Primary and secondary tillage are essential for dryland crop production in Botswana and it contributes more than half of the costs for producing cereals. Soils of Botswana have high bulk density and on drying easily self compact into a massive hard structure and form surface crusts. At least one tillage operation is necessary to control weeds, improve the soil structure and water infiltration and to obtain good seedbed tilth, crop establishment, and crop stands. The Department of Agricultural research, in collaboration with the Arable Lands Development Program (ALDEP) of the Ministry of Agriculture conducted an agronomic research program on dryland sorghum and maize with emphasis on tillage practices. During three cropping seasons, 1989/91 -1990/91 five tillage practices were evaluated at eight sites in Botswana. In general an early primary tillage followed by a secondary tillage at the time of planting resulted in higher yields than the conventional system in which planting is done after ploughing only once. Soil moisture measurements at two sites showed that grain yields per ha. per mm of water used over a specified interval after planting were highest in plots with an earlier tillage treatment. Deep ripping (down to 50 cm) and planting along the rip line consistently performed poorer than the other tillage systems.

Introduction

The main land divisions of Botswana are Hardveld and Sandveld. Smaller areas of Alluvium and Lacustrine occur superimposed on these land systems. Arable soils of Botswana are inherently infertile (table 1). Most arable soil units are arenosols, luvisols and vertisols. The arenosols and luvisols have a relatively high sand content and are classified as either sands, loamy sands, or sandy loams. These soils become dense and hard when dry. Their clays are easily water dispersible and on wetting, clay increases on the surface resulting in aggregate cementation, crust formation and surface sealing. On the other hand the Vertisols have a relatively high clay content that makes it difficult to plough so that high power tillage operations are required. When dry, the soil is too difficult to shutter.

The most important food crop grown under dryland in Botswana is sorghum. It is the staple food source and the stalks are used as feed for livestock. Although sorghum is fairly well adapted to the arid environment still its production is constrained by low, erratic rainfall, low soil fertility and adverse soil physical properties.

Good tillage is necessary to control weeds, improve soil structure and rainfall infiltration and reduce runoff to obtain good seedbed tilth, crop establishment, and crop stands. On average 4-5 ha. are ploughed per family using tractor, or cattle or donkey teams. The single-furrow mouldboard is the most common implement used. Planting is usually by hand broadcasting, but many farmers use row planting. Many fields are not fenced yet animals are usually not herded until many farmers in the area begin ploughing. In this regard many farmers are unwilling to take advantage of the early rains as their unprotected crops may be destroyed by cattle. Crops are weeded once and sometimes not at all. Crop residues are not returned to the soil but are usually left to be grazed by cattle. The use of fertilizers, rotations and pest control methods are rare.

Not withstanding its benefits, tillage can also increase the rate of organic matter decomposition resulting in deterioration of soil structure and a decline in soil productivity. Therefore management practices for dryland regions must be carefully chosen and in particular focus should be on conservation tillage that can sustain both soil moisture and soil quality. Conservation tillage has been shown to be one of the best ways of conserving soil and water in dryland crop production in the states (Meyers, 1983). In the US most forms of conservation tillage involve the use of crop residues for reducing soil and water losses (Greb et al. 1979). In Botswana residues are grazed to livestock and farmers are reluctant to return them to the soil. Conservation tillage here imply the ability to reduce soil and water losses when compared with the traditional or conventional tillage methods.

Table 1. Average chemical analyses of topsoil for major arable soil groups of Botswana
 

Note: pH in 0.01M CaCl₂, P by Bray 2, CEC by Neutral ammonium acetate, OC by Walkley and Black

Research on tillage was a major component of Dryland Farming Research Scheme (DLFRS) activities in Botswana. The work during the 1971-1974 and 1975-1979 phases of DLFRS are summarised in the review paper prepared for initiating the 1980-1984 phases of its activities (DLFRS, 1980). All the tillage research by DLFRS from 1971-1979 were done at Sebele Research station except for some trials initiated in 1978 on Kalahari sands at Motopi. No attempt were made to evaluate tillage systems and practices across soil types.

Tillage research was not abandoned in the 1980-1984 phase of DLFRS. Results of these activities are summarised in vol 4 of the final report on their 1980-1984 activities (DLFRS, 1985). Research during this phase continued mainly at Sebele and Motopi. Efforts were focused on comparing primary tillage using various implements with and without residues. Measurements were made of tillage and its residual effects on soil physical properties. In its entirety, tillage research by DLFRS was orientated more towards production agronomy than soil and water conservation. Much valuable information resulted from the DLFRS activities on tillage. However, no definite and acceptable recommendations resulted that matched tillage practices to soil and crop needs and the environment, economic, and social conditions of the farming systems in Botswana.

The 1980-1984 phase of DLFRS activities in Botswana overlapped with the strart in 1982 of the farming systems-orientated Agricultural Technology Improvement (ATIP). Tillage research was included as part of the ATIP activities. The project’s on-farm activities were concentrated mainly in the Central and Northeast administrative districts especially in villages in the environs of Mahalapye and Francistown. Initially, diagnostic work was undertaken to identify the farmers resources and perceptions that were constraining the adoption of new technology (ATIP, 1985a, 1985b).

With regard to tillage, the concepts of early ploughing and double ploughing proved to be the simplest and most effective (ATIP 1990a, 1990b). And ATIP was able to identify several options for inclusion in the production system.

