Adapting African Agriculture to Climate Change

From 2011 to 2014, the CALESA project was a research-for-development project which coupled integrated climate risk analyses, crop growth simulation modelling and field-based research both on-station and on-the-ground with participatory research with farmers. It comprised research-oriented activities for knowledge and technology creation, and development-oriented activities for information sharing and capacity building. The main purpose of the CALESA project was to develop sound adaptation strategies for future temperature increases associated with greenhouse gas emissions using “analogue locations”, both as learning- and technology-testing sites. This was meant to improve the ability of rainfed farmers in the semi-arid tropics of sub-Saharan Africa, in particular Kenya and Zimbabwe, to adapt to progressive climate change through crop, soil and water management innovation, and appropriate crop genotype choices. Another key feature of the CALESA project was the development and implementation of tailor-made capacity-building activities specifically designed to fulfil the needs of local scientists in the field of climate change adaptation and climate-smart agriculture. To achieve its objectives, the CALESA project used a combination of model-based ex ante analyses and iterative field-based research on station and in farmers’ fields. This facilitated the evaluation of potential agricultural adaptation strategies for rainfed agriculture in the semi-arid and dry sub-humid tropics. In this line, four important crop production zones (two in Kenya and two in Zimbabwe) were identified. Subsequently, the corresponding ‘spatial analogue locations’ for each production zone, providing eight study locations in all, were identified. A strong element of participatory research with small-scale farmers ensured that the perceptions of current and future climate risk and their preferred climate change adaptation strategies was effectively taken into account. In addition, this also ensured that the project activities and outputs remained relevant to their needs and expectations. The main outputs of the CALESA project are as it follows. Firstly, the identification and fully characterisation of four important crop growing areas in Kenya and Zimbabwe which comprise cool/dry, cool/wet, warm/dry and warm/wet growing conditions, and their temperature analogue locations. Secondly, through the combined use of long-term daily climate data, crop growth simulation models and participatory surveys with farmers, the identification and quantification of the implications of both current and future climate change production risk at the study locations. Thirdly, through iterative field research both on station and in farmers’ fields over more than 2 years, the evaluation of potential crop, soil and water management, and crop genotype adaptation options. This was followed by the formulation of adaptation strategies for the target locations. Finally, through the overall implementation of the project activities, the institutional capacity in understanding climate change impacts and the development of effective adaptation responses in Kenya and Zimbabwe was fostered.

Climate change is expected to be one of the major threats to sustained economic growth leading to extended poverty in semi-arid regions of sub Saharan Africa (SSA). The areas of highest vulnerability are the health sector, food production, biodiversity, water resources, and rangelands. Climate change will likely create increasingly high temperatures and dry conditions across much of the globe in the next 30 years, especially along large parts of Eurasia, Africa and Australia. Many of the world’s most densely populated regions will be threatened with severe drought conditions. It will likely have a profound and negative impact on livelihoods of many rural and urban communities, which could lead to changes in land use. It is estimated that the Eastern regions of Africa will experience reduced average rainfall (although some areas may experience increased average rainfall) exposing agriculture to drought stress and a rise in temperature. The situation will be worsened by the interaction of multiple stresses factors occurring at various levels, which will negatively impact agricultural productivity.

