Integrating Local Farmers Knowledge Systems in Rainfall Prediction and Available Weather Forecasts to Mitigate Climate Variability: Perspectives from Western Kenya

Abstract

This chapter examines relevant studies and examples on integrating farmer’s traditional knowledge systems in rainfall prediction with available weather forecasts to mitigate impact of changing climate among rainfall dependent farmers in Western Kenya. The chapter combines the results of a study conducted in Western Kenya among maize and wheat growing farmers in Uasin Gishu County and perspectives from other related studies within the Eastern and Southern part of Africa. The chapter details how farmers have navigated the impact of changing climate on the farming enterprise that is largely dependent on rainfall. The findings reveal that farmers in western Kenya have experienced crop losses during planting and harvesting seasons due to prevailing variations in weather patterns. This is corroborated by over 340 (87.8%) of farmers in Uasin Gishu county of Kenya who agreed so and further stated that they had experienced changes in rainfall patterns and even the timing for maize and wheat growing had become uncertain and contrary to what they have known over time in the recent years. Similarly, like other findings in the reviewed studies in this chapter, the Kenyan farmers (84.9%) agreed strongly that they applied their local indigenous knowledge and experience gained over time to predict rainfall onset and cessation dates thus making key farming decisions. Relying heavily on traditional weather forecasting by farmers is catastrophic now due to changes on the environment associated to environmental degradation; ecosystem disturbance and changing climate which have seen important traditional predictor indicators disappear or lost completely from the environment. Although over 90% of the Kenyan farmers in average belief in use of weather forecast information, integration of this information is not effective because of its adaptability, format and timing challenges. The same is true for farmers in some countries within the region. Importantly, provision of context-specific and downscaled weather forecast information to support farmer’s resilience is crucial. Most studies and programmes reviewed in this chapter agree that there is synergy in integrating local knowledge systems and available weather forecast information for better weather prediction. It is critical that policymakers, practitioners or key stakeholders and forecasters (both from the meteorological services and indigenous groups) converge and agree on weather prediction if they are to support farmers in managing climate risk or uncertainties.

1. Introduction

Due to the prevailing climate variability brought about by a changing climate global phenomena, the devastating impact on various sectors have been felt at regional and local levels more so, the developing countries. The changing climate is having a growing impact on the African continent, hitting the most vulnerable hardest, and contributing to food insecurity, population displacement and stress on water resources. Further, the latest decadal predictions, covering the five-year period from 2020 to 2024, shows continued warming and decreasing rainfall especially over North and Southern Africa, and increased rainfall over the Sahel [1]. Extensive areas of Africa will exceed 2 °C of warming above pre-industrial levels by the last two decades of this century under medium scenarios as reported in the Intergovernmental Panel on Climate Change Fifth Assessment Report [2]. Some countries in the sub Saharan Africa have recently witnessed increased flood disasters, invasion of desert locusts that endanger food security and livelihoods and now face the looming danger of drought due to the likelihood of La Niña event [3].

The Horn of Africa region experienced a combination of very dry conditions, floods and landslides associated with heavy rainfall between 2018 and 2019 period while the Southern Africa region faced drought phenomena. A report by WMO explains that extensive flooding occurred over large parts of Africa in 2020 as rainfall was mainly above average in most of the Greater Horn of Africa region during the March–May season. This followed a similarly wet season in October–December 2019. Sudan and Kenya thus were affected more with 285 deaths reported in Kenya and 155 deaths and over 800,000 people affected in Sudan in addition to disease impacts associated with flooding. Heavy rains in the Arabian Peninsula and East Africa resulted in the largest desert locust outbreak in 25 years across the Horn of Africa [4, 5]. In Ethiopia alone, 200000 hectares of cropland were damaged and over 356000 tons of cereals were lost, leaving almost one million people food insecure [6]. In Somalia, floods were associated with the displacement of over one million people in 2020, mostly inside the country, while drought-related impacts induced a further 80,000 displacements [7].

Based on Remote Sensing data and other observations, Kenya’s Lakes that include Naivasha, Elementaita, Nakuru, Bogoria, Baringo, Turkana, Logipi have been rising since 2018. Most of the lakes have over flown displacing thousands of people from their homes, leaving behind submerged schools, health facilities and social amenities and rendering some of the public infrastructure like roads impassable [8]. The agricultural sector in Eastern and some Southern Africa region remain dominantly a rainfall dependent system and majority of farmers work on a small-scale or subsistence level and have few financial resources, limited access to infrastructure, and disparate need for access to agronomic data and information [9]. For the farmers, any abnormal variation in rainfall onset and cessation dates in addition to extreme temperatures result in serious crop loss or damage. With the prevailing changes in weather patterns due to changing climate, African traditional knowledge systems in weather prediction for agriculture otherwise reliable for centuries could or have been rendered ineffective to some extent. Agriculture, being the backbone of Africa’s economy and accounting for the majority of livelihoods across the continent is both an exposure and vulnerability “hot spot” for climate variability and change impacts [10]. IPCC projections have suggested that warming scenarios risk having devastating effects on crop production and food security. Farmers therefore have suffered and will continue to suffer the most in their farming enterprise due to crop losses related to heat stress and pest damage, crop diseases and food system destabilization due to floods brought about by variations in weather and climatic conditions [10]. To mitigate this phenomena, most farmers have continued to rely on their indigenous knowledge systems in weather prediction in addition to integrating available weather forecasts to create a synergy to enable them navigate such uncertainties related to climate variability. This has created a great need for climate and weather information to be delivered to those farmers engaged in farming activities at the farm level in rainfall dependent farming systems. The information thus needs to be context specific, timely and delivery through the most reliable and accessible modes.

