![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
3
*
School of Fisheries, University of Washington, Seattle, WA 98195 and
Hagerman Fish Culture Experiment Station, University of Idaho, Hagerman, ID 83332
3To whom correspondence should be addressed.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
KEY WORDS: • phosphorus • requirement • urine • response criteria • rainbow trout
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, to optimize growth of the fish with economical diets containing
all essential nutrients above the minimum requirements. An excess of
nutrients in diets has rarely been considered a problem, unless it
leads to an economical disadvantage. Due to an increasing concern for
global environmental pollution in the last few decades, however,
increasing attention has been directed toward the unretained (excess)
portion of nutrients in diets, particularly phosphorus
(P)4
, a critical nutrient that is present in water discharged from
aquaculture facilities. Excess portions of P in diets impact the
environment via an excessive eutrophication of the aquatic ecosystem.
Because of this new consideration, fish nutritionists now specify P in
diets to meet the biological requirement (i.e., maximizing growth) as
well as environmental guidelines (i.e., minimizing excretion). As a
result, it is essential to know the minimum dietary requirement of P
very accurately, and to formulate feeds accordingly.
Numerous researchers have investigated the dietary P requirements for
various fish species using young fish of <10 g in body wt
(NRC 1993
). However, nutrient requirements in most
animals generally decrease as they become older or larger. Little is
known about the specific dietary requirements of P for large fish. In
commercial aquaculture, large fish consume >90% of the total feed
used in a production cycle and excrete a proportional amount of P into
the effluent. Therefore, obtaining an accurate estimation of the
minimum dietary requirement of P for large fish is necessary to
minimize P excretion into the aquatic environment without the risk of
possible deficiency problems in cultured fish.
Numerous response criteria, including growth, tissue saturation levels,
whole body retention, bone calcification or strength, activities of
certain enzymes, disease resistance and/or clinical signs have been
used to estimate P status of young fish and their dietary requirement
(reviewed in Asgard and Shearer 1997
). The problem in estimating the
dietary requirements for large fish is their much slower growth rate
compared to small or young fish. This is mainly due to different feed
intakes. For example, fish weighing 405 g require a 6.7 times
longer period than fish weighing 1.7 g to consume the same weight
of feed per their body wt (NRC 1993
). In addition, since
weight gain per feed consumed (feed efficiency, FE) is considerably
lower in large fish than in small fish, large fish take even longer
time to increase their body wt relative to their initial wt. Feeding
large fish with research diets for such a long period is
time-consuming and expensive. Also, results obtained with large
fish can be misleading if the feeding period is not sufficiently long.
Developing a new and sensitive approach was, therefore, necessary to
accurately estimate the dietary requirements of P for large fish.
Nonfecal P excretion (primarily urinary) is a new criterion that has
not been used previously to estimate the dietary P requirement in any
animal species. Studies with terrestrial animals have shown that
urinary excretion of water-soluble vitamins and free amino acids
increases as intake increases. For example, with increasing intake of
thiamin or riboflavin, the quantity excreted in urine at first
increases only slightly. Then, the excretion begins to rise rapidly
once the intake exceeds a certain level (Boisvert et al. 1993
, Wang and Yudkin 1940
). This critical point
of urinary excretion is thought to represent the level of intake at
which organ saturation occurs (Bro-Rasmussen 1958
).
P-homeostasis also depends on the mechanisms that govern renal
excretion of P (Dennis 1992
, Lee et al. 1981
, Moser et al. 1981
).
Preliminary studies in our laboratory examining various physiological
responses of fish to reduced dietary P concentrations indicated that
urinary excretion of inorganic P (Pi) was a more sensitive indicator
than others such as P, glucose, glucose-6-phosphate, ATP, creatine
phosphate, total lipids, total cholesterol, glycogen and acetoacetate
concentrations in blood, skeletal muscle, liver or feces
(Sugiura et al. 2000
). However, collecting urine
directly from fish has not been successful due to the small volume of
collectable urine (by stripping), large differences of urinary Pi
concentration over time, stresses of the confinement in a metabolic
chamber and the catheterization of urinary bladder, and a high
variability among individual fish. The present method, though not
strictly specific to urinary P, largely eliminated these shortcomings
in the previous studies. This paper presents a new, sensitive method to
estimate the minimum dietary requirement of P for large rainbow trout
based on nonfecal excretion of P and applied this technique to fish of
varying nutritional histories.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Two groups of rainbow trout, Oncorhynchus mykiss (House
Creek strain), differing in size were used to estimate dietary P
requirements. One group of fish, subsequently called small fish, had an
initial mean body wt of 203 ± 0.65 g (n = 117 fish), whereas the other group of fish, called large fish, had an
initial mean body wt of 400 ± 2.88 g (n = 63
fish). Thirteen small fish or 7 large fish were stocked in each of 18
140-L holding tanks receiving spring water at 5 L/min. The water had
constant temperature (15.0–15.5°C), dissolved oxygen (ca. 9.0 mg/L),
P (ca. 0.011 mg/L) and calcium (Ca, 28–36 mg/L) concentrations. A
14:10 h diurnal photoperiod was maintained using fluorescent lighting.
