Journal of Agricultural Science, Cambridge (1992), 119, 243-254.

Primed in Great Britain 243

Intake and digestion in swamp buffaloes and cattle.

2. The comparative response to urea supplements in
animals fed tropical grasses
P. M. K E N N E D Y 1 * , A. N . B O N I F A C E 2 , Z . J . L I A N G ' f , D. M U L L E R 1 AND R. M. M U R R A Y 2
1
Division of Tropical Animal Production, CSIRO, Aitkenvale, Qld. 4814, Australia
2
Department of Tropical Veterinary Science and Agriculture, James Cook University of North Queensland,
Townsville, Qld. 4811, Australia
(Revised MS received 18 November 1991)

SUMMARY
Swamp buffaloes (Bubalus bubalis) and crossbred cattle (B. indicus x B. taurus) were offered two
forages in two experiments. In Expt 1, four animals of each species were offered mature Rhodes
grass hay (Chloris gayana) with a mineral supplement or with a supplement of minerals and urea
(17-6 gN/d). In Expt 2, another group of four buffaloes and four cattle were offered mature spear
grass hay (Heteropogon contortus) with mineral supplementation at intervals of 3 h. Four levels of
urea (0, 5, 21 and 97 g/day) and 35S-sulphate were continuously infused in four periods.
Rhodes grass was consumed in greater amounts by cattle, whereas buffaloes ate more spear grass.
Urea supplements increased intake of Rhodes grass by 12% in buffaloes and 22% in cattle, and of
spear grass by 34% in buffaloes and 41 % in cattle. Digestibility of cell wall constituents and acid-
detergent fibre of spear grass was lower (P < 005) in buffaloes than in cattle (417 v. 499; 471 v.
560 g/kg respectively). In Expt 2, dry matter digestibility progressively increased (P < 005) from 364
to 408 g/kg with increased urea infusion. Rumen dry matter pool increased by 11-21 % on infusion
of 97 g urea/day, compared with no urea. Digestion of ground forages incubated in situ in the rumen
was depressed below rumen ammonia levels of 30-60 mg N/l, while digestion of cotton thread was
depressed below 60-80 mg N/l (Expt 1) or 150-200 mg N/l (Expt 2).
Patterns of N content in cotton thread suggested that more microbes attached in buffaloes and
there was subsequently faster detachment than in cattle, particularly with increased urea infusion.
After ruminal incubation of five ground forages, including spear grass and rice straw, in situ for 72
h, buffaloes digested 2-0-3-5 % units more (P < 0-05) of these substrates than cattle. The concentration
of microbial N per weight of rumen dry matter was less (P < 005) in buffaloes than for cattle, and
increased (P < 001) with greater urea infusion in both species. Similarly, across treatments, the N
content of residues of three out of five forages incubated in situ for 24 h was less (P < 005) in
buffaloes. When animals were infused with 97 g urea/day, and rates of digesta passage from the
rumen were comparable between species, the N content of residues of the five forages incubated for
24 h in buffaloes relative to N content of residues from cattle was positively related to the initial fibre
content of the substrate, suggestive of attachment of more fibrolytic microbes in buffaloes.
It was concluded from rumen microbe content of rumen digesta that urea supplements increased
the digestion rate of ingested spear grass when rumen ammonia levels were increased from values less
than 100 mg N/l, and that there was evidence for a greater fibrolytic activity of adherent microbes
in buffaloes than in cattle. In addition, a component of this population from buffaloes appeared to
be capable of digestion of significantly more highly fibrous forage.

INTRODUCTION
Des ite
• Present address: CSIRO Longpocket Laboratories, P conflicting reports (e.g. Chalmers 1974; Van
Private Bag 3, Indooroopilly, Qld. 4068, Australia. Soest 1982; Devendra 1983; Wanapat 1989) on the
t Present address: Guangxi Animal Husbandry Research relative voluntary intake and digestive abilities of
Institute, Nanning, Guangxi, PR China. cattle and swamp buffaloes (Bubalus bubalis), there
244 P. M. KENNEDY ET AL.

