Innovation in forage evaluation: digestibility and metabolizable energymodeling with canonical correlation analysis and random forest in Colombia
DOI:
https://doi.org/10.15649/2346075X.5664Keywords:
Tropical Forages, Digestibility, Metabolizable energy, Predictive modelAbstract
Introduction. Livestock productivity in Colombia largely depends on forage-based feeding systems, where nutritional quality directly influences digestibility and metabolizable energy in ruminants (ME-r). Objectives. To analyze the AlimenTro database in order to evaluate the relationship between nutrients, digestibility, and ME of tropical forages using Canonical Correlation Analysis (CCA) and the Random Forest algorithm. Materials and Methods. A total of 500 samples of grasses, legumes, and shrubs from the AlimenTro database were evaluated using standardized methodologies, including nearinfrared spectroscopy (NIRS), proximate analysis, and in situ techniques. CCA was applied to identify multivariate associations between nutrients and energy parameters, while the Random Forest algorithm was used to assess predictive performance and variable importance, yielding low root mean square error (RMSE) values. Results. CCA revealed that fiber fractions, including neutral detergent fiber (NDF), acid detergent fiber (ADF), lignin, and hemicellulose, were negatively associated with digestibility and ME, whereas crude protein and starch showed positive associations. The Random Forest model achieved R² values greater than 0.95, with low RMSE and mean absolute error (MAE), confirming its high predictive accuracy and highlighting fiber and protein as key predictors. Conclusions. The integration of CCA and Random Forest provides a robust and applicable approach for predicting the nutritional quality of tropical forages. This methodological framework supports the development of more efficient and sustainable feeding strategies in tropical livestock systems.
References
1. Ganadero. Ganadero Agosto de 2023. Bogotá: Fedegán; 2023. Disponible en: https://bibliotecadigital.ccb.org.co/server/api/core/bitstreams/8084ec30-c67b-4ad8-8ccd-755a0c8e0ac9/content
2. Da Cruz CH, Santos SA, de Carvalho GGP, Azevedo JAG, Detmann E, Valadares Filho SC, et al. Estimating digestible nutrients in diets for small ruminants fed with tropical forages. Livest Sci. 2021;249:104532. https://doi.org/10.1016/j.livsci.2021.104532
3. Francy da Costa Backsman C, Monção FP, Aspiazu I, Júnior VRR, Figueiredo Portugal A, et al. Agronomic traits, fermentation quality, chemical composition, and silage digestibility of different forage sorghum genotypes and biomass in the semi-arid region of Brazil. J Appl Anim Res. 2025;53(1). https://doi.org/10.1080/09712119.2025.2462573
4. López Jara AG, Reta Sánchez DG, Santana OI, Reyes Gonzáles A, Rodríguez Hernández K, et al. Rendimiento de forraje y valor nutritivo del ensilado de forrajes alternativos y tradicionales de otoño-invierno. Rev Mex Cienc Pecu. 2024;16(1):208-23. https://doi.org/10.22319/rmcp.v16i1.6301
5. Hernández-Arboleda X, Ortiz-Grisales S, Vivas-Arturo WF, Fernández-Romay Y, O-León O, et al. Nutritional value and in vitro dry matter degradability in Mexican sunflower: Tithonia diversifolia Helms (Gray). Trop Subtrop Agroecosyst. 2024;27(3). https://doi.org/10.56369/tsaes.5211
6. Smith C, Karunaratne S, Badenhorst P, Cogan N, Spangenberg G, Smith K. Machine learning algorithms to predict forage nutritive value of in situ perennial ryegrass plants using hyperspectral canopy reflectance data. Remote Sens. 2020;12(6):60928. https://doi.org/10.3390/rs12060928
7. Zwick M, Cardoso JA, Gutiérrez-Zapata DM, Cerón-Muñoz M, Gutiérrez JF, et al. Pixels to pasture: Using machine learning and multispectral remote sensing to predict biomass and nutrient quality in tropical grasslands. Remote Sens Appl Soc Environ. 2024; 36: 101282. https://doi.org/10.1016/j.rsase.2024.101282
8. Gallo A, Moschini M, Cerioli C, Masoero F. Use of principal component analysis to classify forages and predict their calculated energy content. Animal. 2013;7(6):930-9. https://doi.org/10.1017/S1751731112002467
9. Culqui L, Huaman Pilco AF, Juarez Contreras L, Vigo CN, Goñas M, et al. Nutritional potential of native shrub species for cattle feeding in northeastern Peru. Rangeland Ecol Manag. 2025;98:600-8. https://doi.org/10.1016/j.rama.2024.11.004
10. Palmonari A, Gallo A, Fustini M, Canestrari G, Masoero F, et al. Estimation of the indigestible fiber in different forage types. J Anim Sci. 2016;94:248-54. https://doi.org/10.2527/jas.2015-9649
11. Ministerio de Agricultura y Desarrollo Rural. Alimentos del trópico para alimentación animal - AlimenTro. Datos Abiertos Colombia. 2020. Disponible en: https://www.datos.gov.co/Agricultura-y-Desarrollo-Rural/Alimentos-Del-Tr-Pico-Para-Alimentaci-n-Animal-Ali/6arb-D547
12. Villalpando P, Castillo JA, Cortez K. Análisis de correlación canónica (ACC) e investigación científica. ResearchGate. 2007. Disponible en: https://www.researchgate.net/publication/315813696
13. Hair J, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. 7ª ed. New Jersey: Pearson Education; 2010.
14. González Ariza A, Navas González FJ, Arando Arbulu A, León Jurado JM, Delgado Bermejo JV, et al. Discriminant canonical analysis as a tool to determine traces of endangered native hen breed introgression through egg hatchability phenomics. Anim Biosci. 2022;35(7):915-24.
