Spatial heterogeneity in the Mediterranean Biodiversity Hotspot affects barcoding accuracy of its freshwater fishes.

Incomplete knowledge of biodiversity remains a stumbling block for conservation planning and even occurs within globally important Biodiversity Hotspots (BH). Although technical advances have boosted the power of molecular biodiversity assessments, the link between DNA sequences and species and the analytics to discriminate entities remain crucial. Here, we present an analysis of the first DNA barcode library for the freshwater fish fauna of the Mediterranean BH (526 spp.), with virtually complete species coverage (498 spp., 98% extant species). In order to build an identification system supporting conservation, we compared species determination by taxonomists to multiple clustering analyses of DNA barcodes for 3165 specimens. The congruence of barcode clusters with morphological determination was strongly dependent on the method of cluster delineation, but was highest with the general mixed Yule-coalescent (GMYC) model-based approach (83% of all species recovered as GMYC entity). Overall, genetic morphological discontinuities suggest the existence of up to 64 previously unrecognized candidate species. We found reduced identification accuracy when using the entire DNA-barcode database, compared with analyses on databases for individual river catchments. This scale effect has important implications for barcoding assessments and suggests that fairly simple identification pipelines provide sufficient resolution in local applications. We calculated Evolutionarily Distinct and Globally Endangered scores in order to identify candidate species for conservation priority and argue that the evolutionary content of barcode data can be used to detect priority species for future IUCN assessments. We show that large-scale barcoding inventories of complex biotas are feasible and contribute directly to the evaluation of conservation priorities.

Occurrence and host preferences of Anopheles maculipennis group mosquitoes in England and Wales.

Mosquitoes of the Anopheles maculipennis Meigen (Diptera: Culicidae) group are of public health concern: five of the 11 morphologically indistinct species have been historically considered as vectors of malaria in Europe. Three members of the An. maculipennis group have been reported in the U.K.: Anopheles atroparvus van Thiel; Anopheles messeae Falleroni, and Anopheles daciae Linton, Nicolescu & Harbach. To study the distribution of the three U.K. species, particularly that of An. daciae, we developed a polymerase chain reaction-Restriction fragment length polymorphism (PCR-RFLP) assay using the nuclear ribosomal internal transcribed spacer 2 (ITS-2) gene. Anopheles daciae was found to be widespread, occurring in four of the five counties surveyed in southern England and on the Welsh island of Anglesey, often in sympatry with the closely related species An. messeae. The host preferences of 237 blood-fed females were determined using either direct sequencing or PCR-based fragment analysis of the mitochondrial cytochrome oxidase b gene with DNA from females' abdomens. All three species were found to be opportunistic, having fed on at least three different hosts. Seventeen individuals contained multiple bloodmeals, including two An. daciae that had fed on humans and birds. Our results show that An. daciae is widespread in England and Wales, occurs in sympatry with other members of the An. maculipennis group, and feeds on humans, which suggests it is a potential vector of disease in the U.K.

Genetic differentiation of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in fragmented alpine streams.

The interpretation of low FST values as evidence of high levels of gene flow among habitat fragments may be confounded by population genetic structures that are indicative of historical rather than present-day levels of gene flow. We examined the genetic structure of 23 populations of Baetis alpinus (Insecta: Ephemeroptera) living in alpine streams fragmented by lakes ( approximately 10 000 years old), reservoirs ( approximately 100 years old), and in nonfragmented streams, to examine if lakes act as barriers to gene flow and to investigate the temporal resolution of allozyme markers. Estimates of gene flow indicated little or no genetic divergence along four nonfragmented reference streams and across two lakes and two reservoirs (FST=0.004-0.041), but marked differentiation across four lakes (FST=0.092-0.362) and across one reservoir that was a lake enlarged by a dam (FST=0.075). Differentiation was unrelated to distance between fragments, but occurred only in lakes found in valleys that have been ice-free throughout the Holocene. We suggest that standing water bodies act as barriers to gene flow in B. alpinus and that low FST values observed between fragments separated by reservoirs do not indicate high levels of gene flow but rather show that genetic differentiation was not detectable within the first 100-1000 years of habitat fragmentation.