Butterfly biogeography and endemism on tropical Pacific islands

Butterfly distributions on 26 tropical Pacific archipelagos were analysed to examine the effects of geography on diversity and endemism. The total butterfly fauna for each archipelago was divided into continental (found also on continental areas), Pacific (found within more than one archipelago but not outside of the study area), and endemic species (restricted to a single archipelago). Numbers and proportions of each species were related to eight geographic variables by stepwise multiple linear regression analysis. Total area of an archipelago and distance from other land masses were important predictors of the number of species within an archipelago. Proportions of butterfly species in each category were related differently to the geographic variables, with endemism being promoted by the number of large islands within an archipelago. Relative to birds, butterflies have been less successful in colonizing remote archipelagos and have much lower levels of endemism. Even if colonization is successful, butterfly speciation may be constrained by the mechanics of coevolution with available host plants.     

Biogeography of mammals on tropical Pacific islands

Aim We examine the influence of geography on species richness and endemism of mammals on tropical Pacific archipelagos to determine the importance of intra‐ and inter‐archipelago speciation in promoting local and regional species richness. Location Thirty tropical Pacific archipelagos. Methods A distributional list of mammals on 30 archipelagos was compiled, and values for 10 geographical variables were estimated for each archipelago. Mammal species were placed in three different categories (continental, Pacific and endemic) based on their distribution. The total number of species and numbers of species within each category were related to the geographical variables using Poisson regression analysis. Results Species richness was related positively to variables describing land area, numbers of large islands and elevation; and negatively to variables describing isolation. Levels of endemism did not differ between volant and non‐volant species, but differed between mega‐ and microchiropterans. Main conclusions Variation in species richness of mammals in the tropical Pacific region can be accounted for by a combination of intra‐archipelago speciation within archipelagos composed of large islands, and inter‐archipelago speciation, particularly among more isolated archipelagos. Mammals were less widely distributed throughout the study area than previously found for butterflies, skinks or birds. However, the level of endemism was similar to that of skinks and birds on the same archipelagos, and was higher than that of butterflies.     

Dispersal and speciation of skinks among archipelagos in the tropical Pacific Ocean

Summary We examined the potential effects of geography on the distribution and speciation of skinks on tropical Pacific archipelagos. The entire tropical Pacific skink fauna was divided into continental (found also in continental areas), Pacific (endemic to the study area but found within more than one archipelago) and endemic (found within only one archipelago) species categories. The number and proportion of skinks within each species category were determined for each of the 27 archipelagos in the study area. Nine geographic variables reflecting archipelago size, isolation and elevation were estimated for each archipelago. Principal components analysis was used to reduce the nine variables to three uncorrelated composite variables that were interpreted as representing archipelago size, isolation and elevation. Numbers and proportions of skinks in each category within an archipelago were related to the composite geographic variables using multiple linear regression analysis. Archipelago size and isolation were important predictors of both skink diversity and endemism. Results were then compared to diversity and endemism of birds within the study area. Skinks showed an archipelago-wide level of endemism similar to that of birds. On an archipelago by archipelago basis, however, large differences between birds and skinks were evident. In particular, the New Caledonia skink fauna was much more endemic than that of birds. The bird faunas of Hawaii and the Marquesas were nearly completely endemic, while no endemic skinks occurred in these two archipelagos. These differences presumably reflect the relative dispersal powers of skinks and birds and, consequently, rates of colonization and speciation. Differences may also be due partly to morphological conservatism among isolated skink populations and the occurrence of cryptic species that have not yet been identified as separate species. The discovery of such cryptic species, however, is unlikely to increase the endemic skink fauna of Hawaii and other distant archipelagos to a level commensurate with that of birds. Differences in endemism between skinks and birds may also be due to unknown local ecological interactions.     

Determinants of bird species richness,endemism, and island network roles in Wallacea and the West Indies: is geography sufficient or does current and historical climate matter?

