One of the recurrent themes in historical biogeography relates to the units of analysis, their definition and identification. Although areas of endemism are usually accepted as the standard units of study, other units have been suggested, as well as several methods for identifying them. There is no consensus on which units are best suited for the studies; however, the effect of the units and area schemes on analytical results is acknowledged. Here, I review the literature on biogeographic units, their uses and recommendations, highlighting the relevance of the use of more than one area-classification scheme in empirical studies.

Conservation biogeography involves the application of biogeographical principles and methods to conservation issues, including the design of protected areas. Bioregionalisation has been central in the implementation of main global conservation strategies, providing the basis for prioritising protected areas and evaluating their representativeness and effectiveness in conservation actions. Traditionally, experts established these bioregionalisations without repeatable methodologies and using only qualitative evidence, which has set constraints in their usefulness. We compared three descriptive bioregionalisations commonly used for conservation decision-making, with a regionalisation produced using quantitative methods (endemicity analysis), so as to assess biases and differences in the representativeness of the existing protected-area system of Argentina. Areas of endemism were detected using NDM/VNDM quantitative methodology on a database consisting of 19 250 distribution records of 116 taxa of snakes, and the results were compared with previous descriptive regionalisations. We recovered 9 quantitative bioregionalisation units (QBU) v. 6–8 descriptive bioregionalisation units (DBU) proposed by previous authors. From this comparison, the following was found: (1) we discovered three new QBU not considered by any previous DBU; (2) other three areas proposed by DBU are not supported by our endemicity analysis; (3) we detected differences comparing the representativeness of protected areas between descriptive v. quantitative bioregionalisations, leaving the first, some areas of conservation relevance largely unprotected. Moreover, DBU were characterised by a high degree of uncertainty and biases, such as the consideration of probably artificial units, the non-recognition of some natural units and mistakes in the representativeness of protected areas. We emphasise the importance of applying quantitative biogeographic methods to identify bioregionalisation units and its fundamental role in conservation biogeography so as to optimise protected-area efficiency and other territorial conservation strategies.

In the present study, we measured spatiotemporal properties of ecological niches of amphibians in China and tested the relative importance of various niche-diversity metrics for explaining the evolutionary distinctiveness-weighted extinction risk (EDGE) of amphibian species. We applied the hierarchical partitioning technique on the phylogenetically independent contrasts of the niche covariates and EDGE of amphibians, for the purpose of removing the influence of evolutionary inertia among species. As a comparison, phylogenetic least-square general regression (PLGS) was also conducted. The results showed that EDGE was high for those amphibian species of China identified as Critically Endangered or Endangered on the IUCN Red List. Niche fragmentation dimension (NFD) and niche position (NP) were the top two predictors across partial correlation analyses, hierarchical variation partitioning, PLGS and multiple regression analyses. Most temporal niche properties were not significantly associated with the EDGE index of amphibians. Variation partitioning analysis showed that the spatial component of niche measures explained the largest proportion of total variation in EDGE (∼31%), whereas the temporal component of niche properties explained ∼8% of the variation. The significantly negative role of NFD and extinction risk of amphibians in China may be attributed to a reduced rescue effect, habitat geometry, and local extinction in species with large and continuous distributional ranges.

Coprosma is perhaps the most ubiquitous plant genus in New Zealand. It belongs to the tribe Anthospermeae, which is distinctive in the family Rubiaceae through its small, simple, wind-pollinated flowers and its southern hemisphere distribution. The tribe comprises four main clades found respectively in South Africa, Africa, Australia and the Pacific. The high level of allopatry among the four subtribes is attributed here to their origin by vicariance. The Pacific clade, subtribe Coprosminae, is widespread around the margins of the South Pacific and also occurs on most of the high islands. Distributions of the main clades in the subtribe are mapped here and are shown to be repeated in other groups. The distribution patterns also coincide with features of regional geology. Large-scale volcanism has persisted in the central Pacific region since at least the Jurassic. At that time, the oldest of the Pacific large igneous provinces, the Shatsky Rise, began to be erupted in the region now occupied by French Polynesia. Large-scale volcanism in the central Pacific continued through the Cretaceous and the Cenozoic. The sustained volcanism, along with details of the clade distributions, both suggest that the Coprosminae have persisted in the central Pacific by survival of metapopulations on individually ephemeral islands. It is also likely that vicariance of metapopulations has taken place, mediated by processes such as the subsidence of the Pacific seafloor by thousands of metres, and rifting of active arcs by transform faults. It is sometimes argued that a vicariance origin is unlikely for groups on young, oceanic islands that have never been connected by continuous land, but metapopulation vicariance does not require physical contact between islands.

