Although the genus Pseudobulweria was described in 1936 for the Fiji Petrel (Ps. macgillivrayi), its validity, phylogenetic relationships, and the number of constituent taxa it contains remain controversial. We tried to clarify these issues with 496 bp sequences from the mitochondrial cytochrome-b gene of 12 taxa representing three putative subspecies of Pseudobulweria, seven species in six other genera of the Procellariidae (fulmars, petrels, and shearwaters), and one species each from the Hydrobatidae (storm-petrels) and Pelecanoididae (diving-petrels). We also include published sequences for two other petrels (Procellaria cinerea and Macronectes giganteus ) and use Diomedea exulans and Pelecanus erythrorhynchos as outgroups. Based on the pronounced sequence divergence (5 to 5.5%) and separate phylogenetic history from other genera that have been thought to be closely related to or have been synonymized with Pseudobulweria, we conclude that the genus is valid, and that the Mascarene Petrel (Pseudobulweria aterrima) and the Tahiti Petrel (Ps. rostrata) are distinct species. In trees constructed with maximum parsimony and maximum likelihood, Pseudobulweria is the sister taxon to Puffinus and Calonectris, and these genera in turn are most closely related to Bulweria (and Procellaria in the maximum-parsimony tree). Pseudobulweria is not closely related to Pterodroma in either tree. Because Ps. r. trouessarti from New Caledonia, and Ps. r. rostrata from Polynesia differ by only 0.6%, these taxa do not deserve species status and should be regarded as valid subspecies.

Populations of the Little Bustard Tetrax tetrax in the farmlands of Europe have declined greatly over the last century. In Western Europe, France now holds the only remaining migratory population, which currently numbers fewer than 300 displaying males. However, the movements of these birds are virtually unknown, in spite of the important implications of this information for the conservation of this species. We identified migratory movements and overwintering areas of French migratory populations, using wild individuals fitted with satellite or radio‐transmitters. Little Bustards completed their migration journey over a relatively short time period (2–5 days), with nocturnal migration flights of 400–600 km per night. All birds overwintered in Iberia. In addition, we tested the consequences of captive rearing on migratory movements. French wild adults and French captive‐bred juveniles behaved similarly with regard to migration, suggesting that hand‐raising does not alter migratory movements. However, birds originating from eggs collected in Spain and reared in western France did not migrate, suggesting a genetic component to migratory behaviour. These results therefore suggest that a conservation strategy involving the release in France of birds hatched from eggs collected in Spain may imperil the expression of migratory movements of the French population. More generally, to maintain the integrity of native populations, the introduced individuals should mimic their migratory movements and behaviour.

Population structure, in terms of the body mass, condition, sex and reproductive status of individuals, has been found to vary in phase with population density in cyclic populations of microtine rodents. Because sustained population cycles involve delayed density dependent changes in the population growth rate, we would expect at least some life history traits also to depend on past densities. Detailed, long‐term studies of changes in vole life history traits are however few, and are largely restricted to northern Europe. In view of the uncertainty as to whether the cyclic microtine populations of western Europe represent the same phenomenon as those of northern Europe, we studied temporal variation in the structure of a clearly cyclic population of the common vole Microtus arvalis Pallas, in the cereal plains of mid‐western France. Our data set contains seasonal, individual‐level data from long‐term, large‐scale trapping covering four entire population cycles. We found considerable cyclic variation in population structure in spring (April), but less so in summer (June). In spring of post‐peak years, animals were of low body weight and body condition (particularly females), litter sizes were smaller and there was a reduction in the proportion of breeders. All of these could be proximal drivers of increased mortality rates, or decreased birth rates, contributing to the population declines. Few life history traits, however, showed direct density dependent variation, and none of the traits studied here showed delayed density dependence. We have shown declines in the fecundity and body condition of voles from a western European population that coincides with, and may be a proximal cause of, cyclic declines in population density. Closer attention to proximal causes, by which ecological processes drive cycles, could clarify the extent to which microtine cycles across Europe represent a single phenomenon.

Audubon's Shearwater Pufhnus lherminieri is a little known species of petrel. The recent descriptions of a possibly new species (P. atrodorsalis) and a new subspecies (l? 1. colstoni) have suggested that the population breeding on Reunion Island, I! 1. bailloni, may represent a key taxon for understanding the phylogeography of this group in the Indian Ocean. We provide the first detailed information on distribution, abundance, habitat characteristics and breeding biology of the species on Reunion, and assess relationships between the Indian Ocean taxa based on biometrics, coloration characters and patterns of breeding phenology. We also studied the behaviour of the birds on Reunion, paying particular attention to vocalizations. Analysing the parameters that affect calling rate, such as moonlight, season and time, we show that calling activity can be used to estimate colony size. Using this method, we sampled more than half of Reunion Island habitats suitable for Audubon's Shearwater, and estimate that the breeding population there is between 3000 and 5000 pairs in at least 235 colonies. Compared to other Indian Ocean populations ( n = 136 specimens), the analysis of biometrics and coloration characters suggests that three main populations exist (Comoro, Reunion and Seychelles), with birds of Maldives and Amirantes being biometrically inseparable from those of Seychelles, and birds from Aldabra being intermediate between Seychelles and Reunion. The proposed new species, atrodorsalis, also falls in the overlap between Seychelles and Reunion birds, which argues against a full species taxonomic status.

The role of glacial refugia in shaping contemporary species distribution is a longstanding question in phylogeography and evolutionary ecology. Recent studies are questioning previous paradigms on glacial refugia and postglacial recolonization pathways in Europe, and more flexible phylogeographic scenarios have been proposed. We used the widespread common vole Microtus arvalis as a model to investigate the origin, locations of glacial refugia, and dispersal pathways, in the group of "Continental" species in Europe. We used a Bayesian spatiotemporal diffusion analysis (relaxed random walk model) of cytochrome b sequences across the species range, including newly collected individuals from 10 Iberian localities and published sequences from 68 localities across 22 European countries. Our data suggest that the species originated in Central Europe, and we revealed the location of multiple refugia (in both southern peninsulas and continental regions) for this continental model species. Our results confirm the monophyly of Iberian voles and the pre-LGM divergence between Iberian and European voles. We found evidence of restricted postglacial dispersal from refugia in Mediterranean peninsulas. We inferred a complex evolutionary