Sulfur (S) deficiency has been recently reported in soybean [Glycine max (L.) Merr.] producing regions across the United States. However, field studies have often failed to demonstrate a strong relationship between yield and S fertilization and generally attributing the lack of yield response to unfavorable weather and high soil S supply. In addition, only a few reports described seed composition changes due to S availability under contrasting field conditions. Therefore, our goals were (i) to implement a meta-analytic model to quantify the effect of S application at different growth stages on yield and seed concentration of protein, oil, essential non-S amino acids, and S amino acids (SAA, cysteine and methionine); ii) identify environmental factors underpinning the response of S to these plant traits. Field experiments were carried out from 2017 to 2019 growing seasons with a total of 44 unique site-years conditions across 18 locations in 8 states. Mineral S fertilizer (sulfate/ elemental S) was supplied depending on the study at sowing, vegetative and/or reproductive stages. A random-effects multilevel meta-analysis was conducted. The effect sizes compared yield and seed composition responses relative to the unfertilized control. A principal component analysis (PCA) separated distinctive environmental conditions and a sub-grouped meta-analysis with the main environmental factors was later executed to understand the response of the plant traits with those factors. Seed protein concentration increased by 0.3 % when S was applied at sowing. The concentration of SAA increased by ca. 1% regardless of the fertilization timing. Sites exposed to drought stress (18–29% reduction of potential transpiration) neither presented changes in yield nor seed composition due to S fertilization. Soils with organic matter between 25 and 32 g kg^-1 (medium cluster) displayed significant responses to S application. This research brings extensive data and provides a comprehensive analysis of weather and soil attributes influencing soybean yield and seed composition responses to S availability.

大豆中最近发现了硫缺乏症[ Glycine max（L.）Merr。]遍布美国的产区。然而，田间研究往往未能证明产量与硫素施肥之间有很强的联系，并且通常归因于不利的天气和土壤高硫素供应对产量缺乏响应。此外，只有少数报告描述了在相反田间条件下由于硫的有效性而导致的种子组成变化。因此，我们的目标是（i）建立一个荟萃分析模型，以量化不同生育期施用S对蛋白质，油，必需非S氨基酸和S氨基酸（SAA，半胱氨酸和蛋氨酸）；ii）确定支撑S对这些植物性状响应的环境因素。在2017年至2019年的生长季节进行了野外试验，在8个州的18个地点共进行了44种独特的站点年条件。根据播种，营养和/或生殖阶段的研究，提供了矿物S肥料（硫酸盐/元素S）。进行了随机效应多级荟萃分析。效应大小比较了相对于未受精对照的产量和种子组成响应。主成分分析（PCA）分离了不同的环境条件，并随后进行了具有主要环境因素的分组荟萃分析，以了解这些因素对植物性状的响应。播种时施S可使种子蛋白浓度增加0.3％。SAA的浓度增加了约。不论施肥时间长短为1％。暴露于干旱胁迫（潜在蒸腾作用减少了18-29％）的站点，既未因硫磷施肥而未导致产量或种子组成发生变化。土壤中的有机物含量在25到32克/千克之间^-1（中等簇）显示出对S应用程序的显着响应。这项研究带来了广泛的数据，并提供了天气和土壤属性的综合分析，这些天气和土壤属性影响大豆产量和种子组成对硫素有效性的响应。