al 1993). Kinetic analysis indicated the Km was 5.4 mg/mL utilizing
hemoglobin as substrate. Maximum activity was observed at 50°C and pH 3.5 which was
similar to eel and ayu, but higher than rainbow trout, dace, and bonito (Yamada et al.
1993). The enzyme had a molecular weight of 54,000 and an isoelectric point of 3.7.
Substrate and inhibitor assays indicated the enzyme was an aspartic acid protease with

pepsin-like activity similar to swine pepsin (Yamada et al. 1993).

In addition to gastric digestion, tilapia exhibit intestinal proteolytic activity
(Nagase 1964; Hofer and Scheirner 1981). Proteolysis has been attributed to both
trypsin- and chymotrypsin—like enzymes (Fish 1960; Moriarity 1973). It was speculated
tilapia lack carboxypeptidase activity (Moriarity 1973). However, this seems unlikely
since tilapia have been found to exhibit the full complement of enzymes found in other
ﬁsh (Nagase 1964), including endoprolylpeptidase, leucinaminopeptidase,
cysteinaminopeptidase, prolylamino peptidase, and aminopeptidase (Barth et al. 1995).

Histological evidence supports the existence of exocrine pancreatic enzymes.
Acinar cells containing zymogen granules have been identiﬁed in 0. niloticus (Kugler
and Pequignot 1988). In addition, two pancreatic proteolytic enzymes from 0. nilotr'cus
intestine have been isolated and characterized (Yamada et al. 1991). Kinetic analysis
indicated the Km for the two enzymes was 0.03 mg/mL utilizing casein as the substrate.
Maximum activity was observed at 55°C and pH 8.5-9.0. The two enzymes, labeled
PA-3 and PB-3 had molecular weights of 32,000 and 21,000, respectively. The effects of
various inhibitors led the authors to predict PA-3 was a serine protease, and PB-3 was a
cysteine protease (Yamada et al. 1991). Trypsin-like enzymes from tilapia behave
similarly to porcine trypsin (El-Shemy and Levin 1997) indicating the potential for
inhibition by phytic acid.

Pretreatment of plant proteins with phytase should render phytate incapable of
sequestering proteins and minerals, potentially increasing their availability to the animal.
The hydrolysis of phytate should decrease the inhibitory effect observed on gastric and
intestinal proteolytic enzymes thereby increasing CP digestibility. Assuming energy is

not a limiting factor, increased amino acid availability would potentially increase protein

accretion, and decrease amino acid catabolism and ammonia excretion. Both solid and
soluble nitrogen would be reduced minimizing nitrogen inputs into the system, while also
allowing for greater substitution of soybean products for ﬁshmeal.

Therefore, the overall objective of this study is to determine whether phytic acid
associated with soybeans is a causative agent responsible for reduced growth and
performance in tilapia fed SBM based diets. More speciﬁcally, the objectives of this
study are two fold. The ﬁrst is to determine what rate of SBM incorporation leads to
equivalent growth and performance relative to a ﬁsh meal control diet, and whether
removal of phytic acid can increase the rate of SBM incorporation. The second objective
is to determine whether phytic acid inhibits the activity and function of gastric and

intestinal proteases in tilapia.

HYPOTHESIS STATEMENTS

Expe_rim§nt l — Optimal Feeding Frequency

In ﬁsh nutrition studies, it is imperative that food availability, in itself, does not
act as a factor in limiting growth and efﬁciency. Nutrient utilization efﬁciencies should
be calculated at the feeding frequency at which rates of growth and efﬁciency are not
suppressed (J obling 1983). Optimum feeding frequencies vary with physiology, diet,
behavior, species, size, and temperature. There is some contention as to optimal feeding
frequencies for tilapia Additionally, recommendations do not differentiate between
species, feed type, or rearing system (NRC 1993). Based on feeding behavior,
physiology, and gastrointestinal morphology of wild ﬁsh it has been reported tilapia
require many frequent small meals to achieve greatest efﬁciency (Moriarity 1973;
Jauncey and Ross 1982). However ﬁsh reared in conﬁnement have different

requirements than wild ﬁsh.

Hypothesis: There is an optimum feeding frequency for 0. niloticus reared in

recirculating systems beyond which growth and efﬁciency are reduced.

Experiment 2 — Tilap_ia Gastric Evacuation Rate

The rate at which food can be consumed and efﬁciently utilized is a prime factor
in determining growth rate. The rate of consumption is a function of environmental
conditions, meal size, ﬁsh size, and feeding frequency (Holmgren et al. 1983). Feeding
frequency has been shown to be strongly correlated with gastric evacuation time

(Holmgren et al. 1983).

Utilizing an inert undigestible marker the rate at which food traverses the GI tract
can be determined (Fange and Grove 1979). The assumption is ﬁsh will eat available
food in amounts depending on stomach fullness and at intervals determined by the rate of
emptying (Holmgren et al. 1983). Additionally, the rate at which food traverses the GI

tract dictates the contact time the intestinal milieu has with digestive enzymes.

Hypothesis: The rate of passage through the GI tract of 0. niloticus is affected by

feeding frequency.

Experiment 3 — Solvent Extracted SBM Growth Trial

Phytate complexes with proteins (Cheryan 1980; Reddy et al. 1989). This
complex formation potentially reduces protein digestion and availability of amino acids
for uptake. Increased protein digestibility, higher biological value, and better protein
efﬁciency ratios have been demonstrated with diets containing lower phytate levels
(Satterlee and Abdul-Kadir 1983). Cain and Gatling (1995) found rainbow trout fed diets
pretreated with phytase exhibited superior growth compared to ﬁsh fed the same diets
without pretreatment. The authors suggested the increased performance may have been

attributable to improved protein quality.

Hypothesis: Removing phytate from SBM with the use of phytase will increase the
biological value of the SBM, increasing growth and efficiency.
Conversely, feeding increasing graded levels of untreated SBM will

decrease the biological value of the SBM, decreasing growth and

10

efﬁciency in a dose-response manner.

E)_rpe_nm' ent 4 — Phﬂic Acid Supplemented Fish Meal Growth Trig]

 

prhytic acid is the causative agent for differences in growth and efﬁciency in 0.
niloticus fed phytase treated and untreated diets, the eﬁ‘ects should be mimicked by
puriﬁed phytic acid. Incorporating puriﬁed phytic acid into a control diet, in graded
levels reﬂecting those in the SBM diets, should mimic a dose-response relationship on
grth and efﬁciency similar to one observed in experiment 3. Such a relationship

would lend support to the hypothesis in experiment 3.

Hypothesis: Incorporating graded levels of purified phytic acid into a control diet
will decrease the biological value of the control diet decreasing growth

and efficiency in a dose-response relationship.

Experiment 5 - Solvent Extracted SBM Digestibility Tug

Decreased protein digestibility of diets supplemented with salts of phytic acid led
to depressed growth and poor performance in rainbow trout (Spinelli et al. 1983),
Chinook salmon (Richardson et al. 1985), and carp (Hossain and Jauncey 1993). The
addition of microbial phytase to swine diets signiﬁcantly increased total tract digestibility
of CP, and all amino acids except cystine and proline. Ileal digestibilty of methionine
and arginine were also signiﬁcantly increased (Mroz et al. 1994). Pretreatment of plant
proteins with phytase should render phytate incapable of sequestering proteins and

minerals, potentially increasing their availability to tilapia.

11

Hypothesis: Removing phytate from SBM with the use of phytase will increase the
digestibility of total N and individual amino acids. Conversely, feeding
increasing graded levels of untreated SBM will decrease digestibility of

total N and individual amino acids in a dose-response relationship.

Emriment 6 — thc Acid Supplemented Fish Meal Digestibility Trial

As in experiment 4, if phytic acid is the causative agent for differences in total N
and individual amino acid digestibilities in 0. niloticus fed phytase treated and untreated
diets, the effects should be mimicked by puriﬁed phytic acid. Incorporating puriﬁed
phytic acid into a control diet, in graded levels reﬂecting those in the SBM diets, should
mimic a dose-response relationship on N and individual amino acid digestibilities in

tilapia.

Hypothesis: Incorporating increasing graded levels of purified phytic acid into a
control diet will decrease digestibility of total N and individual amino

acids in a dose-response relationship.

Expeg'ment 7 — Tilapia Gastrointestinal Trarct pH Proﬁle

Tilapia exhibit tremendous plasticity of the GI tract, which lends itself to
adaptability and a high degree of variability (Smith 1989). There is evidence to suggest
Tilapine stomachs are highly acidic, with pH values as low as 1.0-2.0 (Moriarity 1973).
Maximal inhibition of enzymes by phytic acid occurs near pH 2.0 (Camus and Laporte

1976; Vaintraub and Bulmaga 1991) suggesting the potential for decreased proteolytic

12

digestion efﬁciency in tilapia. However, maximum activity of a protease isolated from
0. nilaticus gastric mucosa was observed at pH 3.5 with much reduced activity at pH
values as low as 2.0, and only 10 % activity at pH 5.5 (Yamada et al. 1993).

In alkaline environments, such as Tilapine intestine (pH 8.5-8.8) protein-cation-
phytate complexes are favored. These protein-phytate and protein-mineral-phytate
complexes are more resistant to proteolytic digestion in vitra and in viva (Singh and
Krikorian 1982; Satterlee and Abdul-Kadir 1983; Grabner and Hofer 1985; Knuckles et
al. 1985; Carnovale et al. 1988; Vaintraub and Bulmaga 1991; Caldwell 1992). In
addition to demonstrating tilapia manifest a suitable environment for protein-phytic acid
complexation, the data gathered can be used to mimic in viva pH conditions in the in

vitra enzyme assays.

Hypothesis: The pH values of the 0. nilaticus GI tract are optimal for protein-phytic

acid complex formation.

13

METHODS AND MATERIALS

Experiment 1 - Optimal Feeding Frequency

A mixed sex population of 0. nilaticus was obtained from Illinois State
University (Normal, Illinois) and transported to Michigan State University’s Fisheries
Research Laboratory. Fish were held in ﬂow-through well water heated to 25 i1.0°C

until stocking.

Experimental Design

Fish were fed a standard commercial trout diet for a four-week preliminary period
prior to stocking in experimental units. The diet was analyzed for gross energy (GE), and
proximate components. The diet contained 46.3 % crude protein (CP); 3.86 % lipid;

18.4 % ash; and 18.02 kJ/g, on a dry matter basis.

During the preliminary period, ﬁsh were fed 2 % wet body weight per day '1
(BW/day) divided between two feedings. After four weeks, ﬁve ﬁsh were randomly
selected for analysis. All analyses were performed in triplicate. Efﬁciency parameters
were determined by difference between ﬁsh analyzed before the experiment and at its
termination.

Five ﬁsh each were stocked into 12 experimental units. Experimental units were
deﬁned as 40 L tanks. The system was a parallel ﬂow-through system receiving water
from a common head tank. Well water in the head tank was heated to 27°C. Fish were

maintained on a 16:8 lightzdark cycle. Mean weight of ﬁsh at stocking was 34.4154 g.

14

Prior to the experimental period, ﬁsh were allowed a one-week acclimation period and
fed as above. The experimental period lasted 29 days.

Experimental units were randomly assigned a daily feeding regimen. The feeding
regimens were once daily (8:00 h); twice daily (8:00 and 17:00 h); three times daily
(8:00, 12:00, and 17:00 h); and ﬁve times daily (8:00, 10:00, 12:00, 15:00, and 17:00 h).
Fish were fed 6 days a week and weighed on the seventh. Fish in each unit were fed to
apparent satiation at each feeding.

Feeding followed the same regimen for each tank and meal. Fish initially
received pellets until they lost interest. After all tanks received feed, a second and third
pass was made offering additional pellets. Satiation was deﬁned as the point at which a
single pellet remained uneaten for l min after the third pass. Consumption was recorded.

Temperature and dissolved oxygen were monitored Parameters were within

acceptable ranges for tilapia (Cherivinski 1982; Papoutsoglou and Tziah 1996).

Proximate Ana_1ys'§

At termination of the experimental period, ﬁsh were weighed collectively and
euthanized in tricainemethane sulfonate (MS-222) at a concentration of 500 mg/L (Post
1983). Euthanized ﬁsh were weighed individually, dissected, and sexed. Visceral tissue
was removed and weighed Visceral tissue was deﬁned as all tissues within the body
cavity including gonadal tissue. Perivisceral fat was removed from the gastrointestinal
tract and weighed. Whole ﬁsh, including tissues and perivisceral fat, were frozen at

-20°C until analyzed.

15

Frozen whole ﬁsh were passed through a meat grinder and further homogenized
by pulverizing with mortar and pestle. Ground tissues were dried at 105°C after which
they were further ground to pass a 850 um screen. A 2500:0003 g subsarnple was
taken from each ﬁsh within an experimental unit and pooled. Pooled samples were
blended by hand mixing until the sample appeared homogenous, but not less than one
minute. The blended samples were used for the remaining proximate analyses. All
analyses were performed in triplicate on pooled samples.

Dry matter was determined by standard methods (AOAC 1990). Feed and whole
body nitrogen content were determined by the Kjeldahl method (AOAC 1990). Crude
protein was calculated as N x 6.25. Gross energy of feed and whole ﬁsh were determined
by bomb calorimetery using the isoperibol method (Parr Instruments, Moline, IL). Crude
lipid content of feed and whole ﬁsh were determined by difference following lipid
extraction with diethyl ether (AOAC 1990). Ash was determined by combustion at

550°C (AOAC 1990).

galaﬂatiané

Fish within treatments were analyzed for grth and efﬁciency parameters.
Efﬁciency parameters evaluated were feed efﬁciency (FE), protein efﬁciency ratio
(PER), speciﬁc grth rate (SGR), and apparent net protein utilization (ANPU).
Efﬁciency parameters were calculated using the standard equations of Jauncey and Ross
(1982)

Feed Efﬁciency:

FE == wet weight gain / total feed fed

16

Protein Efﬁciency Ratio:
PER = wet weight gain / total protein fed
Energy Retention:
ER = [(fo Bf) - (Wo x 130)] / El x 100
Speciﬁc Growth Rate (%lday):
SGR =[ (ln Wf - ln W0) / (T2 - T1)] x 100
Apparent Net Protein Utilization:
ANPU = [(fo Pf) - (W0 x Po)] / PI x 100
where :
Wf = ﬁnal wet weight
W0 = beginning wet weight
Pf = ﬁnal whole body protein
P0 = beginning whole body protein
Bf = ﬁnal whole body energy
E0 = beginning whole body energy
P1 = total protein intake
E1 = total energy intake

(T2 - T1) = number of days during the experiment

Statistical Analysis
A completely randomized design was employed with number of feedings
day ‘1 as the main effect. Proximate components, F CR, ANPU, PER, and SGR were

analyzed as a one-way ANOVA using the general linear method of SAS (SAS, 1979).

17

When signiﬁcant differences were detected, means were separated using Duncan’s
multiple range test. Signiﬁcance was reported at P<0.05 controlling for the type I
comparisonwise error rate.

Cumulative feed fed was analyzed by ANOVA using the regression procedure of
SAS. Slopes were tested with t-tests, and signiﬁcance reported at P<0.05.

Multivariate regression analyses were performed with visceral wet weight and
weight of perivisceral fat as dependent variables, and total wet weight, gender, and
number of feedings a day as the mutlivariate regressors (SAS 1979). The same procedure
was also used with weight of perivisceral fat as the dependent variable, and total wet
weight, gender, and number of daily feedings as the independent variables. Results were
reported in partitioned ANOVA tables. F-statistics and probabilities were reported and

the null hypotheses rejected at P<0.0l.

Experiment 2 - Tilapia Gastric Evacuation Rate

Nile tilapia were obtained from Purdue University and transported to Michigan
State University's Fisheries Research Laboratory. Fish were held in a recirculating
system at 28°C until stocking. Fish were fed a standard commercial catﬁsh diet during a

four-week acclimation period prior to stocking in experimental tanks.

Experimental Diet Preparation
The experimental diet formulation is given in Table 1. Herring meal was obtained
from Zeigler Brothers, Inc. (Gardners, Pennsylvania). Solvent extracted SBM was

obtained from Zeeland Farm Services (Zeeland, Michigan) (Appendix 2). Dextrin (type

18

Table 1. Composition of experimental diets fed to tilapia (Oreachramis nilaticus) to
determine gastric evacuation rate and intestinal motility.

 

 

Ingredient Cr2O3 diet Fe203 diet
(g/kg dry diet) (g/kg div diet)
Herring meal 259.7 259.7
Soybean meal (solv. ext.) 277.5 277.5
Wheat Bran 100.0 100.0
Dextrin 100.0 100.0
Mineral Premix ' 60.0 60.0
Vitamin Premix 2 3.0 3.0
Carboxymethyl cellulose 20.0 20.0
Cellulose 68.0 68.0
Ascorbic acid 1.0 1.0
Choline chloride 0.8 0.8
Menhaden oil 75.0 75.0
Soy oil 25.0 25 .0
Chromic Oxide 10.0 ----
Ferric Oxide --—--- 10.0
Total 1,000.0 1,000.0

 

‘ Mineral premix contained (g/kg dry mix): CaSO4, 350.0; NaHzPO4, 250.0; KH2P0.,
250.0; MgCO; - 51120, 20.0; Znso. - 71120, 3.0; Peso. - 71120, 2.3; Mnso4 - H2O, 2.0;
CuCl2 - 2H2O, 1.0;A1Cl3 - 6H2O, 1.0; KF, 0.5; KI, 0.1; Na2SeO3 o. 1; CoCl2 - 6H2O, 0.1;
NaMoO4 - 2H2O, 0.1.

2 Warmwater ﬁsh performance premix (Hoffmann-La Roche, Inc.,Nutley, NJ) — as
incorporated in the diet: vitamin A, 10,582 IU; vitamin D3, 2,381 IU; vitamin E, 132
IU; vitamin K, 2 mg; BIZ, 4.4 pg; folic acid, 5.3 mg; riboﬂavin, 17.2 mg; pantothenic
acid, 42.3 mg; niacin, 105.8 mg; choline-Cl, 529.1 mg; thiarnin, 11.9 mg; pyridoxine,
13.2 mg; biotin, 165 pg.

19

H from corn), Chromic oxide, ferric oxide, and L-ascorbic acid were obtained from Sigma
Chemical Co. (St. Louis, Missouri). Choline chloride, CMC, and ct-cellulose were
obtained from ICN Biochemicals (Cleveland, Ohio). Alkali reﬁned, bleached, and
pressed menhaden oil stabilized with 200 ppm Coviox was supplied by Zapata Protein,
Inc. (Reedville, Virginia). Feed grade soybean oil was purchased from a local retailer.
All dry ingredients were mixed in a liquid-solids V-mixer for a minimum of 12
hours. Ingredients were then transferred to a Univex mixer (Univex Corp, Salem, New
Hampshire) where water and lipids were added under continuous mixing. The moist diet
was cold extruded using the appropriate die cast for the experimental ﬁsh. Pelleted diets
were dried in a forced air oven at 60°C for 12 hours. Dried diets were stored at —20°C

until fed.

Exxrimental Design

The experimental system was a 4,200 L recirculating system containing eighteen
150 L tanks. The recirculating system was similar in scope and design as in Appendix 1.
Space limitations required the use of three complete blocks. Each block represented a
replicate. One ﬁsh was stocked into each of the 18 tanks. Individual ﬁsh were
considered experimental units. Experimental units were randomly assigned to one of two
feeding regimens, and one of eight sample collection periods. The two feeding regimens
consisted of feeding to satiation either 3 times day ’1 (8:00, 12:00, and 17 :00 hrs), or 5
times day " (8:00, 10:00, 12:00, 15:00, and 17:00 hrs). The eight sample collection
periods were 1, 2, 4, 6, 8, 12, 18, and 24 hrs following the 8:00 meal on the experimental

day. The remaining two tanks in each block were used to sample ﬁsh at 0 (before

20

feeding) and 0.5 hr following the 8:00 feeding and were not used for statistical analysis,
but for descriptive purposes only.

Fish were fed an experimental diet containing chromic oxide (Cr2O3) at 1.0 % of
the dry diet (Table 1). The Cr203 diet was fed for a four day preliminary period prior to
commencement of a one day experimental period Following the fourth day, ﬁsh were
given one meal at 8:00 hrs of a similar diet containing ferric oxide (Fe2O3) at 1.0 % of the
dry diet. Ferric oxide was substituted for Cr2O3 (Table 1). During all meals, ﬁsh were
fed to apparent satiation, which was deﬁned as the point at which‘a single pellet remained
uneaten for 1 min. Consumption was recorded

During the 1 day experimental period, ﬁsh were serially dissected at 0, 0.5, 1, 2,
4, 6, 8, 12, and 24 hrs following the 8:00 hr feeding. Fish were euthanized in MS-222 at
a concentration of 500 mg/L before dissection (Post 1983). The ﬁsh remaining in the
treatments receiving 3 meals day '1 were fed to satiation again at 12:00 hrs and 17:00 hrs.
The ﬁsh remaining in the treatments receiving 5 meals day ‘1 were fed to satiation again
at 10:00, 12:00, 15:00, and 17 :00 hrs. After the initial feeding with the R203 diet all
subsequent feedings were with the Q20; diet.

The entire GI tract was removed and rinsed with cold distilled deionized water
(DDI). Visceral fat and other tissues were removed. The GI tract was divided into seven
segments. The segments consisted of the stomach, and two equal size segments each
from the anterior, middle, and posterior intestine. The segments were excised and
visually inspected for iron containing digesta All digesta and feces from each segment

were removed and rinsed with cold DDI water to remove any potential blood

21

contaminants. The samples were dried at 105°C for 24 hrs. Dried feces were ground and

homogenized.

Iron Analysis

Ground samples 5 100 mg were wet ashed in 5 mL concentrated H2S04 until
charred. Samples were allowed to cool slightly before adding 30 % H202 dropwise until
remaining carbonaceous material was completely oxidized. Samples were returned to
heat. Aﬁer heating for approximately 5 minutes, samples were again allowed to cool
slightly before adding 5 mL DDI water. Samples were again returned to heat for 2
minutes. After cooling, samples were brought to 100 mL with DDI water and ﬁltered
through Whatrnan # 1 ﬁlter paper.

Filtered samples were analyzed for ferric iron colorimetrically at 7t=535nm
(Davies, Bush, and Motzok 1972). The assay was slightly modiﬁed by substituting
bathophenanthroline disulfonic acid for 4,7-diphenyl-l ,10-phenanthroline thereby

obviating the need for a sulfonation step.

Statistical Analysis
ANOVA using the general linear method (SAS 1979) was performed on iron
consumption, total consumption, and amount of iron appearing in the terminal segment.
The data were analyzed as a randomized complete block design with all factors ﬁxed
using the model statement:
Yijkl=I1+Bi+Tj+ Sk+8ijltl

where B = block; T = time; and S = segment.

22

Contrasts between the two feeding regimens were performed on iron appearing in the
terminal segment of the posterior intestine.

Data utilized for determining gastric evacuation rate (GER) were transformed
The transformed data (natural log cumulative iron) was plotted versus time to generate a
linear relationship and the slopes tested by t-test. Similarly a t-test was used to test the
slopes deﬁning the relationship between rate of iron appearance in the terminal segment

and time following initial feeding. Signiﬁcance was reported at P<0.05.

Experiment 3 — Solvent Extracted SBM Growth Trial

Nile tilapia were obtained ﬁom Purdue and transported to Michigan State
University. These ﬁsh were spawned at Michigan State University and reared in aquaria
until stocked for the experiment. Second generation ﬁsh were used for an eight-week
trial to evaluate the effects of graded levels of SBM inclusion on growth and

performance. Fish not utilized during this experiment were saved for spawning.

Solvent Extracted SBM Diet Premtion

Solvent extracted SBM was obtained from Zeeland Farm Services (Zeeland,
Michigan) (Appendix 2). Herring meal was obtained from Zeigler Brothers, Inc.
(Gardners, Pennsylvania). The SBM and herring meal were ground to pass an 850 um
mesh screen.

Dextrin (type II from corn), L-methionine, and L-ascorbic acid were obtained
from Sigma Chemical Co. (St. Louis, Missouri). Choline chloride, CMC, and or-cellulose

were obtained from ICN Biochemicals (Cleveland, Ohio). Alkali reﬁned, bleached, and

23

pressed menhaden oil stabilized with 200 ppm Coviox was supplied by Zapata Protein,
Inc. (Reedville, Virginia). Feed grade soybean oil was purchased from a local retailer.

Soybean meal was substituted, for an isonitrogenous mixture of herring meal and
cellulose, to provide 25, 50, 75, or 100 % of the crude protein (Table 2). Crystalline L-
methionine was added to diets containing SBM substituted at 50, 75, and 100 % of the
CP to meet the Met and total sulfur amino acid (TSAA) requirements for 0. nilaticus
(Santiago and Lovell 1988). Methionine addition was at the expense of cellulose. All
diets were formulated to contain 33 % CP on a dry matter basis, and a P:DE of 25.0 W]
based on predicted digestible energy values for 0. nilaticus (Anderson et al. 1991; NRC
1993). All diets were supplemented with a complete mineral premix (Table 3), and
Roche warmwater ﬁsh performance vitamin premix (Table2).

Microbial phytase (BASF, 5,000 111/ g) was activated by hydration in a 0.1 M
citrate solution (pH 5.0) at room temperature. The enzyme solution was mixed
thoroughly for 15 min. Soybean meal was wetted with phytase solution 1:1 (w/v) and
mixed thoroughly for one hour at room temperature. The resultant mash was covered and
incubated for 6 hrs at 50°C. Following incubation, the SBM was dried in a forced air
convection oven at 60°C. The re-dried SBM was reground to pass an 850 um mesh
screen The SBM used in diets without phytase treatment was wetted 1:1 (w/v) with a
0.1 M citrate buffer (pH 5.0). The sham treated SBM was mixed, incubated, dried, and
reground in the same manner as the phytase treated SBM.

All dry ingredients for each experimental diet were mixed in a liquid-solids
V-mixer for a minimum of 12 hours. Ingredients were then transferred to a Univex mixer

(Univex Corp, Salem, New Hampshire) where water and lipids were added under

24

 

 

o.8o.. 98.... o.o8.. 0.8.... 0.8.... .soe
3N QR on. cam QR 23-8... .8 $2.3m
one one ...? ...? ...? 2.3.3 ..o 53.52
3.. ...Na 3.... 3: 3mm o83.30.e

...o ...o ...o ...o ...o oeeczo e5.2.0
e.. e.. e.. e.. e.. 28 2&8?
SN 98 OS e..... 98 825:8 ......oixoeao
9m 9m 3. 9m em a 582.. 5535
0.8 ode o8 ode o8 _ ago... .2252
.5. can. en. cam. cam. 833... 5on
2 a. no it ..... 8.8.532-..
See 98“ v.2... boo. ..... 2921 2% 8.8.5 2.3.3..
-..: an: «in 3.: 23 25.8% .eoa gram
8. me 8 mm .2230 a coo... .aeeeanE Eczema—I

 

act—533$ .32 anagram «=33.—

 

.ee be 2.. .o e..» ca 56% ao=a> ....oooa 3.5 once a. 8. ..5 .ﬁ .8 .2 3.38.. so... .228 .85 mane. 2:
a... 2% .o 85:35 an... benzene as .8 sea a. 8.... ace .seoanaxo 2mm 38%.. 22.8 .e 82838 .N 2%...

