Modeling climate-based variability in the fatty acid composition of soybean oil in the American diet from public data repositories

Abstract: The variation in fatty acid composition for soybean oil based upon increased temperature was modeled using publicly available data sources. Through a comprehensive literature review 16 papers containing 233 data points were identified. Data with day-time high temperature means from 15 °C to 40 °C and fatty acid composition data for oleic, linoleic, and α-linolenic acid were used to build models for fatty acid composition in response to temperature. The models predict that fatty acid composition will vary non-linearly in response to increased temperature, with oleic acid expected to increase from 15 to 45 g/ 100 g oil, while linoleic and α-linolenic acid decrease from 55 to 30 g/ 100 g oil and 55 to 30 g/ 100 g oil, respectively over the same temperature range. When applied to contemporary meteorological data sets for counties reporting soybean production to the National Agricultural Statistics Service from 2016-2021 these models predict significant intra-year variation in fatty acid composition of soybeans entering R5 developmental stage in early-July when compared mid-August. Additionally, the models anticipate significant geographical and inter-year variation for fatty acid composition. Methods for application and validation of this model are discussed.