search
for
 About Bioline  All Journals  Testimonials  Membership  News  Donations


African Crop Science Journal
African Crop Science Society
ISSN: 1021-9730 EISSN: 2072-6589
Vol. 4, Num. 3, 1996, pp. 305-313
African Crop Science Journal,Vol. 4. No. 3, pp. 305-313, 1996

Simulating the potential effects of reducing runoff and planting date on sorghum yields in Botswana*

C. K. Davis and R. L. Vanderlip^+

Department of Agronomy, Kansas State University, Manhattan, KS 66506 USA

(Received 21 August, 1995; accepted 9 July, 1996)


Code Number: CS96070
Sizes of Files:
    Text: 23.5K
    Graphics: Line drawings (gif) - 30K
              Tables (jpg) - 370K   

ABSTRACT

Extremely variable climatic conditions cause Botswana farmers to plant after each significant rain. Various management techniques, such as double ploughing, have been proposed to reduce runoff and increase sorghum yields. SORKAM, a computer growth model for sorghum, was used to estimate the potential effects of the reduced runoff effect of double ploughing on sorghum yields and yield stability for four locations (six soils) and climatic data for 40 years. Simulated reduced runoff increased sorghum yields up to 32% on a soil with about 50% clay. The smallest yield increase, 1%, was on a soil with 90% sand. Reduced runoff increased yields most on soils that have higher water holding capacities, allowing greater water storage. Although double ploughing, in general, would delay the first planting and early planting had higher average yields than late planting, no date of planting was optimum and ranges in yield were great for all dates. Cumulative frequency distributions showed that the advantage of reduced runoff not only was dependent on location (water holding capacity) but also on yield level. At all locations at low yield levels, reduced runoff either decreased yield or inceased the incidence of total crop failure.

Key Words: Double ploughing, Sorghum bicolor, water holding capacity

RESUME

Les conditions climatiques etant extrement variables permettent aux agriculteurs du Botswana de planter apres chaque pluie abondante. Les techniques des gestions variees, par exemple, le systeme de labourer deux fois a ete propose pour reduire la carence et l'augmentation en production de sorgho. Le modele de l'augmentation du sorgho qui est connu sous le nom de SORKAM, etait utilise pour estimer les effets du potentiel de la reduction de l'effet de la carence du systeme de labourer deux fois sur la stabilite de production du sorgho pour quatre locations (six sols) et les donnees des climats pour une duree de 40 ans. La reduction de carence simultanee a augmentee la production du sorgho a 32% et sur le sol a peu pres de 50% d'argile. La plus petite augmentation de production sur les sols ont de grandes capacites de conserver de l'eau, qui occasionne le stockage d'une grande quantite d'eau. En general, malgre le double systeme de labourer ce qui retarde le premiere saison de planter tot a une moyenne des productions par rapport au planter tot, aucune date de planter n'etait optimum et les variations en production etaient superieures par rapport a toutes les dates. les distributions des frequences cumulatives ont montrees que l'avantage de carence reduite n'etait par seulement dependante en location (capacite de conserver de l'eau), Mais aussi au niveau de production. Dans toutes les locations aux bas niveau de production, a reduit la carence soit diminuer, l'incidence totale de l'echec de la plante.

Mots Cles: Deux cultures, sorgho bicolor, capacite de conserver de l'eau

INTRODUCTION

Sorghum [Sorghum bicolor (L.) Moench] is the primary crop grown by subsistence farmers in Botswana. Although rainfall in eastern Botswana could produce yields over 2000 kg ha^-1 in most years (Lightfoot, 1979) and research shows that yields can exceed 4000 kg ha^-1 (DLFRS, 1978), the average sorghum production is about 160 kg ha^-1 harvested (Anon., 1985b).

Severe moisture stress occurs because of high potential evaporation (1500 to 2000 mm compared to an average of 500 mm precipitation) and high runoff (as much as 70%). In addition, rainfall is unpredictable, and long dry periods occur (Lightfoot, 1981). The sandy soils are prone to crusting and have low infiltration rates (Jones, 1987c).

