Shifting nonbreeding distributions of migratory fauna in relation to climatic change

The distributions of eight out of nine common species of waders (Charadrii) overwintering on UK estuaries have changed in association with recent climate change. These birds represent a high proportion of various populations from breeding grounds as far apart as Greenland to the west to high‐arctic Russia to the east. During warmer winters, smaller proportions of seven species wintered in south‐west Britain. The distributions of the smaller species show the greatest temperature dependence. The opposite was found for the largest species and no relationship was found for a particularly site‐faithful species. In north‐west Europe, the winter isotherms have a broadly north to south alignment, with the east being colder than the west. The average minimum winter temperatures across the UK having increased by about 1.5°C since the mid‐1980s, the temperatures on the east coast during recent winters have been similar to those of the west coast during the mid‐1980s. On average, estuaries on the east and south coasts of Britain have muddier sediments than those on the west coast and thus support a higher biomass of the invertebrate prey of waders. We suggest that, with global climatic change, the advantage gained by waders wintering in the milder west to avoid cold weather‐induced mortality is diminished. Consequently, more choose to winter in the east and thus benefit from better foraging opportunities. The implications of these results are considered in terms of a site‐based approach to wildlife protection used in Europe and elsewhere.     

Heterogeneity, speciation/extinction history and climate: explaining regional plant diversity patterns in the Cape Floristic Region

This paper investigates the role of heterogeneity and speciation/extinction history in explaining variation in regional scale (c. 0.1–3000 km²) plant diversity in the Cape Floristic Region of south‐western Africa, a species‐ and endemic‐rich biogeographical region. We used species‐area analysis and analysis of covariance to investigate geographical (east vs. west) and topographic (lowland vs. montane) patterns of diversity. We used community diversity as a surrogate for biological heterogeneity, and the diversity of naturally rare species in quarter degree squares as an indicator of differences in speciation/extinction histories across the study region. We then used standard statistical methods to analyse geographical and topographic patterns of these two measures. There was a clear geographical diversity pattern (richer in the west), while a topographic pattern (richer in mountains) was evident only in the west. The geographical boundary coincided with a transition from the reliable winter‐rainfall zone (west) to the less reliable non‐seasonal rainfall zone (east). Community diversity, or biological heterogeneity, showed no significant variation in relation to geography and topography. Diversity patterns of rare species mirrored the diversity pattern for all species. We hypothesize that regional diversity patterns are the product of different speciation and extinction histories, leading to different steady‐state diversities. Greater Pleistocene climatic stability in the west would have resulted in higher rates of speciation and lower rates of extinction than in the east, where for the most, Pleistocene climates would not have favoured Cape lineages. A more parsimonious hypothesis is that the more predictable seasonal rainfall of the west would have favoured non‐sprouting plants and that this, in turn, resulted in higher speciation and lower extinction rates. Both hypotheses are consistent with the higher incidence of rare species in the west, and higher levels of beta and gamma diversity there, associated with the turnover of species along environmental and geographical gradients, respectively. These rare species do not contribute to community patterns; hence, biological heterogeneity is uniform across the region. The weak topography pattern of diversity in the west arises from higher speciation rates and lower extinction rates in the topographically complex mountains, rather than from the influence of environmental heterogeneity on diversity.     

Beta diversity of angiosperms in temperate floras of eastern Asia and eastern North America

The diversity of a region reflects both local diversity and the turnover of species (beta diversity) between areas. The angiosperm flora of eastern Asia (EAS) is roughly twice as rich as that of eastern North America (ENA), in spite of similar area and climate. Using province/state‐level angiosperm species floras, we calculated beta diversity as the slope of the relationship between the log of species similarity (S ) and either geographic distance or difference in climate. Distance‐based beta diversity was 2.6 times greater in the north–south direction in EAS than in ENA and 3.3 times greater in the east–west direction. When ln S was related to distance and climate difference in multiple regressions, both distance and climate PC1 were significant effects in the north–south direction, but only geographic distance had a significant, unique influence in the east–west direction. The general predominance of distance over environment in beta diversity suggests that history and geography have had a strong influence on the regional diversity of these temperate floras.     

