Other variables contributed much less, if at all, to the explanation of runoff variance.
Antecedent 24-hour rainfall and soil suction both added a little further explanation,
as did the percentage of ground cover vegetation. However their overall contribution
was small. For example, on Bari 1 lower terrace, the inclusion of both
vegetation cover and antecedent 24-hour rainfall raised the R2 value from 73.0% to
82.3% (each contributed approximately 5%).
The relationships depicted in Fig. 7 indicate that the ranking of terraces in terms
of runoff generation was consistent from 1992 to 1993 for those terraces monitored
in both years. Bari 2 generated substantially more runoff than Bari 1, which generated
greater runoff than Bari 3. Generally the variation between sites was greater than
the variation between the terraces at any one site, although intra-site differences
clearly existed. The differences reflected variations in soil texture, bulk density, infiltration
properties, and possibly downslope terrace length. Bari 2 appeared to owe
its high runoff to a combination of a fine-textured soil (clay loam) with a very low
infiltration capacity (0.42–0.78 cm hr–1
) and poor soil structure. High runoff may
have been a response to the narrowness of the terrace that offered less surface storage
potential. Of the ditched terraces, Bari 10 experienced the lowest runoff, despite
relatively poor crop growth. This possibly reflected the sandier soil texture and high
infiltration rate in these previously heavily eroded soils. Bari 1 was intermediate in
terms of runoff, surface soil texture and infiltration capacity. At present, it is difficult
to explain the low runoff coefficients on the finer-textured soils of Bari 11.