25

.m8 3486 653238055

we me. .o .58... we 2. .oeaoeea we a... .55... as ..amn 6.8.3.: as ”.8. .5an we 2.. .28

oaoesea we 2. aces... as on .28 2.8 as N .oeeeaeoe 5. .3 .m sass 5. 53 an sass 5. and.
.< £8.83 50.6 5 we. can wage—.8 05 3.395 9.2 $0.32 .65 .ogoom 3.58an .2805 Eaton emu 533833 2.8% N

m 2.3 _

3.83 N can...

26

Table 3. Mineral premix composition and dietary mineral supplementation in both the
soybean meal and phytic acid supplemented experimental diets fed to tilapia

 

 

Premix Salt Concentration Dietary Mineral
Mineral Salt (g/kg dry premix) Supplementation
(mg/kadry diet)
CaS04 350.0 6,182
NaH2P04 250.0 3,366
KH2P04 250.0 3,414
MgCO; - 5H20 20.0 167
ZnSO4 - 7H20 3.0 41
Peso. . 7H20 2.8 ‘ 34
MnSOi - H20 2.0 39
CuCl2 - 2H20 1.0 22
AlC13 0 6H20 1.0 6.7
KF 0.5 9.8
KI 0.1 4.5
Na2Se03 0.1 2.8
CoCl2 - 6H20 0.1 2.6
NaMoOi - 2H20 0.1 1.5

 

27

continuous mixing. The moist diet was cold extruded using the appropriate die cast for
the experimental ﬁsh. Pelleted diets were dried in a forced air oven at 60°C for 12 hours.

Dried diets were stored at —20°C until fed.

Expgrimental Desigp

The mixed sex population of 0. nilaticus spawned from ﬁsh obtained from
Purdue University were reared in aquaria until reaching approximately 1-1 .5 g. Fish held
during this period were fed a standard commercial trout diet (Purina Mills, St. Louis,
Missouri).

The experimental system was a 3,750 L recirculating system, with 27
experimental tanks, and similarly conﬁgured as in experiment 2. Temperature was
maintained at 28 21:1.0°C and ﬂow rate to each tank maintained at 1-2 L/min. Fish were
maintained on a 16:8 lightzdark cycle. Dissolved oxygen was monitored three times
daily. Ammonia, nitrite, and nitrate were measured 3 times weekly with a Hach chemical
test kit (Hach Co., Loveland, Colorado). All water quality parameters were within
acceptable limits for tilapia throughout the trial (Cherivinski 1982; Daud et al. 1988;
Papoutsoglou and Tziah 1996).

A completely randomized 2 x 4 factorial design was employed, in addition to a
control group. Phytase treatment and level of SBM substitution served as main effects.
Eight ﬁsh, each, were randomly stocked into 40 L tanks. Each tank was deﬁned as an
experimental unit. Experimental units were randomly assigned a dietary treatment.

Three replicates were run for each treatment and the control.

28

Mean weight at stocking was 1.29 g (0.04 g SBM, n=216). Fish were allowed a
two-week acclimation period during which they were fed the control diet. Fish were fed
to satiation three times per day. Maximum intake was empirically determined as 7.5 %
(SEM 0.15, n=189) BW/day on a dry matter basis. In an attempt to optimize utilization,
ﬁsh were offered 80 % of maximum consumption. This rate corresponded to suggested
feeding rates for tilapia (Jauncey and Ross 1982; NRC 1993).

Following acclimation, the ﬁsh were fed the experimental diets at 6.0 % BW/day
on a dry matter basis, divided between three equal meals. Weights were recorded at two
week intervals to adjust feed rates. Fish were fed 7 days a week, except on days ﬁsh
were weighed

The experiment was terminated after eight weeks. Fish were euthanized in
MS-222 at a concentration of 500 mg/L (Post 1983). Euthanized ﬁsh were weighed,
pooled by experimental unit, and stored at —20°C for analysis. In addition to the pooled
samples, 25 ﬁsh were randomly selected from the population at stocking, and stored at

—20°C for CP, dry matter, and lipid analysis.

Phytate Analysis
Phytate was determined colorimetrically (Latta and Eskin 1980) on phytase

treated and untreated SBM, as well as the experimental diets. Approximately 100 mg
samples were placed in 5.0 mL of 2.4 % HCl in sealed vials. Samples were extracted
overnight on a shaker bath maintained at room temperature. Samples were ﬁltered
through Whatrnan #1 ﬁlter paper under vacuum. Samples were rinsed with DDI water

and brought to 20 mL.

29

Phytate was separated via anion exchange chromatography using 200-400 mesh
AGl-X4 chloride exchange resin (Bio-Rad Laboratories, Richmond, California).
Charged resin was eluted with 5.0 mL of 0.05 M NaCl to remove organic phosphorus.
The resin was recharged with 10.0mL 0.7 M NaCl and rinsed with 15.0 mL DDI water.
Filtered samples were passed over the column and rinsed with 15 .0 mL DDI water. The
rinsed resin was eluted with 0.05 M NaCl to remove inorganic phosphorus. The column
was eluted with 15.0 mL 0.7 M NaCl and eluent collected for analysis.

Standard solutions were prepared from sodium phytate (Sigma Chemical

Company, St. Louis, Missouri). A 1.0 mL aliquot of Wade reagent (0.03 % FeCl . 6 H20

and 0.3 % sulfosalicylic acid) was added to 3.0 mL aliquots of standards and samples for

phytate determination. Absorbance was read at it=500 nm.

Trypsin Inhibitor Activity

Residual trypsin inhibitor activity was determined on solvent extracted SBM
using a modiﬁcation of American Association of Cereal Chemists Method 71-10 (AACC
1983). A 1.0 g sample of SBM was extracted with 50 mL 0.01 N NaOH for 3 hr.
Samples were transferred to 50 mL scintillation tubes and centrifuged at 4000 RPM for
10 min. Aliquots of 0, 0.6, 1.0, 1.4, and 1.8 mL of extract were brought to 2.0 mL, mixed
with 2.0 mL trypsin solution (4 mg Type I-from bovine pancreas, (Sigma Chemical Co.,
St. Louis, Missouri) in 200 mL 0.001 M HCl), and warmed to 37°C. Following addition
of 5.0 mL BAPNA (40 mg BAPNA dissolved in 1.0 mL DMSO and diluted to 100 mL
with 0.05 M tris buffer (pH 8.2)) the solution was incubated at 37°C for exactly 10 min

before stopping the reaction with 1.0 mL 30 % acetic acid. The solution was twice

30

ﬁltered through Whatrnan # 2 ﬁlter paper and absorbance read at A = 410 nm. Activity
was expressed as TIU which was deﬁned as an increase of 0.01 absorbance units at 410
nm per 10.0 mL reaction volume. TIU was converted to trypsin inhibitor activity (TIA)

expressed as mg/ g sample (Hamerstrand et al. 1981).

Proximate Analysis

Frozen whole ﬁsh, pooled by experimental unit, were passed through a meat
grinder and homogenized by pulverizing with mortar and pestle. Ground tissues were
dried at 105°C and further ground to pass an 850 am screen. Feed and whole body
nitrogen contents were determined by the Kjeldahl method (AOAC 1990). Crude protein
was calculated as N x 6.25. Crude lipid was determined by difference following diethyl
ether extraction (AOAC 1990). All analyses were performed in triplicate on pooled

samples.

Efﬁciency Calculations

Fish within treatments were analyzed for grth and efﬁciency parameters.
Efficiency parameters evaluated were feed efﬁciency, protein efﬁciency ratio, speciﬁc
growth rate, and apparent net protein utilization. Efficiency parameters were calculated

using the standard equations of Jauncey and Ross (1982) as described in experiment 1.

Statistical Analysis
A completely randomized 2 x 4 factorial design, with a control group was

employed. Phytase treatment and level of SBM substitution served as main effects.

31

Growth, SGR, selected proximate components, and efﬁciency parameters were analyzed
as a two—way ANOVA utilizing the model statement

Yijk = 11 + Ci + Pj + CPij + Eijk
where C = SBM concentration, and P = phytase treatment.

When signiﬁcant differences were detected, means were separated using Student-
Newman-Keuls multiple range test controlling for the type I experimentwise error rate.
In addition all SBM containing diets, whether phytase treated or not, were tested against
the control group using Dunnett’s t-test. Signiﬁcant differences were reported at the 0.05
level unless otherwise indicated. To evaluate the effect of phytase treatment at each level
of incorporation, orthogonal contrasts were run between treated and untreated groups at
each level (Cody and Smith 1997). All analyses were performed using SAS statistical
software (SAS 1979). Additionally, linear regression and non-linear regression

(quadratic model) were performed on growth and efﬁciency parameters.

Experiment 4 - Phytic Acid Supplemented Fish Meal Growth Trial
Third generation Nile tilapia originating ﬁom Purdue University were used for an
eight-week trial to evaluate grth and performance of ﬁsh fed ﬁsh meal based diets

supplemented with graded levels of phytic acid.

Phytic Acid Supplemented Fish MealDiets
The ﬁsh meal control diet used in the SBM growth trial served as the formulation

for the phytic acid supplemented diets. All ingredients used were the same as previously

described.

32

Dodecasodium salt of phytic acid from corn (Sigma Chemical Co., St. Louis,
Missouri) was incorporated into the control diet in graded levels (Table 4). Levels
incorporated were equivalent to phytic acid concentrations in diets that contain SBM
providing 0, 25, 50, 75, 100, and 200 % of the CP (Table 5). Diets were formulated,

prepared, and stored as previously described for experiment 3.

Emrimental Design

Fry were reared in aquaria until reaching approximately 2.0 g. During this period,
they were fed the same standard commercial trout diet as used above. The experimental
system and conditions were the same as for the SBM growth trial.

A completely randomized design was employed with a control, and ﬁve diets
incorporating graded levels of phytic acid All treatments were run in triplicate. Eight
ﬁsh, each, were randomly stocked into 40 L tanks. Each tank was deﬁned as an
experimental unit. Experimental units were randomly assigned a dietary treatment.

Mean weight at stocking was 2.04 g (0.04 g SBM, n=144). This experiment
followed the same feeding, sampling, and experimental protocol as in experiment 3. At
initiation of the experiment, 33 ﬁsh were collected from the population at stocking and

stored at -20°C for CP, dry matter, and lipid analysis.

Statistical Analysis

A completely randomized design was employed with concentration of phytic acid
supplementation as the main effect. Growth, SGR, selected proximate components, and
efﬁciency parameters were analyzed as a one-way ANOVA (SAS 1979). When

signiﬁcant differences were detected, means were separated using Student-Newman-

33

 

 

 

.38.. ...8.... .88.. 8.88.. 98.... .38.. ...8.... .33
...mm ...mm ...mm ...mm ...mm ...2 can 83...... ..o 83.8
98 ...... .5 can ...... ...? ...? 88.8... ..o 5.8%:
an... em. 3 3 ...m t..- 1.. 25.8.2

2.8 e..: we.” .88 ..mmm 3mm 3.... 8.0.3.er
.... .... ...o .... .... .8 .8 8.8.5 ...:......o
.... .... .... .... .... .... .... 29.. secs...

...8 ...8 ...8 ...8 ...8 ...8 ...8 m 020
...m ...m ...m ...m ...m ...... ...m N 882.. £855
...8 ...8 ...8 ...8 ...8 ...8 ...8 _ ago... ....er
...mm. ...mm. ...mm. ...8. ...nm. ...2. 3.2 8.1.3. .....on
1... l-.. t... t... t... 1... 2mm 8.30 58
t... t... t... t... t... ...: t... 20.8.“ 2% 8.85m .528

68.. 5.0.. :8 28 5.8 3.8 1... 8.3.1 .82 8.5..

8~ 8. me 8 mu .8an 8...... a .ﬂm Beecher—l
—RﬂOmuHEOan—H

 

 

 

 

500 50...... .55... 5.53005“. 5.0 Se 30..» 5 .005 0.0% 3:05.098 005052.33 500 0.23.. we .858..an .v 030...

40.6 be 05 we 3% 0.3 :32... m05e> 35.05000 5.550»...
0.5 8 5.0.0508 5.3303“. 505...? 80.80 8 00.. 00.6 005-2 05 we 50508.88 5 .56.... 8?. 508.... 0.5.5 05
.... o\e com 28 do. .2. .ow .3 $5038.. Sim 5880 .05 $0... 5 8523:0800 58 5%.... 8 50.05.50 33 ..5359005

34

08.....8 ..eoeeoneo .

.w... 3.8... 60.80.308.85

we 8. ... e..... we am. .0e.x€..... we a... 35.... we ......“ 6.8.66 we ...8. ....8... we 2.. .28

o...e.§..e. we N: £502... we 2 ...: 0...... we N 68.3.3 5. .8. .m eeee 5. ...»... .... eee.> .... Sn...
J. 588.5 00.6 >6 ma ..0.. wage—3.. 0:. 60:09.0 9.2 $0.32 :0... .0503. 3.58....05 0:80... 00:08.8..00 :0... .803; 0:03. N
m 030... .

€3.80. 6 030B

35

3038 0.8: 8:365 Q2 N
géﬁaz .25on Be a 8 was 306 2mm 2 53:8 28 253 “Be 3 33.8.8 _

 

 

ii: a ._ Oz .....--- 2mm Eogm 2528
an 92 ........ 8N
as Rd 92 So 2:
3 ”no 92 can 2
3 a2 oz mm. om
...m 86 92 S. 3
ii: nz Loz o o
_ one 5 my 36 5 § 9.6 5 § 33% one”. §
835525 2% 2mm 9% 5 § 83335
aéﬁaz 38%: 38¢ 835 8:82:83 2% 2% mo .23

 

.306 3580.33 28 ogngobcoo .88 mFEu: 05 3.: 388.35 «co—9:93 Bataan—$2
05 a 852% 8?. 5:5 332“ 0833 .8 @8855. 556 358:3 305 .33 58th 5839 a co 6% 35:8 308
mat—o: a 3.: 335:3 Sim @0835 E028 .«o 20>! vovﬂm maquEooE macaw E 95 XL Scrub—50:8 28 BEE .m 033.

36

Keuls multiple range test. In addition, all levels of phytic acid supplementation were
tested against the control group using Dunnett’s t—test. Signiﬁcant differences were

reported at the 0.05 level unless otherwise indicated (SAS 1979).

Experiment 5 - Solvent Extracted SBM Digestibility Trial

Nile tilapia were obtained from Illinois State University (Normal, Illinois) and
transported to Michigan State University. These ﬁsh were spawned and reared at
Michigan State University until stocking. Second generation ﬁsh were used to determine

amino acid and total N digestibility of the SBM diets used in experiment 3.

Exmg'mental design

The mixed sex population of 0. niloticus spawned ﬁom ﬁsh obtained from
Illinois State University were held in ﬂow-through well water heated to 2711.O°C until
stocking. Fish held during this period were fed a standard commercial trout diet.

The experimental system was a 4,700 L recirculating system similarly conﬁgured
as in experiment 2. Temperature was maintained at 28°C, and ﬂow rate to each tank
maintained at 2-3 L/min. Fish were maintained on a 16:8 lightzdark cycle. Dissolved
oxygen, ammonia, nitrite, and nitrate remained within acceptable limits for tilapia
throughout the trial (Cherivinski 1982 ; Daud et al. 1988; Papoutsoglou and Tziah 1996).

The experiment was designed as a randomized complete block design. Two
replicates, for each experimental diet, were run in each of two blocks. Fifteen ﬁsh, each,
were randomly stocked into one of eighteen 125 L tanks. Bach tank was deﬁned as an

experimental unit. Experimental units were randomly assigned a dietary treatment.

37

Mean weight at stocking was 69.9 g (1.04 g SEM, n=270), and 66.9 g (0.63 g
SBM, n=270), for block one and two, respectively. Fish were allowed a four-day
acclimation period during which they were fed a commercial trout diet. Following
acclimation, the ﬁsh were fed the experimental diets for a 10-day preliminary period
The preliminary period allowed for dietary adjustment. The experimental diets were the
same SBM diets used in experiment 3 (Table 2). Fish were fed 2.4 % BW/day on a dry
matter basis, divided between three equal meals (9:00, 13:00, and 18:00). Following the
preliminary period, the experimental period ensued two hours following the aﬁemoon
feeding.

Fish were euthanized in MS-222 at a concentration of 500 mg/L (Post 1983).
Incisions were made along the mid-ventral line anteriorly from the anus. The exposed GI
tract was clamped 10 cm from the anus, and the posterior section excised All fecal
material within the excised section was collected and pooled by experimental unit.
Pooled fecal samples were stored at -20°C for subsequent analysis. Pooled fecal
samples, and feed samples, were freeze dried for amino acid, nitrogen, and hydrolysis
resistant organic matter (HROM) determinations.

Digestibility coefﬁcients were calculated utilizing the indirect method of Jobling
(1983), with HROM serving as an internal marker.

_

ADC = 100 — (100) % HROM in feed X % Nutrient in feces,
% HROM in feces % Nutrient in feedd

 

where ADC = Apparent Digestibility Coefﬁcient

38

HROM M Crude Protein

Freeze dried feed and fecal samples were analyzed for HROM with slight
modiﬁcations to a method described by Buddington (1980). A standard curve was
prepared using or-cellulose (ICN Biochemicals, Cleveland, Ohio). Approximately 50 mg
fecal samples, or 100 mg feed samples were placed in 15.0 mL of 80 % acetic acid and

1.5 mL HNO3 and gently boiled for 20 min. Samples were ﬁltered under vacuum with

glass microﬁber ﬁlter paper (Whatrnan GF/B) with a pore size of lum. Samples were
sequentially washed and ﬁltered under vacuum with 6 mL hot ethanol, 6 mL hot benzene,
6 mL petroleum ether, and 6 mL ethanol to remove residual organic solvents. Filtered
samples were dried at 105°C for 12 hr. Dried samples were weighed and ashed at 525°C
for 16 hr before re-weighing. Hydrolysis resistant organic matter was calculated as
amormt of material lost on ignition expressed as a percentage of the original sample
weight

Approximately 15 mg freeze dried fecal samples were analyzed for N with a
N-analyzer (Leco FP-2000, Leco Corp., St. Joseph, Michigan) following manufacturers

speciﬁcations. Crude protein was determined as N x 6.25.

Amino Acid Analysis

Feed and feces collected during the digestibility trials were analde for amino
acids, except tryptophan. Samples were freeze-dried, pulverized, and hydrolyzed with
6 N HCl at 110°C for 24 hrs for hydrolysate amino acid analysis. Free amino acids were
derivatized with phenylisothiocyanate (PIT C) before analysis (Waters Manual 1989).

Derivatized amino acids were determined on a C-18 reverse phase HPLC column using a

39

Waters HPLC separation system (Waters Chromatography Division, Millipore Corp.,

Milford, Massachusetts).

ﬁatisﬁal An_a_lysi_s_

The design was a randomized complete block design with all factors ﬁxed. The
data were analyzed with two levels of SBM treatment (phytase treated and untreated) and
four levels of SBM incorporation (25, 50, 75, and 100 % replacement), and a control
(0 %) in each of two blocks utilizing the model statement

Yijkl = 14 + Bi + Cj + Pk + Cij + £in
where B = block, C = SBM concentration, and P = phytase treatment. There was
insufﬁcient material for analytical analysis in some samples. Therefore, means were
tested by the least-squares estimates of marginal means (lsmeans) method, by SBM
treatment (SAS 1979).

In addition, all SBM containing diets, whether phytase treated or not, were tested
against the control group using Dunnett’s t-test (SAS 1979). Signiﬁcant differences were
reported at the 0.05 level unless otherwise indicated. To evaluate the effect of phytase
treatment at each level of incorporation, orthogonal contrasts were run between treated

and untreated groups at each level (Cody and Smith 1997).

Experiment 6 - Phytic Acid Supplemented Fish Meal Digestibility Trial
Cohorts of the ﬁsh used in experiment 5 were used to determine amino acid and
total N digestibility of the ﬁsh meal based diets supplemented with graded levels of

phytic acid used in experiment 4.

4O

Exxrimental Design

The digestibility trial was conducted using the same protocol as in experiment 5.
In addition, three replicates in each block were fed a similarly formulated N—free diet
(Table 4) to determine endogenous N and amino acids excretion. Mean weight of tilapia
at stocking was 64.8 g (0.53 g SBM, n=225) and 63.1 g (0.92 g SEM, n=225) for block
one and two, respectively. Fecal collection, sample preparation, and analyses were

carried out as described in experiment 5.

Statistical Analysis

Crude protein and individual amino acid digestibilities were analyzed as a
randomized complete block design (SAS 1979). Due to insufﬁcient material for
analytical analysis in some samples, means were tested by the least-squares estimates of
marginal means (lsmeans) method (SAS 1979). Additionally, all levels of phytic acid
supplementation were tested against the control group using Dunnett’s t-test (SAS 1979).

Signiﬁcant differences were reported at the 0.05 level unless otherwise indicated.

Experiment 7 - Tilapia Gastrointestinal Tract pH Proﬁle

Exmrimental Design

Third generation ﬁsh from those originating from Purdue University were used
for pH measurements of the GI tract. Fish were maintained in a 2000 L tank utilizing
ﬂow-through well water heated to 27°C. Fish were fed a standard commercial trout feed

containing 41.0 % CP, 12.0 % lipid, and 4.0 % ﬁber for a minimum lO-day preliminary

41

period prior to initiation of sampling. During the preliminary period ﬁsh were fed to
satiation three times daily at 4 hour intervals.

On days of sampling, ﬁsh were fed to satiation in the morning. Following the
morning feeding, ﬁve ﬁsh each were netted at random at 0.5, l, 2, 4, 6, and 8 hours post-
prandially. Additionally, ﬁve ﬁsh each were netted at random prior to ﬁrst feeding and
labeled time 0 (control group). Fish were euthanized via hypothermia, by submersion in
an ice/water slurry for 15 minutes, prior to dissection.

Euthanized ﬁsh were blotted dry, individually weighed, and measured for total
length. Incisions were made along the mid-ventral line posteriorly, exposing the coelom.
The GI tract was clamped, and severed at the esophagus anteriorly, and at the anus
posteriorly. The entire visceral cavity was excised and the GI tract teased away from
other visceral components. The external surface of the GI tract was rinsed in distilled
water and blotted dry prior to weighing.

All pH measurements were made with a ﬂat membrane pH microelectrode (model
MI-406) and external micro-reference electrode (model MCI-402) (Microelectrodes, Inc.,
Bedford, New Hampshire). Values were digitally displayed on an Accumet model 25
pH/ISE meter (Fischer Scientiﬁc, Pittsburgh, Pennsylvania) and recorded.

The intestine was clamped just posterior to the pyloric sphincter and the stomach
removed. The stomach was slit anteriorly from the pyloric sphincter to the esophagus,
and ventrally to expose the contents (Figure l). Stomach contents were analyzed on the
surface, on both sides of the mid-ventral line near the top and bottom of the bolus.
Additionally, any food remaining in the esophagus or pylorus was measured for pH.

Stomach contents were collected in 50 mL Erlenmeyer ﬂasks with 30 mL distilled water

42

Pylorus Pylorus

 

 

 

 

2
3 9
0
‘ti 5 7 g
g Ventral g
o 6 8
4 1 0
1
Esophagus Esophagus

Figure 1. Top view of stomach excised from tilapia (0. niloticus). The stomach was slit
anteriorly from the pyloric sphincter to the esophagus and ventrally to expose
the contents. View indicates 10 sites where pH measurements were made on

the surface of the gastric mucosa. The sites corresponding to 1 and 2 represent
the esophagus and pylorus, respectively.

43

and homogenized. Samples were stored at -—20°C for subsequent pH analysis of the total
contents. The mucosa of the emptied stomach was rinsed under a gentle stream of
distilled water to remove any remaining feed particles. The surface of the gastric mucosa
was measured for pH at the esophagus (l), the pylorus (2), and four sites on each side of
the mid-ventral line (3-6, and 7-10) (Figure 1).

The intestine was measured for total length prior to segmenting intolO equal
segments. Each segment was slit along its length and contents exposed. Readings were
collected on the contents at three sites along the most anterior segment (segment 1). The
ﬁrst measurement was taken immediately following the pyloric sphincter and anterior to
the bile duct. The second measurement was taken in the middle of the segment. The
third measurement was taken at the posterior end of the segment (F i gure 2). The
remaining nine segments were read at the middle of the segment. Following these
measurements the intestine was rinsed under a gentle stream of distilled water to remove
any remaining digesta Mucosal readings were taken at sites corresponding to the point

digesta readings were measured.

ﬂatistical Analysis
Analyses were performed as a one-way ANOVA with time as a ﬁxed factor. Due
to missing values where digesta was not available, means were separated using the least-

squares estimates of marginal means (lsmeans) method (SAS 1979).

44

AS - v 823 $5:st main—=8 2:: 05 we 22:8 05 8 35808 29$ 38:8 28 $52.8 3.582: .EoEwom

05 .«o 95 Steamed 05 Am €0.53. 05 Mo 0:.va 05 Am 326 25 05 8 8:35 93 SSE—Em 2.818 05 wage—.8
£03605 2 uses—mom “€85 #68 05 wee—a moi 02$ 8 88:8 98 $538 2: no 38:8 203 mmaveom

.035 29$ 3582:,“qu mg 223 aces—mom 3:585 88%.: E3832 @3382: .9 «is: mo 83 a uoﬂoxm .N 85w:—

 

 

 

 

 

 

 

 

maomcaomm

 

 

95mg... .205ch 0.85”.

45

RESULTS

Experiment 1 — Optimal Feeding Frequency

Tilapia were fed to satiation 1, 2, 3, or 5 times day ’1. Mean daily intake (MDI)
among the 4 treatments increased as the number of feedings day '1 increased. Mean daily
intake increased from 3.04 g at 1 meal day " to 5.02 g at 5 meals day 'l; however, intake
among treatments receiving 2, 3, or 5 meals day '1 were not signiﬁcantly different
(P<0.05). All treatments receiving more than 1 meal day " consumed signiﬁcantly more
feed than those receiving 1 meal day ’1. A curve drawn through the points representing
MDI indicated an asymptote representing maximum daily intake was approached at 3
feedings day " (Figure 3).

A broken line analysis can be satisfactorily ﬁtted to any response that approaches
an asymptote (Robbins 1986). Therefore a broken-line analysis was performed on MDI.
Zero is not a practical value for estimating feeding frequency, particularly for long term
applications. Therefore the model was ﬁtted to the data for 1, 2, 3, and 5 feedings day '1.
The breakpoint in the model predicted the optimum feeding ﬁ'equency to be 3.18 feedings
day '1 (Figure 3).

The cumulative feed consumed increased in a linear fashion (Figure 4). The
predictive equations best describing the increases were y = 2.97(x) +2.00, R 2 = 0.997
(1 meal day "); y = 4.09(x) + 0.93, R 2 = 0.998 (2 meals day "); y = 4.61(x) + 3.97, R 2 =
0.997 (3 meals day "); and y = 4.80(x) + 5.06, R 2 = 0.997 (5 meals day "). A slope ratio

analysis indicated those ﬁsh receiving 1 meal day '1 consumed signiﬁcantly less feed

46

.556 some 88m “5:83 beau—mama 8:

2o? .59 Go. oVB. S3 2:: — mom 5: :3 832m ragga—ma 203 .53 385 m .5 m m tom saw me moo—SE >13 32

..-»3 mmcaoom E. m 33 £89.85 05 3233:. mam—«5 2:28—65 < demon >3 mm a .83 $05 3388 3038 me

muse—mom 025 me comm—>3 33:3 E0852 man Bum. 3.3 amp—.33.“ m “a 35:. 33 8:532: mascomoaoe Sea—Emma
an 353:5.» :33 mm 3.50 23.. $3 3:5 m 8 m N J 3533 9 wow .3232: .0 was: 90 8.85 >13 :32 .m Esmi

55:08“. 9:08“.

o a § 4 «3” a m P e
F b _ m _ _ F 0.0

 

(0) oxuul Allin wow

 

 

 

47

3838 ace; 8823a .28 oz Agave _- 3 mass m a .m 3 see 55 a2 massage a: _- an as: 3 Be 3
E 828388 3:353 83865 mamas ewe 2.2m .on amenam 63.48.?» e5 3 v Resume .3? £91
x3 maouam .8? 58.?» x8 Seen? .8? 43.?» Meat; 2.22532 2: wastage a3 883.8

05 38335 mam—3a commohmom ... 33 88: m 8 .m .N ._ 3:38 8 30m 326:: .Q .3 3383 33m 02.3550 .v Esmﬁ

 

>5
mm mm «N «N ON 2. or 3. Nr 0.. m w v N o
._ a. x a x n .4 a + a a. _. a. .r .l o
.I...\»\\ 8
. . . . u. mum - o
. . . ”(in .. :8. m
. . . .uELWV- . m
. . . . nb.n1n_\\\. :8 m
. . . . .mLu .\ \n e r 8 a
t l l t l J.—
6 ﬂ 1‘"\\\ m
E ..\. -
Nb .\l l 2:.
. .\ 28:3 0 ...
x 4\‘\C . m
a a “‘0 2g” 4 ll GNP
. \ )
\\ £38" 0 m.
... l l .85. o 4o:

 

Lr cup

48

during the course of the trial than those receiving 3 or 5 meals day ". No other
differences were detected (P<0.05).