The ecological significance of the zooplanktivores, snipefish Macroramphosus spp. and boarfish Capros aper, in the food web of the south-east North Atlantic

The diets of slender snipefish Macroramphosus gracilis, longspine snipefish Macroramphosus scolopax and boarfish Capros aper, three very abundant species on the Portuguese coast, were studied from samples collected between July 2002 and April 2003. Variations in the diet with fish size, season and area, as well as diet overlap and diversity, are explored in this study. The diets of slender snipefish and boarfish were mainly composed of copepods, with the main prey being Temora spp. and Calanus spp., respectively. Mysid shrimps were the most important food item in the diet of longspine snipefish. During the summer season, when the availability of different prey items was highest, the two species of snipefish and the boarfish fed on different prey. Temora spp. were the most important prey in the stomachs of slender snipefish in the summer on the south‐west coast, while Calanus spp. started appearing in the stomachs of boarfish in the spring in the north, increasing their abundance in the summer on the south and south‐west coasts. The abundance of mysids appeared to increase in the autumn on the south‐west coast, being at that time an important food item for both longspine snipefish and boarfish. For slender snipefish and boarfish, the differences in stomach species diversity were explained firstly by the season and then by the area and fish size. For longspine snipefish, the area did not explain the species diversity variability, season being the first variable determining the differences. Of all three species, slender snipefish was the one with highest diversity of stomach contents, particularly in spring and summer on the north and south‐west coasts. Diet overlap between species was very low. No predation on eggs and larval stages of fishes was found for any of these fish species. This work is the first to address the diets of snipefish species and boarfish in the south‐east North Atlantic, in a large spatio‐temporal coverage. These species are important prey for many commercial species on the Portuguese coast and, given their abundance, may have a great impact on zooplankton communities, thus assuming a pivotal position in marine food webs.     

Impact of Pleistocene aridity oscillations on the population history of a widespread,vagile Australian mammal,Macropus fuliginosus

Aim Climatic fluctuations during the Pleistocene have shaped the population structure of many extant taxa. However, few studies have examined widespread species inhabiting the Australian continent, where periods of increased aridity characterized the Pleistocene. Here we investigate the phylogeography and population history of a widespread and vagile southern Australian marsupial, the western grey kangaroo (Macropus fuliginosus). Location Southern Australia. Methods We examined the variation of the mitochondrial DNA (mtDNA) control region from 511 individuals of M. fuliginosus sampled throughout their transcontinental distribution. Maximum likelihood and Bayesian analyses were used to investigate the phylogeography and coalescence analyses were then used to test hypothesized biogeographical scenarios. Results The combined results of the phylogeographical and coalescence analyses revealed a complex evolutionary history. Macropus fuliginosus originated in the south‐west of the continent, with north‐western and south‐western populations subsequently diverging as a result of vicariance events during the mid‐Pleistocene. Subsequent arid phases affected these populations differently. In the north‐west, the expansion and contraction of the arid zone resulted in repeated vicariance events and multiple divergent north‐western mtDNA subclades. In contrast, the south‐western population was less impacted by climatic oscillations but gave rise to a major transcontinental eastward expansion. Main conclusions Macropus fuliginosus exhibits the genetic signature of divergence due to unidentified barriers in south‐western Western Australia, while previously identified barriers across southern Australia appear to have had little impact despite evidence of a broad‐scale range expansion prior to the Last Glacial Maximum (LGM). This pattern of localized expansion and contraction is comparable to unglaciated regions in both the Northern and Southern Hemispheres. Furthermore, this study indicates that despite the potential similarities between Northern Hemisphere glaciation and the activation of dune systems in the Australian arid zone, both of which rendered large areas inhospitable, the biotic responses and resultant phylogeographical signatures are dissimilar. Whereas a limited number of major geographically concordant refugia are observed in glaciated areas, the Southern Hemisphere arid zone appears to be associated with multiple species‐specific idiosyncratic refugia.     

Distribution of South African C3 and C4 species of Cyperaceae in relation to climate and phylogeny

