Glufosinate herbicide-tolerant (LibertyLink) rice vs. conventional rice in diets for
growing-finishing swine
G. L. Cromwell, B. J. Henry, A. L. Scott, M. F. Gerngross, D. L. Dusek and D. W. Fletcher
J Anim Sci 2005. 83:1068-1074.
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://jas.fass.org/cgi/content/full/83/5/1068
www.asas.org
Downloaded from jas.fass.org by on January 29, 2009.
Glufosinate herbicide-tolerant (LibertyLink) rice vs. conventional
rice in diets for growing-finishing swine1,2
G. L. Cromwell*3, B. J. Henry†, A. L. Scott†, M. F. Gerngross‡,
D. L. Dusek‡, and D. W. Fletcher§
*University of Kentucky, Lexington 40546; †Bayer CropScience LP, Research Triangle Park, NC 27709;
‡Texas A&M University Food Protein Research Center, Bryan 77801; and
§Genesis Midwest Laboratories, Neillsville, WI 54456
ABSTRACT: Genetically modified (GM) rice (LibertyLink, event LLRICE62) that is tolerant to glufosinate
ammonium (Liberty) herbicide was compared with a
near-isogenic (NI) conventional medium-grain brown
rice (cultivar, Bengal) and a commercially milled longgrain brown rice in diets for growing-finishing pigs. The
GM and NI rice were grown in 2000. The GM rice was
from fields treated (GM+) or not treated (GM−) with
glufosinate herbicide. The GM− and NI rice were grown
using herbicide regimens typical of southern United
States rice production practices. The four rice grains
were similar in composition. Growing-finishing pigs
(n = 96) were fed fortified rice-soybean meal diets containing the four different rice grains from 25 to 106
kg BW. Diets contained 0.99% lysine initially (growing
phase), with lysine decreased to 0.80% (early finishing
phase) and 0.65% (late finishing phase), when pigs
reached 51 and 77 kg, respectively. The percentage of
rice in the four diets was constant during each of the
three phases (72.8, 80.0, and 85.8% for the growing,
early-finishing, and late-finishing phases, respectively).
There were six pen replicates (three pens of barrows
and three pens of gilts) and four pigs per pen for each
dietary treatment. All pigs were slaughtered at the termination of the study to collect carcass data. At the end
of the 98-d experiment, BW gain, feed intake (as-fed
basis), and feed:gain ratio did not differ (P > 0.05) for
pigs fed the GM+ vs. conventional rice diets, but growth
performance traits of pigs fed the GM+ rice diets were
superior (P < 0.05) to those of pigs fed the GM– rice
diet (ADG = 0.86, 0.79, 0.81, and 0.85 kg/d; ADFI =
2.41, 2.49, 2.37, and 2.45 kg/d; feed:gain = 2.80, 3.17,
2.95, and 2.89 for GM+, GM−, NI, and commercially
milled rice, respectively). Carcass traits (adjusted for
final BW) did not differ (P = 0.10) among treatments
(hot carcass yield = 73.5, 72.6, 72.6, and 73.2%; 10thrib backfat = 23.0, 22.7, 21.3, and 23.8 mm; LM area =
38.6, 38.0, 38.2, and 38.1 cm2; carcass fat-free lean =
50.5, 50.5, 51.2, and 50.0%). Gilts grew slower (P < 0.05)
and were leaner (P < 0.05) than barrows. Responses to
type of rice did not differ between barrows and gilts,
with no evidence of a diet × gender interaction (P = 0.50)
for any trait. The results indicate that the glufosinate
herbicide-tolerant rice was similar in composition and
nutritional value to conventional rice for growing-finishing pigs.
Key Words: Biotechnology, Glufosinate, Herbicide-Tolerant, Pigs, Rice
2005 American Society of Animal Science. All rights reserved.
Introduction
Genetically modified (GM) crops offer a variety of
benefits for agriculture (James, 2003). One example is
insect-protected (Bt) corn, released in the mid-1990s,
which is resistant to the European corn borer (Betz et
1
Journal Paper No. 04-07-114 of the Kentucky Agric. Exp. Stn.
LibertyLink and Liberty are registered trademarks of Bayer CropScience, LP.
3
Correspondence: Dept. of Anim. Sci. (phone: 859-257-7534; fax:
859-303-1027; e-mail: [email protected]).
Received August 19, 2004.
Accepted January 14, 2005.
2
J. Anim. Sci. 2005. 83:1068–1074
al., 2000). In addition, several crops, including corn,
soybeans, canola, cotton, wheat, and beets, have been
developed that include a gene that confers tolerance to
glyphosate, the active ingredient in the commonly used
herbicide, Roundup (Padgette et al., 1995). Recently,
herbicide-tolerant rice (LibertyLink) that carries the
bar gene derived from the soil microorganism, Streptomyces hygroscopicus, was developed. The bar gene encodes the enzyme phosphinothricin-N-acetyl transferase, which catalyzes the conversion of L-phosphinothricin, the active ingredient in the herbicide,
glufosinate ammonium (Liberty), to an inactive form
(Thompson et al., 1987).