Mean daily intake partitioned on a per meal basis indicated ﬁsh receiving 1 meal
day '1 consumed signiﬁcantly more feed at the 8:00 hr feeding than did ﬁsh receiving
multiple meals during the day (P<0.05) (Figure 5). Fish receiving 2 meals day '1
consumed signiﬁcantly more feed at 8:00 hr than ﬁsh receiving 5 meals day '1, but not
more than those receiving 3 meals day '1. The amount of feed consumed during the ﬁrst
meal by ﬁsh receiving 3 meals day '1 and 5 meals day " were not statistically different
(P<0.05).

Mean daily intake for all ﬁsh at 8:00 hr was signiﬁcantly higher during the ﬁrst
week than during the ﬁnal week. Although all treatments showed decreased intake at
8:00 hr during the ﬁnal week, the greatest decrease was in the ﬁsh receiving 3 meals
day '1.

Fish receiving a second meal during the day consumed less feed during that meal
the sooner it followed the previous meal. During the second meal, ﬁsh receiving 2 meals
day '1 ($2 = 2.41 g) consumed signiﬁcantly more than did those receiving 3 meals day "
(x = 1.61 g), which was signiﬁcantly more than those receiving 5 meals day "

(>7 = 0.93 g) (P<0.05). Similarly during the third meal ﬁsh, receiving 3 meals day " (x =
1.51 g) consumed signiﬁcantly more than ﬁsh receiving 5 meals day " (>‘< = 0.88 g).

All treatments exhibited an increase in growth expressed as a percent of original

wet weight (Figure 6). Mean increase in weight was 60 %, 49 %, 36 %, and 9 % for ﬁsh

fed 3, 2, 5, and 1 meal day '1, respectively. Weight increase in ﬁsh fed one meal a day

49

.35 x33 .58 a mats—e 30m 838m 33:3. 8.3 we conned—38 38 05 88052 33> menace.“
:3 an commasmeoo :88 88 30:33 _- .33 388 m 8 .m .N J 309 326:: .9 Me 83.: b3 :32 .m oBmE

menus—.95 mﬁeoom

m n N n
- 8.:

- 84
- 3."
T 84”
l 86

r 8%

 

(3) metal ﬁlled "89w

IL

186
00K... oonmrl Gong—U Gonor- comm.

 

L

- 8K

50

 

 

50w-
5o-» .
-°-1moal ’0
/
“ .1- -D 'mes /
g ‘0 -¢-3moals /
8 -o-5moals : '
i 304-
C
a 201. /
3 ° /
.. c/
3 10 . /
/
o . ’ - . .L :
3 4
-10
WEEK

Figure 6. Weight gain in 0. niloticus fed to satiation 1, 2, 3, or 5 times day '1. Fish fed 1
meal day ’1 exhibited a signiﬁcantly lower increase in weight than ﬁsh fed
multiple meals day '1, which were not signiﬁcantly different ﬁ'om each other

(P<0.05).

51

was signiﬁcantly less than in those fed multiple meals (P<0.05). The remaining
treatments were not signiﬁcantly different from each other (P<0.05).

The initial and ﬁnal proximate compositions of the ﬁsh are summarized in Table
5. Fish fed 3 meals day ’1 contained signiﬁcantly more lipid and GE than ﬁsh fed either
1, or 5 meals day ", but were not signiﬁcantly different in GE than ﬁsh fed twice day ",
even though they contained signiﬁcantly more lipid (Table 6). Fish fed 3 meals day ’1
contained signiﬁcantly less CP than the other treatments (P<0.05). Fish fed 1 meal day '1
had signiﬁcantly more CP than ﬁsh fed 5 meals 'l, but were not different from ﬁsh fed 2
meals day '1.

Performance and efﬁciency parameters of the ﬁsh are summarized in Table 7.
Fish fed 2, 3, or 5 meals day ‘1 were not different from each other in performance as
measured by total weight gain, SGR, and FE (P<0.05). Although ﬁsh fed once day "
were not signiﬁcantly different from ﬁsh fed 2 or 5 meals day '1 in terms of total weight
gain, they did exhibit a signiﬁcantly lower SGR and FE (Table 7).

The PER indicated ﬁsh fed 2, 3, or 5 meals day '1 were not different from each
other, but all retained signiﬁcantly more protein than ﬁsh fed once day ‘1. The ANPU
among the treatments mirrored PER.

Fish fed once day '1 performed signiﬁcantly poorer in terms of ER (P<0.001).
Energy retention was signiﬁcantly better in ﬁsh fed 3 meals day ‘1 than those fed 5 meals
day '1, but was not signiﬁcantly different from those fed 2 meals day '1.

Multivariable analysis indicated there was a signiﬁcant correlation between
visceral weight and total wet weight, with an overall correlation coefficient of 0.67

(Table 8 (A)). The relationship was stronger among females than males when the

52

 

 

 

 

a an: 8.2 13.83.” 18.8 3% 82v 8.: e :38 3.: m
1%.: 2 .2 e 8on 36 c Acne 2% 5.8 3.? age 3?. m
a 2ch N2: e Gas e; a :08 choc ans 3.: a. A38 8.: m
e Gas 2;: c 82: can e 698 2.3 see 3S c £2: 8.8 _
:8:
G3: 8.2 82: an.” 3.3 5% :3: a? 8qu 8.: was...
3a 3: En 3e @3 55:3;
n2 2&4 53on 830 25682 3.8m $80 wagon
.chvnc accents enema»:

33%.: 5:28 a 55.3 autoﬁoaa Bogota .Amuav moan—am “co—eon no «$83 8823.: no gmmv 28 508 05 28352
82g 325 $5882“ 88:9: 025 88... ATE moan—am “go—eon :o 388 88:35 mo A58 and :38 05 88292
828/ 2&5 558 be a no 39892 03 35:88.5 =< .35 gem 333.58 a maﬁa—c _- >3 88: n .8 .m .N J

sonata 383% 8 3% Bob 395888 Bataan a com 3.382: .0 .«o 5:52—88 Sagxoi 33 223 Ram 23 3:5 .0 033.

53

 

 

ca 8:8 2 .2 e23: 2.8. a 92.8 2: e c _ .8 as a 8M8 NE a $28 8.2 m
n 25.8 3.8 e 328 8.8 e a. .8 m? e 398 8.0 e 898 «3 e E”: 3.2 m
e 5.8 8.2 a 5.0: :8 e § .8 R; c 398 N2 1282.— a 888 a .2 N
1:882 033:: 22883 c€o8£o 1:8»: c3333” _
$8 88 85338 32.88 @5822
272 mm mmm mm 2% 58 8833 268..”

 

A898 c.8828 38%? 2885 :58 e :85

3828895 82285 .3553: 28:82 025 82..“ SHE 8888 388 .8 A359 Ba :88 05 8223.. 3:3, .35
5.52» 8.83.88,“ a manic _- .98 88: n .8 .m .m ._ 8338 3 v2 828:: .Q E 5.23 883:8. 8088 8: 828%
can .38 82822 $85 Adm—ac 28.. 85850 8088 Ammv 85.650 32 .3me 82 532» 058% 48% Emma}, s 033.

54

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8de $85 ”338
885 9:3 .023
22 .o 823 “$53 158$ 23. :2 mm Sum
~22 3285 386 2:. 2d 93 m 682
o§mmum
3o: Egan 05599 m 3.? m 94 mm mm 83%
mBaEumm “80:83.“ occur—Sr .«o mam—SE
A6
885 5:3 @6on
886 886 3250
8:5 58¢ $63 38. 8.0 3.: 3 85
28.0 a8§£ 88d 8.? 33 3.9. N 382
u§mmum
$82 EASE 2%E> m 83> m an .mlm a 85%
maﬁa—5mm “Begum 3§tu> «a mam—92
3

2% a 88.: gm 8 .85: .03. .28 203 38:35 «ma—:50.“ mo ..ng .mwﬁvoom 58v mo “38:: v5 ...oncow .2903
3383 MO .833 a 3 2E?» Egmﬁcoa AB 98 mmwaqvoom as me 598:: 98 53% Jaws; 63 :33 mo corona
a ma Emma? 388$ 2v maﬁa—axe €282on 033.3222: 05 Ba 838:8 56883 93 838 <>OZ< Baougm .w 293.

55

analysis was separated by gender. The effects of gender and total weight were
confounded with each other. Males were signiﬁcantly larger, but females had a
signiﬁcantly higher wet visceral weight. However, the female visceral weight was
comprised primarily of gonadal tissues, whereas male visceral weight was comprised
primarily of liver and intestinal tract. The number of feedings a day had no signiﬁcant
effect on the total wet visceral weight.

Visceral weight, gender, and number of feedings day '1 had a signiﬁcant effect on
the amount of perivisceral fat (Table 8 (B)). The overall correlation was low, but was
greater for males than for females. When separated by gender, there was no correlation
among females, whereas the correlation among males was highly signiﬁcant. Males had
signiﬁcantly more perivisceral fat than females. Fish fed once day '1 had signiﬁcantly

less perivisceral fat than ﬁsh fed 3 or 5 meals day '1.

Experiment 2 — Tilapia Gastric Evacuation Rate

The rate of gastric evacuation and passage through the GI tract was followed
using Fe203 as an inert marker. There were no signiﬁcant differences in iron
consumption or total consumption between the treatments. Few ﬁsh dissected at time 0
(prior to feeding) had digesta remaining from the previous day’s feeding. The digesta
that was found, was principally localized in segments 5 - 7. No digesta was found in the
gastric region. Where digesta was found, it was analyzed for iron and the results used to
correct for background iron

The contrast between the green chromic oxide marker and red ferric oxide marker

was readily discernable during dissection. At 2 hrs postprandially, iron was observed

56

throughout the GI tract. At 4 hrs postprandially, a red! green mixed bolus was noted in
segment 4. At 6 hrs postprandially, the mixed bolus was observed in the terminal
segment for both treatments. In ﬁsh fed 5 meals day '1, stomachs at 24 hrs were full of
red and green digesta, whereas the stomachs in ﬁsh fed 3 meals day " were empty and
ﬂaccid. Visual inspections were veriﬁed by analytical determination.

The initial rate of gastric emptying for both treatments was similar during the ﬁrst
hour (Figure 7). Following the ﬁrst hour, GER of ﬁsh fed 5 meals day " was slower than
that of ﬁsh fed 3 meals day '1. Gastric evacuation rate for both treatments was curvilinear
and could be described by the exponential function VT = We 'bo‘); where VT = volume
of feed at time T, V0 = volume of feed at time 0. The equation describing GER for ﬁsh

fed 3 meals day " was VT = 67.0 e ”"530” with a correlation coeﬁ'rcient of R2=O.9O, and

for ﬁsh fed 5 meals day '1 was VT = 85 .0 e ””1490“ with a correlation coefﬁcient of

R2=0.97. A t-test following natural log transformation of cumulative iron suggested the
slopes were not signiﬁcantly diﬁ'erent (P<0.05).

The rate at which iron appeared at the termiml segment of the intestine was
slightly more rapid for ﬁsh fed 5 meals day 4 over the ﬁrst hour (Figure 8); however, the
treatments were not signiﬁcantly different. The amount of iron fed appearing in the
terminal segment, between the 2 hr sampling and the 4 hr sampling, and between the 4 hr
sampling and the 6 hr sampling, was signiﬁcantly higher in the group fed 3 meals day '1
(Figure 8). Conversely, the increase between the 12 hr and 18 hr samplings was
signiﬁcantly higher in the group fed 5 meals day ’1 (P<0.05).

The rate at which iron appeared at the terminus was linear over the ﬁrst eight

hours postprandially (Figure 9). A slope analysis indicated the rates were not

57

.n- be $583 a 3on

. 339 o 9.8 Ba A _- 5 “Museum a 8.3m .32.? o 9% a 88:88 a“ $5.58: 95 2a 5 8.8

acreage 059% 25. .mEcoom wage—Em m5 vm “888:8 no: 39 8 0358 m_ =0.: 0333850 .89: use 692 692
.82 .8qu 8:33 a _. 3. agave m s .89: ea .82 .8qu Saga 8 _- 5 “maﬁa m sea 3 so: .3

.o 08: 8 ace—SE 35886 nu ma 58m mew—:38 36 Ecosﬁoaxo 5 wow 3.28:: .0 E gang—85 059% he Sam .5 053m

350: _a_u:u.auao$ 25...

up or 3. N9 0.. a o v N o

 

 

 

. r «mi . .................................................................................................. , 8
.8532 ...IDL

- 35.32 nIIQIW. E.

i. 1 . ...... on

...................................................................................... 1 ca

 

 

........................... °°F

 

(%) uou amelnumo

58

.68“. use? 8%
3:5 so 8238 wave .838? .333 was 2: 332.8 .89: ea .8“: .32 .82 .8an 8:33 3
_- 5 aaBe m 8 .89: as .32 .83 8:83 2 _- 5 an?“ m .220 3 so; a: .o 25. a 5:8. 3585
5 8 68m 365:3 8.6 358530 :a vow can 3 8.5 R2535 $823.3 .0 .«0 33.58 a 85.33% no: mo 3am .w 253m

$3: 155.88% 8.:

 

 

 

 

 

(%) uou annemuma

 

 

59

.mE vm :9: 380:8 :0: 33
8 02:22 a 3: 3:53:50 guy—3:: 350:8 :a ma 58m 363:8 6% 358598 :a Mo 3608 05 ago—3.:

mac: £90 6:.“ 05 mans: 8a.: 3:585 $928.5 .0 mo Eoﬁwom 35:5: 05 “a @525: :9: S 3885 30:3 d onE

.23: 36:89.3“: 8.:

 

a s c m e n a F o
F _ r _ b _ p _ O
. I 2
O O
85.833 . \n. I on
35.83 n o T 8
...\ T. 3

S? u N:

8
(%) no." eAgelnumo

 

o S3 u N:

60

signiﬁcantly different. The time required for 90 % of the recovered iron to appear in ﬁsh

fed 3 meals day “ was 3 hrs, and for ﬁsh fed 5 meals day '1 was 13 hrs.

Experiment 3 — Solvent Extracted SBM Growth Trial

Graded levels of phytase treated SBM, or untreated SBM, were substituted for
FM and in diets fed to juvenile tilapia to evaluate the effects on growth, efﬁciency, and
body composition. Experimental diets were formulated to contain 33 % CP. Analysis
showed actual dietary CP ranged from 32 — 36 %. No relationship was observed between
growth or performance and dietary CP level.

There was a linear increase in phytic acid with increasing levels of SBM
substitution in the untreated diets (Table 5). All diets prepared with ﬁshmeal and phytase
treated SBM had phytic acid concentrations below detectable limits. Trypsin inhibitor
activity in the solvent extracted SBM was 2.8 mg/ g SBM.

During the eight week experimental period, ﬁsh fed the phytase treated SBM, and
untreated SBM, exhibited similar grth patterns. Increase in weight for ﬁsh fed the
phytase treated SBM diets ranged from 402 - 771 %, for the 100 % and 25 % CP as
SBM, respectively (Table 9). Weight gain in ﬁsh fed the untreated SBM diets ranged
from 441 —— 731 %, for the 100 % and 25 % CP as SBM, respectively (Table 9).

Fish grew slightly better with addition of SBM at 25 % CP regardless of SBM
treatment; however, the overall trend was toward lower weight gain with increasing
incorporation of SBM (Figure 10). Diets containing phytase treated SBM resulted in
signiﬁcantly lower growth as the percentage of CP as SBM increased above 25 %

(P<0.05). In contrast, diets containing untreated SBM did not result in signiﬁcantly

61

Table 9. Weight gain, whole body crude protein, whole body lipid, and moisture in
juvenile tilapia fed diets containing untreated, or phytase treated, solvent
extracted SBM substituted into a reference diet at 25, 50, 75, or 100 % of the

dietary protein. Values represent the mean of three replicates of pooled samples

(n=8). Different superscripts in a column, within the phytase treatment, or
within the untreated treatment, represent signiﬁcant differences within that

treatment (P<0.05). Asterisks represent signiﬁcant differences from the control

 

 

 

 

diet (P<0.05).
Weight Gain Crude Protein Lipid Moisture
(% Increase) (% DM) (% DM) (%)
____C°ntr°| 694 56.09 17.61 74.32
Phytase
Treatment
25 % SBM 771 a 58.70 17.51 74.43
50 % SBM 495 b 60.70 * 14.45 75.17
75 % SBM 561 b 58.41 14.24 74.88
100 % SBM 402 b ’ 61.65 ' 15.10 75.04
Untreated
Trgtment b
25 % SBM 731 a 54.75 ° 19.45 72.67
50 % SBM 671 “b 58.21 b 15.48 73.44 “b
75 % SBM 490 3" 59.60 3" 13.55 74.00 a
b * a * a
100 % SBM 441 61.56 15.36 74.47

 

62

.Godve ea 3:8 ea 3 2522 6E 532» ashes 386536 a mega: 3:. 38.5686 3865 $632
Essa Sea 286 as 2: 2a .2. an .2 a 6.6 .838 65 35 33:33 2% 62856 .828 .38: 683%
8 £030.23 3:888 $0.6 Eeoﬁtunxm 40% .2250 a 3 9522 $0.6 358595 new «as: 2.825 Ce Eu» 2303

3.80:. 25.0 3. 5.3.0:. .3! :uoaaom

 

 

 

................................................................................................................................................ .. 6.8-
..................................... r 0.0..-
uﬁn— DOSE:-

66 3.8:- 98.

 

 

............................................................... r 98.
............................................................... T °eO F:

r 0.0

(MP 10mm 30 %) urns inﬁrm

r 06—.

 

................................................................................................................................................ l°°N

 

.2 2:5

63

lower growth until the percentage of CP as SBM surpassed 75 %. Orthogonal contrasts,
between phytase treated and untreated diets at the same level of SBM substitution,
indicated the diets incorporating SBM at 50 % of the CP were the only diets resulting in
signiﬁcantly different growth (Figure 11). Growth was not signiﬁcantly reduced relative
to the ﬁsh meal control diet, for either treatment, until SBM provided all of the dietary
protein (Table 9).

Proximate components in ﬁsh fed the phytase treated SBM diets were not
signiﬁcantly different from each other. However, relative to ﬁsh fed the control diet, ﬁsh
fed diets containing 50 and 100 % of the CP as phytase treated SBM contained
signiﬁcantly more CP (Table 9). Similarly, ﬁsh receiving the diet containing untreated
SBM incorporated at 100 % of the CP contained signiﬁcantly more CP relative to the
control group.

In contrast to the phytase treated diets, ﬁsh fed diets containing untreated SBM
exhibited increasing CP (%DM) with increasing dietary SBM. Fish fed the diet
containing 25% CP as SBM had signiﬁcantly less CP (%DM) than the remaining
treatments (Table 9).

Whole body lipid levels (% DM) ranged from 13.55 — 19.45 %. None of the
treatments were signiﬁcantly different from each other, or from those fed the control diet.
Moisture followed a similar pattern to CP.

Speciﬁc growth rate, feed efficiency, protein efﬁciency ratio, and apparent net
protein utilization are summarized in Table 10. Speciﬁc growth rate ranged from 2.86 —
3.86 %/day. Relationships and signiﬁcant differences within the diets of each SBM

treatment mirrored those of weight gain. However, unlike weight gain, the orthogonal

aoo , a Treated Diet
' IUntmted Dlat

Weight Galn (96 Increase)

§

 

§

 

SBM Substitution (%CP)

Figure 11. Orthogonal contrasts of weight gain in juvenile tilapia fed diets containing
untreated, or phytase treated, solvent extracted SBM substituted into a
reference diet at 25, 50, 75, or 100 % of the dietary protein. Values represent
the mean of three replicates of pooled samples (n=8). Error bars represent
SEM. Different labels within a level of substitution represent signiﬁcant
differences (P<0.05).

65

Table 10. Speciﬁc growth rate (SGR), feed efficiency (FE), protein efficiency ratio
(PER), and apparent net protein utilization (ANPU) in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM substituted into
a reference diet at 25, 50, 75, or 100 % of the dietary protein. Values represent
the mean of three replicates of pooled samples (n=8). Different superscripts in
a column, within the phytase treatment, or within the untreated treatment,
represent signiﬁcant differences within that treatment (P<0.05). Asterisks
represent signiﬁcant differences ﬁ'om the control diet (P<0.05).

 

 

 

 

SGR FE PER ANPU
(% / daY) (%)
9.22M! 3.68 0.79 2.52 36.79
Phytase
Trgtment
25 % SBM 3.86 a 0.82 a 2.53 a 38.65 a
50 % SBM 3.37 b 0.69 a 2.08 b * 32.15 b
75%SBM 3.18b 0.71 a 2.09“ 31.19“
b * b * c * c *
100 % SBM 2.86 0.57 1.58 25.17
Untreated
Treatment
25 % SBM 3.77 a 0.84 a 2.51 a 38.32 a
50 % SBM 3.63 ab 0.85 a 2.49 a 39.40 a
75 % SBM 3.17 “b 0.74 a 2.21 a 35.34 a
b* b* b* b*
100 % SBM 2.99 0.62 1.83 29.91

 

66

contrasts between phytase treated and untreated diets at the same level of SBM
substitution indicated no signiﬁcant differences in SGR (Figure 12).

Feed efﬁciency values ranged from 0.57 to 0.84 (Table 10). There were no
differences in FE between the ﬁsh fed the control diet and diets containing 75 % or less
of the CP as SBM. Fish fed the diets containing 100 % of the CP as SBM exhibited a
signiﬁcantly lower FE than the other treatments (P<0.05). Contrasts between the same
rates of SBM incorporation indicated only the two diets at the 50 % level were different
from each other (P<0.01), with the ﬁsh receiving the untreated SBM performing better
(Figure 13).

Protein efﬁciency ratio and ANPU exhibited similar trends (Table 10). Among
the groups fed the phytase treated SBM diets, signiﬁcant differences were detected in
PER with incorporation of more than 25 % of the CP as SBM. The ﬁsh receiving diets at
the 50 % and 75 % rates were more efficient than ﬁsh receiving the diet with all the CP
supplied by SBM (P<0.05). Signiﬁcant differences were not detected in PER or ANPU
among the groups receiving the untreated SBM until the level of incorporation reached
100 % ofthe CP as SBM.

Relative to the control diet, ﬁsh receiving the phytase treated diets showed
signiﬁcantly lower PER with more than 25 % of the CP as SBM, and signiﬁcantly lower
ANPU with more than 50 % of the CP as SBM. Orthogonal contrasts between the same
rates of SBM incorporation for PER indicated only the ﬁsh fed the 50 % level were
different from each other (Figurel4). However, contrasts between the same rates of SBM
substitution for ANPU indicated signiﬁcantly lower ANPU among ﬁsh fed the phytase

treated diets with more than 25 % incorporation of SBM as CP (Figure 15). Results of

67

4.0 1

D Treated Diet
I Untreated Diet
3.6 ~
g
6
3 3.0 .
I!
o
«n
2.5 ~
2.0

 

 

 

 

 

25 50 75 100
SBM Substitution (%CP)

Figure 12. Orthogonal contrasts of speciﬁc growth rate (SGR) in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM substituted
into a reference diet at 25, 50, 75, or 100 % of the dietary protein. Values
represent the mean of three replicates of pooled samples (n=8). Error bars
represent SEM. No signiﬁcant differences were detected between treatments
at any level of substitution.

68

0.9 *2
‘ El Treated Diet

a a

l a ‘ I Untreated Diet
0.8 J

3 a

a

1 b
0.7 l a

1 a
0.8
0.5 l“ a, «aw - ---.

25 50

75 100

Feed Efﬁciency

 

 

SBM Substitution (%CP)

Figure 13. Orthogonal contrasts of feed efficiency in juvenile tilapia fed diets containing
untreated, or phytase treated, solvent extracted SBM substituted into a
reference diet at 25, 50, 75, or 100 % of the dietary protein. Values represent
the mean of three replicates of pooled samples (n=8). Error bars represent
SEM. Different labels within a level of substitution represent signiﬁcant
differences (P<0. 0 l ).

69

3.0 -

   
  

U Treated Diet
l Untreated Diet

 

Protein Efﬂclency Ratio

 

SBM Substitution (the?)

Figure 14. Orthogonal contrasts of protein efﬁciency ratio in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM substituted
into a reference diet at_25, 50, 75, or 100 % of the dietary protein. Values
represent the mean of three replicates of pooled samples (n=8). Error bars
represent SEM. Different labels within a level of substitution represent
signiﬁcant differences (P<0.05).

70

8
o

8
o

Apparent Net
Protein Utilization (%)

 

 

a
a a
D Treated Diet
a I Untreated Diet
' b
1 b a
1
25.0 i1 » . e.— .
25 50 15 100

SBM Substitution (960?)

Figure 15. Orthogonal contrasts of apparent net protein utilization in juvenile tilapia fed
diets containing untreated, or phytase treated, solvent extracted SBM
substituted into a reference diet at 25, 50, 75, or 100 % of the dietary protein.
Values represent the mean of three replicates of pooled samples (n=8). Error
bars represent SEM. Different labels within a level of substitution represent
signiﬁcant differences (P<0.05).

71

the linear and non-linear regression analyses for growth and efficiency parameters are

given in Appendix 3.

Experiment 4 - Phytic Acid Supplemented Fish Meal Growth Trial

Experimental diets were formulated to contain 33 % CP and graded levels of
phytic acid. As analyzed, the diets contained 32.6 — 35.4 % CP. Phytic acid
concentrations increased linearly (Figure 16).

Weight gain ranged from 430 — 560 % for diets supplemented with Na-phytate at
12.9 and 25.8 g/kg dry diet, respectively (Table 11). An ANOVA indicated ﬁsh fed diets
incorporating Na-phytate at 3.4 g/kg dry diet, and 25.8 g/kg dry diet, grew signiﬁcantly
slower than the ﬁsh fed the diet incorporating 12.9 g/kg dry diet, whereas the other
treatments did not (P<0.01). The only diet which resulted in signiﬁcantly slower growth
than the control diet was the diet incorporating Na-phytate at 25.8 g/kg dry diet (Table
11). There were no signiﬁcant differences among any of the treatments in ﬁnal whole
body CP, whole body lipids, or moisture.

Speciﬁc growth rate followed a similar trend as weight gain (Table 12). Fish fed
the control diet, the diet incorporating Na-phytate at 9.7 g/kg dry diet, and the diet
incorporating Na-phytate at 12.9 g/kg dry diet, performed signiﬁcantly better in terms of
SGR than the diet containing Na-phytate at 25.8 g/kg dry diet (P<0.01). There were no
other differences in terms of SGR (Table 12). Additionally, there were no differences in

the performance parameters FE, PER, and ANPU.