In 1988 the Department of Agricultural Research in collaboration with the Arable Lands Development Program initiated the National Tillage Research Program The purpose of this program was to undertake a comprehensive evaluation of the relative merits of available and promising tillage practices over a wide range of soil and climatic conditions. This program was oriented equally towards understanding tillage effects in relation to production agronomy as well as soil and water conservation. A key guiding concept in formulating this program was to evaluate tillage practices for their ability to conserve soil moisture.

This paper presents the studies and results from trials undertaken during three cropping seasons 1988/89, 89/90 and 90/91.
 

Materials and Methods

The experiments were designed to evaluate several tillage systems on cereal yield. The experiment was initiated in 1988 on 8 on-farm and 1 on station (Sebele) location. The location of the experimental sites and the site coding, the soil types and the rainfall for the three growing seasons are presented in table 2.

The tillage plots were laid out in a randomised complete blocks design replicated twice at each site except S1 where there were four replications. At J 1 tillage treatments were combined with or without 15 kg ha^-1 as single super phosphate 10.5 % P in a split plot design. Plot sizes were 15x40 m. At J1 plots were 80 x 15 m. These were split in half to apply the fertilizer treatment. All cultivation was done by tractor. The test crop was Segaolane sorghum. The five tillage treatments were coded as follows:

T1 - Conventional: single mouldboard ploughing on day of planting. Crop row-planted at 75 cm row spacing to obtain 50,000 plants ha^-1± 10 %.

T2 - Double ploughing: Early spring mouldboard ploughing with first rains followed by second ploughing on day of planting. Crop planted at 75 cm row spacing to obtain 50,000 plants ha^-1 ± 10%.

T3 - Deep ripping: Deep ripping to 50 cm on 150 cm centers, as soon as possible after harvest, followed by land shaping and other secondary tillage as appropriate with this system. Crop row-planted along rip lines to obtain 20,000 plants ha^-1 ± 10 %.

T4. Ploughing and cultivation: Early ploughing as in T2 followed by tined cultivation on day of planting. Crops row-planted at 75 cm row spacing to obtain 50,000 plants ha^-1 ± 10 %

T5 - Conventional with wide row spacing: As for T1 but with 150 cm row spacing as in T3. Crop row-planted to obtain 20,000 plants^-1 ± 10 %.

The same treatments were conducted in some of the sites using animal draught power.

In addition to the above, other treatments were added as desired and varied from site to site.
 

Results and discussion

The detailed reports of the results can be found in (Persaud, 1990; Persaud et al., 1991; Persaud et al., 1992). The data for several of the sites were not complete and were discarded and in some cases different sites on the same soil were used in subsequent years. Additional treatments were included in the ANOVA for the 1989/90 and 1990/91 cropping seasons, however only the means of the five core tillage treatments are presented. The overall yields of sorghum were 3-5 times the national average even in years of below average rainfall. In general there was a high variability in yields as indicated by high CVs (tables 3,4&5).

Data for 1988/89 (table 3) shows that sorghum responded differently to tillage treatments from site to site. Generally the double plough and the Plough and cultivate treatments tended to do better than others at many of the sites but were significant at the 5 % probability level at only two of the sites. The Double plough and the Plough and cultivate were also reported to have the highest water use efficiency (data not shown).

The positive response of yield to early inversion tillage observed in 1988-89 season was less apparent in the 1989-90 season. The double plough and plough and cultivate yields at site F1 were higher compared with other treatments. At sites with overall mean yield lower than 110 kg/ha the Double plough and the Plough and cultivate treatments tended to perform better than the conventional. At those sites with overall mean yield greater than 1100 the conventional tended to perform as well as, or better than these two treatments. The deep rip performed poorly compared with the conventional except at F2. In terms of yield, the conventional with wide row spacing compared well with deep rip at most sites while at M3 and M4 it performed better.

Table 2. Location, soil type and rainfall at the experimental sites
 

* =site code, **=data not available

A significant response to tillage was observed at a few sites in 1990/91 experiments. The results show a possible benefit from an early inversion tillage but with no consistency. However several points can be made against an early inversion tillage. One point is whether the cost of this additional tillage operation would be counterbalanced by the yield benefit. Another consideration is whether it would be worthwhile to forego a planting opportunity for an early inversion tillage operation. Overall mean yields for sorghum (table 5) across sites tended to be lower for animal draught compared to those for tractor draught possibly due to the deeper ploughing by tractor which was also more effective in removing weeds.
 

Conclusion

No definite conclusions can be established on the effects of treatments on yield but early inversion tillage in addition to the tillage operation done at planting time appears to have some advantage.
 

References

Table 3: Effect of 5 tillage treatments on sorghum grain yields at 10% moisture in 1988/89 growing season
 

**, *, and NS = respectively, highly significant P (>F) < 0.01, significant P (>F) > 0.01 and < 0.05, or not significant.

Table 4: Effect of five tillage treatments on sorghum grain yields at 10 % moisture in 1989/90 growing season
 

**, *, and NS = respectively, highly significant P (>F) < 0.01, significant P (>F) > 0.01 and < 0.05, or not significant.

Table 5: Effect of five tillage treatments on sorghum grain yields at 10 % moisture in 1990/91 growing season
 

**, *, and NS = respectively, highly significant P (>F) < 0.01, significant P (>F) > 0.01 and < 0.05, or not significant.

 Table 6: Effect of five tillage treatments across sites.
 

Department of Agricultural research Ministry of Agriculture; Private Bag 0033, Gaborone, Botswana