Do indigenous forecasts and scientific forecasts influence arable farmers’ and agro-pastoralists’ estimation of onset and cessation of rains? Empirical evidence from Rwenzori region, Western Uganda

https://doi.org/10.1016/j.agrformet.2019.107667Get rights and content

Highlights

  • There is a negative relationship between using IF only and estimation of onset and cessation of rains.

  • There is a positive relationship with using both SF and IF and estimation of onset and cessation of rains.

  • Onsets of rains are highly variable implying climate variability while cessation of rains are not highly variable.

Abstract

This study investigated the influence of indigenous forecasts (IF) and scientific forecasts (SF) on arable farmers’ and pastoralists’ estimation of onset and cessation of rains in the Rwenzori region of Western Uganda. Daily rainfall data (1970–2010) from 3 weather stations in different agro-ecological zones was used to determine mean onset and cessation dates using the Sivakaumar method. After obtaining the household survey of number of farmers and pastoralists whose estimation of onset and cessation dates were in agreement with the stations’ dates, the relationship between forecast type used and estimation of the dates was investigated using the Pearson correlation coefficient. The results indicate that onset dates for the 1st season were on 11th, 17th, 19th March while cessation was 8th, 11th, 21st June for lowlands, mountainous and forested areas, respectively. For the 2nd season, forested and mountainous areas had their onsets on 3rd and 20th August, respectively but the lowland area their onset on 3rd September. The cessation date for 2nd season for forested areas was on 8th December, while dates for lowland and mountainous areas were 17th and 24th November, respectively. The results suggest that there is positive relationship between using both IF and SF and estimation of onset and cessation dates but negative relationship with IF had for arable farmers. There was strong negative relationship between using both IF and SF and estimation of onset dates for the 2nd season for arable farmers in forested areas, but strong positive relationship with those using IF only season. The coefficients of variation for the rain onsets were high implying climate variability. The adopted Sivakaumar method is easy to operationalize by farmers and pastoralists with primary level of education, and access to daily rainfall data from established institutions like schools, local clinics and facilities provided by agricultural extension workers.

Introduction

Although the distribution of seasonal climate forecasts (SCFs) by national meteorological services has improved in the recent past with establishment of regional climate outlook fora in the developing world (Ogallo, 2010; Tall, 2010), arable farmers and pastoralists have not made good use of forecasts for onset and cessation rains (Goddard et al., 2010). This has contributed to their vulnerability to the effects of rainfall seasonality change that include loss of crop harvests. The low use of SCFs has negative implications on the choice of crop enterprises that match the length of the rain season for arable farmers engaged in annual crop production. Furthermore, low use of SCFs may result into poor livestock breeding outcomes for pastoralists (such as reduced milk yields, livestock mortality and reduction in herd size) which would adversely affect their livelihoods. Overall, these negative effects due to poor use of SCFs have huge negative social and economic implications on rural development in developing countries.

Climate variability manifests itself in various forms including unpredictable onset and cessation of rains. Increased variability could have negative impacts on rural households, such as through poor crop yields and pasture scarcity which might induce food insecurity and reduced household incomes (Simelton et al., 2013). Pastoralists and arable farmers use indigenous forecasts (IFs) and scientific forecasts (SFs) to predict onset and cessation of rains to reduce their vulnerability to rainfall seasonality (Kalanda-Joshua et al., 2011; Okonya and Kroschel, 2013; Orlove et al., 2010). Several studies have reported that climate information on onset of rains is considered as one of the most important pieces of information among pastoralists and arable farmers in seasonal climate forecasts (Luseno et al., 2003; Speranza et al., 2010). Forecasts on onset of rains are used by pastoralists in their seasonal calendar events such as herd mobility while arable farmers use the climate information for planning of their crop enterprises under rain-fed agriculture (Roncoli et al., 2002; Speranza et al., 2010). Several onset and cessation of rains studies (Marteaua et al., 2011; Mugalavai et al., 2008; Omotosho et al., 2000; Sivakumar, 1988) have focused mainly on determining onset and cessation dates. However, there is very little in the literature about the association between pastoralists’ and arable farmers’ estimation of onset and cessation of rains and the forecast type used.