Fish were fed commercial trout feed (Silver Cup Fish Food, Murray, UT)
for 2 wk before feeding experimental diets in order to acclimate them
to the rearing environment. Nine semipurified diets, differing in P
concentration (Table 1
), were prepared as moist pellets, stored at -20°C, and used within 3
wk of preparation. The commercial and experimental diets were fed once
daily at 0800 h. The daily feeding levels (dry basis), calculated
from the estimated digestible energy (DE) levels of the diets, fish
body wt and the water temperature according to Smith (1989)
, were 9.0
and 9.7 g/kg body wt (small fish) or 8.5 and 9.2 g/kg body wt (large
fish) for the commercial and experimental diets, respectively.
|
), and no
adverse effects were observed. Feces were continuously collected into
the bottom of an unstirred fecal collection column affixed to each
tank. Water (~100 mL) was collected from each tank 24 h after
stocking at d 0, 3, 6, 9 and 12 in small fish or at d 0, 3 and 6 in
large fish on the experimental diets. For the other days, fish stayed
in the holding tanks receiving spring water continuously. The metabolic
tanks were thoroughly cleaned before each of the 24-h collection
periods to minimize bacterial assimilation of P and nitrogen (N) in the
water. In the tanks containing no fish, preliminary tests showed no
measurable loss of ammonia (NH4Cl used) or P
(KH2PO4 used) from the
water. At the end of the feeding trial (d 13, in small fish only), feed
was withheld for 24 h. Blood samples were collected from the
caudal vessels of five randomly sampled anesthetized (tricaine methane
sulfonate, 100 mg/L tank water) fish per treatment into heparinized
syringes, which were immediately chilled on ice. Experiment 2 (Diet History).
The P requirements of three groups of rainbow trout (House Creek
strain) differing in diet history were examined. The first group
(subsequently called P-sufficient fish: initial mean body wt, 390.4
± 4.44 g, n = 120 fish) had received
commercial trout feed (Silver Cup Fish Food) fortified with P for 3 wk
before the experiment. The diet contained 11.5 g total P and
419 g crude protein/kg dry diet with apparent digestibility
coefficients (%) for P, protein and dry matter of 60.9 ± 0.56,
90.9 ± 0.20 and 81.4 ± 0.22 (n = 10 tanks),
respectively. The second group (P-deficient fish: initial mean body wt,
372.9 ± 3.65 g, n = 120 fish) had received a
P-deficient diet for 5 wk before the experiment. The diet had the
same ingredient composition as the experimental diet of the lowest P
concentration (Table 1)
, but the diet was supplemented further with
aluminum hydroxide as a phosphate-sequestering substance at 10 g/kg
diet (Al:P, 1.93:1, mol/mol). The analytical composition of this diet
was 2.01 g total P, 428 g crude protein/kg dry diet with
apparent digestibility coefficients (%) for P, protein and dry matter
of 64.0 ± 0.72, 96.3 ± 0.28 and 83.1 ± 0.57
(n = 10 tanks), respectively. The third group (starved
fish: initial mean body wt, 349.0 ± 3.20 g, n
= 120 fish) had been starved for 8 wk before the experiment.
Before starvation, the fish had an average body wt of 384.4 ± 3.53 g (n = 120 fish). Fish in each group were
visually graded before the experiment to exclude maturing fish.