Abstract In this study the contribution of climatic factors and phylogenetic relationships affecting the geographical distribution of C₃ and C₄ genera of the Cyperaceae in South Africa was investigated. The δ¹³C values of herbarium specimens of 68 southern African species from 22 genera and eight tribes were used to assign the species to either the C₃ or C₄ photosynthetic pathway. Geographical distribution data for the Cyperaceae were used to investigate relationships between climatic factors and the number of species and proportional abundance of C₄ species per region. The number of Cyperaceae species per 2° × 2° square across South Africa varied from less than five in the north‐western regions to more than 15 in the south‐western and north‐eastern regions of South Africa where rainfall exceeds 800 mm y^‐1. Of the 68 species investigated, 28 had C₄ photosynthesis and these were scattered among nine genera of four tribes (Cypereae, Scirpeae, Abildgaardieae and Rhyncosporeae). The proportional abundance of C₄ species ranged from 14% in the winter rainfall regions of the south‐west of South Africa to 67% in the summer rainfall areas of the north‐east. The geographical distribution of species was related to their phylogenetic position such that the distributions of C₃ and C₄ species in Cypereae, Scirpeae and Schoeneae was quite distinct. Linear regression analysis showed that the transition temperatures (equal C₃ and C₄ species numbers) for the Cyperaceae were different to those obtained for the Poaceae from the same region. No strong relationships were found between the proportional abundance of C₄ species and other climate factors such as altitude and rainfall. Our analysis of the current geographical distribution of C₄ Cyperaceae in southern Africa in a phylogenetic context suggests that the ecological advantages conferred by the C₄ pathway differ amongst the different plant groups.     

Leopard tortoises in southern Africa have greater genetic diversity in the north than in the south (Testudinidae)

In contrast to mammals, little is known about the phylogeographic structuring of widely distributed African reptile species. With the present study, we contribute data for the leopard tortoise (Stigmochelys pardalis). It ranges from the Horn of Africa southward to South Africa and westwards to southern Angola. However, its natural occurrence is disputed for some southern regions. To clarify the situation, we used mtDNA sequences and 14 microsatellite loci from 204 individuals mainly from southern Africa. Our results retrieved five mitochondrial clades; one in the south and two in the north‐west and north‐east of southern Africa, respectively, plus two distributed further north. Using microsatellites, the southern clade matched with a well‐defined southern nuclear cluster, whilst the two northern clades from southern Africa corresponded to another nuclear cluster with three subclusters. One subcluster had a western and central distribution, another occurred mostly in the north‐east, and the third in a small eastern region (Maputaland), which forms part of a biodiversity hotspot. Genetic diversity was low in the south and high in the north of our study region, particularly in the north‐east. Our results refuted that translocations influenced the genetic structure of leopard tortoises substantially. We propose that Pleistocene climatic fluctuations caused leopard tortoises to retract to distinct refugia in southern and northern regions and ascribe the high genetic diversity in the north of southern Africa to genetic structuring caused by the survival in three refuges and subsequent admixture, whereas tortoises in the south seem to have survived in only one continuous coastal refuge.     

GIS analysis of the biogeography of beetles of the subgenus Anisodactylus (Insecta: Coleoptera: Carabidae: genus Anisodactylus)

The advent of GIS technology and the World Wide Web, respectively, facilitate analysing geographical relationships and electronically storing and exchanging biogeographic data. This paper illustrates GIS technology with a study of the subgenus Anisodactylus Dejean (Insecta: Coleoptera: Carabidae: genus Anisodactylus). Species are concentrated in three centres of biodiversity in North America and in four in lands near the western Mediterranean. These centres largely correspond to current areas of wetlands. Eurasia has fewer species than expected based on its area, probably because large portions have habitats unfavourable for the subgenus and/or are poorly collected for Carabidae. Members of the subgenus are primarily adapted to areas with January temperatures between ?10 and 10 °C, July temperatures from 10 to 30 °C and mean annual precipitation from 20 to 200 mm. Cold is apparently a major limiting factor because it typically occurs during several consecutive months of winter and is difficult to escape except by hibernation. Heat is less of a stress when moisture is sufficient. The size of geographical ranges is often larger in the North than in the South and correlates with the latitude of the centre of ranges at r =0.42 (level of significance=0.05). Geographic ranges are often smaller in western North America and in the western Mediterranean than elsewhere in the Northern Hemisphere. Explanations for the smaller sizes include portions of western North America having unfavourable desert or montane habitats, the Rocky Mountains and deserts barring eastward dispersal of species, and the smaller size and more patchy distribution of climatic zones and habitats. In North America geographical ranges west of the Rocky Mountains are north–south elongated because they track primarily north–south orientated climatic zones and because mountains and deserts bar eastward extension. Ranges in north-eastern and north-central North America tend to extend east–west along temperature isotherms. In Eurasia many ranges are stretched east–west because of the shape of the continent and because many northern and southern areas lack suitable habitats. Species with relatively high numbers of apomorphic character states cluster in western Eurasia and to a lesser extent in western North America. The North American centres of biodiversity are post-Wisconsin phenomena while those near the western Mediterranean probably date from the Oligocene or Miocene.     