are a number of studies indicating that the N economy DM) 749 CWC, 403 ADF and 56 N. The diet was
of buffaloes is superior to that of cattle (reviewed by offered twice daily at 120% of the previous day's
Devendra 1983, 1987). intake and continuous infusions of urea and sulphate
Voluntary intake and digestion in ruminants is (N:S, 10:1) were made into the ventral rumen at
commonly reduced on fibrous diets of N content nominal rates of 0, 5, 20 and 100 g/day of urea during
< 1% (Minson 1982), attributable to inhibition of the final 16 weeks. Actual infusion rates were 0, 4-6,
rumen microbial digestion by ammonia deficiency. 20-5, and 96-9 g urea/day. The administration of urea
Accordingly, differences in the digestive ability and of was by infusion to allow more efficient utilization and
voluntary intake of buffaloes and cattle fed fibrous precise calculation of urea input (Romero 1976). Each
diets may originate from differences in N transactions rate of urea was infused into each animal during 4
in the rumen, and may be most marked for diets weeks, employing two 4 x 4 Latin squares of different
deficient in N. Such differences were noted by Hunter randomization, one for each species. Animals were
& Siebert (1985o, b) between Hereford (Bos taurus) housed in individual pens with water always available.
and Brahman (Bos indicus) cattle, with the latter
maintaining rumen ammonia at levels sufficiently
Measurements
high to prevent the decline in digestion rate evident
for Hereford steers offered tropical roughages. In Expt 1, voluntary feed consumption (VFC) was
Kennedy et al. (1992) reported that buffaloes fed measured over 10 days during the last 14 days of each
rice straw maintained higher rumen ammonia levels period. In addition, six groups of three nylon bags
through enhanced recycling of blood urea to the (95x170 mm after closure, mesh aperture 45 \im)
rumen. In addition in that study, net synthesis of were inserted just before the daily feeding into the
microbial N was more efficient in the rumen of ventral rumen and incubated for 5, 11, 18, 24, 47 and
buffaloes than of cattle. The present study investigated 72 h, before thorough washing with squeezing until
the comparative digestive responses of buffaloes and the water was colourless. Blanks were prepared by
cattle fed mature tropical grasses to supplementation soaking extra bags in warm water for 30 min before
with urea and minerals, and aimed to draw further washing. Each set of three bags comprised a single
conclusions about the relative activity of the fibrolytic bag containing cotton thread (1 g, 12 gauge, un-
microbes in the two species. Brief preliminary reports bleached; 'Propeller' Coats Paton) and duplicate
of parts of this work have been published (Kennedy & bags of Rhodes grass (3g) ground through a 3 mm
Waterhouse 1987; Kennedy et al. 1987). screen. Dietary marker was prepared by labelling
Rhodes grass (ground through a 1 mm screen) with
ytterbium acetate, using the soaking method of Mader
MATERIALS AND METHODS et al. (1984) and introduced (20 g) into the ventral
Animals, diets and feeding rumen just before feeding. Grab faecal samples were
taken for Yb analysis at regular intervals, between 24
In Experiment 1, four female buffaloes (Bubalus and 192 h after dosing.
bubalis) and four castrate male Brahman x Shorthorn During the final 4 days of each period and separated
(B. indicus x B. taurus) steers, aged 20-21 months and by 2 days, the rumen digesta were removed through
fitted with ruminal cannulas (o.d. 12 cm) were fed ad the cannula just before (To) and 4 h (T4) after feed was
libitum for 10 weeks in individual pens. The diet was offered. The digesta were weighed, mixed and sub-
Rhodes grass hay (Chloris gayana Kunth) containing sampled for DM determination. At the same time,
(g/kg dry matter, DM) 3-6 N, 770 cell wall con- rumen fluid was collected for determination of
stituents (CWC), and 498 acid detergent fibre (ADF) ammonia and volatile fatty acids (VFA) and jugular
which was offered daily at 1200 g/kg of the intake blood was collected for plasma urea.
recorded on the previous day. Minerals (63g/day; In Expt 2, 8 days before the end of each period, a
Siebert & Kennedy 1972) were sprinkled onto the hay. continuous infusion into the rumen, of 35S-sulphate
The diet was offered without or with a supplement of (5-6 MBq/day for animals receiving 0 and 5 g
urea (17-6 gN/d) in two periods each of 4 weeks. In urea/day; 111 MBq/day for animals receiving 21
period 1, two animals of each species received the and 97 g urea/day) was started, following a priming
unsupplemented grass, and two received urea supple- dose (4-6 MBq and 9-2 MBq, respectively). During
ments. In period 2, diets were swapped in a cross-over the following 7 days, VFC was measured, fresh faecal
design. samples (200 g) were taken twice daily from each
In Experiment 2, four female buffaloes and four animal, bulked and frozen. Rumen fluid (150 ml) was
female Brahman x Shorthorn cattle, aged c. 3-5 years, sampled daily at 0900 h (c. 1 h after feed was offered)
each fitted with a cannula in the rumen and abomasum for ammonia analysis and isolation of microbes by
(i.d. 2-5 cm) were offered spear grass (Heteropogon centrifugation. After 2, 3 and 4 days of 35S infusion,
contortus P. Beauv. ex Roem. et Schult.) once daily duplicate nylon bags containing spear grass, ryegrass
for 20 weeks. The spear grass used contained (g/kg (Lolium perenne L.), calopo (Calopogonium mucu-
 Urea in buffaloes and cattle 245

Table 1. Chemical composition (g/kg dry matter) of digestion of cotton thread were calculated from the
ground forages incubated in nylon bags in the rumen linear decline with time of residual DM in the nylon
(Expt 2) bags, and by extrapolation to time for zero digestion.
The rate of digestion, lag time before onset of
Acid digestion, and predicted potential degradability of
Organic Cell wall detergent dietary CWC incubated in situ, were derived by fitting
matter constituents fibre Nitrogen the disappearance of material with time in the rumen,
using a least squares procedure, to the exponential
Spear grass 921 749 403 5-6 equation described by McDonald (1981).
Ryegrass 880 700 340 24 Estimates of two rate constants (k\, k2), transit time
Rice straw 797 660 400 7.8 (TT) and total mean retention time (TMRT) were
Calopo 960 720 490 12
derived by simultaneous, non-linear fitting of the
Verano 937 490 290 25
model of Grovum & Williams (1973) to the con-
centration of Yb in faeces. It was assumed that the
noides Desv.), verano (Stylosanthes hamata cv. slower rate constant, klt was attributable to passage
Verano (L.) Taub.), rice straw (Oryza saliva L.) or of digesta from the rumen. In addition, the multi-
cotton thread were inserted into the ventral rumen, compartmental model of Dhanoa et al. (1985) was
and incubated for 24, 48, and 72 h. The chemical fitted to the data to derive another estimate of kv
composition of these substrates is given in Table 1. In Apparent passage rate of DM from the rumen was
addition, five 2 cm lengths of freeze-dried kikuyu calculated as faecal DM output/rumen DM pool.
grass (Pennisetum clandestinum Chiov.) leaf blades Data were analysed by analysis of variance. The
were incubated in one bag for 24 h, for estimation of model included terms for animal species, animals
the population of anaerobic rumen fungi (Akin el al. within species, period, treatment and interaction
1983) as described previously (Kennedy el al. 1992). between species and treatment, as follows;
On the fifth day, all bags were removed, a dose of' Ytm = fi + species, + (species/animals)w + treatment,.
CrEDTA (2-5 g Cr, prepared by the method of
Binnerts et al. 1968) was injected into the ventral + period, + (species x treatment),,, + error ytl
rumen, and samples of rumen fluid were taken at 3 h The animal within species interaction was used as the
intervals for 24 h for ammonia and CrEDTA analysis. error term to test for species effects (D.F. = 6). Treat-
On the last day, the rumen was evacuated by hand, ment effects and species x treatment interaction were
the contents were weighed, mixed and two subsamples estimated and tested from the animal x period stratum
were taken for determination of DM and 35S content. (D.F. = 15). When treatments differed significantly
The latter sample was divided into fine particles which (P < 005) they were tested for linear, quadratic and
were squeezed through a cloth of 45 urn mesh, and the cubic effects of quantity of infused urea.
residue (coarse particles).

