Abstract
Soil respiration plays a crucial role in the worldwide carbon cycle and displays great sensitivity to shifts in soil temperature and moisture levels. Accurate prediction of soil respiration under different ratios of food waste compost (FWC) amended soil in various variables requires a clear understanding of the processes involved. This research introduces an appropriate model aimed at estimating soil respiration. This paper employed three distinct regression models: multiple linear (Model 1), first-order polynomial (Model 2), and second-order polynomial (Model 3). These models were employed to predict soil respiration by assessing its relationship with various factors. The study examined several factors, including FWC amended ratio (A), pH (B), electrical conductivity (EC) (C), organic matter (OM) (D), carbon-to-nitrogen ratio (C/N) (E), moisture content (F), porosity (G), and microbial count (H). These factors were considered potential influencers of the CO2 efflux response. It was observed that A,B,C and E exhibited p-values below 0.05 signifying their significance in the context of the study. Among the regression models, Model 3 demonstrated the lowest mean squared error of prediction (MSEP) and root mean square error (RMSE), 1.142 % and 0.153, respectively. The suitability of Model 3 for predicting soil respiration was attributed to its capacity to account for interaction effects among independent variables. Conversely, the results indicated that a non-linear model provide a better understanding of soil respiration under different ratios of FWC amended soil due to the smallest MSEP and RMSE, suggesting that the predictive model for CO2 efflux aligned more with second-order behaviour.