In Botswana, 85 percent of the farmers use traditional planting methods: they broadcast the seed on the soil surface and incorporate by ploughing (Lightfoot, 1981). The rainy season coincides with the growing season, lasting from October to May. Farmers have adapted to the stringent conditions by planting after each rain greater than 10 mm (planting rain); this ensures adequate soil moisture for germination and establishment, reduces risk by spreading out the planting dates, and takes advantage of different climatic conditions. Schouwenaars (1988), using a soil water balance model developed for maize [Zea mays (L.)] and sorghum for southern Mozambique, concluded that while the practice of multiple plantings over a relatively long time period might not maximise production over a number of years, it was probably the best way to secure the availability of a minimum food supply every year.

Whiteman (1975) reported that crop yields in Botswana fluctuate with seasonal rainfall, with failure or low yields in about two years of every five. He suggested that since moisture is the most limiting factor, ways should be found to maximise infiltration, improve conservation of stored moisture, and increase efficiency of moisture use. His results showed sorghum yields could be doubled after fallow (270 kg ha^-1 compared to 128 kg ha^-1 without fallow). However, Jones (1987c) suggested that because of the small water holding capacities of Botswana soils, whole season fallow as suggested by Whiteman (1975) is not necessary but short periods at the beginning or ending of the growing season should be used to accumulate soil moisture for the succeeding crop. This is indirectly supported by the data from Whiteman (1975) in which the amount of moisture stored was independent of rainfall over the fallow season and was over 90% of the storage capacity even after two below average rainfall years.

Farming systems teams are working with farmers in Botswana to find more productive agricultural methods. Since rainfall seems to be sufficient for crops, methods to retain precipitation, i.e., water harvesting are sought. Double ploughing, an extra preplant tillage operation in the fall or early spring, seems promising. Reported effects of double ploughing are improved soil moisture content, reduced stubble, fewer weeds, reduced compaction, decreased crusting, lowered runoff, increased infiltration and storage of preplant rains, easier planting, and higher yields (Anon., 1983, 1984, 1985a; ATIP, 1985, 1986).

Double ploughing experiments at Francistown in 1984-85 resulted in a 262 kg ha^-1 increase in grain yield over single ploughing (traditional method) (ATIP, 1985). Similar experiments in the following year showed that double ploughing improved grain yields 109% over single ploughing. In other tests, average yields from four double-ploughed hectares met the grain requirements of an average seven person family; average yields from five hectares traditionally planted did not meet those requirements (G. Heinrich, Personnal Communication).

Early planting (November) may also increase yields by as much as 68% over late planting (late December) (Livingstone, 1979). Late planting is discouraged because plants do not reach physiological maturity and short season varieties are unavailable (Anon., 1984). Early planting alone is not the key; it is important to begin planting early and keep planting throughout the growing season to reduce risks associated with a single planting date.

One method of assisting extension or farming systems teams is to use crop growth models to simulate growth and development of a crop and predict the yield which could be produced by various management strategies under local conditions. Models have been used for north-west Queensland, Australia to determine the effectiveness of water harvesting strategies on sorghum yields (Clewett, 1985). With models, the most efficient or profitable strategies can be chosen by determining for each practice the distribution of yields and returns associated with the climatic conditions at that location (Worman et al., 1988).

SORKAM, a physiological crop growth model for sorghum (Rosenthal et al., 1989), simulates daily growth and development, computes light interception, soil water, leaf area, and dry matter accumulation and partitioning. SORKAM was used to simulate yields from traditional plantings in Botswana; to evaluate the potential yield increased from the reduced runoff component of double-ploughing; and to determine where and when the greatest benefits might occur.

