Influence of livestock grazing on plant diversity of Highland Sourveld grassland in KwaZulu-Natal
Date
2008-07-15T10:27:58Z
Authors
Martindale, Gregory John
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The effect of livestock type (cattle-to-sheep ratio) and stocking rate on grassland species richness and
vegetation composition were tested using two long-term trials in Highland Sourveld Grassland at the
Kokstad Agricultural Research Station in KwaZulu-Natal.
The two-paddock trial (initiated 1996) was grazed by sheep at three different stocking rates,
each of which was replicated twice. The simulation trial (initiated 1989) simulated a four-paddock
rotational grazing system in which one paddock was rested for the entire grazing season and the other
three were rotationally grazed according to a fixed schedule. It consisted of two stocking rates and
five ratios of cattle-to-sheep.
Plant diversity and species composition were sampled using a Modified-Whittaker Plot in
December 2005. For each trial, differences among treatments in species richness, diversity and
evenness were analysed using analysis of variance. Differences in botanical composition were
evaluated using canonical correspondence analysis.
The results from the two-paddock trial suggested that the responses had been confounded by
burning, and by environmental and possibly historical variables. Lower species richness, especially of
non-graminoid species was recorded for treatments that had been burnt during the preceding
September. Despite these issues, an effect of grazing regime was apparent. There was an increase
in mean number of species per 1m2 quadrat in the treatments grazed at the high stocking rate
compared to those grazed at the low and medium stocking rates, which were associated with changes
in species composition that enabled species suited to disturbed areas to colonise these treatments.
As expected, species such as Themeda triandra declined in abundance with increasing grazing
pressure whilst weedy species such as Senecio speciosus increased in abundance.
A higher mean number of non-graminoid species were found in the simulation trial in the
treatments grazed at the low stocking rate but the converse was true of the graminoid species.
Unpalatable species such as Sporobolus africanus and Eragrostis plana increased in response to
increasing grazing intensity. Senecio retrorsus, which is known to increase in grassland that is overutilised
and Tolpis capensis, which is often found in disturbed places, increased in response to grazing
intensity. Consistent with the findings in the other trial, Helichrysum nudifolium and Vernonia
natalensis declined in response to grazing intensity. A general decline in the mean number of species
per 1m2 quadrat and in the total number of species in all 1m2 quadrats was detected in response to an
increasing proportion of sheep in the cattle-to-sheep ratio. The abundance of Alloteropsis semialata
increased significantly at both stocking rates at low ratios of cattle-to-sheep whilst Pentanisia
angustifolia, Aster bakerianus and Monsonia angustifolia declined dramatically in abundance with
decreasing ratios of cattle-to-sheep, in some instances disappearing altogether. The results suggest
that the ratio of cattle-to-sheep influenced both the species composition and species richness in the
trial, resulting in a lower overall species richness at lower ratios of cattle-to-sheep and lower overall
non-graminoid species richness at higher grazing intensities.
A major shortcoming of the data was the lack of statistical replication available to accurately
determine trends and patterns associated with the different grazing regimes. This could have been
overcome by establishing baseline conditions at the start of the trials and monitoring changes over
time. This would have enabled trajectories of change associated with particular plant species to be
identified and a clearer understanding could have been developed of the influence of different ratios of
cattle-to-sheep and of different grazing intensities on species composition and richness over time.
The results of the study illustrate that sheep are not an ideal choice of livestock for maintaining
plant diversity when grazing humid grasslands in KwaZulu-Natal, for which the best choice is a mixedspecies
grazing regime, consisting of a relatively high ratio of cattle-to-sheep.