Oluwatimilehin and Ayanlade  Agric & Food Secur (2021) 10:3 Agriculture & Food Security https://doi.org/10.1186/s40066-020-00275-5 RESEARCH Open Access Agricultural community‑based impact assessment and farmers’ perception of climate change in selected Ecological Zones in Nigeria Isaac Ayo Oluwatimilehin and Ayansina Ayanlade* Abstract Background: The impacts of climate change are affecting sustenance and livelihood of many rural farmers in Africa. The majority of these farmers have low adaptive capacity. This study investigates climate change impacts, farm- ers’ perception, adaptation options and barriers to adaptation in three selected ecological zones in Nigeria using three staple crops. Rainfall and temperature data of over 35 years were analysed using ANOVA, Mann Kendall and Sen’s Slope Analysis. Farmers’ perception of climate change and cropping experiences were assessed with the aid of a well- structured questionnaire, semi-structured interview and focus group discussion. Results: The results of the study revealed high variability in the annual and monthly rainfall and temperature during the study period. The highest annual maximum temperature was recorded in Kwara with Tmax > 32 ℃. Though, there appeared to be spatial and temporal variations in rainfall in the study area, the highest was in Ogun with mean annual rainfall = 1586.9 mm and lowest in Kwara with mean annual rainfall = 1222.6 mm. Generally the Mann Kendall and Sen’s slope analysis revealed general increase in the minimum and maximum temperature, while rainfall revealed generally downward trend. The study revealed a difference in farmers’ perception but nearly 74% of farmers perceived that climate is changing, which is affecting their farming activities. Nearly 70% claimed that lack of financial capital is the major barrier to climate change adaptation. Conclusions: The study concludes that rainfall and temperature variability have significantly impacted cropping and that farmers are aware of long-term changes in temperature and rainfall, but some are unable to identify those changes as climate change. There is a need for affordable and available improved seedlings and variety of crops that can adapt to climate change conditions. Keywords: Climate change, Farmers’ perception, Community-based impacts assessment, Nigeria Background through change in weather events: windstorms, Climate change and associated impacts have been droughts, rainstorms/floods and dust storms which are reported to be capable of posing imperative short-term becoming more frequent and severe [5–7], while other and long-term impacts on crop production and food studies showed that changes in land use and land surface security. The impacts of climate change is evident on temperature, sea level rise, and coastal erosion is already livelihood of many, especially the rural farmers in Africa presenting significant long-term challenges of climate who depend on rainfall for crop cultivation [1–4]. Some change [8–11]. studies have revealed that climate change is manifested Climate impacts crop production with a direct effect noticeable in quality and quantity of yield which is usually *Correspondence: aayanlade@oauife.edu.ng; sinaayanlade@yahoo.co.uk adverse. Among the elements of climate, rainfall and tem- Department of Geography, Obafemi Awolowo University, Ile‑Ife, Nigeria perature are the major determinants for crop production © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​ mmons​.org/licen​ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat​iveco​mmons​.org/publi​cdoma​in/ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 2 of 17 [12, 13]. Thus, climate change has become a menace to value range is between 1400 mm and 1500 mm with a economies of different countries, especially the agrarian relatively high temperature of an average of 30 ℃ [14, 15]. nations. Since the world attention was drawn by Fourier Kwara is located between latitudes 7°45 ′N and 9°30 ′N in 1824, climate change has been an academic discussion and longitudes 2°30 ′E and 6°25 ′E East [24]. Kwara state of vehemence that is inevitable and has been established (guinea savanna) lies within the tropical climate and it to portend threats to agriculture. With many economies is characterized by double rainfall maxima with tropical depending on agriculture, most especially the develop- wet and dry climate with the seasons lasting for about 6 ing countries, the change affecting agriculture invariably months each with annual rainfall ranging from 1000 to affects the economy of the developing countries [14, 15]. 1500 mm and annual mean temperature ranging between Understanding meteorological conditions is very essen- 30° and 34 °C [24]. The study area is shown in Fig. 1. tial [4, 16, 17], because natural variability plays impor- The majority of people of Ondo state are into agricul- tant role in climate which affects agricultural activities. ture with high production of both tuber and root crops. Certain weather experiences are becoming more promi- The tuber crops include yam, cassava, cocoa yam etc. nent and severe with extended odds and natural limits and grain crops like maize and rice. Also cash crops like associated with climate change [18]. According to Food cocoa, coffee, plantain, banana, kolanut, palm oil, among and Agricultural Organization [19], rain-fed agriculture others are also grown. Also, there is a large scale pro- employs the majority of the labour force in Sub-Saharan duction of timbers because of the large forest reserves Africa, which is about 70% of the total population and of about 2008sq km which yields timbers for furniture, accounts for 25% Gross Domestic Product (GDP) in the fuel wood and industrial uses [25, 26]. Ogun state is region and responsible for the nearly 90% of Sub-Saharan also agro based state in which agriculture has employed Africa food production. Thus, forming the major activity many especially those in the rural areas. Arable crops like of livelihood for about 70% of the population [19]. The maize, rice, yam, cassava, cocoyam, groundnut, melon, current change in climatic conditions are becoming obvi- banana, plantain, oranges etc. are grown [25]. Agricul- ous especially in areas where rain fed agriculture is domi- ture is the major occupation of the people of Kwara with nantly practiced. many into food crop (e.g., cereals like rice, maize, millet Despite the effect climate change portends to agricul- and tuber crops like yam, cassava, cocoyam etc.) cash ture, there are limited studies on climate change, percep- crop production like cocoa, coffee, kolanut, palm pro- tion of farmers and their cropping experiences generally duce are also grown [24]. in Nigeria and other part of Africa [20–22]. The pre- sent study focuses on three (3) different states in Nige- Data acquisition and analysis ria which falls within three different ecological zones by This study employs both primary and secondary data. assessing perception of farmers on climate, their crop- The primary data was sourced through Semi-Structured ping experiences and how it has impacted the yield of Questionnaire, Key Informant Interview (KII) and Focus the selected staple crops (maize—zea mays, rice—oryza Group Discussion (FGD) administered to farmers in sativa and cassava—manihot esculenta) which have not the selected agricultural settlements from the selected been well-documented in literature. ecological zones to assess their perception on climate, cropping experiences, level of adaptation and barriers to Materials and methods practice of adaptation. For this study a total of 180 cop- Study area ies of a questionnaire, 6 Key Informant Interviews and 3 The study covers three states: Ondo in rainforest ecologi- Focus Group Discussions were used. 60 copies of a ques- cal zone, Ogun in freshwater and Kwara in guinea savan- tionnaire, 2 KII and a Focus Group Discussion were done nah (Fig. 1). Ondo State is located between longitudes at each of the selected states belonging to different eco- 5° 5′ and 5° 83′ E; and latitudes 7°10 ′N and 7°15 ′N [23]. logical zone which was further divided into three com- Ondo State is of lowland tropical rain forest having two munities. To ensure data quality, data were sourced from distinct seasons of wet and dry with mean temperature farmers who are above 45 years of age. Climatic data of ranging between 27 ℃ in the south and 30 ℃ in the north 37 years (1982–2017) including rainfall, minimum and and mean relative