RESULTS
Chemical and other analyses
Experiment I
DM was determined by loss of weight at 70 °C. CWC
and ADF were determined sequentially (Van Soest & Mean body weight of buffaloes was 281 kg, and cattle
Robertson 1980). Cr concentration in rumen fluid 261kg (difference, P < 0-10). VFC of cattle was
was measured by atomic absorption spectrometry greater than that of buffaloes (P < 005), and urea
(Model AA-975, Varian Techtron, San Fernando, supplementation increased intake of Rhodes grass by
Ca.) using a lean nitrous oxide/acetylene flame after 12% for buffaloes and 22% for cattle (Table 2).
centrifugation and dilution to provide a medium of Rumen fill expressed as digesta per rumen-free body
4 % v/v HNO 3 /1 % KNO3 (Pearton & Mallett 1972). weight (RFW) was c. 46% higher (P < 005) in cattle
Yb, N, and VFA were measured as previously at both sampling times. Supplementation by urea of
described (Kennedy et al. 1992). Plasma urea-N was animals resulted in increases in rumen fill of 9 %
estimated by an autoanalyser (method SE40001FD4, (P < 005) at To, and of 13% (P = 007) at T 4 . Not-
Technicon Instruments Corp, New York). Calcu- withstanding rumen fill differences, rumen DM pools
lation of microbial N content of rumen digesta was were similar for both animal species and dietary
made as described by Kennedy et al. (1992) after treatments, because the DM content of rumen digesta
assay for organic 35S and N in isolated microbes and was c. 24% greater in buffaloes than in cattle
in two (fine and coarse) rumen digesta fractions. (P <005). TMRT of Yb-labelled small dietary par-
ticles in the gastrointestinal tract averaged 62-2 h in
buffaloes, and 730 h in cattle (difference, P < 010)
Mathematical and statistical analyses
and was reduced from an average of 69-4 to 659 h by
The rate of digestion and lag time before the onset of urea supplementation (P < 005). These differences in
10 AGS 119
246 P. M. KENNEDY ET AL.

(Table 3). Four hours after feeding, only ammonia

concentration was increased (P < 005) by supple-
mentation. Pool sizes of VFA and ammonia were not
influenced by species differences (P > 0-05), but did
increase (P < 0-05) with supplementation. Plasma
urea concentration was consistently higher in buf-
faloes (P < 001).
Digestion of dietary CWC incubated in situ was
I
unaffected by animal species or urea supplementation,
10 20 30 40 50 with the exception of lag time, which was reduced
Incubation time in rumen (h) (P < 005) from 7-8 to 4-8 h by supplementation (Table
Fig. 1. Accumulation with time of N on cotton thread 3). In contrast, for cotton thread digestion, lag time
(means+ S.E.) suspended in the rumen and sampled at To, of was not affected by urea, but rate of digestion of
buffaloes, ( • . • ) , cattle ( # , O), without urea supplement cotton thread was increased from 1-6 to 2-7 %/h (P <
( • . • ) , with urea supplement (D, O) (Expt 1). 005). Lag time for cotton thread was less in buffaloes
than in cattle (15 v. 23 h, P < 005). The accumulation
of N on cotton thread was rapid between 5 and 12 h,
TMRT appeared to be attributable to a combination and was more gradual thereafter (Fig. 1). There was
of lower transit times and retention times in both more (P < 0-05) N accumulated on threads in
compartments. The two alternative procedures to buffaloes at all times except 5 h than in cattle, and on
estimate ruminal retention time (\/kJ yielded similar threads from animals given urea supplements at 11
results (Table 2). and 24 h than in animals without urea. The limiting
Concentrations of VFA and ammonia in the rumen concentration of ammonia, measured at T o (which
fluid sampled before feeding were greater in buffaloes would more closely represent the daily mean than the
(P < 005; P < 0-01, respectively), with only ammonia T4 concentration (Romero 1976)), for rate of digestion
increasing with urea supplementation (P < 0-01) of dietary CWC was c. 50 mg N/l, while that required

Table 2. Dry matter (DM) intake, rumen digesta fill (per unit rumen-free body weight, RFW) and DM content
before (To) and 4 h after (Tt) feeding, and marker kinetics (total mean retention time, TMRT; transit time, TT,
and fractional turnover rates of the slow (kj and fast (k2) pools) in buffaloes and cattle given a diet of Rhodes
grass with or without urea supplement (Expt 7)

S.E.D
Comtrol With urea
ftp t WPP n Within
vv l l l l l l l
DCIWCC11
Buffaloes Cattle Buffaloes Cattle species species

Voluntary DM consumption
(kg/day) 3-28 3-44 3-76 419 0-15 016
(g/kg body wt/day) 11-7 131 13-1 15-8 0-70 0-54
Rumen fill (g/kg RFW)
To 169 245 187 264 23 7
194 281 211 324 25 18
Rumen DM pool (kg)
To 5-70 6-13 6-52 6 51 0-73 0-46
T4 6-55 6-47 7-08 7-44 0-64 0-49
Rumen DM content (g/kg digesta)
To 144 147 122 118 10 9
148 144 116 115 14 13
Retention time, Yb (h)
Total TMRT* 63-7 75-1 60-8 71-0 5-4 19
TT* 19-5 22-1 181 22-8 2-5 2-1
\/k* 37-7 46-3 37-3 42-8 4-5 1-6
1/*,* 6-4 6-7 5-4 5-6 2-5 20
I/*it 39-8 450 36-8 42-8 4-5 20

* Data fitted to the model of Grovum & Williams (1973).

t Data fitted to the model of Dhanoa et al. (1985).
 Urea in buffaloes and cattle 247

Table 3. Concentrations before (To) and 4 h after (T4) feeding of volatile fatty acids, rumen ammonia, plasma urea,
and in situ digestion rate, lag time and potential digestibility in buffaloes and cattle given a diet of Rhodes grass
with or without urea supplement {Expt 1)

S.E.D
Control With supplement
Between Within
Buffaloes Cattle Buffaloes Cattle species species

Volatile fatty acids (mM)

To 57 64 47 46 6 4
T4 55 55 43 53 9 9
Rumen ammonia
(mgN/1) T o 33 18 71 32 8 7
T4 40 10 85 107 26 30
Pool (gN) T o 1-13 0-82 2-70 1-58 0-38 0-30
T4 1-52 050 3-58 604 1-25 1 39
Plasma urea (mg N/l)
T
o 171 70 251 65 40 42
T4 213 56 264 114 53 47
In situ digestion of dietary cell
wall constituents
Rate (%/h) 3-5 2-8 3-9 3-6 0-7 0-5
Lag time (h) 7-5 8-2 50 4-6 12 1-5
Potential digestion 0-59 0-57 0-55 0-56 004 0-04
In situ digestion of cotton thread
Rate (%/h) 21 10 2-7 2-6 0-5 0-5
Lag time (h) 17 29 13 18 4-7 4-8