72

Dietary Phytic Acid (glkg dry diet)

 

 

 

000 I T 1 I T I I I I I j I I I

0 2 4 6 810121416182022242628

Ila-phytate Supplementation (glkg dry diet)

Figure 16. Dietary phytic acid concentrations in ﬁsh meal based experimental diets

supplemented with phytic acid as Na-phytate.

73

Table 11. Weight gain, whole body crude protein, whole body lipid, and moisture in
juvenile tilapia fed a ﬁsh meal based diet supplemented with graded levels of
phytic acid as Na-phytate. Also shown is the level of SBM incorporation
(% dietary CP) providing the equivalent phytic acid concentration. Values
represent the mean of three replicates of pooled samples (n=8). Values within
a column with different superscripts are signiﬁcantly different (P<0.01).

 

 

Na-Phytate SBM Weight Crude Lipid Moisture
Supplementation Equivalent Gain Protein (% DM) (%)
(Egg dry diet) (% dietary CP) 1%) «Va DM) ,

0 0 519 3" 52.76 20.26 75.29

3.4 25 460 b° 52.69 19.90 75.50

6.5 50 490 at” 53.71 20.38 74.78

9.7 75 505 at” 54.06 19.38 74.73

12.9 100 560 a 54.56 20.02 75.73

25.8 200 430 ° 53.67 17.87 75.62

 

74

Table 12. Speciﬁc grth rate (SGR), feed efﬁciency (FE), protein efficiency ratio
(PER), and apparent net protein utilization (ANPU) in juvenile tilapia fed a
ﬁsh meal based diet supplemented with graded levels of phytic acid as
Na—phytate. Also shown is the level of SBM incorporation (% dietary CP)
providing the equivalent phytic acid concentration. Values represent the
mean of three replicates of pooled samples (n=8). Values within a column
with different superscripts are signiﬁcantly different (P<0.01).

 

 

Na-Phytate SBM SGR FE PER ANPU
Supplementation Equivalent (%/day) . (%)
(543g dry diet) (% dietary CP)

0 0 2.94 a 0.75 2.20 28.86

3.4 25 2.73 ab 0.62 2.09 27.05

ab

6.5 50 2.83 0.72 2.02 27.88

9.7 75 2.89 a 0.75 2.16 29.41

12.9 100 3.01 a 0.80 2.17 29.25

25.8 200 2.60 b 0.58 1.99 26.23

 

75

Experiment 5 - Solvent Extracted SBM Digestibility Trial

Crude protein and individual amino acid digestibility coefﬁcients were
determined for tilapia fed diets incorporating graded levels of phytase treated, or
untreated SBM. The apparent crude protein digestibility (ACPD) of the ﬁsh meal control
diet was 81.7 % (Figure 17). The ACPD of the phytase treated SBM diets ranged from
78.9 - 84.6 %, and the untreated SBM diets from 80.9 - 82.5 %. There were no
signiﬁcant differences among the treatments.

It should be noted that the mean for the phytase treated SBM incorporated at
100 % of the CP was based on only two values. Two replicates resulted in coefﬁcients of
221 % and -13 % due to problems in HROM recovery. After testing for outliers, by
examining the standardized residuals, they were excluded from the calculation
(Montgomery 1991).

The nitrogen value obtained from ﬁsh fed the N-free diet in experiment 6 was
2.32:0.72 % N/g DM. This value was used in an attempt to correct the ACPD to true
crude protein digestibility. However, the coefficients ranged from 97.7 — 103.4 %
indicating this may not be an appropriate procedure.

The apparent digestibility of individual amino acids is summarized in Table 13.
There was a high SEM among all the amino acids in the group fed the diet containing
phytase treated SBM as the sole source of protein.

The apparent digestibility of Ala was signiﬁcantly lower in the phytase treated
diet incorporating SBM at 100 % CP than the other diets. Additionally, apparent

digestibility of Lys in this diet was lower than that of the ﬁsh meal control diet.

76

100-

ACPD 1%)
a

75 ~

 

70-

Figure 17.

ITreeIedDiet
IUntreetedDht

82.1

82.3

    

SBM Substitution (%CP)

Apparent crude protein digestibility (ACPD) in juvenile tilapia fed diets
containing untreated, or phytase treated, solvent extracted SBM substituted
into a reference diet at 25, 50, 75, or 100 % of the dietary protein Values
represent the mean of four replicates of pooled samples (n=15), except the two
diets with 100 % substitution, where insufﬁcient material limited the analysis
to two and three replicates for the treated, and untreated diets, respectively.
Error bars represent SBM No signiﬁcant differences were detected

among the treatments.

77

 

 

563 A30 :30 363 sad

2“ 5.: at. a: ”.8 6536
$23 26.3 893 Gad 55

E 2” ea 99 a...» 28 28820
$6.5 88.3 $3 cane 5e

9% me. at at ﬂ: 65685
$2: 5 .3 a»: 53 Sad

one New v.8 as 6.8 82.82
695 an: 5.3 can and

9% a: a: a? 6.3 28 sesame.

383 :83 A36 and as:

e 3:. a v.2. a QR a n: a 98 8.52
N 868. _ some _ 8% x mm 65
3325 gem owing 383m cococomom Eo< OPE—an.

 

Agave sesame 856%? as

38 a 888 macaw—case. «cocobﬁ 53> 83.5 .38865 233 E88 Am TS 8388 “cc—eon .«o 88232 See me Gama

98 S88 85 88852 828/ .5885 .0586 85 we e\e o2 .8 .3 dm .3 8 8.6 3588.. a 85 383333 2mm 388.58
2828 €385 3.8.3“. 8 888.55 wages 38.6 new nae—a 8.833 E ﬁeomowmooo bzﬁumowﬁ 28 8.58 “moan—9‘ .2 833.

78

 

 

:28 § .9 Sea avg :83
3.6 S: can one 6Q. 2.263825
526 5.3 and A38 A8. 3
new 2: 2K QR 8.2. 8.8862
same 22V sod 26.8 at:
e 6.8 82 a; ”.8 a 2» 6563
Game 88.3 Gad $98 an:
36 SE at QR SK 6583
some and § .3 55 God
In 868 Sq. ..m. a: 6:88.62
Gee: 23V And sad see
3.6 NE 3: 8.: at. 65888
888_ 882. 88% .x. mm 65
6223 83:: 682: 6835 8868a 22 65:2

 

.8883 2 2%."

79

 

 

36.5 825 and and are

9% NE .81: TR 6.65 85>
8.5 36$ 9.3 see 55

ea. v.8 6.8 «.8 «.8 6585
83.6 688 cod 5.0 Add

new. a? 3.8 36 one 6:685
$63 5.3 26.8 And 82¢

36 6.2. 2:. «see 6.: Seem
Game :35 Rec 863 8an

6.8 $8 38 men «.68 6565
868. as? 88% as R 65
68:5 8&5 Masai 35 8666a 22 65.5.

 

6.663 a 635

8O

 

 

81

SE A _ we § d 36.9 96.3

“.2. QM» QR SR «R 6:620
ans 5: .3 36.: Red 5.3

:6 New 63 Q8 Q8 666 65620
26.6 5.3 5.8 58.3 63

SR 36 NR Q3 5: 6:685
82¢ 2 we 8Q : QQQ Q: .3

eew Q3 36 2.» Q3 866:.
26.3 888 Ed 86.0 and

QR NS RR QR 6.5 2.6 6.66668.
5 .9 ans Gad 3.3 35.:

NR 08 QR ”.8 Q8 6552
:2: 6R 86% .68 65

_ 38253 _ 60:85.5 605253 _ 38253 3:88qu Ro< 05:2

 

968V .2 635

 

 

63 $3: 893 and ﬁns
E 3” 3K a: 25 85.225
5.3 £5 58 A3: ﬁn:
as 53 NR 2; 2% 85.53:
SE 82¢ 8...: and ac:
o3 New ”.3 Sm 3» 2&3
$3: 33: $5 $3 an:
it New SK 3» 2K 0583
82: at: $5 and God
m2. 2» m2. NR ”3 868.2
A85 55 SN: awe avd
a; 3m 3m 2» nt 8685
:8. :2. :8 :8 85
Baobab 3305:: Egbnb @3825 vacuumed 29¢ 8.82

 

APESV 2 2%...

82

$8823.. 28 no 33328 mos—Er N
$8823“ 025 no 33:28 833/ _

 

 

3.3 :33 5.8 93.9 :03

Q: as 35 3:. 35 as;
3.8 82: 5.3 93.3 God

3o mﬂ as. a? v.8 0585
$8 83: so: at: 5 .3

2: ca 2K no» 03 8285
and 5.3 an: Gas 82;

E» a ”.2 2K 3K 3 a: 85m
83: 55 cos 89$ emu

NR 3» NR 0% NS. 8:3.
$§ {E econ $3 85

33225 332:5 380.55 “.8855 8:20mom Eo< oEE<

 

.9253 2 2.3.

83

Conversely, apparent digestibility of Ser in the group fed the control diet was lower than
that in the group fed the diet containing untreated SBM incorporated at 75 % of the CP.
Results of the orthogonal contrasts between phytase treated and untreated diets at
the same rate of SBM incorporation are displayed in Table 14. The only signiﬁcant
differences detected were between the groups fed the two diets containing SBM at 100 %
of the CP. Alanine, aspartic acid, arginine, glycine , histidine, lysine, methionine,
threonine, and valine digestibilities were all lower in the group receiving the phytase

treated dict.

Experiment 6 - Phytic Acid Supplemented Fish Meal Digestibility Trial

Crude protein and individual amino acid digestibility coefﬁcients were
determined for tilapia fed ﬁsh meal based diets incorporating graded levels of phytic acid,
as Na-phytate. The ACPD of the phytic acid supplemented diets ranged from 75.8 % for
the diet supplemented with 25.8 g Na-phytate/kg dry diet to 87.5 % for the diet
supplemented with 12.9 g Na-phytate/kg dry diet (Figure 18). There were no signiﬁcant
differences detected between the supplemented diets and the control diet, nor among the
supplemented diets. The relatively low coefﬁcient and large standard deviation for the
treatment containing Na-phytate at 25.8 g/kg dry diet was due to one replicate with an
exceptionally low digestibility coefﬁcient.

The nitrogen value obtained from ﬁsh fed the N—free diet was 23210.72 % N/ g
DM This value was used to correct the ACPD to true crude protein digestibility. The
coefﬁcients ranged ﬁom 95.6 — 100.0 %. ANOVA indicated a signiﬁcant difference

between the two blocks in N recovered in the feces of the ﬁsh fed the N—free diet

84

Table 14. Orthogonal contrasts between the means of apparent amino acid digestibility
coefﬁcients in tilapia fed diets containing untreated, or phytase treated, solvent
extracted SBM substituted into a reference diet at 25, 50, 75, or 100 % of the
dietary protein. Signiﬁcant differences are shown with resultant P values.

 

 

 

SBM Substitution (%)

Amino Acid 25 Vs 25 50 Vs 50 75 Vs 75 100 Vs 100
In is usable
Arginine __ _ __ 0.0260
Histidine __ __ __ 0.0500
Isoleucine _ _ __ ____
Leucine __ _ _ __
Lysine __ _ _ 0.0054
Methionine _ __ __ 0.01 12
Phenylalanine _ __ __ _
Threonine __ _ _ 0.0275
Valine __ __ __ 0.0321
DI} pgngble
Alanine __ __ _ 0.0285
Aspartic acid __ __ __ 0.0321
Cysteine __ __ __ _
Glutamic acid _ __ __ __
Glycine _ __ __ 0.0315
Proline _ __ __ __
Serine _ __ _ _
Tyrosine

 

85

82.5

75.8

ACPD (%)

          

 

0 3.4 6.5 9.7 12.9 25.8

Ila-phytate Supplementation (glkg dry diet)

Figure 18. Apparent crude protein digestibility (ACPD) in juvenile tilapia fed ﬁsh meal
based diets supplemented with graded levels of phytic acid as Na-phytate.
Values and error bars represent the mean and SEM of four rcplicates of

pooled samples (n=15).

86

(P<0.002). In one block, the mean fecal N value was 2.961015 % N/g DM, and the
other 1.68:0.25 % N/ g DM, indicating feces from one block may have been subject to
contamination more than the other.

The apparent digestibility of individual amino acids is summarized in Table 15.
Signiﬁcant differences were detected among all the amino acid except Ala and Asp.
The apparent digestibilities of Gly and Ser were signiﬁcantly higher in the group fed the
diet supplemented with Na-phytate at 12.9 g/kg dry diet than all the other treatments.
Apparent digestibility of He was higher among this group than those receiving diets with
Na-phytate at 3.4, 6.5, and 25.8 g/kg dry diet. Proline apparent digestibility was also
higher in this group than those receiving the diet supplemented 25.8 g/kg dry diet. Other
than the above exceptions, the apparent digestibility of amino acids was only higher in
the group fed the diet supplemented with Na-phytate at 12.9 g/kg dry diet than the groups

receiving diets with Na-phytate at 3.4 and 25 .8 g/kg dry diet.

Experiment 7 - Tilapia Gastrointestinal Tract pH Profile

GI tracts in tilapia were sampled for pH following a 10 day preliminary period
during which the ﬁsh were fed a standard commercial trout feed containing 41.0% CP,
12.0% lipid, and 4.0% ﬁber. Values for pH were recorded at 10 locations in the stomach,
and 12 locations in the intestine. Measurements were made on the surface of recovered
digesta ﬁ'om the stomach (Appendix 4); on the surface of the gastric mucosa (Appendix
5); and intestinal mucosa (Appendix 6) over an eight hour period following feeding. The
pH values of the mucosa along the intestine were similar to those of the feed; however,

this was not true of the gastric mucosa and feed recovered in the stomach (Figure 19).

87

 

 

awe 89$ Sad 80: etc sad
an: a :5 a 2» no: a 2:. pm; 8620
3. c and and :39 Ed 2a.:
a o...» a can a an» a 3.» a 2; a new 28 25320
as: 5.3 :2 .3 §.8 33v 3”,:
a QR a as a v.5 a ...$ 9 3a a an» 8635
5d Gm: 5d 30.8 and 3.:
a was a an a 3» a :8 a is a 3» Same
Gm : cad $53 an: 863 5 d
3.” «8 3» 2w 2» «a 28 0.832
6an sad sad and 30$ 98.9
3% 0.8 «8 QE an. in 8E2
«.8 ad 2 3 E” .05 caesium 22. 8E2.

 

90% be macaw 83858235 Bannez

Amodvnc Bogota mangwmm 2e 38 a $88 argon?“ EEobE 5.5,
mos—axr ASHE 838$ woo—eon .3 £58.32 Sou me 958 v5 :88 of 28652 32.3 .3539a2 we 28 can.“ mo
£32 35» as, 3552.53 age 83; .88 a: E “as: 220%.. a £96608 $5.303 28 can; 828%. .2 23.

88

 

 

E: 23.: 3.: 23.: 8a.: 28.:

a 2.8 a «.3 a ...? a n? a E... a 2:. 82322822

33.: 86: Ge: a»: G3: a... :

a 8% a $2 2. SK 2. S: a 35 a 9S 85262

as: an: Ev: 5.: 23.8 am:

a 3» a 2.8 a 3» a. :3 a OS 2. 0.8 223

E .3 so: 92.: 3.: at: 6::

a 0.2. a 4.3 a as a 2:. a a? a 5» 86:3

as: an: 32: an: 3.: an:

a ”do a .2.» 2. <2. 2 n: .2 So a $5 028292

E .3 5.: 82 .m: at: an: 28.:

2 ed. a «.3 a $5 a 9:. p 2.: a a: 059.22
”.8 mm. 3 3 an .22 8.22202 22 22:2.

 

so? be 93: catacoanazm Banaéz

Apps—o8 n— 033.

89

 

 

35: as: 8...: an: an: 28.:

a n: a can a at. a ...8 a a: a 8.2 88>

so: 83.: am: 34.: 5.3 an:

2 23 a 35 a OS a 3.0 a 6% a 36 0585

89: AS: a}: 3.8 23.: as:

a *2. a as a «.2. a a: a 0.? a 8.8 82:85

9n: ns..: 82.: GE: 2%.: a»:

a as a .8 a «.3 a 3: a .3 a 2.8 2.5m

3.: so: 3.: 8°: :2: a: .m:

we? a 8.3 a 9R 2. 3k 2. a: a 3: 22.82
«an ad 2 3 an 58 85.202 22 22:2

 

32: be wéw: 838580—83 Banaez

 

3.808 2 033.

9O

53.30nt .8388 m can .N 80852 533 m

and _ Eoﬁwom 0.5me 5088 .8388 m .«o 5:356 235$ 05 28052 .932 Sam .33.: Snow? 3553.. mo costs.“

05 no v5 «838 05 mo coatsm 05 no 038 203 35823802 d530,“ wage 2.8 5 9:325. 52. 95 A852
EcumSEV ng 05 E 8382 S 23 .AcoEB 0583 :88on 95 5 £838. 2 3 v3.58.— 82? In :32 .2 are

22.2.20
ﬁteoasmmvnuveaahom can.
._____.._.___.HH_.._._8._n a."
«on 8» 8.... 2.” «a... 8.” , a."
T3
-3
,3

..mméméé -2

 

.353. .238...

 

T ad
1 o.»
T ad
I a.»
I a.»
a: 8.5 a: no.» on.» p: 2..» I a.»

c253. 0580 r a.»
h 2

emu Hd

 

 

 

9]

The mean pH at each of the 10 sites measured on the feed was signiﬁcantly lower than
the means of the mucosa] pH measurements at 2, 4, and 8 hours postprandially
(P<0.001).

At 0.5 hr following feeding there were no regional differences in pH values
measured on the surface of the digesta. In subsequent samplings, regional differences did
exist. Generally, the differences observed were between one or more sites in the fundic
region, and the site labeled site 9 (Figure 1).

Statistical analysis indicated pH of food recovered from the esophageal, and
pyloric regions, at 2 and 4 hours following feeding were signiﬁcantly higher than in the
gastric region (P<0.05). The differences were most pronounced relative to the fundic
region. Range and mean pH values of the stomach digesta and gastric mucosa are
reported in Table 16. There was a signiﬁcant decrease in pH of the digesta between 1
and 2 hours following feeding (P<0.05). The pH remained low until it began to increase
again at 6 hours following feeding, after which it again decreased The pH of the gastric
mucosa was signiﬁcantly lower as quickly as 0.5 hours following feeding (Table 16).

The decrease in pH of the mucosa closely paralleled that of the digesta; however, the
decrease was not as pronounced as seen in the digesta.

After measuring pH on the surface of the feed, digesta was collected and analyzed
for pH following homogenization. There was little change in the pH of the homogenized
digesta samples. After one hour of digestion, pH measurements taken at the surface of

the digesta were considerably lower than the homogenized samples (Figure 20).

92

Table 16. Range and mean pH values from feed in the stomach and gastric mucosa in 0.
niloticus before feeding, and 0.5, l, 2, 4, 6, and 8 hours post-prandially. Values
in parentheses represent the standard deviation of eight values collected from
ﬁve ﬁsh at each sampling period. Values within a column with different
superscripts are signiﬁcantly different (p<0.0001).

 

 

Post-prandial Range Mean Range Mean
Time (hrs) of Feed of Feed of Mucosa of Mucosa
Samples Samples Samples Samples
Before feeding N/A N/A 5.60-7.28 6.66 (0.13) a
0.5 4.62-6.33 5.36 (0.17) 3 5227.22 6.12 (0.16) b
1.0 3.96-7.30 5.22 (0.45) a 5.66-8.33 6.92 (0.19) a
2.0 2.75-4.68 3.65 (0.13) ° 4.37-5.85 5.03 (0.19) d
4.0 2.84-4.42 3.52 (0.34) ° 4.11-6.31 4.94 (0.15) d
6.0 2.35-5.29 4.22 (0.32) b 4.25-6.71 5.59 (0.31) °
8.0 2,724.54 3.62 (0.19) ° 3.19-6.28 4.77 (0.21) d

 

93

.veton miaﬁam :08 8 380:8 amm o>m

86 2mm 2: aosaa 23 Sum 2.6655qu 8:8 a 9a a .4 .N ._ .3 2a .953 2&3 868% 2: sea 38:8
8838 8380383 28 £8886 05 88m 888:8 8838 no 8883 .8858 0.58» 05 no 38808 83g mg 832 .8 089m

 

 

 

 

 

 

 

 

 

94

The mean pH values measured in each of the 12 segments in the intestine are
presented in Appendix 6. The pH values in the intestine remained relatively consistent

throughout the sampling period, both within a segment, and with time.

95

DISCUSSION

Biological availability of nutrients in ﬁsh is dependent on a number of factors.
Some factors affecting availability to a given species can be categorized as biological
(requirements, size, age, and physiological state), dietary (quality and quantity of protein,
energy, and processing of components), environmental (temperature, DO, water quality,
and photoperiod), and management (stress, feed intake, and feeding frequency).

Fish culture has historically emphasized maximizing intake and growth, which
may not be the most effective approach for ﬁsh production (Seymour, 1989; Kaushik
1990). This approach oﬁen leads to uneaten feed and lower conversion efﬁciency, which
reduces water quality. This is particularly problematic in recirculating systems where
low water exchange occurs. Reduced water quality adversely effects the system, and
diminishes organism health and performance (Spotte 1970; Goddard 1995; Saddiqui and
Al-Harbi 1999).

The general maxim in regards to feeding strategies for tilapia has been “little and
often” (Jauncey and Ross 1982). This principle is grounded in early work done on
captured wild ﬁsh. Based on feeding behavior, physiology, and GI morphology of wild
tilapia, it was reported they require many, frequent meals to achieve greatest efﬁciency
(Moriarity 1973; Moriarity and Moriarity 1973a,b; Balarin and Hatton 1979; Kubaryk
1980; Caulton 1982). In general, ﬁsh that eat small particles on a continuous basis have
small stomachs (Smith 1989). However, tilapia stomachs have the ability to distend to a

large size and function as a storage unit (Fish 195 l ).

96

Misconceptions from this early work still persist in regards to the structure,
function, and relevancy of the tilapia stomach (Avault 1996). However, a review of the
literature indicates there is a sufﬁcient collection of evidence to refute these
misconceptions (Al-Hussaini and Kholy 1953; Fish 1960; Moriarity 1973; Kapoor et al.
1975; Bowen 1976; Bowen 1982; Smith 1989; Boujard and Leatherland 1992; Yamada et
a1 1993).

Wild 0. niloticus preferentially graze on blue-green algae and bacteria (Bowen
1982). Filtering of algae is energetically costly. Wild tilapia exhibit a higher speciﬁc
dynamic activity associated with the seeking and processing of food relative to farm
raised ﬁsh Therefore wild ﬁsh must consume more food to cover this energetic cost
(Gerking 1994).

Blue green algae and bacteria contain about 50 % CP; however, assimilation
efﬁciencies of this material is low in tilapia. Assimilation efﬁciencies range from
50 % to <1 %, with values near 15 % common (Bowen 1982).

Fish reared in intensive production systems have different requirements than those
in the wild. In such systems, natural food is limited. All nutrients must be exogenously
supplied in the form of high nutrient dense pellets.

Omnivorous species, such as tilapia, are readily trained to eat these nutrient dense
diets. The higher quality and consistency of pelleted diets should obviate the need for
frequent feedings. Consequently, there is a need for optimal fwding regimens to
accommodate this capacity to process formulated diets. The fwding regimen should

include an optimum ration, delivered at a rate and frequency that maximizes efﬁciency.

97

Greatest growth rate is realized at the maximum ration, but the optimal ration,
deﬁned by the highest utilization efﬁciency, is at a submaximal level. From a practical
and economic standpoint, satiation feeding is the best method for both the ﬁsh, and the
production unit (Tidwell et al. 1991). With satiation feeding, the optimal ration will be
deﬁned by the feeding frequency that yields the greatest utilization efﬁciency. To
determine this frequency, tilapia were fed to satiation 1, 2, 3, or 5 times day '1 to evaluate
total consumption and utilization efﬁciency.

Mean daily intake among ﬁsh fed 3 and 5 meals day “ was similar to the
suggested feeding rate for the size of ﬁsh and temperature used in this experiment
(Jauncey and Ross 1982; NRC 1993). The MDI approached an asymptote at 3 feedings
day ‘1, which was veriﬁed by broken-line analysis (3.18 feedings day ’1). A similar
response was observed at 4 feedings day " in 5 g Nile tilapia fed to satiation 1, 2, 4, or 8
times day '1 (Kubaryk 1980). An asymptotic response has also been described for sea
bass ﬁngerlings (Tsevis et al. 1992), rainbow trout (Grayton and Beamish 1977; Bergot
1979), channel catﬁsh (Andrews and Page 1975), African catﬁsh (Singh and Srivastava
1984), European eels (Seymour 1989), rockﬁsh (Kono and Nose 1971), and ﬁleﬁsh,
puffer ﬁsh, and yellow tail (Ishiwata 1969 a;b). The response is temperature dependent,
with lower temperatures resulting in lower optimal frequencies (Seymour 1989).

Fish fed once a day consumed more feed during the morning feeding than ﬁsh fed
more than once a day (Figure 5). This was also demonstrated in striped bass (Powell
1972), channel catﬁsh (Andrews and Page 1975), African catﬁsh (Singh and Srivastava
1984), and in ﬁleﬁsh, puffer ﬁsh and yellowtail (Ishiwata 1969 a,b). Similarly,

increasing intervals between feedings led to increased meal intake in winter ﬂounder

98

(Tyler and Dunn 1976), sockeye salmon (Brett 1971), and rainbow trout (Grove et al.
1978). Fish eat available food depending on stomach fullness, and at intervals
determined by the rate of gastric emptying (Holmgren et al. 1983).

Increased feeding frequencies can decrease aggressive social behavior in some
species, resulting in increased grth rate and reduced size variation (Jobling 1994).
This has been demonstrated in rainbow trout (Grayton and Beamish 1977; Holm et al.
1990), and eels (Seymour 1989). However, there is a limit to the frequency of feeding
beyond which an increase in growth is negligible (Ishiwata 1969; Bergot 1979; Tsevis et
al. 1992). In the hybrid tilapia (0. mossambicus x 0. niloticus), growth was not
signiﬁcantly different with increasing frequency beyond twice a day (Siraj et al. 1988).

Growth over the four week period in the current study was slightly higher in ﬁsh
fed 3 meals day ", although not statistically different from the other groups fed more than
once a day (Figure 6). Consumption was also not statistically different. These results are
similar to those observed in rainbow trout (Grayton and Beamish 1977), sea bass (Tsevis
et al. 1992), and winter ﬂounder (Tyler and Dunn 1976), where growth closely paralleled
intake.

Kubaryk (1980), also working with Nile tilapia found signiﬁcant differences in
growth among all groups except those fed the two highest frequencies, 4 and 8 times
day '1. The diets used in Kubaryk’s study were similar to those used in our study.
However, the nutrient density was diluted by adding an additional 33.6 % water before
' feeding (Kubaryk 1980). Total consumption was signiﬁcantly higher in ﬁsh fed 4 times
day ‘1 than ﬁsh fed twice day ". However, it was not reported if consumption was on an

as fed basis, or on a dry matter basis, making direct comparisons difﬁcult.

99

Although ﬁsh eat to meet their energy requirements, there is a physical limit to
the amount of bulk that can be consumed (Boujard and Leatherland 1992). Kubaryk
(1980) suggested bulk limitations may have limited consumption in ﬁsh fed the lower
frequencies.

Additionally, the ﬁsh evaluated by Kubaryk (1980) were smaller (4.8 g) than the
ﬁsh used in our study (34.5 g). Smaller ﬁsh have a higher requirement for protein and
energy. Fish will increase their feeding frequency to maintain a constant energy intake
when diets are diluted (Jobling 1980). Fish fed at the lower frequencies may not have
been able to obtain sufﬁcient protein and energy to maximize growth.