Identification of glacial refugia in south-eastern North America by phylogeographical analyses of a forest understorey plant, Trillium cuneatum

Aim We examine several hypotheses emerging from biogeographical and fossil records regarding glacial refugia of a southern thermophilic plant species. Specifically, we investigated the glacial history and post‐glacial colonization of a forest understorey species, Trillium cuneatum. We focused on the following questions: (1) Did T. cuneatum survive the Last Glacial Maximum (LGM) in multiple refugia, and (if so) where were they located, and is the modern genetic structure congruent with the fossil record‐based reconstruction of refugia for mesic deciduous forests? (2) What are the post‐glacial colonization patterns in the present geographical range? Location South‐eastern North America. Methods We sampled 45 populations of T. cuneatum throughout its current range. We conducted phylogeographical analyses based on maternally inherited chloroplast DNA (cpDNA haplotypes) and used TCS software to reconstruct intraspecific phylogeny. Results We detected six cpDNA haplotypes, geographically highly structured into non‐overlapping areas. With one exception, none of the populations had mixed haplotype composition. TCS analysis resulted in two intraspecific cpDNA lineages, with one clade subdivided further by shallower diversification. Main conclusions Our investigation revealed that T. cuneatum survived the LGM in multiple refugia, belonging to two (western, eastern) genealogical lineages geographically structured across south‐eastern North America. The western clade is confined to the south‐western corner of T. cuneatum’s modern range along the Lower Mississippi Valley, where fossil records document a major refugium of mesic deciduous forest. For the eastern clade, modern patterns of cpDNA haplotype distribution suggest cryptic vicariance, in the form of forest contractions and subsequent expansions associated with Pleistocene glacial cycles, rather than simple southern survival and subsequent northward colonization. The north–south partitioning of cpDNA haplotypes was unexpected, suggesting that populations of this rather southern thermophilic species may have survived in more northern locations than initially expected based on LGM climate reconstruction, and that the Appalachian Mountains functioned as a barrier to the dispersal of propagules originating in more southern refugia. Furthermore, our results reveal south‐west to north‐east directionality in historical migration through the Valley and Ridge region of north‐west Georgia.     

Range expansion and distributional limits of the nine‐banded armadillo in the United States: an update of Taulman & Robbins (1996)

We conducted a new survey of biologists throughout the southern and central United States, in order to update our last analysis of the range expansion and distributional limits of the nine‐banded armadillo (Dasypus novemcinctus) since 1994. While the armadillo's range has remained stationary to the west along a line corresponding to about 50 cm annual precipitation, it has advanced to the north through central Kansas, into central Illinois, south‐western Indiana and western Kentucky, through central Tennessee, covering Alabama and all but the north‐eastern region of Georgia, and into central South Carolina. The population has reached a latitude corresponding to an average minimum daily January temperature of ?8 °C in Kansas . Armadillos may continue to move northwards in states farther east where they do not yet reach the ?8 °C zone. In the eastern seaboard states, other factors besides winter temperature extremes may be limiting the armadillo's range expansion.     

Impact of mountain chains, sea straits and peripheral populations on genetic and taxonomic structure of a freshwater turtle, Mauremys leprosa (Reptilia, Testudines, Geoemydidae)

