Can unmanned aerial vehicle images be used to estimate forage production parameters in agroforestry systems in the Caatinga?

Authors: MARTINS DOS SANTOS, WAGNER - Federal Rural University Of Pernambuco; PINHEIRO COSTA, CLAUDENILDE - Federal Rural University Of Pernambuco; DA SILVA MEDEIROA, MARIA - Federal Rural University Of Pernambuco; FERRAZ JARDIM, ALEXANDRE - Federal Rural University Of Pernambuco; DA CUNHA, MARCIO - Federal Rural University Of Pernambuco; DUBEUX, JOSE - University Of Florida; Jaramillo, David; CEZAR BEZERRA, ALAN - Federal Rural University Of Pernambuco; DE SOUZA, EVARISTO JORGE - Federal Rural University Of Pernambuco

Submitted to: Applied Sciences
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/1/2024
Publication Date: 6/5/2024
Citation: Martins Dos Santos, W., Pinheiro Costa, C., Da Silva Medeiroa, M., Ferraz Jardim, A., Da Cunha, M., Dubeux, J., Jaramillo, D.M., Cezar Bezerra, A., De Souza, E. 2024. Can unmanned aerial vehicle images be used to estimate forage production parameters in agroforestry systems in the Caatinga?. Applied Sciences. https://doi.org/10.3390/app14114896.
DOI: https://doi.org/10.3390/app14114896

Interpretive Summary: The Caatinga biome is the third most degraded biome in Brazil. In part, this degradation is due to inadequate grazing practices, which are driven by the difficulty of monitoring and estimating the yield parameters of forage plants, especially in agroforestry systems (AFS) in this biome. This study compared the predictive ability of various leaf area and biomass indexes for predicting forage biomass using unmanned aerial vehicle technologies. The results indicate positive correlations using normalized green red difference and visible atmospherically resistant indices with forage biomass predictions. Furthermore, removing trees from the images further improved the predictive power of the biomass estimates. Overall, this study provides evidence for using readily available software and technologies for developing biomass prediction tools for grazing management in degraded ecosystems.

Technical Abstract: The Caatinga biome is the third most degraded biome in Brazil. In part, this degradation is due to inadequate grazing practices, which are driven by the difficulty of monitoring and estimating the yield parameters of forage plants, especially in agroforestry systems (AFS) in this biome. This study aimed to compare the predictive ability of different indexes on the biomass and leaf area index of forage crops (bushveld signal grass and buffel grass) in AFS in the Caatinga biome and to evaluate the influence of removing system components on prediction-model performance. The normalized green red difference index (NGRDI) and visible atmospherically resistant (VARI) showed higher correlations (P<0.05) with the variables. In addition, removing trees from the orthomosaics was the approach that most favored the correlation values. The models based on classification and regression trees (CART) showed lower root mean squared error values, presenting values of 3020.86, 1201.75, and 0.20 for green biomass, dry biomass, and leaf area index, respectively, as well as higher concordance correlation coefficient values (0.94). Using NGRDI and VARI, removing trees from the images, and using CART are recommended in estimating biomass and leaf area index in agroforestry systems in the Caatinga biome.