Examinando por Materia "Random forest"

Mostrando 1 - 9 de 9

Agronomic variables outperform multispectral indices for individual plant yield prediction in Andean quinoa
(Elsevier B.V., 2026-04-18) Pizarro Carcausto, Samuel Edwin; García Seguil, Erika Janina; Gavino Lulo, Esthefany Irene; Requena Rojas, Edilson Jimmy; Ortega Quispe, Kevin Abner; Cccopi Trucios, Dennis
Accurate pre-harvest yield estimation is essential for decision-making in high-altitude agriculture. This study evaluated agronomic and multispectral UAV variables for near-harvest prediction of individual quinoa grain weight, with data collected across six phenological stages to identify when predictors achieve reliable performance, under Andean conditions. A total of 374 plants were monitored across six phenological stages at Santa Ana Experimental Station (Huancayo, Peru, 3280 m a.s.l.) during 2024. OLS, Random Forest, Support Vector Machine, and Neural Network models were trained using agronomic-only (AGRO), spectral-only (IND), and combined (COMP) predictor sets, evaluated through 5-fold cross-validation reporting mean ± standard deviation. Agronomic and combined models achieved moderate performance (R² = 0.22–0.25, RPD = 1.10–1.15), suitable for relative plant ranking in breeding programs, while spectral-only models failed across all algorithms (R² ≤ 0.044, CCC ≤ 0.080), constrained by saturation, phenological decoupling, and canopy heterogeneity. Variable importance analysis confirmed that late-season structural traits dominated predictions, while spectral indices contributed marginally despite including red-edge bands. These results challenge spectral-only approaches for individual plant phenotyping in heterogeneous canopies, demonstrating that integrating simple ground measurements with UAV spectral data is essential for reliable quinoa yield estimation.
Assessment of soil fertility variability for maize production in highland agroecosystems of Peru
(Polish Society of Ecological Engineering (PTIE), 2026-04-01) Garcia Seguil, Erika Janina; Ccopi Trucios, Dennis; Requena Rojas, Edilson Jimmy; Sanabria Quispe, Samuel; Arias Arredondo, Alberto Gilmer; Gavino Lulo, Esthefany Irene; Azabache, Andres; Pizarro Carcausto, Samuel Edwin
Maize (Zea mays L.) is central to food, feed, and rural livelihoods, yet the yields in Peru’s highlands remain modest, underscoring the need for spatially explicit soil diagnostics. This study aimed to characterize the spatial variability of soil fertility in a highland maize production area of the southern Mantaro Valley and translate those patterns into site-specific management zones. The authors sampled the arable layer (0–30 cm) at 100 plots and analyzed pH, electrical conductivity, exchangeable acidity, texture, organic matter (OM), total nitrogen (N), available phosphorus (P), available potassium (K), exchangeable cations (Ca, Mg, Na, K), and calcium carbonate (CaCO₃). Laboratory data were integrated with environmental covariates using geostatistics, Random Forest, and GIS to generate high-resolution maps. Results showed uneven distributions in key attributes about 25% of the area with P deficiency, 15% with localized K shortages, and ~20% with OM < 2% while pH and CEC were comparatively stable. Random Forest achieved strong predictive performance for relatively stable properties (e.g., OM, pH, exchangeable cations), whereas mobile nutrients (available P, exchangeable K) were less predictable. The resulting products constitute the first high-resolution soil-fertility baseline for maize in the southern Mantaro Valley. The maps delineate fertilization management zones and provide a practical basis for preliminary rate recommendations that target constraints while avoiding surpluses. Future work will refine these zoned recommendations through yield-response trials, seasonal monitoring of mobile nutrients, and farmercentered decision-support tools, with the goal of improving nutrient-use efficiency, sustaining maize productivity, and reducing environmental risks across the valley.