The objective in this chapter is to determine how farmer’s indigenous knowledge systems in rainfall prediction influence farm level planning and decisions in Kenya and other countries. Perspectives from some Eastern and Southern Africa region on utilization of indigenous knowledge system in rainfall prediction together with available weather forecasts to mitigate climate variability in the agricultural enterprise is captured for evidence based practice.

2. Indigenous knowledge systems and meteorological forecasts use among farmers dependent on rainfall in western Kenya and some countries in Eastern and Southern Africa – some examples

Indigenous knowledge (IK) is generally defined as “knowledge of a people of a particular area based on their interactions and experiences within that area, their traditions, and their incorporation of knowledge emanating from elsewhere into their production and economic systems” [11]. The terms indigenous, traditional and/or local knowledge as commonly referred to is knowledge and know-how that is accumulated over generations and guides human societies in their innumerable interactions with their surrounding environment [12]. Indigenous Knowledge is still important among local communities in many parts of Africa and the global scientific community do acknowledge its value [12]; however, most of this knowledge except a few captured by researchers has not been well documented hence the risk of losing some of this tacit knowledge when holders of such knowledge are incapacitated in any way. Over the years, communities have developed their own systems for monitoring climate conditions, but this information may not be adequate to inform adaptation if the changing climate continues in unprecedented way. An example from Burkina Faso illustrates that farmers who traditionally relied on observation of environmental indicators to predict climate patterns have now lost much confidence in their ability to predict rainfall accurately given increased changing climate thus increasingly seek to incorporate weather forecast information [13]. Socio–cultural changes also account for the shift away from traditional practices such as the use of bio–indicators for agricultural production, even when such practices continue to provide useful information [13, 14]. It is thus important to reflect on the traditional knowledge systems of communities as this provides an important entry point to scientist, researchers and key stakeholders in water, climate change, agriculture, food security and livelihood sectors into understanding how a new type of information about the climate/weather might be accepted and used by the local people to counter prevailing adverse weather conditions [15].

Climate information is a valuable resource for confronting and living with an increasingly uncertain future. Availability of climate information or weather forecasts provides a basis on which people whose livelihoods are affected by climate can make forward looking and flexible plans that are adapted to a range of climate possibilities. Consequently, climate information allows us to move from strategies which react to conditions as or after they occur, to those which seek to build resilience under all possible conditions and ultimately, to proactive strategies informed by forecasts and forecast probabilities [16]. In the agricultural enterprise, crop growth, or crop yield, requires appropriate amounts of moisture, light, and temperature at its correct time hence timing of farmer activities is critical at the local farm level. Detailed and accurate historical, real-time and forecast weather information can help farmers better understand and track the growth status of crop hence being able to make informed decisions. Having access to this critical agronomic data and information can guide farmers in making significant and potentially costly decisions, such as when to start tilling the land and subsequent planting period of crops as this is directly related to rainfall onset and cessation dates in rainfall dependent agriculture [16]. The most useful weather forecast information that can assist farmers in making decisions on agricultural management is the early indication of the characteristics of the rainy season. It should include: onset date of the main rains; quality of the rainy season (rainfall amount); cessation date of the main rains; temporal and spatial distribution of the main rains; timing and frequency of active and dry periods (wet and dry spells) and probability of extreme weather events. These decisions relate to the choice of crops to be made, cultivar (early or late flowering), mix of crops fertilizer use, pest and disease control and also timing of the harvest period [17].

There is evidence that substantial gain to sustainable food security and national development adaptation strategies can be achieved in Africa through provision and integration of improved climate information and prediction products into decision-making systems [18]. Accurate monitoring, prediction and early warning of seasonal rainfall performance can be used to improve planning and management of various rainfall dependent socio-economic activities like agriculture and the same can be used to enhance the livelihoods of the communities and services and support their resilience to adverse weather conditions. According to [13], access to climate information and technologies for adaptation is essential to enable actors to anticipate long–term risks and make the appropriate adjustments to increase their resilience. However, despite significant scientific gains in predicting the climate, often there is a lack of climate information available at the local level due to uncertainty in climate projections and seasonal forecasts, or due to lack of information on particular climate indicators, such as rainfall variability. Even when climate information is available, incorporation of scientific climate information into local decision making may not often occur because of the way such information is communicated and disseminated [19]. Several studies have shown that there is a need to make climate information more accurate, accessible, and useful for rural communities [20].