Research involving numerous GM plants has shown
that they are substantially equivalent in chemical com-
1068
Downloaded from jas.fass.org by on January 29, 2009.
Transgenic vs. conventional rice
position to their near-isogenic, conventional counterparts (Clark and Ipharraguerre, 2001; Flachowsky and
Aulrich, 2001; Faust, 2002). In addition, more than 100
studies with food-producing animals have documented
the nutritional equivalency of GM, insect-protected
and/or herbicide-tolerant corn, soybeans, canola,
wheat, cottonseed, and sugar or forage beets and their
normal counterparts (Cromwell and Hartnell, 2004).
Except for a study conducted in the Netherlands with
broiler chicks (Schat et al., 1999), no other large animal
studies have been conducted with herbicide-tolerant
rice. Thus, the objectives of this study were to assess
the nutritional equivalency of glufosinate herbicide-tolerant and conventional rice in rice-soybean meal diets
for growing-finishing swine.
Experimental Procedures
The GM and conventional rice grains used in this
study were produced and the compositional and nutritional comparisons were conducted following the general guidelines of ILSI (2003).
Rice
A GM, herbicide-tolerant rice grain, a near-isogenic
conventional rice grain, and a commercial rice grain
were evaluated in this study. The GM rice (LibertyLink,
event LLRICE62; Bayer CropScience, Research Triangle Park, NC) contained a genetic sequence that makes
it tolerant to the glufosinate ammonium herbicide, Liberty (Bayer CropScience). The conventional, mediumgrain rice (cultivar, Bengal) was the cultivar used for
the bar gene transformation, from which LLRICE62
was derived and is considered near isogenic to the GM
rice. The GM rice was grown in Texas and the nearisogenic rice was grown in Arkansas in 2000. Half the
field in which the GM rice was grown was sprayed with
Liberty herbicide. This resulted in two types of GM rice
for the study: one from a glufosinate herbicide-treated
field and the other from a field untreated with glufosinate herbicide. The GM and near-isogenic rice were
grown using herbicide regimens typical of southern
United States rice production practices. In addition,
commercially milled, brown rice was purchased from
Riviana Foods (Houston, TX) to include in the study.
This rice was considered as representative of long-grain
rice cultivars typically grown in the southern United
States using a conventional herbicide regimen.
Diets
The experimental diets for the study were prepared
at the Texas A&M University Food Protein Research
Center. The GM and near-isogenic rice grains were
milled to remove the hull and produce brown rice
(brown rice is dehulled white rice with the bran intact).
The brown rice purchased from Riviana Foods had been
commercially milled to match the milling age of the
1069
grain processed by Texas A&M for the study. All of the
rice grains were ground to a medium consistency with
a hammer mill.
Rice-soybean meal diets fortified with minerals and
vitamins to meet or exceed NRC (1998) standards were
prepared in meal form (Table 1). Diets were prepared
for three feeding phases (growing, early finishing, and
late finishing), with the same percentage of rice and
the same percentage of dehulled soybean meal used
in the four diets for the three phases. A common source
of soybean meal was used for all diets and phases. Diets
were formulated to contain 0.99, 0.80, and 0.65% total
lysine (as-fed basis) during the growing, early-finishing,
and late-finishing phases of the experiment, with diet
changes made at an average of 51 and 77 kg BW, respectively. Changes in lysine concentration for the three
phases were made by adjusting the proportions of rice
and soybean meal in the diets. Dietary lysine levels
were the same for both barrows and gilts, and were
sufficient to meet the NRC (1998) estimated requirements for gilts with a medium-high rate of fat-free carcass lean gain (i.e., 325 g/d). Soybean oil was added to
certain diets to equalize dietary fat across all diets.
Proportions of dicalcium phosphate and ground limestone were adjusted so that dietary concentrations of
Ca and P were the same for each diet. An antioxidant
(ethoxyquin) was included, but antimicrobial agents
were not included in the diets. After mixing, diets were
stored at 18 ± 2°C until they were shipped to the animal
research facility, where they were kept in an enclosed
building at 22 ± 3°C with an average relative humidity
of 71%.
Animals
The experiment was conducted from July to October
2002, at a commercial animal research facility (Genesis
Midwest Laboratories, Neillsville, WI). Crossbred pigs
(n = 96) initially averaging 25 kg BW were used in the
study. They were grouped into six blocks by gender
(barrows and gilts), and pigs within each gender were
allotted randomly to four dietary treatments from outcome groups of initial weight. Each pen consisted of
four pigs, and there were three pens of barrows and
three pens of gilts in the study, for a total of six replicate
blocks per dietary treatment.