There are two main types of definition according to onset of rains literature, namely meteorological onset and agro-climatic onset. The agro-climatic onset has been defined as the optimal date when there is sufficient soil moisture during planting and early growing periods to avoid crop failure and without immediate dry spell of 7 days in the next 30 days (Marteaua et al., 2011). This implies that having long dry spells after a rain down pour does not persuade farmers to plant due to the likelihood of crop failure. On the other hand, meteorological onset refers to the first wet day after a dry season with 20 mm of rainfall (Marteaua et al., 2011). This implies that one or two down pours with a total amount of 20 mm symbolizes onset even if there is a dry spell thereafter. Obviously farmers are uncomfortable with meteorological onsets as a basis of crop planting decisions. Agro-climatic onset is associated with crop performance after first rain days with sufficient available water in the root zone while meteorological onset is associated amounts of the first rains (Marteaua et al., 2011). For this study, the agro-climatic definition using historical daily rainfall data was used which relates to crop production. Several onset and cessation studies have taken into consideration crop production in their analysis (Marteau et al., 2011; Mugalavai et al., 2008). Late onset and early cessation lead to reduced growing season which results in reduction of crop yields. Plants have critical periods (such as the grain filling stage for maize) when there is high need for water (Rao et al., 2011). Insufficient supply of water at the critical crop growing stages results in poor crop yield or total failure in extreme cases (Rao et al., 2011; Traore et al., 2013). Occurrence of long dry spells of 5–20 days after planting lead to poor germination and reduction in crop yields of 40% for cotton and maize (Traore et al., 2013). Cessation was defined as last date of the rain season with no rain for a period of 20 days and there is no useful contribution of available water to root zone of the crop sown (Sivakumar, 1988). The word useful that implies the available water can be utilized by the plants after considering water losses due to evapotranspiration and loss from the soil due to prevailing maximum temperatures in the agro-ecological zones. Thus, for this study, the agro-climatic approach was adopted.

Arable farmers and pastoralists have not made good use of SFs for onset and cessation of rains due to its shortcomings in saliency, credibility, trust and legitimacy (Cash et al., 2006, 2003; McNie, 2007; Patt and Gwata, 2002). Furthermore, Goddard et al. (2010) indicated that poor spatial resolution, temporal resolution and local specificity of SFs were major areas of concern to end users. These shortcomings are attributed to the low meteorological station density in African countries including Uganda (Medany et al., 2006; UNECA, 2011). The other constraints of national meteorological systems in Africa include dysfunctional stations and poor maintenance of equipment (Snow et al., 2016; UNECA, 2011). With these challenges of national meteorological services, some farmers have lost confidence in SFs and are using indigenous forecasts (IFs) because of their perceived good spatial and temporal resolution. However, there are some initiatives by the United Nations Development Programme (UNDP) that are intervening on the issue of low station density in Africa through provision of automatic weather stations to countries (Snow et al., 2016).

The Greater Horn of Africa Climate Outlook Forum in conjunction with Uganda National Meteorological Authority (UNMA) provides SCFs for arable farmers and pastoralists. UNMA downscales SCFs which include onset and cessation of rains every February and September for Rwenzori region (UNMA, 2017). UNMA networks with radio and television stations in the dissemination of the forecasts. Despite access to SCFs, arable farmers and pastoralists in Rwenzori region use IFs in estimating onset and cessation of rains (Nganzi et al., 2015; Okonya and Kroschel, 2013). In the Rwenzori region (Fig. 1), the low confidence in SFs has been attributed to the low station density.

Several studies among arable farmers and pastoralists in Africa have demonstrated the influence of IFs and SFs on the estimation of onset and/or cessation of rainfall. Examples include Luseno et al. (2003) in Kenya;Ingrama et al. (2002); Roncoli et al. (2002), Roncoli et al. (2008) in Burkina Faso; and Nanja (2010) in Zambia. Several other studies have reported that indigenous knowledge forecasts are used in predicting onset and cessation of rains in Australia (Green et al., 2010), Kenya (Speranza et al., 2010), Senegal (Roudier et al., 2014) and Uganda (Orlove et al., 2010).

It is thus evident from the literature reviewed that IFs and SFs influence the estimation of onset and cessation of rainfall by African farmers and pastoralist. However, there is still a knowledge gap on the whether the relationship of the forecast type is positive or negative? Thus, the research question addressed by this paper is: ‘Does the forecast type used by pastoralists and arable farmers have positive or negative relationship with their estimation of onset and cessation of rains? Therefore, the paper presents findings of a study that investigated the relationship between using IFs and/or SFs on arable farmers’ and pastoralists’ estimation of onset and cessations of rains in Rwenzori region in Western Uganda. The aim is to give empirical evidence that can be used by agricultural and rural development stakeholders to improve the resilience of arable farmers and pastoralists to climate variability and improve their preparedness for the rainy seasons under rain-fed agriculture. The study contributes to the international literature that relates to the use of forecasts to the estimation of onset and cessation of rains in the developing world.