We present a study of the endemicity patterns in the Brazilian Atlantic Forest on the basis of the distribution of 107 fly species belonging to 24 genera of 15 families. This is the first picture of endemism for Diptera in the Atlantic Forest. Instead of the traditional grid of geographical coordinates, we used a system of topographic units (TUs) for the analysis, delimited after gathering information on rivers and altitude for each state and country. A parsimony analysis of the data matrix with the species records for the TUs was performed, named topographic-unit parsimony analysis (TUPA). The same distributional data was used in a NDM/VNDM analysis. The combination of the resulting patterns from both analyses indicated the existence of the following three major areas of endemism for flies in the Atlantic Forest: a Northern Atlantic Forest, north of Rio Doce; a Southern Atlantic Forest, south of Rio Doce along the coast, extending to the west and to the south at the level of the state of Paraná; and a Semideciduous Seasonal Forest, west to the ombrophilous forest along the coast. None of these areas seems to be shaped solely by vicariance events. They can possibly be the result of biotic fusion of ancestral areas of endemism as a result of barrier collapse and secondary overlap of sister biotas, a hypothesis yet to be tested. The recognition of a separate area of endemism for flies in the Semideciduous Forest agrees with phytogeographical reconstructions and raises an important alert for the scarcity of biological reserves for this vegetation.

The mammals are the biological group initially analysed by Wallace to define the Neotropical region (NR). Their areas of endemism (Ae) are considered historical patterns, which have been used to describe biogeographic schemes. However, the Ae at regional scale are currently unclear. In the present study, we analyse Ae of mammals at the regional scale and compare them with previous biogeographic schemes of the NR. The Ae of Neotropical terrestrial mammals were identified using the endemicity analysis (software NDM/VNDM). Our results showed that the NR is composed of 10 Ae, supported by 82 endemic taxa (6 families, 29 genera, and 47 species). The Ae showed a NR with multiple boundaries and with a core of higher overlap of the areas of endemism (OAE) from Veracruz and the Pacific coasts of Mexico to the southern limit of Amazonia in Brazil. The NR boundaries vary strikingly with latitude, with substantially more overlapping areas of endemism in the tropical biomes than in the temperate biomes of America. This pattern of OAE is consistent with the higher mammal-species richness zone within the tropical biomes and other biogeographic patterns such as higher productivity and spatial heterogeneity.

Biogeographical transition zones are areas of a complex biotic mixture located at the borders between biogeographical units. Climatic, physical and ecological factors should play an important role in allowing coexistence of different biotic elements in the transition zone. Here, we explore the relationship between environmental factors and biogeographical transition zones, defined by Neotropical mammal distributions, by a model selection approach based on the Akaike information criterion and accounting for the spatial structure in the data. We detected three areas of high overlap between mammalian areas of endemism. Two of them corresponded to the well-established regional-level transition zones, namely Mexican (MTZ) and South American (SATZ) transition zones; the third was one located in south-eastern Brazil, approximately between the Paraná and Chacoan dominion that we call The Atlantic Forest integration zone (AF). Only one explicative variable was shared by the three transitions zones (precipitation of the warmest quarter). However, shared variables with great explanatory power indicated two environmental aspects as facilitators for the coexistence of different biotic components in a given geographical area. The first one was the heterogeneity component, either topographic for the SATZ and MTZ or climatic for the AF. The second one was related non-extreme thermal conditions: precipitation of the warmest quarter, interpreted as a thermal buffer, shared by AF and SATZ, and isothermality shared by MTZ and SATZ.

Time-slicing of areas is a novel biogeographic method that helps resolve conflicting area relationships and assess temporal overlap as an explanation for the conflict. The method differs from others currently popular in biogeography in that it does not date nodes before analysis (e.g. divergence dating) to infer area relationships and classification. Here, time-slicing is used as a proof of concept approach to interpret the inter-relationships of Neogene and Palaeogene biotic areas of Wallacea, a well-studied area of biogeographic overlap between South-East Asia and Australasia. We used 18 Palaeogene and 25 Neogene areas within Wallacea, represented in 28 areagrams from 25 published phylogenetic hypotheses. Areas were delimited using palaeogeographical reconstructions and biotic distribution data. Paralogy-free subtree and transparent methods of analysis were used to find a general area cladogram (GA), which was then compared with palaeogeographical reconstructions. Palaeogene areas formed clades different from those of Neogene areas. Area relationships correlated strongly with palaeogeographical reconstructions of the Neogene and the Palaeogene. The new approach demonstrated that Palaeogene and Neogene areas have distinct biogeographic histories. Wallacea is a temporal, as well as a geographic, composite that lies between two inferred barriers of distribution, namely the Palaeogene Wallace’s line and the Neogene Weber’s line.