MATERIALS AND METHODS

SORKAM (Rosenthal et al., 1989) simulations were run for forty years of weather data for four locations (Gaborone, Mahalapye, Francistown, and Maun) with six soil types, making simulations relevant for much of the sorghum growing region of the country. Daily climatic data, and sowing, plant, and soil data are required to run the model. WGEN (Richardson and Wright, 1984) was used to generate unavailable climatic data. WGEN created a file of daily temperature, radiation, and precipitation data with a range and distribution similar to those of actual data at each location. Hybrid 'RS610' was simulated instead of the local Segaolane sorghum variety for which plant input values were unavailable. RS610 is an early, photoperiod-insensitive hybrid with 15 leaves. Plant population was set at 25,000 plants ha^-1. Rees (1986) concluded that while production might be maximised with populations of 40,000 to 80,000, plants ha^-1 a lower range of 10,000 to 20,000 plants ha^-1 would show a reduced frequency of crop failure and greater yield stability. Jones (1987a, b) concluded that plant population for maximum yield ranged from 25,000 to 69,000 plants ha^-1 depending on rainfall. Soils information came from the SCS report on Botswana soils (Soil Conservation Service, 1987). Data from one soil at Gaborone and Maun and two at Francistown and Mahalapye were used (Table 1). Initial available soil water was set to zero for all depths since no rain falls from May until planting, so the soil profile would be dry. Data for maximum available water came from the soils report (Soil Conservation Service, 1987). These inputs were held constant for all runs at a location.

Hydrologic curve numbers from the Soil Conservation Service are used in SORKAM to govern runoff; these numbers represent runoff classes, they are not percentages. To simulate the differences in runoff between single and double ploughing, two values were used. To represent single ploughing, the highest curve number, 91, was used; soils in Botswana have extremely high runoff (Anon., 1984; Jones, 1987c). A low curve number, 72, was chosen for the double ploughing simulations to represent improved soil conditions.

Two planting opportunities were simulated for each year since double-ploughing would likely replace the first planting. For the first planting opportunity, the sowing date was the day after the first planting rain of the growing season (rainfall < 10 mm in one day). Sowing date for the second planting opportunity was the day after the second planting rain of the season (rain > 10 mm falling more than 3 days after the first planting rain but before 1 January).

Therefore, for each soil and year, four SORKAM runs were made:

- High runoff (Traditional ploughing) - first planting
- High runoff (Traditional ploughing) - second planting
- Low runoff (Double ploughing) - first planting
- Low runoff (Double ploughing) - second planting

After running the simulations, means, standard deviations, maximum and minimum values, and coefficients of variation (Snedecor, 1956) were calculated for yield and growing season rainfall. No economic analyses were conducted. Cumulative frequency distributions of yield and the relationships of yield with rainfall and sowing date were examined for each location.

RESULTS AND DISCUSSION

At each location, 3 of the 40 years simulated had no second plantings because a second planting rain did not occur before 1 January. In addition, Maun had 1 year with no planting, and a total of 6 years with no second planting.

Average yields were 2000 to 3000 kg ha^-1, the level many feel Botswana is capable of obtaining (DLFRS, 1978; Lightfoot, 1979). These are much higher than average yields currently produced, but SORKAM assumes no problems with soil fertility, diseases, pests, weeds, or uneven stands, all of which limit production in Botswana. The use of a hybrid instead of the local variety also may partly account for higher yields. In addition, average actual yields were from a drought period in Botswana. Predicted yields from dry years were under 1000 kg ha^-1 and, therefore, were not unrealistic.

Two planting dates were simulated to determine what effect planting date had on yields. Statistical analysis showed no significant effect of date at any location although yields from early planting were slightly higher than late planting for all sites except Gaborone (Table 2). No date of planting was optimum (Fig. 1), thus the simulations agree with farmer practices. In contrast, grain yield increased with increasing rainfall during the crop growth cycle (Fig. 2). From the regression equation we can compute the X intercept which suggests that after approximately 87 mm of rainfall the water use efficiency (WUE) is 10.2 kg of grain per mm of rainfall received. This is much higher than the value of 2.89 reported from field trials with Segoalane at 25,000 plants ha^-1 (Jones, 1987a). His results were based on a fixed (August-March) season as compared to the simulated growing season in the present paper. He also found an X intercept of less than 3 mm which appears to be an unrealistically low value for the start of grain production. The WUE value of 10.2 compares favourably with that reported by Seetharama et al. (1984) of 13.0 for the rainy season, 11.8 for the post-rainy season, and 8.6 for the summer season in India. They did report WUE values for the post-rainy season as low at 2.8 kg/mm rainfall.