Island biogeography has greatly contributed to our understanding of the processes determining species' distributions. Previous research has focused on the effects of island geography (i.e., island area, elevation, and isolation) and current climate as drivers of island species richness and endemism. Here, we evaluate the potential additional effects of historical climate on breeding land bird richness and endemism in Wallacea and the West Indies. Furthermore, on the basis of species distributions, we identify island biogeographical network roles and examine their association with geography, current and historical climate, and bird richness/endemism. We found that island geography, especially island area but also isolation and elevation, largely explained the variation in island species richness and endemism. Current and historical climate only added marginally to our understanding of the distribution of species on islands, and this was idiosyncratic to each archipelago. In the West Indies, endemic richness was slightly reduced on islands with historically unstable climates; weak support for the opposite was found in Wallacea. In both archipelagos, large islands with many endemics and situated far from other large islands had high importance for the linkage within modules, indicating that these islands potentially act as speciation pumps and source islands for surrounding smaller islands within the module and, thus, define the biogeographical modules. Large islands situated far from the mainland and/or with a high number of nonendemics acted as links between modules. Additionally, in Wallacea, but not in the West Indies, climatically unstable islands tended to interlink biogeographical modules. The weak and idiosyncratic effect of historical climate on island richness, endemism, and network roles indicates that historical climate had little effects on extinction‐immigration dynamics. This is in contrast to the strong effect of historical climate observed on the mainland, possibly because surrounding oceans buffer against strong climate oscillations and because geography is a strong determinant of island richness, endemism and network roles.     

Contemporary geography dominates butterfly diversity gradients within the Aegean archipelago (Lepidoptera: Papilionoidea, Hesperioidea)

Aim We compare the influence of contemporary geography and historical influences on butterfly diversity for islands in the Aegean archipelago. Location The Aegean archipelago (Greece) and two islands (Cyprus and Megisti) in the Levantine Sea. Methods Thirty‐one islands were examined. Data are taken from own surveys (Coutsis and Olivier) and from the literature. Stepwise multiple regression is used to determine relationships between species richness, frequency, rarity and endemicity against potential geographical predictors. Stepwise logit regression is used to determine geographical predictors of species incidence on islands. Inter‐island and inter‐species associations have been examined using multivariate ordination and clustering techniques. Results The Aegean butterfly fauna is characterized by decreasing diversity and rarity, and increasing homogeneity, from the periphery to the present geographical centre of the archipelago (Cyclades). Diversity and rarity are shown to relate closely to species richness, and species richness, in turn, is largely explained by contemporary geography, particularly the degree of isolation from the nearest mainland sources of Greece or Turkey, and island dimensions. Islands towards the centre of the archipelago are characterized by a group of mobile species (n ≥ 20 species) with extensive ranges across Europe; species that would have recolonized Santorini (Thira) following the VI6 eruption there c. 1630 bc . Endemic components, indicative of autochthonous evolutionary events, are few (5% of species are endemic) compared to known sedentary organisms (molluscs and isopods), but exceed those for more mobile animals (i.e. birds); their distribution is mainly confined to large isolated islands along the Aegean arc (i.e. Kriti) and in the Dodecanese group. Main conclusions Contemporary geography, i.e. processes currently operating in ecological time, dominates butterfly diversity gradients (species richness, frequency, rarity and incidence) in the archipelago. Two reasons are suggested to account for the lack of endemism and the pattern of decreasing diversity into the Cyclades. First, relict butterfly elements may have become extinct on all but a few larger islands, particularly from environmental changes since the Neolithic (fire and overgrazing). Second, colonization from the continental landmasses is ongoing with more mobile species transferring even to the most isolated islands.     

Not as the crow flies: assessing effective isolation for island biogeographical analysis

