Big data, small explanatory and predictive power: Lessons from random forest modeling of on-farm yield variability and implications for data-driven agronomy

Martin van Ittersum (2023, [Artículo])

Context: Collection and analysis of large volumes of on-farm production data are widely seen as key to understanding yield variability among farmers and improving resource-use efficiency. Objective: The aim of this study was to assess the performance of statistical and machine learning methods to explain and predict crop yield across thousands of farmers’ fields in contrasting farming systems worldwide. Methods: A large database of 10,940 field-year combinations from three countries in different stages of agricultural intensification was analyzed. Random effects models were used to partition crop yield variability and random forest models were used to explain and predict crop yield within a cross-validation scheme with data re-sampling over space and time. Results: Yield variability in relative terms was smallest for wheat and barley in the Netherlands and for wheat in Ethiopia, intermediate for rice in the Philippines, and greatest for maize in Ethiopia. Random forest models comprising a total of 87 variables explained a maximum of 65 % of cereal yield variability in the Netherlands and less than 45 % of cereal yield variability in Ethiopia and in the Philippines. Crop management related variables were important to explain and predict cereal yields in Ethiopia, while predictive (i.e., known before the growing season) climatic variables and explanatory (i.e., known during or after the growing season) climatic variables were most important to explain and predict cereal yield variability in the Philippines and in the Netherlands, respectively. Finally, model cross-validation for regions or years not seen during model training reduced the R2 considerably for most crop x country combinations, while for wheat in the Netherlands this was model dependent. Conclusion: Big data from farmers’ fields is useful to explain on-farm yield variability to some extent, but not to predict it across time and space. Significance: The results call for moderate expectations towards big data and machine learning in agronomic studies, particularly for smallholder farms in the tropics where model performance was poorest independently of the variables considered and the cross-validation scheme used.

Model Accuracy Model Precision Linear Mixed Models CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA MACHINE LEARNING SUSTAINABLE INTENSIFICATION BIG DATA YIELDS MODELS AGRONOMY

Nepal Seed And Fertilizer Project

Dyutiman Choudhary (2022, [Objeto de congreso])

CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA SEED FERTILIZERS SEED INDUSTRY PRIVATE SECTOR MAIZE RICE INTEGRATED SOIL FERTILITY MANAGEMENT COVID-19

Nepal Seed And Fertilizer Project

Dyutiman Choudhary (2021, [Objeto de congreso])

CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA SEED SEED INDUSTRY PRIVATE SECTOR MAIZE RICE INTEGRATED SOIL FERTILITY MANAGEMENT COVID-19

DE INTERFACES Y DISEÑADORES: UNA ETNOGRAFÍA DIGITAL SOBRE EL DISEÑO UX/UI

Santiago López Martínez (2023, [Tesis de maestría])

“La primera ley de Kranzberg es que “la tecnología no es buena ni mala, pero tampoco es neutra” (1986: 545). Esta ley ha sido muy estimulante para muchos teóricos de los medios, a partir de dos interpretaciones principales: la primera sería que el valor moral de la tecnología depende de su uso, y por lo tanto puede ser utilizada para bien o para mal. La segunda, y más interesante para esta investigación, es que la tecnología en su construcción posee características humanas y por lo tanto sociales, en procesos técnicos y supuestamente racionales. Esta segunda premisa es explorada y defendida desde trabajos clásicos como Ciencia y técnica como “ideología” de Jürgen Habermas, donde escribe que “lo mismo antes que ahora son los intereses sociales los que determinan la dirección, las funciones y la velocidad del proceso técnico” (2007: 87), y posteriormente argumenta que la ciencia y la tecnología, en su estrecha relación con el poder y la economía pueden convertirse en ideologías que legitiman y perpetúan sistemas de dominación. Sostiene que el enfoque instrumental de la ciencia, es decir, su utilización como un medio para alcanzar fines predeterminados, puede llevar a una heteronomía del mundo de la vida cotidiana, en el cual las decisiones y la autonomía de los individuos se ven disminuidas frente a procesos tecnificados. Pero también existen estudios recientes de tecnologías digitales que señalan la presencia de ideologías, como el libro Armas de destrucción matemática (2017) escrito por la matemática Cathy O’Neil, en donde presenta varios casos en los que algoritmos discriminan a ciertos grupos humanos; por ejemplo, un programa de calificación crediticia el cual “afecta de forma desproporcionada a los candidatos de renta baja y a los candidatos de color” (O’Neil, 2017: 120)”.

Etnografía digital. Diseño UX. Diseño UI. Programación creativa. Tesis - Maestría en Antropología Social, CDMX. CIENCIAS SOCIALES SOCIOLOGÍA CAMBIO Y DESARROLLO SOCIAL TECNOLOGÍA Y CAMBIO SOCIAL TECNOLOGÍA Y CAMBIO SOCIAL

Research for development approaches in mixed crop-livestock systems of the Ethiopian highlands

Million Gebreyes James Hammond Lulseged Tamene Getachew Agegnehu Rabe Yahaya Anthony Whitbread (2023, [Artículo])

This study presents processes and success stories that emerged from Africa RISING's Research for Development project in the Ethiopian Highlands. The project has tested a combination of participatory tools at multiple levels, with systems thinking and concern for sustainable and diversified livelihoods. Bottom-up approaches guided the selection of technological interventions that could address the priority farming system challenges of the communities, leading to higher uptake levels and increased impact. Joint learning, appropriate technology selection, and the creation of an enabling environment such as the formation of farmer research groups, the establishment of innovation platforms, and capacity development for institutional and technical innovations were key to this study. The study concludes by identifying key lessons that focus more on matching innovations to community needs and geographies, systems orientation/integration of innovations, stepwise approaches to enhance the adoption of innovations, documenting farmers' capacity to modify innovations, building successful partnerships, and facilitating wider scaling of innovations for future implementation of agricultural research for development projects.