Pigs were housed in an environmentally controlled
building in pens that provided approximately 0.85 m2
per pig. The pens had solid concrete floors and were
bedded with wood shavings as needed. Fluorescent
lights were used in the buildings to provide 16 h of light
and 8 h of darkness during each 24-h period. Pigs were
allowed to consume their diets and water ad libitum
from metal, one-hole self-feeders and automatic water
nipples. The pigs were individually weighed, and feed
consumption was determined on a pen basis at weekly
intervals during the experiment. Pigs were switched
from the growing diet to the early-finishing diet, and
then again to the late-finishing diet at the same time.
Downloaded from jas.fass.org by on January 29, 2009.
1070
Cromwell et al.
Table 1. Composition of diets (%, as-fed basis)a
Growing phase
Item
Rice type:b
Treated with herbicide:b
GM rice, treated
GM rice, untreated
Near-isogenic, conventional rice
Commercial rice
Soybean meal, dehulledb
Soybean oil
L-LysineⴢHCl
Dicalcium phosphate
Ground calcitic limestone
Iodized salt
Vitamin-trace mineral premixc
Antioxidantd
Calculated analysese
P, %
Crude fat, %
Lysine, %
Ca, %
P, %
Bioavailable P, %
GE, Mcal/kg
GM
yes
GM
no
NI
no
Early-finishing phase
Comm.
no
72.80
GM
yes
GM
no
NI
no
Late-finishing phase
Comm.
no
80.00
72.80
GM
yes
72.80
24.18
0.52
1.00
0.88
0.40
0.20
0.02
23.92
0.63
0.04
1.21
0.78
0.40
0.20
0.02
18.51
2.61
0.99
0.66
0.62
0.28
3.82
18.95
2.61
0.99
0.66
0.62
0.25
3.82
17.93
2.61
0.99
0.66
0.62
0.29
3.78
Comm.
no
85.80
80.00
72.80
24.87
NI
no
85.80
80.00
24.13
0.63
0.02
1.16
0.64
0.40
0.20
0.02
GM
no
17.06
0.58
0.98
0.73
0.40
0.20
0.02
17.01
0.70
0.02
0.96
0.69
0.40
0.20
0.02
0.78
0.96
0.40
0.20
0.02
16.78
0.70
0.04
1.02
0.84
0.40
0.20
0.02
18.54
2.61
0.99
0.66
0.62
0.25
3.88
15.82
2.83
0.80
0.62
0.56
0.23
3.81
16.30
2.83
0.80
0.62
0.56
0.20
3.80
15.19
2.83
0.80
0.62
0.56
0.24
3.77
85.80
80.00
17.83
11.38
0.62
0.76
0.79
0.40
0.20
0.02
11.32
0.76
0.02
0.85
0.63
0.40
0.20
0.02
0.66
0.92
0.40
0.20
0.02
11.09
0.75
0.04
0.90
0.80
0.40
0.20
0.02
15.86
2.83
0.80
0.62
0.56
0.20
3.88
13.68
3.00
0.65
0.56
0.52
0.21
3.80
14.20
3.00
0.65
0.56
0.52
0.17
3.80
13.01
3.00
0.65
0.56
0.52
0.21
3.76
85.80
12.20
0.64
0.74
0.40
0.20
0.02
13.72
3.00
0.65
0.56
0.52
0.17
3.88
a
The diets were fed during three phases from approximately 25 to 51 kg (growing), 51 to 77 kg (early finishing), and 77 to 106 kg BW (late
finishing), respectively. Within each phase, diets were formulated to have the same levels of lysine, Ca, P, and fat calories, based on actual
analyses of the rice grains and dehulled soybean meal.
b
The four rice grains include: GM = genetically modified herbicide-tolerant brown rice (LibertyLink, event LLRICE62) grown in fields
treated or not treated with glufosinate herbicide; NI = conventional near-isogenic brown rice (Bengal); Comm. = commercially milled longgrain brown rice. All rice grain was produced using conventional herbicide regimens typical of the Southern U.S. rice production practices.
The GM rice is derived from the medium-grain rice cultivar, Bengal.
c
Provided per kilogram of diet: vitamin A, 5,500 IU; vitamin D3, 1,100 IU; vitamin E, 24.2 IU; vitamin K (as menadione sodium bisulfite
complex), 0.7 mg; riboflavin, 4.4 mg; pantothenic acid, 17.6 mg; niacin, 26.4 mg; vitamin B12, 0.026 mg; choline, 66 mg; Zn, 90 mg; Fe (as
FeSO4ⴢH2O), 80 mg; Mn (as MnO), 32 mg; Cu (as CuSO4ⴢ5H2O), 10 mg; I (as CaI2O6), 0.4 mg; and Se (as NaSeO3), 0.3 mg.
d
Provided 150 mg of ethoxyquin per kilogram of diet.
e
Based on actual analyses of rice and dehulled soybean meal and calculated analysis of other ingredients.