Adapting to climate change requires improved understanding of the linkages between climatic conditions and the outcomes of climate sensitive processes or activities; agricultural production for example in a certain region could be influenced by the availability of water resources and their management ways. Information from literature according to [21] explains that adequate use of climate and weather information conditions by farmers’ results in at least 30% increase in crop yields. The utilization of this information reduces farmers’ vulnerability to weather related risks, ensures that informed decisions are made on time, and reduces the risk of agricultural losses as well as indicating to farmers the most marketable crop in respective times. The analysis in the study on economic value of climate forecasts for livestock production in the Northwest Province of South Africa demonstrated that, for the commercial farmers, long term average annual income could potentially be increased through using ENSO predictions [22].

During a World Bank funded workshop in Dar-es-Salaam in 1999 on users responding to seasonal climate forecasts in southern Africa and the lessons learned then, it became apparent that there were communication barriers between the generators of the information and the users of the same information thus there was a need to develop appropriate information channels to relay such information. The second was that there were bottlenecks in the effective use of seasonal climate forecasts by farmers [23, 24, 25]. Users of seasonal climate forecasts have not been able to decode the information disseminated and therefore, users could not make use of the information provided if they did not understand the information provided in the first place [25]. Field studies conducted in the southern part of Africa reveal the existence of a considerable gap between information needed by farmers and that provided by meteorological services. There existed a communication barrier as the two parties have been interacting for a long time but to some extent, they have not been able to communicate effectively. The farmers know what they want and the meteorological services know what they need to give to the farmers, but there is no “shared meaning” [23, 24]. Without a shared meaning in communication, the value attached to particular information availed to the user (farmer) is diminished and may not serve the intended purpose.

In the light of such challenges, adopting an integrated approach where weather forecasts are combined with indigenous knowledge systems locally have shown to be effective in mitigating variability in the farming enterprise. According to [26, 27, 28], the use of scientific weather forecasts and indigenous climate forecast information for farm level decision making has been reported in Kenya and Mozambique. Previous studies in East Africa indicate that both IK and scientific weather forecasts are used for making crop and livestock production decisions, conserve the environment, and deal with other natural disasters. In Malawi and Zimbabwe, communities have combined scientific and indigenous climate forecast information for farm-level decisions hence being able to cope with prevailing drought [29]. Therefore indigenous knowledge system in weather forecasting is crucial in complementing available weather forecast information for improved decision making by farmers.

The Climate Change Adaptation in Africa (CCAA) program, funded by the International Development Research Centre (Canada) and the Department for International Development (United Kingdom) have supported projects that investigated how seasonal climate forecasts might be better integrated into agricultural and pastoral decision-making to strengthen livelihoods and food security [30]. Through these projects, it is apparent that indigenous knowledge forecasts, which have been used by communities for decades, provide information that is complementary to the meteorological forecasts. Indigenous seasonal forecasting and weather forecast information from the meteorological services help assist development challenges related to climate variability brought about by changing climate. Many farmers already use indigenous forecasts in their farm-level decisions and may only need certain information, such as total rainfall expected in the season, to complement what they already have. It is, therefore, important that policymakers, practitioners, and forecasters (both from meteorological services and indigenous groups) target existing gaps and take advantage of opportunities if they are to support farmers and pastoralists in managing climate risk in Africa [30].

Many societies and communities have their own ways of interpreting climate and weather patterns developed over years of experience. Traditional rainfall forecasts/predictions differ across communities, cultural background, and environment around the farm. According to [31], in South-Western Free State and Kwa-Zulu Natal of South Africa, as well as Western Kenya, inhabitants use birds, toads, and white ants to predict the dry season and onset of rains as well as temperatures. In Tanzania, they look at the behavioral patterns of birds and mammals. In a study on climate forecasting among the Basotho in Lesotho, they were asked if there were any ways to predict the coming weather and climate from what they know traditionally and a lot of answers were given that touched on weather conditions (hours and days) rather than climate conditions (weeks to months). The indicators were both environmental and cultural beliefs. According to [15], birds and insects were the most common environmental indicators. People mentioned the ‘squawk or the Makara’ bird as being indicative of rain in the coming days. Winds that blow from a certain direction were thought to bring rain. Plants flowering at certain time, the amount and color of clouds gathering, rising groundwater and frisky animals were all mentioned as indicators of imminent rain.