Aim To evaluate the role of island isolation in explaining the distribution of vascular plant species in a dense freshwater archipelago, specifically comparing conventional measures of island isolation with landscape measures of island isolation. Location Data were collected from 35 islands within Massasauga Provincial Park on the eastern shores of the Georgian Bay, Ontario, Canada. Methods Sampled islands were located using stratified random selection based on location and size variation. The number of species was recorded along stratified random transects. Island isolation variables included distance to the mainland, distance to the nearest island, largest gap in a stepping‐stone sequence, distance to the closest upwind point of land, and a landscape measure of island isolation. The landscape measure of isolation was quantified as the percentage of the land area within 100, 250, 500, 1000, 1500 and 2000 m of each island’s perimeter. The isolation variables were calculated within a geographical information system (GIS). Dependent variables in the regression analyses included species richness, the logarithm of species richness and residuals of the species–area relationship. Independent variables included island isolation variables and their logarithmic transformations. Results Isolation plays a role, albeit small, in explaining species richness in the study area. In the regression analyses, the landscape measure of isolation provided a better fit than conventional measures of island isolation. Islands with less land than water within a 250‐m buffer were more effectively isolated and had fewer species present than islands surrounded by a greater proportion of water. Main conclusions Consistent with the species–isolation relationship, fewer species were present on more isolated islands within the Massasauga study area, as elucidated using a series of island buffers in a GIS. Applying a landscape measure of isolation to similar dense, freshwater archipelagos may elucidate species–isolation patterns not evident through conventional, straight‐line distance measurements of island isolation. The low value of the regression coefficients as well as the isolation history and high density of the Massasauga islands suggests caution in extending the results, especially to dissimilar archipelagos.     

Island area and species diversity in the southwest Pacific Ocean: is the lizard fauna of Vanuatu depauperate?

One island group suggested to be an exception to the species–area Relationship is the Vanuatu Archipelago, a group of 13 large and 80 small islands in the southwest Pacific Ocean. To test the hypothesis that the lizard fauna of the Vanuatu Archipelago does not meet the predictions of the species-area relationship, and thus is depauperate, we compare diversity among several island groups in the southwest Pacific: Fiji, the Loyalty Islands, New Caledonia, Samoa, the Solomon Islands, Tonga, and Vanuatu. We found that the lizard diversity of Vanuatu meets the pattern of diversity predicted by the species-area relationship. The Solomon Islands, the largest and least isolated oceanic archipelago considered, has the greatest species diversity and endemism of the oceanic islands. Inclusion or exclusion of island groups based on factors such as geologic history or faunal source affects the strength of the relationship between diversity, area, and history of emergence, and influences perceptions of diversity within individual archipelagos. In addition to island size, factors such evolutionary time scale, speciation, and archipelago complexity influence species richness on islands.     

Floristic biogeography of the Hawaiian Islands: influences of area, environment and paleogeography

Aim A detailed database of distributions and phylogenetic relationships of native Hawaiian flowering plant species is used to weigh the relative influences of environmental and historical factors on species numbers and endemism. Location The Hawaiian Islands are isolated in the North Pacific Ocean nearly 4000 km from the nearest continent and nearly as distant from the closest high islands, the Marquesas. The range of island sizes, environments, and geological histories within an extremely isolated archipelago make the Hawaiian Islands an ideal system in which to study spatial variation in species distributions and diversity. Because the biota is derived from colonization followed by extensive speciation, the role of evolution in shaping the regional species assemblage can be readily examined. Methods For whole islands and regions of each major habitat, species–area relationships were assessed. Residuals of species–area relationships were subjected to correlation analysis with measures of endemism, isolation, elevation and island age. Putative groups of descendents of each colonist from outside the Hawaiian Islands were considered phylogenetic lineages whose distributions were included in analyses. Results The species–area relationship is a prominent pattern among islands and among regions of each given habitat. Species number in each case correlates positively with number of endemics, number of lineages and number of species per lineage. For mesic and wet habitat regions, island age is more influential than area on species numbers, with older islands having more species, more single‐island endemics, and higher species : lineage ratios than their areas alone would predict. Main conclusions Because species numbers and endemism are closely tied to speciation in the Hawaiian flora, particularly in the most species‐rich phylogenetic lineages, individual islands’ histories are central in shaping their biota. The Maui Nui complex of islands (Maui, Moloka‘i, Lāna‘i and Kaho‘olawe), which formed a single large landmass during most of its history, is best viewed in terms of either the age or area of the complex as a whole, rather than the individual islands existing today.     