Kubaryk (1930) evaluated ﬁsh fed at 2 and 4 feedings day ". If3 feedings
day “ had been evaluated it may have been the optimal feeding frequency. This
assumption is supported by plotting the data from their study; both consumption and
growth plots indicated an asymptote would have been reached at 3 feedings day ".

The ﬁsh fed 3 meals day 4 in the current study contained the highest level of GE.
This was due to higher whole body lipid levels since these ﬁsh also contained the lowest
whole body CP. No differences were detected in proximate components in rainbow trout
fed diﬁ‘erent feeding frequencies (Bergot 1979). However, the analysis of proximate
components was performed on carcass instead of whole body. Feeding frequency
differences did exist in the liver and viscera; therefore, a whole body analysis may have
resulted in signiﬁcant differences.

Grayton and Beamish (1977) also working with rainbow trout evaluated whole
body composition The patterns in moisture and lipid were similar to those observed in

our study. The differences they detected were not statistically signiﬁcant. The

100

researchers suggested feeding frequency may not have a direct effect on body
composition (Grayton and Beamish 1977). However, rainbow trout generally grow
slower than tilapia, and the inability to detect signiﬁcant differences may have been due
to the short term of their study.

Specific grth rate and FE were lower in ﬁsh fed once day '1 than the other
feeding frequencies evaluated in this study. This was also observed in red tilapia (Siraj et
al. 1988) and eels (Seymour 1989). Conversely, SGR and FE were not signiﬁcantly
different among rainbow trout fed one meal day ‘1 to satiation, or multiple meals day " to
satiation (Grayton and Beamish 1977). Unlike tilapia and eels, rainbow trout consumed
as much in one feeding as in multiple feedings and this was reﬂected in grth and
efficiency.

Energy retention in ﬁsh fed 3 meals day '1 was high, and signiﬁcantly greater than
in ﬁsh fed 5 meals day '1. There was a drop in efﬁciency parameters in ﬁsh fed 5 meals
day " compared to those fed 2 or 3 meals day ’1. A decreasing trend in efﬁciency with
increasing feeding frequency beyond optimal was also observed in catﬁsh (Singh and
Srivastava 1984), and rainbow trout (Grayton and Beamish 1977). In comparison, a
plateau in efﬁciency, but not a decrease, was observed in red tilapia (Siraj et al. 1988)
and sea bass (Tsevis et al. 1992) fed feeding frequencies deemed to be beyond optimal.

Ofojekwu and Ejike (1984) proposed an optimal ration and feeding regimen
should take into account a combination of the interplay between feed conversion ratio
(FCR) and SGR. Feed eﬂiciency is the inverse of FCR and can be used for determining
the optimal feeding frequency in a similar manner. Evaluating these parameters for ﬁsh

fed in the present study suggests the optimal feeding frequency for both parameters is

101

either three or four feedings day '1 (Figure 21). This is similar to the general
recommendation of 4 - 5 times day '1 for ﬁngerlings, and 2 — 3 times day '1 for adult

(> 100 g) tilapia (Jauncey and Ross 1982; NRC 1993), and 4 times day '1 for Nile tilapia
(Kubaryk 1980). However, this is more than the 2 times day '1 recommended for red
tilapia (Siraj et al. 1988) suggesting there may be species differences among the tilapia

The determination of an optimum feeding regimen should be evaluated from two
aspects; 1) physiology of the species, and 2) economics of the aquaculture production
unit (Tsevis et al. 1992). The economics of the production unit would suggest 3 feedings
day '1 is superior due to the cost of labor. Although signiﬁcant differences could not be
detected, the efﬁciency parameters may still have biological signiﬁcance. In a
recirculating system where nutrients accumulate over time, ER and ANPU should
warrant as much consideration as growth.

Caution must be exercised in interpreting the results of this study. The data
suggest 3 feedings day “ is optimal for 0. niloticus, but it should be noted the feedings
were given at 4 - 5 hour intervals during the course of the day. The interval between
feedings may be a more important determinant than the total number of feedings. Tsevis
et al. (1992) suggested the interval between feedings was responsible for differences
observed in sea bass consumption, growth, and performance. A management strategy
utilizing a longer daily feeding period, particularly during the summer, may beneﬁt from
additional feedings if spaced at 4 hour intervals. The optimal interval between feedings
will vary with factors inﬂuencing the return of appetite (Gwyther and Grove 1981), and is
related to the capacity of the stomach and the rate of digestion (Brett 1971; Seymour

1989).

102

2.0 -

1.0 4

0.5 ~

Specific Growth Rate
(%Iday)

 

0.0

  

1.5 . -----------------

 

_h-----

 

‘ ﬁ‘-----

a
N
u

Feeding Frequency

 

 

Feed Efﬁciency (FE
O
a

 

 

 

4------

 

 

.J

&
CI

2 3
Feeding Frequency

Figure 21. (A) Speciﬁc grth rate (SGR), and (B) feed efﬁciency (FE) as a ﬁmction of
feeding frequency in 0. niloticus fed to satiation 1, 2, 3, or 5 times day “. The
relationship for SGR is described as SGR = - 0.2209x2 + 1.5187x — 1.0476
(R2 = 0.999), and for FE as FE = - 0.0832x2 + 0.5595x — 0.3027 (R2 = 0.979).

103

Understanding rate of digestion and its relationship to GER is important for
determining the return of appetite. Making food available as soon as appetite has
returned can maximize intake, and increase efﬁciency. Gastric evacuation rate is a
function of temperature, ﬁsh weight, meal size, dietary energy level and composition, and
feeding frequency (Windell et al. 1969; Grove et al. 1978; Flowerdew and Grove 1979;
Grove and Crawford 1980; J obling 1980; Hofer and Schiemer 1981; Holmgren et al.
1983). Demonstrating a consistent relationship between stomach fullness and appetite
will allow an optimal feeding interval to be predicted from factors that control GER

A number of mathematical models have been developed to describe GER, time
for total gastric evacuation (GET), and prediction of consumption. There are two
principal groups of models; volume dependent and surface-area dependent. The two are
similar in describing the relationship between food left in the stomach and time. Both
models describe a curvilinear relationship, but differ in that they are based on different
assumptions.

Volume dependent models (VDM) are based on volume of feed ingested as it
effects distension of the stomach, and modiﬁes the rate at which food is evacuated
(J obling 1981). Volume dependent models describe a relationship where the larger the
original volume, the greater is the initial rate of emptying. Gastric evacuation curves are
linearized by square root transformation of stomach residuum and plotted against
postprandial time. The rationale is that distension of the stomach initiates peristaltic
contraction, and the circumferential tension is proportional to the radius, therefore being

proportional to the square root of the volume of the residuum (J obling 1981 ).

104

Surface-area dependent models (SDM) are based on the assumption that the
surface area of the food inﬂuences digestion and evacuation. Gastric evacuation curves
are linearized by an exponential transformation of stomach residuum and plotted against
postprandial time. The rationale being digestive enzymes attack the outer surface of the
food, therefore the rate of digestion is proportional to the particle surface area (J obling
1981). The rate of evacuation is predicted to be higher for a meal consisting of small
food particles than large particles, for a given size meal.

Plotting gastric evacuation curves for the data obtained in this study indicate
curvilinear relationships (Figure 7). The data from these gastric evacuation curves were
subjected to both square root transformation, and natural log transformation. The
transformed data were plotted against postprandial time, and the two models subjected to
least squares estimates of the residuals. Both feeding frequencies were best described by
the exponential function VT = Voe MO. The slopes described by the term b represent the
instantaneous rate of gastric evacuation (Figure 22). The instantaneous rate for the two
feeding frequencies in this study were nearly identical at 0.153 (3 meals day '1) and 0.149
(5 meals day '1), and were not different from each other.

The accuracy of the model is dependent on two assumptions. The ﬁrst
assumption was the F e203 marker was evacuated from the stomach at the same rate as the
rest of the meal. The second assumption was 95 % evacuation is the practical limit to
which GER and GET can be applied in a culture setting. Since there is no end-point on a
semilogarithrnic plot, a practical limit must be applied (Grove and Crawford 1980). The

information of value to a culturist is not what point the stomach contains zero residuum,

105

£5383 we

aoaa 38352 as announce a: .o can 3 .o mace 2: .9- an ameeooa a Borax .93 + go: .o. u a can 5- an

anemone u gonna .83 + an: .o. u a one 328822 2: wannabe woe: oﬂ .meeooa manage E em 38:8

:9: .33 3 2522 a no: 3.33850 .892 new 692 692 .992 698 8:33. 3 _- >3 mmﬁvoou n 8 .892

98 692 693 5.5323 3 _- >3 mwemooou m .220 com 203 can 6:5 Regen “min? 6 08.5 8 cod—ES 3:35
5 3 58m meageox. 5% 358530 5 com 39:02: .9 82m 380:8 no: 023383 .8 83 Baguio—gum .3 053m

5.3:. 2:: 38:238..

 

     
 

3. 3 3. «w or a o v N c
_ _ a r e _ a _ _ C
r _. .l
0
mr m
- N m
m.
m
A.
- n N
m
u
.lll nae_ue£m_u l v mm

 

35.83 a o

 

106

but at what point appetite returns. Therefore, 95 % evacuation seems a reasonable point
to estimate the practical time at which the gastric residuum is brought to zero.

Fish were the same size (weight), and the variables diet (quality and energy),
temperature, amount to satiation, and particle size were constant between the two
treatments. The only factor that was altered was the frequency at which the two
treatments were fed. The instantaneous rate of evacuation would be expected to be the
same since consumption at ﬁrst feeding was the same.

The two curves are offset by the difference in the constant V0, which represents
the maximum volume at time 0. The model suggests ﬁsh fed 5 feedings day ’1 had a
higher maximum volume. However, maximum stomach volume is proportional to
weight, and when ﬁlled, the stomach stretches isometrically (Flowerdew and Grove
1979; Jobling 1980; Holmgren et a1. 1983). Since ﬁsh between the two treatments were
the same weight, maximum volume would be predicted to be the same. The likely
explanation for this incongruity is that ﬁsh fed 5 feedings day " were fed a subsequent
meal 2 hours following the ﬁrst meal. This corresponds to where the two gastric
evacuation curves begin to deviate. The net effect was to ﬁll the stomach again and shift
the gastric evacuation curve.

The curvilinear relationship would predict a faster evacuation of the gastric
contents with multiple feedings. The rate of evacuation will be faster with greater food
volume in the stomach. This was observed in dab (Fletcher et al. 1984), and catﬁsh
(Andrews and Page 1975). However, this was the opposite observed in bluegill (El-
Shamy 1976), as it was in our study. The rate of evacuation appeared to slow for the

treatment receiving 5 meals day '1 (Figure 7).

107

In some species, food ingested leaves the stomach on a ﬁrst in —— ﬁrst out basis
(Fletcher et al. 1984). This does not appear to be the case with tilapia. Tilapia possess
the ability to by—pass the fundic region of the stomach; food passing directly ﬁ'om the
esophagus to the pylorus and into the intestine (Moriarity 1973 ). Ingested food retained
in the stomach passes down the ventral side and accumulates in the fundic region. At this
point it comes in contact with acid-secreting cells localized to the ventral face of the
mucosa (Moriarity 1973). Ingested food by-passing the stomach passes to the intestine
without the beneﬁt of initial hydrolysis and mixing.

Assuming gastric evacuation is incomplete two hours postprandially, it was
conceivable some of the newly ingested food given to ﬁsh receiving 5 feedings day '1
would pass to the intestine without the beneﬁt of initial hydrolysis. In this context, two
scenarios can be envisioned. The ﬁrst scenario would be, food passing directly to the
intestine undergoes inefﬁcient digestion resulting in lower utilization efﬁciency at the
same level of consumption. This was the indication in the feeding frequency study where
ﬁsh fed 3 meals day ‘1 consumed as much as those fed 5 meals day ‘1, but were more
efﬁcient in converting nutrients. Gwyther and Grove (1981) determined there is a
signiﬁcant positive correlation between feeding frequency and meal size with stomach
emptying time. Gastric evacuation rate increases with more frequent feedings leading to
decreased digestion efﬁciency (Brett et a1 1969; Powell 1972; Tsevis et al. 1992). There
are a number of reports of less efﬁcient digestion and utilization in species fed at short
intervals (Tsevis et al. 1992).

The second scenario would be, feed from the initial feeding is unable to pass to

the intestine because of the newly ingested feed. In this case it would appear as though

108

the feed were being evacuated more slowly because the initial feed is unable to pass to
the intestine. The feed ingested two hours postprandially is not following the ﬁrst in —
ﬁrst out dictum. It would follow then, that a greater portion of ingested Fe203 would
remain in the stomach. This was the indication, with a greater percent of ingested iron
leaving the stomach and appearing at the terminus in the group fed 3 meals day “,
following the feeding at two hours postprandially. Additionally, ﬁsh fed 5 meals day '1
were observed to have residual Fe203 in the stomach at 24 hours, where those fed 3 meals
day '1 did not.

Grove et al. (197 8) estimated that it requires 15 hours for rainbow trout to
evacuate a l % BW meal. According to the evacuation curve used, it was estimated 80 -
9O % evacuation would require 6 hours, which they felt corresponded to the return of
appetite (Grove et al. 1978). The evacuation curves constructed ﬁom the data collected
in our study would predict 8 hours are required to attain 80 % evacuation.

Fish should be fed when appetite has returned, and not before (Grove et al. 1978).
Fish receiving meals at 2, 3, 4, or 5 hour intervals are predicted to have evacuated 27, 36,
64, and 70 % of their initial meal to satiation, respectively. Conversely, during their
ensuing meal they consumed the equivalent of 52, 58, 73, and 69 % of their original
meal, respectively, to once again reach satiation (Table 17). Fish being fed at intervals of
4 -— 5 hours appear to be consuming as much as they have evacuated. This would imply
ﬁsh receiving meals at 2 — 3 hour intervals are evacuating more quickly than predicted, or
feed is being fed before it can be efﬁciently utilized.

Seymour (1989) suggested the optimal feeding frequency in eels is one in which

the feeding interval corresponds to the volume and rate of emptying of the stomach.

109

Table 17. Hours between feedings, predicted gastric evacuation (%) from evacuation
curves, and consumption at the ensuing meal as a percent of satiation.

 

 

 

 

3 Feedi gg
Predicted Consumption
Hours Between Meals Gastric Evacuation( %) (% Satiation)
4 64 73
5 70 69
5 Feeding;
Predicted Consumption
Hours Between Meals Gastric Evacuation( %) (% Satiation)
2 27 53
2 27 49
3 36 58
2 27 55

 

110

Assuming the optimum interval between feedings is the point at which evacuation of the
previous meal is matched by consumption, then the optimum interval between fwdings
for 0. niloticus appears to be 4 — 5 hours, depending on the energy and composition of
the diet. This is slightly longer than the previously reported food passage rate of 2.5 — 3.0
hrs at 30°C in ﬁsh feeding on phytoplankton (Hargreaves et al. 1988).

It was determined from the growth and efﬁciency parameters evaluated in the ﬁrst
preliminary trial that 3 feedings day '1 leads to optimum growth and efﬁciency in tilapia
when fed over a 12 hour day. Additionally, the 4 — 5 hour interval implied by this
feeding strategy corresponds to return of appetite in tilapia. Therefore, this feeding
management strategy was employed during all subsequent studies evaluating the ANF
phytic acid.

There are a number of ANFs in SBM reported to decrease growth and efﬁciency
in ﬁsh These include proteinase inhibitors (Krogdahl et al. 1994; Olli et al. 1994),
antigens (Kaushik et al. 1995; Rumsey et al. 1995), alcohol soluble components (Amesen
et al. 1989; Olli and Krogdahl 1995), lectin and agglutinin (Hendricks et al. 1990),
oligosaccharides (Ramsey et al. 1995), and phytic acid (Spinelli et al. 1983).

Effects of phytic acid on mineral availability in ﬁsh are well known. They are
documented in rainbow trout (Spinelli et al. 1983; Cain and Garling 1995; Riche and
Brown 1996; Ramseyer et al. 1999), channel catﬁsh (Satoh et al. 1989), carp (Hossain
and Jauncey 1993), and tilapia (McClain and Gatlin 1988). Decreased protein
digestibility of diets supplemented with salts of phytic acid led to depressed grth and
poor performance in rainbow trout (Spinelli et al. 1983), Chinook salmon (Richardson et

al. 1985), and carp (Hossain and Jauncey 1993). Cain and Gatling (1995) found rainbow

lll

trout fed diets pretreated with phytase, to hydrolyze phytic acid, exhibited superior
growth relative to diets without pretreatment. Increased performance was attributed to
improved protein quality, as has been demonstrated in terrestrial animals (Atwal et al.
1980; Satterlee and Abdul-Kadir 1983; Mroz et al. 1994; Martin et al. 1998; Sebastian et
al. 1997). In this investigation, tilapia were fed graded levels of phytase treated SBM, or
untreated SBM, substituted into a FM based diet to determine the effects of phytic acid
on growth, performance, and digestibility of CP and individual amino acids.

The overall growth and SGR during this study was slightly better than those
reported for similar size tilapia, fed similar diets (Wee and Shu 1989). Weight gain
among ﬁsh fed diets containing SBM exhibited a distinct trend. There was a lower
percent gain with increasing SBM inclusion, regardless of treatment. Fish fed diets
substituting phytase treated SBM, or untreated SBM, at 25 % of the CP, resulted in
higher growth than the FM control, although the differences were not signiﬁcant. The
same outcome was observed in red drum (Reigh and Ellisl992), and Mozambique tilapia
(Jackson et al. 1982).

A small inclusion of plant protein in tilapia diets results in increased grth and
efficiency. A 25 % incorporation of the legume Phaseolus aureus (DeSilva and
Gunasekera 1989) resulted in increased growth. However, decreased performance
occurred at the 50 % incorporation rate. The researchers concluded Phaseon aureus
could be substituted up to 37 % of the dry diet without deleterious effects. Substitution
of copra, groundnut, SBM, and rapeseed, at 25 % of the CP increased weight gain and

performance of Mozambique tilapia over a PM diet (Jackson et al. 1982). Increased

112

growth and performance was attributed to improved IAA patterns with plant protein
incorporation at 25 % of the CP (Dabrowski et al. 1989).

There was a linear decrease in weight gain, relative to the control, with inclusion
of SBM beyond 25 % of the CP. Correlation coefficients for this trend were 0.94 and
0.74 for the untreated, and phytase treated diets, respectively. However, weight gain
relative to the control diet was not signiﬁcantly depressed until SBM was substituted at
100 % CP.

A large depression in growth occurred in ﬁsh fed the phytase treated diet
incorporating SBM at 50 % of the CP relative to those fed phytase treated SBM at 25 %
and 75 % substitution. It is unclear why this occurred. However, Jackson et al. (1982)
reported similar ﬁndings at the same level of substitution suggesting these observations
may have biological relevance and are not spurious.

Signiﬁcant differences in growth and performance were observed in tilapia fed
diets of varying CP (15 -- 36 %), with graded levels of SBM up to 100 % substitution for
FM (Davis and Stickney 1978). The differences occurred at the lower protein levels, but
not the higher protein levels, and only among the diets with 100 % CP as SBM.
However, the diets used in their study contained considerably more energy and Met than
those used in the current study. Methionine levels in the diets used by Davis and
Stickney (1978) were supplemented to match the level in their 36 % CP FM diet. It is
possible at the lower CP levels some other IAA, such as Lys, became limiting.

In contrast, tilapia fed a low protein diet (24 % CP) substituting 33, 67, and 100 %
of the PM with SBM only showed signiﬁcant differences relative to the FM control diet

when the diet contained SBM as 100 % of the CP (Shiau et al. 1989). It was suggested

113

the decreased performance was due to a deﬁciency in IAA requirements other than Met
(Shiau et al. 1989). The differences in growth, FE, and PER occurred with, or without
Met supplementation (Shiau et al. 1989). The researchers concluded SBM could replace
up to 67 % of the FM in a 24 % CP diet. This level of substitution was equivalent to 37
% of the dry diet, which was similar to the 50 % CP substitution in the current study.

Davies et al. (1989) reported they could substitute SBM or soy protein
concentrate (SPC) up to 75 % of the FM in their tilapia diets (50 % dietary CP) without
any detriment to growth or performance. They observed better performance with SPC
and attributed it to higher energy retention. The same has been reported for rainbow trout
(T acon et al. 1983). Incorporation of SPC in place of SBM was also reported to reduce
antigenic effects associated with the globular proteins glycinin, and beta-conglycinin
(Rumsey et al. 1995).

The grth data was used to ﬁt two different dose-response curves (Jobling 1994)
to estimate maximum level of incorporation of phytase treated and untreated SBM into
diets for tilapia. The ﬁrst curve was ﬁtted using the two slope broken-line method
(Robbins 1986). This model suggests SBM should be restricted to 31 % and 14 % of the
CP for untreated, and phytase treated SBM, respectively (Figure 23). The second curve
was ﬁtted using a quadratic equation, where 95 % of the maximum response details
maximum rate of incorporation (Lanari et al. 1998). This model suggests maximum rates
of 38 % and 17 % of the CP for untreated and phytase treated SBM, respectively (Figure
24).

The values from these models represented 20 -— 25 % of the diet on a DM basis for

untreated SBM, which was similar to suggested limits for tilapia (Jackson et al. 1982;

114

ﬂ
8 5
H

 

1

L

Mﬁm)
gages

ﬁ

 

 

 

d

 

r r

o as 30.0 an 15 100
Soybean Meal Inclusion or Crud. Protein)

 

 

 

 

o ”'7 a so 75 too

mmnmmcmmm

Figure 23. Grth data (% increase) of juvenile tilapia fed graded levels of (A) untreated
solvent extracted SBM, and (B) phytase treated solvent extracted SBM, as a
percent of the crude protein. Data were ﬁtted with a two-slope broken line
model utilizing a non-linear procedure as described in methods. Dotted lines
indicate model parameter estimates. Diminishing returns and lower growth
are predicted with greater than 30.6 % of the CP as untreated solvent extracted
SBM (A), and 13.7 % of the CP as phytase treated solvent extracted SBM (B).

115

W i (A)

 

 

 

 

 

 

§ 700 6 \u o y . accent” + 0.3mm + 712.83
g 550 _ R1-0.9071
; coo -
g 550 «
(9 500 A
450 d
m l 38 l l 1
0 25 50 75 100
Soybean Meal Inclusion (96 Crude Protein)
800

  
 
  

O (B)

    

 

 

 

 

8 700 ...................
§
2 650 , ya-o.0147x’-1.7oeexe7u.er
E R’IOJm
3a 600*
g 550~
0 500~

450 a

400 r r

o ‘7 25 so 75 100

Soybean Meal Inclusion (96 Crude Protein)

Figure 24. Growth data (% increase) of juvenile tilapia fed graded levels of (A) untreated
solvent extracted SBM, and (B) phytase treated solvent extracted SBM, as a
percent of the crude protein. Data were ﬁtted with a quadratic model. Dotted
lines indicate model parameter estimates. Diminishing returns and lower
grth are predicted with greater than 38 % of the CP as untreated solvent
extracted SBM (A), and 17 % of the CP as phytase treated solvent extracted

SBM (B).

116

Shiau et al. 1990), and in agreement with the industry standard limit of 25 % of the dry
diet (John Stanley, Star Milling Co., personal communication). In comparison, the
models suggested phytase treated SBM should be restricted to no more than 10 % of the
diet on a DM basis.

The whole body proximate composition analysis indicated the values for CP,
lipid, and moisture were the same as reported for similar size tilapia fed defatted SBM at
52 % of the dry diet (Wee and Shu 1989). In smaller tilapia fed graded levels of SBM,
CP and moisture were similar, but lipid levels higher, relative to the current study (Davies
et al. 1989).

The dietary effects on proximate composition showed a clear trend in increasing
CP with increasing levels of untreated SBM. There were no dietary effects on CP or on
whole body lipid levels in ﬁsh fed the phytase treated SBM. This is similar to Davis and
Stickney (1978) who observed no dietary effects on proximate composition of Tilapia
aurea fed graded levels of SBM. In contrast, others reported no effect on CP (Wee and
Shu 1989; Davies et al. 1989). However, contradictory dietary effects on whole body
lipid levels were reported (Wee and Shu 1989; Davies et al. 1989).

Both phytase treated, and untreated diets, showed the same trend in SGR and
efficiency as observed for growth. There was a slight increase in performance at 25 %
SBM substitution relative to the control diet, followed by a steadydecline. Signiﬁcant
differences were not detected among ﬁsh fed the untreated SBM diets until the level of
incorporation surpassed 75% of the CP as SBM. This is similar to the ﬁndings of others

(Davis and Stickney 1978; Shiau et al. 1989; Davies et al. 1989). In contrast, signiﬁcant

117

differences were detected in all efﬁciency parameters, except FE, in ﬁsh fed the phytase
treated diets once the level of substitution surpassed 25 % of the CP as SBM.

Speciﬁc growth rate, FE, PER, and ANPU for all SBM diets, except those at
100 % of the CP, were higher in this experiment than previously reported for SBM diets
fed to Nile tilapia (Wee and Shu 1989), and Mozambique tilapia (Davies et al. 1989), but
slightly lower than reported for red tilapia (DeSilva et al. 1991). In the study with Nile
tilapia, the highest level of SBM incorporated with low TIA activity comprised 52 % of
the dry diet (Wee and Shu 1989). The highest level in the present study was 60 % of the
dry diet At these levels, the results between the two were comparable. Nile tilapia are
sensitive to soybean trypsin inhibitors. However, the TI analysis indicated TIA was
below levels detrimental to tilapia (Wee and Shu 1989; Shiau et al. 1990).

The FE was similar, but PER slightly lower than reported for similar size tilapia
fed SBM substituted at 30 % of the CP (Shiau et al.1987; Shiau et al. 1990). This is not
surprising as the researchers were feeding a suboptimal CP diet, and PER increases when
ﬁsh are fed lower levels of protein. Additionally, the efﬁciency parameters reported here
are the same as those reported for 2.9 g tilapia fed diets incorporating graded levels of the
legume Phaseolus aureus (DeSilva and Gunasekera 1989).

Fish fed the untreated SBM diets exhibited slightly better FE and PER than those
fed phytase treated diets, but the differences were only signiﬁcant at 50 % of the CP. The
ANPU was slightly higher with phytase treated SBM at 25 % of the CP. At higher levels
of substitution the ANPU from untreated diets was signiﬁcantly higher.

The phytase treated SBM diets, and untreated SBM diets, resulted in ANPU

reduction with increasing levels of substitution. However, only ﬁsh receiving untreated

118

SBM at 100 % of the CP was signiﬁcantly reduced relative to the FM control diet. In
contrast, the phytase treated SBM resulted in signiﬁcantly reduced ANPU relative to the
FM control beyond 50 % of the CP. This would indicate better bioavailability of IAA or
energy from the untreated SBM beyond 25 % of the CP. The inference from ANPU is
phytase treated SBM should be restricted to less than 17 % of the diet on a dry matter
basis. This is in agreement with the values predicted from the grth models.
Increasing ANPU is important for economic success of any aquaculture venture,
but is critical in recirculating systems. As a direct measure of dietary protein retention, it
is an implied measure of IAA bioavailability. When confronted with excess, or
improperly balanced AA, ﬁsh readily degrade the unusable AA resulting in NH3/NH4+
excretion (Walton 1985; Cai et al. 1996). A major limiting constraint associated with
recirculating systems is the accumulation of nutrients, particularly nitrogenous products.
Phytic acid reduces digestion and availability of amino acids for uptake (Cheryan
1980; Reddy et al. 1989). Diets supplemented with salts of phytic acid were believed to
depress growth and decrease protein digestibility in rainbow trout (Spinelli et al. 1983),
Chinook salmon (Richardson et al. 1985), and carp (Hossain and Jauncey 1993).
Additionally, elimination of phytic acid from swine diets signiﬁcantly increased total
tract digestibility of CP, all IAA , and ileal digestibilty of Met and Arg (Mroz et al.
1994). Increased protein digestibility, higher biological value, and increased N retention
have been demonstrated with diets containing lower phytic acid (Satterlee and Abdul-
Kadir 1983; Mroz et al. 1994). Therefore, a digestibility study was conducted to
determine if removal of phytic acid would increase the digestibility of CP and individual

AA from SBM fed to tilapia.