Among the Nganyi community in Western Kenya, the traditional weathermen observe the flora and fauna in the Nganyi forest shrine to predict weather conditions. The forest, which lies on just one acre of land, has pristine biodiversity that has helped the local Bunyore community predict weather conditions for generations. According to [32], researchers with the Climate Change Adaption in Africa (CCAA) program, a collaboration between international organizations and Kenyan scientists recorded data from a meteorological weather station near Bunyore in western Kenya for two seasons. They then compared its results with predictions made by indigenous forecasters who use the forest shrine as their main tool. The two findings were similar in all aspects. Based on this outcome, the researchers recommended the combination of both meteorological data and indigenous knowledge to facilitate accurate predictions that are acceptable scientifically and by the local community [32]. The Nganyi shares the consensus forecast for the coming season with the local community, in local languages, through radio, in churches and other community gathering points. The Nganyi community weather predictions are based on close observation of natural phenomena, like the budding or flowering of specific plant species and the behavior of local insects, birds and animals, associated with seasonal changes. A colony of bees migrating from downstream to the upper land clearly means that long rains are approaching and the vice versa symbolizes dry season [32].

Among farmers in the Kalenjin community in the North Rift region of Kenya, there are indicators or rainfall predictors that have been observed over the years and have been perceived as very important. The indicators can be classified as those related to the plant species, meteorological, animal and universe indicators. According to [33], some of the meteorological indicators include wind direction blowing eastwards signifying rainfall near onset, clouds thickening at the horizon and wind veering or breaking towards the east and cloud darkening in color; this signify rainfall onset and also cloud movement from eastern to western side of their farms all indicating rainfall onset. High sunshine intensity during the day and warm nights or high temperatures at night and low temperature in the evening signify onset of rains. Lightning strikes in near vertical position in three specific locations indicate near onset of rainfall.

3. Methodology and data

3.1 Study site and study design

The study was carried out in Uasin Gishu County; one of the 37 counties in Kenya. The county is made up of six sub counties namely:-Turbo, Soy, Moiben, Anapkoi, Kesses and Kapseret. According to [34], the total area of the County is 3327.8 Km2 with arable land covering 2603.2 Km2 and non-arable land covering 682.6 Km2. The County extends between longitude 34° 50′ and 35 ° 37′ east and 0° 03′ and 0° 55′ north. The headquarters of Uasin Gishu County is Eldoret town located on the main highway serving Kenya, Uganda and other countries in the great lakes region areas that include Uganda, Tanzania, Rwanda, Burundi, Democratic Republic of Congo, Zambia, and stretching all the way to Cape Town in the Republic of South Africa. The town is located at an altitude of 2085 m above sea level with a relatively cool climate experiencing daily mean maximum temperatures of 23.7 ° C and a mean minimum of 9.5 ° C. Eldoret town is traversed approximately latitude 00° 30’ North and Longitude 35° 15′ East of the Equator [35]. Uasin Gishu County is basically an agricultural district producing more than a third of the total wheat production in Kenya. Similarly, maize ranks second both as food and cash crop. A report compiled by [36] shows that the 2009 long rain maize production season was about 1.84 million Metric Tonnes, which was 28 percent below normal. There was a growing apprehension that the production could further be revised downwards due to insufficient and erratic rains in some parts of the main maize producing areas in North Rift including Uasin Gishu County due to the changing climate. The crop production has never been steady with each year having different challenges related to rainfall variations hence impacting positively or negatively to overall maize and wheat production.

The study utilized both qualitative and quantitative techniques. The use of both quantitative and qualitative research techniques are known to complement each other especially where exploration of indigenous knowledge among farmers is important. The qualitative aspect helped consolidate the themes emerging from the interview or survey. The target population was all farmers engaged in maize and wheat production resident in Uasin Gishu County at the time of study. Because of its nature of utilizing both qualitative and quantitative techniques, a representative sample was picked considering the levels of stratifications. The sampling frame for the study was 129, 384 farmers distributed as follows: - Soy Sub-County = 61, 138, Moiben Sub-County = 38, 950 and Kesses Sub-County = 29, 296. A minimum of 399 farmers were included in the study. In addition 12 key informants were interviewed; one from each ward totalling to 9 and also 2 from Directorate of Agriculture and Directorate of Meteorology in Uasin Gishu County respectively. One other key informant from the Kenya Meteorological Services in Nairobi was interviewed as well. Purposive, stratified and random sampling procedure was adopted to be able to capture a representative sample of farmers and based on this criterion, 3 sub-counties of Moiben, Soy and Kesses were selected for exhibiting both maize and wheat production (Figure 1).

Figure 1.

Map of Uasin Gishu County showing the study area.

4. Results and discussion

4.1 Farmers perception on local weather changes in the recent years

In order to understand what farmers perceive or belief in relation to weather changes and its resultant effects to maize and wheat growing activities, they were asked to use rainfall parameter; a familiar phenomena to gauge what they percieved as changes that may have occurred at their local level. They were asked to state whether “in the recent years there were any changes in rainfall patterns experienced by them and that even the timing for maize and wheat growing had become uncertain”. A total of 142 (36.7%) farmers strongly agreed with the statement and a further 198 (51.2%) agreeing as well. They further affirmed that the change in rainfall pattern experienced had led to declines and losses in maize and wheat production as captured in Figure 2.