The aim of this study is to assess the extent to which scientific information has been used to inform climate change adaptation policies, plans and strategies in Kenya; and also to assess the effectiveness of existing platforms for sharing climate change information in the country. Two major policy documents guiding climate change adaptation planning in Kenya, the National Climate Change Response Strategy (NCCRS) and the National Climate Change Action Plan (NCCAP), were analysed for use of scientific information in their formulation through literature review; and interviewing policy makers using an open-ended questionnaire to determine the extent to which they accessed and applied scientific-based evidence of climate change impacts in development planning. Both documents, the NCCRS and NCCAP, made fairly good use of evidence contained in technical reports, especially the UNFCC, World Bank and FAO reports. However, they made very minimal, less than 20 %, reference to the hard scientific facts offered by journals, books and workshop proceedings. Similarly, only about 6 % of the respondents used the climate change information to develop mitigation and adaptation plans, training curricula, and Research and Development programs. The rest, over 76 %, rarely used it for planning purposes. This could be attributed to limited knowledge of appropriate methodology to distil relevant decision-relevant information from the spectrum of available information on climate change projections, availability of the information in user-unfriendly formats, and lack of information sharing protocols. There is need to reverse this trend. Most respondents (42 %) preferred the agricultural extension system in delivering climate change information. This was followed by stakeholders meetings with 29 % of the respondents’ preference, conferences and workshops with 5 %, media (4 %), and climate change networks and internet with less than 1 % each. However, the national agricultural system is severely constrained by staff and facilities, and is therefore very limited in its reach. There is therefore need to strengthen it and also take full advantage of recent advances in ICT if the war against climate change is to be won. Meanwhile, majority of the respondents (50 %) were ignorant of the existence of any climate change databases. But about 17 % of the respondents were aware of and accessed databases hosted by Consultative Group on International Agricultural Research (CGIAR) and other international research centres. Another 10 % of the respondents relied on databases managed by donor agencies whilst about 8 % of the respondents each accessed databases established by Government Departments and National Agricultural Research Institutions (NARIs). Finally, about 7 % of the respondents relied solely on the FAO-based databases. The preference by respondents for databases managed by CGIAR centres may be attributed to the richness and accessibility of these databases due to very active participation of these centres in climate change research. There is need to enrich NARIs databases and those of Government Departments and make them more accessible to enhance sharing and application of climate change information by policy makers and other stakeholders.

Given that climate change and variability have become one of the greatest threats to food security and livelihoods, a baseline study and some literature synthesis were conducted to understand the current situation of CC scenarios in Kenya. The study sought to determine the current status of CC projects that have been undertaken in Kenya in the past five years. Major CC themes and sensitive productive sectors to CC were conceptualized in which the study was based. The baseline survey targeted key informants in academic, research and policy arenas. It was observed that adaptation, mitigation and capacity building accounted for 60, 17 and 23 % of the projects sampled. Agricultural sector (crops) accounted for most of CC projects, accounting for 36 % as well as 40 % of all projects on adaptation. Agriculture, livestock and environment sectors accounted for 30 % each of the mitigation projects. It is established that most projects undertaken in Kenya on CC arena have been on adaptation, capacity building and mitigation. CC projects undertaken in Kenya were in agriculture and livestock sectors. Although considerable efforts appear to have been put in adaptation to CC, more needs to be done, especially in agriculture and water sectors, which are important in Kenya’s economy.

Drier parts of Embu County, Eastern Kenya, endure persistent crop failure and declining agricultural productivity which have been attributed, in part, to prolonged dry-spells and erratic rainfall. Nonetheless, understanding spatial-temporal variability of rainfall especially at seasonal level, is an imperative facet to rain-fed agricultural productivity and natural resource management (NRM). This study evaluated the extent of seasonal rainfall variability and the drought characteristics as the first step of combating declining agricultural productivity in the region. Cumulative Departure Index (CDI), Rainfall Anomaly Index (RAI) and Coefficients-of-Variance (CV) and probabilistic statistics were utilized in the analyses of rainfall variability. Analyses showed 90 % chance of below cropping-threshold rainfall (500 mm) exceeding 213.5 mm (Machanga) and 258.1 mm (Embu) during SRs for one year return-period. Rainfall variability was found to be high in seasonal amounts (CV = 0.56 and 0.38) and in number of rainy-days (CV = 0.88 and 0.27) at Machang’a and Embu, respectively. Monthly rainfall variability was found to be equally high even during April (peak) and November (CV = 0.42 and 0.48 and 0.76 and 0.43) with high probabilities (0.40 and 0.67) of droughts exceeding 15 days in Embu and Machang’a, respectively. Dry-spell probabilities within growing months were high (81 %) and (60 %) in Machang’a and Embu respectively. To optimize yield in the area, use of soil-water conservation and supplementary irrigation, crop selection and timely accurate rainfall forecasting should be prioritized.

