ROLE OF PLANT COVER AND STOCK TRAMPLING ON RUNOFF AND SOIL-EROSION FROM SEMIARID WOODED RANGELANDS

Relationships between plant cover, runoff and erosion of a massive red earth were investigated for a runoff zone of an intergrove area in a semi-arid wooded rangeland in eastern Australia. The measurements were carried out in small experimental paddocks with different stocking rates of sheep and kangaroos. A trailer-mounted rainfall simulator was used to apply rainfall at a time averaged rate of 30 mm h-1 to obtain runoff rates and sediment concentrations. There was a significant negative relationship (r2 = 0.58; P < 0.01) between final runoff rate and plant cover. It is probable that the plants increase infiltration and decrease runoff by (i) funnelling water down their stems and (ii) providing macropores at the base of the plant through which water can rapidly enter the soil. However, there was no significant effect of plant cover on sediment concentration. Probable reasons for this are: (i) even though plant cover will absorb raindrop energy and decrease the erosive stress on the soil, the nature of the plants investigated is such that they may not be 100% effective in protecting the soil beneath them, and (ii) the distribution of contact cover provided by the base of the plants is highly patchy and thus relatively inefficient at reducing sediment concentration. At zero cover final runoff rates from paddocks with a high and low stocking rate were similar, i.e. 23.4 and 22.3 mm h-1 respectively. However, at zero cover, the sediment concentration from the high stocking rate paddock was significantly (P < 0.01) greater than that from the low stocking rate paddock. Greater hoof activity and lower organic matter (and hence lower structural stability) of the 0.20 mm layer in the high stocking rate paddock caused the soil surface to be more susceptible to erosion. These results show that grazing by removing perennial grasses and pulverizing the surface soil can have a major impact on local water balances and erosion rates respectively within the intergrove areas. The implications of these results for the long-term stability of semi-arid mulga woodlands is briefly discussed.