Figure 2.

Farmers perception on prevailing local weather changes affecting their crop production in Uasin Gishu County.

The findings reveal that in total, over 340 (87.8%) of the farmers agreed that in the recent years they had experienced changes in rainfall patterns and even the timing for maize and wheat growing had become uncertain and contrary to what they have known over time. This clearly show that local farmers can identify with the fact that the changing climate is areal phenomena and has been experienced locally at their farm level. The poor maize crop shown in Figure 3 as captured during the field study help to illustrate what farmers are reporting in relation to losses they have incurred during their farming activities.

Figure 3.

Poor maize crop due to erratic rains in Kesses Sub-County of Uasin Gishu County, Kenya.

The finding in this study are similar to finding from [37] where farmers indicated that climate has been changing in the previous five years, in that rains start earlier, rains end latter and the maximum number of dry spells has increased. A comparison of the meteorological records with farmers’ assessment of climate change showed a large disparity, with few of the stated changes being evident in the long term record. It is apparent that changes occurring in local and global weather patterns as a result of the changing climate will challenge to a great extend the indigenous knowledge systems in weather prediction otherwise reliable for centuries. A great need for climate and weather information to be delivered to those engaged in farming activities to supplement indigenous knowledge systems is critical in mitigating this phenomena.

4.2 Factors influencing a farmer’s decision at the farm level in maize and wheat production in Uasin Gishu, county, Kenya

In an effort to understand what influences a farmers decision to commence an activity in the maize and wheat growing calendar (when to start land preparation, planting, type of crop to grow, weeding, top dressing, spraying of weeds, fungus or bacterial infection on crops and even harvesting), the following responses were gathered from the farmers:-Those who commence their activities just because those are the dates known to them through their experience over time in maize and wheat growing were 333 (84.9%). Another group of farmers 141 (36%) rely on looking around for “signs that rains are about to fall (wind direction from east to western side, cloud movement from eastern to western side of their farms and high sunshine intensity during the day and warm nights)”. Advice from the Ministry of Agriculture Livestock and Fisheries officials was mentioned by 29 (7.4%) of the farmers as having assisted them make decisions on the start of farming activities in their respective farms as shown in Figure 4.

Figure 4.

What factors influence a farmer’s decision at the farm level in maize and wheat production in Uasin Gishu, county, Kenya.

The results reveal a farming population solely dependent on their own indigenous knowledge systems and experience gained over time in maize and wheat growing to make certain farming decisions. The fact that 333 (84.9%) farmers commence their activities just because those are the dates known to them through their experience over time and another 141 (36%) looking around at some indigenous knowledge system indicators that rains are about to fall is a clear indication that framers do base their faming decisions on their own indigenous knowledge systems and experience gained over time. However, depending solely on indigenous knowledge indicators to predict onset dates of the rains and experience gained over time practicing maize and wheat creates a big challenge to the farmer especially with the prevailing changing climate that keeps on distorting seasons known to farmers. This makes farmers vulnerable and may incur much losses related to unpredictable weather patterns brought about by climatic changes. The findings in this study correspond with findings in [24] which asserts that in different parts of the world, farmers depending on rain-fed cultivation have developed complex cultural models of weather and may be able to cite local predictors of seasonal climate. Similarly, [13] argue that even when climate information is available, incorporation of scientific climate information into local decision making may not often occur because of the way such information is communicated and disseminated. There is need to identify clear channels of information delivery and also downscale the information to make sense for the farmer for effective uptake of such information.

4.3 Indigenous knowledge system indicators in rainfall prediction for farmers in Uasin Gishu County, Kenya

In an effort to understand what farmers use as their local indigenous knowledge indicators for weather prediction, farmers were asked to state what they thought were signs that rains were about to fall in their respective areas based on their experiences gained over time as maize and wheat farmers. The farmers who affirmed and believed that the real sign for rain commencement in their farms is when there are consistent lightning flashes around the Lake Region or Tindiret area in Nandi South as affirmed by 245 (62.7%). Those farmers who belief that heavy cloud cover and intense sunshine during the day was a sign that rains were about to commence were 178 (45.5%). The farmers relying on wind direction (eastwards and sometimes westwards) to alert them that the rains were about to commence were 159 (40.7%). Farmers who belief that very warm nights was a sign that rains were about to fall in their area were 195 (49.9%). Those who affirmed that in the recent years, it has been quite difficult to predict when the rains are about to fall were few farmers 42 (10.7%). The responses are as shown in Figure 5.

Figure 5.

Indigenous knowledge system indicators for predicting rainfall in Uasin Gishu County, Kenya. (*multiple responses - farmers were allowed to note as many reasons as were applicable).