Action Research Systems Thinking CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA INNOVATION PARTNERSHIPS SCALING UP INTEGRATED CROP-LIVESTOCK SYSTEMS

Complex traits and candidate genes: estimation of genetic variance components across multiple genetic architectures

GIOVANNY COVARRUBIAS-PAZARAN Hans-Peter Piepho (2023, [Artículo])

Average Semivariance Linear Mixed Model Variance Component Estimation Polygenic Inheritance Oligogenic Inheritance Mendelian Inheritance CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA MENDELISM GENETIC VARIANCE GENOME-WIDE ASSOCIATION STUDIES PHENOTYPES CHROMOSOME MAPPING

Weed management and tillage effect on rainfed maize production in three agro-ecologies in Mexico

Simon Fonteyne Abel Jaime Leal González Rausel Ovando Ravi Gopal Singh Nele Verhulst (2022, [Artículo])

Maize (Zea mays L.) is grown in a wide range of agro-ecological environments and production systems across Mexico. Weeds are a major constraint on maize grain yield, but knowledge regarding the best weed management methods is lacking. In many production systems, reducing tillage could lessen land degradation and production costs, but changes in tillage might require changes in weed management. This study evaluated weed dynamics and rainfed maize yield under five weed management treatments (pre-emergence herbicide, post-emergence herbicide, pre-emergence + post-emergence herbicide, manual weed control, and no control) and three tillage methods (conventional, minimum and zero tillage) in three agro-ecologically distinct regions of the state of Oaxaca, Mexico, in 2016 and 2017. In the temperate Mixteca region, weeds reduced maize grain yields by as much as 92% and the long-growing season required post-emergence weed control, which gave significantly higher yields. In the hot, humid Papaloapan region, weeds reduced maize yields up to 63% and pre-emergence weed control resulted in significantly higher yields than treatments with post-emergence control only. In the semi-arid Valles Centrales region, weeds reduced maize yields by as much as 65%, but weed management was not always effective in increasing maize yield or net profitability. The most effective weed management treatments tended to be similar for the three tillage systems at each site, although weed pressure and the potential yield reduction by weeds tended to be higher under zero tillage than minimum or conventional tillage. No single best option for weed management was found across sites or tillage systems. More research, in which non-chemical methods should not be overlooked, is thus needed to determine the most effective weed management methods for the diverse maize production systems across Mexico.

Corn Integrated Weed Management Manual Weed Control CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA MAIZE WEED CONTROL MINIMUM TILLAGE ZERO TILLAGE

Unsupervised segmentation and clustering time series approach to Southern Africa rainfall regime changes

Lovemore  Chipindu Walter Mupangwa Isaiah Nyagumbo Mainassara Zaman-Allah (2023, [Artículo])

Autoregressive Integrated Moving Average Facebook Prophet Hidden Markov Model Regression Regression with Hidden Logistic Process CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA COASTAL AREAS SEMIARID ZONES SUBHUMID ZONES RAINFALL CLIMATE CHANGE

Using homosoils for quantitative extrapolation of soil mapping models

Andree Nenkam Alexandre Wadoux Budiman Minasny Alex McBratney Pierre C. Sibiry Traore Gatien Falconnier Anthony Whitbread (2022, [Artículo])

Cubist Digital Soil Mapping Model-Based Validation Soil Spatial Variation Soil-Forming Factors CIENCIAS AGROPECUARIAS Y BIOTECNOLOGÍA LAND USE ORGANIC CARBON SOIL SURVEYS SPATIAL VARIATIONS

GENERACIÓN DE HACES ESTRUCTURADOS PARCIALMENTE COHERENTES

Maria Fernanda Arvizu Soto (2024, [Tesis de maestría])

"Mediante la manipulación de la intensidad, polarización y fase de un campo luminoso es como podemos obtener luz estructurada. Los estudios han demostrado que los haces parcialmente coherentes son más resistentes a las fluctuaciones atmosféricas. En esta tesis presentamos el modelo matemático, la generación computacional del Haz Vectorial Parcialmente Coherente (PCVB) y dos configuraciones experimentales para generarlos con la ayuda de un Dispositivo Micro Espejo Digital (DMD). Los resultados teóricos esperados concuerdan con las simulaciones y allanan el camino para otros resultados teóricos y experimentales, que se dejan como trabajo futuro."

Luz estructurada Haces vectoriales Coherencia parcial Dispositivo digital de microespejos CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA FÍSICA ÓPTICA OPTICA FÍSICA OPTICA FÍSICA