Co-evolution between herbivores and plants is believed to be one of the processes creating Earth's biodiversity. However, it is difficult to disentangle to what extent diversification is really driven by herbivores or by other historical-geographical processes like allopatric isolation. In the cruciferous plant Barbarea vulgaris, some Danish individuals are resistant to herbivory by flea beetles (Phyllotreta nemorum), whereas others are not. The flea beetles are, in parallel, either resistant or susceptible to the plants defenses. To understand the historical-evolutionary framework of these interactions, we tested how genetically divergent resistant and susceptible plants are, using microsatellite markers. To test whether they are reproductively fully compatible, resistant and susceptible plants were grown intermixed in an outdoor experiment, and the paternity of open-pollinated offspring was determined by analysis of molecular markers. Resistant and susceptible Danish plants were genetically strongly differentiated and produced significantly fewer hybrids than expected from random mating or nearest neighbour mating. Our results suggest that the two types belong to different evolutionary lineages that have been (partly) isolated at some time, during which genetic and reproductive divergence evolved. A parsimonious scenario could be that the two plant types were isolated in different refugia during the previous ice age, from which they migrated into and met in Denmark and possibly neighbouring regions. If so, resistance and susceptibility has for unknown reasons become associated with the different evolutionary lineages.

Two subspecies of Barbarea vulgaris are taxonomically recognized as ssp. vulgaris and ssp. arcuata. In addition, two types of Barbarea vulgaris ssp . arcuata occurs in Denmark. Th e G-type is resistant to an herbivorous fl ea beetle ( Phyllotreta nemorum ) whereas the P-type is susceptible. A previous study suggested that the P-type evolved by a loss of resistance from a resistant progenitor. We analyzed the genetic relatedness among eight Barbarea taxa: B. vulgaris spp. vulgaris , B. vulgaris ssp. arcuata G-and P-types, hybrids between the types, B. verna , B. intermedia , B. stricta , B. orthoceras and B. australis , using AFLP and SSR markers. A clear distinction between the G-and P-types was revealed. Both were distinct from B. vulgaris ssp. vulgaris , the G-type less so than the P-type. Barbarea verna and B. intermedia formed unambiguous clusters, whereas the remaining taxa produced less discrete groupings. Possible evolutionary scenarios for fl ea-beetle resistance and susceptibility are discussed, including lineage sorting from a polymorphic ancestral population, and de novo loss of resistance in the P-type of B. vulgaris ssp. arcuata.

Interactions between herbivores and plants are believed to have been important drivers of biodiversity. However, to drive an initial resistance divergence into different evolutionary lineages and taxa, these interactions have probably been embedded in other processes of divergence, like allopatric isolation. The cruciferous plant Barbarea vulgaris ssp. arcuata occurs in Denmark in two types: one (G) is resistant to most genotypes of the flea beetle Phyllotreta nemorum, the other (P) is susceptible. The two types additionally differ in hairiness and glucosinolates, they are genetically strongly divergent, and reproduction between them is reduced. To determine whether the two plant types and their resistance polymorphisms are also present outside Denmark, and to understand how they have evolved, we analysed 33 European populations of B. vulgaris for resistance, hairiness, glucosinolates, and microsatellite markers. Most populations had traits indicative of the G type, including the already characterized Danish G populations. In contrast, only two populations outside Denmark were of the P type; one from northern Sweden and one from Estonia. Genetically, the G populations formed two genetic clusters that were strongly divergent from a genetic cluster containing P populations. A fourth genetic cluster, which contained only a single population and no Danish plants, belonged morphologically to the subspecies ssp. vulgaris. The divergence found in Denmark between a resistant G and a susceptible P type is thus part of a larger divergence in Europe. Judging from the trait correlations, genetic divergence, and partial reproductive incompatibility, the plant types must have been isolated from each other for quite some time. The two P populations outside Denmark came from the north and east, suggesting a more eastern distribution. If so, resistant and susceptible types could have diverged during the ice age and later met in Scandinavia. However, more samples from Eastern Europe are needed to clarify this.

Crops were domesticated from wild plants not too long ago and have subsequently diverged from the wild ones, especially in traits used by humans. Whether divergence between the cultigen and wild forms has also lead to reduced reproductive compatibility is unknown for many species. Chicory (Cichorium intybus L.) has been bred as a crop at least since Roman times. To test if this has led to a loss in reproductive compatibility with wild chicory, we planted cultivar, wild, and F 1 hybrid plants into two field plots, and let them pollinate freely. On 2 days, in the beginning and middle of the flowering season, we counted the numbers of flowering capitula and open flowers per capitulum, which in combination with counts of viable pollen per flower were used to estimate the expected proportion of seeds fathered by cultivar, wild, and hybrid plants. Open capitula on wild and hybrid plants were marked, and when seeds were ripe we determined whether cultivar, wild or hybrid plants had pollinated the seeds, using AFLP markers. Cultivar plants fathered much fewer seeds than expected, both on wild and hybrid plants, suggesting that some degree of incompatibility has evolved between cultivar and wild chicory. Hybrid plants fathered more seeds than expected on some of the hybrid plants, indicating that hybrids do not suffer much from outbreeding depression. Our study thus suggests that cultivated and wild chicory, even though they belong to the same species, have diverged to an extent where reproduction between them is somewhat impaired.

No abstract