The experiment was terminated after 98 d, at which
time the pigs averaged 106 kg BW.
Carcass Evaluation
All pigs were transported from the research facility
to a local packing plant (Pinter’s Packing Plant, Dorchester, WI) within 2 h after the end of the experiment.
The pigs were humanely slaughtered (electrically
stunned followed by exsanguination), skinned, and
eviscerated. Hot carcass weight (head on) was then determined. The weight of the skinless, head-on carcass
was assumed to be similar to that of a skin-on, headless
carcass (Buege and Russell, 2000). Following a 24-h
chill at 2°C, the carcasses were split between the 10th
and 11th rib, and backfat depth (three-fourths of the
distance from the midline to the end of the LM), and
LM area were measured. From the carcass data, the
percentage of fat-free lean in the carcass was determined using the NPPC (2000) equation (adapted to metric units) as follows:
Carcass fat-free lean, % = 100 × [3.899 + (0.465 × hot
carcass weight, kg) − (3.914 × 10th-rib backfat, cm)
+ (0.21146 × LM area, cm2)]/HCW, kg
Chemical Analysis of Rice, Soybean Meal, and Diets
Representative samples of the rice, soybean meal,
and mixed diets were analyzed for DM after oven drying
to a constant weight, for CP by a N analyzer (N × 6.25),
for crude fat based on ether extraction, and for ash
and crude fiber; all methods were based on standard
procedures (AOAC, 1995). Nitrogen-free extract was
determined by difference. Acid detergent fiber, NDF,
and acid detergent lignin were analyzed by methods
described by Robertson and Van Soest (1981) and modified by Jeraci et al. (1988). Calcium was analyzed with
atomic absorption spectrophotometry after wet-ashing,
and P was determined by a colorimetric procedure
(AOAC, 1995). Amino acids were analyzed with ionexchange chromatography after acid hydrolysis. Methionine and cystine were oxidized to methionine sulfone
and cysteic acid by treatment with performic acid before
hydrolysis. Tryptophan was analyzed after alkaline hydrolysis. Rice and diets also were analyzed for GE, trace
minerals (including heavy metals), tocopherols, and
pesticides (AOAC, 1995), and for aerobic plate counts,
Clostridium perfringens, and Salmonella spp. (Vanderzant and Spllittstoesser, 1992). Assays were conducted
Downloaded from jas.fass.org by on January 29, 2009.
1071
Transgenic vs. conventional rice
at the Woodson-Tenent Laboratories, Des Moines, IA,
or at Bayer CropScience, Research Triangle Park, NC.
Table 2. Composition of ground rice (air-dry basis)
Rice
Statistical Analyses
The data were analyzed as a randomized complete
block design (Steel and Torrie, 1980) using the GLM
procedure of SAS (SAS Inst., Inc., Cary, NC). The statistical model included the effects of block, dietary treatment, and block × treatment (error, 15 df). Gender was
nested within blocks and was tested with replication
within gender (4 df). Nested within block × treatment
was the gender × treatment interaction (3 df), which
was tested with replication within gender × treatment
(12 df). In addition, final BW was included in the model
as a covariate in the analysis of the carcass data. Treatment mean separation was with single-df comparisons
of the herbicide-treated GM rice treatment vs. each of
the other three dietary treatments. In all instances,
pen was considered the experimental unit. An alpha
level of P < 0.05 was considered statistically significant.
Results and Discussion
Composition of Rice Grains and Diets
The proximate components (DM, CP, crude fat, crude
fiber, nitrogen-free extract, and ash) were similar in
concentration among the four rice grains (Table 2). The
various fiber components (NDF, ADF, lignin) also were
similar in the rice grains, as were the Ca and P concentrations. The lysine concentrations in the herbicidetreated and untreated GM rice averaged 0.35% and
were similar to those in the near-isogenic and commercial rice grains (0.33%). All other AA were essentially
equivalent for the three rice grains, and reasonably
close to calculated values for a blend of 85% polished
and broken rice (i.e., white rice) and 15% rice hulls
(NRC, 1998). This blend of white rice and hulls would
be approximately equivalent to brown rice.
The nutrient concentrations (i.e., trace minerals and
tocopherols) in all the diets were very close to targeted
levels (data not shown). Bacterial analyses of the rice
and diets revealed that aerobic plate counts, Clostridium, and Salmonella, were within acceptable limits, as
were pesticides and heavy metal concentrations (data
not shown).
