Bari 3 consistently experienced the lowest runoff, both in 1992 when the farmer
failed to weed the terrace beneath the maize crop, and in 1993 when the maize crop
was weed as normal. This plot differed from the others in that the terraces possessed
a sandy, well-draining soil with no rise ditches. The soil texture encouraged
infiltration and the plot configuration, of low-angle terraces without ditches, gave
more potential for on-plot depression storage and a longer residence time as the
water moves across the three terraces monitored (total length of 16.8 m). Variability
in response was also great for individual rainfall events, as well as seasonal loss and
runoff. This is why meaningful comparisons are only possible through the use of
empirically derived relationships between the key variables at each location.
Regression relationships between total rainfall and runoff, using a complete set
of pre-monsoon and monsoon rainfall events, standardized for each site, have enabled
annual mean runoff coefficients to be calculated. Estimated total runoff as a percentage
of total rainfall was 5–10% for Bari 1, 10 and 11, approximately 3% for Bari
3, and 18–26% for Bari 2 (Fig. 8). For comparison, the monsoon runoff coefficients
for good secondary forest cover and highly degraded forest cover were found to be
1% and 33% respectively (Gardner & Gerrard, 2002)