The objective of the current study is to address the possible potential impacts of climate change and variability on agricultural crops and water resources in Pennar river basin, of Southern India. As part of the study Integrated Modelling Assessment (IMS) was developed by establishing functional links between hydrological model Soil Water Assessment Tool (SWAT), agricultural crop simulation model Environmental Policy Integrated Climate (EPIC) and regional climate model Providing REgional Climates for Impacts Studies (PRECIS). Database pertaining to climatic parameters, hydrological and agro-meteorological inputs to run integrated assessment systems are synthesized to run the model for study area. The model in general aim at major driver of this study is HadRM3 (Hadley Centre third generation regional climate model)—The Hadley Center Regional Climate Models resolution, which is 0.44° × 0.44° (approx. 50 km cell–size) on ground covering an average size of typical Indian districts/sub-basins. For regional levels the results are obtained by aggregating from the sub-basin/district level. The assessment will include the following components: (1) Baseline climatology, (2) Under global warning HadRM3 derived climate change scenarios, (3) Water Resources (Hydrological) analysis including irrigation water, and (4) agro-meteorological analysis including soil-water regime, plant growth and cropping pattern. Overall in Pennar region results revealed that the mean annual flows in the river system would increase by 8 % in A2 and 4 % in B2 whereas, increase in evapotranspiration losses were found to be about 10 % in A2 and 12 % in B2. Impacts on crop yields is the combined effect of increased surface temperatures, decreased rainfall and higher ambient atmospheric CO₂. Three rain-fed crops (Groundnut, Sorghum, Sunflower) show decreased yields under A2, whereas B2 seemed to be relatively better than A2. The decrease is significant for groundnut (−38 % for A2 and −20 % for B2), but compared to groundnut impact were less detrimental for other two rain-fed crops (Sorghum and Sunflower). Rice being an irrigated crop in the region showed decrease in yield by −15 and −7 % for A2 and B2 scenarios respectively. Negative simulated crop yields in the region are predominantly due to increased surface temperatures in the future climate change scenarios.

Climate analogues, based on 30 years meteorological data, were identified in smallholder areas of Zimbabwe. The sites were Kadoma (722 mm annual mean rainfall; 21.8 °C annual mean temperature) which was the higher temperature analogue site for Mazowe (842 mm annual mean rainfall; 18.2 °C annual mean temperature) for wetter areas, and Chiredzi (541 mm annual mean rainfall; 21.3 °C annual mean temperature) which was the higher temperature analogue site for Matobo (567 mm annual mean rainfall: 18.4 °C annual mean temperature) for drier areas. At each site and for each crop, three varieties were laid out in a randomized complete block design with three replications. The trials were conducted for two seasons (2011/2012 and 2012/2013). Maize and groundnut yields were higher at the cooler and wet sites and decreased significantly at the warmer and dry sites. In case of sorghum and cowpea, yields at the hotter site remained high implying that these crops are more tolerant to warmer temperatures predicted for 2050. At the drier sites, yields for all crops were significantly lower at the hotter site implying that crop production in the 2050s climate of the cooler site will be more difficult. The hypothesis that with increasing surface temperatures in a climate change scenario short duration genotypes can perform better compared with long duration was not confirmed.

Three experiments were conducted to evaluate the performance of maize, sorghum, common bean and pigeon pea varieties under different water management in a cool and wet region of Central Kenya, as a part of the studies at analogue sites. The first experiment evaluated the growth and performance of three varieties (early maturing: EM, medium maturing: MM and late maturing: LM) of maize (Zea mays L), sorghum (Sorghum bicolor L.), pigeon pea (Cajanus cajan) and common bean (Phaseolus vulgaris L.). The second experiment evaluated maize and sorghum response to water conservation and three fertiliser rates (0, 20 and 40 kg N/ha). The third experiment assessed the effect of water conservation measures on crop yields of common bean and pigeon pea grown under three plant densities (low, medium and high). Tied ridge tillage was used as the water conservation measure and disc plough as the control in the second and third experiments. Maize, sorghum, pigeon pea and common bean took more than 180, 245, 217 and 95 days respectively, to reach physiological maturity. The MM maize variety (DK8031), EM pigeon pea variety (ICPL 84091) and LM common bean variety (GLP 24) yielded the greatest grain of 4,938, 881 and 620 kg/ha respectively, among the respective crop varieties. The sorghum varieties were attacked by fungal and rust diseases leading to yield losses in all seasons. Soil water conservation in general did not have a significant effect on crop yield though there were yield improvements. In the plant density trial, the medium plant densities of pigeon pea (33,333 pl/ha) and common bean (148,148 pl/ha) resulted in the greatest grain yields. The highest grain yield of maize (4,184 kg/ha) and sorghum (47 kg/ha) was obtained in plots with 20 kg/ha of nitrogen fertilizer. Based on the results of this study, pigeon pea and common bean can be introduced in the farming systems to improve crop diversity. The production of the tested sorghum varieties should be discouraged in this region because they are prone to fungal and rust diseases due to the cold and wet weather conditions.