Performance Data
Average daily gain, daily feed intake (as-fed basis),
and efficiency of feed utilization (expressed as feed:gain
ratio) for pigs fed the four rice grain diets are shown
in Table 3. Weight gain and feed intake did not differ
significantly for the four treatment groups during the
initial growing phase, but the feed:gain of pigs fed the
herbicide-treated GM rice was less (P < 0.05) than for
pigs fed the untreated GM rice. During the early- and
late-finishing phases, pigs fed the herbicide-treated GM
Item
Herbicide treated:d
DM, %
CP, %
Crude fat, %
Crude fiber, %
NDF, %
ADF, %
Lignin, %
Nitrogen-free extract, %
Ash, %
Ca, %
P, %
Amino acids, %
Arginine
Histidine
Isoleucine
Leucine
Lysine
Methionine
Cysteine
Phenylalanine
Tyrosine
Threonine
Tryptophan
Valine
GE, Mcal/kg
GMa
Yes
GMa
No
Near
isogenicb
No
Commercialc
No
86.4
9.68
2.55
0.9
3.5
1.4
1.5
72.8
1.40
0.01
0.33
85.5
10.25
2.71
1.4
3.9
1.6
1.1
71.0
1.57
0.02
0.37
86.1
9.03
2.56
1.4
3.6
1.9
1.4
73.2
1.33
0.01
0.32
86.6
9.24
3.42
1.9
3.6
2.0
1.8
75.4
1.47
0.01
0.37
0.70
0.23
0.33
0.67
0.34
0.22
0.19
0.43
0.28
0.31
0.11
0.49
3.81
0.70
0.24
0.34
0.71
0.36
0.30
0.22
0.45
0.29
0.34
0.14
0.51
3.82
0.63
0.21
0.30
0.61
0.33
0.22
0.18
0.39
0.25
0.29
0.12
0.44
3.78
0.65
0.21
0.32
0.66
0.33
0.23
0.19
0.41
0.28
0.31
0.11
0.48
3.94
a
GM = genetically modified medium-grain brown rice (LibertyLink,
event LLRICE62) that was tolerant to glufosinate herbicide; Bayer
CropScience, Research Triangle, NC.
b
Near-isogenic conventional medium-grain brown rice (Bengal).
c
Commercially milled long-grain brown rice.
d
The GM rice was from fields treated or not treated with glufosinate
herbicide.
rice gained faster than those fed the untreated GM rice
(P < 0.05 during the early-finishing phase) or the nearisogenic conventional rice (P < 0.05 during both phases),
and their feed:gain ratios were improved during both
finishing phases (P < 0.05 during the late-finishing
phase). However, gain, feed intake, and feed:gain of pigs
fed the herbicide-treated GM rice were not significantly
different from those of pigs fed the commercial rice
diet during the two finishing phases. Over the entire
experimental period, pigs fed the herbicide-treated GM
rice performed better (P < 0.05) than those fed the untreated GM rice, but the growth performance traits did
not differ significantly from those of pigs fed the two
conventional rice diets.
The reason for the statistical difference in the herbicide-treated GM rice and the untreated GM rice is not
clear, but the differences could be simply due to chance
alone because of the high statistical power of the study.
But even if the differences were real, the effects are
of no practical significance because herbicide-tolerant
crops would likely be treated with herbicides.
Carcass Data
Hot carcass weight was affected (P < 0.05) by dietary
treatment and generally reflected the differences in fi-
Downloaded from jas.fass.org by on January 29, 2009.
1072
Cromwell et al.
Table 3. Growth rate, feed intake (as-fed basis) and feed:gain of pigs fed rice-dehulled
soybean meal diets containing conventional or herbicide-tolerant ricea
Rice
Item
GMb
Yes
Herbicide treated:
Avg initial BW, kg
24.7
Avg final BW, kge
109.1
Growing phase (25 to 51 kg)
ADG, kg
0.76
ADFI, kg
1.79
Feed:gaine
2.37
Early finishing phase (51 to 77 kg)
0.96
ADG, kgef
ADFI, kg
2.57
Feed:gain
2.67
Late finishing phase (77 to 106 kg)
ADG, kgf
0.88
ADFI, kg
2.88
Feed:gaine
3.30
Entire test (25 to 106 kg)
ADG, kge
0.86
2.41
ADFI, kge
Feed:gaine
2.80
GMb
No
Near isogenicc
No
Commerciald
No
CV
25.0
102.2
24.7
103.7
25.1
108.1
3.61
4.21
0.70
1.94
2.83
0.76
1.85
2.44
0.74
1.92
2.59
8.58
6.62
11.99
0.89
2.59
2.95
0.88
2.52
2.86
0.94
2.60
2.76
5.58
6.06
9.03
0.80
2.95
3.71
0.79
2.78
3.52
0.87
2.85
3.27
8.26
4.38
6.21
0.79
2.49
3.17
0.81
2.37
2.95
0.85
2.45
2.89
5.59
2.61
6.45
a
Based on six replications of four pigs per pen (three replications of barrows and three replications of
gilts).
b
GM = genetically modified herbicide-tolerant rice from fields that were treated or not treated with
glufosinate herbicide.
c
Near-isogenic, conventional medium-grain brown rice (Bengal).
d
Commercially milled long-grain brown rice.
e
Herbicide-treated GM rice vs. untreated GM rice, P < 0.05.
f
Herbicide-treated GM rice vs. near-isogenic conventional rice, P < 0.05.
nal BW among the four treatment groups; however,
covariance analysis eliminated the treatment differences (Table 4). Hot carcass yield, backfat, LM area,
and calculated carcass fat-free lean did not differ significantly among pigs fed the four rice grain diets.