humidity is over 75% [24]. The rain falls maximum temperatures were sourced from the archive throughout the year with relative dryness from Novem- of Nigerian Meteorological Agency (NiMet) Oshodi. This ber to January with annual rainfall of 2000 mm. Ogun data was daily data collected from the synoptic stations state is located between latitudes 6°12′0 ″N and 7°47′60 located at Akure, Abeokuta and Ilorin in Ondo, Ogun ″N and longitudes 3°0′0 ″E and 5°0′0 ″E [25]. Ogun State and Kwara states, respectively. NiMet carried out the (freshwater) is typical of tropical climate consisting of data recording at these stations using the British Stand- two different seasons of wet and dry. The annual rainfall ard Rainguage and Dine’s tilting siphon rainfall recorder Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 3 of 17 Fig. 1 Map of the study area showing weather stations and sample points for rainfall and thermometer. These three stations were Mann Kendall (MK) and Sen’s slope analysis was used in selected first because of their spatial location in different this study to assess the trend and the slope of climatic data ecological zones and second, connectedness. Crop yield in the study area. MK test is a widely used technique for data was sourced from the archive of Agriculture Devel- the detection of trends in climatic time series [27]. This test opment Programme (ADP) offices in Ondo, Ogun and is used for assessing the monotonic upward and downward Kwara states. The yield is calculated by dividing the total trend of climatic variables derived as given in Eqs. 1–4. output of the year by the land area, measured in metric Also, the Thiel Sen’s slope non-parametric test was done to tons per hectares. assess the trend in climatic time series and it is calculated The primary data from the questionnaire was analysed as shown in Eqs. 5, 6. using tables, charts and content analysis, the secondary n−1 n data (climatic data) was analysed using ANOVA, Mann   S= i=1 j=i+1 sig Xj − Xi (1) Kendall and Sen’s Slope Test. ANOVA test was carried out to show the difference in the mean of the climatic param-  � � eters in the three stations in the study area. One-way � �  +1 if �Xj − Xi � > 0 ANOVA assume that a group of data is derived from nor- sig Xj − Xi = 0 if �Xj − Xi � = 0 (2) mal distribution with a roughly constant variability within −1 if Xj − Xi < 0  the group. Statistically, one-way ANOVA is a simplification of the two sample t-test with F statistic comparing the vari- ability that exists between and within the groups. Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 4 of 17 1 q    Table 1 Annual climate variability in the study area V (S) = [n(n − 1)(2n + 5) − tp tp − 1 2tp + 5 18 p=1 Variable N Mean ± SD df F P (3)  Minimum temperature  √ S−1 if, S > 0  VAR(S) 2 148.87 0.00 Z= 0 if S = 0 (4) Ondo 36 21.78 ± 0.59 √ S+1 if, S < 0   VAR(S) Ogun 36 23.38 ± 0.39 99 Kwara 30 21.64 ± 0.38 In these equations, the chronological order of the Maximum temperature observations is Xi and Xj series, where n is the length of 2 31.97 0.00 time series, tp is the number of ties for pth value, and q is Ondo 36 31.34 ± 0.71 the number of tied values. Positive Z values indicate an Ogun 36 31.55 ± 0.07 99 upward trend in the climatic time series. If /Z/ > Zi-α/2, Kwara 30 32.38 ± 0.08 ­(H0) is rejected and a statistically significant trend exists Rainfall in the climatic time series. The critical value of /Z/ > Zi- 2 15.77 0.00 α/2 for a p value of 0.05 from the standard normal table is Ondo 36 1484.5 ± 232.7 1.96. A value of Z greater than 1.96 confirms an increas- Ogun 36 1586.9 ± 284.4 102 ing trend, while a negative value smaller. Sen’s slope non- Kwara 33 1222.6 ± 309.0 parametric method was used to analyse the slope of the time series data: xj − xk Ogun (23.38 ℃), while it was lowest in Kwara (21.64 ℃). T1 = (5) The results of ANOVA for the minimum temperature j−k showed that the minimum temperature varied spatially In this equation, xj and xk represent data values at time and temporally in the stations. The variation in mini- j and k, respectively: mum temperature can be explained by several factors.  The temperature is lowest in Ondo state as revealed by T(N + 1) N is odd   /2 statistics. Qi 1 . (6) The maximum temperature patterns showed spatial � �  2 TN + T(N + 2) N is even changes in annual maximum temperature also. The high-  /2 /2 est was recorded in Kwara (32.38 °C) and lowest in Ondo Multiple regression statistics was done to find the com- (31.34 °C). The maximum temperature is expected to be bined impact of climate on each of the selected crops. highest in the northern part of the country and lowest in With the use of climatic (temperature and rainfall) data, the southern part and in agreement with the ecological impact of climate on yield was established. The formula distribution of the vegetation zones in the country. The for regression is as shown in Eq. 7. rainfall also varied spatially and temporally as shown by  2      the analysis. It is highest in Ogun (1586.9 mm) and low- x2 x1 y − x1 x2 x2 y b1 =  2  2   2 , (7) est in Kwara state (1222.6 mm). x1 x2 − x1 x2 The crop yields also vary over time and space. Figure 2 reveals the change in the yield of maize, rice and cas- where x1 = Independent variable 1 (temperature); sava. For maize, the yield was relatively stable for some x2 = Independent variable 2 (rainfall); y = Dependent var- years and significantly increased in 2013 in Ondo state. iable (crop yield). The year 2013 was marked with rainfall below or about normal with minimum and maximum temperature above normal in all stations. For Ogun and Kwara states, Results and discussion there is a slight fluctuation in the yield of maize. Rice on Climatic variation and crop yields in the selected ecological the other hand was only steady in Ogun state but fluc- zones tuation was observed in the yield of rice for Ondo and The results reveal significant difference in the annual Kwara throughout the study period. Cassava yield has minimum temperature [f (2, 99) = 148.87, P < 0.05], maxi- an upward trend in all stations though with fluctuations mum temperature [f (2, 99) = 31.97, P < 0.05] and rain- but there is a sharp drop in the yield of cassava in Ondo fall [f (2, 102) = 15.77, P < 0.05] which are all statistically state since 2014. Increase in land devoted to maize and significant (Table 1). The overall minimum temperature rice production may bring no corresponding increase in spatial behaviour unveiled spatial changes in annual yield as rainfall amount and frequency during the period minimum temperature. The highest was recorded in Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 5 of 17 4.0 7 Ondo Ondo 3.5 Ogun 6 Ogun Kwara Kwara 3.0 Rice Yield (MT/Hac) 5 Maize Yield (MT/Hac) 2.5 4 3 2.0 2 1.5 1 1.0 0 0.5 1985 1990 1995 2000 2005 2010 2015 2020 1985 1990 1995 2000 2005 2010 2015 2020 Year Year a Changes in the yield of maize b Changes in the yield of rice 30 Ondo Ogun 25 Kwara Cassava Yield (MT/Hac) 20 15 10 5 1985 1990 1995 2000 2005 2010 2015 2020 Year c Changes in the yield of cassava Fig. 2 Changes in crop yield: Maize (a), Rice (b) and Cassava (c) shortly after planting affects yield. Though maize and rice Mann–Kendall (MK) and Sen’s slope tests were used on can be less sensitive to rainfall at the onset and cessation monthly scale to detect trends in the minimum tempera- of the growing season, several factors can account for ture record at different stations. Table 2 shows the result reduction in yield of maize as both reduction and erratic of MK test and from the result MK and Sen’s slope tests rainfall affects maize yield, rainfall variability influences were different in pattern for the minimum temperature maize yield but it depend on the level of inputs such as in the stations of Ondo, Ogun and Kwara states. From fertilizer in maize cultivation [24]. Prolonged dry spells Table 2, the results of the MK test for Ondo state, Janu- can also affect yield of maize negatively, the impacts of ary has a random trend of z value of 1.66 in minimum rainfall and temperature on tuber crops are obvious espe- temperature with a slope of 0.044 which is statistically cially for cassava which can even be up to 95% probabil- not significant at p > 0.05. In February, there exists uncer- ity levels [24]. There is possibility of increased yield even tainty in MK trend of z value of 1.26, while the Sen’s slope with reduction in land area with introduction of superior test showed a slope of 0.028, but in March, a random maize varieties and the liberalization of seed value chain trend of z value of 0.71 and a slope of 0.19 that is statis- that increased the availability of improve seed to farmers tically significant was detected. In April, a random MK [24]. trend of value 0.82 and a Sen’s slope of 0.010 that is not statistically significant. For May, a random MK trend of z Variation in climate and time series from Mann Kendall value of 1.65 and a Sen’s slope of 0.031 that is statistically and Sen’s Slope Detection significant was detected. In June, a positive MK trend The results of monthly trend tests showed a mix of with a z value of 2.00 and a positive Sen’s slope value of randomness and upward trend in different stations. 