A wildfire in an afforested research catchment presented the rare oppo
rtunity to compare the hydrological effects of wildfire with the effec
ts of clearfelling in the same catchment in the Jonkershoek Valley, in
the south-western Western Cape Province of South Africa. The timber p
lantation, which occupies 57% of the 2 km(2) catchment, had been clear
felled and re-planted to Pinus radiata roughly five years before the f
ire. The effects of the two treatments on total flow, stormflow and qu
ick-flow volumes, peak discharge and storm response ratio were determi
ned by means of multiple regression analysis, employing the dummy vari
able method to test for the significance of treatments. Both clearfell
ing and wildfire caused significant increases in all the stream-flow v
ariables analysed. But the clearfelling effect was dominated by large
increases in total flow (96% over three years), of which storm-flow an
d quick-flow volumes formed only minor parts. After the wildfire, by c
ontrast, increases in total flow were small (12%) but the storm flow i
ncreases were three- to fourfold in the first year and roughly double
in the second year. The wildfire caused fire-induced water repellency
in the soils which led to overland flow on mid-slope sites, where soil
infiltrability normally far exceeds local rainfall intensities. It is
argued that these results support the hypothesis that stream-flow gen
eration processes were changed by the wildfire in that overland flow m
ade a direct contribution to storm flows, but that clearfelling had no
such effect. (C) 1997 by John Wiley & Sons, Ltd.