119

Apparent crude protein digestibility among the diets containing SBM ranged from
78.9 -— 84.6 %. No signiﬁcant differences were detected, as was observed in hybrid
striped bass (Papatryphon et al. 1999). The values reported here are similar to those
reported for tilapia fed diets with SBM substitution up to 67 % of the CP (Shiau et al.
1989). Beyond 67 % of the CP, ACPD dropped to 70 % with Met supplementation, and
60 % without Met supplementation (Shiau et a1 1989). In two other studies, the ACPD
were the same or slightly higher than in this study, depending on dietary CP level (Shiau
et al. 1987; Shiau et al. 1990).

In the present study, Ala digestibility was lower in the phytase treated SBM
substituted at 100 % of the CP than the reference, or other phytase treated SBM diets.
The reason for differences in Ala digestibility is unclear. A survey of the literature did
not uncover a suitable explanation for differences in Ala digestibility.

Digestibility of Lys in the reference diet was also greater than in the phytase
treated SBM diet substituting SBM at 100 % of the CP. Despite low digestibility
coefﬁcients for the phytase treated diet with SBM at 100 % of the CP, no other
differences in AA digestibility were detected among the phytase treated SBM diets, or the
reference diet. This was likely a result of the large SEM for the diet with SBM at 100 %
of the CP. An insufﬁcient amount of feces from ﬁsh fed the diet with SBM at 100 % of
the CP limited statistical analysis to two samples. Therefore, it is difﬁcult to draw
conclusions about this treatment as the biological relevance may be confounded by
sampling error.

The only difference in AA digestibility among the diets incorporating graded

levels of untreated SBM was a higher digestibility of Ser in the diet at 75 % than the

120

reference diet. As with Ala in the phytase treated diets, the reason for the difference in
digestibility is unclear

Amino acid digestibility coefﬁcients for the untreated SBM diets were similar to
values reported by Sadiku and Jauncey (1995) for tilapia (0. niloticus) fed graded levels
of SBM in a SBszoultry meat meal blend diet (Table 18). Amino acid digestibilities
were similar except for Met, which were lower in our study. However, it is possible the
Met values were underestimated in our study due to loss upon hydrolysis for analysis.

Values reported by Sadiku and J auncey (1995) became more similar with higher
rates of SBM incorporation. This was likely due to the poor digestibility of poultry meat
meal (Cho 1991; Alexis et al. 1985; Sadiku and Jauncey 1995). Dilution of the poultry
meat meal with increasing SBM incorporation increased the digestibility of the diet.

All diets were formulated to contain sufﬁcient IAA to meet the known
requirements of tilapia (0. niloticus). Analysis of dietary IAA indicated all diets
contained sufﬁcient levels, except Met in the two diets containing 75 % of the CP as
SBM, and Met in the diet containing untreated SBM at 50 % of the CP. Cysteine can
contribute up to at least 50 % of the Met requirement in tilapia (Jackson and Capper
1982). Therefore the Met requirement is more appropriately termed a requirement for
TSAA. If at least 20 % of dietary Cys could be used to meet the TSAA requirement, then
the diets met all the known requirements for tilapia.

Incorporating sufﬁcient IAA into the diet does not ensure an adequate supply to
meet the needs of the animal. Incompletely digested AA are lost to the environment in

feces, and uptake of digested AA in excess of immediate needs are catabolized and

121

.533 .3253. use 853m Soc mo=_m> m
.c «eoﬁtoaxo Seam £5 Sea 828,. _

 

 

 

 

no» one E 2e 3e a? mi
3.» «.3 men 2: can e.» «mm
new new .3» NR 3: men 9a
can c3 2» a? CS was Se
”.8 New as we» 3: 3e 3e
2” an» 3» NR em» as $6
a: can 3” has a? 3: am<
2: «is E can Sn ”.8 5...
<< ozaaeoaa:
can 3» Sn E man 3K ,3,
new 0.2 n: S: n: no» m5.
2” 3» men one can In a:
Se 3.» n; at we” CS .52
new New 3» ”.3 at. new $5
«.8 New one 2: E 3» :5
o. 8 as 98 men 3» at md
93 e3 2» n8 can «S an
era 9% as 3.» men «.3 92
<< 2.353....—
aoz bosom no: ea .82 58a .82 5E .82 Face .82 ,2...
secs 8.5 as 3 2% see; oeao :3 2% seen 830 as 2. 2%
.a be cons sea 938

a .8 ._ «6% women :88 saw a 33 vegaﬁnnm A5865 09:0 65 we e\e mh use 6m 63 2mm me £32 33% to.“ A3628? .9
«Ea—u E A<<V man 658“ uEemeamE 98 ozemeamqvﬁ Sm $56530 chum—mnemomﬁ 828% we :SEnEeU .3 29¢

122

unavailable for protein synthesis (Walton 1985). Dependence of protein synthesis upon
dietary protein supply in ﬁsh is greater than in mammals (Fauconneau 1985). Apparent
digestibility coefﬁcients were applied to dietary IAA concentrations to determine if
available IAA met requirements (Table 19).

The availability of all IAA was higher from the two SBM diets at 25 % of the CP
than from the reference control diet. This is likely the reason for the slightly better
grth and conversion efficiency observed in ﬁsh fed diets containing 25 % of the CP as
SBM. Available Phe in the reference diet was 17 % below the requirement, and 25 %
lower than in the two SBM diets. Therefore, in addition to the overall IAA availability it
is possible Phe became the limiting amino acid relative to the SBM diets at 25 % of the
CP.

Available TSAA and Thr were below reported dietary requirements in all diets.
However, as stated previously, the TSAA values may be underestimated due to potential
loss of Cys and Met upon hydrolysis for analysis. A deﬁciency of available Thr with a
high level of SBM incorporation were also noted in rainbow trout (Davies and Morris
1997)

The availability of IAA from diets containing phytase treated SBM was similar to
that of the untreated SBM, with a few exceptions. The availability of Lys was lower in
phytase treated SBM diets than the untreated SBM diets at each level of SBM
substitution. The availability of Met and Thr were lower in phytase treated SBM diets
than the untreated SBM diets once the level of SBM substitution surpassed 50 % of the

CP. This trend appeared to be similar for most of the IAA.

123

Table 19. Requirements and available dietary indispensable amino acids (IAA) for tilapia
fed graded levels of untreated, or phytase treated, SBM as a percent of dietary

 

 

 

 

 

 

crude protein.

Untreated SBM (% Dietary CP)
IAA Requirementl Control 25 5o 75 100
Arg 1.18 1.55 1.89 1.87 2.21 2.30
His 0.48 0.49 0.60 0.58 0.66 0.69
Ile 0.87 1.00 1.15 1.09 1.24 1.22
Leu 0.95 1.91 2.14 2.02 2.21 2.16
Lys 1.43 1.67 1.92 1.70 1.71 1.67
Met 0.75 0.67 0.70 0.55 0.61 0.61
TSAA 0.90 0.81 0.87 0.70 0.80 0.81
Phe 1.05 0.87 1.10 1.11 1.30 1.37
Thr 1.05 0.85 0.96 0.89 1.00 0.93
Val 0.78 1.24 1.34 1.21 1.36 1.26
Phytase Treated SBM (% Dietary CP)
IAA Requirement 25 50 75 100
Arg 1.18 1.83 1.86 2.03 1.81
His 0.48 0.57 0.60 0.62 0.56
Ile 0.87 1.12 1.16 1.19 0.98
Leu 0.95 2.14 2.15 2.11 1.73
Lys 1.43 1.76 1.55 1.47 1.14
Met 0.75 0.66 0.61 0.55 0.42
TSAA 0.90 0.81 0.77 0.73 0.58
Phe 1.05 1.08 1.13 1.29 1.15
Thr 1.05 0.92 0.88 0.89 0.66
Val 0.78 1.29 1.30 1.25 0.89

 

1 Requirements taken from NRC (1993).

124

 

A correlation analysis was run between weight gain and individual available
amino acids from the SBM diets fed to tilapia during the eight week growth trial. The
same analysis was performed on the efﬁciency parameters SGR, FE, PER, and ANPU.
All AA exhibiting a signiﬁcant correlation (P<0. 10) are reported along with their Pearson
correlation coefﬁcient for the relationship (Table 20). Two trends are readily apparent
for both growth, and efﬁciency parameters.

In ﬁsh fed the untreated SBM diets, there was a signiﬁcant negative correlation
between Arg, His, and Phe with weight gain, SGR, and PER Increasing SBM
substitution led to higher dietary levels of Arg, His, and Phe, with resulting depression in
growth. Although the Pearson correlation coefficients indicate a strong relationship, their
statistical relevance were marginal (0.05 < P < 0.10). The possible relationship observed
with His is unclear. Feeding His and Thr well beyond their requirements in chum salmon
had no adverse effect on growth or conversion (Akiyama et al. 1985). Although
suggestive of a relationship, correlation analysis does not necessarily indicate a cause and
effect.

A possible explanation for the relationship observed with Arg is an Arg/Lys
antagonism. It has been demonstrated in ﬁsh that Arg and Lys compete for the same
intestinal transporters, as they do in mammals (Ash 1985), but an antagonistic
relationship has yet to be convincingly demonstrated in ﬁsh (Wilson 1989; NRC 1993).
Similarly, there is evidence for competition between Phe and Met for transport into the
enterocyte (Ash 1985). Higher levels of Phe with increasing SBM incorporation may

have exacerbated any Met deﬁciency.

125

Table 20. Correlations between weight gain and efﬁciency parameters, and available
individual amino acids in phytase treated SBM, and untreated SBM,
experimental diets fed to tilapia. Parameters include speciﬁc grth rate
(SGR), feed efﬁciency (FE), protein efﬁciency ratio (PER), and apparent net
protein utilization (ANPU). Experimental diets incorporated SBM at 0, 25 , 50,
75, and 100 % of the dietary crude protein. Shown are IAA and associated
Pearson correlation coefﬁcients for all relationships with (P<0. 10).

 

 

Treatment IAA Pearson Correlation P — value
Coefﬁcient
Weight Gain Untreated ARG -0.867 0.057
(% Increase) Untreated HIS -0.805 0.098
Untreated PHE -0.852 0.067
Phytase LYS 0.908 , 0.033
Phytase MET 0.853 0.066
Phytase TSAA 0.836 0.078
SGR Untreated ARG -0.863 0.060
(% /day) Untreated HIS -0.806 0.099
Untreated PHE -0.849 0.069
Phytase LYS 0.930 0.022
Phytase MET 0.879 0.050
Phytase TSAA 0.871 0.055
FE Phytase LYS 0.974 0.005
Phytase MET 0.948 0.014
Phytase TSAA 0.946 0.015
Phytase THR 0.840 0.075
Phytase VAL 0.820 0.089
PER Untreated ARG -0.851 0.068
Untreated HIS -0.819 0.090
Untreated PHE -0. 840 0.07 5
Phytase LYS 0.975 0.005
Phytase MET 0.964 0.008
Phytase TSAA 0.953 0.012
Phytase THR 0.809 0.097
ANPU Phytase LYS 0.985 0.002
(%) Phytase MET 0.962 0.009
Phytase TSAA 0.947 0.015
Phytase THR 0.81 1 0.096

 

126

In ﬁsh fed the phytase treated SBM diets, there was a strong correlation between
Lys, Met, and TSAA and weight gain. Similar correlations were observed between Lys,
Met, and TSAA, and all efﬁciency parameters (Table 20). Additionally, there was a
statistically marginal effect of Thr on FE, PER, and ANPU, and of Val on FE. Davies
and Morris (1997) found depressed ANPU in rainbow trout, which they attributed to
deﬁciencies in Met, Lys, and Thr.

Although correlations with available TSAA were signiﬁcant, the effects were due
to Met, as the correlation coefﬁcients decreased and P values increased aﬁer including
Cys in the relationship. The high correlations observed with available Lys and Met
suggest these two IAA are responsible for depressed weight gain and poor performance
of ﬁsh fed increasing levels of phytase treated SBM.

A plot of dietary and available Met in relation to dietary CP, and as a function of
SBM substitution, indicates dietary and available Met decreased with increasing
substitution of phytase treated SBM. This occurred even as dietary CP increased (Figure
25 (A)). At the 100 % substitution level, availability decreased substantially in relation to
dietary Met and CP. Similar observations were made in rainbow trout (Dabrowski et al.
1989).

In contrast, dietary and available Met in the untreated SBM diets paralleled each
other and were constant following crystalline Met supplementation (Figure 25 (B)).
These plots would suggest a factor associated with the phytase treated SBM was
responsible for the observed decrease in Met availability and performance in ﬁsh fed

diets incorporating phytase treated SBM at 75 % and 100 % of the CP.

127

- o- AnilableMet (A)

 

 

 

 

 

 

 

 

0.9 T “'4' Mr! 1‘“ -~ 30
b- . - — A IWMUCP
0.8 ~»
3 + 30 3
o 0.7 —~ car
5 o
25 0.0 ~~ 2‘.
g ~~ 34 a.
o
0.5 ..
0.4 32
0 25 50 75 100
SBM Substitution (96 CP)
- o - Available Not (B)
0.0 e- _, 30
. . ___ A -A- DietaryMet
0.8 «a ‘
S ~~ so A
r; n- 2
o E
5 0.6 .. E
g ‘1' 3‘ o
0.5 ~—
0.4 - 32
0 25 50 75 100
SBM Substitution (96 CP)

Figure 25. Dietary Met and available Met in relation to dietary CP. Diets contained
graded levels of phytase treated SBM (A), or untreated SBM (B) as a percent
of CP in the diet.

128

A plot of dietary and available Lys in relation to dietary CP, and as a function of
SBM substitution, indicates dietary and available Lys follow an identical trend to Met
(Figure 26). However, whereas Met availability began to decrease relative to dietary Met
at 100 °/o substitution, the Lys availability began to decrease relative to dietary Lys when
the level of substitutionsurpassed 25 % of the CP. This pattern in decreased Lys
availability was the same pattern observed for weight gain, SGR, PER, and ANPU.
Although a Lys/Arg antagonism has not been conclusively demonstrated in ﬁsh, there is
insufﬁcient evidence to rule out the possibility that high levels of Arg did not exacerbate
the Lys deﬁciency observed in phytase treated SBM diets.

Soybean meal has a favorable IAA proﬁle for replacing FM in formulated diets
for tilapia, although it has a relatively low chemical score (Tacon and Jackson 1985). At
high rates of incorporation the diets become deﬁcient in one or more IAA. The ﬁrst two
IAA considered limiting for tilapia, and most ﬁsh species in general, are Met and Lys,
although Thr is also mentioned. In a number of studies, SBM diets have been
supplemented with IAA to compensate for this deﬁciency. Reports of success with this
approach vary (Rumsey and Ketola 1975; Viola et al. 1981;Walton et al. 1982; Viola et
al. 1983; Shiau et al. 1989; El-Sayed 1989; El-Dahhar and El-Shazly 1993; Davis et al.
1995; Ng et al. 1996; Davies and Morris 1997; Coyle et al. 2000).

Tilapia fed diets with SBM at 100 % of the CP supplemented with Lys and Met
did not show an increase in growth (Viola and Arieli 1983). Additionally, tilapia fed
diets containing 33, 67, and 100 % of the CP as SBM supplemented with Met exhibited

no differences in growth except at the 100 % inclusion level. At this level there was a

129

A
2-4 r - 0- Available Lys ( ) T 38

 

 

 

 

 

 

 

 

2 2 —A- Dietary Lw
' I A -O—Dietery OF
A 2 .0. A
g .. 30 E
n
g 1 8 "F E.
a
33 1.6 + 35
g 1.4 ~»— + 3‘ 3
1.2 ..
1 32
0 26 50 75 100
SBM Substitution (%CP)
(B)
2.4 T - o- Available Lys ,_ 33
—A- Dietary Lys
2.2 ~~ , Ac . —-O—Dletary cr
3 2 “P «- as 3
a 1 a -. °
5 - s
35 1.6 ~~ 35
a a.
E 1.4 __ 34 o
1.2 e»
1 - v 32
0 25 50 75 100
SBM Substitution (“lo CP)

Figure 26. Dietary Lys and available Lys in relation to dietary CP. Diets contained
graded levels of phytase treated SBM (A), or untreated SBM (B) as a percent
of CP in the diet.

130

signiﬁcant increase over the unsupplemented diet, but still signiﬁcantly lower than the
other diets with or without supplementation (Shiau et al 1989).

When the experimental diets were formulated, it was apparent dietary Met was
below requirements for tilapia. Crystalline L-methionine was incorporated in the diets
containing 50 % or more of he CP as SBM to meet the requirement. Although dietary
Met met the requirement, on an available basis, the level of Met was below the
requirement. The crystalline Met appeared to be available to tilapia in the diets
containing untreated SBM (Figure 27). As Met in the intact protein decreased with
increasing levels of SBM substitution, available Met as a percent of the dietary
requirement improved slightly, suggesting crystalline Met was utilized.

Conversely, Met supplementation in the diets containing phytase treated SBM
appeared to be completely unavailable to tilapia (Figure 27). Diets containing phytase
treated SBM resulted in severely depressed available Met despite Met supplementation.
However, a closer examination of the available Met values from these diets indicates the
supplemental Met may be as effective as the intact protein at supplying Met. The
measured total available Met was similar to the dietary Met supplied as intact protein,
appearing as though crystalline Met is unavailable (Table 21). Applying the apparent
digestibility coefﬁcient to both the Met supplied in the intact protein, and the
supplementation, results in nearly identical values for calculated and measured available
Met (Table 21).

There is evidence that crystalline amino acids are readily leached from diets
(Wilson 1989). The potential for leaching would be high from the experimental diets

used in this study. The large surface area to volume ratio implied by the feed size

131

dang—me

03800:: 3 33¢ 20.5 025.0 .0305 .8 0:080:30.— 0500306 @806 05 808 8 5085 09:0 .9806 05 .«0

a. 2: 05 an .8 a 2% 058.088... 0% 20m 03825 05 .20m Base 083% 25 35.823 as 05830:
033580303 .9006 05002008 05:50.974 .«0 530:3 a 00 3:082:62 50% {av 0:801:08 030=0>< Hm 0.59m

30:. be 9:3
:ozﬁcoﬁoansm 05.3.50! 05:83.8-..

 

o.” 0.0. a.“ 0.. 3 0.0 a...
r a _ _ a _ cm
088: $5200 - mm )
F n a“ lo
000003—50 o V
- 8 m u
9 I.
a m
I no ..M m
H
m m
1 eh m. M
.... m m
m .u.
- 3 (

 

 

 

FIN“ in”

132

 

.A 303 DMZ 52.: :83: 2mm 0:0 000:: m5.:.:0m :0: mo:_0> 05:05:02 _

 

N06 m0.o 5 .o omd wdm and 36 co:
mmd who .2 .o 93 N00 2 .o wmd mh
Ed $6 86 00o Wm: wed :6 on
09¢ 36 0 mod Qt. o vwd mm
:06 Sue o Rd 5.2. c cad o

 

02:: 0:: 0:0: 00.0 00.300
85802 032800 020 5 0:0: 02:: .5 000: 822.080 020 00 0: E 00.6 .00 0:0:
:02 :02 3052090 522: :83: 0050805 0582.35 00:80:08.: 80 0.0200
230.2 08:. 2:203: .08.: :02 2020.2 :02 2.303.. .8303. as: : :02

05:05:05
03502: 0050005 0:0 4:050:58 0:: 0:0 50:05: 80:5 50:: 05:05:05 :0 5:00:02: 00:53—00 0:: 50:0 05 :0:
05065000 5.500050 520000 0:: 0:0 850:0. 00?. 50:05 000:0 b.3050 .:0 ::00:0: 0 00 2mm 00:00:: 000:3: :0 £0.6—
0008m m5:0:0&005 30:0 5 50050503000 05:02:05 05:0:mboé 0:0 50:05 80:5 .3 00:00:: 05:02:05 .9305 .— N oB£

133

required by 1.3 g ﬁsh, coupled with feeding behavior of tilapia make leaching a strong
possibility. It can not be determined if available Met was solely from the intact protein or
partitioned between the two sources. This is a needed area for future investigation.

It appears crystalline L-Met supplementation was effective in improving growth,
performance, and digestibility of the untreated SBM diets used in this study, relative to
the phytase treated SBM diets. This is contrary to other ﬁndings in tilapia (Shiau et al.
1989; Robinson et al. 1984); however, effectiveness of supplementation may be
dependent on dietary CP levels (Shiau et al. 1987). Suitability of crystalline amino acid
supplementation has been demonstrated in hybrid striped bass (Grifﬁn et al. 1992), red
drum (Brown et al. 1988), rainbow trout (Davies and Morris 1997), Atlantic salmon (Olli
et al. 1995), carp (Nose et al. 1974), Tilapia zilli (El-Sayed 1989), and channel catﬁsh
(Wilson et al. 1977).

Care should be taken in interpreting the results of Met and TSAA availability. It
is possible the feed and fecal values of both Cys and Met were underestimated Cysteine
and Met have been shown to be unstable under the conditions used to hydrolyze the feed
and fecal samples. Typically, to protect them ﬁom destruction, Cys and Met are
converted to the more stable derivatives cysteic acid and methionine sulfone before
hydrolysis (Waters 1989). A lack of sufficient material precluded taking these
preventative measures before hydrolysis. Although it is possible some of the Met and
Cys were destroyed before analysis, measured dietary Met was virtually the same as
calculated from the diet formulation, suggesting losses may have been minimal.

A number of hypotheses lend themselves to an explanation for reduced

availability of AA and energy from the diets containing phytase treated SBM. Principal

134

 

among these is the feeding behavior of tilapia. Tilapia have been described as
preferential grazers as opposed to meal eaters, such as trout or salmon that swallow food
whole. Tilapia are equipped with palatine teeth and pharyngeal gill rakers allowing them
to grind feed and ﬁlter material before swallowing (J auncey and Ross 1982; Bowen
1982). Feeding tilapia repeatedly pick up food items and expel them numerous times
before consumption (Hanley 1985; personal observations). This process eﬁecﬁvely
creates exceedingly small particles with high surface to volume ratio to facilitate
digestion. In addition, tilapia will take hours to ﬁnish a meal while the feed sits on the
bottom. This feeding activity enhances the potential for leaching of soluble components.
Phytic acid acts to decrease solubility of nutrients (Reddy et al. 1989) and its removal
may lead to a loss of nutrient availability.

The behavioral response of the tilapia to the diet substituting phytase treated SBM
at 100 % of the CP may also indicate dietary deﬁciencies or imbalances. The tilapia
displayed avoidance to the diet. The behavior was originally attributed to poor
palatability. However, this behavior may also have been a response to an AA deﬁciency,
or imbalanced amino acids, as has been exhibited in terrestrial animals (Gietzen 1993).

Phytic acid may also serve in a protective capacity. Phytic acid binds to the
globular proteins glycinin, and B-conglycinin (Reddy et al. 1989), both of which have
been found to be ANFs in ﬁsh (Kaushik et al. 1995; Rumsey et al. 1995). Additionally,
phytic acid may protect sensitive AA, such as Lys, from degradation during processing or
pelleting.

Finally, the phytase treatment process itself may be responsible for altering

availability of AA, energy, or some other essential nutrient. Tilapia appear to adapt to the

135

increased levels of SBM substitution with time, possibly by inducing increased activity of
amylase (Anderson et al. 1991). In ducks, phytase addition to the diet decreased amylase
activity (Martin et al. 1998). Tilapia amylase activity is high for ﬁsh species (Nagase
1964) allowing them to make better use of carbohydrates than other species (Anderson et
al. 1991). The potential reduction of amylase activity may have decreased available
energy. Unfortunately, the limitation of small fecal samples precluded the analysis of
digestible energy.

The results from the SBM grth and digestibility trials would suggest phytase
was successful in hydrolyzing the phytic acid associated with SBM. Incorporating
phytate pretreated SBM beyond 25 % of the CP led to a depression in weight gain and
decreased efﬁciency. In comparison, incorporation of untreated SBM into the reference
diet did not decrease weight gain, or diminish efﬁciency, until the level of substitution
surpassed 75 % of the CP. This would suggest phytic acid may play a role in sustaining
protein integrity and availability of AA in tilapia Less conclusive is whether phytic acid
is beneﬁcial to amino acid availability, or the absence of phytic acid causes a decrease in
amino acid availability. If phytic acid effects grth and efﬁciency, the results should be
reproducible in a dose-response manner using a puriﬁed, or semi-puriﬁed form. This
approach has been used successfully in determining the effects of phytic acid on mineral
availability (Spinelli et al. 1983; Satoh et al. 1989; Hossain and Jauncey 1993). To
evaluate the effect phytic acid may have on growth, efﬁciency, and amino acid
digestibility, tilapia were fed a ﬁsh meal based diet with graded levels of phytic acid as

Na-phytate.

136

Fish fed the ﬁsh meal diets supplemented with Na-phytate exhibited differences
in growth and SGR. Only the treatment receiving the highest level of supplementation
grew less than the control treatment (without supplementation). No signiﬁcant
differences were detected in FE, PER, ANPU, or whole body proximate composition,
regardless of rate of supplementation. Weight gain and all efﬁciency parameters were
lower in this trial than in the SBM growth trial. This applied to the reference control diet
as well, which was the same diet for both trials, suggesting the differences were
independent of dietary effects.

The data suggest a superﬁcial trend toward increased growth with increasing
levels of phytic acid supplementation. However, dietary CP in these diets ranged from 31
— 37 % of the dry diet, and shows an almost identical trend to that of growth (Figure 28).
A correlation analysis indicated a strong relationship between growth and all available
IAA. Pearson Correlation coefﬁcients ranged from 0.90 for Thr to 0.95 for His, with an
overall mean of 0.94 (P<0.01). This would suggest grth was independent of dietary
phytic acid, lending support to the conclusion increased grth was due to dietary CP, or
the DE:P ratio.

The CP digestibilities of these diets showed no signiﬁcant differences when
dietary Na-phytate ranged from 0 — 25.8 g/kg of the dry diet. As was noted for the
growth response, there appeared to be a superﬁcial trend of increasing ACPD with
increasing phytic acid supplementation. A multiple regression analysis on ACPD of the
Na-supplemented FM diets indicated ACPD was correlated to level of CP and

independent of dietary phytic acid.

137

 

.020 00 000: 20:00.02
00 308550508 0:00 00.30 E0650 .:0 50000.: 0 00 0:0 .30: 50:8: 000:0 50:0 0: 500—0: 5 500 5503 .mm 0:005

030 00 9:2 503552230 30:20-02

 

 

 

 

«.mu 0.: 5.0 mi 0.0 0

on can

an + .1 03 M
0 w.

:0 .. L- 03 1
H 9
% m.
o 00 i + 02. m.
.m .l.
m w.
m 00 1- -- 8.: a
.. m

0:. + :0 basal? 1- an» (

5.020.251? .
NV 1r Lr 600

138

Apparent digestibility of individual amino acids in the Na-supplemented FM diets
exhibited a parallel trend to that of growth and ACPD. Other than a few exceptions, the
AA digestibility coefficients followed the same pattern for each IAA. The diet
containing 12.9 g Na- phytate/g dry diet had the highest digestibility coefficient for all
AA. However, in general they were only signiﬁcantly different from the two diets
containing 3.4 and 25.8 g Na-phytate/kg dry diet. No significant differences were
observed for Ala and Asp among any of the diets. Additionally, Gly and Ser were
signiﬁcantly higher in the diet containing Na-phytate at 12.9 g /kg dry diet than the other
diets.