Endemic regions of the vascular flora of the peninsula of Baja California,Mexico

Question: Can we recognize areas of high endemism and high endemic richness, using data from collections, and what are the ecological variables that best explain these areas? Location: Peninsula of Baja California, Mexico. Methods: We analysed the distribution of 723 endemic vascular plants species along the peninsula of Baja California and neighbouring islands distributed in 218 cartographic cells 15’ x 20’ in size. By means of a residual analysis, we identified areas of significantly high endemic species richness, and we calculated the degree of endemicity (or rarity) in each cell by giving to each species a weight factor inversely proportional to the land area it covers. Results: Nine regions of high‐endemicity and/or high endemic species richness were found. Discussion and conclusions: The analyses of rarity and endemic species richness showed two contrasting scenarios: High endemicity values in oceanic and sky islands accounts for a high number of species with a restricted distribution, promoted most likely by genetic isolation and high environmental heterogeneity. High endemic richness along the peninsular coast is related to ecotonal transition along vegetation types. After correcting for collection effort (i.e. the number of specimens collected within a cell), we found the phytogeographic region and altitudinal heterogeneity to be the variables that best predicted endemic richness. Both high endemism and high endemic richness have distinct geographic patterns within our study region. The nine endemic regions provide elements for priority definitions in future conservation programs.     

Species richness and structure of ant communities in a dynamic archipelago: effects of island area and age

Aim To assess how ant species richness and structure of ant communities are influenced by island age (disturbance history) in a dynamic archipelago. Location Cabra Corral dam, Salta Province, north‐west Argentina (25°08′ S, 65°20′ W). Methods Ant species richness on remaining fragments (islands) of a flooded forest was determined, as well as island area, isolation and age. Simple linear regressions were performed to assess relationships between ant species richness and those insular variables. Furthermore, a stepwise multiple linear regression analysis was conducted in order to determine the relative influence of each insular variable on ant species richness. Islands were categorized in two age classes (old and young) and co‐occurrence analyses were applied within each class to evaluate changes in community structure because of interspecific competition. Results Simple regression analyses indicated a moderate, positive effect of island area on ant species richness. Weak, marginally non‐significant relationships were found between ant species richness and both island isolation and island age, showing the tendency for there to be a decrease in ant species richness with island isolation and that ant species richness might be higher in old islands. The multiple regression analysis indicated that island isolation and age had no significant effects on the number of ant species, island area being the only independent variable retained in the analysis. On the contrary, whereas a random pattern of species co‐occurrence was found on young islands, ant communities in old islands showed a significantly negative pattern of species co‐occurrence, suggesting that the effect of competition on community structure was stronger on older islands than on younger islands. Main conclusions Island area was the most important variable explaining ant species richness on the islands of Cabra Corral dam. However, both island isolation and island age (or disturbance history) might also contribute to shape the observed community patterns. The present study also shows that island age significantly affects the strength with which interspecific interactions structure ant communities on islands.     

Genetic evaluation of marine biogeographical barriers: perspectives from two widespread Indo-Pacific snappers (Lutjanus kasmira and Lutjanus fulvus)

Aim In the Indo‐Pacific, the mass of islands of the Indonesian archipelago constitute a major biogeographical barrier (the Indo‐Pacific Barrier, IPB) separating the Pacific and Indian oceans. Evidence for other, more localized barriers include high rates of endemism at the Marquesas and other isolated peripheral islands in the Pacific. Here we use mitochondrial‐sequence comparisons to evaluate the efficacy of biogeographical barriers on populations of the snappers Lutjanus kasmira and Lutjanus fulvus across their natural ranges. Location Pacific and Indian oceans. Methods Mitochondrial cytochrome b sequence data were obtained from 370 individuals of L. kasmira and 203 individuals of L. fulvus collected from across each species’ range. Allele frequency data for two nuclear introns were collected from L. kasmira. Phylogenetic and population‐level analyses were used to determine patterns of population structure in these species and to identify barriers to dispersal. Results Lutjanus kasmira lacks genetic structure across the IPB and throughout 12,000 km of its central Indo‐Pacific range. In contrast, L. fulvus demonstrates high levels of population structure at all geographical scales. In both species, highly significant population structure results primarily from the phylogenetic distinctiveness of their Marquesas Islands populations (L. kasmira, d = 0.50–0.53%; L. fulvus, d = 0.87–1.50%). Coalescence analyses of the L. kasmira data indicate that populations at opposite ends of its range (western Indian Ocean and the Marquesas) are the oldest. Coalescence analyses for L. fulvus are less robust but also indicate colonization from the Indian to the Pacific Ocean. Main conclusions The IPB does not act as a biogeographical barrier to L. kasmira, and, in L. fulvus, its effects are no stronger than isolating mechanisms elsewhere. Both species demonstrate a strong genetic break at the Marquesas. Population divergence and high endemism in that archipelago may be a product of geographical isolation enhanced by oceanographic currents that limit gene flow to and from those islands, and adaptation to unusual ecological conditions. Lutjanus kasmira shows evidence of Pleistocene population expansion throughout the Indo‐central Pacific that originated in the western Indian Ocean rather than the Marquesas, further demonstrating a strong barrier at the latter location.     

