André Nogueira Bozza
Research Subjects
André Nogueira Bozza has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Diego Campos Vasconcelos
Departamento Regional MS, Serviço Nacional de Aprendizagem Industrial (SENAI), Campo Grande 79005-901, MS, Brazil
2 shared publications - João Gabriel Chaib
Departamento Regional MS, Serviço Nacional de Aprendizagem Industrial (SENAI), Campo Grande 79005-901, MS, Brazil
2 shared publications - Maximiliano Guimarães Sales
Departamento Regional MS, Serviço Nacional de Aprendizagem Industrial (SENAI), Campo Grande 79005-901, MS, Brazil
2 shared publications - Renato Maurício Barbosa Tavares
Departamento Regional MS, Serviço Nacional de Aprendizagem Industrial (SENAI), Campo Grande 79005-901, MS, Brazil
2 shared publications - Rodney Murillo Peixoto Couto
UNESP
2 shared publications - Rosa Helena da Silva
Instituto de Ciências, Laboratório de Ecologia e Botânica, Universidade Federal de Mato Grosso do Sul, Campo Grande, MS, Brasil
2 shared publications
Latest Publications
- Open Access
Validation Rigor Determines Apparent Predictive Skill of UAV-LiDAR Carbon Models: A Cautionary Case Study in a Heterogeneous Tropical Savanna
Machine-learning models trained on remote sensing data are widely used to map ecosystem carbon stocks, but their validation often ignores spatial autocorrelation, inflating apparent predictive skill. We tested whether this concern, established for large-scale satellite-based biomass mapping, also holds for UAV-LiDAR carbon estimation at the plot scale, comparing continuous structural metrics, segmentation-derived crown metrics, and Sentinel-2 spectral variables for predicting aboveground biomass (AGB) and soil organic carbon (SOC) across 29 inventory plots in a Cerradão-dominated Pantanal–Cerrado ecotone, Brazil. Random Forest models were evaluated under fourteen predictor scenarios and two allometric formulations. A single spatial partition yielded seemingly reasonable accuracy for AGB (best R2 = 0.27), whereas SOC showed weak apparent skill, even under a single partition (best rRMSE = 33.7%, R2 = −0.14); repeating validation across 20 independent spatial partitions, combined with leakage-free nested variable selection, showed predictive skill collapsing to near-zero or negative R2 for all 98 tested scenario–response combinations (mean R2: −0.44 to −0.003), a pattern statistically indistinguishable from a null model built on permuted response data. Variable selection stability differed from full-dataset importance rankings: raw spectral bands and canopy vertical-complexity metrics were the most consistently retained predictors across partitions for AGB and SOC, respectively, even though a derived spectral index and a lower-canopy structural metric dominated single-model importance for each response. Cross-pool coupling tests showed no meaningful gain from using AGB to predict SOC or vice versa. Empirical variogram analysis further indicated that the spatial blocking scheme adequately separated the two response variables, although residual spatial dependence could not be excluded for a subset of predictors. Validation strategy, rather than sensor combination or vegetation representation, was the primary determinant of apparent predictive skill, cautioning against overinterpreting single-partition accuracy in small, heterogeneous tropical landscapes.
Open AccessValidation Rigor Determines Apparent Predictive Skill of UAV-LiDAR Carbon Models: A Cautionary Case Study in a Heterogeneous Tropical Savanna
Machine-learning models trained on remote sensing data are widely used to map ecosystem carbon stocks, but their validation often ignores spatial autocorrelation, inflating apparent predictive skill. We tested whether this concern, established for large-scale satellite-based biomass mapping, also holds for UAV-LiDAR carbon estimation at the plot scale, comparing continuous structural metrics, segmentation-derived crown metrics, and Sentinel-2 spectral variables for predicting aboveground biomass (AGB) and soil organic carbon (SOC) across 29 inventory plots in a Cerradão-dominated Pantanal–Cerrado ecotone, Brazil. Random Forest models were evaluated under fourteen predictor scenarios and two allometric formulations. A single spatial partition yielded seemingly reasonable accuracy for AGB (best R2 = 0.27), whereas SOC showed weak apparent skill, even under a single partition (best rRMSE = 33.7%, R2 = −0.14); repeating validation across 20 independent spatial partitions, combined with leakage-free nested variable selection, showed predictive skill collapsing to near-zero or negative R2 for all 98 tested scenario–response combinations (mean R2: −0.44 to −0.003), a pattern statistically indistinguishable from a null model built on permuted response data. Variable selection stability differed from full-dataset importance rankings: raw spectral bands and canopy vertical-complexity metrics were the most consistently retained predictors across partitions for AGB and SOC, respectively, even though a derived spectral index and a lower-canopy structural metric dominated single-model importance for each response. Cross-pool coupling tests showed no meaningful gain from using AGB to predict SOC or vice versa. Empirical variogram analysis further indicated that the spatial blocking scheme adequately separated the two response variables, although residual spatial dependence could not be excluded for a subset of predictors. Validation strategy, rather than sensor combination or vegetation representation, was the primary determinant of apparent predictive skill, cautioning against overinterpreting single-partition accuracy in small, heterogeneous tropical landscapes.