Experimental fish had unique body compositions reflecting these
previous diet regimens as shown in Table 2
.
|
, were
prepared as dry compressed pellets. Subsequently, the diets were stored
at 0–5°C and used within 2 mo of manufacture. Twelve fish
(P-sufficient, P-deficient or starved fish) were placed in the
metabolic tank. The fish were fed experimental diets (dry basis) at
either 7.8 g (for P-sufficient and P-deficient groups) or
8.5 g (for starved group)/(kg body wt · d). The starved fish
received more diets because their feeding responses were more active
than those of the other fish. The other feeding procedures were the
same as described in expt. 1. Water and fecal collection procedures
were also the same as described in expt. 1, except that fish stayed in
the metabolic tank and were not transferred from the holding tank after
feeding to avoid possible handling stress. Thus, water was sampled
initially (immediately after feeding) and at the end (24 h after that)
to measure net amounts of P and N excreted by the fish. In addition, a
199-L auxiliary tank was connected to each metabolic tank to reduce
possible stresses of ammonia during the 24-h period. The number of fish
stocked in each tank was larger in expt. 2 than in expt. 1. Water and
fecal samples were collected at d 0, 3, 6, 9 and 12 (and 15 in starved
group) from tanks receiving the experimental diets. For the other days
when samples were not collected, clean spring water was continuously
supplied to the tanks. The entire system was thoroughly cleaned before
each of the 24-h collection periods. There was no mortality or sign of
disease during the experiments. All fish were treated in accordance
with the guidelines approved by the Animal Care and Use Committee of
the University of Idaho. Analytical methods.
Feed and fecal samples were dried (105°C for 6 h) and analyzed
for total N by a LECO FP-428 Nitrogen determinator (Leco Instruments,
St. Joseph, MI) and for gross energy by a 1241 adiabatic oxygen bomb
calorimeter (Parr Instrument Company, Moline, IL). Subsamples of dried
feeds and feces were analyzed for the concentrations of chromium
(Cr) (Bolin et al. 1952
), P (Taussky and Shorr 1953
) and Ca (Sigma Diagnostics, St. Louis, MO). In expt. 2,
initial (whole) body composition was determined on seven randomly
sampled fish per treatment. Total N, P and Ca concentrations were
measured as mentioned above; fat was extracted by Goldfisch extractor
using methylene chloride as the solvent and weighed; ash was obtained
by igniting dried samples at 550°C for 12 h and weighed. In
expt. 1, water samples were stored at 0–4°C after collection and
analyzed within 1 d for Kjeldahl-N and Pi (using stannous
chloride) according to American Public Health Association et al. (1989)
. In expt. 2, water samples were immediately acidified after
collection and analyzed within 1 wk for Kjeldahl-N. Total P and Pi
were determined on non-acidified samples stored at -20°C for up to 2
wk. For total P, water samples were digested with sulfuric and nitric
acids, neutralized and analyzed for P by the stannous chloride method.
Since collected water samples were apparently free of suspended solids,
they were not filtered. Plasma and serum were obtained by centrifuging
the blood at 1000 x g for 10 min within 1 h of
the collection and assayed for Pi (Taussky and Shorr 1953
) and Ca (Sigma Diagnostics).
Calculations and statistical procedures.
The fecal losses of P and N and the DE values of diets were estimated
based on their net absorption (apparent digestibility), using chromic
oxide as an indicator and the feed (P and N) consumed by the fish. The
actual amount of diet consumed by the fish was calculated by collecting
the uneaten feed pellets immediately after each feeding. Available P in
diets was calculated by subtracting the fecal loss from the total P of
each diet based on apparent absorption. Thus, available P (g P/kg diet)
in this study is defined as net absorbed P or apparently available P
(=truly available P –endogenous fecal P). N retention (gain) by fish
was estimated by subtracting N loss in water and in feces from total N
consumed by fish. The body weight gain was calculated based on N
(protein) content of normal fish (Table 2)
.
The requirement of P was calculated based on the following criteria: (i) Nonfecal excretion, which was based on the Pi concentration in the tank water, primarily represented urinary excretion. Concentrations of Pi (rather than total P) were used to estimate the minimum dietary requirement based on the assumption that Pi is the primary form of urinary P. The value indicated the X-intercept of the linear regression line of "floating" (nonzero) points. When "floating points" were not linear, the regression line was determined in the linear region. (ii) The 95%-body saturation was the concentration of dietary P that was necessary to achieve 95% of the maximum P retention based on polynomial regression lines (either the second or the third order was used). Retention was estimated as Intake - (fecal loss + nonfecal loss). Total P was used to estimate the retention. (iii) The 95%-plasma saturation was the concentration of dietary P that was required to achieve 95% of the plasma saturation. These values are presented on a per diet basis (g/kg dry diet), N gain basis (g/g N gain) and DE basis (g/MJ DE). The values expressed on a per diet basis are standardized based on FE (standardized value or requirement coefficient = measured value/FE) to present the requirement values when FE is 1 (FE = g weight gain of fish/g dry feed consumed).
