Search

Understanding the transition of epidemic to endemic dengue transmission remains a challenge in regions where serotypes co-circulate and there is extensive human mobility. French Polynesia, an isolated group of 117 islands of which 72 are inhabited, distributed among five geographically separated subdivisions, has recorded mono-serotype epidemics since 1944, with long inter-epidemic periods of circulation. Laboratory confirmed cases have been recorded since 1978, enabling exploration of dengue epidemiology under monotypic conditions in an isolated, spatially structured geographical location. A database was constructed of confirmed dengue cases, geolocated to island for a 35-year period. Statistical analyses of viral establishment, persistence and fade-out as well as synchrony among subdivisions were performed. Seven monotypic and one heterotypic dengue epidemic occurred, followed by low-level viral circulation with a recrudescent epidemic occurring on one occasion. Incidence was asynchronous among the subdivisions. Complete viral die-out occurred on several occasions with invasion of a new serotype. Competitive serotype replacement has been observed previously and seems to be characteristic of the South Pacific. Island population size had a strong impact on the establishment, persistence and fade-out of dengue cases and endemicity was estimated achievable only at a population size in excess of 175 000. Despite island remoteness and low population size, dengue cases were observed somewhere in French Polynesia almost constantly, in part due to the spatial structuration generating asynchrony among subdivisions. Long-term persistence of dengue virus in this group of island populations may be enabled by island hopping, although could equally be explained by a reservoir of sub-clinical infections on the most populated island, Tahiti., Author summary Dengue virus is the most significant arthropod-borne virus infecting man. Understanding how long dengue virus can persist in populations of varying size is key to understanding its epidemiology. This is, however, impossible to achieve in settings where dengue is endemic, because of continued human movement and is further complexified by the occurrence of several co-circulating serotypes. By contrast, French Polynesia, an isolated group of 72 inhabited islands in the South Pacific, has had intermittent majoritarily monotypic dengue epidemics since the 1940s and offers a unique opportunity to address questions of viral persistence, turnover and the importance of spatial sub-structure in determining dengue epidemiology. Collating and analyzing a database of laboratory-confirmed dengue cases from across French Polynesia over a 35 year period we were able to show that dengue virus die-out can occur with or without replacement by a new serotype, monotypic transmission of dengue viruses fails to be maintained within small island populations but can persist for years among isolated islands connected via air and sea links. This remarkable long-term persistence of dengue virus in French Polynesia could be maintained by asynchronous viral transmission among connected islands and/or by repeated seeding from a reservoir of sub-clinical infections in the most populated island, Tahiti.

International audience; Multifurca is a small genus newly established to accommodate lactarioid and russuloid species with some characters reminiscent of corticoid members of Russulaceae. It shows an amphi-pacific distribution with strong preference for the tropical zone of the Northern Hemisphere and thus has particular significance for biogeographical study. Using worldwide samples and three loci (ITS, 28S rDNA and rpb2), we demonstrated that Multifurca is split into two highly supported major clades that are here recognized at the subgeneric level: subg. Furcata subg. nov. exclusively includes lactarioid species, while subg. Multifurca includes species with a russuloid habit. Using phylogenetic species recognition and comparison of genetic distances we recognize five new and six previously described species, almost double the known number of species before this study. Molecular dating using a Bayesian method suggested that Multifurca originated in early Paleocene and diversified in the Eocene. The most recent interspecific divergences occurred both in Asia and America, roughly at the same time around the Pliocene. Ancestral area reconstruction and comparisons of genetic distances and morphology suggested an early divergence within Australasia or tropical Asia. From the early Miocene to Pliocene, multiple dispersals/migrations to Australasia and North America by island hopping or land bridge likely happened. Vicariance at the late Tertiary might be the most likely mechanism accounting for the eastern Asia—southeastern North America and Australasia—tropical Asia disjunct distributions. The shared polymorphisms in the ITS alignment, numerous degenerated base pairs in the rpb2 sequences and weak conflict between the ITS and LSU genealogies of M. subg. Furcata suggest recent speciation. Host specificity of Multifurca species or species pairs is relatively low. Host shifts are believed to have aided establishment in new territories during the dispersals and migrations.