Remote sensing is essential in precision agriculture as this approach provides high-resolution information on the soil’s physical and chemical parameters for detailed decision making. Globally, technologies such as remote sensing and machine learning are increasingly being used to infer these parameters. This study evaluates soil fertility changes and compares them with previous fertilization inputs using high-resolution multispectral imagery and in situ measurements. A UAV-captured image was used to predict the spatial distribution of soil parameters, generating fourteen spectral indices and a digital surface model (DSM) from 103 soil plots across 49.83 hectares. Machine learning algorithms, including classification and regression trees (CART) and random forest (RF), modeled the soil parameters (N-ppm, P-ppm, K-ppm, OM%, and EC-mS/m). The RF model outperformed others, with R² values of 72% for N, 83% for P, 87% for K, 85% for OM, and 70% for EC in 2023. Significant spatiotemporal variations were observed between 2022 and 2023, including an increase in P (14.87 ppm) and a reduction in EC (−0.954 mS/m). High-resolution UAV imagery combined with machine learning proved highly effective for monitoring soil fertility. This approach, tailored to the Peruvian Andes, integrates spectral indices and field-collected data, offering innovative tools to optimize fertilization practices, address soil management challenges, and merge modern technology with traditional methods for sustainable agricultural practices.

Precision agriculture aims to improve crop management using advanced analytical tools. In this context, the objective of this study is to develop an innovative predictive model to estimate the yield and morphological quality, such as the circularity and length–width ratio of potato tubers, based on phenotypic characteristics of plants and data captured through spectral cameras equipped on UAVs. For this purpose, the experiment was carried out at the Santa Ana Experimental Station in the central Peruvian Andes, where advanced potato clones were planted in December 2023 under three levels of fertilization. Random Forest, XGBoost, and Support Vector Machine models were used to predict yield and quality parameters, such as circularity and the length–width ratio. The results showed that Random Forest and XGBoost achieved high accuracy in yield prediction (R² > 0.74). In contrast, the prediction of morphological quality was less accurate, with Random Forest standing out as the most reliable model (R² = 0.55 for circularity). Spectral data significantly improved the predictive capacity compared to agronomic data alone. We conclude that integrating spectral indices and multitemporal data into predictive models improved the accuracy in estimating yield and certain morphological traits, offering key opportunities to optimize agricultural management.

The growing global loss of genetic diversity, phenotypic characterization becomes essential for identifying resilient varieties capable of diversifying and strengthening the agricultural production of underutilized crops such as tarwi (Lupinus mutabilis S.). This study aimed to characterize the phenotypic variability of 41 tarwi accessions conserved in the germplasm bank of the National Institute of Agricultural Innovation (INIA) of Peru. The accessions were evaluated over two consecutive agricultural seasons at the Santa Ana Agrarian Experimental Station under local conditions. Thirty morphological descriptors (17 qualitative and 13 quantitative) were used following IBPGR guidelines. Data were analyzed using descriptive statistics, principal component analysis, hierarchical clustering and correlation analysis for quantitative descriptors, as well as frequency tables and the Shannon-Weaver diversity index for qualitative descriptors. The results revealed high phenotypic variability, particularly in traits related to yield, plant architecture and floral attributes. The accessions were grouped into three morpho-agronomic types: (1) highly productive accessions, (2) accessions with vigorous vegetative development, and (3) short-cycle plants with moderate yields. Yield per plant was significantly associated with the total pod number, total seed mass in hundred seeds and seed thickness. The study revealed considerable phenotypic diversity, characterized by significant correlations among key agronomic traits, the delineation of three distinct phenotypic clusters, and the identification of valuable qualitative attributes, which reinforces their potential for conservation and breeding programmes. However, expanded germplasm evaluation and multi-environment trials are required to validate genotype stability and refine selection criteria. However, additional accessions and further analyses are needed to validate the observed patterns.