Table 4. Intake {kg/day and g/kg body weight {W)/day) and digestibility of dry matter {DM), cell wall
constituents {CWC) and acid detergent fibre {ADF) of spear grass in buffaloes and cattle infused with various levels
of urea {Expt 2)

Dry matter intake Digestibility (g/kg)

1 lrp:i
Species (g/day) (kg/day) (g/kg W/day) DM CWC ADF
Buffaloes 0 4-35 15-6 337 391 448
Cattle 0 3-30 13-2 390 497 539
Buffaloes 4-6 4-52 163 352 417 453
Cattle 4-7 3-45 13-6 406 515 571
Buffaloes 20-8 500 17-8 360 415 467
Cattle 20-3 400 15-6 415 499 560
Buffaloes 931 5-83 19 5 398 445 514
Cattle 100-7 4-66 181 418 485 571
S.E.D.
Between species 203 0-32 1-2 30 45 41
Within species 203 0-25 0-8 25 37 29

for minimal residue of cotton thread after 24 h ratic effects, P < 0-05) (Table 4). Voluntary intake
incubation, was 60-80 mg N/l. of buffaloes without urea was identical (15-6 g/kg
BW/day) to that of cattle receiving 20 g/day of urea.
VFC tended to reach a maximum for both species
Experiment 2
when N intake was c. 25 g per kg of digestible DM
Mean body weight of buffaloes was 302 kg and cattle (Fig. 2), when the amount of urea infused represented
272 kg (difference, P = Q\2). Buffaloes ate more spear 8-5 g and 11-5 g N per kg DM intake for buffaloes and
grass than cattle (kg/day, P < 0 0 1 ; g/kg BW/day, cattle, respectively. Overall, apparent DM digestibility
P < 0-10), but VFC of both species responded simi- increased {P < 0-05) with increasing urea infusion
larly to urea infusions (linear effects, P < 0-001, quad- from 364 to 408 g/kg intake. Digestibility of DM,
10-2
248 P. M. KENNEDY ET AL.

CWC, ADF and hemicellulose (CWC-ADF) was

higher in the cattle (P < 010, P < 005, P < 005 and
P < 010, respectively). Separate analyses of variance
indicated tendencies (/" < 0'10) in buffaloes for in-
creased DM and ADF digestibilities with increasing
urea infusion, while in cattle, treatment effects were
Q a a"a not statistically significant (P > 038) for digestibility
of DM or any fibre component. Within both species,
DM and hemicellulose digestibility differed between
o animals (P < 005).
Rumen digesta fill per RFW and rumen DM pool
10 15 20 25 30 35 40 were similar in both animal species (P > 005) (Table
Intake of N/digestible DM (g/kg) 5). Infusion of urea increased rumen fill (/• < 0-10)
Fig. 2. Relationship of voluntary intake of spear grass with and rumen DM pool (P < 005). The DM content of
N content of the diet expressed per unit of digestible DM rumen digesta was greater in buffaloes than in cattle
intake for buffaloes ( • ) and cattle ( • ) in Expt 2. (150 v. 139 g/kg; P < 0-05). With increasing amounts

Table 5. Rumen fill (g/kg rumen-free body weight (RFW)), dry matter (DM) content in rumen, and apparent
fractional fluid and passage rates in buffaloes and cattle infused with various levels of urea {Expt 2)

Rumen

Passage rate (%/h)

TJrC3 Fill DM pool DlVf contpnt
Species (g/d) (g/kg RFW) (kg) (g/kg digesta) Apparent DM Fluid

Buffaloes 0 195 7-40 152 •64 5-62

Cattle 0 231 6-75 134 •29 4-30
Buffaloes 4-6 204 7-60 153 •64 6-58
Cattle 47 217 6-53 144 •32 400
Buffaloes 20-8 222 8-05 149 •67 610
Cattle 20-3 242 7-42 132 •34 3-61
Buffaloes 93-1 249 895 144 •66 5-62
Cattle 100-7 226 7 51 146 •56 4-26
S.E.D.
Between species 20 0-75 6 0 21 0-64
Within species 15 0-50 5 012 0-67

Table 6. Concentrations of ammonia, volatile fatty acids (VFA) and microbial nitrogen pool and content of dry
matter (DM) in the rumen, and sporangia of fungi on grass blades in buffaloes and cattle (Expt 2)

Microbial N
Ammonia (mg N/l) Fungal
Urc3 VFA snnrHTiPis pool pnnlpn t
Species (g/ d a y) 0900 h 24 h (mM) (no./mm 2 ) (g) (g/kg DM)

Buffaloes 0 16 22 69 0 59 7-92
Cattle 0 18 27 69 0 68 9-62
Buffaloes 4-6 13 24 70 0-2 65 8-70
Cattle 4-7 26 31 75 1-2 67 10-34
Buffaloes 20-8 41 51 71 0-5 76 9-47
Cattle 20-3 76 132 78 0-2 80 11-81
Buffaloes 931 140 190 76 01 86 10-24
Cattle 100-7 163 232 75 0-8 89 11-91
S.E.D.
Between species 16 23 7 0-7 11 0-10
Within species 14 "25 6 0-7 8 009
 Urea in buffaloes and cattle 249

