Abstract Closing the gap between yields currently achieved on farms and those that can potentially be achieved with best practice and current technology (the yield gap) is a key strategy to intensify grain production without expanding cropland. Much research has been done to quantify the yield gap of wheat, maize and rice globally and of wheat, barley, canola, sorghum and pulse crops in Australia. However, crops are grown in rotations (recurring crop sequences) that vary in their cropping intensities and in the diversity of their species. Little is known about yield gaps at the cropping system level, especially in regions where there are many possible combinations of crop types and fallow periods. This prompted us to investigate crop rotations in Australia's subtropical grains region where current crop rotations include winter and summer cropping with cereal, pulse and oilseed crops interspersed with fallow periods ranging from nil to 18 months duration. To determine the system level yield gaps, we simulated the water-limited yield potential of 26 locally practiced crop rotations for over 800 weather stations by up to 3 soil types per station. We captured the impact of climate variability with 30–35 years by 2–7 fields per rotation for each site. We expressed the results in terms of energy, protein and revenue per hectare per year and mapped the results of the optimal rotations over the cropping zone. Surprisingly, a single rotation (sorghum/fallow/mungbean/wheat/fallow/chickpea rotation; with 4 crops in 3 years, balanced between summer and winter crops and between cereal and pulse crops) was optimal for revenue over almost the whole subtropical grain zone. Using revenue as the metric for yield gaps at statistical local area scale we found, over the whole subtropical zone, a mean revenue gap of 970 $/ha/yr. This represents a relative revenue (Revenue% = 100 x (actual revenue/water-limited revenue)) of 34% which is much lower than expected from the 40–60% relative yields achieved by individual crops. We investigated whether growers may select rotations that have lower revenue than the optimal rotation in response to economic factors such as profit and risk. We found that for much of the area the same rotation that optimised revenue also optimised profit. However, for some of the cropping zone, particularly in the south western portion, a different, less intensive and more winter dominant, rotation was most profitable. Similarly, risk averse farmers may choose less productive and profitable rotations with less risk.

中文翻译：

---

澳大利亚亚热带旱地的轮作可以通过更优化的轮作来缩小作物系统产量差距

摘要 缩小目前农场产量与最佳实践和当前技术可能实现的产量之间的差距（产量差距）是在不扩大农田的情况下加强粮食生产的关键策略。已经进行了大量研究来量化全球小麦、玉米和水稻以及澳大利亚小麦、大麦、油菜、高粱和豆类作物的产量差距。然而，作物轮作（重复作物序列）种植，其种植强度和物种多样性各不相同。人们对种植系统层面的产量差距知之甚少，尤其是在作物类型和休耕期有多种可能组合的地区。这促使我们调查澳大利亚的作物轮作 s 亚热带谷物地区，目前作物轮作包括冬季和夏季作物，谷物、豆类和油籽作物散布着从零到 18 个月的休耕期。为了确定系统水平的产量差距，我们模拟了 800 多个气象站的 26 种当地实施的作物轮作的限水产量潜力，每个气象站最多 3 种土壤类型。我们通过每个站点每轮转 2-7 个田地捕获了 30-35 年气候变化的影响。我们以每年每公顷的能量、蛋白质和收入来表示结果，并绘制了种植区最佳轮作的结果。令人惊讶的是，单轮轮作（高粱/休耕/绿豆/小麦/休耕/鹰嘴豆轮作；3 年 4 次作物，夏季和冬季作物之间以及谷物和豆类作物之间的平衡）对几乎整个亚热带粮食区的收入而言是最佳的。使用收入作为统计局部区域范围内产量差距的指标，我们发现在整个亚热带地区，平均收入差距为 970 美元/公顷/年。这代表了 34% 的相对收入（收入% = 100 x（实际收入/水限制收入）），这远低于单个作物实现的 40-60% 相对产量的预期。我们调查了种植者是否会根据利润和风险等经济因素选择收益低于最佳轮换的轮换。我们发现，在大部分地区，优化收入的轮换也优化了利润。然而，对于一些种植区，特别是在西南部分，不同的，强度较低且冬季占主导地位，轮换是最有利可图的。同样，规避风险的农民可能会选择风险较低、生产率较低、利润较低的轮作。