Climate analogues can be used to assess climate–induced risks and adaptation options for smallholder farmers. Surveys were carried out in smallholder areas at two 2050s climate analogue sites to assess smallholder climate-induced risks, farmers’ perceptions, and adaptation options, with a gender perspective. Pairs of sites selected had similar annual rainfall totals but differed in mean annual temperature by 2–4 °C. For drier areas Chiredzi was hypothesised to represent Matobo, and for wetter areas Kadoma was hypothesized to represent Mazowe/Goromonzi 2050s climates. Differences in crop management strategies and gender issues vary across sites. At the drier analogue pair, higher proportions of households grew small grains in Chiredzi compared to Matobo. Implications are for increased uptake of small grains, in 2050s climates for Matobo farmers. Gender issues include labour for production and processing of the small grains, against a background of male labour migration. For wetter climates, soil and water management strategies are important options for smallholders. Accesses to draft power, labour, agricultural assets, social and financial capital in differently managed households are important for increasing adoption of effective crop management strategies.

This paper presents the results from a study conducted to examine adaptation measures initiated by farmers in response to variability and changes in climate and develop a model for climate change adaptation. Drop in rainy days and amount of rainfall concomitant with raise in daily temperature after the year 2000 has been acknowledged widely by farmers. This is corroborated by the actual rainfall and temperatures recorded in the local meteorological observatory. These factors exacerbated groundwater extraction for irrigation and domestic needs. In order to cope with the predicament, farmers shifted to late sowing varieties, dairying and adopted soil and water conservation measures such as farm ponds, ridges and dead-furrows and mulching. Those using conventional irrigation, shifted to drip irrigation, cultivating horticulture crops. Case study of a vulnerable woman farmers indicated that by participating in the groundwater market, the farmer purchased groundwater to cultivate vegetables and keeping three local cows which provided them with steady flow of income from milk. Thus, coping mechanisms were to sustain incomes and resilient due to climate change crisis. To create awareness and understanding of climate change and the challenges for social scientists an integrated model has been proposed by using synergies from program development, program delivery and program impact.

The salient results of in situ soil water conservation technologies that have been extensively tested and found suitable for increasing soil moisture for increased land productivity in the arid and semi-arid lands (ASALs) of eastern Kenya are reviewed in this paper. The technologies reviewed are fanya juu terraces, contour bunds, negarims, trapezoidal bunds, on-farm micro-catchments, Zai pits, tumbukiza, tied ridges, deep tillage and sub-soiling and ripping. These technologies hold rainwater on the soil surface thereby allowing it more infiltration time leading to enhanced soil moisture status, which would not be attained in the absence of these interventions. Zai pits, tumbukiza and deep tillage when used together with soil fertility improvement can increase crop yields by 4–10 times when compared to other similar fields cultivated conventionally. When tied-ridging tillage is used together with fertilizer, manure or their combination it can increase crop yields by 100–300 %. Sub-soiling and ripping increases crop yields by 50–100 % when used together with soil fertility improvement. Micro-catchment technology at 1:1 and 2:1 catchment to cultivated land ratio can increase crop yields, but is not practised due to land limitation. Use of fertilizers and or manures with in situ soil moisture conservation leads to improved water use efficiency by crops planted in the semi-arid eastern Kenya. It is, therefore, proposed that in situ rainwater conservation technologies should be an integral part of the farming systems for increased soil moisture conservation, crop production and food security in the semi-arid Eastern Kenya.