Inventaire des moustiques (Diptera : Culicidae) des îles du sud-ouest de l’océan Indien,Madagascar excepté — Une revue critique

Inventory of the mosquitoes (Diptera: Culicidae) of the islands of southwestern Indian Ocean, Madagascar excluded – A Critical Review. The biodiversity of mosquitoes in the islands of southwestern Indian Ocean is the concern of numerous publications. Here, we propose a synthetic inventory and the analysis of the mosquito diversity, based on the available literature. A comprehensive annotated checklist of mosquito species has been recently published on Madagascar; this is the reason why this land is excluded from our work. Studied area encompasses 28 tropical islands in the southern hemisphere: 4 islands in the Comoros archipelago, 5 Scattered Islands (îles Éparses), 5 in Mascarene, 11 in the Seychelles and 3 in the Chagos archipelago. In total, the mosquito list presents 73 valid species, of which 10 are Anophelinae and 63 Culicinae. The number of species that are distributed in these islands only is 19, i.e. 26%, which is a remarkable level for endemism. The richness in mosquito species in these islands is analysed through several aspects including geography, local speciation and natural or human dissemination. This updated inventory increases by 33% the number of known species by regard to the previous inventory published by Julvez & Mouchet in 1994. The historical responsibility of humans in the introduction of new mosquito species in these islands is strongly documented. For instance, the species with the highest distribution among islands are Aedes aegypti, Ae. albopictus and Culex quinquefasciatus. The islands belong to the afrotropical biogeographic area and, logically, the majority (63%) of mosquito species present phylogenetic affinities with continental Africa and/or Madagascar; interestingly, the number of species present in these islands and in Madagascar but absent in continental Africa is higher than the number of species present in these islands and in continental Africa but absent in Madagascar (respectively 12 and 2 species). Thanks to valuable increase in the sampling effort, our knowledge of the culicidian fauna is increasing in these islands that constitute indisputably hotspots of biodiversity.     

Climate drives community‐wide divergence within species over a limited spatial scale: evidence from an oceanic island

Geographic isolation substantially contributes to species endemism on oceanic islands when speciation involves the colonisation of a new island. However, less is understood about the drivers of speciation within islands. What is lacking is a general understanding of the geographic scale of gene flow limitation within islands, and thus the spatial scale and drivers of geographical speciation within insular contexts. Using a community of beetle species, we show that when dispersal ability and climate tolerance are restricted, microclimatic variation over distances of only a few kilometres can maintain strong geographic isolation extending back several millions of years. Further to this, we demonstrate congruent diversification with gene flow across species, mediated by Quaternary climate oscillations that have facilitated a dynamic of isolation and secondary contact. The unprecedented scale of parallel species responses to a common environmental driver for evolutionary change has profound consequences for understanding past and future species responses to climate variation.     

