Nutritional Evaluation of Low Glucosinolate Mustard

Meals (Brassica juncea) in Broiler Diets

R. W. NEWKIRK,* H. L. CLASSEN,*,1 and R. T. TYLER†

*Department of Animal and Poultry Science, 72 Campus Drive, University of Saskatchewan, Saskatoon, Saskatchewan,
Canada S7N 5B5, and †Department of Applied Microbiology and Food Science, 51 Campus Drive,
University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5A8

ABSTRACT Experiments were conducted to evaluate
the nutritional value of meal derived from low
glucosinolate cultivars of mustard (Brassica juncea) in
comparison to samples of canola meal (Brassica napus,
Brassica rapa). Samples of Brassica seed (four B. juncea,
one B. napus, and one B. rapa) were processed using
laboratory procedures to produce oil-extracted meals,
which were examined for composition (DM basis), and
nutritional value for broiler chickens as judged by
nutrient retention (AMEn, ileal protein digestibility) and
performance. Meals derived from B. juncea contained
more CP and less total dietary fiber (TDF) on a dry basis
than either B. napus or B. rapa, 45.9 vs 44.6 and 43.1% CP
and 27.22 vs 29.47 and 29.67% TDF, respectively. Acid
detergent fiber (ADF) and neutral detergent fiber (NDF)
(Key words: canola meal, mustard meal, low glucosinolate,
nitrogen-corrected apparent metabolizable energy, broiler chick)
levels for B. juncea and B. rapa meals were similar to
each other, but lower than those of B. napus, 12.79 and
13.20 vs 20.6% ADF, and 21.15 and 19.58 vs 29.47% NDF,
respectively. Brassica juncea meals contained more
glucosinolates than B. napus and B. rapa, 34.3 vs 21.8 and
25.5 mmol/g total glucosinolates, respectively. Brassica
juncea meals were equal or superior to B. napus and B.
rapa meals for AMEn and apparent ileal protein
digestibility. Similarly, broilers fed B. juncea meals grew
as quickly and converted feed to BW gain as efficiently
to 21 d of age as those birds fed B. napus and B. rapa
meals. Feeding meal from B. rapa reduced growth rate
and gain to feed ratio. In conclusion, the nutritional
value of meal from low glucosinolate mustard was equal
or superior to that of canola meal samples derived from
B. napus and B. rapa cultivars.
1997 Poultry Science 76:1272–1277

INTRODUCTION demonstrated that the nutritional value of MM was

Mustard (Brassica juncea) has been grown as a
specialty crop in western Canada for many years but has
not been used extensively for oil production despite the
relatively high oil content of its seed. In comparison to
rapeseed (canola), mustard has agronomic advantages,
such as drought tolerance and disease resistance, which
give this crop considerable oil production potential. The
recent development of cultivars with oil characteristics
similar to canola and low glucosinolate levels (Love et
al., 1990) has enhanced the potential of this crop. If this
crop is to be used extensively for oil production, the
nutritional value of the meal derived after oil extraction
should be ascertained for livestock.
The nutritional value of mustard meal (MM) has not
been extensively studied. However, earlier research has
Received for publication October 7, 1996.
Accepted for publication March 10, 1997.
1To whom correspondence should be addressed.
2103-111 Research Drive, Saskatoon, SK, Canada, S7N 3R2.
similar to canola meal (CM) if the MM was first treated
with ammonia to reduce its glucosinolate content (Bell et
al., 1984; Blair, 1984). Presumably, the genetic reduction
of glucosinolate content would produce similar results
but the nutritional value of this type of MM has not
been accurately assessed.
The objective of this research was to compare the
nutritional value of four samples of low glucosinolate
MM to two samples of CM (Brassica napus and Brassica
rapa) for broiler chickens. All seed samples were
processed in the same manner to eliminate the potential
effects of different oil extraction procedures.
MATERIALS AND METHODS
Meal Preparation
Six Brassica samples were tested. One B. napus (Excel),
one B. rapa (Parkland), and four B. juncea samples
arbitrarily called MM1 to MM4. All samples were grown
in Canada during the 1994 growing season and were
supplied by the Saskatchewan Wheat Pool.2 Strains MM1