All diets were formulated to meet or exceed the known IAA requirements for
tilapia The only difference in preparation was the level of incorporation of Na-phytate at
the expense of cellulose. Analysis indicated dietary levels of all IAA were sufficient to
meet the requirements for tilapia. Apparent digestibility coefﬁcients were applied to
dietary IAA concentrations to determine if available IAA met requirements (Table 22).
The apparent availability was low for TSAA, Phe, and Thr in the diets containing Na-
phytate at 3.4, 6.5, 25.8 g/kg dry diet. Not surprisingly these were the slowest growing
treatments. The data from the phytic acid supplemented diets suggest there was neither a
beneﬁcial, or detrimental effect due to the supplementation in terms of growth,
efficiency, body composition, or digestibility.

Tilapia exhibit plasticity of the GI tract providing variability and adaptability.
Moriarity (1973) observed highly acidic stomachs in wild Nile tilapia, with pH values as
low as 1.0-2.0. These values are similar to values reported for other wild captured tilapia

species (Bowen 1982). Low gastric pH in wild tilapia assists digestion by hydrolysis of

139

0000:: 0:2 80:: 09:0: 00052300: :

 

140

 

00.: :0: 00.: 00.: 00.: 00.: 00.0 00>
00.0 00.: 00.: 00.0 00.0 00.: 00.: :00
00.: 0:.0 00.: 00.: 00.: :0: 00.: E. + 000
00.0 00.: 00.: 00.0 :0.0 00.: 00.: 00.:
00.0 0:: 00.0 00.0 00.0 00.: 00.0 02.00
00.0 00.0 00.0 00.0 :00 00.0 00.0 :02
00.: 00.0 20 00.0 00.: 0:.0 00.: 00.:
00.0 00.0 00.0 :0 00.0 00.0 00.0 00:
00.: 00.: 0:: 0:: 00.: :0: 00.0 0::
00.0 00.0 00.0 00.0 00.0 00.0 00.0 0::
00.: :0 00.: 00.: 00.: 00.: 0:: 02
0.00 0.0: 0.0 0.0 0.0 0 : 80:02:08: <<:

 

90:0 b0 3%: 30050520006 05030.02

 

.0:0:.E0.0Z 00 0:00 0530 00 205. 00003 :53 0050505000 0:50 0000::
:85 02.: 00.: 0:020 a: 20:: 00:8 20:5 2000000000: 030:0 200::020 000 0:853:03: .00 200:.

phytoplankton and bacterial cell walls, and in mineral decomposition of periphytic
detrital aggregate (Moriarity 1973; Bowen 1976; Bowen 1982).

Maximal inhibition of acidic proteases by phytic acid occurs near pH 2.0 (Camus
and Laporte 1976; Vaintraub 1991). Gastric pH reported by Moriarity (1973) suggest an
environment favorable for enzyme inhibition. However, maximum activity of a protease
isolated from 0. niloticus gastric mucosa was observed at pH 3.5, with minimal activity
reported at pH values outside the range of 2.0 - 5.5 (Yamada et al. 1993). In the pH
range reported for wild tilapia, and favorable for phytic acid inhibition, proteolytic
activity would likely be severely diminished.

Measurements of pH were taken at 10 deﬁned sites on the surface of digesta
removed from the gastric region at 0.5, 1, 2, 4, 6, and 8 hours postprandially. The lowest
pH recorded on the digesta was 2.35, 6 hours postprandially. Values as low as 2.5 - 2.75
were recorded on sites in contact with the acid-secreting cells localized to the ventral face
of the mucosa However, the mean values of the 10 sampling sites ranged ﬁ'om 5.36
shortly following feeding, to 3.52 at 4 hours postprandially. The pH values following the
feeding of a high nutrient dense diet were closer to the optimal pH for enzyme activity
than were pH values from wild ﬁsh.

The range and mean pH values measured on the gastric mucosa were higher than
measured on the digesta, providing further evidence for acid secretory cells described by
Fish (1960) and Kapoor et al. (1975). Although pH on the surface of the feed decreased
with time, the pH values of the homogenized digesta remained relatively constant over
time. This implied that gastn'c mixing of the digesta was minimal when ﬁsh were fed

pelleted diets. The small abundance of phytoplankton, coupled with its small size,

141

provides a large surface area for acidic degradation, mixing, and proteolytic activity
observed by Moriarity (1973).

The small surface area to volume ratio of feed pellets, short gastric retention time,
and minimal mixing, potentially decreases the efﬁciency of gastric hydrolysis and
digestion This interpretation is further supported by the inability of the volume
dependent model of GER to adequately describe GER in tilapia fed pelleted diets.

The pH values in the intestine remained relatively consistent throughout the
sampling period, both within a segment and with time. The pH values increased in
intestinal segments posterior to segment 1, which is the site receiving the bile contents.
The mean pH values measured in these posterior segments ranged from near 7.0 — 8.3.
The values were similar (pH 6.8 - 8.8) to those reported for wild tilapia (Fish 1960;
Nagase 1964; Moriarity 1973). These values are optimal for protein-cation-phytate
complex formation (Cheryan 1980; Reddy et al. 1989). In addition to demonstrating
tilapia manifest a suitable environment for protein-phytic acid complex formation, the

data gathered can be used to mimic in viva pH conditions in in vitro enzyme assays.

142

SUMMARY AND CONCLUSIONS

Fish reared in intensive production systems have different requirements than those
in the wild. In such systems natural food is limited. All nutrients must be exogenously
supplied in the form of high nutrient dense pellets. Consequently there is a need for
optimal feeding regimens to accommodate this capacity to process formulated diets.

An optimal ration and feeding regimen should take into account a combination of
the interplay between FCR and SGR. Evaluating these parameters for ﬁsh fed in this
study suggests the optimal feeding frequency for both parameters is either three or four
feedings day '1. The determination of an optimum feeding regimen should be evaluated
from two aspects; 1) physiology of the species, and 2) economics of the aquaculture
production unit (Tsevis et al. 1992). The economics of the production unit would suggest
3 feedings day '1 is superior due to the cost of labor.

Understanding rate of digestion and its relationship to GER is important for
determining the return of appetite. Making food available as soon as appetite has
returned can maximize intake, and increase efﬁciency. Demonstrating a consistent
relationship between stomach fullness and appetite will allow an optimal feeding interval
to be predicted from factors that control GER. Therefore, GER and return of appetite
were evaluated in ﬁsh fed to satiation 3, and 5 times day '1.

Gastric evacuation curves indicated curvilinear relationships. Both feeding
frequencies were best described by the exponential frmction VT = Voe 4""). The slopes
described by the term b represent the instantaneous rate of gastric evacuation. The

instantaneous rate for the two feeding frequencies in this study were nearly identical.

143

Optimal feeding frequency corresponds to the volume and rate of emptying of the
stomach. If the optimum interval between feedings is the point at which evacuation of
the previous meal is matched by consumption, then the optimum interval for 0. niloticus
is 4 - 5 hours, depending on the energy and composition of the diet.

Decreased protein digestibility of diets supplemented with salts of phytic acid led
to depressed growth and poor performance in rainbow trout, and carp. In this
investigation, tilapia were fed graded levels of phytase treated SBM, or untreated SBM,
substituted into a F M based diet to determine the effects of phytic acid on growth,
performance, and digestibility of CP and individual amino acids.

The growth data was used to ﬁt two different dose-response curves to estimate
maximum level of incorporation of phytase treated and untreated SBM into diets for
tilapia The two models suggested similar levels of restriction, 31 — 38 % and 14 — 17 %
of the CP for untreated, and phytase treated SBM, respectively. The values for the
untreated SBM represented 20 -— 25 % of the diet on a DM basis, which was similar to
suggested limits for tilapia, and in agreement with the industry standard limit of 25 % of
the dry diet. In comparison, the models suggested phytase treated SBM should be
restricted to no more than 10 % of the diet on a DM basis.

Increased protein digestibility, higher biological value, and increased N retention
have been demonstrated in terrestrial animals with diets containing lower phytic acid.
Therefore, a digestibility study was conducted to determine if removal of phytic acid
would increase the digestibility of CP and individual AA from SBM fed to tilapia.

Apparent crude protein digestibility among the diets containing SBM ranged from

78.9 - 84.6 %, and no signiﬁcant differences were detected. Apparent digestibility

144

coefﬁcients were applied to dietary IAA concentrations to determine if available IAA met
requirements.

In ﬁsh fed the untreated SBM diets, there was a signiﬁcant negative correlation
between Arg, His, and Phe with weight gain, SGR, and PER. Increasing SBM
substitution led to higher dietary levels of Arg, His, and Phe, with resulting depression in
growth. Although the Pearson correlation coefﬁcients indicated a strong relationship,
their statistical relevance were marginal (0.05 < P < 0.10). Although suggestive of a
relationship, correlation analysis does not necessarily indicate a cause and effect.

In ﬁsh fed the phytase treated SBM diets, there was a strong correlation between
Lys, and Met, with weight gain. Similar correlations were observed between Lys, Met,
and TSAA, and all efﬁciency parameters. Additionally, there was a statistically marginal
effect of Thr on FE, PER, and ANPU, and of Val on FE.

Crystalline L-methionine was incorporated in the diets containing 50 % or more
of he CP as SBM to meet the requirement. Although dietary Met met the requirement,
the calculated available Met was below the requirement The L-crystalline Met appeared
to be available to tilapia in the diets containing untreated SBM and was effective in
improving growth, performance, and digestibility of the untreated SBM diets, relative to
the phytase treated SBM diets used in this study.

Conversely, evidence suggested Met supplementation in the diets containing
phytase treated SBM appeared to be unavailable to tilapia A closer examination of the
available Met values indicated the evidence may be misleading. It could not be
determined if available Met was solely from the intact protein or partitioned between the

two sources. This is a needed area for future investigation.

145

A number of hypotheses were presented to explain the reduced availability of AA
and energy from the diets containing phytase treated SBM. Principal among these was
that the feeding behavior of tilapia enhances the potential for leaching of soluble
components. Additionally, phytic acid may serve in a protective capacity by reducing the
effects of other ANFs, or protecting sensitive AA, such as Lys, from degradation during
processing or pelleting. Finally, the phytase treatment process itself may be responsible
for altering availability of AA, energy, or some other essential nutrient.

Incorporating phytate pretreated SBM beyond 25 % of the CP led to a depression
in weight gain and decreased efﬁciency. In comparison, incorporation of tmtreated SBM
into the reference diet did not decrease weight gain, or diminish efﬁciency until the level
of substitution surpassed 75 % of the CP. This would suggest phytic acid may play a role
in sustaining protein integrity and availability of AA in tilapia.

If phytic acid effects growth and efﬁciency, the results should be reproducible in a
dose-response manner using a puriﬁed, or semi-puriﬁed form. To evaluate the effect
phytic acid may have on growth, efﬁciency, and amino acid digestibility, tilapia were fed
a ﬁsh meal based diet supplemented with graded levels of phytic acid as Na-phytate.

No signiﬁcant differences were detected in FE, PER, ANPU, or whole body
proximate composition, regardless of rate of supplementation The CP digestibilities of
these diets showed no signiﬁcant differences. A multiple regression analysis on ACPD
of the Na-supplemented FM diets indicated ACPD was independent of dietary phytic
acid. Apparent digestibility of individual amino acids in the Na-supplemented FM diets
was similar to ACPD. Other than a few exceptions, the AA digestibility coefﬁcients

followed the same pattern for each IAA. The data from the phytic acid supplemented

146

diets suggest there was neither a beneﬁcial, or detrimental effect due to the
supplementation in terms of growth, efﬁciency, body composition, or digestibility.

Wild Nile tilapia have been reported to have gastric pH values as low as 1.0-2.0.
Low gastric pH in wild tilapia assists digestion by hydrolysis of phytoplankton and
bacterial cell walls, and in mineral decomposition of periphytic detrital aggregate.
Maximal inhibition of acidic proteases by phytic acid occurs near pH 2.0 suggesting an
environment favorable for enzyme inhibition. However, maximum activity of a protease
isolated from 0. niloticus gastric mucosa was observed at pH 3.5, with minimal activity
reported at pH values outside the range of 2.0 - 5.5. Measurements of pH were made in
the tilapia GI tract, following feeding, to determine if pH in ﬁsh fed nutrient dense diets
were similar to those in the wild.

Values as low as 2.5 — 2.75 were recorded on sites in contact with the acid-
secreting cells localized to the ventral face of the mucosa. However, the mean values
ranged from 5.36 shortly following feeding, to 3.52 at 4 hours postprandially. Values
following the feeding of a high nutrient dense diet were closer to the optimal pH for
enzyme activity than were pH values from wild ﬁsh. The resultant pH values provided
evidence that gastric mixing of the digesta was minimal.

The pH values in the intestine remained relatively consistent throughout the
sampling period, both within a segment and with time. The mean pH values measured in
the posterior segments of the intestine ranged from near 7 .0 - 8.3. These values are

Optimal for protein-cation-phytate complex formation in the intestine.

147

REFERENCES

AACC (American Association of Cereal Chemists). 1983. A proved Methods of the
American Association of Cereal Chemists, Method 71-10, edition. American
Association of Cereal Chemists, Inc., St. Paul, Minnesota, USA.

Akiyama, T. S. Arai, T. Murai and T. Nose. 1985. Threonine, histidine, and lysine
requirements of chum salmon fry. Bulletin of the Japanese Society of Scientiﬁc Fisheries
5 1 1635-639.

Alexis, M.N., E. Papaparaskva-Papoutsoglou and V. Theochari. 1985. Formulation of
practical diets for rainbow trout (Salmo gairdneri) made by partial or complete
substitution of ﬁsh meal by poultry by-products and certain plant by-products.
Aquaculture 50:61-73.

Al-Hussaini, AH. and AA. Kholy. 1953. On the functional morphology of the
alimentary tract of some omnivorous teleost ﬁsh. Proceedings of the Egyptian Academy
of Science 4:17-39.

Anderson, B.J., B.S. Capper and NR. Bromage. 1991. Measurement and prediction of
digestible energy values in feedstuffs for the herbivorous ﬁsh tilapia (Oreochromis
niloticus Linn). British Journal of Nutrition 66:37-48.

Andrews, J .W. and J .W. Page. 1975. The effects of ﬁequency of feeding on culture of
catﬁsh. Transactions of the American Fisheries Society 104:317-321.

AOAC (Association of Ofﬁcial Analytical Chemists). 1990. Oﬁicial Methods of Analysis
of the Official Association of A nalytical Chemists, 15‘” edition. Association of the Ofﬁcial
Analytical Chemists, Arlington, Virginia, USA.

Amesen, P., L.E. Brattas, J. Olli and A. Krogdahl. 1989.Soybean carbohydrates appear to
restrict the utilization of nutrients by Atlantic salmon (Salmo salar, L.). In: Proceedings
of the Third International Symposium on Feeding and Nutrition of Fish. Toba, Japan,
Aug. 28 — Sept. 1, 1989, pp 273-280.

Ash, R. 1985. Protein digestion and absorption. In: Nutrition and Feeding in Fish. (ed.
C.B. Cowey, A.M. Mackie and J .G. Bell), pp. 69-93, Academic Press, Inc., Orlando,
Florida, USA.

Atwal, A.S., N.A.M. Eskin, B.E. McDonald and M. Vaisey-Genser. 1980. The effects of

phytate on nitrogen utilization and zinc metabolism in young rats. Nutrition Reports
International 21:257-267.

148

Avault, J .W., Jr. 1996. Fundamentals of Aquaculture. AVA Publishing, Co., Inc., Baton
Rouge, Louisiana, USA, pp 889.

Balarin, J .D. and J .P.Hatton. 1979. T ilapia: A Guide to Their Biology and Culture in
Aﬁica. University of Stirling, Stirling, Scotland pp 174.

Barth, T.L., L. Hauzerova, J. Kouril, J. Hamackova and R. Vachta. 1995. Extraction of
proteolytic enzymes of gastrointestinal tract extracted from offal viscera of carp and
tilapia. Aquaculture 129:251-252.

Bergot, F. 1979. Carbohydrate in rainbow trout diets: effects of the level and source of
carbohydrate and the number of meals on growth and body composition. Aquaculture
18: 157-167.

Boujard, T. and J. F. Leatherland. 1992. Circadian rhythms and feeding times in ﬁshes.
Environmental Biology of Fishes 35:109-131.

Bowen, SH. 1976. Mechanism for digestion of detrital bacteria by the cichlid ﬁsh
Sarotherodon mossambicus (Peters). Nature 260: 137-138.

Bowen, SH. 1982. Feeding, digestion and growth -— qualitative considerations. In: The
Biology and Culture of Tilapias. (eds. R.S.V. Pullin and RH. Lowe-McConnell), pp.
141-180, ICLARM, Manila, Philippines.

Brett, J .R. 1971. Satiation time, appetite and maximum food intake of sockeye salmon
(Oncrhynehus nerka). Journal of the Fisheries Research Board of Canada 27: 1767-1779.

Brown, P.B., DA Davis and EH. Robinson. 1988. An estimate of the dietary lysine
requirement of j uvenile red drum (Sciaenops ocellatus). Journal of the World
Aquaculture Society 19: 109-1 12.

Buddington, R.K. 1980. Hydrolysis resistant organic matter as a reference for

measurement of ﬁsh digestive efﬁciency. Transactions of the American Fisheries Society
109:653-656.

Cai, Y., J. Wermerskirchen and IR. Adelman. 1996. Ammonia excretion rate indicates
dietary protein adequacy for ﬁsh Progressive Fish Culturist2581124-127.

Cain, KB. and BL. Gatling. 1995. Pretreatment of soybean meal with phytase for
Salmonid diets to reduce phosphorus concentrations in hatchery effluents. Progressive
Fish Culturist 57:114-119.

Caldwell, RA. 1992. Effect of calcium and phytic acid on the activation of trypsinogen
and the stability of trypsin. Journal of Agricultural Chemistry 40:43-46.

149

Camus, M. and 1C. Laporte. 1976. Inhibition de la protéolyse pepsique in vitro par le blé.
Rele de l’acide phytique des issues. Annales de Biologie Animale, Biochimie,
Biophysique 16:719-729.

Carnovale, E, E. Lugaro and G. Lombardi-Boccia. 1988. Phytic acid in faba bean and
pea: effect on protein availability. Cereal Chemistry 65: 1 14-1 17.

Caulton, MS. 1982. Feeding, metabolism and growth of tilapia: some quantitative
considerations. In: The Biology and Culture of T ilapias. (eds. R.S.V. Pullin and RH.
Lowe-McConnell), pp. 157-180, ICLARM, Manila, Philippines.

Cherivinski, J. 1982. Environmental Physiology of Tilapias. In: The Biology and Culture
of Tilapias. (eds. R.S.V. Pullin and RH. Lowe-McConnell), pp. 119-128, ICLARM,
Manila, Philippines.

Cheryan, M. 1980. Phytic acid interactions in food systems. Critical Reviews in Food
Science and Nutrition 13:297-335.

Cho, CY. 1993. Digestibility of feedstuffs as a major factor in aquaculture waste
management. In: Fish Nutrition in Practice. 1V"I International Symposium on Fish
Nutrition and Feeding, Biarritz, France, June 24—2 7, I 993. (ed. INRA) pp. 365-374,
lNRA, Paris, France.

Cody, RP and J .K. Smith. 1997. Applied Statistics and the SAS Programming
Language, 4‘” edition. Prentice Hall, Inc., Upper Saddle River, New Jersey, USA, 445 pp.

Coyle, S.D., J .H. Tidwell and CD. Webster. 2000. Response of large mouth bass
Micropterus salmoides to dietary supplementation of lysine, methionine, and highly
unsaturated fatty acids. Journal of the World Aquaculture Society 31:89-95.

Dabrowski, K, P. Poczyczynski, G. KocK and B. Berger. 1989. Effect of partially or
totally replacing ﬁsh meal protein by soybean meal protein on growth, food utilization
and proteolytic enzyme activities in rainbow trout (Salmo gairdneri). New in vivo test for
exocrine pancreatic secretion Aquaculture 77 :29-49.

Daud, S.K, D. Hasbollah and AT Law. 1988. Effects of unionized ammonia on red
tilapia (Oreochromis mossambicus/O. niloticus hybrid) fry. In: The Second International
Symposium on T ilapia in Aquaculture. ( ed. by R.S.V. Pullin, T. Bhukaswan, K.
Tonguthai & J .L. Maclean), pp. 411-413, ICLARM, Manila, Philippines.

Davies, M.I., K Bush and I. Motzok. 1972. Determination of iron in low-iron diets by

colorimetric and atomic absorption procedures. Journal of the Association of Ofﬁcial
Analytical Chemists 55:1206-1210.

150

Davies, SJ. and PC. Morris. 1997. Inﬂuence of multiple amino acid supplementation on
the performance of rainbow trout, Oncorhynchus mykiss (Walbaum), fed soya based
diets. Aquaculture Research 28:65-74.

Davies, S.J., N. Thomas and KL. Bateson. 1989. The nutritional value of a processed

soya protein concentrate in diets for tilapia fry (Oreochromis mossambicus, Peters).
Barnidgeh 41 :3-1 1.

Davis, AT. and RR. Stickney. 1978. Growth responses of Tilapia aurea to dietary
protein quality and quantity. Transactions of the American Fisheries Society 1072479-
483.

Davis, D.A., D. Jirsa and CR. Arnold. 1995. Evaluation 0 soybean proteins as
replacement for menhaden ﬁsh meal in practical diets for the red drum Sciaenops
oceIlatus. Journal of the World Aquaculture Society 26:48-5 8.

DeSilva, SS. and RM. Gunasekera. 1989. Effect of dietary protein level and amount of
plant ingredient (Phaseolus aureus) incorporated into the diets on consumption, growth,
performance, and carcass composition in Oreochromis niloticus (L.) ﬁy. Aquaculture
80: 12-133.

DeSilva, S.S., R.M. Gunasekera and KF. Shim. 1991. Interactions of varying dietary
protein and lipid levels in young red tilapia: evidence of protein sparing. Aquaculture
95:305-318.

El-Dahhar, AA. and El-Shazly, K. 1993. Effect of essential amino acids (methionine and
lysine) and treated oil in ﬁsh diet on growth performance and feed utilization of Nile
tilapia, Tilapia nilotica (L.). Aquaculture and Fisheries Management 24:731-739.

El-Sayed, AM. 1989. Evaluation of semipuriﬁed test diets for Tilapia zilli ﬁngerlings.
Journal of the World Aquaculture Society 20:240-244.

El-Shamy, F .M. 1976. Analyses of gastric emptying in bluegill (Lepomis macrochirus).
Journal of the Fisheries Research Board of Canada 33: 1630-1633.

El-Shemy, MG. and RE Levin. 1997. Characterization of afﬁnity-puriﬁed trypsin from
hybrid tilapia (T ilapia nilotica/aurea). Journal of Food Biochemistry 21 :163-15.

Fauconneau, B. 1985. Protein synthesis and protein deposition in ﬁsh In: Nutrition and
Feeding in Fish. (ed C.B. Cowey, A.M. Mackie and J .G. Bell), pp. 1745, Academic
Press, Inc., Orlando, Florida, USA.

F ange, R. and D. Grove. 1979. Digestion. In: Fish Physiology, Vol. VII]. Bioenergetics

and Growth (ed. W.S. Hoar, D.J. Randall, and J .R. Brett), pp. 161-260, Academic Press,
Inc., New York, New York, USA.

151

Fish, GR. 1951. Digestion in T ilapia esculenta. Nature 167:900-901.

Fish, GR. 1960. The comparative activity of some digestive enzymes in the alimentary
canal of T ilapia and perch. Hydrobiologia 15:161-179.

Fletcher, D.J., D.J. Grove, R.A. Basimi and A. Ghaddaf. 1984. Emptying rates of single
and double meals of different food quality from the stomach of the dab, Limanda limanda
(L.). Journal of Fish Biology 25:435-444.

F lowerdew, MW. and DJ. Grove. 1979. Some observations of the effects of body
weight, temperature, meal size and quality on gastric emptying time in turbot,
Scopthalmus maximus (L.) using radiography. Journal of Fish Biology 14:229-238.

Gerking, SD. 1994. Feeding Ecology of Fish. Academic Press, Inc., San Diego, CA,
pp 416.

Gietzen, D.W. 1993. Neural mechanisms in the responses to amino acid deﬁciency.
Journal of Nutrition 123:610-625.

Goddard, S. 1995. Feed Management in Intensive Aquaculture. Chapman and Hall, New
York, New York, USA, pp 194.

Grabner, M. and R Hofer. 1985. The digestibility of the proteins of broad bean (Vicia
faba) and soya bean (Glycine max) under in vitro conditions simulating the alimentary
tracts of rainbow trout (Salmo gairdneri) and carp (Cyprinus carpio). Aquaculture
48:111-122.

Grayton, ED. and F .W.H. Beamish 1977. Effects of feeding frequency on food intake,
growth and body composition of rainbow trout (Salmo gairdneri). Aquaculture 11:159-
172.

Griﬂin, ME, P.B. Brown and AL. Grant. 1992. The dietary lysine requirement of
juvenile hybrid striped bass. Journal of Nutrition 122: 1332-1337.

Grove, DJ. and C.Crawford. 1980. Correlation between digestion rate and feeding
frequency in the stomachless teleost, Blennius pholis L. Journal of Fish Biology 16:235-
247.

Grove, D.J., L. Lozoides and J. Nott. 1978. Satiation amount, frequency of feeding and
gastric emptying rate in Salmo gairdneri. Journal of Fish Biology 12:507-516.

Gwyther, D. and DJ. Grove. 1981. Gastric emptying in Limanda limanda and the return
of appetite. Journal of Fish Biology 18:245-259.

Hamersrand, G.E., L.T. Black and J .D. Glover. 1981. Trypsin inhibitors in soy products:
modiﬁcation of the standard analytical procedure. Cereal Chemistry 58:42-45.

152

Hanley, F. 1987. The digestibility of foodstuffs and the effects of feeding selectivity on
digestibility determinations in tilapia, Oreochromis niloticus (L). Aquaculture 66: 163-
179.

Hargreaves, J.A., J .E. Rakocy and A. Nair. 1988. An evaluation of ﬁxed and demand
feeding regimes for cage culture of Oreochromis aureus. In: The Second International
Symposium on T ilapia in Aquaculture. ( ed by R.S.V. Pullin, T. Bhukaswan, K
Tonguthai & J .L. Maclean), pp. 335-339, ICLARM, Manila, Philippines.

Hendriks, H.G., T.S. Van den Ingh, A. Krogdahl, J. Olli and J.F. Koninkx. 1990. Binding
of soybean agglutinin to small intestinal brush border membranes and brush border
membrane enzyme activities in Atlantic salmon (Salmo salar). Aquaculture 91:163-170.

Holm, J .C., T. Refstie and S. Be. 1990. The effect of ﬁsh density and feeding regimes on
individual growth and mortality in rainbow trout (Oncorhynchus mykiss). Aquaculture
89:225-232.

Holmgren, S., D.J. Grove, and DJ. Fletcher. 1983. Digestion and the control of
gastrointestinal motility. In: Control Processes in Fish Physiology. (ed. J .C. Rankin, TJ.
Pitcher, and RT. Dugan), pp. 23-40, John Wiley and Sons, New York, New York, USA.

Hofer, R. and F. Scheimer. 1981. Proteolytic activity in the digestive tract of several
species of ﬁsh with different feeding habits. Oecologia 48:342-345.