of infused urea, DM content of rumen digesta in

buffaloes declined, while that in cattle increased
(species x treatment interaction, P < 005).
Fractional passage rate of rumen fluid, measured
by CrEDTA kinetics, was higher in buffaloes than in
cattle (60 v 4 0 % / h ; /><0001), but there was no
effect of urea infusion on fractional passage rate from
the rumen of fluid or of DM (Table 5). CWC
digestibility ( / , g/kg CWC intake) was negatively
related to apparent DM passage rate (X, %/h),
according to the following equation: 30 40 50 60 70 80
Time (h)
Y= 764-202 X (R.S.D. =4-93; r2 = 0-66; /><0001) Fig. 4. Patterns of digestion with time of incubation in
Rumen VFA concentration was similar in both species nylon bags of cotton thread (A), verano ( • ) , calopo ( • ) ,
and was unaffected by urea infusion (Table 6). ryegrass (O), rice straw (A) and spear grass ( • ) in Expt 2
(overall means across urea infusions).
Ammonia concentration in rumen fluid was similar in
samples taken 1 h before feeding (at 0900 h) for 7
days and from samples taken frequently during a 24-h 20 (difference, P < 001). The N content of fine par-
cycle. Low ammonia levels (20-27 mg N/l) were ticles (Y, g N/kg DM) was related to the propor-
measured in animals without infused urea, but these tion of rumen microbes in that fraction (X), calculated
increased linearly (/>< 0-001) to c. 200 mg N/l by reference to 35 S, by the following equation:
in animals infused with the highest level of urea. Y= 0-746 + 680 X
Buffaloes had lower concentrations than cattle of
(R.S.D. = 2 1 8 ; r2 = 0-55; P < 0001)
rumen ammonia measured at 0900 h over 7 days and
throughout 24 h (68 v. 91 mg N/l, / ) < 0 0 5 ; 72 v. Total microbial N in the rumen digesta increased
105 mg N/l, P < 001, respectively), but when the with urea infusion (lineareffects, P < 0001; quadratic
means were covariance adjusted for amount of urea effects, P <005) with no species differences (Table
infused, significant (P < 005) species differences were 6). Microbial N in the fine particles comprised 11-2%
observed for the 24-h sampling only. Pool sizes of of the rumen microbial pool and was not affected
ammonia and VFA in the rumen were not different (P > 0-5) by species or urea infusion. In contrast,
between species {P > 005) but were increased (P < microbial N content per unit of DM was lower in
0-05; linear effects, P < 001) by urea infusion. buffaloes than in cattle (91 v. 10-9 g N/kg DM; P <
Rumen fines comprised c. 50 g/kg of rumen 005). The fungal sporangial numbers counted were
particulate DM. The non-ammonia N content (g/kg low (0-1-2 per mm2 of leaf blade) and not affected by
DM) of the fines and coarse particles was increased animal species or by treatment.
with urea infusion (linear effects, P < 001 ; quadratic The mean recoveries of five forages and cotton
effects, / > <0-01). In cattle, N content of coarse thread after 24 h incubation in situ (Fig. 4) indicate
particles was higher (10-2 v. 86 g/kg DM; P < 001) that c. 500-550 g/kg of verano and ryegrass had been
than in buffaloes. The relationship between N content digested, compared with 150-300 g/kg for the other
of fine and coarse particles is shown in Fig. 3. The substrates. Between 24 h and 72 h of incubation,
concentration of N in fines was 2-46 that of coarse cotton thread was extensively digested, with spear
particles in buffaloes, whereas in cattle the ratio was grass, ryegrass, rice straw, calopo and verano showing
progressively less digestion. Of the six substrates
incubated in situ, only the cellulose of cotton was
digested more completely with increased infusion of
5 urea for every incubation period (linear effects, P <
o. 0001; quadratic effects, P < 005). The minimum
10
rumen ammonia level (determined from samples taken
over 24 h) for optimal digestion of cotton thread after
24 h incubation was c. 150-200 mg N/l (Fig 5 a). With
increasing ammonia concentrations, only calopo and
ryegrass incubated for 24 h showed similar increases
in digestion that achieved statistical significance (treat-
18 19 20 21 22 23 ment effects, / " < 0 0 5 and />< 0-001, respectively).
N content of Tine particles (g/kg DM) Using the plateau model of Mehrez el al. (1977)
applied to amount of residue after 24 h incubation
Fig. 3. Relationship of mean N content of coarse and fine
rumen particles in the rumen of buffaloes ( • ) and cattle ( • ) relative to that after 72 h, the optimal ammonia
receiving four rates of infused urea in Expt 2. concentration (using samples taken over 24 h) for
250 P. M. KENNEDY ET AL.

1000 r (a)
a •
900

"a
800
jj
(N
700
toi

o
U 600 10 20 30 40 50 60 70 80 90 100
50 100 150 200 250 300
Urea infused (g/day)
Rumen ammonia (mg N/l)
u (b) Fig. 7. Concentration of N in cotton thread removed from
idu

1-6 the rumen of buffaloes, relative to cattle values, for all

S treatments in Expt 2. Incubation times; 24 h ( • ) , 48 h ( • ) ,
1-5
and 72 h (O).
14

1 1-3
• 8
to 97 g/day for spear grass and calopo (Fig. 6). For
resi due

« 1-2 A animals receiving the highest rate of urea infusion,

1-1 there was a linear relationship between the relative
o A
A
I
B content of N in residues incubated for 24h (Y, ratio
1 1 1
10 buffalo: cattle) and original CWC content of the
o: 0 50 100 150 200 250 forages incubated (X, g/g DM) as described by the
Rumen ammonia (mg N/l) following equation:
Fig. 5. Relationship after ruminal incubation for 24 h of Y= 0-72 + 0-432 X
residue weight of (a) cotton thread in buffaloes ( • ) and 2
cattle ( • ) and (b) roughages with concentration of ammonia (R.S.D. =001213; r = 0-95; P <QQ\)
in rumen fluid sampled over 24 h (Expt 2). For (6), values are The N content of cotton thread from buffaloes was
expressed as relative to residue recovery at 72 h, and symbols
as for Fig. 4.
higher (P < 005) than that from cattle after 24, but
not after 48 or 72 h incubation. The N content of
cotton threads at all incubation times was positively
related to level of urea infusion (treatment differences,
P < 0001), and N content in cotton threads recovered
from buffaloes after 48 and 72 h relative to the N
content in treads recovered from cattle at the same
time was related in a curvilinear manner to the
amount of urea infused (Fig. 7).
After 72 h incubation of spear grass and rice straw
in nylon bags, 2-0-3-5% less DM (P < 005) was
recovered from buffaloes than from cattle. CWC
content (per weight DM) of these residues was similar.
Urea infused (g/day) There was also a tendency (/ ) <010) for lower
recovery from buffaloes of verano incubated for 48
Fig. 6. Concentration of N in roughage residues incubated
for 24 h in the rumen of buffaloes, relative to cattle values,
and 72 h.
for all treatments in Expt 2. Symbols as for Fig. 4.
DISCUSSION
digestion of forage substrates appeared to be c.
VFC, digestibility and passage rates
30-60 mg N/l (Fig. 5 b).
Across treatments, the N content of residues from The responses in V F C to urea supplementation
in situ 24-h incubation was lower for buffaloes than indicated that there was a primary N deficiency in
cattle for verano (P < 0001), calopo {P < 001) and both diets when sulphur and minerals were provided.
rice straw (P < 005). Increasing infusion of urea was The increase in Expt 1 in V F C ( 1 8 % ) obtained by
associated with greater N content of these residues in supplementation with urea when sulphur and minerals
both animal species for calopo (/* < 0-001), spear were available was similar to that (20-29 % ) reported
grass (/»< 0001) and ryegrass (P<0-05); these previously for Brahman crossbred and Hereford cattle
increases were relatively greater for buffaloes for urea fed spear grass (Siebert & Kennedy 1972; Hunter &
infusion levels up to 20 g/day for all forages, and up Siebert 1985b). T h e V F C response to urea and sul-
 Urea in buffaloes and cattle 251

