Climate variability and related disasters can be mitigated by temporary or permanent protective measures or by avoidance strategies that try to escape the peak values or their consequences. These are all aspects of preparedness strategies. We have indicated above that i. heavy moisture flows or the lack of water, ii. changing heat flows and related temperatures, iii. cropping seasons' climate distributions are the meteorological/climatological factors we should particularly deal with in this paper on traditional knowledge and indigenous technologies that mitigate consequences of climate variability in LEISA farming systems.
The IPCC reports clearly indicate rainfall variability and related disasters as the single most determining factor endangering agricultural production in developing countries. Drought being already a serious threat, indications for longer dry spells in rainy seasons and longer sequences or higher frequencies of abnormal rainfall seasons, with respect to total rainfall and rainfall distribution, make ways of coping with drought situations even more important.
As early as 1986 the FAO/UNEP/UNESCO/WMO Interagency Group on Agricultural Biometeorology had Orev publishing his "Practical Handbook on Desert Range Improvement Techniques", containing two long chapters devoted to the problems of mobilizing, managing and utilizing water resources for local technicians in local agro-pastoral populations in the drier parts of Africa, starting from local experience. Most recently, Das (2001) has reviewed examples in which prosperity of districts and villages in India were directly related to preservation of traditional water harvesting methods and technologies of the use of underground water. A related technology of which also IPCC advocates more intensive use is that of water impoundment, surface storage for later use. This is for example contemplated in Indonesia to make the country again self-sufficient in rice production (Syarifudin Karama, personal communication, 1997), which is at present becoming increasingly lower (Paltridge and Ma'shum, 2002), and in El Salvador after the most recent ENSO triggered drought (Zimmerman, 2001). In Sri Lanka the traditional so called "bethma" practice combines such reservoirs with temporary land redistribution and sometimes field rotation, and attempts are made to revitalise this old practice (MOST/CIRAN, running database (1)).
In Niger, traditional planting pits were improved by making them into water collecting reservoirs imitating part of a soil improvement technology traditionally used in other parts of the country and in Burkina Faso (MOST/CIRAN, running database (2)). From Burkina Faso, it has most recently been reported that villages that adopted land reclamation techniques such as this pitting through crusted soils, filling the pits with manure and water, have seen crop yields rise by 60%, while villages that did not adopt these techniques realized much smaller gains in crop yields under very recent rainfall increases (Reij, cited in Katz, 2002). In north Nigeria small pits in sandy soil are filled with manure for keeping transplanted tree seedlings wet after the first rains. This is tried in China by stony structures in pits, diminishing soil evaporation. Permaculture, water harvesting and infiltration pits, together with the use of drought tolerant crops, have been more recently extended in Zimbabwe, particularly by women, with the help of NGOs, in reply to the recurrent droughts (Shumba, 2001). In semi-arid Nigeria water-harvesting constructions with gutters and bonds are traditionally used around Cassava plots. Again for West Africa, Slikkerveer (1999) mentions a project case study of the successful re-introduction of indigenous "demi-lunes" for better water harvesting. This method was also successfully used in Sudan by the TTMI-project for tree establishment in an arid area near the White Nile (Adil Ahmed Abdalla, personal communication). The earlier example in Niger and these latter two examples further demonstrate the significance of integration of indigenous knowledge and practices in development co-operation projects aiming at increasing resilience (e.g. LEISA, 2001; Stigter and Ng'ang'a, 2001).
Traditional methods and farmer innovations of using occult precipitation under very dry conditions have been dealt with by Acosta Baladon (1995). Further evidence that many of the current traditional adaptation strategies with agrometeorological components also hold for the situations of increasing climate variability is the following quotation from Lin Erda in Zheng et al. (2001) on future measures in China:
"the response strategies include changing the land topography to reduce run off, improve water uptake and reduce wind erosion, introducing artificial systems to improve water availability and to control soil erosion, changing farming practices to conserve soil moisture and nutrients, changing farm operations timing to fit new climatic conditions and using different crops or varieties to match variations in the water supply and temperature conditions. (...) In the course of time new technologies may have to be developed to cope with anticipated impacts and to reduce the costs of adaptation".
It is of course not always a(n) (increasing) variability of climate leading to innovative water use. Changes in cultivation due to population pressure, such as being forced to use sloping land prone to water erosion (Ong et al., 1996; Zheng and Tuo, 2000), as well as income and market considerations have also led to extending or replacing old practices and using new practices that at the same time increase the resilience of the farming systems (Tchawa, 2000) and protection against consequences of drought (Nasr et al., 2000).