Hossain, MA. and K. Jauncey. 1993. The effects of varying dietary phytic acid, calcium
and magnesium levels on the nutrition of common carp, Cyprinus carpio. In: Fish
Nutrition in Practice. I 1"” International Symposium on Fish Nutrition and Feeding,
Biarritz, France, June 24-27, 1993. (ed. INRA) pp. 705-715, INRA, Paris, France.

Ishiwata, N. 1969a. Ecological studies on the feeding of ﬁshes - VII. Frequency of
feeding and satiation amount Bulletin of the Japanese Society of Scientiﬁc Fisheries
35:979-984.

Ishiwata, N. 1969b. Ecological studies on the feeding of ﬁshes — VIII. Frequency of
feeding and growth. Bulletin of the Japanese Society of Scientiﬁc Fisheries 35:985-990.

Jackson, A.J. and BS. Capper. 1982. Investigations into the requirements of the tilapia
Sarotherodon mossambicus for dietary methionine, lysine, and arginine in semi-synthetic
diets. Aquaculture 29:289-297.

Jackson, AJ., B.S. Capper and A.J. Matty. 1982. Evaluation of some plant proteins in
complete diets for the tilapia Sarotherodon mossambicus. Aquaculture 27197-109.

Jauncey, K. & Ross, B. 1982. A Guide to Tilapia Feeds and Feeding. Institute of
Aquaculture, University of Stirling, Stirling, Scotland, 111 pp.

153

Jobling, M. 1980. Gastric evacuation in plaice, Pluronectes platessa L.: effects of dietary
energy level and food composition. Journal of Fish Biology 17 287-196.

Jobling, M. 1981. Mathematical models of gastric emptying and the estimation of daily
rates of food consumption for ﬁsh. Journal of ﬁsh Biology 19:245-257

Jobling, M. 1983. A short review and critique of methodology used in ﬁsh growth and
nutrition studies. Journal of Fish Biology 23:685-703.

Jobling, M. 1994. Fish Bioenergetics. Chapman and Hall, London, UK, pp 309.

Kapoor, B.G., H. Smit and LA. Verighina 1975. The alimentary canal and digestion in
teleosts. Advances in Marine Biology 13:109-239.

Kaushik, S. 1990. Analytical and empirical methods in diet formulation. In:
Mediterranean Aquaculture. (ed. Rosa Flos, Lluis Tort and Pere Torres), pp. 141-161,
Ellis Horwood, New York, New York, USA.

Kaushik, 8.1, J .P. Cravedi, J .P. Lalles, J. Sumpter, B. Fauconneau and M. Laroche. 1995.
Partial or total replacement of ﬁsh meal by soybean protein on growth, protein utilization,
potential estrogenic or antigenic effects, cholesterolemia and ﬂesh quality in rainbow
trout, Oncorhynchus mykiss. Aquaculture 133:257-274.

Knuckles, B.E., D.D. Kuzrnicky and AA Betschart 1985. Effect of phytate and partially
hydrolyzed phytate on in vitro protein digestibility. Journal of Food Science 50:1080—
1082.

Kono, H. and Y. Nose. 1971. Relationship between the amount of food taken and growth
in ﬁshes — 1. Frequency of feeding for a maximum daily ration. Bulletin of the Japanese
Society of Scientiﬁc Fisheries 37: 169-175.

Krogdahl, A., TE. Lea and Jan Olli. 1994. Soybean proteinase inhibitors affect intestinal
trypsin activities and amino acid digestibilities in rainbow trout (Oncorhynchus mykiss).
Comparative Biochemistry and Physiology 107(A):215-219.

Kubaryk, J .M 1980. Effects of diet, feeding schedule, and sex on food consumption,
growth. and retention of protein and energy by tilapia. Ph.D. Dissertation, Auburn
University, Auburn, Alabama.

Kugler, C.O.J. and J. Pequignot. 1988. Growth and histological studies on the liver and
anterior intestine of Nile tilapia (Oreochromis niloticus) fed on mixed foods: Daphnia
magna, Chlorella vulgaris and commercial carp pellets. In: The Second International
Symposium on T ilapia in Aquaculture. ( ed by R.S.V. Pullin, T. Bhukaswan, K
Tonguthai & J .L. Maclean), pp. 361-365, ICLARM, Manila, Philippines.

154

Lanari, D., E.D’Agaro and C. Turri. 1998. Use of nonlinear regression to evaluate the
effects of phytase enzyme treatment of plant protein diets for rainbow trout
(Oncorhynchus mykiss). Aquaculture 161:345-356.

Latta, M. and M. Eskin 1980. A simple and rapid colorimetric method for phytate
determination. Journal of Agricultural and Food Chemistry 28: 1313-1315.

Liener, LE. 1994. Implications of antinutritional components in soybean foods. Critical
Reviews in Food Science and Nutrition 34:33-67.

Martin, E.A., J .V. Nolan, Z. Nitsan and DJ. Farrell. 1998. Strategies to improve the
nutritive value of rice bran in poultry diets. IV. Effects of addition of ﬁsh meal and
microbial phytase to duckling diets on bird performance and amino acid digestibility.
British Poultry Science 39:612-621.

McClain, W.R. and DM. Gatlin, III. 1988. Dietary zinc requirement of Oreochromis
aureus and effects of dietary calcium and phytate on zinc bioavailability. Journal of the
World Aquaculture Society 19:103-108.

Mroz, 2., AW. Jongbloed and PA. Kemme. 1994. Apparent digestibility and retention
of nutrients bound to phytate complexes as inﬂuenced by microbial phytase and feeding
regimen in pigs. Journal of Animal Science 72:126-132.

Montgomery, DC. 1991. Design and Analysis of Experiments, 3"! edition. John Wiley
and Sons, New York, New York, USA, pp 649.

Moriarity, D.J.W. 1973. The physiology of digestion of blue-green algae in the cichlid
ﬁsh Tilapia nilotica. Journal of Zoology, London 171125-40.

Moriarity, CM. and D.J.W. Moriarity. 1973a. Quantitative estimation of the daily
ingestion of phytoplankton by T ilapia nilotica and Haplochromis nigripinnis in Lake
George, Uganda Journal of Zoology, London 171 : 15-23.

Moriarity, D.J.W. and CM. Moriarity. 1973b. The assimilation of carbon from
phytoplankton by two herbivorous cichlid ﬁshes: T ilapia nilotica and Haplochromis
nigripinnis. Journal of Zoology, London 171241-45.

Nagase, G. 1964. Contribution to the physiology of digestion in T ilapia mossambica
Peters: digestive enzymes and the effects of diets on their activity. Zeitschriﬁ ﬁir
Vergleichende Physiologie 49:270-284.

Ng, WK, S.O. Hung and MA. Herold. 1996. Poor utilization of dietary free amino acids
by white sturgeon. Fish Physiology and Biochemistry 15:131-142.

155

Nose, T., S. Arai, D. Lee and Y. Hashimoto. 1974. A note on amino acids essential for
growth of young carp. Bulletin of the Japanese Society of Scientiﬁc Fisheries
40:903-908.

NRC (National Research Council). 1993. Nutrient Requirements of Fish National
Academy Press, Washington, DC, USA, pp 114.

Ofojekwu, PC. and C. Ejike. 1984. Growth response and feed utilization in the tropical
cichlid Oreochromis niloticus niloticus (Linn) fed on cottonseed-based artiﬁcial diets.
Aquaculture 42:27-36.

Olli, J .J , K. Hjelrneland and A. Krogdahl. 1994. Soybean trypsin inhibitors in diets for
Atlantic salmon (Salmo salar, L.): effects on nutrient digestibilities and trypsin in pyloric

caeca homogenate and intestinal content. Comparative Biochemistry and Physiology
109(A):923-928.

Olli, J .J . and A. Krogdahl. 1995. Alcohol soluble components of soybeans seem to reduce
fat digestibility in ﬁsh-meal-based diets for Atlantic salmon, Salmo salar L. Aquaculture
Research 26:831-835.

Olli, J .J ., A. Krogdahl and A. VAbeno. 1995. Dehulled solvent-extracted soybean meal as
a protein source for Atlantic salmon, Salmo salar L. Aquaculture Research 26: 167-174.

Papatryphon, E., R.A. Howell and J .H. Scares, Jr. 1999. Growth and mineral absorption
by striped bass Morone saxatil is fed a plant feedstuff based diet supplemented with
phytase. Journal of the World Aquaculture Society 30: 161-173.

Papoutsoglou S.E. & Tziah G. 1996. Blue tilapia (Oreochromis aureus) growth rate in
relation to dissolved oxygen concentration under recirculated water conditions.
Aquacultural Engineering 15, 181-192.

Post, G.W. 1983. Textbook of Fish Health. TFH Publications, Inc., Neptune City, New
Jersey, USA, pp 256.

Powell, MR. 1972. Cage and raceway culture of striped bass in brackish water in
Alabama. In: Proceedings of the 26'” Annual Conference of the Southeast Association of
Game and Fish Commissioners 26:553-565.

Rackis, J .J . 1974. Biological and physiological factors in soybeans. Journal of American
Oil Chemist's Society 51:16lA-174A.

Ramseyer, L., D. Garling, Jr., G. Hill and J. Link. 1999. Effect of dietary zinc
supplementation and phytase pre-treatrnent of soybean meal or corn gluten meal on
growth, zinc status and zinc-related metabolism in rainbow trout, Oncorhynchus mykiss.
Fish Physiology and Biochemistry 20:251-261.

156

Reddy, N.R., M.D. Pierson, S.K. Sathe and UK. Salunkhe. 1989. Phytate in Cereals and
Legumes.CRC Press, Inc., Baton Raton, Florida, USA, pp 152.

Reigh, RC. and SC. Ellis. 1992. Effect of soybean and ﬁsh-protein ratios on growth and
boy composition of red drum (Sciaenops ocellatus) fed isonitrogenous diets. Aquaculture
104:279-292.

Richardson, N.L., D.A. Higgs, RM Beames, J .R. McBride. 1985. Inﬂuence of dietary
calcium, phosphorus, zinc and sodium phytate level on cataract incidence, growth and
histopathology in juvenile Chinook salmon (Oncorhynchus tshawytscha). J oumal of
Nutrition 115:553-567.

Riche, M. and PB. Brown. 1996. Availability of phosphorus from feedstuffs fed to
rainbow trout, Oncorhynchus mykiss. Aquaculture 142:269-2 82.

Robbins, K. 1986. A method, SAS program, and example for ﬁtting the broken-line to
growth data. University of Tennessee Agricultural Experiment Station, University of
Tennessee, Knoxville, Tennessee, USA, pp 8.

Robinson, E.H., S.D. Rawles, P.W. Oldenburg and RR. Stickney. 1984. Effects of
feeding glandless or glanded cottonseed products and gossypol to Tilapia aurea
Aquaculture 38: 145-154.

Rumsey, G.L., J .G. Endres, P.R. Bowser, KA. Eamst-Koons, D.P. Anderson and AK
Siwicki. 1995. Soy protein in diets of rainbow trout: effects on growth, protein
absorption, gastrointestinal histology, and nonspeciﬁc serologic and immune response.
In: Nutrition and Utilization Technology in Aquaculture. (ed by C.E. Lim and J. Sessa),
pp 166-188, AOCS Press, Champaign, Illinois, USA.

Rumsey, G.L. and HG. Ketola. 1975. Amino acid supplementation of casein in diets of
Atlantic salmon (Salmo salar) fry and of soybean meal for rainbow trout (Salmo
gairdneri) ﬁngerlings. Journal of the Fisheries Research Board of Canada 32:422-426.

Saddiqui, A.Q. and AH. Al-Harbi. 1999. Nutrient budgets in tanks with different
stocking densities of hybrid tilapia. Aquaculture 170:245-252.

Sadiku, S.O.E. and K. J auncey. 1995. Digestibility, apparent amino acid availability and
waste generation potential of soybean ﬂour: poultry meat meal blend based diets for
tilapia, Oreochromis niloticus (L.), ﬁngerlings. Aquaculture Research 26:651-657.

Satterlee,L.D. and R. Abdul-Kadir. 1983. Effect of phytate content on protein nutritional
quality of soy and wheat bran proteins. Lebensmittel-Wissenschaﬂ and Technologie
16:8-14.

SAS (1979) Statistical Analysis System User {s Guide. SAS Institute, Inc., Cary, North
Carolina, 278 pp.

157

Santiago, CB. and RT. Lovell. 1988. Amino acid requirements for growth of Nile
tilapia. Journal of Nutrition 118:1540-1546.

Satoh, S. W.E. Poe and RP. Wilson. 1989. Effect of supplemental phytate and/or
tricalcium phosphate on weight gain, feed efﬁciency and zinc content in vertebrae of
channel catﬁsh. Aquaculture 80:155-161.

Sebastian, 8., SP. Touchburn, ER. Chavez and PC. Lagué. 1997. Apparent digestibility
of protein and amino acids in broiler chickens fed a com-soybean diet supplemented with
microbial phytase. Poultry Science 76: 1760-1769.

Seymour, EA 1989. Devising optimum feeding regimes and temperatures for the
warmwater culture of eel, Anguilla anguilla L. Aquaculture and Fisheries Management
20:311-323.

Shiau, S., J. Chuang and C. Sun 1987. Inclusion of soybean meal in tilapia (Oreochromis
niloticus x 0. aureus) diets at two protein levels. Aquaculture 65:251-261.

Shiau, S., C. Kwok, J. Hwang, C. Chen and S. Lee. 1989. Replacement of ﬁshmeal with
soybean meal in male tilapia (Oreochromis niloticus x 0. aureus) ﬁngerling diets at a
suboptimal protein level. Journal of the World Aquaculture Society 20:230-235.

Shiau, S., S. Lin, S. Yu, A. Lin and C. Kwok. 1990. Defatted and full-fat soybean meal as
partial replacements for ﬁsh meal in tilapia (Oreochromis niloticus x 0. aureus) diets at
low protein level. Aquaculture 86:401-407.

Singh, M and AD. Krikorian. 1982. Inhibition of trypsin activity in vitro by phytate.
Journal of Agricultural Food Chemistry 30:799-800.

Singh, RP. and AK Srivastava 1984. Effect of feeding frequency on the growth,
consumption and gross conversion efﬁciency in the siluroid catﬁsh, Heteropneustes
fossilis (Bloch). Bamidgeh 36:80-91.

Siraj, S.S., Kamaruddin, Z., Satar, MKA. & Kamaruddin, MS. (1988) Effects of
feeding frequency on growth, food conversion and survival of red tilapia (0.
mossambicus x 0. niloticus) hybrid fry. In: The Second International Symposium on
T ilapia in Aquaculture. (ed by RSV. Pullin, T. Bhukaswan, K. Tonguthai & J .L.
Maclean), pp. 383-386, ICLARM, Manila, Philippines.

Smith, LS. 1989. Digestive functions in teleost ﬁshes. In: Fish Nutrition, 2'” edition. (ed.
by J .E. Halver), pp. 331-421, Academic Press, Inc., San Diego, CA, USA.

Spinelli, J ., C.R. Houle and J .C. Wekell. 1983. The effect of phytates on the growth of

rainbow trout (Salmo gairdneri) fed puriﬁed diets containing varying quantities of
calcium and magnesium. Aquaculture 30:71-83.

158

Spotte, SH. 1970. Fish and Invertebrate Culture: Water Management in Closed Systems.
John Wiley and Sons Inc., New York, New York, USA, pp 145.

Tacon, A.G.J., J .V. Haaster, P.B. Featherstone, K Kerr and A.J. Jackson. 1983. Studies
on the utilization of full-fat soybean and solvent extracted soybean meal in a complete
diet for rainbow trout. Bulletin of the Japanese Society of Scientiﬁc Fisheries 49: 1437-
1443.

Tacon, A.G.J. and A.J. Jackson. 1985. Utilisation of conventional and unconventiorml
protein sources in practical ﬁsh feeds. In: Nutrition and Feeding in Fish. (ed. C.B.
Cowey, AM. Mackie and J .G. Bell), pp. 119-145, Academic Press, Inc., Orlando,
Florida, USA

Tidwell, J .H., C.D. Webster and RS. Knaub. 1991. Seasonal production of rainbow trout,
Oncorhynchus mykiss (Walbaum), in ponds using different feeding practices.
Aquaculture and Fisheries Management 22:335-341.

Tsevis, N., S. Klaoudatos and A Conides. 1992. Food conversion budget in sea bass,
Dicentrarchus labrax, ﬁngerling under two different feeding frequency patterns.
Aquaculture 101:293-304.

Tyler, AV. and RS. Dunn. 1976. Ration, growth, and measures of somatic and organ
condition in relation to meal frequency in winter ﬂounder, Pseudopleuronectes

americanus, with hypotheses regarding population homeostasis. Journal of the Fisheries
Research Board of Canada 23:63-75.

Vaintraub, LA. and VP. Bulmaga. 1991. Effect of phytate on the in vitro activity of
digestive proteinases. Journal of Agricultural Food Chemistry 39:859-861.

Viola, S. and Y. Arieli. 1983. Nutrition studies with tilapia (Sarotherodon).
1. Replacement of ﬁshmeal by soybeanmeal in feeds for intensive tilapia culture.
Barnidegeh 3529-17.

Viola, S., J. Arieli, U. Rappaport and S. Mokady. 1981. Experiments in the nutrition of
carp: replacement of ﬁshmeal by soybeanmeal. Bamidegeh 33:35-49.

Viola, S., S. Mokady and Y. Arieli. 1983. Effects of soybean processing methods on the
growth of carp (Cyprinus carpio). Aquaculture 32:27-38.

Walton. M.J. 1985. Aspects of amino acid metabolism in teleost ﬁsh. In: Nutrition and
Feeding in Fish. (ed. C.B. Cowey, AM. Mackie and J .G. Bell), pp. 47—67, Academic
Press, Inc., Orlando, Florida, USA.

Walton, Ml, C.B. Cowey and J .W. Adron. 1982. Methionine metabolism in rainbow

trout fed diets of differing methionine and cystine content. Journal of Nutrition 112:1525-
1535.

159

Waters Manual. 1989. Pico-Tag Method (ed. by S.A Cohen, M. Meys & T.L. Tarvin),
Waters Chromatography Division, Millipore Corp, Milford, Massachusetts.

Wee, KL. and S. Shu 1989. The nutritive value of boiled full-fat soybean in pelleted
feed for Nile tilapia Aquaculture 81 1303-314.

Wheaton, F.W. 1993. Aquacultural Engineering. Krieger Publishing Co., Malabar, FL,
pp 708.

Wilson, RP. 1989. Amino acids and Proteins. In: Fish Nutrition, 2", edition. (ed. by J .E.
Halver), pp. 111-151, Academic Press, Inc., San Diego, CA, USA.

Wilson, R.P., D.E. Harding and UL. Garling, Jr. 1977. Effect of dietary pH on amino
acid utilization and the lysine requirement of ﬁngerling channel catﬁsh. Journal of
Nutrition 107:166-170.

Windell, J .T., D.O. Norris, J .F. Kitchell and J .S. Norris. 1969. Digestive response of
rainbow trout, Salmo gairdneri, fed three pelleted diets. J oumal of the Fisheries Research
Board ofCanada 26:1801-1812.

Yamada, A, K. Takano and I. Kamoi. 1993. Puriﬁcation and properties of protease from
tilapia stomach. Bulletin of the Japanese Society of Scientiﬁc Fisheries 5911903-1908.

Yamada, A, K. Takano and I. Karnoi. 1991. Puriﬁcation and properties of proteases from
tilapia intestine. Bulletin of the Japanese Society of Scientiﬁc Fisheries 571155 1-1557.

160

APPENDICES

APPENDIX 1

RECIRCULATING SYSTEM DIAGRAM

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

     
   
  

A ‘ " "’ ——————— " " " ’1
A l
r l n | '
I
0 I
Last . Solids
Culture Unit I Removal
L 1/ II
I-"".—""“—‘:‘J Filtrimon
r I ‘ Settling
«I
First 1
1
Culture Unit 0
t. J
A
Rotating
Biological

 

Contactor

 

 
  

+ _ -
\NH4->N02 >N03

\

V

Figure 29. Conceptual ﬂow diagram of the recirculating system used in the experiments.

161

APPENDIX 2

Processing Steps for Solvent Extracted Soybean meal obtained from Zeeland Farm
Services, Zeeland, MI.

Prepgation:

Conditioning -— heat to 150 F

Dry soybeans — heat to 150 - 190 F

Crack beans in half

Dehull beans by aspiration

Crack beans to V4 and crack beans again to 1/8
Dehull and recondition by heating to 150 F
Flaking — down to 15/ 1000 of an inch

NQMPP’N?‘

Extraction:

Extract with hexane solvent
Desolventize
Toast

Dry
Cool

MPPNS‘

162

APPENDIX 3

Table 23. Predictive equations and R2, (n=12) for linear regression and non-linear
regression models applied to growth and efﬁciency parameters in juvenile

tilapia fed diets containing untreated, or phytase treated, solvent extracted SBM

substituted into a reference diet at 25, 50, 75, or 100 % of the dietary protein.
Parameters include speciﬁc growth rate (SGR), feed efﬁciency (FE), protein

efﬁciency ratio (PER), and apparent net protein utilization (ANPU).

 

 

 

Parameter Model Predictive Equation RI

Weight Gain

Phytase Treated Linear y = -4. 16x + 817.2 0.621
Quadratic y = 0.047rr2 — 10.07x + 965 0.661

Untreated Linear y = -4.20x + 846.5 0.623
Quadratic y = 0.004x2 — 4.71 + 859 0.624

_s_G_1;

Phytase Treated Linear y = -0.0113x + 4.02 0.600
Quadratic y = 700%2 — 0.02x +4.24 0.611

Untreated Linear y = -0.0114x + 4.10 0.610
Quadratic y = -1x10'5x2 —0.01x + 4.06 0.610

Feed Efﬁciency

Phytase Treated Linear y = -o.ooosx + 0.872 0.709
Quadratic y = 400%2 — 0.002x + 0.859 0.710

Untreated Linear y = -0.003x + 0.957 0.715
Quadratic y = -5rr10‘5x2 + 0.003x +0813 0.795

Egg

Phytase Treated Linear y = -0.0114x +2.78 0.824
Quadratic y = -2x10'5x2 — 0.0085x +2.71 0.825

Untreated Linear y = -0.0093x +2.84 0.727
Quadratic y = -lx10'3x2 +0.0086x +2.40 0.813

M

Phytase Treated Linear y = -0.166x + 42.14 0.857
Quadratic y = -0.0002x2 — 0.189x + 42.74 0.857

Untreated Linear y = -0.117x +4306 0.624
Quadratic y = .0.0026x2 + 0.208x + 34.90 0.779

 

163

 

:00 :00 :00 :00 :00 :00 :0: :00 :00 :00

00.00000 00.00000 :0.0a00.0 00.00000 0000000 :00200 ::.0a:0.0 00.00000 00000.0 02000.0 050::0
:00 :00 :00 :00 :00 :00 :00

00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.0...000 <2 $2 $2 03000
:0: :00 :00 :00 :00 :00 :00 :0: :00 :00

000000 00.00000 00.00000 00.00000 00.00000 0:00:00 00.0000 00.0000 ::.:a00.0 00000.0 05000
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00

00.00000 0:.00000 00.00000 00.00000 00000.0 00000.0 00.00000 00.00000 00.00000 00.00000 000::0
:00 :00 :00 :00 :00 :00 :00 :0:

00020.0 00030.0 0:00:00 00000.0 00000.0 :0.:a00.0 00000:.0 00.00000 :07: S2 50::
:00 :00 :00 :00 :00 :00 :00 :00

00000:.0 00.0000 00.00000 00.00000 00.00000 00.00000 00.0000 00.0200 <2 <2 5200

$0000.:

007: <2 <2 <2 $2 $2 <2 <2 <2 <2 2000::
0: 2:0 0 9:0 0 3:0 0 3:0 0 3:0 0 000 08:0 0 3:0 0303 30000000:

 

0 Nun—lam:

.00—030000000 .0008 000000 0 3:30:00 0.50:: 0 000 N 020600
203 032: 0 000 0 02000: 00020 05 60.0 000080 00:0: 00 00:32:00 0.00:: 0 05 .0 .0 .0 .: .00 05 00.680 0.0000
088:8 0003 00:08.3 00:00:00 00:00:00 00 00:80:: «00000000 000050200 5505 0000802 .0000 5 00000000 00 0:000:80
0000000 E0005 000:0 20:0 000 00000200 000000 a0000:0000 0:0 00 088:8 000.: .00 000.050 00 Amm 0V 00::0> I0 0002 .00 0300.

164

 

:00 :00 :00 :00 :00 :00 :00 , :00 :00 :00

00000.0 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.0200 00.0020 05000
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00

00.00000 00.00000 00.00000 00.00:.0 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 03000
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00

00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.0000 00.00000 05000
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00

00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00000.0 00000.0 00.00000 00.00000 000000
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00

00.00000 00.00000 00.00000 00.00000 00.00:.0 00.00000 00.00000 00.00000 00.00000 00.00000 500:
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00

00.00000 00.00000 00.00000 20000.0 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 500 0.0
:00 :00 :00 :00 :00 :00 :00 :00 :00 :00 000.80

00.00000 0.00000 0.00000 00.00000 00.00000 00.00000 00.00000 00000.0 00.00000 00.00000 200000
0: 000 0 000 0 000 0 000 0 000 0 000 0 000 0 000 0.00000 $000000

 

590000000 000:0 000000 0 350:8 0.500 0 000 m 003.000

203 0000 0 05 0 0030. 00020 030 00.0 000080 00.0.0 5 0:50:00 0500 0 05 .0 .0 .0 .0 .00 05 00.0080 8008
0800:00 0003 00:90.3 0080:00 020.800 .00 0000000 000000000 00000000000 0.00.05 00000002 .0000 0: 000000000 00 00000080

0000000 0.0003 0000 00000 000 000000000 00008 .00—$000000 000 00 0080:00 0000000 00.0000 .00 0mm 00 00300, $0 0002 .3 0.00.0.

n Nun—73::

165

 

00.00 :00 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 0. 005000
0000000 00.000: .0 00.00000 00.00000 00.00000 0: 0000.0 0.0000 .0 0 080000
0000000 00.00 00.0 0: 00000 00.00000 00.00000 0: 0000.0 00.00000 0 085000
00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 0 055000
00.00000 00.00000 00.00000. 00.00000 00.00000 00.00000 0000000 0080.000
00.00000 00.0000 .0 0000000 0: .00 :00 00.00000 00.00000 00.00000 0 058000
00.00000 00.00000 00.00000 00.00000 00.00000 00.00 0 2. 0000000 0 085000
00.00000 00.00000 00.00000. 00 0000.0. 00.00000 0: .00 0 0 .0 0.00: .0 0 0005000
00.00000 00.00000 00.00000 00.00000 00.00000 00.00000 00.000: .0 0 008000
0: 0000.0 00.00000 00.00000 00.00 :00 00.00000 00.00000 00.00000 30880
000000 00.00 .00 00.00000 0: 0000.0 00.00000 00.00000 00.00000 0.0002
00 0000.0 00.00000 00.00000 0: 0000.0 00.00000 0: 0000.0 0: 0000.0 00020.0.
043%

080.08.. 080.080: 080.0000 080-000 080-000 080-080 05080

008$ w 0000: 0 050m 0 0.50: N 000: 0 00000.: We 000.0000

 

00000000 0000 0: 00000000000000 0 000000000 000_0>
05080 00000 00 0:36:00 0500 0 000 .0 .0 .0 .0 .00 000 00080 20000 0008 20; 000820082 .0 2:000 .0 0000080
00 0000 00000 00.0. 000 000000000 0000000 00000 05 :0 0000 00000 00 00000000. 000000. 30000000 .00 0mm 00 002000 :0 0002 .3 200,—.

0 Nun—Zak:

166

"ITIEﬂ'ﬂﬂﬂjﬂmﬂiﬂ'MﬂﬁWs