Going Beyond the Plot: Accelerating Innovation in Agriculture.

Development and rapid adoption of technologies are the most promising avenues to increase food production in a sustainable manner. However, applying agricultural information to broader spatial scales requires agronomy to transform the classical plot-level approach to an integration of big data volumes on a broad scale. This crucial change requires a robust spatial framework that enables upscaling of results from specific locations to larger areas. Here, we present a technology extrapolation (TED) domain framework for technology transfer in agriculture based on a combination of weather and soil information. Validation of the spatial framework indicated that the inclusion of climate and soil properties in the TED framework allowed aggregation of regions with similar biophysics features, and only three biophysical factors embedded into the TED framework (annual total growing degree-days, aridity index, and plant-available water holding capacity) accounted for 56% and 37% of the variation in average yield and CV among regions. An application example of the TED framework demonstrated its capacity for extrapolating a more intensified cropping system technology (2-yr pulse-wheat-pulse rotation) from Argentina to Australia, representing an income increase of 75% relative to the current wheat-fallow system. The TED framework was also used for improving site selection efficiency of a real 96-site trial network located within 19 TEDs, which, in turn, accounted for 42% of total US maize area. A better selection of the testing locations increased maize area coverage (from 42 to 96%) with the same number of sites. The TED concept presented here addressed a fundamental, missing aspect of agricultural research, that is, to know where to test a technology, how to extrapolate it, and how to predict responses. Without a robust spatial framework, our capacity to evaluate technologies and ensure local impact will remain fragile.