0.018 that is statistically significant. In July, a positive MK Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 6 of 17 Table 2 Mann Kendall and Sen’s Slope analysis of minimum temperature Minimum temperature (℃) Month Ondo Ogun Kwara S Z Trend Slope S Z Trend Slope S Z Trend Slope January 123 1.66 0.04 31 0.43 0.04* 25 0.43 0.01 February 93 1.26 0.03 183 2.49 + 0.02 53 0.98 0.04* March 51 0.71 0.02* 102 1.43 0.20* 108 1.92 0.03 April 61 0.82 0.01 189 2.67 + 0.03* 71 1.26 0.01 May 117 1.65 0.03* 141 1.99 + 0.04* 110 2.05 + 0.03* June 142 2.0 + 0.02* 166 2.34 + 0.03* 110 1.96 0.02 July 141 1.99 + 0.02* 207 2.93 + 0.03* 203 3.64 + 0.02 August 56 0.78 0.01* 157 2.13 + 0.01 129 2.30 + 0.02 September 160 2.18 + 0.03 158 2.23 + 0.03* 176 3.15 + 0.03 October 140 1.97 + 0.03 165 2.33 + 0.05* 36 0.63 0.00 November 80 1.08 0.01 173 2.44 + 0.04* −34 −0.59 −0.01 December 197 2.67 + 0.08 143 1.94 0.02 −16 −0.27 −0.01 Overall 134 1.89 0.03* 187 2.64 + 0.03* 163 2.91 + 0.03 ( +), Positive Trend; (-), Negative Trend; (), Random; (*), Significant at 0.05 trend with a positive value of 1.99 and a positive Sen’s In June, a positive MK trend with a z value of 2.34 and a slope value of 0.024 that is statistically significant was positive Sen’s slope of 0.033 that is statistically significant detected. In August, a random MK trend with a value of was detected. In July, a positive MK trend with a value 0.78 and a slope of 0.10 that is statistically significant was of 2.93 and a positive Sen’s slope of 0.029 that is statisti- detected. In September, a positive MK trend with a value cally significant was detected. In August, a positive MK of 2.18 and a positive Sen’s slope of 0.016 that is not sta- trend with a value of 2.13 and a Sen’s slope of 0.014 that tistically significant was detected (Table 2). In October, a is statistically not significant was detected. In September, positive MK trend with a value of 1.97 and a Sen’s slope a positive MK trend with a value of 2.23 and a positive of 0.026 that is statistically not significant was detected. Sen’s slope of 0.031 that is statistically significant was In November, a random MK trend with a value of 1.08 detected. In October, a positive MK trend with a z value and a Sen’s slope of 0.014 that is not statistically signifi- of 2.33 and a positive Sen’s slope of 0.046 that is statis- cant was detected. In December, a positive MK test trend tically significant was detected. In November, a positive with a value of 2.67 and Sen’s slope value of 0.075 that MK trend with a z value of 2.44 and a positive Sen’s slope is not statistically significant was detected. In general, of 0.040 that is statistically significant was detected. In throughout the study period, a random MK trend with a December, a random MK test trend with a z value of 1.94 value of 1.89 and a Sen’s slope of 0.025 that is not statis- and a Sen’s slope value of 0.024 that is not statistically sig- tically significant was detected for all the years covering nificant was detected. the study period (Table 2). In general, throughout the years and months covering From the results of the MK test for Ogun state the study period, a positive MK trend with a value of 2.64 (Table 2), January had a random MK trend of z value of and a positive Sen’s slope of 0.029 that is statistically sig- 0.43, while the Sen’s slope showed a value of 0.43 that nificant was detected for the station in Ogun state which is statistically significant in minimum temperature at can be concluded to be an increasingly positive trend in p > 0.05. In February, there exists a positive MK trend of minimum temperature. In Ogun state, there exists posi- z value of 2.49, while the Sen’s slope test showed a value tive trends with upward movement and significant slope of 0.02 that is not statistically significant. In March, a values in the increase in the slope for all the months of positive MK trend of z value 1.43 and a Sense’s slope of the year excluding January, March and December. And 0.20 that is statistically significant was detected. In April, there exists statistically a significant trend at p < 0.05 in a positive MK trend of value 2.67 and a Sen’s slope of almost all the months excluding February, August and 0.032 that is statistically significant was detected. For December. In Kwara state though, a random MK trend of May, a positive MK trend of z value of 1.99 and a Sen’s Z value of 0.43 and with Sen’s slope value of 0.014 that slope of 0.041 that is statistically significant was detected. is not statistically significant at p > 0.05 was observed in Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 7 of 17 January minimum temperature (Table 2). In February, to have a random trend in the MK test, the Sen’s slope there exists a random MK trend of z value 0.98, while the analysis shows a significant increase in the slope of mini- Sen’s slope test showed 0.0044 that is statistically signifi- mum temperature only in February and May and Mann cant. In March, a random MK trend of z value of 1.92 and Kendall Test shows a positive trend in July, August and a Sen’s slope of 0.027 that is statistically not significant September with the general trend throughout the study was detected. In April, a random MK trend of value 1.26 period to be positive trend (Table 2). and a positive Sen’s slope of 0.014 that is not statistically Table 3 shows the result of the Mann–Kendall and significant was detected. For May, a positive MK trend of Sen’s slope test for maximum temperature in the stations z value of 2.05 and a positive Sen’s slope of 0.025 that is of Ondo, Ogun and Kwara states. From the results of the statistically significant was detected. In June, a random MK test for Ondo state, January has a positive random MK trend with a z value of 1.96 and a positive Sen’s slope trend of z value of 0.15 in maximum temperature with a of 0.022 that is statistically not significant was detected. slope of 0.00 which is statistically significant at p > 0.05. In In July, a positive MK trend with a value of 3.64 and a February, there exists negative trend of z value of −0.04, positive but not statistically significant Sen’s slope of while the Sen’s slope test shows a slope of 0.00 that is 0.023 was detected. In August, a positive MK trend with statistically significant. In March, a negative trend of a z value of 2.30 and a positive Sen’s slope of 0.017 that is z = −0.05 and a slope of 0.00 that is not statistically sig- not statistically significant was detected. In September, a nificant was detected. In April, a negative value of −0.61 positive MK trend with a value of 3.15 and a Sen’s slope and a slope of 0.00 that is not statistically significant. For of 0.025 is not statistically significant was detected. In May, a positive trend of z value of 2.47 and a slope of 0.00 October, a random MK trend with a z value of 0.63 and that is not statistically significant was detected. In June, without a Sen’s slope of value 0.004 that is not statisti- a positive trend with a z = 2.55 and a slope of 0.00 that cally significant was detected. In November, a negative is statistically significant. In July, a random trend with a MK trend with a z value of −0.59 and a negative slope of positive value of 2.22 and a slope of 0.00 that is statisti- −0.011 that is statistically not significant was detected. In cally significant was detected. In August, a random trend December, a negative MK trend with a z value of −0.27 with a positive value of 1.90 and a slope of 0.056 that is and a with negative Sen’s slope value of −0.007 that is statistically significant was detected. In September, a statistically not significant was detected. In general, positive trend with a positive value of 