Molecular and archaeological evidence on the geographical origin of domestication for Camelina sativa

Abstract

Premise: Camelina (gold-of-pleasure or false flax) is an ancient oilseed crop with emerging applications in the production of sustainable, low-input biofuels. Previous domestication hypotheses suggested a European or western Asian origin, yet little genetic evidence has existed to assess the geographical origin for this crop, and archaeological data have not been systematically surveyed.

Methods: We utilized genotyping-by-sequencing of 185 accessions of C. sativa and its wild relatives to examine population structure within the crop species and its relationship to populations of its wild progenitor, C. microcarpa; cytotype variation was also assessed in both species. In a complementary analysis, we surveyed the archaeological literature to identify sites with archaeobotanical camelina remains and assess the timing and prevalence of usage across Europe and western Asia.

Results: The majority of C. microcarpa sampled in Europe and the United States belongs to a variant cytotype (2n = 38) with a distinct evolutionary origin from that of the crop lineage (2n = 40). Populations of C. microcarpa from Transcaucasia (South Caucasus) are most closely related to C. sativa based on cytotype and population structure; in combination with archaeological insights, these data refute prior hypotheses of a European domestication origin.

Conclusions: Our findings support a Caucasus, potentially Armenian, origin of C. sativa domestication. We cannot definitively determine whether C. sativa was intentionally targeted for domestication in its own right or instead arose secondarily through selection for agricultural traits in weedy C. sativa, as originally proposed by Vavilov for this species.

Keywords: Brassicaceae; Camelina; allopolyploid; biofuel; crop mimicry; domestication; population genetics; weed; wild crop relatives.

Figure 1

Growth habit of (A) Camelina sativa and (B) C. microcarpa (2n = 40). Rapidly developing and inflated seed pods and substantially larger seeds can be observed in the domesticate in comparison to C. microcarpa. Scale bars = 1 mm.

Figure 2

PCA generated from the final SNP data set for 183 Camelina individuals that passed filtering. Colored dots represent individual's country of origin and ellipses represent taxonomic groups, as indicated including observed karyotypes within those groups.

Figure 3

PCA generated from the final data set of 10,737 SNPs from 110 individuals, including only 2n = 40 Camelina microcarpa and C. sativa. Colored dots represent individual's country of origin.

Figure 4

Analysis of relatedness and population structure of 2n = 40 Camelina microcarpa and C. sativa individuals. (A) Neighbor‐joining tree with individuals colored according to population genetic identity >70%; C. sativa, red; C. microcarpa Armenia, light blue; C. microcarpa Georgia, dark blue; hybrid (<70% population identity to any genetic group), green. Admixed C. sativa individuals denoted with ‘*Cs’. (B) Admixture plots at K = 2 and K = 3 with morphologically determined C. sativa (left) and C. microcarpa (right) split.

Figure 5

Map of archaeological records for Camelina across Europe and Western Asia. Points are mapped according to the approximate location of archaeological sites. Triangles = C. sativa; squares = C. microcarpa; circles = Camelina sp. or unidentified. Letters above and to the left or right of sites indicate that the plants are likely cultivated (C) or weedy/wild (W). Numbers below sites represent approximate number of Camelina seeds recovered at sites for which this information was available. Sites are shaded to denote approximate age of the site or sample, from black (oldest) to white (youngest).