Mauremys leprosa, distributed in Iberia and North‐west Africa, contains two major clades of mtDNA haplotypes. Clade A occurs in Portugal, Spain and Morocco north of the Atlas Mountains. Clade B occurs south of the Atlas Mountains in Morocco and north of the Atlas Mountains in eastern Algeria and Tunisia. However, we recorded a single individual containing a clade B haplotype in Morocco from north of the Atlas Mountains. This could indicate gene flow between both clades. The phylogenetically most distinct clade A haplotypes are confined to Morocco, suggesting both clades originated in North Africa. Extensive diversity within clade A in south‐western Iberia argues for a glacial refuge located there. Other regions of the Iberian Peninsula, displaying distinctly lower haplotype diversities, were recolonized from within south‐western Iberia. Most populations in Portugal, Spain and northern Morocco contain the most common clade A haplotype, indicating dispersal from the south‐western Iberian refuge, gene flow across the Strait of Gibraltar, and reinvasion of Morocco by terrapins originating in south‐western Iberia. This hypothesis is consistent with demographic analyses, suggesting rapid clade A population increase while clade B is represented by stationary, fragmented populations. We recommend the eight, morphologically weakly diagnosable, subspecies of M. leprosa be reduced to two, reflecting major mtDNA clades: Mauremys l. leprosa (Iberian Peninsula and northern Morocco) and M. l. saharica (southern Morocco, eastern Algeria and Tunisia). Peripheral populations could play an important role in evolution of M. leprosa because we found endemic haplotypes in populations along the northern and southern range borders. Previous investigations in another western Palearctic freshwater turtle (Emys orbicularis) discovered similar differentiation of peripheral populations, and phylogeographies of Emys orbicularis and Mauremys rivulata underline the barrier status of mountain chains, in contrast to sea straits, suggesting common patterns for western Palearctic freshwater turtles.     

Faunal relationships and zoogeographical affinities of mammals in north-west Africa

The nonmarine mammal fauna of the Maghreb region of north‐west Africa is related to that from three potential source areas: the northern Palaearctic (Europe and south‐west Asia; here referred to as the European fauna), subsaharan Africa (the African fauna) and the arid Palaearctic (Sahara Desert: the desert fauna). On the basis of geographical distribution patterns, this fauna divisible into two groups: the bats, whose affinities are most closely related to southern Europe and south‐west Asia, and nonflying species, most closely related to subsaharan Africa but with an appreciable northern Palaearctic element. These affinities are even more pronounced if desert fauna are removed from the analysis. The nonflying European fauna probably colonized via south‐west Asia and north Africa, rather than direct from western Europe. The results demonstrate that terrestrial habitat barriers are less of an impediment to dispersal, for all mammals except bats, than even narrow stretches of water. The fauna of the Maghreb may be undergoing faunal relaxation, following immigration from tropical Africa and south‐west Asia during mesic phases in the Late Pleistocene and early Holocene.     

Climatic and biogeographical associations of southern African dung beetles (Coleoptera: Scarabaeidae s. str.)

Climatic and biogeographical associations of southern African dung beetles (Scarabaeinae, Coprinae) were analysed from a collection amassed between 1971 and 1986. Endemism to Africa south of 15°S was much greater in southwesterly climates (winter rainfall, bimodal spring/autumn rainfall, arid late summer rainfall) than to the north‐east in the moist, mid‐summer rainfall region. Major biogeographical groups centred to the south‐west comprised predominantly southern African endemics, Western/Eastern Cape coast, Karoo, Karoo/Namib, northern Namibia and the south‐western Kalahari. Biogeographical groups centred on the south‐eastern highlands and the subtropical east coast (mid‐summer rainfall) also comprised predominantly southern African endemics. All other major groups centred to the north‐east in the mid‐summer rainfall region comprised predominantly species with widespread tropical biogeographical affiliations, pan‐southern Africa but centred in the east, pan‐mid‐summer rainfall region, eastern mid‐summer rainfall region, tropical east Zimbabwe/central Moçambique, subtropical/tropical game reserves (non‐ruminant dung specialists). There were cross‐climatic differences in taxonomic composition of the fauna. Within the winter rainfall region, percentage species composition of Scarabaeinae was greater whereas that of the coprine tribe, Onthophagini, was lower than within the other three climatic regions. Percentage species richness of most other tribes of Coprinae differed little between most climates but that of Scarabaeinae declined from west to east (Canthonini, Scarabaeini), east to west (Sisyphini) or to either side of the late summer rainfall region (Gymnopleurini).     

Spatial variation in density, mean size and physiological condition of the holarctic amphipod Diporeia spp. in Lake Michigan

