Sympatric divergence of the ergot fungus, Claviceps purpurea, populations infecting agricultural and nonagricultural grasses in North America

Abstract

The ergot diseases of agricultural and nonagricultural grasses are caused by the infection of Claviceps spp. (Hypocreales, Ascomycota) on florets, producing dark spur-like sclerotia on spikes that are toxic to humans and animals, leading to detrimental impacts on agriculture and economy due to the downgrading of cereal grains, import-export barriers, reduced yield, and ecological concerns. At least seven phylogenetic lineages (phylogenetic species) were identified within the premolecular concept of C. purpurea s.l. (sensu lato) in agricultural areas and vicinities in Canada and the Western United States. Claviceps purpurea s.s (sensu stricto) remained as the most prevalent species with a wide host range, including cereal crops, native, invasive, and weedy grasses. The knowledge on genetic diversity and distribution of C. purpurea s.s. in North America is lacking. The objective of the present study was to shed light on genetic differentiation and evolution of the natural populations of C. purpurea s.s. Multilocus DNA sequences of samples from Canada and the Western USA were analyzed using a phylogenetic network approach, and population demographic parameters were investigated. Results showed that three distinct genetically subdivided populations exist, and the subdivision is not correlated with geographic or host differentiations. Potential intrinsic mechanisms that might play roles in leading to the cessation of gene flows among the subpopulations, that is, mating and/or vegetative incompatibility, genomic adaptation, were discussed. The neutrality of two house-keeping genes that are widely used for DNA barcoding, that is, translation elongation factor 1-α (TEF1-α) and RNA polymerase II second largest subunit (RPB2), was challenged and discussed.

Keywords: Ascomycota; house‐keeping gene; multilocus haplotype; neutrality; phylogenetic network; population structure; selective sweeping.

FIGURE 1

Claviceps purpurea sclerotia on grasses Elymus repens in the field heads (background), the asexual stage in honeydew after inoculating a barley plant in greenhouse (upper inset), and the sexual stage produced from a germinating sclerotium in a controlled environment (bottom inset)

FIGURE 2

Geographic locations of studied samples, and designation of geographic regions: western Canada (CAW), eastern Canada (CAE) and western USA (USW). Larger sizes of rhombus marks indicate more samples

FIGURE 3

Network analyses based on 146 haplotypes of four‐locus concatenated sequences using SplitsTree4 V4.14.8 (a) and BEAST v2.5 (b)

FIGURE 4

Principal coordinate analyses (PcoA) of 146 haplotypes using GenAlex 6.5. Populations were assigned based on genetic cluster (a), geographical regions (b), and host groups (c)

FIGURE 5

Percentages of three genetic clusters in three geographic regions (a), and three host groups (b)

FIGURE 6

Mantel test for isolation by distance (IBD) showing no evidence of a correlation between genetic distance versus log geographic distance

FIGURE 7

Allelic pattern across three genetic clusters. N _a, number of different alleles; N _e, number of effective alleles = $1/(\sum P_i^2)$, I Shannon's information index = −1×∑(P_iLn(P_i)), h, haploid genetic diversity = $1‐\sum P_i^2$, where P _i is the frequency of the ith allele for the population (cluster)