As to floods, the technological literature is less abundant and the solutions have most often little to do with agrometeorology as such. For example, traditional drainage ditches and tunnels have been reported from wheat fields in China (Cheng Yanian, personal communication, 2000). In some cases evaporation and occasionally soil conservation and shading by water absorbing trees play a role. Mitigation by reforestation is an often-mentioned aspect (e.g. MOST/CIRAN, running database (3)). However, preparedness and post-disaster measures are more often referred to because, large infrastructural measures apart, there are few ways to counteract serious floods (e.g. Berg et al., 2001).
As to mechanical impacts of rain and hail, although they are forecasted by IPCC to increase in several regions during peak rainfall, we have not found any examples in the literature dealing with increasing protection attempts. The usual protection of crops and soil by the cover from trees, bushes, crops, crop residues left standing, and grass cover and/or mulching will be increasingly necessary, where these problems of mechanical damage of crops and soil are most serious, depending on the specific crops and soils concerned (Stigter, 1994). Classic work from China reviews various traditional adaptations after serious hail damage, assisting plant recovering, through management and compensation measures, or planting follow-up crops in accordance with the length of the remaining growing season (SAAS, 1977). Rivero Vega (2002) reviews other evidence on traditional adaptation measures for hail protection. Such technologies are applicable elsewhere when such damages are increasing (like in India: V.R.K. Murthy, 2002).
Even small changes in the frequency of extreme temperature events may have disproportionate effects. Salinger et al. (2000) mention the life cycle of perennial plants and the stability of forage supplies as well as the balance between temperature and sub-tropical species as examples. It appears that response farming, as we will deal with it in Section 3.3, should not only be considered with respect to fitting the cropping seasons to variable rainfall patterns, but also for fitting it to variable temperature patterns (Van Viet, 2001). This shows that heat is another important factor to be considered in strategies to cope with climate variability. In this case study (Van Viet, 2001), using seasonal temperature forecasting, recommendations could be given on planting date or a combination of planting date and variety, to make sure that rice was flowering in decades for which the required optimal temperatures had been forecasted. Contemporary scientific knowledge has taken over from traditional knowledge here. For example, the detailed knowledge available, as reviewed, on the influence of temperature, temperature extremes and temperature distributions on growth, development and yield of rice (Salinger et al., 1997) makes this possible. Temperature may be involved in flowering peaks of plants used as a traditional forecast for monsoon arrival (Anonymous, 2001).
Farmers near Beijing adapted their sayings on the best seeding time since the 1980s, because of the warming of which they observe the agronomical consequences. So also traditional weather lore may change. Where temperature is a limiting factor to photosynthesis, traditional farmers may react to cooling/warming by changing their cropping system. This is exemplified by the North China Plain, where originally a change from double cropping to more traditional intercropping of early maize with late wheat took place. In southern parts intercropping, that gave higher degree-days for maize, was after a decade again replaced by double cropping, while in cooler mountainous areas and further North the intercropping was kept (Zheng et al., internal publications in the 1990s). However, in many other cases the protection will again very likely not change with any increase in variability of temperature. We are then back to the relevant examples of microclimate management and manipulation (Stigter, 1988,1994), among which there is the classic example of too severe heat flow/temperature modification, by traditional grass mulching against water erosion, leading to subsequent death of young tea (Othieno et al., 1985). Another example of this kind is the traditional furrow sowing of winter wheat in northern China, giving stronger seedlings less suffering from winter damage, due to more soil moisture and higher temperatures in the furrows (Zheng et al., internal publications in the 1980s).
Protection of crops against hot air by shelterbelts was reported by Onyewotu et al. (1998) for reclaiming a desertified area under highly variable climate conditions (Onyewotu et al., 2003). However, it is indeed in traditional parkland agroforestry and other stabilising intensive management of scattered or clumped or alleyed trees that such risk management may be most efficiently found when risks increase (Arnold and Dewees, 1998; Boffa, 1999; Mungai et al., 2001; Onyewotu et al., 2003). It is generally accepted that the weather conditions that create the infamous drought and flammable forests under Indonesian conditions are quite natural, even when their frequency and intensity have increased. However, the factors that have turned this into a disaster, are man-made because most fires are deliberately lit for various reasons. They are due to deliberate policies of non-preparedness and inaction in the face of warnings of extreme fire dangers (Byron and Shepherd, 1998). That is why the appropriate policy environments occur in the B Domain in Figure 1. With the appropriate policies in place, preparedness using meteorological forecasts for grading fire danger has been shown to be a good solution under highly varying conditions (WMO, 1993).