Analytical data pertaining to plasma and fish samples, net absorption of P and N and DE values were subjected to paired t tests, single-factor ANOVA followed by Newman-Keuls multiple comparison test and linear and polynomial regression analyses. Variance homogeneity was tested by Bartlett’s test before ANOVA. Data were not transformed since heterogeneous variance was primarily due to outliers (not excluded). Outliers were considered true responses rather than errors in sampling or analyses. Statistical analyses were performed using Prism, version 2.01 (GraphPad Software, San Diego, CA). Treatment effects were considered significant at P < 0.05. Mean values are expressed as mean ± SEM (n) throughout the text unless otherwise stated.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There was no appreciable increase of P in the tank water of small fish
fed diets containing up to 5.85 g available P/kg dry diet or of
large fish fed up to 4.42 g available P/kg dry diet. P excretion
increased abruptly in the tank water when fish were fed diets of higher
P concentrations (Fig. 1
). This pattern of P excretion observed on d 3 of consuming the test
diets remained constant for the subsequent sampling days (d 6, 9 and
12).
|
Although total fecal N was unaffected by the dietary P concentration,
nonfecal N excretion was higher in fish consuming diets of very low P
concentration than in fish consuming diets of higher P concentration.
Mean N retention in small fish (excluding the lowest P group,
n = 8 tanks) was 44.0 ± 0.70% of dietary intake
or 24.7 ± 0.39 g N/kg dry diet consumed. Average N retention
in large fish (excluding 2 lowest P groups, n = 7
tanks) was 45.8 ± 0.55% of dietary intake or 25.7 ± 0.28 g N/kg dry diet consumed. The net absorption of N ranged from
97.4 to 98.0% (n = 9 tanks). The maximum P/N retention
ratio (mmol P/mol N) of all treatments was 100.4 in large fish and
128.7 in small fish. The apparent digestibility of energy in small fish
ranged from 81.5 to 84.5% (n = 9 tanks), which did not
correlate with the dietary P concentration (P = 0.24,
linear effect). Estimated weight gain (kg/kg dry diet consumed or FE)
was 0.924 ± 0.014 in small fish or 0.958 ± 0.010 in large
fish (excluding the lowest P group that had a lower value). Estimates
of requirements of available P based on the plasma or body saturation
(95% level) were higher than those determined from nonfecal excretion
(Table 3
).
|
Concentrations of Pi in tank water remained very low when fish were fed
diets containing available P in an amount up to 3.65 g/kg dry diet (in
P-sufficient fish and starved fish) or 4.79 g/kg dry diet (in
P-deficient fish) (Fig. 2
). Above this threshold concentration of dietary P, P-sufficient
fish and starved fish excreted excess portions of dietary P linearly
with dietary P levels; thus the regression coefficient (slope) was
close to 1 (i.e., no further retention of dietary P). Conversely,
P-deficient fish continued to retain dietary P even after the point
at which they started to excrete P. Therefore, the regression
coefficient was < 1, and the response of nonfecal excretion to
dietary P concentration was not linear. The maximum P/N retention ratio
(mmol P/mol N) of all treatments was 60.5 in P-sufficient fish,
107.2 in P-deficient fish and 65.4 in starved fish. At high P
concentrations in the diet, P retention (per N retention) decreased as
was seen in expt. 1 (Fig. 3
).
|
|
). Growth rate did not differ significantly among treatments (P
concentration in the diet) within each group except that the
P-sufficient fish that consumed the diet of the lowest P level had
lower weight gain (P = 0.03) than those that consumed
the diet containing 10.09 g available P/kg dry diet. This was due to a
slight difference in feed consumption since N retention (per N intake)
did not differ (P = 0.14) among treatments.
