COMPARATIVE EVALUATION OF MACHINE LEARNING MODELS FOR CACAO YIELD PREDICTION USING SOIL PHYSICOCHEMICAL PROPERTIES
DOI:
https://doi.org/10.69980/zc8fsm72Keywords:
Cacao yield prediction, Soil physicochemical properties, Machine learning, Random Forest Regression, Precision agricultureAbstract
Cacao productivity is influenced by soil physicochemical characteristics, making accurate yield prediction essential for sustainable agricultural management. This study aimed to evaluate relationships between soil properties and cacao yield, compare machine learning models, and identify influential soil predictors. Secondary agricultural data were obtained from the publicly available Mendeley Data repository, comprising 150 cacao plantations across three climatic zones in Ghana. Following data preprocessing, 149 complete observations were included in predictive modelling. Descriptive statistics, Pearson correlation analysis, and four regression models like Linear Regression, Decision Tree Regression, Random Forest Regression, and Support Vector Regression were applied. Model performance was evaluated using the coefficient of determination (R²), root mean square error (RMSE), and mean absolute error (MAE). Results revealed significant correlations between cacao yield and exchangeable acidity (r = 0.601), soil pH (r = -0.600), and sand content (r = -0.504). Random Forest Regression demonstrated the best overall predictive performance, achieving an R² of 0.487 and RMSE of 375.07 kg/ha/year. Decision Tree Regression achieved the lowest MAE of 274.23 kg/ha/year. Feature importance analysis identified soil pH, exchangeable acidity, and sodium as influential predictors. The findings demonstrate the potential of soil-based machine learning approaches for cacao yield estimation, although moderate predictive accuracy highlights the need for additional environmental and management variables to improve agricultural prediction and decision-making.
References
1.yield gaps of smallholder cocoa systems and the main determining factors along a climate gradient in Ghana. Agricultural Systems, 181, 102812.
2.Abdulai, I., Jassogne, L., Graefe, S., Asare, R., Van Asten, P., Läderach, P., & Vaast, P. (2018). Characterization of cocoa production, income diversification and shade tree management along a climate gradient in Ghana. PloS one, 13(4), e0195777.
3.Abdulai, I., Muilerman, S., Jassogne, L., Hoffmann, M. P., van Asten, P., Rötter, R. P., Tao, H. H., Vaast, P., Asare, R., Läderach, P., & Graefe, S. (2020b). Data for: Variations in yield gaps and soil fertility along a climatic gradient in smallholder cocoa systems of Ghana (Version 1) [Data set]. Mendeley Data. https://doi.org/10.17632/bcrd9g226g.1
4.Adet, L., Rozendaal, D. M., Zuidema, P. A., Vaast, P., & Anten, N. P. (2024). Cocoa tree performance and yield are affected by seasonal rainfall reduction. Agricultural Water Management, 302, 108995.
5.Agbotui, D. K., Ingold, M., & Buerkert, A. (2024). Carbon and nutrient cycling in cocoa agroforests under organic and conventional management. Nutrient Cycling in Agroecosystems, 129(1), 7-20.
6.Amponsah-Doku, B., Daymond, A., Robinson, S., Atuah, L., & Sizmur, T. (2022). Improving soil health and closing the yield gap of cocoa production in Ghana–A review. Scientific African, 15, e01075.
7.Anokye, J., Abunyewa, A. A., Jørgensen, U., Kaba, J. S., Twum-Ampofo, K., Dawoe, E., ... & Ulzen, J. (2024). Mitigation of greenhouse gas emissions through shade systems and climate-smart soil fertility interventions in cocoa landscapes in the semi-deciduous ecological zone of Ghana. Soil Advances, 1, 100001.
8.Ariza-Salamanca, A. J., Navarro-Cerrillo, R. M., Crozier, J., Stirling, C., Mancini, A., Blaser-Hart, W., & González-Moreno, P. (2024). Drivers of cocoa yield and growth in young monoculture and agroforestry systems. Agricultural Systems, 219, 104044.
9.Asitoakor, B. K., Vaast, P., Ræbild, A., Ravn, H. P., Eziah, V. Y., Owusu, K., ... & Asare, R. (2022). Selected shade tree species improved cocoa yields in low-input agroforestry systems in Ghana. Agricultural Systems, 202, 103476.
10.Cheng, E., Zhang, B., Peng, D., Zhong, L., Yu, L., Liu, Y., ... & Yang, S. (2022). Wheat yield estimation using remote sensing data based on machine learning approaches. Frontiers in Plant Science, 13, 1090970.
11.de Freitas Cunha, R. L., & Silva, B. (2020, March). Estimating crop yields with remote sensing and deep learning. In 2020 IEEE Latin American GRSS & ISPRS Remote Sensing Conference (LAGIRS) (pp. 273-278). IEEE.
