Thousands of plants have been selected as crops, yet, only a few are fully domesticated. The lack of adaptation to agroecological environments of many crop plants with few characteristic domestication traits potentially has genetic causes. Here, we investigate the incomplete domestication of an ancient grain from the Americas, amaranth. Although three grain amaranth species have been cultivated as crop for millennia, all three lack key domestication traits. We sequenced 121 crop and wild individuals to investigate the genomic signature of repeated incomplete adaptation. Our analysis shows that grain amaranth has been domesticated three times from a single wild ancestor. One trait that has been selected during domestication in all three grain species is the seed color, which changed from dark seeds to white seeds. We were able to map the genetic control of the seed color adaptation to two genomic regions on chromosome 3 and 9, employing three independent mapping populations.Within the locus on chromosome 9, we identify a MYB-like transcription factor gene, a known regulator for seed color variation in other plant species. We identify a soft selective sweep in this genomic region in one of the crops species but not in the other two species. The demographic analysis of wild and domesticated amaranths revealed a population bottleneck predating the domestication of grain amaranth. Our results indicate that a reduced level of ancestral genetic variation did not prevent the selection of traits with a simple genetic architecture but may have limited the adaptation of complex domestication traits.

Out of the almost 2,000 plants that have been selected as crops, only a few are fully domesticated, and many intermediates between wild plants and domesticates exist. Genetic constraints might be the reason why incompletely domesticated plants have few characteristic crop traits, and retained numerous wild plant features. Here, we investigate the incomplete domestication of an ancient grain from the Americas, amaranth. We sequenced 121 genomes of the crop and its wild ancestors to show that grain amaranth has been selected three times independently from a single wild ancestor, but has not been fully domesticated. While only few domestication traits have been adapted during amaranth domestication, the seed color converted from dark seeds to white seeds between the ancestor and the crops. To investigate the seed color adaptation in amaranth we produced a mapping population between wild and domesticated amaranth, a whole genome sequenced bulked segregant population and perform genome wide association mapping in our divers panel. All three methods agree on two quantitative trait loci controlling the trait. We identify a MYB-like transcription factor gene, a known regulator for seed color variation in other plant species, within one of the significant regions and shows that the trait was independently converted in Central and South America. We propose that a low effective population size at the time of domestication might have contributed to the lack of adaptation of complex domestication traits. Our results show how genetic constraints influenced domestication and might have set the fate of hundreds of crops. SignificanceEarly farmers cultivated hundreds of plant species and altered their morphological and physiological characteristics during domestication to be well suited for human consumption. Despite long cultivation histories, many crops remained of minor importance, because they were limited in their adaptation to agro-ecological systems. We investigate the domestication history of the ancient minor crop amaranth and show that it was independently domesticated three times from the same ancestor. In all three domesticates, white seed color, which is controlled by only two genes, was independently selected. In contrast, more complex domestication traits like seed size were not changed during domestication, and our analyses suggest that this may have resulted from a lack of functional genetic variation in the ancestor during domestication.

Setosphaeria turcica is a major fungal pathogen of maize and causes the foliar disease Northern corn leaf blight (NCLB). It originates from tropical regions and expanded into Central Europe since the 1980s, simultaneously with a rapid increase of maize cultivation area in this region. To investigate evolutionary processes influencing the rapid expansion of S. turcica we sequenced 121 isolates from Central Europe, Western Europe and Kenya. Population genetic inference revealed five genetically distinct clusters that differ by their geographic distribution and emergence dates. One genetically diverse cluster is restricted to Kenya, and the four European clusters consist of three distinct clonal lineages with low genetic diversity and one genetically diverse cluster with several clonal sublineages. A comparison of two different coalescent models for genetic diversity in the most frequent and geographically widespread clonal lineage in Europe supported a model of neutral, strongly exponential population growth over models accounting for different types of selection. In contrast to Kenyan isolates, European isolates did not show sexual recombination despite the presence of both mating types MAT1-1 and MAT1-2 in Europe. Within clonal lineages phenotypic variation in virulence to different monogenic resistances likely originated from repeated de novo mutations in virulence genes of S. turcica. k-mer based association mapping between genetic clusters did not identify genomic regions associated with pathogen races but few genomic regions that are significantly differentiated between two clonal lineages and contain putative effector genes. Our results suggest that the rapid colonization of Europe by different clonal lineages of S. turcica was not driven by selection of virulent races but reflects a neutral demographic process of fast pathogen population growth fostered by a rapid expansion of the maize cultivation area in this region.

Modern agricultural practices, climate change, and globalization foster the rapid spread of plant pathogens, such as the maize fungal pathogen Setosphaeria turcica, which causes Northern corn leaf blight and expanded into Central Europe during the twentieth century. To investigate the rapid expansion of S. turcica, we sequenced 121 isolates from Europe and Kenya. Population genomic inference revealed a single genetically diverse cluster in Kenya and three clonal lineages with low diversity, as well as one cluster of multiple clonal sublineages in Europe. Phylogenetic dating suggests that all European lineages originated through sexual reproduction outside Europe and were subsequently introgressed multiple times. Unlike isolates from Kenya, European isolates did not show sexual recombination, despite the presence of both MAT1-1 and MAT1-2 mating types. For the clonal lineages, coalescent model selection supported a selectively neutral model with strong exponential population growth, rather than models with pervasive positive selection caused by host defense resistance or environmental adaptation. Within clonal lineages, phenotypic variation in virulence to different monogenic resistances, which defines the pathogen races, suggests that these races may originate from repeated mutations in virulence genes. Association testing based on k-mers did not identify genomic regions linked to pathogen races, but it did uncover strongly differentiated genomic regions between clonal lineages, which harbor genes with putative roles in pathogenicity. In conclusion, the expansion and population growth of S. turcica in Europe are mainly driven by an expansion of the maize cultivation area and not by rapid adaptation.

No abstract