Effects of deep tillage and fertilization on soya yields in a compacted Ustochrept during seven cropping seasons, Santa Cruz, Bolivia

A study was carried out in tropical eastern Bolivia to determine the profitability of deep tillage practices and fertilization on soya in a compacted sandy loam Ustochrept. The experimental site is characterized by summer and winter cropping seasons with average rainfalls of 807 and 235 mm respectively.

A field trial was conducted over four summer and three winter cropping seasons (1985–1989) using a randomized complete block design with four tillage treatments and four replicates which was converted into a split-plot design by with- and without-fertilizer subtreatments in the second year. The tillage treatments (and mean tillage depth) were:

1. (1) Conventional — generally two passes of a heavy disc harrow (0.15 m) followed by two passes of a light disc harrow.

2. (2) Disc ploughing in 1985 — one pass of a disc plough (0.30 m) in 1985 only, followed by conventional tillage.

3. (3) Subsoiling in 1985 — two overlapping passes of a subsoiler (0.40 m) to give 0.37 m tine separation distances in 1985 only, followed by conventional tillage.

4. (4) Annual subsoiling — as for subsoiling in 1985 but with subsoiling every year.

For the seven cropping seasons, 1985–1989, the percentage yield increases from subsoiling in 1985, disc ploughing in 1985 and annual subsoiling compared with yields from conventionally tilled (compacted) plots were 14%, 19% and 25% respectively, equivalent to partial gross margins of US$ 230, 306 and 378 ha⁻¹. Significant differences were found between both the disc ploughing in 1985 and annual subsoiling treatments compared with conventional tillage; the residual effects of disc ploughing annual subsoiling in 1985 were observed to last 3 and at least 4 years respectively.

Significant fertilizer responses were obtained in 2 of the 3 years but were not economic owing to high fertilizer costs. No significant tillage×fertilizer interaction effect was found.

Soya responses due to annual subsoiling varied from 0 to 90% and were shown to be significantly and inversely related to seasonal rainfall when the latter was <760 mm, and to be approximately zero at higher seasonal rainfalls, provided that yields were not influenced by population differences. The higher responses of subsoiling at low seasonal rainfalls were attributed to increased rooting depth and hence to increased moisture supply. It was estimated that the minimum soya yield responses expected 7 years in 10 would be 0% in summer and 56% in winter, equivalent to a partial gross margin of US$ 98 ha⁻¹ for the first year.