Agriculture is the growing, processing and distribution of food and other products through intensive plant cultivation and animal husbandry in and around cities. It includes green belts around cities, farming at the city edge, vegetable plots in community gardens and food production in thousands of vacant inner - city lots (CFSCNAAC, 2003). In agriculture the choice of what to produce and how to produce is determined by the culture, traditions, market, water supply, climate, soil condition, plot size and distance from home (AbdulAziz, 2002; Wiebe, 2003). In view of the foregoing, climate has been undoubtedly identified as one of the fundamental factors that determine both crop cultivation and livestock keeping.

The variability of rainfall, temperature and relative humidity has been a topical issue in a sustainable environment as the crop yield and production is very important to the economy and livelihood of the people of Nigeria and also the world at large. The sub-humid climatic zone of Africa permits the cultivation of a variety of crops in a pattern that emerged in earlier centuries in response to local conditions (Onyekwelu et al., 2006, Ziervogel et al., 2008). It follows therefore that any change in climate may have effect in agricultural sector in particular and other socio-economic activities in general.

Climate change could have both positive and negative impacts and these could be measured in terms of effects on crop growth, availability of soil water, soil fertility and erosion, incidents of pests and diseases, and sea level rise (Onyekwelu et al., 2006; Ziervogel et al., 2008; Semenov, 2009; Butterworth et al., 2010). Temperatures throughout Nigeria are generally high with annual mean of about 27°C while diurnal variations are more pronounced than seasonal differences (Salami and Matthew, 2009). During wet seasons, major portion of the country comes under the influence of moisture-laden tropical maritime air. Spatial variability is evident in the irregular distribution of rainfall at both short-time scale and average conditions while the temporal variability tends to be greater in the Northern and Southern parts of the country (Olaniran, 1991b, Omotosho and Abiodun, 2007; Odiana and Ibrahim, 2015).

The regularity of drought periods has been among the most notable aspects of Nigerian climate in recent years, particularly in the drier regions in the north (Olaniran, 1991a; Salami and Matthew, 2009; Osang et al., 2015). These drought periods are indications of the great variability of climate across tropical Africa and the most serious effects of which are usually felt at the drier margins of agricultural zones or in the regions occupied primarily by pastoral groups. The high degree of spatial variability of Nigerian rainfall is associated with the intense randomness of the convective process, which is the dominant rain-producing mechanism in the country with the attendant effects on local features such as topography, vegetation and land cover type among others (Adefolalu, 1986; Omotosho and Abiodun, 2007). The relationship between climate change and food security is complex. Many factors influence food security, which means that often the link is not even made between failed crops and changing weather patterns. One of the problems of farming is the frequent complete loss of crops due to adverse weather conditions or pests. Changing weather patterns or extreme weather events, such as floods or droughts, can have negative consequences for agricultural production (Tadross et al., 2005; Misselhorn, 2005; Ziervogel et al., 2008; Mthulisi, 2013, IPCC, 2014). Rural communities’ dependent on agriculture in a fragile environment are continuously facing an immediate risk of increased crop failure and loss of livestock. Consequently, there is less access to food, which forces the price of the little available food product out of reach of the common man. This translates to secondary effect on the living standard of farmers, fishers and forest-dependent people who are already vulnerable to poor living conditions.

Few studies have been done in Ondo State, Nigeria to investigate the pattern and trend of rainfall, temperature, relative humidity and other parameters on yield of cassava, yam, pepper and tomato. However, previous records show that very few of them have intensively examined the relationship between climate change and crop production. Moreover, different crops might be affected non-uniformly. In order to ensure food security in south-western Nigeria, a region that feeds more than 45% of the nation's population, there are needs to examine the trends in the climate of this region, therefore, the purpose of this research is to determine the co-efficient of variation in climatic variables and examine the effect of climatic variables on crop yield in Ondo State, south west, Nigeria.