The findings in this study reflect a farming community that has developed its own indicators in the traditional knowledge system on rainfall prediction. The set of indicators of rainfall used by farmers include consistent lightning flashes around the Lake Region or Tindiret area in Nandi South, heavy cloud cover and intense sunshine during the day, blowing of wind towards eastern side and sometimes westwards and very warm nights. During the key informant interview to supplement and corroborate some of the response, the following were identified as other key indicators used by farmers especially among the Nandi community and include examining the behavior of certain plants or trees to determine rainfall near onset. Among the plant species known include the tree Erithrina Abyssinica (Kakarwet). The tree starts flowering red with full leaves regained, Vernonia Auriculifera (Tepengwet) tree starts flowering, Flacourtia indica (Tungururwet) start budding is an indicator for rainfall near onset and farmers prepare to commence dry planting. The Fig tree (Ficus sycomorus) known locally by farmers as (Mogoiywet) start shading leaves is an indicator for rainfall near onset. One other small plant (herb) growing in thickets or bushes Scadoxus multiflorus of the Amaryllidaceae family (Ngotiotet) starts flowering red around March period and found in thickets or bushes near river banks is a real indicator for rainfall near onset.

Other indicators include migratory birds among them the White Stock (Kaptalaminik) moving or flying towards the north side signifying near rainfall onset and when flight changes towards the south, this signifies rainfall cessation. In addition to the indicators, the community had a unique way or prayer asking their God for rainfall or water for their domestic use during drought manifestation. According to the key informants interviewed, the traditional prayer song (ingoo) was sang by mature women and those at child bearing age at night and in several groups (3 to 4 groups of women) from various locations. They could join in the prayer song at night converging at one place in the open near a watering point away from homesteads as participants sang without cloths on them. They carried with them cooking items that include cooking sticks (mukanget ak Kurbet) and some finger millet flour (Beek ab Kipsongik). This form of prayers according to key informants and other farmers participating in this study all agreed that the women prayers at night were answered almost instantly there and then because before mid-night during the singing, heavy rains would fall on the singing women as they retreat back to their homesteads happy.

The study by [18] also supports the findings in this study and explain that majority of farmers prefer indigenous forecasting knowledge more than contemporary forecasting. The reasons being that indigenous information is more compatible with local culture and it has been tested, tried and trusted. In addition, it is more specific and is in a language that can be understood better by communities. There is a clear view here that it is important to document indigenous knowledge system used by maize and wheat farmers and ultimately integrate the knowledge with science based climate forecasting. Integrating both scientific and traditional knowledge will enable uptake and ownership of climate information by communities hence helping farmers avoid losses as a result of the changing climate. This is because farmers consider the fact that their perspectives have been taken into consideration when their indicators for predicting rainfall near onset have been taken into account by the weather forecasts. This creates ownership and sustainability in the long run.

The study finding compares well with findings by [38] explaining that in Uganda, indigenous knowledge forecasters associate the onset of rainfall with the appearance of clouds. The appearance of nimbostratus and cumulonimbus clouds indicates a high probability of rainfall. Biological indicators focus on the behavior and activities of domestic and wild animals, insects and different species of plants for weather forecasting. For instance, in Uganda, the Mvule tree indicates onset of the rainy season. In Kenya like Ethiopia, the intestines of sheep and goats are used to forecast about the magnitude, severity, and duration of drought, drought-affected places, disease outbreak, the prospect of peace, and/or conflict. In Tanzania, the occurrence of large flocks of swallows and swans, roaming from the South to the North during the months of September to November, is an indication of onset of short rains [38]. From the afore going, to effectively mainstream access to climate and weather information in key sectors, it is important to understand the local perspectives in relation to the changing climate and coping strategies in place. Ignoring this fact might hinder uptake of weather forecasts by the farmers hence leaving them completely vulnerable to the changing climate as they would rely only on their indigenous knowledge systems to mitigate prevailing weather conditions.

4.4 Do famers belief in the use of climate information/weather forecasts or their indigenous knowledge system indicators in rainfall prediction

Farmers were further tested on whether they relied on the use of climate and weather information or their indigenous knowledge/practice in their farming decisions and the responses were varied as outlined in Table 1.

Statement	All the time	Most of the time	Sometimes	Not at all
Often belief in seasonal climate and weather information	68 (17.3)	80 (20.4)	219 (55.7)	26 (6.6)
Often rely on experience/indigenous knowledge in maize production	89 (22.7)	148 (37.8)	141 (36)	14 (3.6)
Often rely on experience/indigenous knowledge in wheat production	85 (22.8)	163 (43.7)	111 (29.8)	14 (3.6)

Table 1.

Belief in use of climate and weather information or indigenous knowledge among farmers in Uasin Gishu County.

