Statistical Analysis of Climate Change Effects on Crop Yield in Benue State
Notice: This is a sample project for study and reference. Submitting it as your own work violates most universities' academic integrity policies.
Abstract
About This Research Topic
Benue State feeds Nigeria. Producing roughly 60% of the nation's yams and significant shares of rice, maize and sorghum, it employs over 80% of its workforce in agriculture. Yet the very climate that makes the Guinea Savannah productive is shifting. Farmers across its 23 local government areas now report later onset of rains, shorter growing seasons, and more frequent floods like those in 2012 and 2022.
This study provides a 39-year statistical assessment of those shifts, using secondary data from 1985 to 2023 on annual rainfall and mean temperature from the Nigerian Meteorological Agency Makurdi Station and crop yields from the Benue State Agricultural Development Programme. Rather than relying on anecdote, it applies a complete inferential chain: descriptive decade analysis, Mann-Kendall trend testing with Sen's slope, Pearson and Spearman correlation, simple and multiple linear regression, and classical additive time series decomposition.
The approach matters because most existing Nigerian studies use national aggregates or single-crop correlations without trend significance testing. By combining non-parametric trend detection robust to outliers with regression models that isolate joint effects of rainfall and temperature, this analysis delivers evidence that extension services and policymakers can act on. Readers unfamiliar with climate-agriculture linkages can start with our guide to climate change adaptation strategies in African agriculture.
Main Abstract
This research conducts a comprehensive statistical evaluation of climate change impacts on major crop yields in Benue State, Nigeria, known as the Food Basket of the Nation. Utilizing 39 years of annual observations from 1985 to 2023, the study integrates climate data on total annual rainfall and mean annual temperature from NiMet Makurdi and yield data for maize, rice, sorghum and yam from BNARDA. Descriptive analysis shows mean annual rainfall declined from 1,487 mm in 1985-1994 to 1,312 mm in 2015-2023, a drop of 175 mm (11.8%), while mean temperature rose from 27.1°C to 28.4°C (+1.3°C). Non-parametric Mann-Kendall tests confirm a significant declining rainfall trend (Kendall's tau = -0.341, p = 0.008, Sen's slope = -4.87 mm/year) and a significant rising temperature trend (tau = 0.487, p < 0.001, slope = 0.034°C/year). Pearson correlation indicates significant positive associations between rainfall and yields of maize (r = 0.712), rice (r = 0.689), sorghum (r = 0.624) and yam (r = 0.541), all p < 0.001, while temperature correlates negatively with all four crops. Multiple linear regression with rainfall and temperature as joint predictors explains 63.7% of maize yield variance (F = 31.49, p < 0.001), 57.8% for rice, 48.7% for sorghum and 41.2% for yam, with both predictors independently significant. Time series decomposition reveals declining trend components for maize, rice and sorghum, accelerating post-2010. The findings support promotion of drought-tolerant varieties, smallholder irrigation expansion, and strengthened agrometeorological advisory systems in Benue State.
Chapter One Preview
Background to the Study
Climate change is now unequivocally human-driven. The IPCC Sixth Assessment Report concluded that global mean surface temperature is about 1.1°C above pre-industrial levels and that 1.5°C will likely be exceeded in the early 2030s without deep emission cuts. For agriculture, warming shortens crop cycles, raises evapotranspiration, disrupts pollinator phenology, expands pest ranges, and intensifies droughts and floods. Shifts in precipitation timing are especially disruptive in rain-fed systems where planting calendars depend on monsoon onset.
Sub-Saharan Africa is disproportionately exposed. West Africa is projected to warm 1.5 to 3 times the global average under high-emission scenarios, while the West African Monsoon becomes more variable. Farming is predominantly rain-fed, with limited irrigation, insurance, or improved seed access. In Nigeria, agriculture contributes about 24% of GDP and livelihoods for an estimated 36 million smallholder households, yet NiMet records show warming of 0.25–0.30°C per decade since 1960. Studies project maize yield declines of 16–40% by 2050 without adaptation.