Section snippets

The study area

The study was done in Rwenzori region in Western Uganda that has mountainous, lowland, wetland, forested, mountainous and forested agro-ecological areas (Fig. 1). Orographic features in the study area include Mountain Rwenzori, natural and planted Forest Protected Areas (FPAs). Literature shows that topographic features such as mountains1

Results

The results indicate that onset dates for the 1st season were on 11th, 17th, 19th March while cessation was on the 8th, 11th, 21st June for lowlands, mountainous and forested areas respectively. For the 2nd season, forested and mountainous areas had their onsets on 3rd and 20th August respectively, but lowland area had its onset on 2nd September (Table 1). Based on FGDs, it became evident that IFs and SFs were in general agreement in terms of predicting onset of rains for the 2nd rainy season.

Discussion

The findings are in agreement with an earlier study done in Western Uganda (Breytenbach, 2013) that revealed that onset of rains in neighborhoods of Kibale National park was in March for the first season and August for the second season. Earlier research elsewhere has also revealed that onset of rains in the equatorial tropics was in the month of March (Amekudzi et al., 2015; Odekunle, 2004) in Ghana and Nigeria, cessation of rains took place from October to December depending on the location

Conclusion

The study findings suggest that there was a negative relationship between using IF only and estimation of onset and cessation of rains but positive relationship with using both SF and IF for arable farmers. There was a strong negative relationship between using both IF and SF and estimation of onset dates for the 2nd season for arable farmers’ in forested areas but strong positive relationship with using IF only. The coefficients of variation for the onsets of both 1st and 2nd seasons for

Acknowledgement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Family savings of the lead author were used for data collection and special thanks to Nassali Mercy Nkuba for moral support.

References (62)

  • E.M. Mugalavai et al.

    Analysis of rainfall onset, cessation and length of growing season for western Kenya

    Agric. For. Meteorol.

    (2008)
  • L. Ogallo

    The mainstreaming of climate change and variability information into planning and policy development for Africa

    Procedia Environ. Sci.

    (2010)
  • A. Patt et al.

    Effective seasonal climate forecast applications: examining constraints for subsistence farmers in Zimbabwe

    Glob. Environ. Chang. Part A

    (2002)
  • P. Roudier et al.

    The role of climate forecasts in smallholder agriculture: Lessons from participatory research in two communities in Senegal

    Clim. Risk Manage.

    (2014)
  • M. Sivakumar

    Predicting rainy season potential from the onset of rains in Southern Sahelian and Sudanian climatic zones of West Africa

    Agric. For. Meteorol.

    (1988)
  • A. Tall

    Climate forecasting to serve communities in West Africa

    Procedia Environ. Sci.

    (2010)
  • B. Traore et al.

    Effects of climate variability and climate change on crop production in southern Mali

    Eur. J. Agron.

    (2013)
  • C.D. Ahrens

    Meteorology Today: An Introduction to Weather, Climate, and the Environment

    (2012)
  • R. Allen et al.
    (1998)
  • L.K. Amekudzi et al.

    Variabilities in rainfall onset, cessation and length of rainy season for the various agro-ecological zones of Ghana

    Climate

    (2015)
  • R.G. Barry et al.

    Atmosphere, Weather and Climate

    (2009)
  • P. Bauer et al.

    Effect of seeding date, plant density, moisture availability, and soil nitrogen fertility on maize kernel breakage susceptibility1

    Crop Sci.

    (1986)
  • E. Breytenbach

    Following the Rains: Evidence and Perceptions Relating to Rainfall Variability in Western Uganda

    (2013)
  • J. Byakatonda et al.

    Analysis of rainfall and temperature time series to detect long-term climatic trends and variability over semi-arid Botswana

    J. Earth Syst. Sci.

    (2018)
  • P. Camberlin et al.

    Components of rainy seasons’ variability in Equatorial East Africa: onset, cessation, rainfall frequency and intensity

    Theor. Appl. Climatol.

    (2009)
  • D.W. Cash et al.

    Countering the loading-dock approach to linking science and decision making: comparative analysis of El Niño/Southern Oscillation (ENSO) forecasting systems

    Sci. Technol. Human Values

    (2006)
  • D.W. Cash et al.

    Knowledge systems for sustainable development

    Proc. Natl. Acad. Sci.

    (2003)
  • W.G. Cochran

    Sampling Technique

    (1963)
  • S.U. Ewansiha et al.

    Performance of cowpea cultivars when grown as an intercrop with maize of contrasting maturities

    Arch. Agron. Soil Sci.

    (2014)
  • A. Farrow et al.

    Assessing the risk of root rots in common beans in East Africa using simulated, estimated and observed daily rainfall data

    Exp. Agric.

    (2011)
  • D. Green et al.

    Indigenous Australians’ knowledge of weather and climate

    Clim. Change

    (2010)
  • Cited by (6)

    View full text