phate in Expt 2 occurred even when rumen ammonia bility and apparent passage rate implies that CWC
was in excess of the concentration considered as digestion in the rumen for spear grass retained for
limiting to digestion (50-100 m g N / 1 ; Satter & 72 h would be c. 14% greater than that retained for
Slyter 1974) or to intake of sheep (50-80 m g N / 1 ; 48 h. In contrast, the prediction from nylon bag data
McMeniman & Elliott 1984). This result was in indicated a value of 9%. The discrepancy may have
contrast to the response observed in young Hereford been due to different conditions applying to digestion
steers offered low-quality hay sprayed with various of particles in the rumen digesta and in nylon bags,
amounts of urea (Hennessy 1984). In the latter attributable to increased rate of digestion or by
experiment, urea intakes > 50 g/day did not result in reduction of lag time caused by grinding prior to in
increased intakes. However, the increase in VFC in situ incubation, or to differences in microbial popu-
the present experiment was similar, but less in lations or activities. Alternatively, the estimation of
magnitude, to that reported when rumen ammonia apparent passage rate may have substantially under-
levels were increased from below 2 to 140 mg N/l in estimated the true passage rate by attributing all
the study of Boniface et al. (1986), using similar cattle disappearance of digesta DM to passage, whereas a
and infusion regimen, but with spear grass hay proportion would be digested in the rumen. The latter
containing 25 % less N than that used in the present alternative is supported by the lower mean values of
experiment. apparent DM passage, ranging from 1'3 to 1-7%/h
In Expt 2, VFC for both species increased until in Expt 2, compared to the higher values of kl of
N intake was c. 25 g/kg digestible DM (Fig. 2), 2 1 - 2 - 7 % / h in Expt 1, obtained by use of indigestible
equivalent to c. 30 g/kg DM digested in the rumen, markers.
assuming that the proportion of digestible DM Faster passage of plant residues from the rumen of
apparently digested in the rumen was similar to that buffaloes has been attributed to greater duration of
(0-82) observed for OM in cattle fed spear grass and rumination and more forceful stomach contractions
urea/sulphate supplements (Hunter & Siebert 1980). in buffaloes (McSweeney & Kennedy 1989; Kennedy
The VFC response to additional urea can therefore be et al. 1992) and is even more marked for fluid passage
accounted for by the requirements of rumen microbes rate (Table 5).
for N, which may reach 30-35 g N/kg OM apparently
digested in the rumen (Kennedy et al. 1992).
Ammonia levels, rates of digestion, and urea
VFC of buffaloes relative to cattle was 25-30%
recycling
higher, and digestibility was lower, with CWC
digestibility inversely related to apparent DM passage Hunter & Seibert (1985a) found digestion rates of
rate. This result was in agreement with general ground dietary CWC were depressed at low ammonia
negative relationships between intake and digestibility concentrations below 30-40 mg/1 in Hereford steers.
within animals, which reflects the positive relationship A similar depression of roughage digestion in Expt 2
between intake and digesta passage rate (Warner was detected, although not as marked as in previous
1981). However, in contrast to expectations from reports in cattle (Boniface et al. 1986) or in sheep
these overall relationships, as VFC increased with (Perdock & Leng 1989). It was of interest that these
infusion of greater amounts of urea, digestibility decreases in rates of digestion were associated with
tended to increase in buffaloes or remain constant in decreases in the population of adherent microbes, to
cattle. The advantage in CWC digestibility enjoyed by the extent indicated by the N content of the residues,
cattle compared to buffaloes was progressively re- and also by decreased microbial content of rumen
duced from 127% in animals without infused urea to DM measured by 3SS. Optimal digestion rate of cotton
109% for animals with maximal urea infusion. The appeared to require c. 30 (Expt 1) to 100 mgN/1
association of faster passage with increased urea (Expt 2) more ammonia than the roughage substrates,
infusion and higher intake of cattle, but not of and data for Expt 2 is in agreement with the report
buffaloes, in Expt 2 was similar to the pattern of of Krebs & Leng (1984) who found enhancement of
retention time (1/fcJ of the digesta marker (Yb) in cellulose digestibility with increases of ammonia levels
Expt 1. The ability of buffaloes to increase VFC in up to 210 mg N/l. It is likely that the rate of digestion
Expt 2 without change in apparent DM passage rates of particulate material in the rumen is poorly
was related to a 25 % expansion of the rumen DM mimicked by digestion of ground dietary material
pool in buffaloes infused with maximal urea infusion, incubated in nylon bags, and indeed Perdock & Leng
whereas only a 11 % increase was observed in cattle. (1989) suggested that conclusions about the critical
Thus it appears that the increase in digestibility ammonia concentration could be affected by an
observed when buffaloes were supplemented with artifact caused by differential loss from the bags of
urea was masked in cattle by the counteracting small particles during washing, implying that recom-
depression of digestibility caused by faster rates of mendations on N requirements on the basis of
digesta passage. digestibility of ground dietary material in nylon bags
The relationship described between CWC digesti- or in vitro may be invalid. On the basis that microbial
252 P. M. KENNEDY ET AL.

