Human development must ensure the welfare and survival of an ever-growing population. To assess the impact of agroecosystems on sustainability in Nariño (Colombia), variables related to sustainable development were evaluated through a semi-structured survey conducted with 122 producers from 6 municipalities in the Nariño department. The association between the categorical variables associated with the pillars of sustainability was analyzed using multiple correspondence analysis (MCA). From the MCA, 23 indicators were selected, representing quantitative information on the economic, social, and environmental characteristics of the participants. The municipalities were characterized according to the needs and strengths of the producers. Five groups encompass most characteristics of the producers, which are largely independent of geopolitical conditions. On the other hand, the population characterization indicates that the economic well-being of fruit growers can promote the conservation of natural resources. Finally, a principal component factor analysis (PCFA) was conducted to construct the General Sustainability Index (GSI). A network graph for each municipality revealed that the indicators of the production systems are in an unfavorable environment in almost all aspects related to sustainability. The GSI values reveal unstable agroecological conditions that are causing environmental deterioration, with critical levels in the municipalities of Arboleda, Sandoná, and Providencia. Social investment and public policies must be implemented to improve the economic conditions of the producers, accompanied by legislation on the proper management of natural resources.

Companion cropping offers a potential solution to the challenges of sustainable agriculture, such as optimizing resource use and reducing reliance on chemical inputs. The problem of achieving higher yields while maintaining environmental health remains critical. This practice enhances natural resource conservation, improves fertilization, and optimizes nutrient cycling through the balanced use of chemical and organic sources. Studies, such as those involving tree tomato and Hass avocado, have demonstrated a significant yield increase compared to monocultures, underscoring the viability of this practice. In addition to their environmental benefits, companion crops provide economic advantages by allowing producers to harvest multiple products simultaneously, thereby strengthening food security and the rural economy. This study evaluated three levels of fertilization and interactions between fruit trees at different altitudes, observing differential behavior in the variables evaluated. The combination of cape gooseberry and blackberry showed significantly positive results, with more leaves and fewer pests, demonstrating the benefits of companion plants. A trend towards the combined use of chemical and organic fertilizers was observed, a potential strategy to reduce costs and improve crop growth. The results indicated that the UF system (P. peruviana and P. vulgaris) had the highest plant height, while TF (tree tomato and bean) showed the best stem perimeter development. The incidence of pests was also significant, with Trialeurodes vaporarioum being most prevalent in the P. peruviana companion. These findings support companion cropping as a viable and promising strategy for more efficient and sustainable agriculture, offering both environmental and economic benefits.

Rhizoctonia solani es un hongo que afecta la producción de tubérculos y vigor de las plantas de papa Solanum tuberosum L. Las consecuencias ambientales que genera su control han impulsado la búsqueda de alternativas menos contaminantes, en donde los aceites esenciales tienen gran potencial. El arbusto silvestre Lippia origanoides H.B.K. (Verbenaceae) ha sido estudiado por sus propiedades fungicidas en su extracto esencial, por ello, el objetivo de esta investigación fue evaluar la actividad biocida del aceite esencial de L. origanoides sobre R. solani en condiciones in vitro. La extracción del aceite se realizó mediante arrastre con vapor. La composición química se determinó por cromatografía de gases-espectrometría de masas (GC-MS) y la actividad biocida del aceite se evaluó mediante la técnica de dilución en medio de cultivo, complementando los resultados con la técnica de bioautografía, la cual permitió identificar al timol como el compuesto mayoritario del aceite (60,7%) y responsable de la inhibición del hongo. La concentración 110 µL mL-1 limitó el crecimiento de R. solani al afectar sus estructuras, con ello se demostró el efecto biocida del aceite esencial y el potencial agente antifúngico natural.

Esta investigación estuvo orientada a estimar el estado de la diversidad específica, a través de los índices de riqueza, diversidad y equitabilidad en los sistemas productivos de café, en cuatro municipios que se encuentran ubicados en la subregión del Río Mayo, así como en la del Centro y en la del Occidente en el departamento de Nariño (Colombia). La obtención de la información se efectuó mediante visitas de campo, durante el primer semestre del año 2019, en 100 sitios de muestreo en cada municipio; en cada uno de los predios se identificaron las especies arbóreas y se realizó su respectiva clasificación taxonómica. Se encontraron 76 especies arbóreas dentro de los 400 sistemas productivos de café identificados; de estas, el 35,5% estaban presentes en los cuatro municipios, y eran pertenecientes a 37 familias, siendo la familia Fabaceae la más representativa y la naranja (Citrus x aurantium) la especie de mayor frecuencia, la cual fue encontrada en 299 fincas del total de las evaluadas. El municipio con el mayor número de especies y valor de índices de riqueza fue el de La Unión, con presencia de 60 especies diferentes, mientras que los municipios con los menores índices fueron los de La Florida y Sandoná. De igual forma, el municipio de La Unión obtuvo los mayores valores de diversidad, mientras que el de La Florida ocupó el último lugar. En cuanto a la estimación de equitabilidad, el municipio de Sandoná obtuvo la mayor calificación, seguido de los de La Unión, Consacá y La Florida. A su vez, los índices muestran que los sistemas productivos de café estudiados son diversos, asunto que contribuye a la mitigación del cambio climático.

The LeafArea package is an innovative tool for estimating leaf area in six Andean fruit species, utilizing leaf length and width along with species type for accurate predictions. This research highlights the package’s integration of advanced machine learning algorithms, including GLM, GLMM, Random Forest, and XGBoost, which excels in predictive accuracy. XGBoost’s superior performance is evident in its low prediction errors and high R² value, showcasing the effectiveness of machine learning in leaf area estimation. The LeafArea package, thus, offers significant contributions to the study of plant growth dynamics, providing researchers with a robust and precise tool for informed decision making in resource allocation and crop management.

Leaf area estimation is a critical component in the study of plant growth and productivity within agricultural systems. This research introduces the LeafArea package, a specialized tool designed to calculate the leaf area of six distinct Andean fruit species: S. quitoense, S. betaceum, P. peruviana, R. fruticosus, P. ligularis and P. edulis. Leveraging response variables such as species type, leaf length and width, the package employs advanced machine learning algorithms to estimate leaf area accurately. The primary focus of the study is to identify the most effective model for describing the relationship between leaf width, length, and area for each plant species. Currently, the LeafArea package utilizes four different machine learning algorithms, namely generalized linear model (GLM), generalized linear mixed model (GLMM), Random Forest and XGBoost. Among these, XGBoost stands out as a top-performing algorithm, exhibiting exceptional predictive accuracy. The evaluation metrics employed in the program provide valuable insights for researchers, aiding in informed decision-making. Specifically, XGBoost demonstrates significantly lower prediction errors and approaches a near-perfect R2 value, emphasizing its potential to enhance predictive accuracy. These results underscore the efficacy of machine learning techniques, as a compelling choice for researchers seeking precise and robust predictions in leaf area estimation. The LeafArea package thus represents a valuable tool for advancing our understanding of plant growth dynamics, resource allocation, and overall productivity within agricultural ecosystems.