More than half of the farmers 219 (55.7%) sometimes belief in seasonal climate and weather information, 163 (43.7%) rely on experience/indigenous knowledge in wheat production most of the time and 148 (37.8%) rely on experience/indigenous knowledge in maize production most of the time as outlined in Table 1. The results show a group of farmers who tend to belief in the use of climate and weather information to conduct their business of maize and wheat crop growing however, they remain influenced by their traditional knowledge system which they have relied upon over the years. This arrangement becomes very challenging at present with the prevailing changing climate. There is need to balance access to the two sets of information by farmers as stated by farmers in this study. The findings in the study agree with those by [13] in a paper on community based adaptation in climate change which reveal that farmers in Burkina Faso traditionally rely on observation of environmental indicators to predict climate patterns, but they lost confidence in their ability to predict rainfall given the increased climate variability and increasingly seek to incorporate meteorological forecasts/climate information.

To understand further the usage of climate and weather information and the influence of farmer’s indigenous knowledge systems in maize and wheat production over time, indicators used in rainfall prediction were examined together with usage of climate and weather information as shown in Table 2. The results portray existence of a significant relationship between indigenous knowledge system indicators for rainfall prediction and the use of climate and weather information in maize and wheat production activity as shown in Table 2.

Indigenous Knowledge indicators	Climate & Weather Information Use		Chi	P
	Yes No
Lightning flashes around lake region/Tindiret area in Nandi	110 (44.7%)	136 (55.3%)	0.476	0.490
Cloud cover and intense sunshine	70 (39.1%)	109 (60.9%)	6.391	0.011
Wind direction	83 (52.2%)	76 (47.8%)	4.059	0.046
Very warm nights	70 (35.7%)	126 (64.3%)	16.832	<0.001
Difficulty in prediction recently	23 (53.5%)	20 (46.5%)	1.074	0.300

Table 2.

Relationship between climate information and indigenous knowledge indicators usage in rainfall prediction in Uasin Gishu County, Kenya.

A higher proportion of the farmers relying on their indigenous knowledge system indicators to predict rainfall hence making decision on farming activities do not use climate and weather information as indicated in Table 2. A greater negative influence is created on utilization of climate information by farmers if they are left alone to decide whether to embrace climate information or stick to their indigenous knowledge systems in their farming decisions. Farmer’s indigenous knowledge indicators and experience gained over time practising maize and wheat growing influence a lot their ability to use climate information. There is need therefore for the County Directorates of Agriculture in Kenya to undertake greater sensitization meetings during farmers village meetings, agricultural shows and farmer’s field day outlets to educate farmers on the need to consider using climate information in addition to their great local knowledge and experience learned over time in farming decisions as this will cushion them against extreme weather variations that may lead to crop losses hence impacting negatively on their livelihoods.

The traditional weather forecasters form part of the decision making process sometimes when farmers experience delayed onset of rainfall when they look at their farming calendar. Farmers listen to what known traditional weather forecasters can deduct from reading the arrangement of stars in the solar system and the “positioning of male and female star”. During the key informant interview with the traditional elder shown in Figure 6, he predicted that “there was going to be plenty of rain in December 2013 and that rains would continue to March 2014. He advised the farmers to grow short duration crops like beans that take 3 months to mature”.

Figure 6.

Explanation of weather phenomena by a traditional weather forecaster in Moiben Sub County of Uasin Gishu, Kenya.

From the diverse responses by the farmers, it is evident that indigenous knowledge system plays a crucial role in farmer’s decision making process; ultimately, it has a significant impact on their activities. With changing climate being real in the region, farmers relying on “known signs” or meteorological indicators derived in their traditional knowledge systems are at a greater risk of losing out on benefiting from prevailing positive conditions or avoiding bad weather leading to crop loss. Farmers and key stakeholders in the agricultural sector need to incorporate climate and weather information or weather forecasts in their decision making process. Indigenous knowledge weather prediction methodologies are now facing serious challenges related to environmental degradation and interference of the natural ecosystem balance by man. According to the Director of Meteorology in Uasin Gishu County and some traditional weather forecasters interviewed, most indigenous trees have disappeared completely and are being replaced by exotic trees which are alien to indigenous weather predictors. They have thin leaves thus their ability to sequestrate carbon in the atmosphere is reduced hence does not help much in arresting greenhouse gasses. Burning of farms has not only destroyed micronutrients and shrubs but also destroy insects and their migratory paths which traditional weather forecasters use to predict near onset of rains.

Some bird species have migrated elsewhere where they can still find a natural habitat hence rendering traditional weather forecasters using birds to predict onset of rains helpless. It is however important to integrate both scientific and traditional knowledge system as the scientific forecast diverges from traditional farmer’s prediction in scale and to some extent on predictors [30]. In any case, some of the principles of the predictors like wind flow, temperature changes converge with the scientific forecast. Farmers have been using combination of various biological, meteorological and astronomical indicators to predict the rainfall. While the scientific forecast are developed using the predictors such as wind, sea surface temperatures and others which are primarily meteorological indicators.