The study area

Ondo State is located in the south western part of Nigeria, with the capital city in Akure as shown in Figure 1. It has acoordinates of 7°10′N 5°05′ E; this means that the states lies entirely in the tropics. It is classified as wetland of flat terrain whose hydrological cycle is generally affected by its location. Ondo State has a population of 3,440,000 estimated by the 2006 population census. The soil of the study area is a sandy clay loam soil according to USDA textural classification of soil. Ondo state has a land area of about 15,500 km² and is situated in the western upland area within the humid region of Nigeria. The general elevation is 300 to 700 m. Local peaks rise to 1000 m; other hill-like structures which are less prominent rise only a few hundred meters above the general elevations (Fasinmirin and Konyeha, 2009).The pattern of rainfall is bimodal, the first peak occurring in June and July, and the second in September, with a little dry spell in August (Fasinmirin, 2009). The mean annual rainfall ranges from 1300 to 1500 mm. More than half of humid zone of Nigeria is covered by Pre-Cambrian basement complex, principally composed of metamorphic and igneous materials (Asiwaju–Bello, 1999). The soils are light textured, fine sandy loam to fine sandy clay loam. The soil is moderately well supplied by organic matter and nutrients. Moisture holding capacity is moderately good. The soil of the environment is however subject to seasonal water logging for varying periods, but generally it becomes dry during the dry seasons which fall within November and March (Fasinmirin and Konyeha, 2009). Agriculture is the main occupation of the people of Ondo states, and it is the major source of income for many in the states. Agriculture provides income and employment for more than 75% of the population of the people.

Some of the agricultural produce is: Cash crops such as Cocoa, Oil Palm, Kolanut, Plantain, Bananas, Cashew, Citrus and Timber; Arable /Food Crops such as Rice, Yam, Cassava, Maize and cowpea.

Data collection and analysis

Temperature, precipitation, and solar radiation are the three most widely used climate variables to assess climate change and its impact. However, solar radiation has a closer positive correlation with maximum temperature. In general, higher solar radiation leads to a higher maximum temperature and lower solar radiation leads to a lower minimum temperature because of radioactive cooling (Peng et al., 2004). This shows the direct correlation between temperature and solar radiation. Therefore, to overcome the possible correlation among the independent variables, this study considers only temperature, rainfall and relative humidity. Rainfall is the most important form of precipitation in terms of meeting water requirement of agricultural crops.

Secondary data on average climatic variables such as (temperature, relative humidity, rainfall) from 1996 to 2014 for Ondo state in the South-western Nigeria were obtained from the Agro-climatological department of the Agricultural Development programme (ADP), Akure, Ondo State, crop yield data were also collected. The limitations of the crop-yield data availability and accessibility influenced the selection of the period for the climatic variability study. The yield/production rates for some common staple food crops (cassava, yam) and fruits (Pepper and Tomato) in the region were obtained from Akure Agricultural Development Programme Office, Ondo State. In order to adequately examine the response of different classes of food crops to varying climatic conditions, the crops under investigation were grouped into two classes as follows: Tubers (Yam and Cassava), and fruits (Pepper and Tomato) all expressed in metric tons per hectare (mt/ha). The mean annual temperature, rainfall and relative humidity were computed from the climatic data collected.

Data obtained from the experiments were analyzed using Statistical Package for Social Sciences (SPSS). Both descriptive and multiple regression were employed in data analysis while simple correlation, stepwise multiple regressions and analysis of variance (ANOVA) were used in showing the relationship between climatic parameters and crop yield and showing the trend and variation in crop yield. These statistical techniques were employed in the analysis of both crop yield data and climatic parameters because of their peculiarity in revealing the relationship and variation among variables

The regression model is specified thus:

Where: Y = Crop output (mt); ₁ , ₂ and ₃ = Coefficient of variables X₁, X₂ and X₃ respectively; X₁ = Annual mean rainfall; X₁ = Mean temperature; X₁ = Mean temperature and e =unexplained variables.

Table 1 and Figure 2 show the descriptive analysis of cassava output and yield data in Ondo State from 1996 to 2014. Out of the four selected crops, Cassava has a total sum of production of 38745.35 mt and a mean value of 2039.23 mt. This shows that within the years under study, Cassava had the highest yield value. 2014 recorded the highest cassava yield of 23.86 mt/ha compared with the previous year which recorded a yield of 23.41 mt/ha. The table also shows that the mean annual rainfall increased from 1880.20 mm to 2002.41 mmin 2014, relative humidity reduced from 78.30 to 77.40%, minimum temperature reduced from 20.40 to 20.10°C and maximum temperature increased from 30.50 to 30.64°C. The increase in yield recorded may be as a result of the increase in rainfall and maximum temperature. Year 1996 recorded the lowest Cassava yield of 16.25 mt/ha; the year 1997 recorded an increase in yield from the 16.25 to 17.50 mt/ha, also mean annual rainfall increase from 1834.63 mm to 2011.14 mm, relative humidity reduced from 76.46 to 73.43%, minimum temperature increased from 19.63 to 20.14°C and maximum temperature reduced from 32.75 to 29.08°C. The increase in yield recorded in 1997 may be as a result of the increase in rainfall and minimum temperature.