Two studies were conducted during the long rains of 2010 and 2011 and short rains of 2011 and 2012 to evaluate the performance of drought tolerant maize varieties under different water harvesting technologies (zai pits, tied ridges and conventional). The treatments were laid out in a split plot design with water harvesting methods as the main plots and maize varieties as the sub-plots. Four maize varieties (DK8031, DUMA 43, KDV1 and PH4) were evaluated under the three water harvesting technologies for the first experiment. For the second experiment, four maize plant population treatments of 3, 5, 7 and 9 plants per pit were used. Maize variety DUMA 43 was used alongside the four plant populations. The results for both experiments indicated that the maize yields in zai pits and tied ridges treatments were significantly (P < 0.05) higher than conventional treatment and the population of 5 plants per pit had significantly (P < 0.05) higher grain yield than the rest of treatments.

Scenario analysis using data generated from APSIM model was conducted to investigate the effect of soil and water conservation practices (tied ridges and mulching) on grain yield of improved maize varieties (Katumani and Makueni) generated with and without N fertilizers under below normal (<250 mm), normal (≥250 < 350 mm) and above normal seasons (≥350 mm) in two sites, Katumani and Makindu in Machakos and Makueni counties Eastern Kenya. Results indicate that the yields were significant (<0.01) under the different seasons and treatments with the magnitude of the yields response varied. Highest yields in Katumani (3,370 kg/ha) were obtained during below normal seasons and when both fertilizer and tied ridges were used. In Makindu, however, under all treatments, highest yields were obtained during above normal seasons with 3,708 kg/ha yield when 40 kg N/ha fertilizer was applied. Lowest yields on the other hand, were obtained during normal seasons in both sites with 507 kg/ha in Katumani and 552 kg/ha under tied ridges and mulching + fertilizers in Makindu. Compared with farmers practice (control), the yield increment obtained was 4 kg/ha (0.6 %) and 5 kg/ha (0.7 %) in Katumani; 32 kg/ha (4.6 %) and 33 kg/ha (4.7 %) in Makindu under mulching and tied ridges respectively during below normal seasons otherwise the yield decreased during normal and above normal seasons with up to 19 % in Makindu when tied ridges was practised. Fertilization increased the yields of maize by as high as 2,552 kg/ha (433 %) and 2,319 kg/ha (166 %) in Katumani and Makindu respectively during above normal seasons. However, during normal seasons, there was yield decrease in Makindu by 42 %. When both fertilization and soil and water conservation practices was done, yield increase was 2,335 kg/ha (456 %) and 2,382 kg/ha (465 %) in Katumani during normal seasons under mulching +40 kg N/ha and tied ridges +40 kg N/ha respectively. In Makindu, yields declined during normal seasons, however, increase was by 2,229 kg/ha (160 %) and 2,108 kg/ha (152 %) during above normal seasons under mulching +40 kg N/ha and tied ridges +40 kg N/ha respectively. The results indicate that the use of fertilizers and soil and water conservation are indispensable for ensuring food security in semi-arids where rainfall is very variable.

A study was carried out to understand the adoption levels of water conservation practices in Meru County. The influence of water resource accessibility on adoption of water conservation (WC) practices and constraints were assessed. Tree planting, roof catchment and bench terraces were the major WC practices in use. Multiple regression analysis revealed that lack of technical know how could explain 83.5 % variations of adoption level of WC practices. One sample t-test comparing the means of WC practices among respondents’ was significant at P < 0.01. Spearman’s rank test revealed a decreasing trend during the long rains (March–May) for the period 1986–2008 at P < 0.05. The disparity between the levels of adoption among water users coupled with the decreasing seasonal rainfall calls for urgent and better management of water resources in the study area.