- Open Access
- 24 July 2026
- Published by MDPI AG in Preprints.org
Validation Rigor Determines Apparent Predictive Skill of UAV-LiDAR Carbon Models: A Cautionary Case Study in a Heterogeneous Tropical Savanna
Machine-learning models trained on remote sensing data are widely used to map ecosystem carbon stocks, but their validation often ignores spatial autocorrelation, inflating apparent predictive skill. We tested whether this concern, established for large-scale satellite-based biomass mapping, also holds for UAV-LiDAR carbon estimation at the plot scale, comparing continuous structural metrics, segmentation-derived crown metrics, and Sentinel-2 spectral variables for predicting aboveground biomass (AGB) and soil organic carbon (SOC) across 29 inventory plots in a Cerradão-dominated Pantanal–Cerrado ecotone, Brazil. Random Forest models were evaluated under fourteen predictor scenarios and two allometric formulations. A single spatial partition yielded seemingly reasonable accuracy for AGB (best R² = 0.27), whereas SOC showed weak apparent skill even under a single partition (best rRMSE = 33.7%, R² = −0.14); repeating validation across 20 independent spatial partitions, combined with leakage-free nested variable selection, showed predictive skill collapsing to near-zero or negative R² for all 98 tested scenario–response combinations (mean R²: −0.44 to −0.003). Variable selection remained comparatively stable across partitions, with a Sentinel-2 shortwave-infrared band and canopy vertical-complexity metrics consistently retained, and cross-pool coupling tests showed no meaningful gain from using AGB to predict SOC or vice versa. Validation strategy, rather than sensor combination or vegetation representation, was the primary determinant of apparent predictive skill, cautioning against overinterpreting single-partition accuracy in small, heterogeneous tropical landscapes.
- 24 July 2026
- Published by MDPI AG in Preprints.org
Open AccessValidation Rigor Determines Apparent Predictive Skill of UAV-LiDAR Carbon Models: A Cautionary Case Study in a Heterogeneous Tropical Savanna
Machine-learning models trained on remote sensing data are widely used to map ecosystem carbon stocks, but their validation often ignores spatial autocorrelation, inflating apparent predictive skill. We tested whether this concern, established for large-scale satellite-based biomass mapping, also holds for UAV-LiDAR carbon estimation at the plot scale, comparing continuous structural metrics, segmentation-derived crown metrics, and Sentinel-2 spectral variables for predicting aboveground biomass (AGB) and soil organic carbon (SOC) across 29 inventory plots in a Cerradão-dominated Pantanal–Cerrado ecotone, Brazil. Random Forest models were evaluated under fourteen predictor scenarios and two allometric formulations. A single spatial partition yielded seemingly reasonable accuracy for AGB (best R² = 0.27), whereas SOC showed weak apparent skill even under a single partition (best rRMSE = 33.7%, R² = −0.14); repeating validation across 20 independent spatial partitions, combined with leakage-free nested variable selection, showed predictive skill collapsing to near-zero or negative R² for all 98 tested scenario–response combinations (mean R²: −0.44 to −0.003). Variable selection remained comparatively stable across partitions, with a Sentinel-2 shortwave-infrared band and canopy vertical-complexity metrics consistently retained, and cross-pool coupling tests showed no meaningful gain from using AGB to predict SOC or vice versa. Validation strategy, rather than sensor combination or vegetation representation, was the primary determinant of apparent predictive skill, cautioning against overinterpreting single-partition accuracy in small, heterogeneous tropical landscapes.