15. Bock H, Baier AD, Gaul CW, Critchley KF, Vichi MM, et al. Data Analysis, Machine Learning and Knowledge Discovery. New York. Springer, 2014.
16. Díaz Monroy LG, Morales Rivera MA. Análisis estadístico de datos multivariados. 1ª ed. Bogotá: Universidad Nacional de Colombia, 2012.
17. Breiman L. Random forests. Mach Learn. 2001;45(1):5-32. https://doi.org/10.1023/A:1010933404324
18. Jeong JH, Resop JP, Mueller ND, Fleisher DH, Yun K, et al. Random forests for global and regional crop yield predictions. PLoS One. 2016;11(6):e0156571. https://doi.org/10.1371/journal.pone.0156571
19. Pepeta BN, Moyo M, Adejoro FA, Hassen A, Nsahlai IV. Techniques used to determine botanical composition, intake, and digestibility of forages by ruminants. Agronomy. 2022;12(10):2456. https://doi.org/10.3390/agronomy12102456
20. Lee M, Kim DH, Seo S, Tedeschi LO. Development of machine learning models for estimating metabolizable protein supply from feed in lactating dairy cows. Animals. 2025;15(5):1068. https://doi.org/10.3390/ani15050687
21. Sherry A, Henson RK. Conducting and interpreting canonical correlation analysis in personality research: A user-friendly primer. J Pers Assess. 2005;84(1):37-48. https://doi.org/10.1207/s15327752jpa8401_09
22. Jiang MZ, Aguet F, Ardlie K, Chen J, Cornell E, et al. Canonical correlation analysis for multi-omics: Application to cross-cohort analysis. PLoS Genet. 2023;19(5):e1010517. https://doi.org/10.1371/journal.pgen.1010517
23. Martínez Lobo DS. Análisis de la relación entre las pruebas Saber Pro y los cursos realizados por estudiantes de licenciatura en matemáticas utilizando correlación canónica. Bucaramanga: Universidad Industrial de Santander; 2013.
24. Quintero E, Martínez D. Data Science in Secondary Education: Exploring Correlations and Predicting Saber 11 Test Results from the Formative Process. Comunicaciones en Estadística. 2025; 18(2): 33-42. https://doi.org/10.15332/23393076.11820
25. Diel MI, Dal'Col Lúcio A, Lambrecht DM, Vinícius M, Pinheiro M, et al. Canonical correlations in agricultural research: Method of interpretation used leads to greater reliability of results. Int J Innov Educ Res. 2020;7:1-10. https://doi.org/10.31686/ijier.vol8.iss7.2464
26. Lourencon RV, Patra AK, Ribeiro LPS, Puchala R, Wang W, et al. Effects of the level and source of dietary physically effective fiber on feed intake, nutrient utilization, heat energy, ruminal fermentation, and milk production by Alpine goats. Anim Nutr. 2024;17:312-24. https://doi.org/10.1016/j.aninu.2024.02.002
27. Leishman EM, Sahar M, Cieslar S, Darani P, Ellis JL. What the hay: predicting equine voluntary forage intake using a meta-analysis approach. Animal. 2024;18(9):101266. https://doi.org/10.1016/j.animal.2024.101266
28. Jung HG, Allen MS. Characteristics of plant cell walls affecting intake and digestibility of forages by ruminants. J Anim Sci. 1995 Sep;73(9):2774-90. https://doi.org/10.2527/1995.7392774x
29. Gomes DI, Detmann E, Valadares Filho SC, Fukushima RS, de Souza MA, et al. Evaluation of lignin contents in tropical forages using different analytical methods and their correlations with degradation of insoluble fiber. Anim Feed Sci Technol. 2011;168(3-4):206-22. https://doi.org/10.1016/j.anifeedsci.2011.05.001
30. Maskaľová I, Vajda V, Timkovičová Lacková P. Estimation of ruminal digestibility of nutrient and intestinal digestibility of un-degradable proteins at different feedstuffs. Acta Fytotechn Zootechn. 2024;27(1):8-17.. https://doi.org/10.15414/afz.2024.27.01.8-17
31. Guo X, Sun L, Zheng Z, Diao X, He L, et al. Study on rumen degradability and intestinal digestibility of mutton sheep diets with different concentrate-to-forage ratios and nonfiber carbohydrates/neutral detergent fiber ratios. Animals. 2024;14(19):2816. https://doi.org/10.3390/ani14192816
32. Shishir MSR, Cullen B, Brodie G, Zhong R, Cheng L. Potential of feeding microwave-treated forage hays to improve sheep intake, digestion, nitrogen partitioning, and metabolism. Anim Feed Sci Technol. 2024;315:116008. https://doi.org/10.1016/j.anifeedsci.2024.116008
33. Raffrenato E, Nicholson CF, Van Amburgh ME. Development of a mathematical model to predict pool sizes and rates of digestion of 2 pools of digestible neutral detergent fiber and an undigested neutral detergent fiber fraction within various forages. J Dairy Sci. 2019;102(1):351-64. https://doi.org/10.3168/jds.2018-15102
Downloads
Published
How to Cite
Downloads
Issue
Section
License
Copyright (c) 2026 Innovaciencia
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
All articles published in this scientific journal are protected by copyright. The authors retain copyright and grant the journal the right of first publication, with the work simultaneously licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0), which permits sharing the work with authorship recognition and without commercial purposes.
Readers may copy and distribute the material from this journal issue for non-commercial purposes in any medium, provided the original work is cited and credit is given to the authors and the journal.
Any commercial use of the material from this journal is strictly prohibited without written permission from the copyright holder.
For more information on the copyright of the journal and open access policies, please visit our website.
