Tropical Grasslands (2006) Volume 40, 6569
Diet selection of steers grazing Stylosanthes hamata cv. Veranograss
pastures in north Queensland and its potential influence
on botanical composition
R.J. JONES1 AND F.D. HU1,2
1 CSIRO Tropical Ecosystems, Davies Laboratory,
Townsville, Queensland, Australia
2 Present address: CSIRO Plant Industry, Davies
Laboratory, Townsville, Queensland, Australia
The diets of steers grazing 3 Stylosanthes hamata
cv. Verano-grass pastures, each stocked continuously
at 0.65, 0.95 and 1.25 steers/ha, were studied over
a 3-year period. Dietary C4 (grass) and C3 (legume)
components were estimated using δ13C measurements
of the faeces and pasture components on 14 occasions.
In May each year, pasture yield and botanical composition
were also measured. The % legume in the pastures
of Verano–Urochloa mosambicensis cv. Nixon
(Nixon) and Verano–Bothriochloa pertusa cv. Bowen
(Bowen) were then compared with the % legume in
the diets on the dates closest to the May pasture sampling
to calculate a selection index (SI) for legume.
Associated grass species had a greater effect on
dietary legume content than did stocking rate (SR)
and there was no species x SR interaction (P > 0.05).
Overall, steers on Nixon pastures consumed diets
with a lower proportion of legume (25%) than those
on Bowen pastures (47%), with marked differences in
the first year of the study, smaller but still significant
differences in the second year and no differences in
the third year. Diets of steers on rhodes pastures were
intermediate (37%). Dietary legume percentages of
steers at high SRs were generally lower than those at
the other SRs, but on only 3 of 14 sampling occasions
were the differences significant (P < 0.05).
Dietary legume content varied widely over the
year on all treatments, with highest values in the late
wet–early dry season (May–July) and lowest in the
late dry–early wet season (November–February).
Although, overall, the SI for legume was higher
(more legume) with Bowen, only in the final year
was the difference signifi cant (P < 0.01). This index
appeared to be related to pasture composition and not,
as its name suggests, to animal preference.
It is concluded that the associated grass had a
greater effect than stocking rate on diet selection, and
that the trend for steers on Bowen pastures to select a
diet higher in Verano over a longer period of the year
than those on Nixon pastures, could lead to botanical
changes that favour the grass component.