Gender Effects
There was no evidence of a gender × diet interaction
(P = 0.50) for any of the growth performance or carcass
traits (Table 5). As expected, barrows gained weight
Table 4. Carcass characteristics of pigs fed rice-dehulled soybean meal diets containing
conventional or herbicide-tolerant ricea,b
Rice
Item
Herbicide treated:
Hot carcass weight, kg
Actualfg
Adjusted
Hot carcass yield, %
10th-rib fat depth, mm
10th-rib LM area, cm2
Carcass fat-free lean, %
GMc
Yes
GMc
No
Near isogenicd
No
Commerciale
No
CV
80.0
77.8
73.5
23.0
38.6
50.5
74.4
76.8
72.6
22.7
38.0
50.5
75.3
76.7
72.6
21.3
38.2
51.2
79.0
77.5
73.2
23.8
38.1
50.0
4.23
1.62
1.57
10.13
5.53
2.84
a
Based on six replications of four pigs per pen (three replications of barrows and three replications of
gilts).
b
Carcass data adjusted by covariance for final BW.
c
GM = genetically modified herbicide-tolerant brown rice from fields that were treated or not treated with
glufosinate herbicide.
d
Near-isogenic conventional medium-grain brown rice (Bengal).
e
Commercially milled long-grain brown rice.
f
Herbicide-treated GM rice vs. untreated GM rice, P < 0.05.
g
Herbicide-treated GM rice vs. near-isogenic conventional rice, P < 0.05.
Downloaded from jas.fass.org by on January 29, 2009.
Transgenic vs. conventional rice
Table 5. Performance and carcass characteristics of barrows and gilts fed rice-dehulled soybean meal diets containing herbicide-tolerant rice or conventional ricea
Gender
Item
Initial BW, kg
Final BW, kg
Entire test
ADG, kgc
ADFI (as-fed basis), kgc
Feed:gain
Hot carcass weight, kg
Adjusted carcass datab
Hot weight, kg
Carcass yield, %
10th-rib fat depth, mm
10th-rib LM area, cm2
Carcass fat-free lean, %c
Barrows
Gilts
CV
23.9
109.5
25.8
102.1
14.13
7.21
0.87
2.54
2.94
79.3
0.78
2.31
2.97
75.1
4.16
4.94
5.42
7.89
76.8
72.5
24.3
36.6
49.4
77.6
73.4
21.2
39.8
51.7
7.89
1.26
10.05
5.25
1.12
a
Each mean represents 12 pens of four pigs (three replications of
four pigs per pen per dietary treatment.
b
Adjusted by covariance for final BW.
c
Effect of gender, P < 0.05.
faster (P < 0.05) and consumed more feed (P < 0.05)
than gilts; however, feed:gain responses were similar
for the two genders. Gilts tended to have less backfat
and larger longissimus areas than barrows, but the
differences were not significant. The calculated fat-free
carcass lean was greater in gilts than in barrows (P <
0.05). The calculated fat-free lean gains (NPPC, 2000)
of the barrows (317 g/d unadjusted; 305 g/d adjusted)
and gilts (309 g/d unadjusted; 320 g/d adjusted) indicate
that the pigs were of medium to medium-high lean
growth capacity (NRC, 1998).
General Discussion
Transgenic crops developed through biotechnology
represent a new tool in the production of food, feeds,
and fiber that can make a vital contribution to an everincreasing global need. Domestic and worldwide adoption of GM crops has increased dramatically since becoming commercially available. Global adoption of GM
crops from 1996 to 2003 increased 40-fold, from 1.7
million to 67.7 million ha, and land area for GM crop
production has increased by approximately 15% per
year during this 8-yr period (James, 2003).
Of the four major GM crops (soybeans, corn, canola,
cotton), herbicide tolerance is the most prominent GM
trait, representing 73% of the total (James, 2003). Most
GM crops with herbicide tolerance contain a gene that
makes the plant tolerant to glyphosate, the active ingredient in a commonly used herbicide, Roundup (Padgette
et al., 1995). This gene encodes a glyphosate-tolerant
5-enolphyuvyl-shikimate-3-phosphate synthase from
Agrobacterium sp. Strain CP4 (CP4 EPSPS). Our study
involved a rice cultivar that carried the bar gene derived
from the soil microorganism, Streptomyces hygroscopicus. The bar gene encodes the enzyme phosphi-
1073
nothricin-N-acetyl transferase, which catalyzes the
conversion of L-phosphinothricin, the active ingredient
in the herbicide, glufosinate ammonium (Liberty), to
an inactive form (Thompson et al., 1987).