How island size and isolation affect bee and wasp ensembles on small tropical islands: a case study from Kepulauan Seribu,Indonesia

Aim To study the effects of isolation and size of small tropical islands on species assemblages of bees (superfamily Apoidea) and wasps (superfamily Vespoidea). Location Twenty islands in the Kepulauan Seribu Archipelago off the coast of west Java, Indonesia. The size of surveyed islands ranges between 0.75 and 41.32 ha; their distance from the coast of Java varies between 3 and 62 km. Methods Field work was conducted from February to May 2005. Bees and wasps were caught with a sweep net during sampling units of 15 min, continuing until four consecutive samples revealed no new species. Total species richness was quantified by the estimators Chao 2, first‐order jackknife and Michaelis–Menten. The software binmatnest was used to test for nestedness of species assemblages. Similarities of species composition between islands were quantified by Sørensen’s similarity index. Results Eighty‐two species were recorded on the 20 surveyed islands. Species richness declined with increasing isolation of islands from the source area, Java. Although the size of the largest island exceeded that of the smallest island by a factor of almost 60, island size only very weakly affected species richness of bees; no effect of island size was found for wasps. Mean body size of species decreased with increasing island isolation. Nestedness of island faunas was only weakly developed. Species composition of both superfamilies was affected by island isolation, but not by island size. Main conclusions While the species–isolation relationship on the very small islands of Kepulauan Seribu followed the prediction of MacArthur and Wilson’s equilibrium theory, the absence of a species–area relationship indicated a weak ‘small‐island effect’, at least in wasps. The combination of an only weakly developed pattern of nested species subsets, the shift in species compositions and the decline of mean body size with increasing island isolation from the source area indicates that biotic interactions and different species traits contribute to the shaping of communities of bees and wasps within the archipelago. The potential of biotic interactions for generating distribution patterns of species within the archipelago is also emphasized by the observed restriction of some species with apparently high dispersal abilities to outer islands.     

An integrative species delimitation approach reveals fine-scale endemism and substantial unrecognized avian diversity in the Philippine Archipelago

The Philippine archipelago is recognized as a biodiversity hotspot because of its high levels of endemism and numerous threatened species. Avian lineages in the Philippines feature morphologically distinct allopatric taxa, which have been variably treated either as species or subspecies depending on species concepts and recognition criteria. To understand how alternative species limits would alter diversity metrics and patterns of endemism in the Philippines, we selected 19 focal lineages of birds, each containing multiple described taxa within the Mindanao Island Group. We delimited species in an integrative, lineage-based framework using three operational criteria: species must (1) form well-supported, geographically circumscribed clades, (2) be monophyletic with significant genetic differentiation identified by a coalescent model, and (3) feature fixed differences in phenotypic characters. Our criteria identified 40 species from the original 19 focal lineages, a 50–74% increase over recent comprehensive taxonomic treatments. Genetic criteria in isolation identified an additional 10 populations that could be cryptic species in need of further study. We identified fine-scale endemism within the Mindanao Island Group, with multiple unrecognized avian endemics restricted to Samar/Leyte, Bohol Island, and the Zamboanga Peninsula. Genetic and phenotypic information support the hypothesis that polytypic bird species in the Philippines tend to be composed of evolutionarily distinct, range-restricted, allopatric replacements rather than widespread and variable “superspecies”. We conclude that lack of species recognition has resulted in underestimates of species diversity and overlooked fine-scale endemism in the Philippines. Recognizing this diversity would alter conservation priorities, shifting efforts to protect microendemics on smaller islands and finer scale endemic areas within larger islands.     

舟山群岛蝶类群落嵌套分布格局及其影响因素

片断化生境中群落的物种组成常呈现嵌套分布格局。2013年7-8月, 我们在浙江舟山群岛采用截线法对28个岛屿上的蝴蝶群落进行了野外调查, 探讨了岛屿物种嵌套分布格局及其影响因素。通过测量采集标本获得蝶类的生活史特征(最小需求面积、翅展和体重), 查阅文献资料获得蝶类的栖息地特征(岛屿面积、距最近大陆距离和距最近大岛距离), 分析了影响蝶类群落嵌套结构的因素。研究结果显示: (1)舟山群岛蝶类群落符合嵌套分布格局; (2)岛屿面积和物种最小需求面积对嵌套格局的形成有显著影响; (3)舟山群岛蝶类群落嵌套格局的形成支持选择性灭绝假说; (4)随机检验零模型结果显示该嵌套分布格局并非采样偏差造成的。因此, 在制定舟山群岛区域蝶类保护措施时, 应优先考虑那些分布在面积较大岛屿的和最小需求面积较大的物种。     