1.97 and a slope of throughout the study period, a positive MK trend with 0.00 that is statistically significant was detected. In Octo- a value of 2.91 and a positive Sen’s slope of 0.025 that is ber, a random trend with a positive value of 1.17 and a statistically not significant was detected for the station slope of 0.00 that is statistically significant was detected. in Kwara state. In Kwara state, most months were found In November, a negative trend with a value of −0.74 and Table 3 Mann Kendall and Sen’s slope analysis of maximum temperature Maximum temperature (℃) Month Ondo Ogun Kwara S Z Trend Slope S Z Trend Slope S Z Trend Slope January 11 0.15 0.00* 0 0.00 0.01* 94 1.65 0.03 February −4 −0.04 0.00* −101 −1.42 − −0.05* 66 1.22 0.05* March −36 −050 0.00 −128 −1.73 − −0.02 155 2.75 + 0.06 April −43 −0.61 0.00 54 0.75 0.00* 117 2.08 + 0.04 May 167 2.47 + 0.00 151 2.13 + 0.03* 63 1.16 0.04* June 172 2.55 + 0.00* 155 2.19 + 0.06* 75 1.39 0.03* July 155 2.22 + 0.00* 112 1.58 0.09* 153 2.72 + 0.04 August 133 1.90 0.06* 183 2.59 + 0.08* 59 1.04 0.01 September 136 1.97 + 0.00* 155 2.19 + 0.08* 110 1.96 0.02 October 81 1.17 0.00* 161 2.27 + 0.05* 15 0.25 0.00 November −51 −0.74 0.00 104 1.41 0.01 8 0.13 0.01* December −14 −0.19 0.00 185 2.51 + 0.02 104 1.93 0.07* Overall 83 1.22 0.00 159 2.24 + 0.04* 110 2.05 + 0.04* ( +), Positive Trend; (-), Negative Trend; (), Random; (*), Significant at 0.05 Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 8 of 17 a Sen’s slope of 0.00 that is not statistically significant was detected. For May, a random MK trend of z value of 1.16 detected. In December, a negative MK test trend with a and a positive Sen’s slope of 0.038 that is statistically sig- value of −0.19 and a Sen’s slope value of 0.00 that is not nificant was detected. In June, a random MK trend with statistically significant was detected. In general, through- a z value of 1.39 and a positive Sen’s slope of 0.030 that out the study period, a random trend with a value of 1.22 is statistically significant was detected. In July, a positive and a Sen’s slope of 0.00 that is not statistically significant MK trend with a value of 2.72 and a positive but not sta- was detected. tistically significant Sen’s slope of 0.038 was detected. In From the results of the MK test for Ogun state August, a random MK trend with a positive value of 1.04 (Table 3), there is uncertainty in the trend for January and a positive Sen’s slope of 0.09 that is not statistically z = 0.00 of MK test and same with Sen’s slope test with significant was detected. In September, a random MK a value of 0.00 in maximum temperature but the Sen’s trend with a value of 1.96 and a Sen’s slope of 0.020 that slope analysis shows the value is statistically significant at is not statistically significant was detected. In October, a p > 0.05. In February, there exists negative trend of z value random MK trend with a z value of 0.25 and without a of −1.42, while the Sen’s slope test shows also a negative Sen’s slope of value 0.00 that is not statistically significant slope of −0.048 that is statistically significant. In March, was detected. In November, a random MK trend with a a negative trend of z value of −1.73 and a negative Sense’s z value of 0.13 and a positive Sen’s slope of 0.014 that is slope of −0.016 that is not statistically significant was statistically significant was detected. In December, a ran- detected. In April, a random MK trend of value 0.75 and dom MK test trend with a z value of 1.93 and a positive a Sen’s slope of 0.00 that is statistically significant was Sen’s slope value of 0.071 that is statistically significant detected. For May, a positive MK trend of z value of 2.13 was detected. In general, throughout the study period, and a Sen’s slope of 0.029 that is statistically significant a positive MK trend with a value of 2.05 and a positive was detected. In June, a positive MK trend with a z value Sen’s slope of 0.035 that is statistically significant was of 2.19 and a positive Sen’s slope of 0.057 that is statisti- detected for the station in Kwara state. cally significant was detected. In July, a random MK trend Table 4 shows the result of the Mann–Kendall and with a value of 1.58 and a positive Sen’s slope of 0.092 Sen’s slope test for rainfall in the stations of Ondo, Ogun that is statistically significant was detected. In August, and Kwara states. From the results of the MK test for a positive MK trend with a positive value of 2.59 and a Ondo state, January has a random trend of z value of 0.91 positive Sen’s slope of 0.083 that is statistically significant in rainfall with a slope of 0.00 which is statistically not was detected. In September, a positive MK trend with a significant at p > 0.05. In February, there exists random value of 2.19 and a slope of 0.080 that is statistically sig- MK trend of z value of 0.43, while the Sen’s slope test nificant was detected. In October, a positive MK trend shows a slope of 0.00 that is statistically not significant. with a z value of 2.27 and a positive Sen’s slope of 0.053 In March, a random trend of z value of 0.37 and a slope of that is statistically significant was detected. In November, 0.529 that is not statistically significant was detected. In a random MK trend with a z value of 1.41 and a positive April, a negative MK trend of value −0.61 and a negative Sen’s slope of 0.013 that is not statistically significant was Sen’s slope of −0.725 that is not statistically significant. detected. In December, a positive MK test trend with a z For May, a negative MK trend of z value of −0.12 and a value of 2.51 and a positive Sen’s slope value of 0.024 that negative Sen’s slope of −0.920 that is not statistically sig- is not statistically significant was detected. In general, nificant was detected. In June, a random MK trend with a throughout the study period, a positive MK trend with a z value of 1.01 and a Sen’s slope value of 1.053 that is not value of 2.24 and a positive Sen’s slope of 0.036 that is sta- statistically significant. In July, a random MK trend with tistically significant was detected for the station in Ogun a positive value of 1.22 and a positive Sen’s slope value state. of 2.746 that is statistically significant was detected. In In Kwara state (Table 3), January has a random MK August, a random MK trend with a positive value of 0.87 trend of z value of 1.65 and with Sen’s slope value of 0.033 and a slope of 0.361 that is not statistically significant was in maximum temperature but the Sen’s slope analysis detected. In September, a random MK trend with a posi- shows the value is statistically not significant at p > 0.05. tive value of 0.34 and a positive Sen’s slope of 0.361 that In February, there exists random MK trend of z value of is not statistically significant was detected. In October, a 1.22, while the Sen’s slope test shows also a positive slope negative MK trend with a value of −0.90 and a negative of 0.050 that is statistically significant. In March, a posi- Sen’s slope of −1.146 that is statistically not significant tive MK trend of z value of 2.75 and a positive Sen’s slope was detected. In November, a random MK trend with of 0.056 that is not statistically significant was detected. a value of 1.32 and a positive Sen’s slope of 0.563 that is In April, a positive MK trend of value 2.08 and a positive not statistically significant was detected. In December, a Sen’s slope of 0.038 that is not statistically significant was negative MK test trend with a value of −0.52 and Sen’s Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 9 of 17 Table 4 Mann Kendall and Sen’s Slope analysis of rainfall Rainfall (mm) Month Ondo Ogun Kwara S Z Trend Slope S Z Trend Slope S Z Trend Slope January 65 0.91 0.000 −4 −0.04 – 0.000 −5 −0.07 0.000 February 32 0.43 0.000 65 0.87 0.167 −20 −0.31 0.000 March 28 0.37 0.529 91 1.23 0.890 −81 −1.24 −0.690 April −46 −0.61 −0.725 11 0.14 0.000* 84 1.29 1.429 May −83 −1.12 −0.920 83 1.12 1.317 8 0.11 0.058 June 75 1.01 1.053 73 0.98 1.742 42 0.64 0.704 July 87 1.22 2.746* 40 0.53 0.938 58 0.88 0.974 August 65 0.87 1.523 29 0.38 0.628 −21 −0.31 −0.324 September 26 0.34 0.361 127 1.72 3.026 66 10.1 + 1.529 October −67 −0.90 −1.146 101 1.36 2.000 160 2.46 + 3.994 November 98 1.32 0.563 185 2.51 + 1.375 1 0.00 0.000 December −36 −0.52 0.000 42 0.59 0.000 −11 −0.22 0.000 Overall 48 0.64 1.297 113 1.59 12.667* 69 1.05 4.000 ( +), Positive Trend; (-), Negative Trend; (), Random; (*), Significant