Other than a company report to the USDA (USDA,
1999), our study is the first peer-reviewed report documenting that the composition of glufosinate-tolerant
rice grain is substantially equivalent to that of nearisogenic conventional rice grain (Bengal, mediumgrain). Furthermore, the GM rice was similar in composition to an unrelated commercial rice grain. Similar
results of no substantial compositional effects have
been reported for other GM and unmodified grains
(Clark and Ipharraguerre, 2001; Aumaitre et al., 2002;
Faust, 2002).
Our study also is the first to demonstrate that the
nutritional characteristics of herbicide-treated, glufosinate-tolerant rice are essentially equivalent to a nearisogenic conventional rice grain or to a commonly produced commercial rice grain when fed to growing-finishing pigs. A previous study with broiler chicks conducted in the Netherlands found results similar to ours
(Schat et al., 1999). In that study, broilers fed 30%
rice diets containing glufosinate-tolerant rice or nearisogenic conventional rice for 42 d had essentially equivalent growth rate (58.8 vs. 59.1 g/d), feed intake (101.4
vs. 100.4 g/d), gain efficiency (0.580 vs. 0.588 G:F), and
percentage of carcass (70.8 vs. 70.7%), breast (19.8 vs.
19.8%), and abdominal fat (2.0 vs. 1.8%, respectively).
Our results also agree with other reports in the scientific literature in which GM and conventional crops
were assessed. Cromwell and Hartnell (2004) summarized more than 100 animal studies that were conducted primarily in the United States and Europe. The
studies involved insect-protected and/or herbicide-tolerant soybeans, corn, canola, wheat, cottonseed, and
beets. In all cases, the composition of the grain and
forage from the GM grains or forages was substantially
equivalent to those produced from conventional crops.
Similarly, the nutritional value of the GM and conventional feeds has been found to be essentially the same
when tested in a number of animal species, including
poultry, swine, dairy and beef cattle, sheep, water buffalo, rabbits, and catfish. Other researchers have conducted studies or produced similar reviews and drew
similar conclusions (Clark and Ipharraguerre, 2001;
Aumaitre et al., 2002; Cromwell et al., 2002).
We did not attempt to verify the presence or absence
of transgenic DNA or protein in tissues of the pigs fed
the GM rice; however, assessment of muscle, milk, eggs,
and other products from animals fed GM crops have
shown that foreign DNA or the specific modified protein
responsible for the effect in the plant are not detected
in tissues (Khumnirdpetch et al., 2001; Ash et al., 2003;
Jennings et al., 2003a,b). Other current publications
on the feeding of GM crops to livestock and poultry
and the assessment of detecting transgenic DNA and
protein in meat, milk, or eggs of animals fed these crops
are available (FASS, 2004).
Downloaded from jas.fass.org by on January 29, 2009.
1074
Cromwell et al.
Implications
The results of this study indicate that genetically
modified, glufosinate herbicide-tolerant (LibertyLink)
rice grain was essentially equivalent in composition
and nutritional value to a near-isogenic rice grain and
to a common commercial rice grain for growing-finishing swine. This study demonstrates that genetic modification of rice plants to make them tolerant to glufosinate herbicide did not affect the nutritional composition
of the grain or the growth performance of swine fed the
genetically modified rice grain in rice-soybean meal
diets.
Literature Cited
AOAC. 1995. Official Methods of Analysis. 16th ed. Assoc. Off. Anal.
Chem., Arlington, VA.
Ash, J., C. Novak, and S. E. Scheideler. 2003. The fate of genetically
modified protein from Roundup Ready soybeans in laying hens.
J. Appl. Poult. Res. 12:242:245.
Aumaitre, A., K. Aulrich, A. Chesson, G. Flachowsky, and G. Piva.
2002. New feeds from genetically modified plants: Substantial
equivalence, nutritional equivalence, digestibility, and safety for
animals and the food chain. Livest. Prod. Sci. 74:223–238.
Betz, F. S., B. G. Hammond, and R. L. Fuchs. 2000. Safety and
advantages of Bacillus thuringiensis-protected plants to control
insect pests. Regul. Toxicol. Pharmacol. 32:156–173.
Buege, D. R., and R. Russell. 2000. Pork Carcass Evaluation. Dept.
of Muscle Biology, Univ. of Wisconsin, Madison.
Clark, J. H., and I. R. Ipharraguerre. 2001. Livestock performance:
Feeding biotech crops. J. Dairy Sci. 84(Suppl. E):E9-E18.
Cromwell, G. L., and G. F. Hartnell. 2004. Animal nutrition studies on
biotechnology-derived foods/crops. Proc. 11th Anim. Sci. Cong.
Asian-Aust. Assoc. Anim. Prod. Soc., Kuala Lumpur, Malaysia.
Cromwell, G. L., M. D. Lindemann, J. H. Randolph, G. R. Parker, R.