• 1 We examined spatial patterns in population characteristics (density, biomass, mean body length) and physiological condition (lipid content, length‐weight) of the amphipod Diporeia spp. in Lake Michigan by collecting samples at up to 85 sites in late summer 1994 and 1995. Variables were examined relative to water depth and three lake regions: south, central and north. Most major river systems are found in the south, and this region is more nutrient‐enriched compared to the north.
• 2 Over all sites, mean density was 5240^‐2, biomass was 4.1g dry wt m^‐2, and mean body length was 5 mm. While maximum densities were related to depth, with a peak at 30–70 m, greatest densities occurred on the west side of the lake, and low densities were found in the south‐east, north‐east and lower Green Bay. High densities in the west probably resulted from upwelling, and reduced densities in the south‐east may reflect food competition with Dreissena polymorpha (zebra mussel).
• 3 Lipid content, weight per unit length, and mean length declined with increased water depth, but depth‐related trends were most evident in the south. Overall, mean lipid content and weight per unit length were significantly lower in the south (16.6% dry wt, 0.59 mg at 5 mm body length) compared to the north (23.7% dry wt, 0.78 mg at 5 mm body length). These regional differences may have resulted from greater diatom availability in the north and competition from D. polymorpha in the south. Triacylglycerols and phospholipids were the dominant lipid classes in all three regions. Although the mean proportion of triacylglycerols, the energy‐storage lipid, was lower in the south than in the north, regional differences in proportions of lipid classes were not significant.
• 4 Mean lipid content and weight per unit length of Diporeia in the south were lower than values found in the late 1980s prior to the establishment of Dreissena. Mean lipid content of mature individuals is now at levels considered a minimum for successful reproduction.

     

Multi-scale determinants of dung beetle assemblage structure across abiotic gradients of the Kalahari–Nama Karoo ecotone, South Africa

Aim Changing conditions across spatial gradients are primary determinants of biotic regions, local habitats, and distributional edges. We investigate how a climatic gradient and edaphic mosaic interact as multi‐scale drivers of spatial patterns in scarabaeine dung beetles. The patterns are tested for congruency with ecoregion and floral boundaries over the same gradient, as responses to physical factors often differ among higher taxa. Location Southern Africa and the Nama Karoo–Kalahari ecotone, Northern Cape, South Africa. Methods Data consisted of the climatic distributions of 104 species and their abundances at 223 sites in two ecoregions/floral biomes, four bioregions, and 13 vegetation units. Factor analyses determined the biogeographical composition of the species, and regional‐ to local‐scale patterns in species abundance structure. Hierarchical analysis of oblique factors determined the proportional contribution of spatial variance to patterns. One‐way anova was used to test for significant separation of patterns along factor axes. Stepwise multiple regression was used to determine correlations of five physical attributes with species richness, Shannon‐Wiener diversity, and factor loadings for the study sites. Results Four biogeographical influences overlap in the study region, although rank contribution declines from south‐west arid through north‐east savanna to widespread and south‐east highland taxa. Species abundance structure comprises five subregional patterns, two centred to the north‐east (Kalahari, Isolated Kalahari Dune) dominated by Kalahari influence, and three to the south‐west (Nama Karoo subdivisions: Bushmanland, ‘Upper’, ‘Stony Prieska’) dominated by south‐west arid influence. Kalahari deep sands are characterized especially by a warmer, moister climate, whereas the Nama Karoo mosaic of deep or stony soils is characterized especially by north‐west aridity (Bushmanland), south‐east cooler temperatures (‘Upper’), or excessively stony soils (‘Stony Prieska’). Four of the subregional patterns each comprised three localized patterns related primarily to relative stoniness, edge effects from geographical position, or incidence of rainfall. Species richness and diversity declined with decreasing rainfall and increasing stoniness. Main conclusions Climatic and edaphic factors are important multi‐scale determinants of spatial patterns in dung beetle assemblage structure, with edaphic factors becoming more important at local spatial scales. The patterns are roughly congruent with the Kalahari Savanna–Nama Karoo ecotone at the floral biome or ecoregion scale, but show limited coincidence with finer‐scale floral classification.     

Mitochondrial differentiation and biogeography of Hyla meridionalis (Anura: Hylidae): an unusual phylogeographical pattern