at 0.05 slope value of 0.00 that is not statistically significant was Sen’s slope of 1.375 that is not statistically significant was detected. In general, throughout the study period, a ran- detected. In December, a random MK test trend with a z dom MK trend with a value of 0.64 and a Sen’s slope of value of 0.59 and a Sen’s slope value of 0.00 that is not sta- 1.297 that is not statistically significant was detected. tistically significant was detected. In general, throughout From the results of the MK test for Ogun state the study period, a random MK trend with a value of 1.59 (Table 4), January had a negative MK trend of z value of and a positive Sen’s slope of 12.667 that is statistically sig- −0.04 0.00 of MK test, while the Sen’s slope showed a nificant was detected for the station in Ogun state. value of 0.00 that is not statistically significant in rainfall In Kwara state (Table 4), a negative MK trend of z value at p > 0.05. In February, there exists random MK trend of of −0.07 and with Sen’s slope value of 0.000 that is not z value of 0.87, while the Sen’s slope test showed a posi- statistically significant at p < 0.05 was observed in Janu- tive value of 0.167 that is not statistically significant. In ary rainfall. In February, there exists negative MK trend March, a random MK trend of z value 1.23 and a positive of z value −0.31 while the Sen’s slope test showed 0.000 Sense’s slope of 0.890 that is not statistically significant that is statistically not significant. In March, a negative was detected. In April, a random MK trend of value 0.14 MK trend of z value of −1.24 and a negative Sen’s slope and a Sen’s slope of 0.00 that is statistically significant of −0.690 that is not statistically significant was detected. was detected. For May, a random MK trend of z value In April, a random MK trend of value 1.29 and a posi- of 1.12 and a positive Sen’s slope of 1.317 that is statisti- tive Sen’s slope of 1.429 that is not statistically significant cally not significant was detected. In June, a random MK was detected. For May, a random MK trend of z value of trend with a z value of 0.98 and a positive Sen’s slope of 0.11 and a positive Sen’s slope of 0.058 that is statistically 1.742 that is not statistically significant was detected. In not significant was detected. In June, a random MK trend July, a random MK trend with a value of 0.53 and a posi- with a z value of 0.64 and a positive Sen’s slope of 0.7040 tive Sen’s slope of 0.938 that is not statistically signifi- that is statistically not significant was detected. In July, cant was detected. In August, a random MK trend with a random MK trend with a value of 0.88 and a positive a positive value of 0.38 and a positive Sen’s slope of 0.628 but not statistically significant Sen’s slope of 0.974 was that is statistically not significant was detected. In Sep- detected. In August, a negative MK trend with a z value tember, a random MK trend with a value of 1.72 and a of −0.31 and a negative Sen’s slope of −0.324 that is not positive Sen’s slope of 3.026 that is not statistically signifi- statistically significant was detected. In September, a ran- cant was detected. In October, a random MK trend with dom MK trend with a value of 10.1 and a Sen’s slope of a z value of 1.36 and a positive Sen’s slope of 2.000 that 1.529 that is not statistically significant was detected. In is not statistically significant was detected. In November, October, a positive MK trend with a z value of 2.46 and a positive MK trend with a z value of 2.51 and a positive without a positive Sen’s slope of value 3.994 that is not Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 10 of 17 statistically significant was detected. In November, there no significant increase in Ondo and Kwara but there is a was randomness in MK test with a z value of 0.00 and statistically significant increase at p < 0.05 in Ogun state. no slope with Sen’s slope of 0.00 that is statistically not The perception of farmers (Table 5) about the tem- significant was detected. In December, a negative MK perature of the hottest months as opined by the major- test trend with a z value of −0.22 and with no slope, as ity of the farmers in Ondo and Ogun states, have been Sen’s slope value of 0.00 that is statistically not significant reported to have increased significantly while the major- was detected. In general, throughout the study period, ity think that it has remained the same over the years in a random MK trend with a value of 1.05 and a positive Kwara state. The rain generally and the rain during the Sen’s slope of 4.00 that is statistically not significant was rainy season which spans usually April to October have detected for the station in Kwara state. been reported by farmers to be higher than it has been in The result of monthly trend tests showed a mix of ran- the past in Ondo and Ogun while a higher proportion of domness; and a positive trend in different stations over farmers perceived that the rain has remained the same in the years. Mann–Kendall (MK) and Sen’s slope tests Kwara states and this was found to corroborate with the were used on monthly scale to detect slope in the mini- analysis of rainfall data. The rain during the dry season mum temperature, maximum temperature and rainfall has been reported to be higher in Ondo while it has gen- recorded at different stations. The result is as presented erally remained the same in Ogun and Kwara states. in Tables 2, 3 and 4. Generally, the study found out that The occurrence of the extreme flood was perceived by while there is general increase or upward trend in the farmers to be more frequent in Ondo and Kwara while minimum and maximum temperature as shown by the the majority of the farmers in Kwara believe that it has Mann Kendall and Sen’s slope analysis, there is generally remained the same over the years. For the occurrence of negative trend for rainfall. The decrease in rainfall can extreme droughts, it was reported to be less frequent in be noticed in many months and this is enough to bring Ondo, more frequent in Ogun and same in Kwara state about poor yield. Climate change has taken a centre point (Table 5). The rainy season duration was reported to be in the midst of diverse threatening environmental chal- earlier in Ondo while it has remained the same in Ogun lenges of our time which have promoted discourses on and Kwara states as opined by the majority of the farm- causes, long term effects and how to forestall the linger- ers. The length of the dry season was by majority of the ing and frustrating impact as many tropical and sub-trop- farmers to have remained the same in Ondo, Ogun and ical countries are more vulnerable to the threat because Kwara states. Figure 2a shows farmers’ perception on of their dependence on rain fed agriculture [24]. temperature generally. Majority of the farmers in Ondo state are of the opinion that the temperature is warmer while majority in Ogun and Kwara states are of the Community‑based perception of climate change opinion it has remained the same over time. Figure 2b Majority of rural farmers in Ondo and Ogun states shows perception on rainfall for which the majority in opined that the temperature in general has become Ondo and Ogun opined has been higher while majority warmer while in Kwara states, the majority believed it has in Kwara said it has remained the same. Occurrence of remained the same (Table 5) and Fig. 3. Though farmers extreme floods is depicted in Fig. 2c for which majority perceived it has remained the same on a general note, said it has remained the same in Ondo meanwhile it has the climatic data of the Ogun and Kwara states showed been more frequent in Ondo and Ogun states. However, that it has been warmer since the year 2000 to the end studies have shown that there is usually variation in the of the study period in 2017 in Ogun state and for Kwara, farmers’ perception of climate and climatic record as the it was warmer between 1982 and 1986, colder between farmers’ perception was different from the climate data. 1987 and 1997 and have generally remained warmer This is probably because of the cooling effect of the pre- since. The perception of the temperature in Ondo state sent climatic condition and intense rainfall over the past is in line with the climate data which showed that it has few years. This may not be significant enough to be put remained warmer since 2010 with fluctuation in the pre- into cognizance [28, 