Table 2 and Figure 3 show the descriptive analysis of yam products yield data in Ondo State from 1996 to 2014. Out of the four selected crops, Yam has a total sum of production of 31600.79 mt and a mean value of 1663.20 mt. This shows that within the years under study, after cassava, yam had the next highest yield value. Year 2014 also recorded the highest Yam yield of 19.27mt/ha compared with the previous year which recorded a yield of 18.97mt/ha. The table show that in 2014 mean annual rainfall increased from 1880.20 to 2002.41 mm, relative humidity reduced from 78.30 to77.40%, minimum temperature reduced from 20.40 to 20.40°C and maximum temperature increased from 30.64 to 30.64°C.

The increase in yam yield recorded in year 2014 may be as a result of the increase in rainfall and maximum temperature from year 2013. Year 2000 recorded the lowest yam yield of 13.16 mt/ha, the year before recorded a slight increase in yield 13.82 mt/ha, also annual mean rainfall increase from 1740.71 to 1932.63 mm, relative humidity reduced from 77.87 to 77.06%, minimum temperature increased from 18.37 to 20.17°C and maximum temperature reduced from 32.10 to 31.70°C in year 1999 and 2000, respectively. The reduction in yield recorded in year 2000 may be as a result of the reduction in relative humidity and maximum temperature.

Table 3 and Figure 4 show the descriptive analysis of pepper products yield data in Ondo State from year 1996 to 2014. Out of the four selected crops, pepper has a total sum of production of 222.19 mt and a mean value of 11.70 mt. This shows that within the years under study, pepper had the second lowest yield value. Year 2007 also recorded the highest pepper yield of 4.96 mt/ha compared with the previous year which recorded a yield of 1.44 mt/ha. The table shows that in 2007, mean annual rainfall increased from 1835.66 to 1974.11 mm, relative humidity reduced from 77.70 to75.51%, minimum temperature increased from 19.36 to 20.36°C and maximum temperature increased from 31.46 to 31.89°C. The increase in pepper yield recorded in year 2007 may be as a result of the increase in rainfall, increase in minimum and maximum temperature from year 2006 to 2007. Year 1996 recorded the lowest yam yield of 0.73 mt/ha; the next year recorded an increase in yield from the 0.73 to 1.01 mt/ha, also annual mean rainfall increase from 1834.63 to 2011.14 mm, relative humidity reduced from 76.46 to 73.43%, minimum temperature increased from 19.63 to 20.14°C and maximum temperature reduced from 32.75 to 29.08°C. The increase in yield recorded in 1997 may be as a result of the increase in rainfall and minimum temperature.

Table 4 and Figure 5 show the descriptive analysis of tomatoes products yield data in Ondo State from 1997 to 2014. Out of the four selected crops, tomatoes has a total sum of production of 111.467 mt and a mean value of 5.87 mt .This shows that within the years under study, tomatoes had the lowest yield value. Year 2014 also recorded the highest tomatoes yield of 3.89 mt/ha compared with the previous year which recorded a yield of 3.74 mt/ha. The table shows that in 2014 mean annual rainfall increased from 1880.20 to 2002.41 mm, relative humidity reduced from 78.30 to 77.40%, minimum temperature reduced from 20.40 to 20.40°C and maximum temperature increased from 30.64 to 30.64°C.

The increase in tomatoes yield recorded in year 2014 may be as a result of the increase in rainfall and maximum temperature from the previous year 2013. Year 1996 recorded the lowest tomatoes yield of 1.27 mt/ha, the next year recorded an increase in yield from the 1.27 to 1.75 mt/ha, also annual mean rainfall increase from 1834.63 to 2011.14 mm, relative humidity reduced from 76.46 to 73.43%, minimum temperature increased from 19.63 to 20.14°C and maximum temperature reduced from 32.75 to 29.08°C. The increase in yield recorded in 1997 may be as a result of the increase in rainfall and minimum temperature.

Regression results on the effect of climate variability on crop yield

Table 5 shows the regression result for the effect of annual mean rainfall, temperature and relative humidity on the yield of cassava in the study area. After some econometric considerations, the result from the linear function was chosen as the lead equation. The result from the regression analysis showed that the regression coefficient of determination R² was 0.207, it can be noted that about 20.7% of variation in cassava could be explained by means of rainfall, relative humidity and mean temperature. The remaining 79.3% were largely due to other variables outside the regression model that also have effect on cassava yield. The regression result also reveals that for every unit increment in rainfall there is a 0.04% positive effect on the yield of cassava also for every unit increment in temperature there is a 29.9% positive effect in the yield of cassava, and lastly for every unit increment in relative humidity there is a 1% positive effect on the yield of cassava during the study period. The implication of these findings is that the volume of rainfall, mean temperature and relative humidity may not necessarily determine the output of cassava in the study area.The important determinant is the spread of the rainfall within the year. According to Ojo (2000), one-month drought when maize is tasseling can result in serious reduction in the output of maize, and thus the same principles apply to the other climatic variables.