Climate change and tree cover loss affect the habitat suitability of Cedrela angustifolia evaluating climate vulnerability and conservation in Andean montane forests
(PeerJ Inc., 2025-02-27) Ames Martínez, Fressia N.; Capcha Romero, Ivan; Guerra, Anthony; Inga Guillen, Janet Gaby; Quispe Melgar, Harold Rusbelth; Galeano, Esteban; Rodríguez Ramírez, Ernesto C
Background. Because of illegal logging, habitat fragmentation, and high value timber Andean montane forest Cedrela species (such as Cedrela angustifolia), is endangered in Central and South America. Studying the effects of climate change and tree cover loss on the distribution of C. angustifolia will help us to understand the climatic and ecological sensitivity of this species and suggest conservation and restoration strategies. Methods. Using ecological niche modeling with two algorithms (maximum entropy (MaxEnt) and Random Forest) under the ecological niche conservatism approach, we generated 16,920 models with different combinations of variables and parameters. We identified suitable areas for C. angustifolia trees under present and future climate scenarios (2040, 2070, and 2100 with SSP 3-7.0 and SSP 5-8.5), tree cover loss, and variables linked to soil and topography. Results. Our results demonstrated 10 environmental variables with high percentage contributions and permutation importance; for example, precipitation seasonality exhibited the highest contribution to the current and future distribution of Cedrela angustifolia. The potential present distribution was estimated as 13,080 km2 with tree cover loss and 16,148.5 km2 without tree cover loss. From 2040 to 2100 the species distribution will decrease (from 22.16% to 36.88% with tree cover loss variation). The results indicated that Bolivia displayed higher habitat suitability than Ecuador, Peru, and Argentina. Finally, we recommend developing conservation management strategies that consider both protected and unprotected areas as well as the impact of land-use changes to improve the persistence of C. angustifolia in the future.
Comprehensive spatial mapping of metals and metalloids in the Peruvian Mantaro Valley using advanced geospatial data Integration
(Elsevier, 2024-12-12) Pizarro Carcausto, Samuel Edwin; Pricope , Narcisa G.; Vera Vilchez, Jesús Emilio; Cruz Luis, Juancarlos Alejandro; Lastra Paucar, Sphyros Roomel; Solórzano Acosta, Richard Andi; Verástegui Martínez, Patricia
The quality and safety of soil are crucial for ensuring social and economic development and providing contaminant-free food. The availability and quality of soil data, particularly for multiple metals and metalloids, are often insufficient for comprehensive analysis. Soil formation and the distribution of metals are shaped by various factors such as geology, climate, topography, and human activities, making accurate modeling highly challenging. Additionally, agricultural intensification, urban expansion, road construction, and mining activities frequently result in soil pollution, posing serious risks to ecosystems and human health. This study aims to integrate diverse geospatial datasets with machine learning for high resolution soil contamination mapping (10 m spatial resolution) in a major agricultural region of Peruvian highlands. This study mapped 25 elements (Ca, Mg, Sr, Ba, Be, K, Na, As, Sb, Se, Tl, Cd, Zn, Al, Pb, Hg, Cr, Ni, Cu, Mo, Ag, Fe, Co, Mn, V) in the Peruvian Mantaro Valley using a training dataset of 109 topsoil samples combined with various geospatial datasets (remote sensing, climate, topography, soil data, and distance). The model provided satisfactory results in predicting the spatial distribution of the selected elements, with R² values ranging from 0.6 to 0.9 for most elements. Edaphic, climate, and topographic covariates were the most significant predictors, particularly for croplands near rivers, whereas spectral variables were less important. The results reveal As, Pb, and Cd concentrations significantly above permissible limits, highlighting urgent health risks. These findings suggest that it is feasible to identify polluted soils and improve regulations based on widely available geospatial datasets with minimal training data. The study contributes to the development of models to assess the impact of pollutants on environmental and human health in the short-to-medium term, emphasizing the need for further research on the translocation of toxic metals into food crops and the implications for public health.
Critical edaphic and altitudinal factors influencing cation exchange capacity in coffee-growing soils of northeastern Peru: implications for sustainable fertility management
(Frontiers Media SA, 2026-05-05) Díaz Chuquizuta, Henry; Manrique Gonzales, Luis Fernando; Sánchez Ojanasta, Martín; Cuevas Giménez, Juan Pablo; Carbajal Llosa, Carlos Miguel; Cuellar Condori, Néstor Edwin; Martínez Zapata, Boris Guillermo; Vallejos Torres, Geomar
Introduction: Effective cation exchange capacity (ECEC) is a key indicator of soil fertility and sustainable soil management assessment in coffee-growing systems. Methods: This study aimed to identify the principal edaphic and altitudinal factors explaining ECEC variability in 69 soil samples collected from coffee farms in northeastern Peru. Results: ECEC results exhibited substantial variation, ranging from 0.14 to 55.49 cmol(+)·kg⁻¹ (mean = 15.21; SD = 12.47), and were significantly correlated with organic matter (r = 0.71), clay content (r = 0.62), exchangeable acidity (r = -0.63), and altitude (r = 0.33). Principal component analysis accounted for 64.3% of the edaphic variability, identifying Ca²⁺, pH, Mg²⁺, and exchangeable acidity as the most influential variables. The Random Forest model demonstrated high predictive accuracy (R² = 0.93; root mean square error (RMSE) = 2.1 cmol(+)·kg⁻¹), outperforming the generalized additive model (GAM) and identifying Ca²⁺ as the most important predictor (IncMSE% = 3177.37). A functional altitudinal gradient was also evident: areas above 1150 m.a.s.l. showed higher acidity and aluminium content, whereas areas below 900 m.a.s.l. exhibited greater base saturation and higher ECEC. Discussion: These findings support the development of site-specific fertilization strategies and soil–climate zoning, emphasizing the value of integrating multivariate analyses with machine-learning models as key tools for optimizing fertility management and coffee crop productivity in tropical mountain ecosystems; where soil texture represents a key factor influencing coffee sustainability, as greater nutrient retention capacity and improved nutritional balance are associated with enhanced potential for sustainable production and reduced environmental impact.