4.5 Value attached to utilization of climate and weather information

If weather forecasts/weather and climate information were to be integrated with indigenous knowledge system for synergy prediction among farmers, it is critical to understanding how such farmers perceive as value attached to utilization of such climate and weather information in relation to their maize and wheat growing. Farmers were asked to rank the value they attached to climate and weather information use in their farming enterprise and their responses were captured either as very important, important, unsure, and not important. As indicated in Figure 7, more than half of the farmers view climate and weather information as very important in both maize and wheat growing 215 (54.8%) and 206 (54.4%) respectively. Similarly, a large number of farmers still considered the same information as important 143 (36.7%) for maize growing and 136 (36.1%) for wheat growing.

Figure 7.

Value attached to utilization of climate and weather information in Uasin Gishu County, Kenya.

As seen earlier farmers in this study have agreed that climate and weather information is important but whether this translates into action or use remains to be tested. A mere belief in some variable (climate and weather information) does not necessarily translate into adoption of the same especially where traditional knowledge systems influence decisions as seen elsewhere in the results of this study. This calls for greater efforts to educate farmers to be able to embrace weather forecasts as well in their farming decisions as mere belief in it would not alter losses they will incur especially with the prevailing climate variability which has distorted order of activities and seasons known to farmers in maize and wheat growing areas of Kenya. It is important to emphasize that climate information or weather forecasts is key to understanding climate as a major influence on lives, livelihoods, resources, ecosystems and development. According to [39], weather forecasts supports decision making on which option to invest in, when and how much to invest. Flexible and proactive planning enabled by climate information helps vulnerable communities especially the farming communities to cope with the risks inherent. It provides a way of analyzing the nature and scale of impacts due to past and current climate and the potential future impacts as the climate continues to vary and change. In the end, actors can then make informed and appropriate decisions and plans to deal with climate-related impacts through adaptation, risk reduction and development actions.

5. Conclusions

Farmers in Kenya and some countries in eastern and southern Africa who depend on rain-fed cultivation have developed complex cultural models of weather prediction and are able to cite local indicators of seasonal weather. From this study majority of farmers prefer indigenous forecasting knowledge because they have a belief that indigenous information is more practical as it has been tested, tried and trusted over time.

The dependency on traditional indicators alone without much uptake of available weather forecasts can and has predisposed farmers to crop losses as they try to navigate the risks of changing climate. Although farmers do trust climate and weather information, its uptake will require more sensitization and demonstration if farmers are to navigate risks associated to the changing climate in the farming enterprise.

There is need to integrate both traditional knowledge systems and weather forecast information for synergy and reliable weather prediction. This will act as a motivator to farmers to embrace the use of climate and weather information because their traditional knowledge system perspectives have been incorporated hence creating ownership and sustainability.

The changes on the environment associated to environmental degradation; ecosystem disturbance and changing climate has seen some important traditional predictor indicators disappear or lost completely from the environment. Integrating both scientific and traditional knowledge system will supplement the loss on traditional indicators of rainfall prediction that used to support farmer decisions hence mitigating against losses that are bound to occur. It is important to note that some of the principles of the traditional predictors like wind flow, temperature changes converge with the scientific forecasts.

A national policy guideline on a closer working relationship should be established between the farmers, agricultural extension officers and meteorological scientists including traditional weather predictors. A strong link, including feedback loops between scientists, advisory agents and farmers will help in communicating downscaled climate information and facilitating access by local farmer communities. Format, delivery mode and timing of the information is key. All key stakeholders in agriculture, livelihoods and climate change sectors during their participatory weather scenario planning action should be aware of farmers needs whether they are real or perceived because farmers do know what information they need at particular point in time at their local farm level.

5.1 Recommendation

This study suggests that there is need to integrate both traditional knowledge systems and weather forecasts at local farmer levels for synergy forecasting that will generate reliable climate information useful to farmers. Incorporating farmers perspectives by slightly modifying and using them to meet current needs and situations will help address the needs of farmers and may be a motivator in embracing the use of climate and weather information as their traditional knowledge system perspectives have been incorporated. Climate information services must be embedded in local, national and regional processes to enable scaled-up support for widespread adaptation activities.

Acknowledgments

Gratitude to University of Nairobi and IDRC Collaborative project on Innovative Application of ICTs in Addressing Water-related Impacts of Climate Change (ICTWCC) in The School of Computing and Informatics and the Department of Land Resource Management and Agricultural Technology (LARMAT) for providing the much needed research grants led by Prof. Timothy Waema and team members Dr. laban Mc’Opiyo and Anuradha Khodha. Prof. Grace Cheserek, Professor Gilbert Nduru and Dr. Victor Odenyo for their valuable consultation.

Conflict of interest

The Author declare that there is no conflict of interest to state. I certify that the submission is original work and is not under review at any other publication.

Thanks

Thank you Lord for granting me abundant life and good health in all times. Thanks to Professor Vincent Sudoi and University of Eldoret for the space and moral support. Kiprop Murgor, Chemutai Murgor and Brian Murgor for simply their loyalty.