Table 6 shows the regression result for the effect of annual mean rainfall, temperature and relative humidity on yam yield in the study area. After some econometric considerations, the result from the linear function was chosen as the lead equation. The result from the regression analysis showed that the regression coefficient of determination R² was 0.186, it can be noted that about 18.60% of variation in yam could be explained by means of rainfall, relative humidity and temperature, the remaining 81.40% were largely due to other variables outside the regression model that also affects the yield of yam in the study area. The regression result also reveal that for every unit increment in rainfall there is a 0.03% positive effect on the yield of yam, also for every unit increment in temperature there is a 36.50% positive effect in the yield of yam, and for every unit increment in the relative humidity there is a 1.20% positive effect on the yield of yam during the study period in Ondo State.

Table 7 shows the regression result for the effect of mean annual rainfall, temperature and relative humidity on the yield of pepper in the study area. After some econometric considerations, the result from the linear function form was chosen as the lead equation. The result from the regression analysis showed that the regression coefficient of determination R² was 0.268; it can be noted that about 26.80% of variation in pepper could be explained by means of rainfall, relative humidity and temperature. The remaining 73.20% were largely due to other variables outside the regression model that also has effect on the pepper yield in the study area. The regression result also reveal that for every unit increment in rainfall there is a 0.01% positive effect on the yield of pepper, also for every unit increment in temperature there is a 1.70% positive effect in the yield of pepper, and for every unit increment in the relative humidity there is a 2.0% positive effect on the yield of pepper during the study period in Ondo State.

Table 8 shows the regression result for the effect of annual mean rainfall, temperature and relative humidity on the yield of tomatoes in the study area. After some econometric considerations, the result from the linear function form was chosen as the lead equation. The result from the regression analysis showed that the regression coefficient of determination R² was 0.155. It can be noted that about 15.50% of variation in tomatoes could be explained by means of rainfall, relative humidity and temperature. The remaining 84.50% were largely due to other variables outside the regression model that has effect on the yield of tomato in the study area. The regression result also reveal that for every unit increment in rainfall there is a 0.01% positive effect on the yield of tomato, also for every unit increment in temperature there is a 6.50% positive effect in the yield of tomato, and for every unit increment in the relative humidity there is a 3.40% positive effect on the yield of tomato during the study period in Ondo State.

In conclusion, this project analyzed the effect of current climate trend on yield of four different crops in Ondo State. The study showed that there were variations in the climatic parameters (rainfall, temperature and relative humidity) over the nineteen years’ period of study (1996-2014). There were also variations in the output and yield of the crops under study (cassava, yam, pepper, tomatoes) during the same period, with cassava having the highest output mean value of 2039.23 mt, and tomatoes having the lowest output mean value of 2039.23 mt. From the result of the analysis, it was concluded that climate variability had a 20.7% effect on the yield of cassava, 18.6% effect on the yield of yam, 26.8% effect on the yield of pepper and 15.5% effect on the yield of tomatoes in Ondo State during the nineteen years of study. However, the empirical evidence established strong credibility that, overall, climate variability and change adversely impacted yield and cropping area of major these fruit and tuber crops in Ondo State. The project research also shows that some other factors affects the output and yield of crops in Ondo State Nigeria, non-climatic factors, such as farm management practices, soil fertility, pests, seed type and quality and planting period may contribute significantly to variations in crop yield. Inappropriate management practices such as soil compaction during the site clearing and preparation, topsoil and litter repositioning, burning of debris, harvesting methods and management of harvest residues (Chen et al., 2004; Onyekwelu et al., 2006) have been reported to influence crop yield. As a result of these, the concerned authority should take proper plan to combat against climate change impacts on agriculture in the country for ensuring food security for the ever increasing population through implementing sustainable agricultural development. In summary, data on crops yield and climate might not depict the true scenario for the different agro-ecological zones of any country about climate change impacts on crops yield (Amin et al., 2015). Therefore, region-specific research should be conducted to highlight regional differences and to guide intensive measures on the perspective of climate change and cultivation of major food crops in agro-ecological zone of Nigeria.

The authors have not declared any conflict of interests.