Peculiarities of island endemism in Lasiocampidae (Lepidoptera)

The distinctive features of island endemism in Lasiocampidae are extensive adaptive radiation and vicariance. These phenomena are especially typical of archipelagoes, where different taxa occur on closely located islands. The diverse ecological conditions of the islands determine the presence of a great number of subordinate taxa. Owing to the reduced flight activity, the species of Lasiocampidae are characterized by a high level of endemism on isolated islands. Islands of continental origin (Sumatra, Borneo) have a lower level of endemism, with centers on the mountain ranges. The endemism level is much higher in islands of biogenic origin (the Philippines). The Pacific islands have a higher index of specific endemism. At the same time, the islands of mixed and oceanic origin have low species diversity by a large fraction of endemic taxa. The boundaries of species distribution are determined not only by the landscape or climatic factors but also by biological ones.     

Are island‐like systems biologically similar to islands? A review of the evidence

Islands are geographically defined as land masses completely surrounded by water, and island systems have been used as models for many biogeographic, ecological, and evolutionary theories ever since Darwin's pioneering efforts. However, their biological definition is complex. Over the past few decades these theories have been applied to many study systems that only share some geographic features with island systems. These features include spatial fragmentation, limited area, spatial and temporal isolation from adjacent parts of the system, and low connectivity between different parts within the system, to mention just a few. These systems vary in their form, the matrix that surrounds them, the factors defining their borders, the extent of insularity they impose on the different taxa, and their geological similarity to different types of actual islands. Here, I seek to understand whether such island‐like systems (ILS) function biologically as true islands. In the first part, I describe the wide diversity of ILS suggested in the literature and the variation in the features that define their insularity. In the second part, I review the extent to which the main theories of island biology are applicable to these systems: species–area and species–isolation relationships, community composition, evolutionary radiations, and the extent of endemism and genetic diversity. In the third and final part, I suggest a new conceptual framework within which to classify and study the biology of ILS, as well as practical future research directions. I conclude that the term ‘biological island’ is a multi‐faceted concept, loosely related to its geographical definition. As ILS are often less isolated than true islands, and their biological patterns are only partly similar to those of true islands (and even this is true only for some ILS) the use of the term ‘island’ to describe any isolated habitat is therefore inappropriate.     

Nested distribution of amphibians in the Zhoushan archipelago, China: can selective extinction cause nested subsets of species?

Because of their poor dispersal ability, amphibians are well suited for testing the selective extinction theory on islands. Amphibian fauna in the Zhoushan archipelago, China, exhibit a high level of nestedness (C = 0.893), and the species number is lower on islands than on similar sized areas on the mainland. No correlation was found between island-specific species richness and the nearest distance from a larger island, distance from the mainland or density of human population. These results suggest that no amphibian colonisation has occurred in the archipelago since island isolation 7000–9000 years ago. Furthermore, the results imply that selective extinction contributes to the nestedness of amphibians in the Zhoushan archipelago. The incidence of a species on the islands is significantly correlated with log area of the smallest island occupied by the species and the number of provinces on the Chinese mainland in which the species occur. However, there is no correlation with average body length of adults and island occurrence. It is concluded that (1) the area of the smallest island occupied by a species is a good estimate of the minimum area for a viable population of the species and a good predictor of species incidence on islands, (2) species with a restricted distribution range are more vulnerable to extinction from islands than those with a wide distribution range and (3) the effect of body size on occurrence on the islands is uncertain, and may be specific to the archipelago and taxa studied. The observed nestedness of amphibian assemblages has two implications for conservation: (1) not only can all the species found in several small reserves be found on a large reserve of the same total size, but additional species can be found on the single large reserve; (2) for a reserve to maintain viable populations of all species in a region it should be at least as large as the smallest island occupied by the most vulnerable species. Received: 16 December 1996 / Accepted: 22 September 1997