29]. Importantly, farmers have per- ceding decade. The coldest season is between December ceived delayed in onset of events, increased temperature and January during harmattan. Majority of the farmers and these are enough for reduction in yield. in the three states believe that the temperature of the coldest season has remained the same over the years. Cropping experiences of farmers and perceived change Whereas, Mann Kendall and Sen’s slope analysis of cli- in climate matic data showed that there is an positive trend in In Onibode, Ogun state, the FGD conducted revealed December temperature in Ondo, fluctuation in Ogun and that farmers have perceived climate change and are a negative trend for Kwara. While in January there was of the opinion that the drivers of climate change are Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 11 of 17 Table 5 General perception of farmers on climate change Variables Variables Perception of change (%) Ondo Ogun Kwara Over all Temperature generally Same 24.1 49.2 70.9 48.2 Warmer 44.4 35.6 14.5 31.5 Colder 31.5 15.2 14.5 20.3 Temperature of the coldest season Same 35.2 50.8 65.5 50.6 Warmer 29.6 38.6 16.4 29.2 Colder 31.5 10.6 18.2 20.2 Temperature of the hottest season Same 27.8 35.6 52.7 38.7 Warmer 46.3 49.2 40 45.2 Colder 25.9 15.3 7.3 16.1 Rain generally Same 20.8 23.7 50.9 31.5 Lower 14.4 8.5 12.7 11.9 Higher 64.8 67.8 36.4 56.5 Rain during rainy season Same 24.1 27.1 52.7 34.5 Lower 12.9 5.1 14.5 10.7 Higher 63.0 67.8 32.7 54.8 Rain during dry season Same 37.0 45.8 58.2 47.0 Lower 20.4 10.1 16.4 15.5 Higher 42.6 44.1 25.5 37.5 Extreme floods Same 25.9 18.6 50.9 31.5 Less frequent 13.0 30.5 12.7 19.0 More frequent 42.6 44.1 20 35.7 More intensive 18.5 6.8 16.4 13.7 Extreme drought Same 29.6 16.9 50.9 32.1 Less frequent 42.6 27.1 25.5 31.5 More frequent 16.7 44.1 16.4 26.2 More intensive 11.2 11.9 7.2 10.1 Rain onset Same 29.6 83.1 70.9 61.9 Earlier 55.6 10.2 20.0 28.0 Later 14.8 6.7 9.1 10.1 Rainy season duration Same 31.5 81.4 70.9 61.9 Shorter 9.2 6.7 9.1 8.3 Longer 59.3 11.9 20 29.8 Dry season duration Same 51.9 79.7 72.7 68.5 Shorter 33.3 13.6 12.7 19.6 Longer 14.8 6.7 14.5 11.9 Awareness of climate change Yes 81.5 88.1 81.8 83.9 No 18.5 11.9 18.2 16.1 natural rather than human-induced. There were observed December and last till March (about 4 months). Also, changes in the mean temperature, frequency of cold he stated that the number of rainy days is now more and days, sunny days, the intensity of solar radiation and farmers now grow new crops that are not native to their heavy rainfall events were generally agreed to be on the land. Examples of such crops include cucumber, water increase. The frequency of warm days with changes in melon, golden melon, carrot and a specie of cassava (Ege mean rainfall was agreed to be higher. However, a par- Nuru) while cocoa, beans and white seed melon (Egunsi ticular farmer in a separate interview opined that heat Itoo) have been abandoned as a result of climate change. starts from December, therefore, they start prepar- A group of farmers in Shao, Kwara state having their ages ing land for cultivation and that dry season starts from in the range of 45–68 (the youngest in his mid-forties and Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 12 of 17 80 60 70 Same 50 Warmer 60 Colder 40 50 Ondo Ogun 40 30 Kwara 30 20 20 10 10 0 0 Ondo Ogun Kwara Same Less frequent More frequent More intensive a General perception of temperature b General perception on rain 80 Same 70 Lower 60 Higher 50 40 30 20 10 0 Ondo Ogun Kwara c Perception on Extreme Floods Fig. 3 Farmers’ perception about temperature, rain and extreme floods the oldest in his late sixties with over 30 years of farm- extreme events have not been observed for a long time. ing experiences) in the FGD conducted stated that there Over 65-year-old farmer in Kobape who claimed to have is a great change in climatic events. However, the opin- been farming for over 40 years said he used to plant few ions about the changes are different as the temperature crops over an area of about 2–3 hectares but changes in was argued to be on the increase by few while the rainy climate has so much affected the yield and that made him season duration was agreed to be on the decrease. On increase the area under cultivation, now ranging from the intensity of rain, no agreement was reached as two of 5–10 hectares on yearly basis with more crops of differ- four maintained that the rain is still the same as it used to ent species planted to enhance yield. According to him, be while others believed it has changed. farmers who want high yield must increase the area of However, the perception of the other two that disa- land cultivated and as well plant many crops. Many of greed was probably influenced by the distribution and the species of crops we used to know and planted are not intensity of rain in the year just before the field sur- very productive again. Some even with the application vey (2019). Generally, it was agreed that the tempera- of fertilizer, the yield is not encouraging, hence, there ture is on the increase, heat period, intensity, wind and have been changes in the species of crops planted. For other climatic elements have really changed in the past cassava which used to be 2–3 years before harvest, we years. There is now less flooding except for this year and now have a new variety which we can plant and harvest Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 13 of 17 in 6–8 months. New varieties of maize were introduced Generally, there were differences in the opinion and by the extension workers of the ADP which we don’t perceptions of farmers on climate and as have been need to do a thorough tilling or ploughing before a bet- argued by many of the farmers, some of their arguments ter yield is harvested. Many aging farmer said that they were corroborated with climatic data analysis. However, used to plant rice in the past but not anymore because studies have shown that there is usually a difference in of changes in the biophysical environment probably asso- climatic data analyses and farmers perception on climate ciated with climate change. They added that there were change [22, 30]. birds when they were growing up which used to disturb rice farm and clothes were hanged on rice farms to scare Adaptive capacity of farmers toward climate change birds away (scare crows). Now the population of the bird and barriers to adaptation (rice sucker) has increased and there are other birds not Figure 4 shows the farmers coping or adaptive capac- feeding on rice then that do now. ity toward extreme climatic conditions. 16.7% of the “The clothes we used to hang in rice farm to scare respondents are engaged in agricultural diversification the birds doesn’t work again, the birds observe and while 83.3% are not, 17.3% of the total respondents are if there is no noticeable movement in the clothes or engaged in changing the crops they cultivate while 82.7% in the farm, they invade in their large numbers and are not. 20.2% are practicing crop rotation while 79.8 are can finish 10 hectares of rice in a day with farmers not, 19.6 are practicing mixed farming while 79.8 are not. having nothing to show forth at the end of the farm- 19% for changes in the size of farmland while 81% are not, ing season. We stopped farming rice simply because 14.9% are engaged in different crop composition while of this”. 