D. Coffey, K. M Laurent, C. L. Armstong, W. B. Mikel, E. P.
Stanisiewski, and G. F. Hartnell. 2002. Soybean meal from
Roundup Ready or conventional soybeans in diets for growingfinishing swine. J. Anim. Sci 80:708–715.
FASS. 2004. Animal Biotechnology. Fed. Anim. Sci. Soc., Savoy, IL.
Available: http://www.animalbiotechnology.org. Accessed: Nov.
26, 2004.
Faust, M. A. 2002. New feeds from genetically modified plants: The
U.S. approach to safety for animals and the food chain. Livest.
Prod. Sci. 74:239–254.
Flachowsky, G., and K. Aulrich. 2001. Nutritional assessment of
feeds from genetically modified organisms. J. Anim. Feed Sci.
10:181–194.
ILSI. 2003. Best Practices for the Conduct of Animal Studies to Evaluate Crops Genetically Modified for Input Traits. Int. Life Sci.
Inst., Washington, DC.
James, C. 2003. Preview: Global Status of Commercialized
Transgenic Crops: 2003. ISAAA Briefs No. 30. ISAAA, Ithaca,
NY.
Jennings, J. C., L. D. Albee, D. C. Kolwyck, J. B. Surber, M. L. Taylor,
G. F. Hartnell, R. P. Lirette, and K. C. Glenn. 2003a. Attempts
to detect transgenic and endogenous plant DNA and transgenic
protein in muscle from broilers fed YieldGard Corn Borer corn.
Poult. Sci. 82:371–380.
Jennings, J. C., D. C. Kolwyck, S. B. Kays, A. J. Whetsell, J. B.
Surber, G. L. Cromwell, R. P. Lirette, and K. C. Glenn. 2003b.
Determining whether transgenic and endogenous plant DNA
and transgenic protein are detectable in muscle from swine fed
Roundup Ready soybean meal. J. Anim. Sci. 81:1447–1455.
Jeraci, J. L., T. Hernandez, J. B. Robertson, and P. J. Van Soest.
1988. New and improved procedure for neutral detergent fiber.
J. Anim. Sci. 66(Suppl. 1):351. (Abstr.)
Khumnirdpetch, V., U. Intarachote, S. Treemanee, S. Tragoonroong,
and S. Thummabood. 2001. Detection of GMOs in the broilers
that utilized genetically modified soybean meals as a feed ingredient. [Poster 585, Abstr.] Available: http://www.intl-pag.org/
pag/9/abstracts/P08_38.html. Accessed Feb. 11, 2005.
NPPC. 2000. Pork Composition and Quality Assessment Procedures.
Natl. Pork Prod. Counc., Des Moines, IA.
NRC. 1998. Nutrient Requirements of Swine. 10th ed. Natl. Acad.
Press, Washington, DC.
Padgette, S. R., K. H. Kolacz, X. Delannay, D. B. Re, J. B. LaVallee,
C. N. Tinius, W. K. Rhodes, I. Y. Otero, G. F. Barry, D. A.
Eichholtz, V. M. Peschke, D. L. Nida, N. B. Taylor, and G. M.
Kishore. 1995. Development, identification and characterization
of a glyphosate-tolerant soybean line. Crop Sci. 35:1451–1461.
Robertson, J. B., and P. J. Van Soest. 1981. The detergent system of
analysis and its application to human foods. The Analysis of
Dietary Fiber. W. P. T. James and O. Theander, ed. Marcell
Dekker, New York.
Schat, B., G. M. Beelen, and B. P. M. Janszen. 1999. Transgenic rice
(LL Rice 62): Feeding values of transgenic rice (Event 62) in
poultry-broilers. TNO report V99.034. TNO Nutrition and Food
Res. Inst., Zeist, The Netherlands.
Steel, R. G. D., and J. H. Torrie. 1980. Principles and Procedures of
Statistics: A Biometrical Approach. 2nd ed. McGraw-Hill Publishing Co., New York.
Thompson, C. J., N. R. Movva, R. Tizard, R. Crameri, J. E. Davies,
M. Lauwereys, and J. Botterman. 1987. Characterization of the
herbicide-resistance gene bar from Streptomyces hygroscopicus.
EMBO J. 6:2519–2523.
USDA. 1999. Determination of non-regulated status for rice genetically engineered for glufosinate herbicide tolerance. Fed. Regis.
64:22595–22596.
Vanderzant, C., and D. F. Spllittstoesser. 1992. Compendium of Methods for the Microbiological Examination of Foods. 3rd ed. Am.
Public Health Assoc., Washington, DC.
Downloaded from jas.fass.org by on January 29, 2009.
References
This article cites 13 articles, 5 of which you can access for free at:
http://jas.fass.org/cgi/content/full/83/5/1068#BIBL
Downloaded from jas.fass.org by on January 29, 2009.