1272
 NUTRITIONAL VALUE OF MUSTARD MEAL FOR BROILERS 1273
TABLE 1. Composition and calculated analysis of the experimental diets

Nutrient retention study Performance study

Ingredients and content Test diet Reference diet Brassica Soybean meal
(%)
Test ingredient 40.00 0 20.00 0
Soybean meal 6.97 12.00 15.51 32.71
Wheat 46.67 80.38 54.38 58.69
Canola oil 1.74 3.00 6.23 4.47
L-lysine HCl 0 0 0.070 0
DL methionine 0 0 0.172 0.215
Limestone 1.847 1.847 1.505 1.712
Dicalcium phosphate 1.558 1.558 1.382 1.465
Salt 0.345 0.345 0.381 0.368
Vitamin premix1 0.100 0.100 0.100 0.100
Mineral premix2 0.070 0.070 0.070 0.070
Choline chloride 0.100 0.100 0.100 0.100
Avizyme Tx3 0.100 0.100 0.100 0.100
Chromic oxide 0.500 0.500 0 0
Calculated nutrient content4
CP 26.8 16.16 23.3 23.3
AMEn, kcal/kg 2,595 3,091 3,050 3,050
Available P 0.519 0.450 0.450 0.450
Ca 1.27 1.02 1.00 1.00
1Supplied per kilogram of diet: vitamin A (retinyl acetate + retinyl palmitate), 15,700 IU; vitamin D, 3,140 IU;
vitamin E (dl-a-tocopheryl acetate), 40 IU; menadione, 2.5 mg; thiamine, 2.5 mg; riboflavin, 10 mg; niacin, 84 mg;
pyridoxine, 4 mg; vitamin B12, 0.016 mg; pantothenic acid, 15.43 mg; folic acid, 1.1 mg; and biotin, 0.14 mg.
2Supplied per kilogram of feed: iron, 56 mg; zinc, 77 mg; manganese, 94.5 mg; copper, 7 mg; iodine, 0.91 mg
and selenium, 0.21 mg.
3Finnfeeds International Ltd., Marlborough, Wiltshire, SN8 1AA, UK.
4Only approximate, because each of the meals varied in composition. Calculated based on composition of
MM3.

