Ráðunautafundur - 15.02.1989, Side 23
-15-
of an Agrostis - Festuca site at Llyn Llydaw, Perkins et al. (1978) have recorded a
range of 1990-4000 kg DM ha"1 in harvestable net herbage accumulation between
years and an average utilisation of grasses by grazing sheep of about 50% of
current production. Mean biomass of grasses in 1969, which was a year of
comparatively low production, was 760 kg DM ha'1 (Agrostis tenuis) together with
280 kg DM ha"1 (Festuca ovina) and the combined biomass rarely approached 500 kg
DM ha"1 below which overgrazing would result.
Interpretation and application of these results for other rangeland sites is
complicated by the effect the proportion and spatial variability of different
vegetation communities has on grazing preference and hence utilisation.
Over-all management changes, which involve increasing or decreasing average
stocking rate, are further complicated by the existence of family groups of ewes
with strong associations with specific vegetation communities and/or topographic
regions. If selective adjustment is made to the average stocking rate, based, for
example, on weaning weight, the impact on the voluntary stocking density of
different vegetation communities may also be selective.
Further research is necessary before these problems can be resolved and accurate
management guidelines developed.
Progress could be made from further sheep census studies over a wider
geographical range. The practical difficulties associated with this may be partly
tackled through the use of remote-sensing, which depends on a close association
between plant community type and productivity and sheep density, combined with
more traditional ground land observations. Williams (1986, 1987, 1988) has
demonstrated the effectiveness of the Landsat - 5 Thematic Mapper or Airborne
Thematic Mapper for identifying vegetation type and biomass while Thompson and
Milner (1988) have demonstrated a close correspondence of sheep density and near
infra-red reflectance (wavelength 0.60-0.63 m) with correlation coefficients ranging
from 0.81 - 0.89 when data was grouped according to additional response to wave-
band 5 (0.63-0.69 m) and 7 (0.70-0.90 m) (Figure 2).
While such observations on stocking density are helpful there is a need to
associate animal performance (intake of herbage and live weight gain) to easily
identifiable aspects of vegetation state as well plant community area and
distribution. The use of hand-held radiometers, sward height and other simple
morphological measurements could be involved.
A major difficulty in such studies is, and has been for some time, the problem of
determining quantitatively the selective consumption by grazing animals of different
plant species. Although undesirable on animal welfare ground, the use of fistulae is
perhaps the most reliable technique. Recent developments in the use of n-alkanes as
markers for the determination of herbage intake have proved useful in two species