12.Doe, E. K., Attua, E. M., Obour, P. B., Quaye, A. K., & Fosu-Mensah, B. Y. (2023). Soil health and synergy of ecological determinants of green cocoa productivity in different soil ecotypes in Ghana. Frontiers in Sustainable Food Systems, 7, 1169015.
13.Dogbatse, J. A., Awudzi, G. K., Arthur, A., Owusu-Ansah, F., Amoako-Attah, I., Quaye, A. K., ... & Amaning, A. A. (2024). Assessment of calcium nitrate fertilizer as a suitable nitrogen source for cacao (Theobroma cacao L.) cultivation in Ghana. International Journal of Agronomy, 2024(1), 5578534.
14.Goudsmit, E., Rozendaal, D. M., Tosto, A., & Slingerland, M. (2023). Effects of fertilizer application on cacao pod development, pod nutrient content and yield. Scientia Horticulturae, 313, 111869.
15.Iniyan, S., Varma, V. A., & Naidu, C. T. (2023). Crop yield prediction using machine learning techniques. Advances in Engineering Software, 175, 103326.
16.Jabed, M. A., & Murad, M. A. A. (2024). Crop yield prediction in agriculture: A comprehensive review of machine learning and deep learning approaches, with insights for future research and sustainability. Heliyon, 10(24).
17.Kaneko, H. (2022). Cross‐validated permutation feature importance considering correlation between features. Analytical Science Advances, 3(9-10), 278-287.
18.Khaki, S., Pham, H., & Wang, L. (2021). Simultaneous corn and soybean yield prediction from remote sensing data using deep transfer learning. Scientific Reports, 11(1), 11132.
19.Konaté, N. G., Ouattara, Y., Kouakou, A. K., & Barima, Y. S. (2024). Effects of traditional agroforestry practices on cocoa yields in Côte d’Ivoire. Sustainability, 16(22), 9927.
20.Leukel, J., Zimpel, T., & Stumpe, C. (2023). Machine learning technology for early prediction of grain yield at the field scale: A systematic review. Computers and Electronics in Agriculture, 207, 107721.
21.Mendoza-Meneses, C. J., Feregrino-Pérez, A. A., Guevara-González, R. G., & García-Trejo, J. F. (2023). Implementation of pre-harvest techniques in emerging agroforestry systems to increase the yield of cocoa tree (Theobroma cacao L.). Heliyon, 9(3).
22.Mwafulirwa, L., Sizmur, T., Daymond, A., Atuah, L., Quaye, A. K., Coole, S., ... & Hadley, P. (2024). Cocoa pod husk-derived organic soil amendments differentially affect soil fertility, nutrient leaching, and greenhouse gas emissions in cocoa soils. Journal of Cleaner Production, 479, 144065.
23.Oikonomidis, A., Catal, C., & Kassahun, A. (2023). Deep learning for crop yield prediction: a systematic literature review. New Zealand Journal of Crop and Horticultural Science, 51(1), 1-26.
24.Paudel, D., Boogaard, H., de Wit, A., van der Velde, M., Claverie, M., Nisini, L., ... & Athanasiadis, I. N. (2022). Machine learning for regional crop yield forecasting in Europe. Field Crops Research, 276, 108377.
25.Saputra, D. D., Khasanah, N. M., Sari, R. R., & van Noordwijk, M. (2024). Avoidance of tree-site mismatching of modelled cacao production systems across climatic zones: Roots for multifunctionality. Agricultural Systems, 216, 103895.
26.Schmidt, J., Addo-Danso, S. D., Asare, R., Tettey, A., & Isaac, M. E. (2024). Soil quality reflects microbial resource availability and drives rhizosphere microbiome variation in Ghanaian cocoa farms. Applied Soil Ecology, 198, 105378.
27.Shawon, S. M., Ema, F. B., Mahi, A. K., Niha, F. L., & Zubair, H. T. (2025). Crop yield prediction using machine learning: An extensive and systematic literature review. Smart Agricultural Technology, 10, 100718.
28.Tripathi, A., Tiwari, R. K., & Tiwari, S. P. (2022). A deep learning multi-layer perceptron and remote sensing approach for soil health based crop yield estimation. International Journal of Applied Earth Observation and Geoinformation, 113, 102959.
29.Van Klompenburg, T., Kassahun, A., & Catal, C. (2020). Crop yield prediction using machine learning: A systematic literature review. Computers and electronics in agriculture, 177, 105709.
30.Vasquez-Zambrano, E., Woittiez, L. S., van Heerwaarden, J., Rusinamhodzi, L., Hauser, S., & Giller, K. E. (2025). Deriving fertiliser recommendations for cocoa: An offtake model approach. European Journal of Agronomy, 164, 127463.
31.Xu, H., Zhang, X., Ye, Z., Jiang, L., Qiu, X., Tian, Y., ... & Cao, W. (2021). Machine learning approaches can reduce environmental data requirements for regional yield potential simulation. European Journal of Agronomy, 129, 126335.