and MM2 were the same strain grown on two different
locations, whereas MM3 and MM4 were strains of low
glucosinolate mustard that had undergone further selec-
tion.
Oil was extracted from seed samples (approximately 25
kg per sample) at POS Pilot Plant Inc.3 using a bench top
process designed to mimic commercial processing. The
seed was thoroughly cleaned to remove foreign material
and undeveloped seeds prior to processing. Water was
sprayed on the seed to increase the moisture content to 8
to 10% and the samples were then placed in a sealed
container and allowed to equilibrate for 16 h. The samples
were then flaked, heated to 90 to 100 C in a microwave
oven (4.5 to 7 min) and transferred to a convention oven
for 20 to 30 min (final temperature was 89 to 93 C). The
samples were then pressed with a Komet Oil Expeller4
and extracted with hexane over a period of 4 h. The meal
was desolventized in a fume hood for 2 to 3 d. The meal
was once again heated to 97 to 106 C in a microwave oven
(7 to 8 min) and toasted in a convection oven for 40 min
(final temperature was 97 to 103 C). An exception to the
above procedure was that one-third of the B. rapa seed was
not pressed prior to extraction because of a breakdown of
3118 Veterinary Road, Saskatoon, SK, Canada, S7N 2R4.
4IBG Monforts Gmbh and Co., Mönchengladback, Post fach 20
D-41 208, Germany.
the press. As a result, the meal from B. rapa had a slightly
higher oil content and underwent slightly less heat
treatment (about 2 min of 80 to 100 C).
Nutrient Retention
Nitrogen-corrected apparent metabolizable energy was
determined according to the methods of Sibbald and
Slinger (1963) using commercial broiler cockerels raised to
28 d of age on commercially prepared feed and housed in
raised wire floor cages. After 28 d, the birds were
randomly placed two per cage in cages designed for feces
collection and were fed the test diet for 7 d. The test
ingredient was included in the diet at a level of 40%.
Chromic oxide was added to the diet as an indigestible
marker. The diet AMEn was multiplied by a factor of
1.0484 as described by Sibbald and Slinger (1963) to
compensate for the test diet and reference diet having 4.84
g of premix (limestone, dicalcium phosphate, salt, vitamin
premix, choline chloride, Avizyme Tx, and chromic oxide)
added per 100 g of macroingredients (soybean meal,
wheat, and canola oil). The AMEn of the test ingredient
was calculated as: AMEn of reference diet – [(AMEn of
reference diet – AMEn of test diet)/0.4]. The composition
of the test and reference diets are shown in Table 1. Each
treatment was replicated six times with two birds per
08 53, replicate. Feed and water were consumed ad libitum. Feces
were collected daily for the last 3 d and frozen upon
1274 NEWKIRK ET AL.
TABLE 2. Composition of low glucosinolate Brassica napus (Excel), Brassica rapa (Parkland),
Brassica juncea (MM1 to MM4), and soybean (SBM) meals on a dry matter basis

Composition B. napus B. rapa MM1 MM2 MM3 MM4 SBM

(%)
Crude protein 44.6 43.1 46.3 47.2 45.0 45.1 51.7
Oil content 0.5 1.0 0.6 0.4 0.4 0.4 ND1
Acid detergent fiber 20.06 13.20 12.24 12.52 13.29 13.12 6.67
Neutral detergent fiber 25.72 19.58 21.93 19.82 21.46 21.38 9.06
Total dietary fiber 29.47 29.67 27.90 25.55 27.76 27.68 16.76
(mmol/g)
Glucosinolates
3-Butenyl 3.4 4.6 22.2 22.6 20.3 19.6 ND
4-Pentenyl 0.4 3.5 1.3 1.7 0.8 0.9 ND
2-OH-3-Butenyl 7.6 9.2 1.5 3.5 1.5 1.6 ND
2-OH-4-Pentenyl 0.1 1.4 0 0.1 0 0 ND
Total aliphatics 11.5 18.7 25.0 27.9 22.6 22.1 ND
3-Methylindolyl 1.1 0.2 0 0.1 0.2 0.2 ND
4-OH-3-Methylindolyl 9.2 6.4 4.7 4.0 3.2 3.1 ND
Total indolyls 10.3 6.6 4.7 4.1 3.4 3.3 ND
Allyl 0 0.2 6.9 2.0 7.1 7.4 ND
4-OH-Benzyl 0 0 0 0 0.3 0.2 ND
1ND = not determined.