The distribution of cadmium in soil and cacao beans in Peru
(Elsevier, 2023-04-11) Thomas, Evert; Atkinson, Rachel; Zavaleta, Diego; Rodriguez, Carlos; Lastra Paucar, Sphyros Roomel-Luciano; Yovera, Fredy; Arango, Karina; Pezo, Abel; Aguilar Zapata, Javier Neptali; Tames, Miriam; Ramos, Ana; Cruz, Wilbert; Cosme, Roberto; Espinoza, Eduardo; Chavez, Carmen Rosa; Ladd, Brenton
Peru is the eighth largest producer of cacao beans globally, but high cadmium contents are constraining access to international markets which have set upper thresholds for permitted concentrations in chocolate and derivatives. Preliminary data have suggested that high cadmium concentrations in cacao beans are restricted to specific regions in the country, but to date no reliable maps exist of expected cadmium concentrations in soils and cacao beans. Drawing on >2000 representative samples of cacao beans and soils we developed multiple national and regional random forest models to develop predictive maps of cadmium in soil and cacao beans across the area suitable for cacao cultivation. Our model projections show that elevated concentrations of cadmium in cacao soils and beans are largely restricted to the northern parts of the country in the departments of Tumbes, Piura, Amazonas and Loreto, as well as some very localized pockets in the central departments of Huánuco and San Martin. Unsurprisingly, soil cadmium was the by far most important predictor of bean cadmium. Aside from the south-eastern to north-western spatial trend of increasing cadmium values in soils and beans, the most important predictors of both variables in nation-wide models were geology, rainfall seasonality, soil pH and rainfall. At regional level, alluvial deposits and mining operations were also associated with higher cadmium levels in cacao beans. Based on our predictive map of cadmium in cacao beans we estimate that while at a national level <20 % of cacao farming households might be impacted by the cadmium regulations, in the most affected department of Piura this could be as high as 89 %.
From rangelands to cropland, land-use change and its impact on soil organic carbon variables in a Peruvian Andean highlands: a machine learning modeling approach
(Springer, 2024-09-09) Carbajal, Mariella; Ramirez, David A.; Turin Canchaya, Cecilia Claudia; Schaeffer, Sean M.; Konkel, Julie; Ninanya, Johan; Rinza, Javier; De Mendiburu, Felipe; Zorogastua, Percy; Villaorduña, Liliana; Quiroz, Roberto
Andean highland soils contain significant quantities of soil organic carbon (SOC); however, more efforts still need to be made to understand the processes behind the accumulation and persistence of SOC and its fractions. This study modeled SOC variables—SOC, refractory SOC (RSOC), and the 13C isotope composition of SOC (d13CSOC)—using machine learning (ML) algorithms in the Central Andean Highlands of Peru, where grasslands and wetlands (‘‘bofedales’’) dominate the landscape surrounded by Junin National Reserve. A total of 198 soil samples (0.3 m depth) were collected to assess SOC variables. Four ML algorithms—random forest (RF), support vector machine (SVM), artificial neural networks (ANNs), and eXtreme gradient boosting (XGB)—were used to model SOC variablesusing remote sensing data, land-use and landcover (LULC, nine categories), climate topography, and sampled physical–chemical soil variables. RF was the best algorithm for SOC and d13CSOC prediction, whereas ANN was the best to model RSOC. ‘‘Bofedales’’ showed 2–3 times greater SOC (11.2 ± 1.60%) and RSOC (1.10 ± 0.23%) and more depleted d13CSOC (- 27.0 ± 0.44 &) than other LULC, which reflects high C persistent, turnover rates, and plant productivity. This highlights the importance of ‘‘bofedales’’ as SOC reservoirs. LULC and vegetation indices close to the near-infrared bands were the most critical environmental predictors to model C variables SOC and d13CSOC. In contrast, climatic indices were more important environmental predictors for RSOC. This study’s outcomes suggest the potential of ML methods, with a particular emphasis on RF, for mapping SOC and its fractions in the Andean highlands.