content per weight of rumen digesta (Table 6) is a low (Ho & Abdullah 1987;G. L. R. Gordon, personal
better measure of ammonia needs, the optimal communication). On the other hand, the proportion
ammonia concentration for both cattle and buffaloes of cellulolytic organisms was probably higher in the
given continuous urea infusions was c. 100 mgN/1. buffalo microbial population, as shown by the greater
The results of Expt 1 indicated that buffaloes accumulation of N on cotton thread in Expt 1; these
maintained higher levels of rumen ammonia than microbes also seemed to detach or lyse at a faster rate
cattle when fed Rhodes grass without urea; little in buffaloes (see Fig. 7). Faster detachment of
difference between species was seen in animals without microbes and subsequent association with the fluid
urea in Expt 2, although across treatments, cattle and small particle pool of the rumen may explain the
maintained higher ammonia levels than did buffaloes. relative enrichment in buffaloes of this pool in Expt 2,
Although greater recycling in buffaloes in Expt 1 and also that in the abomasum (Kennedy et al. 1992).
would be expected, given their higher levels of blood In other studies (Homma & Ichikawa 1983; Homma
urea, there was little evidence from relative pool sizes 1986), specific cellulase activity of microbes from
of rumen ammonia of the superior recycling of blood buffaloes was similar to or lower than from cattle.
urea to the rumen measured in these species given rice However, patterns of concentration of microbes in
straw with 5 % leucaena (Kennedy et al. 1992). In Expt 2 differed from those of Homma (1986), who
another study with buffaloes and cattle offered alkali- reported that buffaloes maintained higher concen-
treated straw pellets, Pearson & Archibald (1990) also trations of microbes than cattle when fed rice straw,
found no evidence for differences in urea recycling. In but not when the diet was timothy hay.
the present experiments, the species with the higher The adherent microbe population in buffaloes did
intake had the lower rumen ammonia levels, but it appear to be more able than cattle to respond to high
was not possible to attribute causal relationships, fibre diets by greater colonization, under conditions
especially as the two species maintained their relative of high infusion of urea and sulphate when passage
intakes when given large amounts of supplemental rates of digesta from the rumen of both species were
urea. comparable, and by greater fibre digestion. This
concept is supported by the finding that digestion of
palm press fibre (c. 80% CWC) in buffaloes was
Microbial concentration and activity markedly superior to that in cattle (Jainudeen 1983).
The lower concentration of microbes per weight of It is suggested that buffaloes harbour an organism, or
rumen digesta DM in buffaloes than in cattle found in a consortium of organisms, with fibrolytic capability
Expt 2 agreed with earlier observations in animals superior to that in cattle under comparable conditions
given a rice straw-based diet (Kennedy et al. 1992). of rumen digesta turnover, and that the degree of
In contrast, Homma (1986) reported that buffaloes colonization of fibre by these microbes is more marked
maintained higher concentrations of microbes than with highly fibrous substrates.
cattle when fed rice straw, but not when the diet was In conclusion, it appears that interactions exist
timothy hay. Despite a lower concentration of between ruminal N supply, digesta dynamics, com-
microbes in rumen digesta DM, buffaloes exhibited position of the diet, and microbial populations and
an ability to digest dietary and other substrates dynamics which may be manifested in differences in
immobilized in nylon bags, which was equal to that of digestion and microbial yield between Australian
cattle, and indeed was superior for several substrates. swamp buffaloes and cattle. These interactions merit
From the differences between buffaloes and cattle in further study with modern techniques to quantify
distribution of N on large and small rumen particles population sizes, fibrolytic activities and turnover of
(Fig. 3), it appears that the time available for microbial rumen microbes.
colonization and growth of adherent microbes may
have been affected by the enhanced comminution and
passage of particles in buffaloes. If greater digestion G. L. R. Gordon kindly provided data on numbers
of forages after 72 h of ruminal incubation in buffaloes of fungal sporangia. Partial funding was supplied by
than in cattle reflected more efficient microbial the Australian Centre for International Agricultural
digestion, this was not due to differences in popu- Research and the Australian Meat and Livestock
lations of anaerobic fungi in Expt 2 as detected in our Research and Development Corporation. The authors
experiment, and indeed the fungal populations were thank B. Stevens for technical assistance.
 Urea in buffaloes and cattle 253