collection. Fecal samples from each replicate were pooled
on the final day and dried in a forced air oven at 50 C prior
to grinding and chemical analysis. Upon final collection
the birds were killed by cervical dislocation and the
contents of the distal ileum (defined as the last half of the
section between the Meckel’s diverticulum and the ileo-
ceco-colonic junction) were collected and freeze-dried for
determination of ileal protein digestibility as described by
Ten Doeschate et al. (1993).
Growth Performance Study
An experiment was carried out to examine the effect of
inclusion of 20% experimental meal in the diet on broiler
performance. Day-old commercial (Hubbard · Hubbard)
male broiler chicks were selected to minimize variability
in chick weight. Chicks of acceptable weight were
randomly allocated (four chicks per pen) to Jamesway5
battery brooders. Treatments were assigned to pens
according to a completely randomized experimental
design with each treatment replicated six times. Initial
room temperature was set at 32 C and was gradually
reduced throughout the experiment according to stan-
dard husbandry practice. Feed and water were consumed
ad libitum throughout the experiment.
During the experiment, dead birds were recorded and
weighed, as well as average intake up to the time of death.
Final body weight and feed consumption was determined
on Day 21. Average gain was determined by subtracting
the average initial weight from the average final weight.
Gain to feed was calculated by dividing average weight
gain, including weight gain of any dead birds, by average
feed intake.
5Jamesway Manufacturing Co., Ft. Atkinson, WI 53538.
Treatments consisted of two canola meals and four B.
juncea meals added to a wheat based diet at a level of 20%.
The Brassica-based test diets were formulated to meet or
exceed NRC (1994) requirements based on the nutrient
specifications of MM3, as it contained approximately
average CP level and AMEn relative to those of the other
Brassica meals. Another treatment was based on a diet
containing soybean meal as the sole protein concentrate
and was formulated to have nutrient specifications
identical to the Brassica-based diets. Diet compositions are
shown in Table 1.
Analytical Procedures
Acid detergent fiber (ADF), CP, moisture, and gross
energy were measured according to the methods of the
Association of Official Analytical Chemists (1980). Diet
and fecal chromic oxide content was measured according
to the methods of Fenton and Fenton (1979) to determine
digestibility. Neutral detergent fiber (NDF) and total
dietary fiber (TDF) were determined using the methods of
Mongeau and Brassard (1979) and Theander and Wester-
lund (1986), respectively. Glucosinolates were measured
according to the methods of Daun et al. (1989).
Statistical Analysis
Data were analyzed by one-way ANOVA using the
General Linear Models (GLM) procedure of the SAS
Institute (1989). Variables with significant F tests (P £ 0.05)
were compared using Duncan’s multiple range test
(Duncan, 1955).
RESULTS
The composition of Brassica and soybean meals are
shown in Table 2. The B. juncea meals contained slightly
 NUTRITIONAL VALUE OF MUSTARD MEAL FOR BROILERS 1275

FIGURE 2. Regression of protein digestibility and total dietary fiber

FIGURE 1. Regression of AMEn and total dietary fiber in Brassica
for Brassica meals. Protein digestibility = –1.567 TDF + 121.6, P < 0.03,
meals. AMEn = –183 TDF + 7,179.9, P < 0.02, R2 = 0.78 and SD of the
R2 = 0.73 and SD of the coefficient and intercept being 0.477 and 13.37,
coefficient and intercept being 48.77 and 1,367.5, respectively.
respectively.

more protein (average 45.9%) than the B. napus (44.6%)
and B. rapa meals (43.1%). The B. juncea meals had
similar levels of ADF and NDF fiber (average 12.80%
ADF, 21.2% NDF) in comparison to B. rapa (13.2% ADF,
19.58% NDF) but lower than B. napus (20.1% ADF, 25.7%
NDF). Brassica juncea meals had slightly lower levels of
TDF (average 27.2%) than meals from B. napus (29.5%)
and B. rapa meal (29.7%).
Brassica juncea meals were higher in total aliphatic
glucosinolates (average 24.2 mmol/g) than meals from B.
napus (11.5 mmol/g) and B. rapa (18.7 mmol/g). Meals
from B. juncea contained more 3-butenyl glucosinolates
(average 21.2 mmol/g) than B. napus (3.4 mmol/g) and B.
rapa (4.6 mmol/g) and accounts for the difference in the
level of aliphatic glucosinolates between the Brassica
species. Brassica juncea meals were lower in indol
glucosinolates (3.9 mmol/g) than either B. napus (10.3
mmol/g) or B. rapa (6.6 mmol/g). All meals except MM3
and MM4 were free of 4-OH-benzyl glucosinolates.
All meals from B. juncea had numerically if not
significantly higher AMEn (2,216 kcal/kg average) than
either of the meals from B. napus (1,832 kcal/kg) or B.
rapa (1,557 kcal/kg) (Table 3). The AMEn was negatively
affected by TDF in Brassica meals (Figure 1). Meal MM2
resulted in higher ileal protein digestibility (82.99%)
than B. napus (75.34%) and B. rapa (76.72%; Table 3).
Meal MM1 had significantly higher protein digestibility
(78.02%) than B. napus but not B. rapa. Strains MM3 and
MM4 had similar protein digestibility values to those of
B. napus and B. rapa. Protein digestibility was negatively
affected by TDF (Figure 2) in the Brassica meals.
Broilers fed the four samples of B. juncea meals and
the B. napus meal had similar weight gains (average 628
g vs 617 g, respectively) (Table 3). Strains MM2, MM3,
MM4, and B. napus meal resulted in weight gain similar
to soybean meal (640, 617, and 654 g for B. juncea, B.
napus, and soybean meal, respectively). All meals except
B. rapa resulted in a similar gain to feed ratio (Table 3).
The B. rapa meal resulted in a gain to feed ratio of 0.673
g:g, whereas the average for the other meals was
0.724 g:g.