85.1% are not. 33.3% are engaged in agricultural intensifi- cation while 66.7% are not, 30.4% are engaged in bush fal- The chief of Onibode (Baale) an adult in his early 70 s lowing while 69.6% are not. 29.8% are changing the daily age and a farmer in his 50 s in Onibode who did not state working time while 70.2% are not. 47.6% are presently categorically his age but have been in farming for about engaged in agroforestry while 52.4% are not. 27.4% are 30 years is of the opinion that the rain is still falling well changing the seasonal timing of sowing while 72.6% are but not exactly like it used to be both in the rainy and dry not. 41.1% are changing the harvest time while 58.9% are seasons. The temperature, the wind, the cloud formation not. 35.7% are into irrigation while 64.3% are not. 31.5% among others has remained the same over the years. The are into use of fertilizer while 68.5% are not 32.1% are date for the first rain of the year is 15th of March and that into the use of pesticide while 67.9% are not. 56.5% are still remain intact. However, because of changes associ- into getting loan and credit facilities while 43.5% are not. ated with market demand, some staple crops have been 32.7% are into getting other sources of income so as not abandoned while new ones have been adopted. Generally, to depend on the farm only while 67.3% are not. 44% are in the land, they don’t grow water melon, cucumber, egg changing the method of storing their food while 56% are plants, lettuce and some species of tomato and pepper not. 27.4% are changing the quantity of food consumed which are now cultivated on large scale by many farmers. by the family while 72.6% are not. And 24.4% are chang- He stated that the rain is not like it used to be, as there ing their residence by moving to another area for farming is a significant reduction in the amount of rainfall and while 75.6% are not changing their residence (Fig. 4). the distribution through the year. In the past, the rain do Twenty adaptation options were selected for this come in different ways, we have that of 3 days, 7 days and study and farmers are adapting to extreme climatic con- even 15 days of continuous rainfall. ditions but with generally low ability to cope with the “In 2015, I used the first rain in February to plant extreme events. The comparative analysis for each of maize in expectation that the rainy season has set the selected ecological zones is depicted in Fig. 8 b–d in in. Few days later there was second rain of the year the annex. Figure 4 shows the percentage error graph of and we were so happy, because the first maize of farmers coping or adaptive capacity toward extreme cli- the year is usually expensive and not many farm- matic conditions. 16.7% of the respondents are engaged ers do have that on their farms. Normally, the maize in agricultural diversification, 17.3% are engaged in should be due for harvest in April but not until changing the crops they cultivate, 20.2% are practic- early May before we had the third rain of the year. ing crop rotation, 19.6% are practicing mixed farming. It amounted to great loss for me, waste of money, 19% for changes in size of farmland, 14.9% are engaged energy and time. Since then, I’ve been careful about in different crop composition. 33.3% are engaged in early cultivation of maize”. agricultural intensification, 30.4% are engaged in bush fallowing. 29.8% are changing the daily working time. Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 14 of 17 120 100 80 Frequency (%) 60 40 20 0 n si ed ha so ry da us ific n s g iff in Mi Ro tion n e lic n in nsu ge in h ha red plic ion e w allo on es ing for e Lo stic r ap iga e nt po d an ge es fac on C of fo ood nco s ha l t Ag ing g ng d sto e ric cr o far n ra co rm g f i ie Fe Use rve win C cult tio in B ns sitio Pe tiliz of ir t tim id p tio n nc l I m lan ng m ro tim in m Ag ren Siz ed tatio in of est re m es co ra ul op f fa mi ily h f ati qu an ng it ati tit s in in ilit or w t ca de p a & ap a v op ifi i r cr rs k in ive e o x r e e ha o es l D c i ng ra r tu ha ltu C a t C a y on D es C ricu s as e ng ge C Ag se ha an in C h C es es ng ng ha ha C C Practice Adaptation Options Fig. 4 General adaptation practices by farmers in the region 47.6% are presently engaged in agroforestry, 27.4% are however impeded by certain barriers from practic- are changing the seasonal timing of sowing. 41.1% are ing adaptation [24]. This is against the report that rural changing the harvest time. 35.7% are into irrigation. farmers in Africa recognizes even subtle changes in cli- 31.5% are into use of fertilizer, 32.1% are into the use of mate parameters and take steps to respond [22, 30]. pesticide. 56.5% are into getting loan and credit facili- The study further investigated the reasons for low adap- ties. 32.7% are into getting other source of income so as tation level. The results of the general barriers to adapta- not to depend on the only farm. 44% are changing the tion is presented in Fig. 5 while for each of the ecological method of storing their food. 27.4% are changing the zone is shown in Fig. 5. The result showed that 41.1% quantity of food consumed by the family. And 24.4% are hindered from adapting, because they have low risk are changing their residence by moving to another area to climate. Low probability to climate change accounted for farming (Fig. 4). Importantly, Fig. 4 shows seeking for 42.3%, low severity of climate change accounted for for loans and credit facilities, changes in food storage, 43.5%, low benefits from adapting measures accounted agroforestry, changing the time of harvest, use of irriga- for 50.6%, low ability to adapt accounted for 63.7%, lack tion, use of fertilizer, agricultural intensification among of financial capital accounted for 70.2% lack of physi- others are the most practiced adaptation options. Stud- cal capital accounted for 73.8%, lack of human capital ies have reported that many farmers especially in Africa accounted for 82.7% lack of natural capital accounted for are aware of climate change and the risk associated but 83.3%, lack of social capital and lack of time accounted Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 15 of 17 120 100 80 Frequency (%) 60 40 20 0 La of sica ap t or y o efe n La ow ng cha ct ve ad d La of bilit asu e s ad lim e im isk ta Le tura t/cu ce l su e te co r p or l n e ef ffs ck at ca al so l c al of L l c ital si ew n/s efe ion is ib se ula s ith am io /pr olat t m a l C Hab ista ern of al s ck hu l c ita ck hy l c ap n al rel om is or ck fin y t re ia p po eg m ks ilit pr tio ct ad f tr rre ck a me ng a tim po In ila ork cia enc pa La of n an pit of ura pit t k it rs eo tio g l/ st or pp m c ng e r ci ap ex ac ap fe nt m ur r or La f p ncia ad o r fro ty o ha ng o a o e a c ha L ti e ap at ne ev a e c be s lim at n fit eri te w ow of c clim f w fr ig l r at D ck ilit Low i m La s y ul ab L ob of pr Lo ar ap w Fe Lo ad Hindered er th O General Adaptation Barriers Fig. 5 General barrier to adaptation for 86.9% and 32.7% respectively. Lack of external sup- Conclusion port 56%, distance or isolation especially from fam- Analysis of climate data using Mann Kendall and Sen’s ily members accounted for 23.2%, habits, custom and slope test showed generally warming temperatures local preferences accounted for 34.5%, cultural, religion with reduced rainfall marked with irregularities. The and social norms accounted for 21.4%, other adapta- perception of farmers were assessed. Generally, tem- tion options with similar purpose preferred accounted perature of the coldest month, rain during the rainy for 23.8%, incompatible tradeoffs of adaptation options season and rain during the dry season were believed to accounted for 26.2 while fear of potential risk or adverse have remained the same in Ogun and Kwara by higher effects accounted for 10.7% (Fig. 5). These results imply numbers of respondents, whereas statistical analyses that adaptation practices dependent on age of farmers, showed that there is variation while in Ondo state. The farm size and level of production. There has been expan- farmers are of the opinion that the temperature is sion on adaptation to current and projected impacts warmer generally, temperature of the hottest season of climate change to include barriers to adaptation and is warmer, rain is higher in the rainy and dry seasons these have raised questions around social, financial, with extreme drought less frequent, onset earlier, rainy cultural, environmental and ecological conditions and season duration longer. Though, temperature of the changes which can hamper ability of a farmer to adapt to hottest month was believed to be warmer by a higher climate change [22, 30]. proportion of the respondents. Many farmers are of the opinion that extreme floods are more frequent while Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 16 of 17 extreme drought, duration of the rainy and dry seasons Availability of data and materials The underlying data can be made available upon request to the correspond- have remained the same. In general, the farmers are ing author. Correspondence and requests for materials should be addressed aware of long-term changes in climatic factors (temper- to A.A. ature and rainfall) but some are unable to identify those Ethics approval and consent to participate changes as climate change [4, 31, 32]. This study conforms to the ethics principles of Obafemi Awolowo University. This study revealed that the effect of climate on Ethical