Benue State lies in the Guinea Savannah agroecological zone with a single 5-7 month rainy season. Its ferruginous soils support yam, cassava, rice, maize, sorghum, millet and soybean. Despite its productivity, continuous records from Makurdi Station since 1975 show rising temperatures, falling and more erratic rainfall, and extreme floods. Farmer perceptions of later onset and earlier cessation align with instrumental data, increasing crop failure risk.
Existing literature on Nigeria often uses national aggregates, focuses on one crop or climate variable, or stops at descriptive statistics without Mann-Kendall testing or multivariate regression. This study closes that methodological and geographic gap by applying an integrated framework to state-level multi-decadal data. The U.S. National Oceanic and Atmospheric Administration provides baseline context on how climate change alters precipitation extremes globally, while FAO highlights why such changes threaten food security in Africa. For students designing similar work, see our resource on agricultural research methods and data analysis.
NOAA – Climate Change Impacts on Weather Extremes
FAO – Climate Change and Food Security
Statement of the Problem
Benue State's agriculture faces an escalating climate crisis that threatens food security, rural incomes, and its national Food Basket role. Multi-decadal climate and yield records exist at NiMet and BNARDA, yet no published study has systematically linked observed trends to crop-specific outcomes using robust trend tests and multivariate models for Benue State alone. Without quantified estimates of direction, magnitude and significance, extension officers and policymakers lack evidence on which crops are most vulnerable, whether rainfall decline or temperature rise drives losses, and what adaptation investments are justified. Methodologically, many Nigerian studies omit Mann-Kendall non-parametric trend detection, which is best suited to hydro-meteorological series with outliers and non-normality, and omit multiple regression that controls for confounded climate predictors. This study addresses both substantive and methodological gaps.
Aim and Objectives
The aim is to statistically analyse climate change effects on yields of four major crops in Benue State using 39 years of secondary data (1985–2023).
1. Describe trends in annual rainfall and mean temperature using descriptive statistics and decade-by-decade decomposition.
2. Test for significant monotonic trends in climate variables and crop yields using Mann-Kendall test and quantify magnitude with Sen's slope.
3. Quantify bivariate relationships between climate variables and crop yields using Pearson and Spearman correlation.
4. Develop simple linear regression models for individual effects of rainfall and temperature on each crop yield.
5. Develop multiple linear regression models examining joint effects of rainfall and temperature on maize, rice, sorghum and yam yields.
6. Decompose crop yield series into trend, seasonal and irregular components via classical additive decomposition.
7. Propose evidence-based adaptation recommendations for Benue State's agricultural sector.
Research Questions
1. Have annual rainfall and mean temperature in Benue State shown statistically significant monotonic trends between 1985 and 2023?
2. What are the magnitudes and directions of associations between climate variables and yields of maize, rice, sorghum and yam?
3. Do annual rainfall and mean temperature jointly and significantly predict crop yields when modeled together?
4. What are the long-term trend components in crop yield series after controlling for cyclical and irregular variation?
Significance of the Study
This study provides the first comprehensive multi-decade statistical linkage for Benue State, informing BNARDA on variety promotion – which crops need drought and heat tolerance most – and on whether rainfall scarcity or heat stress dominates yield loss. For policy, it supplies quantitative justification for irrigation investment, drought-tolerant seed systems and weather advisory services under Benue's agricultural development plan and Nigeria's Nationally Determined Contribution to the Paris Agreement, which prioritizes agricultural adaptation. Academically, it demonstrates a replicable integrated framework combining Mann-Kendall, correlation, regression and decomposition applicable to other Nigerian states. Students can compare templates in our collection of statistics project topics for agriculture and environmental studies.
Scope of the Study
The analysis covers 39 years (1985–2023) of state-level secondary data. Climate variables are annual total rainfall (mm) and mean annual temperature (°C) from NiMet Makurdi Station. Crop yield variables are annual yields (kg/ha) for maize, rice paddy, sorghum and yam from BNARDA records. The study uses annual aggregates; disaggregation by local government area, inclusion of solar radiation, humidity, CO2, soil properties, and lagged soil moisture effects are beyond scope due to data constraints.
Operational Definition of Terms
Crop Yield: Agricultural output per unit area, kg per hectare, averaged at state level for each year.
