Agroecological location of farms and choice of drought coping strategies of smallholder farmers in Swaziland

This study uses data from Swaziland to test whether variations in local agro-ecological regions levels of drought susceptibility and other socioeconomic factors significantly determine farmer selected drought coping and adaptation strategies. This was in response to the policy need to understand how livelihoods of poor, rural, smallholder farming communities can be made more resilient in the face of recurrent droughts. Swaziland’s agroecological regions were divided into those that were highly susceptible (Lubombo and Lowveld) and those that were relatively less susceptible (Highveld and Middleveld) to drought. Using structured questionnaires and face-to-face interviews, the study compared 115 randomly selected farmers from the former and 50 farmers from the latter region based on the following household level indicators: behavioural responses to perceived long-term changes in temperature and precipitation; the impact and behavioural responses to the most recent drought event; how farmers would have responded if they had ex ante information on the most recent drought event; ex ante private investment in anticipation of future drought events; and finally farmer preferences for ex ante public investments in anticipation of future drought events. Regarding farmers’ behavioural responses to perceived long-term changes in temperature and precipitation; all farmers in the two agro-ecological zones clearly perceived climate change variability and in response adopted strategies to mitigate the effects. There were, however, marked differences in strategy selection between the two regions. On perceiving these changes; farmers in the more susceptible regions were more interested in adopting adaptation strategies, whereas those in the less susceptible regions were much more reluctant vi to apply any drought coping and adaptation strategies. To determine the significance and validity of these differences in adaptation; chi square tests were conducted on each strategy applied in the two regions and those that were significantly different between the two zones were further analysed using probit analysis to determine socio-economic, biophysical, and policy variables that contributed to the selection of these strategies. The results of the chi square test revealed that adaptation strategies that were significantly selected between the two regions in the past ten years were purchasing water (51%), construction of livestock shelter (51%), purchasing hay (50%), changing livestock type to more drought tolerant animals such as goats (41%), and migration of livestock to areas with better grazing (35%). Factors that influenced the decision to change livestock type were region of farmer location (p=0.001), income source (p=0.084), availability of extension (p=0.049), the decision to proceed with farming (p=0.007), and the need for credit (p=0.007). The decision to purchase hay, on the other hand, was influenced by region of farmer location (p=0.007), whether the farmer owned a trailer or not (p=0.042), availability of maize fields (p=0.012), availability of fields for other crops besides maize (p=0.012), extension services (p=0.004), and the need for credit (p=0.050). Factors that influenced farmers’ decision to migrate their livestock to better pastures were region of farmer location (p=0.007), income source (p=0.022), gender (p=0.022), occupation (p=0.044), number of children (p=0.034), extension services (p=0.024), the decision to proceed with farming (p=0.008), and the need for credit (p=0.032). With regards to purchasing water as a coping strategy; factors such as region of farmer location (p=0.08) influenced the decision, together with income source (p=0.088), ownership of a trailer (p=0.016), livestock units (p=0.073), availability of extension services (p=0.046), and government support (p=0.021). Lastly, factors that contributed to farmers’ decision to construct livestock shelter were region of farmer location (p=0.019), average income (p=0.070), gender (p=0.096), availability of fields for crops other than maize (p=0.087), availability of extension services (p=0.050), and the decision to proceed with farming (p=0.010). With respect to impacts of the most recent drought; farmers were severely affected in their crop and livestock production. In livestock production, farmers mainly experienced cattle deaths (53%); but relatively less deaths in chicken and goat production (21% and 3% respectively). In crop production; extensive losses were experienced in maize production. On average, farmers in the more susceptible region lost E964/ plot and those in the less susceptible region lost E648/plot. In response to the drought; farmers in both regions responded in a reactive manner as they were ill prepared to deal with the impacts of the drought. The coping strategy that was used by 84% of the farmers in crop production was changing planting dates; which basically means that farmers waited for first rains which in turn delayed planting as the drought resulted in late first rains. The second most applied strategy (by 10% of farmers) was that of not planting at all; as some farmers took late first rains to be a sign of an imminent drought. The rest of the farmers irrigated (2%), applied more chemicals (2%), changed crop type (1%), replanted (1%), and started practicing conservation farming (1%). Farmers did not adopt any strategies with respect to livestock production when the drought hit. Strategies that farmers would have selected if they had received ex-ante information on the drought were shifting planting dates to correspond with first rains (94%), changing type of crop to more tolerant crop types such as root crops and legumes (67%), purchasing hay for livestock (58%), construction of livestock shelter to protect animals (58%), changing livestock mix to drought resistant species such as goats and chickens (54%), and livestock migration to better areas (38%). Factors that influenced the decision to shift to drought tolerant crops were age (p=0.004), average income (p=0.007), ownership of goats (p=0.020), availability of livestock shelter (p=0.010), and growing crops other than maize (p=0.001). The decision to change type of livestock to more drought tolerant species such as goats and chickens was influenced by region of farmer location (p=0.000), level of education (p=0.041), average income (p=0.050), ownership of a trailer (p=0.0042), ownership of goats (p=0.048), extension services (p=0.003), need for credit (p=0.001), and the availability of governmental support (p=0.000). Lastly; purchasing hay was influenced by region of farmer location (p=0.002), cattle ownership (p=0.048), availability of livestock shelter (p=0.071), access to credit (p=0.050), willingness to purchase supplementary inputs for livestock (p=0.018), and the need for credit (p=0.019). With respect to private investments in anticipation of future droughts; farmers were willing to adopt a few strategies in preparation. There was more interest in preparing for future droughts mainly from farmers in drought susceptible areas compared to those that were in less susceptible areas. Strategies that more than 50% of farmers in the drought prone areas were willing to adopt include changing crop type to those that are less susceptible to drought (82%), changing livestock from cattle to more drought tolerant livestock such as goats (64%), construction of livestock shelter to protect livestock from climate extremes (65%), and purchasing hay to supplement grazing (70%). Farmers in the region were not willing to use an alternative water supply (74%) as they stated that severe droughts in their areas compromise water from all sources therefore such a strategy is not a viable option. In the less susceptible region; farmers were willing to adopt only two strategies, these being changing crop type (60%) and utilizing alternative sources of water (50%). Regarding public investments farmers would like implemented in anticipation of future droughts; farmers were mostly interested in relief measures such as food, water, and money. Only 15% of farmers mentioned the need for inputs and reservoirs for irrigation (12%). Other less popular strategies included community grain storage facilities for rationing during drought periods, education on drought coping and adaptation, and assistance with nonagricultural business start-up costs. Farmers in middle class households were also concerned about their exclusion in relief measures; thus they recommended that the government should reconsider their selection criteria. In conclusion; small scale farmers are severely impacted by drought, and on perceiving such impacts, they adopt strategies to reduce the impacts on their enterprises. The severity and nature of drought impact differs across agro-ecological locations, thus bringing to light the importance of area specific strategies when dealing with drought. It is therefore recommended that all stakeholders involved in developing drought resilience measures for small holder farmers should put into consideration agro-ecological location and farmer specific characteristics when formulating policies that address drought incidences. Farmers, non-governmental organizations, and the government should work hand in hand in this regard to increase farmer resilience to future droughts in Swaziland.