TABLE 3. Nutrient retention and growth performance (between 0 to 21 d) of broiler chicks

fed low glucosinolate Brassica napus (Excel), Brassica rapa (Parkland),
Brassica juncea (MM1–MM4) and soybean meals

Ileal protein
Feedstuff AMEn digestibility Weight gain Gain to feed
(kcal/kg)1 (%) (g) (g:g)
Brassica napus 1,832d 75.34c 617ab 0.711ab
Brassica rapa 1,5573 76.72bc 536c 0.673b
MM1 2,171bc 78.02b 589bc 0.720a
MM2 2,382ab 82.99a 637ab 0.731a
MM3 2,011cd 76.65bc 640ab 0.731a
MM4 2,302ab 76.39bc 645ab 0.732a
Soybean meal 2,476 83.22a 654a 0.719a
Pooled SEM 79 0.82 20 0.016
a–dMeans in the same column with no common superscript differ significantly (P < 0.05).
1Dry matter basis.
1276 NEWKIRK ET AL.
DISCUSSION
Meals produced from B. juncea had considerably
higher levels of glucosinolates than B. napus or B. rapa;
however, the levels would still fall within the definition
of canola meal (less than 30 mm aliphatic glucosinolates/
g of dry, oil-free seed; Bell, 1993). Strains MM3 and
MM4 were low in 4-OH-benzyl glucoinsolates (0.3 and
0.2 mmol/g, respectively) as compared to commercial
meals that range from 0.25 to 7.05 mmol/g (Bell and
Keith, 1991) and the other meals had undetectable levels.
4-OH-benzyl glucosinolates are found in certain weed
seeds commonly found in canola and so the low levels
indicate that the samples used in this experiment were
relatively free of weed seeds.
Brassica juncea meals appear to be slightly higher in
CP than B. napus or B. rapa and this is consistent with
the findings of Simbaya et al. (1995). The CP levels of the
resulting meals were slightly higher than commercially
processed meals with the average crude protein content
of canola meals (B. napus and B. rapa) being 43.9%,
whereas Bell and Keith (1991) in a survey of commercial
canola meals found an average of 41.9%. The higher
levels of protein may be a result of superior cleaning
prior to processing. The elimination of small seeds and
other non seed material would result in increased
protein content (Jensen et al., 1995, Simbaya et al., 1995).
Meals derived from B. juncea had higher levels of
protein and lower levels of fiber than the canola meals
indicating that protein is higher at the expense of fiber.
Simbaya et al. (1995) also demonstrated that TDF and
protein level are negatively correlated (r = –0.82) in B.
juncea meals.
The B. napus meal was higher in ADF and NDF but
not TDF than the B. rapa meal. The B. napus sample used
was a brown-seeded variety, whereas the B. rapa was
yellow-seeded. The fiber composition of brown- and
yellow-seeded varieties are dissimilar. Yellow-seeded
Brassica species are lower in lignin and polyphenols but
contain more nonstarch polysaccharides (NSP) (Slomin-
ski et al., 1994). As ADF and NDF underestimate cell
wall residues, particularly neutral detergent soluble
polysaccharides, but because TDF more accurately
reflects the total fiber residue content (Carré and
Brillouet, 1986), it is understandable that yellow-seeded
Brassica species have lower ADF and NDF but not TDF.
The low NDF:TDF ratio of B. rapa (0.66) relative to B.
napus (0.87) and the B. juncea meals (0.78) indicate this B.
rapa sample was high in neutral detergent soluble
polysaccharides.
The superior energy utilization of B. juncea is
probably related to the reduced levels of TDF, as the
AMEn was negatively related to TDF. This result is in
agreement with the findings of Jensen et al. (1995). One
