Oat domestication - understanding the origin of a European cereal

燕麦可以说是唯一的“欧洲”谷物。先前的研究表明，在土耳其小麦和大麦最初被驯化的地方附近发现了它最近的野生亲戚，但考古证据表明，在农业发展的后期，它才首次被用作一种作物，然后在中欧，而不是在它的原始环境中。对这种转移的一个合理的解释是，一种野生的近亲（Avena sterilis）在破碎的土地上定居，已经适应了作为农业杂草的生活。这种新杂草（A. fatua）在野外传播的能力较弱，但在大麦等谷类作物的耕地上非常成功。a . fatua可以随着这些作物传播到整个欧洲，然后在其他作物失败的困难时期成为一种自己的作物。事实上，燕麦传统上是在贫瘠的土地上种植的，在土壤肥力较低的轮作结束时种植，这可能是这种二次驯化的遗产。在某些阶段，完全不碎的燕麦会被选择用于更容易的谷物收获和储存，并且会施加更大的压力来进行完全驯化（失去芒，更大的颗粒尺寸等等）。这个模型的一个复杂之处在于燕麦似乎被驯化了两次。现在大多数品种类似于“白”燕麦（A. sativa），在春季播种，与土耳其中部的野生种群最密切相关。然而，在地中海周围，传统的地方品种通常是“红”燕麦（A. byzantina），在秋季播种，与土耳其西南部的野生种群关系最密切。现代育种计划主要基于这两种类型的杂交，英国独特的冬燕麦品种可能特别依赖于A. byzantina的性状。我们对了解红燕麦和白燕麦的起源以及a. fatua的作用有着浓厚的兴趣。这不仅是由于对农业发展的好奇，也是由于需要为现代作物增加新的变种。每一步驯化都造成了遗传瓶颈，可能有价值的种质已经丢失。更好地了解这些瓶颈以及野生种群的自然变异，将有助于找到有用的变异，并将其纳入育种计划。基因序列分型（GbS）是近年来一种揭示全基因组遗传变异的高通量方法。它对特定限制性内切酶位点附近的短序列进行采样，并且不需要事先知道哪些变体存在。可以处理数百个样品，使其成为多样性屏幕的理想选择。我们有超过500种杂草、野生和本地植物的选种，已经经过了GbS的筛选。不幸的是，燕麦基因组几乎和小麦一样庞大和复杂，目前还没有将GbS标签与邻近基因联系起来的参考资料。即使当第一批参考资料（预计在2018年）可用时，它们也将来自现代品种，可能与我们正在研究的祖先品种不同。在这个项目中，我们将弥合GbS和基因组参考文献之间的差距，直接或通过改进可用的参考文献来评估基因变异。在第一步，我们将使用一种新开发的测序方法来降低构建红燕麦参考的成本。在第二个项目中，我们将与一位德国大麦专家合作，他开发了外显子组捕获（一种只对目标基因区域进行取样和测序的方法），并与一位加拿大合作伙伴合作，他已经开发了燕麦的外显子组捕获（部分基于我们的基因组数据）。我们将使用改进的设计从220个关键GbS面板中恢复全面的基因序列集合。最后，我们将获得祖先的序列，以便与现代栽培品种进行基因含量和顺序的比较，并将加强外显子组和GbS数据的分析。我们将与一位创造杂交种群的波兰伙伴合作，确定支持驯化的关键基因。

Oats can claim to be the only 'European' cereal. Previous work indicated that its closest wild relatives are found around Turkey near sites where wheat and barley were first domesticated, but archaeological evidence points to first use as a crop only late in the development of agriculture and then in Central Europe, rather than in its original environment. A plausible explanation for this transfer is that a population of the wild relative (Avena sterilis), which colonises broken ground, had become adapted to life as an agricultural weed. The new weed (A. fatua) is less able to spread in the wild but is very successful at colonising tilled fields of cereals such as barley. A. fatua could have spread across Europe with those crops and then become a crop in its own right in hard times where others failed. Indeed, oats have traditionally been grown on poor ground and at the end of rotations when soil fertility is low, a possible legacy of this secondary domestication. At some stage completely non-shattering oats would have been selected for easier grain harvest and storage, and greater pressure would have been applied for full domestication (loss of awns, larger grain size and so on). A complication for this model is that oats appear to have been domesticated twice. Most cultivars now resemble 'white' oats (A. sativa), sown in the spring and most closely related to wild populations in central Turkey. However, around the Mediterranean, traditional landraces are often 'red' oats (A. byzantina), sown in autumn and most closely related to wild populations in south western Turkey. Modern breeding programmes are largely based on crosses between these two types, and the UK's unique winter oat cultivars may be particularly dependent on A. byzantina traits. We have a strong interest in understanding the origin of both red and white oats, and the role of A. fatua. This is not only driven by curiosity about the development of agriculture but also by the need to add new variation to the modern crop. Each domestication step has created genetic bottlenecks where potentially valuable germplasm has been lost. Better understanding of these bottlenecks, and of natural variation in wild populations, will help find useful variation to incorporate into breeding programmes.Genotyping-by-sequence (GbS) is a recent high throughput method to reveal genetic variation across entire genomes. It samples short sequences adjacent to specific restriction enzyme sites, and does not require prior knowledge of which variants are present. Hundreds of samples may be processed, making it ideal for diversity screens. We have panels of over 500 weedy, wild and landrace accessions already screened by GbS. Unfortunately, the oat genome is almost as large and complex as that of wheat, and no reference is yet available to associate GbS tags with neighbouring genes. Even when the first references become available (expected in 2018), they will have been derived from modern cultivars which may differ from the ancestral stocks we are studying. In this project we will bridge the gap between GbS and genome references to assess gene variation either directly or by improving the references available. In the first step we will use a newly developed sequencing approach to cut the cost of building a red oat reference. In the second we will work with a German barley expert who developed exome capture, a method that samples and sequences only targeted gene regions, and with a Canadian partner who has created an exome capture for oats, based in part on our genome data. We will use an improved design to recover comprehensive collections of gene sequences from 220 key GbS panel accessions. Finally we will obtain sequence from progenitors that will allow comparisons of gene content and order with modern cultivars, and will enhance analysis of exome and GbS data. Working with a Polish partner who has created hybrid populations, we will identify key genes underpinning domestication.