Aim To study the patterns of genetic variation and the historical events and processes that influenced the distribution and intraspecific diversity in Hyla meridionalis Boettger, 1874. Location Hyla meridionalis is restricted to the western part of the Mediterranean region. In northern Africa it is present in Tunisia, Algeria and Morocco. In south‐western Europe it is found in the south of France, north‐western Italy and north‐eastern and south‐western Iberian Peninsula. There are also insular populations, as in the Canaries and Menorca. Methods Sampling included 112 individuals from 36 populations covering the range of the species. We used sequences of mitochondrial DNA Cytochrome Oxidase I (COI) for the phylogeographical analysis (841 bp) and COI plus a fragment including part of tRNA lysine, ATP synthase subunits 6 and 8 and part of Cytochrome Oxidase III for phylogenetic analyses (2441 bp). Phylogenetic analyses were performed with paup *4.0b10 (maximum likelihood, maximum parsimony) and Mr Bayes 3.0 (Bayesian analysis). Nested clade analysis was performed using tcs 1.18 and Geo Dis 2.2. A dispersal‐vicariant analysis was performed with diva 1.0 to generate hypotheses about the geographical distribution of ancestors. Results We found little genetic diversity within samples from Morocco, south‐western Europe and the Canary Islands, with three well‐differentiated clades. One is distributed in south‐western Iberia and the High Atlas, Anti‐Atlas and Massa River in Morocco. The second is restricted to the Medium Atlas Mountains. The third one is present in northern Morocco, north‐eastern Iberia, southern France and the Canaries. These three groups are also represented in the nested clade analysis. Sequences from Tunisian specimens are highly divergent from sequences of all other populations, suggesting that the split between the two lineages is ancient. diva analysis suggests that the ancestral distribution of the different lineages was restricted to Africa, and that an explanation of current distribution of the species requires three different dispersal events. Main conclusions Our results support the idea of a very recent colonization of south‐western Europe and the Canary Islands from Morocco. South‐western Europe has been colonized at least twice: once from northern Morocco probably to the Mediterranean coast of France and once from the western coast of Morocco to southern Iberia. Human transport is a likely explanation for at least one of these events. Within Morocco, the pattern of diversity is consistent with a model of mountain refugia during hyperarid periods within the Pleistocene. Evaluation of the phylogenetic relationships of Tunisian haplotypes will require an approach involving the other related hylid taxa in the area.     

Relationship between breeding and wintering ranges in Palaearctic-African migrants

During the northern winter, Palaearctic migrant species are not evenly distributed within sub-Saharan Africa. Species numbers are greatest in a belt of savannah, lying south of the Sahara, and decline southwards. For any one latitude, species numbers are also greater in the east of Africa than in the west. Only about 3% of 187 species winter exclusively south of the equator, but other species migrate from north to south during the course of the northern winter.
For 62 Palaearctic species which winter entirely in Africa, the areas of breeding and wintering ranges are strongly correlated. With some exceptions, species with the largest breeding ranges also have the largest wintering ranges. However, in 69% of species, the breeding range is larger than the wintering range, whereas in 31% of species the wintering range is larger. On average, the wintering ranges of 57 landbird species cover about two-thirds the area of their breeding ranges, and in many species only parts of the wintering range may be occupied at any one time. This implies that the per area carrying capacity of African wintering areas is greater than that of Eurasian breeding areas.
The general correlation between the sizes of breeding and wintering ranges may have its basis in ecology, with generalists able to occupy wider areas than specialists in both breeding and winter quarters. At the same time, the correlation may result partly from an effect of numbers on range size, in that species which have a wide range at one time of the year may then achieve large numbers which occupy a wide range at the other time of year.     

Diversity, endemism and species turnover of millipedes within the south-western Australian global biodiversity hotspot

Aim To examine how current and historical environmental gradients affect patterns of millipede (Diplopoda) endemism and species turnover in a global hotspot of floristic diversity, and to identify regions of high endemism and taxonomic distinctness for conservation management. Location South‐western Australia. Methods Museum database records of millipedes (subclasses Pentazonia and Helminthomorpha), supplemented with extensive fieldwork, were used to map species richness, species turnover (β‐diversity), weighted endemism, average taxonomic distinctness and variation in taxonomic distinctness in half‐degree grid squares (c. 2500 km²). Generalized linear models were used to examine relationships between these parameters with rainfall (present day and historical), topography and human disturbance (clearing for agriculture and urbanization). Results Millipede species richness, particularly within the order Spirostreptida, and millipede endemism were positively associated with large within‐cell differences in elevation (mountainous regions). Large variation in taxonomic distinctness (unevenness in the taxonomic tree) in higher‐rainfall areas was mainly due to speciation within the Spirostreptida genus Atelomastix. Hotspots of millipede endemism and taxonomic distinctness were identified within three categories of importance: primary (Stirling Range East, Cape Le Grand, Cape Arid, Walpole, Porongurups), secondary (Mount Manypeaks, Bremer Bay, Stirling Range West, Duke of Orleans Bay, Ravensthorpe, Albany, Busselton) and tertiary (Nornalup). A species turnover boundary was positively associated with rainfall, broadly located in the transition zone of 300–600 mm year^?1. Main conclusions The current lack of knowledge on the endemism of invertebrates hampers their incorporation into conservation planning. With this knowledge we can identify global biodiversity hotspots and, at a smaller scale, significant conservation areas within a region. Here we have shown that weighted endemism and taxonomic distinctness are useful tools in identifying centres of high endemism and speciation for millipedes within the south‐west Australian hotspot. Moreover, it is unlikely that either vertebrates or vascular plants will be useful surrogates for identifying significant areas for invertebrate conservation. While other workers have shown that vascular plants, mammals and frogs have different centres of endemism within south‐west Australia, our results show that centres of endemism for millipedes encompass all of these plus other areas.     