|
The minimum dietary requirement of available P estimated based on
nonfecal excretion was higher for P-deficient fish than for
P-sufficient fish or starved fish; however, the difference was
smaller than when body saturation (retention) was used as the response
criteria to estimate the minimum requirement (Table 5
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). This is especially
the case for studies of short feeding durations such as the present
study. P-deficient fish continued to retain a portion of dietary P
above the minimum dietary requirement (X-intercept) as indicated by the
regression line, which had a coefficient < 1 (Fig. 2)
. This
suggests that estimating the dietary P requirement based on the balance
or on the maximum retention (body saturation) results in a serious
overestimation (Fig. 3)
. Nonfecal excretion of Pi seems to be most
resistant to differing P-status of fish; however, P-deficient
fish showed a higher requirement estimate than P-sufficient or
starved fish. It was, therefore, necessary to examine the P status of
the experimental fish. P status can be measured directly by analyzing
the P level in the body (or specific tissues) of the fish. A more
accurate way, used in this study, is to measure the P/N retention
ratio. When normal fish are growing normally, the P/N retention ratio
should be consistent with the P/N ratio of the whole body. If fish
retain P more than N (high P/N ratio), the fish are in a
P-deficient state, and the reverse suggests that the fish are in a
P-excess or malnourished state. In expt. 2, P-deficient fish
indeed had a much higher P/N retention ratio than the other groups of
fish or P/N ratio of the whole body of normal fish. In expt. 1,
the P/N retention ratio and the estimated P requirement per gN gain
were unexpectedly high, especially in small fish (Table 3)
. In
addition, the linear regression of the nonfecal excretion had lower
slope, and the values for 95% body P saturation were higher than those
of nonfecal excretion (all similar to P-deficient fish in expt. 2).
These data suggest that the fish in expt. 1 were deficient in P. The
estimated values in expt. 1, therefore, must be interpreted as the
minimum dietary requirement of P for P-deficient fish rather than
that for normal fish.
The minimum dietary requirement is defined as the amount of a nutrient
required to prevent overt deficiency signs including reduced growth.
This is evidently different from the saturation requirement.
Animals can grow for a certain period when fed diets deficient in
essential nutrients by utilizing body stores. For example, juvenile
trout (initial body wt 8.3 g) fed P-deficient diets grew
normally until whole body P levels were depleted by growth dilution
(Hardy et al. 1993
). These authors estimated the
threshold concentration to be around 17 g/kg dry whole body, which is
ca. 75% of normal. This percentage is in close agreement with the body
P level of P-deficient fish compared to that of normal fish (Table 2)
. If normal (initial) fish contain X g P/kg fish body, and if FE is
1, the diet should contain at least the same amount of available P (X
g/kg diet) to maintain the normal P level in the body, whereas the
theoretical minimum requirement of available P can be as low as 0.75X
g/kg diet. However, due to initial body P stores, if
growth is used as a response variable, the estimated minimum dietary
requirement of P can be much lower than this level, depending on the
magnitude of increase in size. In the study of Rodehutscord (1996)
, the
fish grew from 50 to 200 g. Theoretically, the dietary requirement
of available P, if FE is 1, will be (200X–50X)/150 = X g/kg diet
to maintain a normal P level in the body, while growth will not be
reduced as long as the diet contains (200X · 0.75–50X)/150 = 0.67X g/kg diet. If starved, underfed or thin fish are
used, the value will be even lower (200X · 0.75–50X · 1.14)/150
= 0.62X g/kg diet, due to their high initial body P
content (Table 2)
. These values are considerably lower than the
maintenance requirement, X g/kg diet, or the theoretical minimum
requirement 0.75X g/kg diet described above. Quadrupling
body wt (from 50 to 200 g) is the highest magnitude of increase
reported thus far in fish P requirement studies; however, according to
the foregoing calculation, a three-fold gain may not be sufficient
if growth is the response criterion.
For growing fish, specific dietary nutrient requirements depend on the
growth rate. Fish grow faster when fed high-grade (high
performance) feeds rather than low-grade feeds. This indicates that
the former feed requires more P per unit dry diet than the latter. When
fish require 6 g available P/kg in a diet that has a FE of 1.4,
the requirement will be 3 g/kg diet when the FE decreases to 0.7.
Therefore, expressing dietary requirements on per unit dry diet basis,
currently the most common way to express nutrient requirements in fish
diets, is not accurate. Unless the measured requirement is reported
correctly, it cannot be accurate even if it is measured accurately.
Since body P and N levels are fairly constant throughout the life
stages of rainbow trout (Shearer 1984
,
1994
), the available P requirement per weight gain or
per N gain should be similar, regardless of fish size. Standardizing
measured requirement values with weight gain or FE (gain/feed) is the
way to eliminate confounding factors associated with different
protocols among experiments, including different growth rate, FE, feed
composition, fish size and rearing conditions. This standardization,
however, does not eliminate the effect of the different diet history or
P status of fish (Tables 3
and 5)
.