exception appears to be the B. rapa meal, as it had the
lowest AMEn but a TDF value very similar to that of the
B. napus sample.
The protein in meal derived from B. juncea was
utilized as effectively as the meal from B. napus and B.
rapa. However, protein digestibility varied considerably
between meal types. Reasons for the variation are not
clear but TDF and protein digestibility were negatively
related, which is consistent with the findings of Jensen et
al. (1995) using mice. This finding implies dietary fiber is
impeding protein utilization; however, neither NDF or
ADF were correlated with protein digestibility, suggest-
ing that TDF is a more useful predictor of dietary fiber
effects. Reasons for dietary fiber affecting protein
utilization are not clear, but the indigestible protein
fraction in the meal may be bound to, or encapsulated
by, fibrous components of the meal (Jensen et al., 1995).
It is important to note that this relationship was
calculated using a small number of samples (n = 5)
derived from three different Brassica species. As a result,
conclusions cannot be drawn until confirmed in experi-
ments specifically designed to study the effects of fiber
on protein utilization.
Strains MM1 and MM2 were derived from seed of the
same cultivar but grown in different locations. MM2 had
a slightly higher protein level and significantly higher
ileal protein digestibility than MM1. This result indicates
that both nutrient content and availability can be
affected by growing environmental conditions. Both
NDF and TDF were higher for MM1 than MM2; it can
be speculated that environmental conditions affect these
characteristics, which, in turn, influence protein utiliza-
tion. Although previous research has examined the
variability of protein or amino acid availability from
canola meal samples, these data are confounded by
potential processing effects. Research on genotype and
environmental effects on meal quality is required for
both canola and mustard meals.
Inclusion of B. juncea meals at a level of 20% resulted
in weight gain and feed conversion at least equal to B.
napus and soybean meals and superior to B. rapa meal
(Table 3). An exception were broilers fed MM1, which
did not grow faster than those fed the B. rapa meal. Both
B. rapa and MM1 meals produced slower growth rate
than birds fed soybean meal. The depression in
performance due to the inclusion of B. rapa may, in part,
be due to the low utilization of energy. Reductions in
gain to feed ratio with reduced AMEn were expected but
the decrease in weight gain was not anticipated.
Meals derived from four samples of B. juncea seed
were equal or superior to meals derived from B. napus
and B. rapa in terms of protein content, nutrient
utilization (AMEn, protein digestibility) and feeding
value for broiler chickens. Based on this research, meal
derived from mustard meal presumably could be
combined with canola meal to produce a product of
comparable quality. As there is no evidence to indicate
that mustard meal and canola meals would be affected
differentially by processing technique, the results of this
research should also be applicable to meal derived from
commercial processing.
An influence of genotype and environment on the
nutritional value of meal is suggested from this research.
 NUTRITIONAL VALUE OF MUSTARD MEAL FOR BROILERS
Experiments designed to more accurately assess this
interaction are required to more fully understand the
