Yield gaps, which reflect the difference between the yield a grower achieves in a field, and the water-limited yield potential of that field, have been extensively discussed in the last decade. To date, most analyses have occurred at regional levels, and derived insight through surveys, remote sensing, or small scale targeted case studies. Here, we provide an analysis based on a survey of 250 fields from 2015 to 2018 inclusive. Biotic factors, nutrients and crop yields are all monitored. Crop models were calibrated locally at each site. Regression and Classification Trees (RCART) were used to determine the key drivers of the yield gap in wheat, and again used to identify attributes contributing to a yield gap for a field regardless of crop type. The mean yield gap from 697 wheat crops, collected over 4 years, was 1.03 t/ha (MSE =2.23 t/ha). This equated to a mean wheat yield gap of 20.3 % ± S.D 36.2 %. 18 % of fields had repeated yield gaps of 20 % or more, irrespective of crop type. The RCART analysis demonstrated that for wheat, the yield potential of the crop was the most important predictor, where fields with high yield potentials were most likely to have a higher yield gap. Nitrogen was the second most important predictor. Yield gaps of wheat crops grown in high yielding regions were also related to crop rotation, leaf diseases and weed populations. In low rainfall zones, wheat yield gaps were related to the presence of soil pathogens including Pratylenchus sp., Pythium sp. and Fusarium sp. In conclusion, yield potential and nitrogen are important predictors of the yield gap. Increasing nitrogen inputs would address the yield gap in 25 % of wheat crops. In 22 % of wheat crops, increasing nitrogen inputs will not correct the gap, as other biotic stresses are often present, that require sophisticated agronomic intervention. In the remaining 53 % of wheat fields yield gaps were less than 0.37 t/ha.

在过去的十年中，已经广泛讨论了产量差距，反映了一个田间种植者获得的产量与该田的水分受限产量潜力之间的差异。迄今为止，大多数分析都发生在区域级别，并且通过调查，遥感或小规模目标案例研究得出了见解。在此，我们基于2015年至2018年（含）的250个领域的调查结果进行了分析。生物因子，养分和农作物产量都受到监测。作物模型在每个站点都进行了本地校准。回归树和分类树（RCART）用于确定小麦单产缺口的关键驱动因素，并再次用于识别造成田间单产差异的属性，而与作物类型无关。四年间收集的697种小麦作物的平均单产差距为1.03吨/公顷（MSE = 2.23吨/公顷）。这相当于小麦平均单产差为20.3％±SD 36.2％。不论作物类型如何，有18％的田地反复出现20％或更高的产量差距。RCART分析表明，对于小麦而言，农作物的单产潜力是最重要的预测指标，在这种情况下，单产潜力高的田地最有可能出现更高的单产差距。氮是第二重要的预测因子。高产地区小麦作物的产量缺口还与轮作，叶病和杂草种群有关。在低降雨地区，小麦产量缺口与土壤病原体的存在有关，包括 作物的单产潜力是最重要的预测指标，高单产潜力的田地最有可能出现更高的产量缺口。氮是第二重要的预测因子。高产地区小麦作物的产量缺口还与轮作，叶病和杂草种群有关。在低降雨地区，小麦产量缺口与土壤病原体的存在有关，包括 作物的单产潜力是最重要的预测指标，高单产潜力的田地最有可能出现更高的产量缺口。氮是第二重要的预测因子。高产地区小麦作物的产量缺口还与轮作，叶病和杂草种群有关。在低降雨地区，小麦产量缺口与土壤病原体的存在有关，包括鼠疫菌属，腐霉菌属 和镰刀菌。总之，单产潜力和氮素是单产差距的重要预测指标。增加的氮投入将解决25％的小麦作物的单产缺口。在22％的小麦作物中，增加的氮输入量无法纠正这一差距，因为其他生物胁迫经常存在，需要进行复杂的农业干预。在剩下的53％的麦田中，产量差距小于0.37吨/公顷。