Many researchers have studied the dietary P requirement of trout or
salmon using small fish. Ogino and Takeda (1978)
reported that the
dietary P requirement for rainbow trout (initial wt 1.2 g, final
wt ca. 3.5 g) was 7–8 g P/kg diet for maximum weight gain.
However, if their data are reanalyzed using a broken line regression,
the break point is about 3–4 g P/kg diet (they tested only four
dietary P levels). Because the FE of their diet was 1.1, the
standardized requirement value (requirement when FE is 1) will be as
low as 3/1.1 = 2.7 g P/kg diet. Since this value is for total
P, it will be lower if it is expressed as available P, especially for
their high Ca diet since Ca can reduce intestinal P absorption in many
species, including fish. Ketola (1975)
and Ketola and Richmond (1994)
used diets containing nonpurified ingredients and a substantial amount
of Ca, but did not measure the availability (absorption) of P. Unless
dietary P content is expressed as available (absorbed) P,
interpretation of the results is difficult. Asgard and Shearer (1997)
reported the dietary requirement of P for Atlantic salmon to be 10–11
g/kg diet. Their estimation was based on maximum P retention or body P
saturation. If the estimation is made for the dietary P level that
maintains the initial level of P in the body, the value will be lower
(i.e., ca. 9 g P/kg diet). In addition, if this value is expressed as
available P instead of total P, then the value will be even lower
(i.e., ca. 7.7 g available P/kg diet). The diet they used had a
high FE (ca. 1.45). Therefore, the standardized requirement value will
be 7.7/1.45 = 5.3 g available P/kg diet. Concerning fish
growth, maintaining the initial body P level is not necessary since, as
mentioned earlier, body P concentration can be as low as 75% of the
initial (normal) level without reducing growth. Therefore, the minimum
dietary requirement of P is 5.3 x 0.75 = 4.0 g
available P/kg diet, and this value is slightly lower but close to the
value determined in the present study based on nonfecal excretion of P
(Table 5
). Rodehutscord (1996)
constructed exponential
curves for many response variables, including growth, to present an
array of requirement values for rainbow trout, which differed with the
criteria and with the level of the plateau (ranging from 1.9 to
7.4 g P/kg diet). There is no doubt that the minimum requirement
value should differ depending on the response, e.g., mortality, growth,
and disease resistance. The time required to produce an expected
effect, however, differs from one variable to another. The minimum
dietary P requirement that maintains tissue P levels (e.g., bone P) and
that supports optimum growth may be the same, or could be different if
the feeding duration is short. For many response variables that show
dose-response relationships, researchers are required to set
(arbitrarily), for example, 95% of the plateau level on various model
curves or to use an empirical broken-line procedure to derive
specific or reasonable values. By selecting different model equations
and different plateau levels for numerous response variables with
various durations of feeding, it is possible to select virtually any
point as the requirement. Selecting an "objective" method is,
nonetheless, a highly subjective procedure. Obtaining a very clear
(sensitive) response is, therefore, critical to minimize subjectivity.
An important consideration in determining dietary requirement of nutrients is the magnitude of increase in the size of animals. Many months of feeding a large quantity of semipurified diets are required for large fish to increase their initial body size several times, making it very difficult to study the nutrient requirements of large fish with existing methods. Urinary or nonfecal excretion of P is a new response variable for estimating the dietary requirement of P, which was found to be a far more sensitive and rapid indicator than any other response criteria reported thus far. By closely monitoring nonfecal excretion of P, the minimum dietary requirement of P for large fish can be estimated within a few days. This rapidity and sensitivity will be particularly useful to estimate the requirements of P, and probably of some other nutrients, for large fish at various physiological stages and under various environmental conditions.
|
FOOTNOTES |
|---|
2 Present address: 3059F, National Fish Hatchery Rd., Hagerman, ID 83332.
4 Abbreviations used: Ca, calcium; DE, digestible energy; FE, feed efficiency; N, nitrogen; P, phosphorus; Pi, inorganic phosphorus.
Manuscript received September 2, 1999. Initial review completed October 22, 1999. Revision accepted December 14, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1. American Public Health Association, American Water Works Association & Water Pollution Control Federation Standard Methods for the Examination of Water and Wastewater 17th ed. 1989 APHA Washington, DC.