Average simulated yields show that reducing runoff increased sorghum yields (Fig. 3). The greatest increase in mean yields (32%) was at Shoshong (the soil with the highest clay content, Table 1) and the smallest (1%) at Marapong (the soil with the highest sand content). Statistical analysis showed significant differences in effect of rainfall only among the treatments at Gaborone, Maun, and Shoshong, locations with the deepest soils of the six used. Thus, reduced runoff generally had the same yield effect as increased rainfall. Differences in runoff were approximately 20 to 40 mm or about 6-12% of the growing season rainfall (Table 2). Land and Water Management Project (1990) reported preliminary results of 13-14% runoff from crop land in Botswana. They also found that runoff from double ploughed plots was negligible compared to bare fallow. These results support our use of low runoff to represent one aspect of double ploughing.

Like mean yields, cumulative frequencies showed that reduced runoff had a greater effect on yields than date of planting (data not shown). Comparison of the two simulation treatments that most closely approximate the traditional (early planted, high run-off) and double ploughed (late planted, low run-off) management systems are shown in Figure 4. As frequently occurs, the cumulative frequency distributions provide much more information than do mean values. In Figure 3 reduced runoff showed increased yields at each location except Marapong. However, in looking at the cumulative frequencies in Figure 4, at each location including Marapong, on the lower end of the yield distribution, reduced runoff reduced yields or increased the frequency of complete failure. In addition, the cumulative frequencies show considerable differences among locations in the yield level at which reduced runoff appears to be advantageous and the consistency of its effect. For example, the cross over point appeared to be at less an 1,000 kg ha^-1 at Shoshong and over 2,000 kg ha^-1 at Mathangwane. The consistency of expected yield increase also was very different among locations. For example, at Makwate, although the average yield was higher for reduced runoff, there was not very much difference over the entire yield range. In contrast at Shoshong at yield levels above 1,000 kg ha^-1 reduced runoff increased yields and fairly consistently did so over the remainder of the yield range. At Gaborone the effect was primarily in the centre of the yield distribution with little effect at either low or high yield levels. The effects at Mathangwane and Maun were intermediate in that no large yield increases occurred but they were fairly consistent from the midportion of the yield distribution upward. These results illustrate the ability of a crop growth model to integrate the major limitations in a cropping environment, in this case, the rainfall pattern and the water holding capacity of the soil. At Marapong, with only 3.1 cm of water holding capacity (Table 1) the likelihood of increased infiltration of water was not matched with an unfilled soil profile. Likewise at low yield levels the effect of decreased runoff is more than offset by the loss of the rainfall which followed the first ploughing and the delayed planting which generally resulted in the crop running out of moisture at an earlier developmental stage.

These simulation results suggest that the results of double ploughing which are a result of runoff differences would be extremely variable from location to location and from year to year and are consistent with the conclusions of Sigwele and Norman (1993). They concluded that the double ploughing technology, among other things, should be on soils that are relatively deep with a good water holding capacity, that the traditional plough planting operation may be the best under certain circumstances, and that farmers need to be provided with a series of options rather than a ÒblanketÓ recommendation.

REFERENCES

Anon. 1983. Annual Report for the Division of Arable Crops Research 1981-82. Ministry of Agriculture. Gaborone, Botswana.

Anon. 1984. Annual Report for the Division of Arable Crops Research 1982-83. Ministry of Agriculture. Gaborone, Botswana.

Anon. 1985a. Annual Report for the Division of Arable Crops Research 1983-84. Ministry of Agriculture. Gaborone, Botswana.

Anon. 1985b. 1985 Botswana Agricultural Statistics. Planning and Statistics, Ministry of Agriculture. Central Statistics Office, Ministry of Finance and Development Planning. Gaborone, Botswana.

ATIP. 1985. Agricultural Technology Improvement Project. Annual report No. 3. Ministry of Agriculture. Gaborone, Botswana.