Growth and reproductive biology of the foxfish Bodianus frenchii,a very long‐lived and monandric protogynous hermaphroditic labrid

Samples of the foxfish Bodianus frenchii, collected over reefs on the lower west and south coasts of Western Australia, contained individuals ranging up to 78 years old. Although B. frenchii is far smaller than many other species within the Labridae, its maximum age is the greatest yet recorded for this highly speciose family and, together with Achoerodus gouldii, provides an example of a temperate hypsigenyine with exceptional longevity. Length and age compositions of females and males and the histological characteristics of gonads of a wide length range of individuals demonstrated that B. frenchii is a protogynous hermaphrodite. Furthermore, as, on both coasts, the length of the smallest male was greater than that at which all females had become mature, B. frenchii is a monandric protogynous hermaphrodite, i.e. all of its males are derived from functional females. Attainment of maturity by females is related more to length than age, whereas the reverse is true for sex change. On the basis of Schnute growth equations and length‐to‐body mass regression equations, the predicted length at age and body mass at length of fish on the south coast were greater than those on the west coast throughout life. Although B. frenchii spawns daily during the main spawning season, which extends from October to February on both coasts, its fecundity at any given length is substantially greater on the south than on the west coast. The more rapid growth of juveniles and earlier attainment of maturity by B. frenchii on the south coast than on the warmer west coast, together with maturation at a similar size on both coasts, run counter to the trends observed in many species and certain ecological theories regarding the relationships between life‐cycle traits and latitude and temperature. The attainment by B. frenchii of a larger body length at age, of greater body mass at length and of greater fecundity at both length and body mass in fish on the south than on the west coast strongly suggests that conditions on the former, cooler coast are more favourable for this labrid, which belongs to a sub‐genus whose other species typically live in cool, deep, temperate waters.     

Squamate richness in the Brazilian Cerrado and its environmental–climatic associations

We investigate patterns of species richness of squamates (lizards, snakes, and amphisbaenians) in the Brazilian Cerrado, identifying areas of particularly high richness, and testing predictions of large‐scale richness hypotheses by analysing the relationship between species richness and environmental climatic variables. We used point localities from museum collections to produce maps of the predicted distributions for 237 Cerrado squamate species, using niche‐modelling techniques. We superimposed distributions of all species on a composite map, depicting richness across the ecosystem. Then, we performed a multiple regression analysis using eigenvector‐based spatial filtering (Principal Coordinate of Neighbour Matrices) to assess environmental–climatic variables that are best predictors of species richness. We found that the environmental–climatic and spatial filters multiple regression model explained 78% of the variation in Cerrado squamate richness (r² = 0.78; F = 32.66; P < 0.01). Best predictors of species richness were: annual precipitation, precipitation seasonality, altitude, net primary productivity, and precipitation during the driest quarter. A model selection approach revealed that several mechanisms related to the different diversity hypothesis might work together to explain richness variation in the Cerrado. Areas of higher species richness in Cerrado were located mainly in the south‐west, north, extreme east, and scattered areas in the north‐west portions of the biome. Partitioning of energy among species, habitat differentiation, and tolerance to variable environments may be the primary ecological factors determining variation in squamate richness across the Cerrado. High richness areas in northern Cerrado, predicted by our models, are still poorly sampled, and biological surveys are warranted in that region. The south‐western region of the Cerrado exhibits high species richness and is also undergoing high levels of deforestation. Therefore, maintenance of existing reserves, establishment of ecological corridors among reserves, and creation of new reserves are urgently needed to ensure conservation of species in these areas.