«JM Thamaga-Chitja March 2008 Submitted in fulfilment of the requirements for degree: Doctor of Philosophy (Food Security), African Centre for Food ...»
5.1 Introduction The developed and developing countries have, in recent times, demonstrated a growing interest in improving the livelihoods of smallholder farmers, using organic agriculture, among other approaches and technologies (Duram, 1999). The complexity of organic farming management demands a well-developed knowledge system that promotes biological harmony encompassing biodiversity, biological cycles and soil biological activity, while discouraging the use of off-farm inputs (NOSB, 1995). This complexity is exacerbated by the fact that organic agriculture also encompasses other non-agricultural factors, such as those related to the social, economic and institutional dimensions (Scialabba, 1999). Decision-making for smallholder farmers is characteristically complex because of the close interactions between household and farming decisions (FAO, 2006). Decision-making so complex and can be a big challenge for smallholder farmers in poor countries who, not only have poor access to resources and information, but are also faced with literacy constraints. The need for a decision support tool for organic production is crucial to support both the evaluation of potential for organic farming in South Africa and to support decisions of aspirant organic farmers.
Introduction of new technologies, including organic agriculture, in situations of need, such as in many poor African countries, can be viewed as answers to a wide variety of problems (Freyer et al, 1994). There are many studies that have investigated adoption of different technologies but studies that relate to the adoption and successes or failure of organic agriculture in Africa are not easily available. Nevertheless, it is important that any technology should be appropriate to the context within which it is intended.
This chapter presents the components of a farmer-oriented decision support tool (Appendix C) by presenting and discussing the stages that were followed during the development of the tool.
The analysis of organic farming constraints (Chapter 4), led to the development of a decision support making tool to provide farmers with crucial information to guide organic production decisions. The farmers’ main desire was to improve productivity and prosper in organic farming. However, they needed to establish what was required to achieve this goal. Vigorous engagement with the farmer groups led to their main question being expanded into four sub-problems. These four questions were adopted as the study’s sub-problems. As emphasised earlier, the tool’s development was limited to organic production but recognised that other areas of organic farming, such
as marketing, are important. The production related sub-problems were as follows:
Sub-problem 1: What crops can be grown organically in the three chosen areas based on climatic data?
Sub-problem 2: Do farmers concur that these are the most suitable potential organic crops?
Sub-problem 3: How useful do the farmers find the decision making tool?
Sub-problem 4: What constraints threaten commercial production of the identified crops for these farmers?
It was agreed that the tool needed to be as simple as possible for the end-user.
Consequently, the computer programme used was Microsoft Office Excel (version
2003) instead of a complex programme that would require the user to be well versed in computer usage as this could be a deterrent. Due to the multidisciplinary nature of the study, a team of experts from various agricultural specialisations was consulted at various stages of the model development to verify that the approach and stages of development of the model were sound. The methodology applied during the expert workshops is discussed later in this chapter (results and discussion).
The decision support tool produces a two-page printout. The first page contains the output for high moisture-induced crop diseases. The low moisture-induced crop diseases are listed on the second page. Ideally, both high moisture and low moistureinduced diseases should be on a single page printout. However, the Excel program is not sophisticated enough for this. The model could thus be developed further to suit 55 field conditions, using a higher level of computing and programming with the support of computer programming specialists (Voges, 2006). This can be undertaken in the future. Descriptive headings are used for the output to keep the tool simple. Detailed results of the tool are presented later in this chapter. In chapter six, an application of the decision support tool is presented. The following section discusses how the model, upon which the decision support tool is based, was developed. A presentation of results and discussion follows and concludes this chapter.
5.3 Development of the model
The decision support tool was developed in two main stages subsequent to the FFA process. A desktop exercise using existing primary data for calibrations and the development of the user interface in three steps was the first action undertaken.
Several important assumptions were made in the development of the model and they
• It is assumed that for satisfactory crop growth to take place, minimum climatic conditions have to be met.
• Crop nutrient needs were based on maximum quantity to fulfil the argument that organic nutrient needs (based on manure) are based on the most limiting nutrient.
• It was further assumed that rainfall was a correlate of moisture.
• An assumption that rainfall can be predicted was made.
• Rainfall based moisture was used to predict onset of disease.
These rainfall-related assumptions were based on the confidence from graphical correlation of predicted and observed rainfall values as demonstrated in chapter 5 (figure 5.5). Graphical depictions of this assumption, in Figure 5.5, show a very high correlation, yielding confidence that the assumption made is sound. All farmers largely practice rainfall-dependent agriculture with no effective irrigation. It is acknowledged that humidity and mist may play a role in disease onset, but lack of data on these two factors for the study areas resulted in the use of rainfall as the sole source of moisture data.
Numerous sources of information were consulted to aggregate the relevant information in answering the study’s questions. Table 5.1 summarises the various sources of primary data consulted in a quest to answer the second sub-problem namely; what crops can be grown in the chosen areas?
5.2.1 Stage 1: Selection of climatic data and loading of agro-ecological information per crop into specially created Excel spreadsheets (see Appendix C-decision support tool) The first sub-problem was mainly concerned with what crops could be grown in the
three chosen areas. The following steps were taken to respond to this sub-problem:
• Creation of a manageable list of crops.
• Identification of normal physiological growth conditions for crops on the list.
• Use of various computations to link physiological growth conditions and other data located in different Excel spreadsheets.