- Open Access
- 1 September 2010
- Published by FapUNIFESP (SciELO) in Biota Neotropica
- Vol. 10 (3) , 217-226
Uso de recursos alimentares por peixes imaturos e adultos de espécies piscívoras em uma planície de inundação neotropical
Este estudo teve como objetivo avaliar o uso dos recursos alimentares por indivíduos imaturos e adultos de Acestrorhynchus lacustris, Hemisorubim platyrhynchos, Hoplias aff. malabaricus, Plagioscion squamosissimus, Pseudoplatystoma corruscans e Salminus brasiliensis, da planície de inundação do alto Rio Paraná, coletados trimestralmente entre março de 2007 e dezembro de 2008. Foi avaliada a composição quantitativa da dieta através da obtenção do volume das presas (% volume) e amplitude de nicho trófico de cada uma das espécies, bem como as inter-relações tróficas entre elas, através da similaridade entre as dietas e tamanho das presas. A dieta das seis espécies foi composta por peixes-presa (42 espécies), crustáceos e insetos, sendo que todas elas, com exceção de P. corruscans, mostraram alterações ontogenéticas. A amplitude de nicho foi maior para H. aff. malabaricus adultos e menor para P. corruscans adultos e P. squamosissimus imaturos. A similaridade na dieta foi maior para aquelas espécies que consumiram Characidae e menor para P. squamosissimus imaturos, os quais consumiram preferencialmente crustáceos. Houve correlação positiva entre o tamanho da presa e do predador, porém cada espécie consumiu presas de tamanhos diferentes, mostrando que elas não são correlacionadas. Os resultados demonstraram que as seis espécies consumiram peixes desde as fases imaturas, porém com pouca similaridade intra e inter-específica na dieta para a maioria delas, e que o tamanho das presas consumidas aumentou em relação ao tamanho do predador e foi variável entre as espécies. Assim, a dieta dos predadores depende da distribuição e colonização das espécies-presas no ambiente, morfologia e comportamento alimentar tanto das presas quanto dos predadores.
- 1 September 2010
- Published by FapUNIFESP (SciELO) in Biota Neotropica
- Vol. 10 (3) , 217-226
Open AccessUso de recursos alimentares por peixes imaturos e adultos de espécies piscívoras em uma planície de inundação neotropical
Este estudo teve como objetivo avaliar o uso dos recursos alimentares por indivíduos imaturos e adultos de Acestrorhynchus lacustris, Hemisorubim platyrhynchos, Hoplias aff. malabaricus, Plagioscion squamosissimus, Pseudoplatystoma corruscans e Salminus brasiliensis, da planície de inundação do alto Rio Paraná, coletados trimestralmente entre março de 2007 e dezembro de 2008. Foi avaliada a composição quantitativa da dieta através da obtenção do volume das presas (% volume) e amplitude de nicho trófico de cada uma das espécies, bem como as inter-relações tróficas entre elas, através da similaridade entre as dietas e tamanho das presas. A dieta das seis espécies foi composta por peixes-presa (42 espécies), crustáceos e insetos, sendo que todas elas, com exceção de P. corruscans, mostraram alterações ontogenéticas. A amplitude de nicho foi maior para H. aff. malabaricus adultos e menor para P. corruscans adultos e P. squamosissimus imaturos. A similaridade na dieta foi maior para aquelas espécies que consumiram Characidae e menor para P. squamosissimus imaturos, os quais consumiram preferencialmente crustáceos. Houve correlação positiva entre o tamanho da presa e do predador, porém cada espécie consumiu presas de tamanhos diferentes, mostrando que elas não são correlacionadas. Os resultados demonstraram que as seis espécies consumiram peixes desde as fases imaturas, porém com pouca similaridade intra e inter-específica na dieta para a maioria delas, e que o tamanho das presas consumidas aumentou em relação ao tamanho do predador e foi variável entre as espécies. Assim, a dieta dos predadores depende da distribuição e colonização das espécies-presas no ambiente, morfologia e comportamento alimentar tanto das presas quanto dos predadores.