Inorganic fertilizers have become extremely important in correcting declining soil fertility in seed potato (Solanum tuberosum L.) production in Kenya. Unreliable rainfall has also limited seed production. Knowledge on water and nutrient use efficiencies in potato grown under different irrigation regimes with different nitrogen and phosphorus levels will help predict the best application rates for optimal seed potato production and yield. A study was conducted at Egerton University, Horticultural Research and Teaching Farm to determine the effect of integrated application of irrigation water, nitrogen (N) and phosphorus (P) use efficiencies of water, N and P. In a split-split plot design, the irrigation water was applied to maintain soil water at 40, 65 and 100 % field capacity in the main plots, N (0, 75, 112.5 and 150 kg N/ha) to subplots and P (0, 115, 172.5 and 230 kg P₂O₅/ha), which translated into 0, 50.6, 75.9, 101.2 kg P/ha) to sub-subplots, with each treatment replicated three times and the trial repeated once. The irrigation water was applied throughout the potato growth period through drip tube lines, with N supplied as urea (46 % N) in two splits, and P as triple superphosphate (46 % P₂O₅) at planting time. Data on seed potato yield was collected from each treatment at harvest and used to calculate water, N and P use efficiencies. High irrigation water at 100 % compared to 65 and 40 % rate resulted in relatively high N and P use efficiencies, but decreased water use efficiency. Application of intermediate to high N and P nutrient increased the water, N and P use efficiencies. It is recommended to apply low to intermediate irrigation water, intermediate to high N and P to increase their use efficiencies during seed potato production.

Potential for promoting sorghum crop as a climate change adaptation strategy for rain-fed agriculture in Embu County, Kenya was evaluated using farmer perceptions and scientific methods. Three hundred and sixty six smallholder farmers participated in the evaluation. The treatments which were overall rated as ‘good’ are tied ridges with a mean score of 2.9 and mean rank (2,873.87). Under this treatment sorghum grain yield of 3.7 t ha⁻¹ was recorded with application of 40 kg P ha⁻¹ + 20 kg N ha⁻¹ + Manure 2.5 t ha⁻¹. This was closely followed by tied ridges and contour furrows overall rated ‘good’ best three under the same soil fertility management options with a mean score ranging from 2.65 to 2.8 and yielding 2.7–3.7 t ha⁻¹. However, the treatments which were rated as ‘poor’ were experiment controls with a mean score below (1.43), mean rank (1,101.24) and yielding as low as (0.7 t ha⁻¹). Therefore, integration of organic and inorganic inputs under various water harvesting technologies could be considered as an alternative option towards food security under climate change for semi-arid areas of Embu County.

Evaluation of six maize varieties under four different water harvesting and tillage technologies was undertaken with the aim of determining their effect on the performance of maize genotypes and their effectiveness in improving nutrient and water use efficiency. The work was carried out at Mariakani site (one of the KARI centres) representing the arid and semi-arid lands of coastal Kenya in the long rains and short rains seasons of 2005 and 2006 respectively. The results indicate that rainwater harvesting is not critical when the season is wetter than normal in the arid and semi-arid environments. This was demonstrated by the high yields that were recorded from the maize varieties (Pwani Hybrid 4-PH4, Coast Composite Maize-CCM and the local check-Mdzihana) which usually require relatively high rainfall amounts in order for them to produce better yields. Despite the excellent performance of PH4, CCM and Mdzihana, these maize varieties cannot be recommended for the semi-arid areas since the high yields were realized under above normal rainfall. There is need for further research to identify the maize varieties that would be appropriate for the areas that normally receive low rainfall.

Farmers in Mukuyuni can easily adopt drought mitigating technologies like cassava and sweetpotato. They are mainly propagated through stem cuttings or vines to produce starchy tuberous roots, which would provide the much needed carbohydrates if promoted properly. The results showed that about 90 % of the farmers put between 0.125 and 0.25 acres under cassava cultivation. The number of years under cassava and sweetpotato production ranged between 1 and 20 years with only a few farmers indicating to have been growing cassava for a period above 20 years. The majority of farmers were growing local cultivars of the two crops. The main method of utilizing cassava was boiling and eating as a snack (45 %). Over 77.8 % of the respondents indicated the origin of their cassava and sweetpotato cultivars as other farmers. The main method of utilizing sweetpotato was boiling and eating as a snack (56.3 %). About 3.1 % of the farmers mixed sweetpotato with beans and maize and consumed it as a stew while another 3.1 % have not utilized sweetpotato at all. Only 3.1 % fed sweetpotato to livestock. The rest, 28.1 % sold sweetpotato in the local market. This study established that there is a lot of room for commercializing cassava and sweetpotato production in Mukuyuni Division Makueni County through processing and promoting their utilization as a meal.