Crop Residue Management, Rice Productivity and Emission Intensity in Nigeria: An ARDL Bound Testing Approach A Dual-Equation Time Series Analysis, 1981-2022

Abstract

This study investigates the relationship between crop residue management, rice productivity, and emission intensity in Nigeria using annual data from 1981 to 2022. A Composite Crop Residue Management Index (CRMI) was developed from the share, intensity, and scale of residue burning relative to total residue nitrogen turnover using FAOSTAT crop-residue emissions data and complementary FAOSTAT and World Bank datasets. Two Autoregressive Distributed Lag (ARDL) models were estimated, with rice yield (t/ha) and emission intensity (tCO₂e per tonne of rice) as dependent variables, while CRMI, rainfall, fertilizer use, harvested area, and agricultural land share served as explanatory variables. Stationarity tests confirmed a mixed order of integration, supporting the application of the ARDL bounds-testing approach. Bounds test results indicated no evidence of a long-run cointegrating relationship in either model, a finding further supported by insignificant error correction terms. Diagnostic tests confirmed that both models were free from serial correlation, heteroskedasticity, and ARCH effects, with normally distributed residuals and stable parameters. Despite the absence of long-run equilibrium relationships, short-run estimates showed that CRMI significantly increased emission intensity (coefficient = 0.744, p = 0.042), indicating that residue burning generates immediate greenhouse gas emissions. In contrast, the effect of CRMI on rice yield was negative but statistically insignificant (coefficient = −0.508, p = 0.107). The results suggest that while residue burning does not exert a persistent long-run influence on rice productivity or emission intensity, it imposes a measurable short-run environmental cost without delivering significant productivity gains.