Prediction of biomass and nutritional quality of tropical pastures using multispectral analysis and machine learning models
(Elsevier B.V., 2026-05-17) Tafur Culqui, Josué; Atalaya Marin, Nilton; Gómez Fernandez, Darwin; Taboada Mitma, Víctor Hugo; Cruz Luis, Juancarlos Alejandro; Neyra, Henri; Anchayhua Torres, Janella Jelin; Quichua Baldeon, Rosalía; Sánchez Fuentes, Teiser; Olano Camán, Yadhira Milagros; Barrazueta Campos, Mauro Adel; Tineo Flores, Daniel; Goñas Goñas, Malluri
Determining pasture productivity and nutritional value through non-destructive approaches aimed at optimizing forage resource management and improving efficiency in livestock systems has become an urgent priority. In this context, the objective of this study was to evaluate the performance of machine learning models in predicting biomass production and the nutritional contribution of different pasture species, as well as to assess the role of vegetation indices (VIs) in these predictions. To this end, a multispectral sensor mounted on a DJI Matrice 350 RTK platform was used, together with agronomic, yield, and nutritional variables. The curated dataset was subsequently analyzed using linear and polynomial models, as well as tree-based algorithms and support vector machines. Model validation was performed using a group-constrained random partitioning scheme (Group Shuffle Split), with species considered as the grouping variable. Model interpretability was addressed through the SHAP (SHapley Additive Explanations) framework. The results indicated better predictive performance for yield-related variables compared to nutritional attributes. In particular, the Extra Trees model achieved the highest coefficients of determination (R²). SHAP analysis revealed that the Visible Atmospherically Resistant Index (VARI) contributed more strongly to yield-related predictions, whereas the Normalized Difference Red Edge (NDRE) showed a more consistent contribution to nutritional variables. In conclusion, these findings highlight the potential of integrating vegetation indices and machine learning models as effective tools for forage management, supporting informed decision-making in livestock production systems.
Varietal Identification and Yield Estimation in Potatoes Using UAV RGB Imagery in the Southern Highlands of Peru
(MDPI, 2026-02-12) Tueros Munive, Miguel Luis; Galindo Sánchez, Malú Massiel; Alvarez Martínez, Jean; Pozo Huacha, Jesús; Condezo Márquez, Patricia Kelly; Gutierrez Ruti, Rusbel; Bautista Gómez, Rolando; Mateu Mateo, Walter Rolando; Paitamala Campos, Omar; Matsusaka Quiliano, Daniel Claudio
The cultivation of potatoes is essential for rural food security, and the use of Unmanned Aerial Vehicle Red-Green-Blue (UAV-RGB) imagery allows for precise and cost-effective estimation of yield and identification of varieties, overcoming the limitations of manual assessment. We evaluated four INIA varieties (Bicentenario, Canchán, Shulay and Tahuaqueña) by integrating agronomic measurements (height, number and weight of tubers, leaf health) with color and textural indices derived from RGB orthomosaics. Yield prediction was modeled using Random Forest (RF) and Gradient Boosting (GB); varietal identification was approached with (i) a Convolutional Neural Network (CNN) that classifies RGB images and (ii) classical models such as Random Forest, Support Vector Machines (SVMs), K-Nearest Neighbors (KNNs), Decision Trees and Logistic Regression trained on EfficientNetB0 embeddings. The results showed significant genotypic differences in yield (p < 0.001): Tahuaqueña 13.86 ± 0.27 t ha⁻¹ and Bicentenario 6.65 ± 0.27 t ha⁻¹. The number of tubers (r = 0.52) and plant height (r = 0.23) correlated with yield; RGB indices showed low correlations (r < 0.3) and high redundancy (r > 0.9). RF achieved a better fit (Coefficient of determination, R² = 0.54; Root Mean Square Error, RMSE = 2.72 t ha⁻¹), excelling in stolon development (R² = 0.66) and losing precision in maturation due to foliar senescence. In classification, the CNN and RF on embeddings achieved F1-macro ≈ 0.69 and 0.66 (Receiver Operating Characteristic—Area Under the Curve, ROC AUC RF = 0.89), with better identification of Bicentenario and Shulay. We conclude that UAV-RGB is a cost-effective alternative for phenotypic monitoring and varietal selection in high Andean contexts. These findings support the integration of UAV-RGB imagery into breeding and monitoring pipelines in resource-limited Andean systems.