2. Asgard T., Shearer K. D. Dietary phosphorus requirement of juvenile Atlantic salmon, Salmo salar L. Aquaculture Nutr 1997;3:17-23
3. Baker D. H. Problems and pitfalls in animal experiments designed to establish dietary requirements for essential nutrients. J. Nutr. 1986;116:2339-2349
4. Boisvert W. A., Mendoza I., Castaneda C., Portocarrero L. D., Solomons N. W., Gershoff S. N., Russell R. M. Riboflavin requirement of healthy elderly humans and its relationship to macronutrient composition of the diet. J. Nutr. 1993;123:915-925
5.
Bolin D. W., King R. P., Klosterman E. W. A simplified method for the determination of chromic oxide (Cr2O3) when used as an index substance. Science 1952;116:634-635
6. Bro-Rasmussen F. The riboflavin requirement of animals and man and associated metabolic relations. Nutr. Abstr. Rev. 1958;28:1–23, 369–386
7. Dennis V. W. Phosphate homeostasis. Windhager E.E. eds. Handbook of Physiology, Section 8: Renal Physiology 1992;vol. 2:1785-1815 Oxford University Press New York, NY.
8. Hajen W. E., Beames R. M., Higgs D. A., Dosanjh B. S. Digestibility of various feedstuffs by post-juvenile chinook salmon (Oncorhynchus tshawytscha) in sea water. 1. Validation of technique. Aquaculture 1993;112:321-332
9. Hardy R. W., Fairgrieve W. T., Scott T. M. Periodic feeding of low-phosphorus diet and phosphorus retention in rainbow trout (Oncorhynchus mykiss). Kanshik S.J. Luquet P. eds. Fish Nutrition in Practice, Colloq. INRA no. 61 1993:403-412 INRA Paris, France.
10. Ketola H.G. Requirement of Atlantic salmon for dietary phosphorus. Trans. Am. Fish. Soc. 1975;104:548-551
11. Ketola H. G., Richmond M. E. Requirement of rainbow trout for dietary phosphorus and its relationship to the amount discharged in hatchery effluents. Trans. Am. Fish. Soc. 1994;123:587-594
12. Lee D. B. N., Brautbar N., Kleeman C. R. Disorders of phosphorus metabolism. Bronner F. Coburn J.W. eds. Disorders of Mineral Metabolism 1981;vol. III:283-421 Academic Press New York, NY.
13. McCay C. M. Goals in nutrition. Trans. Am. Fish. Soc. 1927;57:261-265
14. Moser S., Brautbar N., Gura V., Kleeman C. R., Massry S. G. On the mechanism of hypophosphaturia during acute and chronic dietary phosphate restriction. Mineral Electrolyte Metab. 1981;5:39-48
15. NRC (National Research Council) Nutrient Requirements of Fish 1993 National Academy Press Washington, DC.
16. Ogino C., Takeda H. Requirements of rainbow trout for dietary calcium and phosphorus. Bull. Jpn. Soc. Sci. Fish. 1978;44:1019-1022
17. Piper R. G. eds. Fish Hatchery Management 1982 U.S. Dept. of the Interior, Fish and Wildlife Service Washington, DC.
18. Rodehutscord M. Response of rainbow trout (Oncorhynchus mykiss) growing from 50 to 200 g to supplements of dibasic sodium phosphate in a semipurified diet. J. Nutr. 1996;126:324-331
19. Shearer K. D. Changes in elemental composition of hatchery-reared rainbow trout, Salmo gairdneri, associated with growth and reproduction. Can. J. Fish. Aquat. Sci. 1984;41:1592-1600
20. Shearer K. D. Factors affecting the proximate composition of cultured fishes with emphasis on salmonids. Aquaculture 1994;119:63-88
21. Smith R. R. Nutritional energetics. Halver J.E. eds. Fish Nutrition 2nd ed. 1989:1-29 Academic Press San Diego, CA.
22. Sugiura, S. H., Dong, F. M. & Hardy, R. W. (2000) Primary responses of rainbow trout to dietary phosphorus concentrations. Aquaculture Nutr. In press.
23.
Taussky H. H., Shorr E. A microcolorimetric method for the determination of inorganic phosphorus. J. Biol. Chem. 1953;202:675-685
24. Wang Y. L., Yudkin J. Assessment of the level of nutrition. Urinary excretion of aneurin at varying levels of intake. Biochem. J. 1940;34:343-352[Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