The first activity in stage one of the decision support tool development involved creating a list of crops from which to identify suitable crops using the Natural Resource Database from the Department of Agriculture and Environmental Affairs in KwaZulu-Natal (DAEA) (Camp, 1999). Since organic farming is a growing niche market with opportunities for smallholder farmers (Barett et al, 2002), it was important to establish which crops were in demand. Therefore, a list of sought-after organic crops was obtained from Woolworths’ buying division as a market leader in retailing a wide range of organic vegetables, fruit and dry products in South Africa (Ferreira, 2004).
Figure 5.1 illustrates a decision support process employed to respond to the first subproblem.
A consequential series of questions was posed for each crop to assess if agro-ecological conditions met crop requirements. The four conditions were set as the minimum and essential requirements for the normal physiology of plants (Bidwell, 1974, pg. 3-4). These are the annual rainfall (mm); the length of the rainy season (days/annum); mean annual temperature (minimum and maximum) and photoperiod, all of which were sourced from FAO’s (2003) Ecocrop website. The first two questions in figure 5.1 were related to water requirements because water is a critical element for plant growth (Bidwell, 1974).
• the length of the rainy season must be equal to or longer than the growth cycle of each crop so that the crop would have enough water during its physiological development;
• the minimum growth temperatures required by each crop had to be fulfilled;
• there had to be adequate sunshine during the photoperiod. A positive answer to all four questions in Figure 5.1 meant that the crop could grow, given optimal conditions. The output at this stage was to list all crops that had potential to grow in the chosen area.
The main assumption in this sub-problem was that for satisfactory crop growth to take place, minimum climatic conditions had to be met. However, it is accepted that a certain level of growth that will lead to a certain level of yield will take place but the decision support tool cannot quantify this variation. This is because mean values of climatic parameters were applied. Therefore, it is accepted that climatic value below the mean may lead to some crops being rejected by the model as not being suitable.
5.3.2 Stage 2: Loading of crop nutrition requirements
Agrochemicals in conventional agriculture have two main roles, which are disease and pest control and crop nutrient supply. On the contrary, organic farming, according to the OFRF (2001), is the exclusion of all external inputs of agrochemicals (pesticides and fertilisers) in agricultural production and related activities. This definition underpins the second sub-problem, which questions whether farmers can grow the selected crops organically. In the case of smallholder farmers (including groups participating in this study), livestock manure is the most common source of soil and crop nourishment (Kuepper, 2003; van Averbeke & Yoganath, 2003).
60 Various government extension sources were consulted to obtain soil nutrient requirements for each crop, including the Guide for Extension Officers (Smith, 1998);
the Vegetable Production Manual (Alleman & Young, 2001) and the Fruit Production Manual (Sheard & Jele, 2002). Other sources including Chadha and Shimansky (1999), Salunke and Kadam (1998) and Salunke & Kadam (1995) were also consulted. Lastly, telephonic communication was held with vegetable and fruit specialists and research papers were consulted for crop nutrition information.
Crop nutrient removal norms also indicate how the soil would be depleted further if no soil nourishment plan is in place. Hygrothech’s (2005) vegetable production guide was used to obtain vegetable nutrient removal norms, which were used to calculate NPK requirements. Furthermore, Manson (2006), Conradie (2005), Kilby (1998), Salunke & Kadam (1995), Agata (1992), Askew (1992) and Kabeerathumma (1992) were consulted to obtain the nutrient withdrawal norms of vegetables, fruit, root and maize crops using equation 1.
The number of wheelbarrow loads of manure required to meet the removal norms for each crop was calculated for N, P and K in turn, yielding three quantities based on equation 2. A load of a wheelbarrow is assumed to be 75kg (van Averbeke & Yogananth)
These quantities of wheelbarrows were then compared for N, P and K. The highest number of wheelbarrow loads for N, P or K was then chosen (see equation 3) as the required manure input per crop, ensuring that all nutrient requirements are met.
It was also important to estimate the amount of manure that could be produced by one animal (beast, sheep or goat). The number of animals is directly related to the availability of manure and thus crop yields. Equation 4 illustrates a series of formulae used to calculate the amount of manure from one animal, the area that can be fertilised and the possible yield for each crop per annum based on one grazing beast. Using this element of the model, it is possible to evaluate available manure or potential for manure production based on the number of animals accessible to the farmers. Manure was calculated using the formula in equations 4, 5 and 6. The assumptions made in equations 4, 5, and 6 were based on USAD (1996) & van Averbeke & Yoganath, (2003).
It was further assumed that the most limiting nutrient between NPK was used as a basis for the calculation of manure requirements. The number of wheelbarrows of manure was based on this nutrient. However, it was expected that some level of soil nutrients will be available, even though nutrition may not be optimum. The decision support tool indicated a very large number of wheelbarrows of manure which cannot be practically applied due to the large volumes, health contamination and nutrient imbalances that may be caused by large applications of certain nutrients. It is doubtful that even the most astute management, including crop rotations, can overcome such nutrient accumulation due to the sheer volumes of manure indicated by the model. It is critical to note that other organic nutrient provision methods, such
5.3.3 Stage 3: Loading of crop disease information The purpose of this third stage was to ascertain if the climatic conditions of the study areas were conducive or detrimental to organic farming. Both temperature and moisture are important for disease occurrence (Agrios, 2004). It was not necessary to program temperature into the decision support tool because summer temperatures are conducive for onset of disease. However, moisture plays a critical role in disease setting (e.g. spore germination and penetration) and disease spreading. As a result, moisture was deemed the single most important indicator of disease risk in this study.