Climate Change: Long-term shifts in temperature, precipitation and extremes attributable to anthropogenic greenhouse gas increases, as assessed by the IPCC.
Mann-Kendall Trend Test: Non-parametric test for monotonic trend in time series, robust to non-normality and outliers; widely used in climatology.
Sen's Slope: Median of all pairwise slopes, providing robust estimate of trend magnitude per year.
Pearson Correlation: Parametric measure of linear association, r ranging -1 to +1.
Spearman Rank Correlation: Non-parametric rank-based measure of monotonic association, used as robustness check.
Time Series Decomposition: Separation of series into trend, seasonal/cyclical and irregular components: Yt = Tt + St + It (additive model).
Guinea Savannah Zone: Agroecological zone with 5-7 month unimodal rainy season in north-central Nigeria supporting both cereals and tubers.
IPCC AR6 Working Group I – Physical Science Basis
NOAA NCEI – Climate Data Access
Short Conclusion
Over 39 years, Benue State shows a statistically significant drying and warming signal: -4.87 mm rainfall per year and +0.034°C per year, totaling -11.8% rainfall and +1.3°C between first and last decades. This climate shift is significantly associated with reduced yields of maize, rice, sorghum and yam, with rainfall positively and temperature negatively correlated. Joint regression models explain 41-64% of yield variance, strongest for maize. Trend decomposition confirms declining yield trends accelerating after 2010. Adaptation must prioritize drought-resistant and heat-tolerant varieties, expansion of smallholder irrigation, timely agrometeorological advisories, and integration of climate-smart practices into Benue's agricultural plan. Future work should incorporate LGA-level spatial data, additional climate variables, and non-climatic agronomic controls. Explore more evidence-based guides on crop production and climate adaptation on our platform.
Frequently Asked Questions
Q: Why focus on Benue State?
A: Benue produces about 60% of Nigeria's yam and large shares of rice, maize and sorghum, employing over 80% of its workforce in agriculture, making climate impacts there nationally significant for food security.
Q: What data sources were used?
A: Annual rainfall and mean temperature from NiMet Makurdi Station and crop yields for maize, rice, sorghum and yam from Benue State Agricultural Development Programme (BNARDA), 1985-2023.
Q: What is Mann-Kendall test and why use it?
A: A non-parametric test detecting monotonic upward or downward trends without assuming normality, robust to outliers and missing values, ideal for hydro-meteorological series.
Q: What did Mann-Kendall results show?
A: Significant declining rainfall trend (tau -0.341, p 0.008, Sen slope -4.87 mm/yr) and significant rising temperature trend (tau 0.487, p <0.001, slope 0.034°C/yr); thus H01 and H04 rejected.
Q: How strongly does rainfall correlate with yields?
A: Pearson r = 0.712 for maize, 0.689 rice, 0.624 sorghum, 0.541 yam, all p<0.001 positive; temperature negative for all crops; H02 rejected.
Q: Do rainfall and temperature jointly predict yields?
A: Yes. Multiple regression R² = 63.7% maize, 57.8% rice, 48.7% sorghum, 41.2% yam, both predictors significant; H03 rejected.
Q: Why use both Pearson and Spearman correlation?
A: Pearson captures linear relationships; Spearman provides non-parametric robustness check for monotonic relationships insensitive to outliers and non-normality.
Q: What are the limitations?
A: State-level aggregates mask LGA heterogeneity, only two climate variables included, contemporaneous annual model ignores lagged soil moisture, yield data includes non-climatic influences like fertilizer and Fall Army Worm, single weather station introduces spatial uncertainty.
Q: What adaptations are recommended?
A: Adopt drought-resistant and heat-tolerant varieties, expand smallholder irrigation, strengthen agricultural weather advisory services, and embed climate adaptation in state agricultural plans.
Q: How can this method be replicated elsewhere?
A: Use same chain: descriptive decade analysis, Mann-Kendall + Sen slope, Pearson/Spearman correlation, simple and multiple regression, additive time series decomposition, with state-level NiMet and ADP data.
Purchase to unlock the full material.