REFERENCES
AKIN, D. E., GORDON, G. L. R. & HOGAN, J. P. (1983). voluntary food intake and rumen characteristics. British
Rumen bacterial and fungal degradation of Digitaria Journal of Nutrition 53, 649-656.
pentzii grown with or without sulphur. Applied and JAINUDEEN, M. R. (1983). The water buffalo. Perlanika 6
Environmental Microbiology 46, 738-748. (Rev. Suppl.), 133-151.
BINNERTS, W. T., VAN'T KLOOSTER, A. T H . & FRENS, A. M. KENNEDY, P. M. & WATERHOUSE, D. (1987). Utilization of a
(1968). Soluble chromium indicator measured by atomic poor quality diet by swamp buffalo and cattle. In Herbivore
absorption in digestion experiments. Veterinary Record Nutrition Research (Ed. M. Rose), pp. 179-180. Bris-
82, 470. bane, Australia: Australian Society of Animal Production
BONIFACE, A. N., MURRAY, R. M. & HOGAN, J. P. (1986). KENNEDY, P. M., BONIFACE, A. N., LIANG, Z. J. & MURRAY,
Optimum level of ammonia in the rumen liquor of cattle R. M. (1987). Responses to infusions of urea and sulphate
fed tropical pasture hay. Proceedings of the Australian in cattle and swamp buffaloes given mature Heteropogon
Society of Animal Production 16, 151-154. conlortus (spear grass). In Herbivore Nutrition Research
CHALMERS, M. I. (1974). Nutrition. In The Husbandry and (Ed. M. Rose), pp. 191-192. Brisbane, Australia: Aus-
Health of the Domestic Buffalo (Ed. W. R. Cockrill), pp. tralian Society of Animal Production.
167-194. Rome: Food and Agriculture Organization. KENNEDY, P. M., MCSWEENEY, C. S., FFOULKES, D.,
DEVENDRA, C. (1983). The utilisation of nutrients, feeding JOHN, A., SCHLINK, A. C , LEFEUVRE, R. P. & KERR, J. D.
systems and nutrient requirements of swamp buffaloes. In (1992). Intake and digestion in swamp buffaloes and
Current Development and Problems in Swamp Buffaloes cattle. 1. The digestion of rice straw (Oryza saliva).
(Ed. H. Shimizu), pp. 73-106. Ibaraki, Japan: University Journal of Agricultural Science, Cambridge 119, 227-242.
of Tsukuba. KREBS, G. & LENG, R. A. (1984). The effect of supplemen-
DEVENDRA, C. (1987). Herbivores in the arid and wet tropics. tation with molasses/urea blocks on ruminal digestion.
In The Nutrition of Herbivores (Eds J. B. Hacker & J. H. Proceedings of the Australian Society of Animal Production
Ternouth), pp. 23-46. Marrikville, Australia: Academic 15, 704.
Press. MADER T. L , TEETER, R. G. & HORN, G. W. (1984).
DHANOA, M. S., SIDDONS, R. C , FRANCE, J. & GALE, D. L. Comparison of forage labeling techniques for conducting
(1985). A multicompartmental model to describe marker passage rate studies. Journal of Animal Science 58,
excretion patterns in ruminant faeces. British Journal of 208-212.
Nutrition 53, 663-671. MCDONALD, I. (1981). A revised model for the estimation of
GROVUM, W. L. & WILLIAMS, V. J. (1973). Rate of passage of protein degradability in the rumen. Journal of Agricultural
digesta in sheep. 4. Passage of marker through the Science, Cambridge 96, 251-252.
alimentary tract and the biological relevance of rate MCMENIMAN, N. P. & ELLIOTT, R. (1984). The influence of
constants derived from the changes in concentration urea infusions on the intake of roughages by sheep.
of marker in faeces. British Journal of Nutrition 30, Proceedings of the Australian Society of Animal Production
313-329. 15, 719.
HENNESSY, D. W. (1984). Response of cattle on low quality MCSWEENEY, C. S. & KENNEDY, P. M. (1989). Reticulo-
pasture hay to increments of urea or to a supplement of ruminal motility in cattle (Bos indicus) and water buffaloes
protein. Proceedings of the Australian Society of Animal (Bubalus bubalis) fed a low quality diet. Comparative
Production 15, 388-391. Biochemistry and Physiology 94A, 635—638.
Ho, Y. W. & ABDULLAH, N. (1987). Digestion of fibrous MEHREZ, A. Z., 0RSKOV, E. R. & MCDONALD, I. (1977).
feed materials by rumen fungi in Kedah-Kelantan cattle Rates of ruminal fermentation in relation to ammonia
and swamp buffalo. Proceedings of the \0lh Annual concentration. British Journal of Nutrition 38, 437-443.
Conference of the Malaysian Society of Animal Production, MINSON, D. J. (1982). Effects of chemical and physical
pp. 265-267. composition of herbage eaten upon intake. In Nutritional
HOMMA, H. (1986). Cellulase activities of bacteria in liquid Limits to Animal Production from Pastures (Ed. J. B.
and solid phases of the rumen digesta of buffaloes and Hacker), pp. 167-182. Farnham Royal, Commonwealth
cattle. Japanese Journal of Zootechnical Science 57, Agricultural Bureaux.
336-341. PEARSON, R. A. & ARCHIBALD, R. F. (1990). Effect of
HOMMA, H. & ICHIKAWA, T. (1983). Cellulolytic activity, ambient temperature and urea supplementation on the
nitrogenous components and volatile fatty acids in the intake and digestion of alkali-treated straw by Brahman
rumen digesta of cattle and buffaloes. Japanese Journal of cattle and swamp buffaloes. Journal of Agricultural
Zootechnical Science 54, 690-696. Science, Cambridge 114, 177-186.
HUNTER, R. A. & SIEBERT, B. D. (1980). The utilization of PEARTON, D. C. G. & MALLETT, R. C. (1972). The de-
spear grass (Heteropogon contorlus). 4. The nature and termination of chromium, manganese, iron, cobalt, and
flow of digesta in cattle fed on spear grass alone and with nickel by atomic-absorption spectrophotometry using the
protein or nitrogen or sulphur. Australian Journal of nitrous oxide-acetylene flame. Report No. 1435, National
Agricultural Research 31, 1037-1047. Institute for Metallurgy of South Africa.
HUNTER, R. A. & SIEBERT, B. D. (1985 a). Utilization of low- PERDOCK, H. B. & LENG, R. A. (1989). Rumen ammonia
quality roughage by Bos laurus and Bos indicus cattle. I. requirements for efficient digestion and intake of straw by
Rumen digestion. British Journal of Nutrition 53, 637-648. cattle. In The Roles of Protozoa and Fungi in Ruminant
HUNTER, R. A. & SIEBERT, B. D. (19856). Utilization of low- Digestion. (Eds J. V. Nolan, R. A. Leng& D. I. Demeyer),
quality roughage by Bos taurus and Bos indicus cattle. 2. pp. 291-293. Armidale, NSW: Penambul Books.
The effect of rumen-degradable nitrogen and sulphur on ROMERO V., A., SIEBERT, B. D. & MURRAY, R. M. (1976). A
254 P. M. KENNEDY ET AL.

study on the effect of frequency of urea ingestion on the VAN SOEST, P. J. & ROBERTSON, J. B. (1980). Systems of
utilization of low quality roughage by steers. Australian analysis for evaluating fibrous feeds. In Standardization
Journal of Experimental Agriculture and Animal Husbandry of Analytical Methodology for Feeds (Eds W. J. Pigden,
31, 393-400. C. C. Balch & M. Graham), pp. 49-60. Ottawa, Canada:
SATTER, L. D. & SLYTER, L. L. (1974). Effect of ammonia International Development and Research Centre.
concentration on rumen microbial protein production in WANAPAT, M. (1989). Comparative aspects of digestive
vitro. British Journal of Nutrition 32, 199-208. physiology and nutrition in buffaloes and cattle. In
SIEBERT, B. D. & KENNEDY, P. M. (1972). The utilization of Ruminant Physiology and Nutrition in Asia (Eds C.
spear grass (Heteropogon contortus) 1. Factors limiting Devendra & E. Imaizumi), pp. 27-43. Sendai, Japan:
intake and utilization by cattle and sheep. Australian Japanese Society of Zootechnical Science.
Journal of Agricultural Research 23, 35-44. WARNER, A. C. I. (1981). Rate of passage of digesta through
VAN SOEST, P. J. (1982). Nutritional Ecology of the Ruminant. the gut of mammals and birds. Nutrition Abstracts and
Corvallis: O & B Books. Reviews 51(B), 789-820.