ATIP. 1986. Agricultural Technology Improvement Project. Annual report No. 4. Ministry of Agriculture. Gaborone, Botswana.

Clewett, J. F. 1985. Shallow storage irrigation for sorghum production in north-west Queensland. Queensland Department of Primary Industries, No. QB85002.

DLFRS. 1978. British Dryland Farming Research Scheme, Phase II. Third administrative report. Agricultural Research Station, Botswana. Government Printer. Gaborone, Botswana.

Jones, M.J. 1987a. Plant population, rainfall and sorghum production in Botswana I. Results of experiment station trials. Experimental Agriculture 23:335-347.

Jones, M.J. 1987b. Plant population, rainfall and sorghum production in Botswana II. Development of farmer recommendations. Experimental Agriculture 23:349-356.

Jones, M.J. 1987c. Soil water and crop production in Botswana. Soil Use and Management 3:74-79.

Land and Water Management Project. 1990. Second Annual Report. January 1990. SACCAR, Gaborone, Botswana.

Lightfoot, C.W.F. (Ed.). 1979. Arable Crops Research in Botswana. Proceedings of a meeting held at the Botswana Agricultural College, 20-21 September 1979. Gaborone, Botswana.

Lightfoot, C.W.F. 1981. An evaluation of improved and traditional technologies: Broadcast planting in perspective. Ministry of Agriculture, Gaborone, Botswana.

Livingstone, I. 1979. Botswana: Preliminary country study. Norwich, England: Overseas Development Group, University of East Anglia.

Rees, D.J. 1986. Crop growth development and yield in semi-arid conditions in Botswana. I. The effects of population density and row spacing on Sorghum bicolor. Experimental Agriculture 22:153-167.

Richardson, C.W. and Wright, D.A. 1984. WGEN: A model for generating daily weather variables. USDA, ARS, ARS-8, 83 pp.

Rosenthal, W.D., Vanderlip, R.L., Jackson, B.S. and Arkin, G.F. 1989. SORKAM: a grain sorghum crop growth model. Research Center Program and Model Documentation, MP-1669. Texas Agric. Expt. Stn., College Station, TX.

Schouwenaars, J.M. 1988. Rainfall irregularity and sowing strategies in southern Mozambique. Agricultural Water Management 13:49-64.

Seetharama, N., Mahalakshmi, V., Bidinger, F.R. and Singh, S. 1984. Response of sorghum and pearl millet to drought stress in semi-arid India. In: Agri meteorology of Sorghum and Millet in the Semi-arid tropics. Proceedings of the International Symposium, 15-20 November, 1982. Virmani, S.M. and Sivakumar, M.V.K. (Eds.). ICRISAT Center, India. Patancheru, A.P. 502324 INDIA.

Sigwele, H.K. and Norman, D.W. 1993. Rural development in Botswana: A case study. Staff paper, Department of Agricultural Economics, Kansas State University, Manhattan, KS 66506.

Snedecor, G.W. 1956. Statistical Methods. The Iowa State College Press, Ames. pp. 199-202.

Soil Conservation Service-USDA. 1987. Final report presented to ATIP USAID. Gaborone, Botswana.

Whiteman, P.T.S. 1975. Moisture conservation by fallowing in Botswana. Experimental Agriculture 11:305-314.

Worman, F.O., Biere, A.W., Hooker, W.L., Vanderlip, R.L. and Kanemasu, E.T. 1988. Simulation analysis using physiological crop response models: alternative cropping strategies for southwest Kansas. Kansas Agricultural Experimental Station Bulletin. 653. Manhattan, KS.

Copyright 1996 The African Crop Science Society


The following images related to this document are available:

Photo images

[cs96070e.jpg] [cs96070f.jpg]

Line drawing images

[cs96070b.gif] [cs96070c.gif] [cs96070a.gif] [cs96070d.gif]
Home Faq Resources Email Bioline
© Bioline International, 1989 - 2022, Site last up-dated on 11-May-2022.
Site created and maintained by the Reference Center on Environmental Information, CRIA, Brazil
System hosted by the Internet Data Center of Rede Nacional de Ensino e Pesquisa, RNP, Brazil