Long-distance gene flow outweighs a century of local selection and prevents local adaptation in the Irish famine pathogen Phytophthora infestans

Abstract

Sustainably managing plant resistance to epidemic pathogens implies controlling the genetic and demographic changes in pathogen populations faced with resistant hosts. Resistance management thus depends upon the dynamics of local adaptation, mainly driven by the balance between selection and gene flow. This dynamics is best investigated with populations from locally dominant hosts in islands with long histories of local selection. We used the unique case of the potato late blight pathosystem on Jersey, where a monoculture of potato cultivar 'Jersey Royal' has been in place for over a century. We also sampled populations from the coasts of Brittany and Normandy, as likely sources for gene flow. The isolation by distance pattern and the absence of genetic differentiation between Jersey and the closest French sites revealed gene flow at that spatial scale. Microsatellite allele frequencies revealed no evidence of recombination in the populations, but admixture of two genotypic clusters. No local adaptation in Jersey was detected from pathogenicity tests on Jersey Royal and on French cultivars. These data suggest that long-distance gene flow (∼ 50/100 km) prevents local adaptation in Jersey despite a century of local selection by a single host cultivar and emphasize the need for regional rather than local management of resistance gene deployment.

Keywords: clonal lineage; evolution; gene flow; host resistance; local adaptation; microsatellites; potato late blight; selection.

Figure 1

Isolation by distance pattern between genetic differentiation, measured as F_ST/(1 − F_ST), and geographic distance (natural logarithm of the distance in km) for pairwise Phytophthora infestans populations.

Figure 2

Minimum-spanning network showing the relationships among the Phytophthora infestans multilocus genotypes (MLGs 1 to 23) of both clusters (1 in white and 2 in gray). Branch sizes are proportional to genetic distance (i.e., the number of different alleles) and circle areas to the numbers of isolates.

Figure 3

Geographic position of each Phytophthora infestans population used in this study and mean assignment percentage for each population to each genetic clusters. The populations linked by black lines are not genetically differentiated (P > 0.05) based on an F_ST test, and those not linked by black lines are genetically differentiated (P < 0.05).

Figure 4

Mean lesion size and spore production of each Phytophthora infestans population, from Jersey, Paimpol, Saint-Malo, and Val de Saire, on the different potato cultivars, Bintje, Europa, and Jersey Royal. Letters represent the homogenous groups identified with the Tukey's HSD test at the 5% threshold.