Lyme disease is caused by bacteria of the Borrelia burgdorferi genospecies complex and transmitted by Ixodid ticks. In North America only one pathogenic genospecies occurs, in Europe there are several. According to the dilution effect hypothesis (DEH), formulated in North America, nymphal infection prevalence (NIP) decreases with increasing host diversity since host species differ in transmission potential. We analysed Borrelia infection in nymphs from 94 forest stands in Belgium, which are part of a diversification gradient with a supposedly related increasing host diversity: from pine stands without to oak stands with a shrub layer. We expected changing tree species and forest structure to increase host diversity and decrease NIP. In contrast with the DEH, NIP did not differ between different forest types. Genospecies diversity however, and presumably also host diversity, was higher in oak than in pine stands. Infected nymphs tended to harbour Borrelia afzelii infection more often in pine stands while Borrelia garinii and Borrelia burgdorferi ss. infection appeared to be more prevalent in oak stands. This has important health consequences, since the latter two cause more severe disease manifestations. We show that the DEH must be nuanced for Europe and should consider the response of multiple pathogenic genospecies.