The main reasons for increased surface runoff and reduced infiltration are plowing along the slope and the formation of a plow pan at shallow depths.

Conservation tillage is seen as a way to alleviate these problems. The widely advocated zero-tillage, however, is not feasible for smallholder farmers in semi-arid regions of Ethiopia because of difficulties in maintaining adequate soil cover, the practice of communal grazing, and high costs of herbicides. Strip tillage systems, on the other hand, may offer a solution. This study was initiated to test strip tillage systems and to evaluate the impacts of new tillage systems on the water balance and grain yields of maize.

Experiments have been conducted in a semi-arid area called Melkawoba in the central Rift Valley of Ethiopia during 2003—2005. Strip tillage systems involved cultivation along planting lines at a spacing of 0.75 m using the Maresha plow followed by subsoiling along the same lines (STS) or without subsoiling (ST). Results have been compared with traditional tillage involving 3—4 overpasses with the Maresha plow (CONV). Soil moisture has been monitored to a depth of 1.8 m using a Time Domain Reflectometer (TDR) while surface runoff has been measured using a specially designed rectangular trough installed at the bottom of each plot.

STS resulted in the least surface runoff (Qs = 18 mm season?1) and the highest grain yields (Y = 2130 kg ha?1) followed by ST (Qs = 26 mm season?1, Y = 1840 kg ha?1) and CONV (Qs = 43 mm season?1, Y = 1720 kg ha?1) provided sowing was carried out within a week after subsoiling. Thus, STS resulted in the highest water productivity, WP = 0.60 kg m?3, followed by ST (WP = 0.52 kg m?3) and CONV (WP = 0.48 kg m?3).

The main conclusion of the paper is that even in dry areas reasonable yields can be obtained provided moisture conservation in the root zone is guaranteed. In this regard subsoiling is essential. Moreover, it is concluded that the time between subsoiling and planting is a key factor and should not exceed one week.