factors affecting the nutritional quality of canola and
mustard meals.
ACKNOWLEDGMENTS
This work was supported by the Saskatchewan Wheat
Pool (Saskatoon, SK, Canada, S7N 3R2). The technical
assistance of Randy Kruger, Dawn Abbot, and Robert
Gonda is gratefully acknowledged. The authors appreci-
ate the help of Doug Korver with manuscript prepara-
tion.
REFERENCES
Association of Official Analytical Chemists, 1980. Official
Methods of Analysis of Analytical Chemists. 13th ed.
Association of Official Analytical Chemists, Arlington, VA.
Bell, J. M., 1993. Factors affecting the nutritional value of
canola meal: A review. Can. J. Anim. Sci. 73:679–697.
Bell, J. M., and M. O. Keith, 1991. A survey of variation in the
chemical composition of commercial canola meal
produced in Western Canadian crushing plants. Can. J.
Anim. Sci. 71:469–480.
Bell, J. M., M. O. Keith, J. A. Blake, and D. I. McGregor, 1984.
Nutritional evaluation of ammoniated mustard meal for
use in swine feeds. Can. J. Anim. Sci. 64:1023–1033.
Blair, R., 1984. Nutritional evaluation of ammoniated mustard
meal for chicks. Poultry Sci. 63:754–759.
Carré, B., and J.-M. Brillouet, 1986. Yield and composition of
cell wall residues isolated from various feedstuffs used for
non-ruminant farm animals. J. Sci. Food Agric. 37:341–351.
Daun, J. K., D. R. DeClercq, and D. I. McGregor, 1989. Analysis
of Glucosinolates in Canola and Rapeseed. Agriculture
Canada, Canadian Grain Commission, Winnipeg, MB,
Canada.
1277
Duncan, D. B., 1955. Multiple range and multiple F tests.
Biometrics 11:1–42.
Fenton, T. W., and M. Fenton, 1979. An improved procedure
for the determination of chromic oxide in feed and feces.
Can. J. Anim. Sci. 59:631–634.
Jensen, S. K., Y.-G. Liu, and B. O. Eggum, 1995. The effect of
heat treatment on glucosinolates and nutritional value of
rapeseed meal in rats. Anim. Feed Sci. Technol. 53:17–28.
Love, H. K., G. Rakow, J. P. Raney, and R. K. Downey, 1990.
Development of low glucosinolate mustard. Can. J. Plant
Sci. 70:419–424.
Mongeau, R., and R. Brassard, 1979. Determination of neutral
detergent fiber, hemi-cellulose, cellulose and lignin in
breads. Cereal Chem. 56:437–441.
National Research Council, 1994. Nutrient Requirements of
Poultry. 9th rev. ed. National Academy Press, Washing-
ton, DC.
SAS Institute, 1989. SAS/STAT User’s Guide. Version 6,
Fourth Edition. SAS Institute Inc., Cary, NC.
Sibbald, I. R., and S. J. Slinger, 1963. A biological assay for
metabolizable energy in poultry feed ingredients together
with findings which demonstrate some of the problems
associated with the evaluation of fats. Poultry Sci. 42:
313–325.
Simbaya, J., B. A. Slominski, G. Rakow, L. D. Campbell, R. K.
Downey, and J. M. Bell, 1995. Quality characteristics of
yellow-seeded Brassica seed meals: protein, carbohydrates,
and dietary fiber components. J. Agric. Food Chem. 43:
2062–2066.
Slominski, B. A., L. D. Campbell, and W. Guenter, 1994.
Carbohydrates and dietary fiber components of yellow
and brown seeded canola. J. Agric. Food Chem. 42:
704–707.
Ten Doeschate, R.A.H.M., C. W. Scheele, V.V.A.M. Schreurs,
and J. D. van der Klis, 1993. Digestibility studies in broiler
chickens: Influence of genotype, age, sex and method of
determination. Br. Poult. Sci. 34:131–146.
Theander, O., and E. A. Westerlund, 1986. Studies on dietary
fiber. 3. Improved procedures for analysis of dietary fiber.
J. Agric. Food Chem. 34:330–336.