3.3. FITTING CROPPING PERIODS TO THE VARYING SEASONS
The oldest way of coping with climate variability is trying to fit cropping to the ongoing season, using in risk management any possible indigenous forecasts on its behaviour or adapting to what is experienced in the ongoing season. Flexibility and resilience of farming systems with respect to rates of change is a recurrent factor in such attempts. There are ample examples of permanent, slow and fast traditional adaptations to seasonal variability for reasons of risk management and food security (e.g. Bunting, 1975; Stewart, 1988; Blench, 1999; Clemens and Nashrullah, 1999; Gadgil et al., 2000). In fact, these adaptations may be seen as the oldest examples of response farming in the most direct meaning of the words. However, there are no expectations of improvement of these traditional "fitting" methods per se under the presently fast changing conditions. Their blending with more scientific mete-orological/climatological and agronomical/breeding approaches appears the only way forwards (Blench, 1999; Gadgil et al., 2000), as also implied by the end-to-end scheme of Figure 1.
Stewart (1988, 1991) defined response farming in a more limited way, with respect to adapting cropping to the ongoing rainy season by guidance of agronomical operations, using experiences of the past, preferably from interpretations of meteorological rainfall records with support from traditional expert knowledge where available. Given the indications for increasing variability and change of the climate in terms of rainfall, this will have to be adapted to those new conditions, limiting the period in the past over which the experience can be used (Ati et al., 2002). In agrometeorology, pilot projects with the use of on-line agrometeorological information for farmers to respond to, successfully exist already for two decades in West Africa (e.g. Traore et al., 1992; Diarra, 2001). When such changes in definitions of response farming are accepted, it is only a little step to include other parameters like temperature, a possibility earlier mentioned above in Section 3.2, due to better probabilistic seasonal forecasting techniques (Van Viet, 2001).
The situation described above also has another policies related face. Blench and Marriage (1998) have noted that in rain-fed farming areas of eastern and southern Africa, governments and development projects have encouraged high-input, high-risk strategies such as planting hybrid maize instead of sorghum and millet. This, although long experience of uncertainty about weather patterns had induced farmers to develop complex cultivar mixtures to ensure yields under all conditions. The effects of the prolonged drought of the early nineties could have been less, if the risks had been spread across a range of crops with greater tolerance of low-rainfall regimes, as that had been traditionally done. In another example, the dominance of a few seed companies combined with commercial pressure on farmers and an extremely negative attitude to "old" crops and open-pollinated varieties, as well as the replacement of many traditional livestock breeds with "modern" breeds, has massively increased small farmers' vulnerability to climate shock events. Because the high risks under adverse conditions are more important for poor farmers than the opportunities of better years, here the right policy environment is again surfacing as a necessary condition for services towards sustainable livelihood systems (Blench and Marriage, 1998).
In an even more recent paper, Blench (1999) has noted that multi-lateral agencies are urging that climate forecasts be made available to small-scale farmers. Disaster preparedness strategies, both of governments and NGOs, have begun to take account of such forecasts, and there is considerable interest in assigning them an economic value. However, field studies of the impact of recent forecasts in southern Africa suggest that there is a considerable gap between the information needed by small-scale farmers and that provided by the meteorological services (Blench, 1999). This was confirmed by investigating the role of intermediaries such as Agricultural Demonstrators in Botswana (Stigter, 2002b) and Provincial Agrometeorologists in Vietnam (Stigter, 2002a). Risk-aversion strategies in LEISA production systems do pose a problem for adapting forecast information. Low-income farmers are interested in a broader range of characteristics of precipitation, notably: total rainfall, patchiness of rainfall, intensity, starting date, distribution of rainfall, end of the rains and prospects for dry spells and their length (Blench, 1999; Ati et al., 2002). The use of this information then has to be adapted to local soils and topography.
It is exactly here, where scientific quantification/extensions and improvements of Stewart's response farming approach would bring highly needed solutions (Stewart, 1991; Gadgil et al., 2000). There are recent attempts to define conceptual strategies for demonstration projects of this kind, demanding strategic and tactical interactions between physical, agricultural, social and economic systems, with a long list of elements (Manton, 2001). Carefully organized, but less science driven pilot projects of that kind are highly needed, in which other experience referred to in this paper could be of much use as well (Gadgil, 2001; Stigter, 2002a, b).
Was this article helpful?