approval for this study was obtained with standard ethics as the par- crops is not the same across the ecological zones ticipants were not vulnerable in anyway, data was processed in anonymised procedure, and survey participants had the possibility to skip questions. under consideration. The yield of cassava and rice is negatively affected in all the ecological zones though Consent for publication greatly affected in Ondo (rainforest) and Kwara (guinea Not applicable, because there are no data contained within our manuscript from which individual patients or participants may be identified. savanna) while the yield of maize is significantly affected only in Ondo state. Therefore, there is need to Competing interests enhance maize production in Ondo, rice and cassava The authors declare that they have no competing interests. production in the three ecological zones under con- Received: 27 May 2020 Accepted: 15 October 2020 sideration. According to the farmers’ report, improved seedlings of maize, rice and cassava have been intro- duced by the Agricultural Development Programme (ADP) to enhance their practices. However, many farm- References ers are now changing their agricultural and farming 1. Ayanlade A, et al. Rainfall variability and drought characteristics in two practices which can be said to be passive response to agro-climatic zones: an assessment of climate change challenges in Africa. Sci Total Environ. 2018;630:728–37. climate change. The key finding from this stduy is that 2. Mkuhlani S, et al. Classification of small-scale farmers for improved rainfall the major barrier to adaptation is centered on capital variability management in South Africa. Agroecol Sustain Food Syst. which influences who can access farm inputs. 2020;44(1):7–29. 3. Kotir JH. Climate change and variability in Sub-Saharan Africa: a review of current and future trends and impacts on agriculture and food security. Environ Dev Sustain. 2011;13(3):587–605. Abbreviations 4. De Longueville F, et al. Comparing climate change perceptions and KII: Key Informant Interview; FGD: Focus Group Discussion; MK: Mann Kendall; meteorological data in rural West Africa to improve the understanding of ANOVA: Analysis of variance. household decisions to migrate. Climatic Change, 2020. 1–19. 5. Ayanlade A, et al. Drought Characteristics in Two Agro-Climatic Zones in Acknowledgments Sub-Saharan Africa. Climate Prediction S&T Digest. 2018: 140. We thank the Nigerian Meteorological Agency (NiMet) for proving data and 6. Chen H, Sun J. Anthropogenic warming has caused hot droughts more the participating rural households for their candor. frequently in China. J Hydrol. 2017;544:306–18. 7. Clarke C, Shackleton S, Powell M. Climate change perceptions, drought Authors’ contributions responses and views on carbon farming amongst commercial livestock AA and IAO. developed the conception and every aspect of this study; AA and game farmers in the semiarid Great Fish River Valley, Eastern Cape played the leading role in the supervision of the research project, through province, South Africa. Afr J Range Forage Sci. 2012;29(1):13–23. which the paper was developed, While IAO. analysed the data. All authors read 8. Ayanlade A, Howard MT. Land surface temperature and heat fluxes over and approved the final manuscript. three cities in Niger Delta. J Afr Earth Sc. 2019;151:54–66. 9. Agboola AM, Ayanlade A. Sea level rise and its potential impacts on Authors’ information coastal urban area: a case of Eti-Osa, Nigeria. Analele Universitatii din Isaac Ayo Oluwatimilehin is a graduate student in the Department of Geog- Oradea Seria Geografie. 2016;27(2):188–200. raphy, Obafemi Awolowo University, Ile-Ife, Nigeria. His Postgraduate research 10. Cazenave A, Cozannet GL. Sea level rise and its coastal impacts. Earth’s work mainly involves climate change impacts and adaptive capacity of rural Future. 2014;2(2):15–34. farmers, supervised by A. Ayanlade. 11. Toimil A, et al. Managing coastal erosion under climate change at the regional scale. Coast Eng. 2017;128:106–22. Ayansina Ayanlade is a researcher and lecturer in the Department of Geog- 12. Sowunmi F, Akintola J. Effect of climatic variability on maize production in raphy, Obafemi Awolowo University, Ile-Ife, Nigeria. He earned his PhD from Nigeria. Res J Environ Earth Sci. 2010;2(1):19–30. King’s College London, University of London, UK. His research areas of interest 13. Renard D, Tilman D. National food production stabilized by crop diversity. are Applied Climatology with specific interest in climate change studies, Nature. 2019;571(7764):257–60. land-climate processes, and application of remote sensing and GIS to land- 14. Moylan H. The impact of rainfall variability on agricultural production atmospheric interactions. and household welfare in rural Malawi, in Urbana-Campaign, U.o. Illinois, Editor 2012, University of Illinois: Illinois. Funding 15. Tyagi NK, et al. Role of development policies in combating climate This paper is part of research project, carried out with the aid of a grant from change issues in Indian agriculture: an assessment of irrigation and the International Development Research Centre, Ottawa, Canada, and with fertilizer policies, in climate smart agriculture in South Asia. Germany: financial support from the Government of Canada, provided through Global Springer; 2019. p. 205–26. Affairs Canada (GAC), under the framework of the Mathematical Sciences 16. Ayanlade A, et al. Communicating climate change impacts as manifested for Climate Change Resilience (MS4CR) program administered by AIMS. in extreme weather: a case of newspapers’ reports in Nigeria, in hand- Research was also partly funded by an ERC Consolidator Grant to Reyes-García book of climate services. Germany: Springer; 2020. p. 401–21. (FP7-771056-LICCI). 17. Hasan MK, Kumar L. Meteorological data and farmers’ perception of coastal climate in Bangladesh. Sci Total Environ. 2020;704:135384. Oluwatimilehin and Ayanlade Agric & Food Secur (2021) 10:3 Page 17 of 17 18. Loum A, Fogarassy C. The effects of climate change on cereals yield of 28. Kemausuor F, et al. Farmers’ perception of climate change in the Ejura- production and food security in Gambia. APSTRACT: Applied Studies in Sekyedumase district of Ghana. ARPN J Agric Biol Sci. 2011;6(19):26–37. Agribusiness and Commerce, 2015. 9(1033–2016–84322): 83. 29. Deressa TT, Hassan RM, Ringler C. Perception of and adaptation to 19. FAO, Report of the World Food Summit, in World Food Summit. Rome; climate change by farmers in the Nile basin of Ethiopia. J Agric Sci. 1996. 2011;149(1):23–31. 20. Ayanlade A. Seasonal rainfall variability in Guinea Savanna part of Nigeria: 30. Thomas DS, et al. Adaptation to climate change and variability: farmer a GIS approach. Int J Climate Change Strategies Manag. 2009;1(3):282–96. responses to intra-seasonal precipitation trends in South Africa. Climatic 21. Ajetomobi J. Sensitivity of crop yield to extreme weather in Nigeria. No. Change. 2007;83(3):301–22. 310–2016–5403. 2016. 31. Tripathi A, Mishra AK. Knowledge and passive adaptation to cli- 22. Ayanlade A, et al. Inter-annual climate variability and crop yields anoma- mate change: an example from Indian farmers. Climate Risk Manag. lies in Middle Belt of Nigeria. Adv Nat Appl Sci. 2009;3:452–65. 2017;16:195–207. 23. Ogunleye OS. Urban mobility in Ekiti State: options for transformation. 32. Ayanlade A, Radeny M, Morton JF. Comparing smallholder farmers’ per- Am Int J Social Sci. 2014;3(5):200–6. ception of climate change with meteorological data: a case study from 24. Oladimeji Y, Abdulsalam Z, Abdullahi A. Determinants of participation southwestern Nigeria. Weather Climate Extremes. 2017;15:24–33. of Rural farm households in non-farm activities in Kwara state, Nigeria: a Paradigm of poverty alleviation. Ethiopian J Environ Studies Manag. 2015;6(6):635–49. Publisher’s Note 25. Solanke M. Intra-Urban work and school trips characteristics in Ogun Springer Nature remains neutral with regard to jurisdictional claims in pub- state, Nigeria. Int J Dev Sustain. 2013. 2(2). lished maps and institutional affiliations. 26. Owoeye R, Sekumade A. Effect of climate change on cocoa production in Ondo State, Nigeria. J Soc Sci Res. 2016;10(2):2014–2025. 27. Khalid B, Jahangir M, Ali SM. Annuual Minimum Temperature Varia- tions in early 21st century in Punjab, Pakistan. J Atmos Solar Terr Phys. 2016;137:1–9. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions