Archipelago Scale Research Articles

Scale-dependent influences of environmental, historical, and spatial processes on taxonomic and functional beta diversity of Japanese bat assemblages.

This study investigated the relative influences of environmental, spatial, and historical factors, including the island-specific history of land connectivity, on bat assemblages in the Japanese Archipelago. We collected bat distribution data from 1408 studies and assigned them to Japan's First Standard Grid (approximately 6400 km2). Japanese bat assemblages were analyzed at two scales: the entire Japanese Archipelago comprised 16 islands and exclusively the four main islands. At first, we calculated taxonomic and functional total beta diversity (β total) by Jaccard pairwise dissimilarity and then divided this into turnover (β repl) and richness-difference (β rich) components. We conducted hierarchical clustering of taxonomic beta diversity to examine the influence of the two representative sea straits, Tsugaru and Tokara, which are considered biogeographical borders. Variation partitioning was conducted to evaluate the relative effects of the three factors on the beta diversity. Clustering revealed that the Tokara Strait bordered the two major clades; however, the Tsugaru Strait did not act as a biogeographical border for bats. In the variation partitioning, shared fraction between spatial and historical factors significantly explained taxonomic and functional β total and taxonomic β repl at the entire archipelago scale, but not at the four main islands scale extending only Tsugaru Strait but not Tokara Strait. Pure environmental factors significantly explained functional β total at both scales and taxonomic β total only at the four main islands scale. These results suggest that spatial and historical factors are more pronounced in biogeographical borders, primarily structuring assemblage composition at the entire archipelago scale, especially in taxonomic dimension. However, current environmental factors primarily shape the assemblage composition of Japanese bats at the main island scale. The difference in results between the two scales highlights that the primary processes governing assemblages of both dimensions depend on the quality of the dispersal barriers between terrestrial and aquatic barriers for bats.

Weak phylogenetic effect on specialist plant assemblages and their persistence on habitat islands

AbstractAimThe influence of species phylogenetic relatedness on the formation of insular assemblages remains understudied in functional island biogeography, especially for terrestrial habitat islands (i.e. distinct habitat patches embedded in a matrix that differ in the prevailing environmental conditions). Here, we tested three eco‐evolutionary hypotheses: (1) functional specialization of species (i.e. specialism) is associated with phylogenetic clustering at the habitat archipelago scale, (2) such clustering increases with insularity at the habitat island scale and (3) traits indicative of effective local persistence strategies shape island specialism.LocationTerrestrial habitat islands, Europe (Fens in the Western Carpathians, Outcrops in Moravia and Mountaintops in the Cantabrian Range).TaxonAngiosperms.MethodsWe assessed the phylogenetic relatedness of habitat specialists in three different archipelagos composed of terrestrial habitat islands based on phylogenetic signals and phylogenetic diversity (PD) measures. We estimated the effect of insularity on PD using linear models and the effect of persistence traits on specialism using phylogenetic logistic regressions.ResultsOur hypotheses were largely not supported. Outcrop and mountaintop specialist assemblages did not exhibit any phylogenetic structuring, whereas fen specialists were clustered at the archipelago scale. Therefore, insularity seems not to act as a selective force for phylogenetic structure, and ecologically important persistence traits do not operate as precursors of specialism.Main ConclusionsOur results show that species phylogenetic relatedness plays a minor role in shaping habitat island specialist assemblages. Furthermore, the effects of phylogenetic relatedness on assemblages of island specialists are system and scale dependent. Finally, accounting for species' phylogenetic relatedness on persistence traits yielded results similar to previous studies, which corroborates the positive relationship between insularity and functional traits (indicative of enhanced plant persistence abilities with increasing within‐archipelago insularity).

Economic mapping and assessment of Cymodocea nodosa meadows as nursery grounds for commercially important fish species. A case study in the Canary Islands

Cymodocea nodosa seagrass meadows provide several socio-economically ecosystem services, including nurseries for numerous species of commercial interest. These seagrasses are experiencing a worldwide decline, with global loss rates approaching 5% per year, mainly related to coastal human activities. Cymodocea nodosa, the predominant seagrass in the Canary Archipelago (Spain), is also exposed to these threats, which could lead to habitat loss or even local disappearance. In this case study, we estimated the potential economic value of Cymodocea nodosa seagrass meadows for local fisheries at an archipelago scale. Habitat suitability maps were constructed using MAXENT 3.4.1, a software for modelling species distributions by applying a maximum entropy machine-learning method, from a set of environmental variables and presence and background records extracted from historical cartographies. This model allows characterising and assessing the C. nodosa habitat suitability, overcoming the implicit complexity derived from seasonal changes in this species highly dynamic meadows and using it as a first step for the mapping and assessment of ecosystem services. In a second step, value transfer methodologies were used, along with published economic valuations of commercially-interesting fish species related to C. nodosa meadows. We estimate that the potential monetary value of these species can add up to more than 3 million euros per year for the entire Archipelago. The simplicity of the proposed methodology facilitates its repeatability in other similar regions, using freely available data and hence, being suitable for data-scarce scenarios.

Flightlessness in insects enhances diversification and determines assemblage structure across whole communities.

Dispersal limitation has been recurrently suggested to shape both macroecological patterns and microevolutionary processes within invertebrates. However, because of potential interactions among biological, environmental, temporal, and spatial variables, causal links among flight-related traits, diversification and spatial patterns of community assembly remain elusive. Integrating genetic variation within species across whole insect assemblages, within a simplified spatial and environmental framework, can be used to reduce the impact of these potentially confounding variables. Here, we used standardized sampling and mitochondrial DNA sequencing for a whole-community characterization of the beetle fauna inhabiting a singular forested habitat (laurel forest) within an oceanic archipelago setting (Canary Islands). The spatial structure of species assemblages together with species-level genetic diversity was compared at the archipelago and island scales for 104 winged and 110 wingless beetle lineages. We found that wingless beetle lineages have: (i) smaller range sizes at the archipelago scale, (ii) lower representation in younger island communities, (iii) stronger population genetic structure, and (iv) greater spatial structuring of species assemblages between and within islands. Our results reveal that dispersal limitation is a fundamental trait driving diversity patterns at multiple hierarchical levels by promoting spatial diversification and affecting the spatial configuration of entire assemblages at both island and archipelago scales.

Large-scale mapping of live corals to guide reef conservation

Coral is the life-form that underpins the habitat of most tropical reef ecosystems, thereby supporting biological diversity throughout the marine realm. Coral reefs are undergoing rapid change from ocean warming and nearshore human activities, compromising a myriad of services provided to societies including coastal protection, fishing, and cultural practices. In the face of these challenges, large-scale operational mapping of live coral cover within and across reef ecosystems could provide more opportunities to address reef protection, resilience, and restoration at broad management- and policy-relevant scales. We developed an airborne mapping approach combining laser-guided imaging spectroscopy and deep learning models to quantify, at a large archipelago scale, the geographic distribution of live corals to 16-m water depth throughout the main Hawaiian islands. Airborne estimates of live coral cover were highly correlated with field-based estimates of live coral cover (R2 = 0.94). Our maps were used to assess the relative condition of reefs based on live coral, and to identify potential coral refugia in the face of human-driven stressors, including marine heat waves. Geospatial modeling revealed that water depth, wave power, and nearshore development accounted for the majority (>60%) of live coral cover variation, but other human-driven factors were also important. Mapped interisland and intraisland variation in live coral location improves our understanding of reef geography and its human impacts, thereby guiding environmental management for reef resiliency.

Fishing, environment, and the erosion of a population portfolio

AbstractMany organisms exhibit tremendous fluctuations in population abundance and experience unexpected collapse. Conservationists seeking to minimize region‐wide variability in resources and reduce extinction risk often seek to preserve a metapopulation portfolio of spatially asynchronous subpopulations connected by dispersal. However, portfolio properties are not necessarily static, and the erosion of a portfolio can fundamentally alter the population dynamics and services a species provides. In the Northeast Pacific, a portfolio of spatially asynchronous herring populations has historically provided regional reliability of herring to mobile predators and commercial fishermen as well as local subsistence and ceremonial harvest. Here, we fit a mechanistic time‐series model to herring spawn and catch records from 1950 to 2015 to quantify how population growth, climate, and fishing have contributed to a major shift in the herring portfolio over time. We document the erosion of the herring portfolio and a severe decline in herring population growth. Commercial harvest historically played a key role in herring dynamics, hovering around typical annual exploitation rates (15%) at the archipelago scale, but local harvest rates were much higher when fishing occurred (as high as 65%). Additionally, the Pacific Decadal Oscillation and population growth had equally strong effects on local and regional herring population dynamics. Our results highlight how spatially structured populations can undergo major shifts following disturbance and emphasize how ecological systems do not always rapidly recover and provide services following disturbance. Developing herring management strategies at a finer scale may ensure greater regional resource reliability by recovering previous levels of spatial population asynchrony. However, doing so may require higher implementation and monitoring costs in order to yield higher ecological, social, and economic benefits. Such place‐based solutions that match the spatial scale of governance to the spatial scale of ecological dynamics have the potential to improve future management and conservation in an increasingly dynamic world.

Understanding biological and socioeconomic tradeoffs of marine reserve planning via a flexible integer linear programming approach

Analyzing tradeoffs among ecological, economic, and management goals with respect to marine reserve network design is an important facet of systematic conservation planning. We designed an integer linear programming model to quantify tradeoffs among five marine reserve network aspects: ecological conservation value, economic opportunity cost, geographic domain size, total reserve area, and reserve spatial compactness. Using ecological and economic data from the Hawaiian deepwater bottomfish fishery as a case study, an integer linear programming model was designed to choose areas that 1) maximize conservation value and 2) minimize opportunity cost, defined as foregone fisheries revenue. Compromise solutions that equally weighted conservation value and opportunity cost resulted in solutions with dramatically lower foregone fisheries revenue and a relatively small loss in conservation value compared to solutions with the maximum conservation value. When opportunity cost was assumed uniform across the spatial domain, solutions had considerably higher foregone revenue for a given level of conservation value, highlighting the drawback of not including a spatially explicit metric of opportunity cost in reserve selection models. Inclusion of only indicator species, rather than the entire species complex, in the optimization led to considerable representation gaps in conservation value for non-included species. We found that optimizations performed at the archipelago scale provided geographically disproportionate reserve allocations and thus disproportionate conservation benefits and socioeconomic impacts across geopolitically distinct island regions. We showed how reserve selection models can be used to support systematic conservation planning exercises characterized by many diverse and conflicting objectives and parties.

Establishing archipelagic landscape ecological network with full connectivity at dual spatial scales

An archipelagic landscape ecological network (ALEN) is important for optimizing the bird migration route and providing reference for island biodiversity maintenance. The ALEN with full connectivity at dual spatial scales was established in this study. Ecological nodes and routes at archipelago scale, as well as ecological sources and corridors at island scale, were identified to constitute the ALEN based on the habitat suitability and integrating the approaches of Euclidean distance and Least cost distance. Then, the landscape structures of the ALEN were analyzed and the optimization measures were proposed to improve the habitat quality and network connectivity. Miaodao Archipelago, an important and typical archipelago in North China, was selected as the study area. Results indicated that the ALEN was composed of 32 nodes and 81 routes at archipelago scale, and 42 sources and 76 corridors at island scale. At archipelago scale, the ecological efficiency of the ALEN was determined by the landscape resistance and position of different islands, and the landscape structure optimization should focus on the islands that involved the indispensable routes. At island scale, the number and spatial distribution of the sources, as well as the landscape structure of the corridors, contributed the most to the ecological efficiency. More ecological sources with uniform distribution should be added, and the corridor width of 50 m was practical to implement the ALEN. The ALEN realized the full connectivity, that is, the bird species could flow (flight or walk) all over the nodes and sources along the routes and corridors across the dual spatial scales, and the balance of efficiency and cost was achieved. The ALEN could be widely applied in different archipelagos with distinct ecological functions and external disturbances.

How repeatable is microevolution on islands? Patterns of dispersal and colonization-related plant traits in a phylogeographical context.

Archipelagos provide a valuable framework for investigating phenotypic evolution under different levels of geographical isolation. Here, we analysed two co-distributed, widespread plant lineages to examine if incipient island differentiation follows parallel patterns of variation in traits related to dispersal and colonization. Twenty-one populations of two anemochorous Canarian endemics, Kleinia neriifolia and Periploca laevigata, were sampled to represent mainland congeners and two contrasting exposures across all the main islands. Leaf size, seed size and dispersability (estimated as diaspore terminal velocity) were characterized in each population. For comparison, dispersability was also measured in four additional anemochorous island species. Plastid DNA data were used to infer genetic structure and to reconstruct the phylogeographical pattern of our focal species. In both lineages, mainland-island phenotypic divergence probably started within a similar time frame (i.e. Plio-Pleistocene). Island colonization implied parallel increases in leaf size and dispersability, but seed size showed opposite patterns of variation between Kleinia and Periploca species pairs. Furthermore, dispersability in our focal species was low when compared with other island plants, mostly due to large diaspore sizes. At the archipelago scale, island exposure explained a significant variation in leaf size across islands, but not in dispersability or seed size. Combined analyses of genetic and phenotypic data revealed two consistent patterns: (1) extensive within-island but very limited among-island dispersal, and (2) recurrent phenotypic differentiation between older (central) and younger (peripheral) island populations. Leaf size follows a more predictable pattern than dispersability, which is affected by stochastic shifts in seed size. Increased dispersability is associated with high population connectivity at the island scale, but does not preclude allopatric divergence among islands. In sum, phenotypic convergent patterns between species suggest a major role of selection, but deviating traits also indicate the potential contribution of random processes, particularly on peripheral islands.

Small size does not restrain frugivory and seed dispersal across the evolutionary radiation of Galápagos lava lizards.

Frugivory in lizards is often assumed to be constrained by body size; only large individuals are considered capable of consuming fruits, with the potential of acting as seed dispersers. However, only one previous study has tested the correlation of frugivory with body and head size at an archipelago scale across closely related species. All nine lava lizards (Microlophus spp.) were studied on the eleven largest Galápagos islands from 2010 to 2016 to investigate whether frugivory is related to body and head size. We also tested whether fruit abundance influences fruit consumption and explored the effect of seed ingestion on seedling emergence time and percentage. Our results showed that across islands, lava lizards varied considerably in size (64–102 mm in mean snout–vent length) and level of frugivory (1–23%, i.e., percentage of droppings with seeds). However, level of frugivory was only weakly affected by size as fruit consumption was also common among small lizards. Lava lizards consumed fruits throughout the year and factors other than fruit abundance may be more important drivers of fruit selection (e.g., fruit size, energy content of pulp). From 2,530 droppings, 1,714 seeds of at least 61 plant species were identified, 76% of the species being native to the Galápagos. Most seeds (91%) showed no external structural damage. Seedling emergence time (44 versus 118 days) and percentage (20% versus 12%) were enhanced for lizard-ingested seeds compared to control (uningested) fruits. De-pulping by lizards (i.e., removal of pulp with potential germination inhibitors) might increase the chances that at least some seeds find suitable recruitment conditions. We concluded that lizards are important seed dispersers throughout the year and across the whole archipelago, regardless of body size.

Regional forcing explains local species diversity and turnover on tropical islands

AbstractAimTo determine the role of regional forcing on plot‐level species diversity and composition, and to quantify the relative importance of biogeographical and climatic factors in explaining woody plant diversity and composition at the local‐, island‐ and archipelago‐scale.LocationForty‐one tropical islands of the Indo‐Pacific region from Madagascar to Hawai‘i Island.MethodsWe analysed the diversity and composition of tropical woody plant communities located across 113 plots, 41 islands and 19 archipelagos. We used generalized linear mixed‐effects models and generalized dissimilarity models to determine the role of regional forcing at the island and archipelago scale and to assess the relative importance of biogeographical (area and isolation of islands or archipelagos, geographical distance between plots) and climatic factors in explaining differences in local diversity and composition (species turnover). Analyses were conducted at different geographical scales (local, island and archipelago) and taxonomic levels (species, genus and family).ResultsVariation in local (plot‐level) diversity (as species density, the number of species per 100 woody plants) was primarily explained by island and archipelago identity. Maximum species density was positively correlated with the area of an island (or archipelago) and negatively correlated with the isolation of an archipelago. Local climatic variability was also a significant predictor of species density, but less important than regional forcing. Climate variables explained < 20% of the variation in species turnover across all plots. The importance of geographical distance between plots relative to climate in driving species turnover decreased from the species to family level, and from the regional to island level.Main conclusionsRegional forcing was the key driver of local diversity and composition on islands. Island area and archipelago isolation are likely driving local diversity through their effects on the pool of island species. Geographical distance between plots is the main factor explaining species turnover, while at higher taxonomic levels, climatic factors and niche conservatism are the main drivers.

Evaluation on island ecological vulnerability and its spatial heterogeneity

The evaluation on island ecological vulnerability (IEV) can help reveal the comprehensive characteristics of the island ecosystem and provide reference for controlling human activities on islands. An IEV evaluation model which reflects the land–sea dual features, natural and anthropogenic attributes, and spatial heterogeneity of the island ecosystem was established, and the southern islands of Miaodao Archipelago in North China were taken as the study area. The IEV, its spatial heterogeneity, and its sensitivities to the evaluation elements were analyzed. Results indicated that the IEV was in status of mild vulnerability in the archipelago scale, and population pressure, ecosystem productivity, environmental quality, landscape pattern, and economic development were the sensitive elements. The IEV showed significant spatial heterogeneities both in land and surrounding waters sub-ecosystems. Construction scale control, optimization of development allocation, improvement of exploitation methods, and reasonable ecological construction are important measures to control the IEV.

Assessment of island beach erosion due to sea level rise: the case of the Aegean archipelago (Eastern Mediterranean)

Abstract. The present contribution constitutes the first comprehensive attempt to (a) record the spatial characteristics of the beaches of the Aegean archipelago (Greece), a critical resource for both the local and national economy, and (b) provide a rapid assessment of the impacts of the long-term and episodic sea level rise (SLR) under different scenarios. Spatial information and other attributes (e.g., presence of coastal protection works and backshore development) of the beaches of the 58 largest islands of the archipelago were obtained on the basis of remote-sensed images available on the web. Ranges of SLR-induced beach retreats under different morphological, sedimentological and hydrodynamic forcing, and SLR scenarios were estimated using suitable ensembles of cross-shore (1-D) morphodynamic models. These ranges, combined with empirically derived estimations of wave run-up induced flooding, were then compared with the recorded maximum beach widths to provide ranges of retreat/erosion and flooding at the archipelago scale. The spatial information shows that the Aegean pocket beaches may be particularly vulnerable to mean sea level rise (MSLR) and episodic SLRs due to (i) their narrow widths (about 59 % of the beaches have maximum widths < 20 m), (ii) their limited terrestrial sediment supply, (iii) the substantial coastal development and (iv) the limited existing coastal protection. Modeling results indeed project severe impacts under mean and episodic SLRs, which by 2100 could be devastating. For example, under MSLR of 0.5 m – representative concentration pathway (RCP) 4.5 of the Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate change (IPCC) – a storm-induced sea level rise of 0.6 m is projected to result in a complete erosion of between 31 and 88 % of all beaches (29–87 % of beaches are currently fronting coastal infrastructure and assets), at least temporarily. Our results suggest a very considerable risk which will require significant effort, financial resources and policies/regulation in order to protect/maintain the critical economic resource of the Aegean archipelago.

Quality Trait Signature at Archipelago Scale of the Cultured Pearls Produced by the Black-Lipped Pearl Oyster (Pinctada margaritiferaVar.cumingi) in French Polynesia

ABSTRACT In French Polynesia, the aquaculture of Pinctada margaritifera (Linnaeus, 1758) covers a large maritime exploitation area, spread over nearly 20 degrees latitude and longitude, with numerous pearl farms located in three archipelagos (Gambier, Society, and Tuamotu). As these archipelagos have specific seasonal temperature patterns each year, pearl oysters are subject to disparate and contrasting environmental regimes. This study aimed to examine the specificity of commercial pearl quality traits (n = 2,236 samples) at the archipelago scale, in such a way as to provide preliminary data to design the most appropriate strategy for the distribution of hatchery-produced phenotypes. A large and standardized grafting experiment using the same donor phenotype was designed and carried out over six grow-out locations, covering the three archipelagos. Results revealed significant differences in commercial pearl quality traits among archipelagos, giving these groups of growing sites distinctive “signatures”: ...

Using species abundance distribution models and diversity indices for biogeographical analyses

We examine whether Species Abundance Distribution models (SADs) and diversity indices can describe how species colonization status influences species community assembly on oceanic islands. Our hypothesis is that, because of the lack of source-sink dynamics at the archipelago scale, Single Island Endemics (SIEs), i.e. endemic species restricted to only one island, should be represented by few rare species and consequently have abundance patterns that differ from those of more widespread species. To test our hypothesis, we used arthropod data from the Azorean archipelago (North Atlantic). We divided the species into three colonization categories: SIEs, archipelagic endemics (AZEs, present in at least two islands) and native non-endemics (NATs). For each category, we modelled rank-abundance plots using both the geometric series and the Gambin model, a measure of distributional amplitude. We also calculated Shannon entropy and Buzas and Gibson's evenness. We show that the slopes of the regression lines modelling SADs were significantly higher for SIEs, which indicates a relative predominance of a few highly abundant species and a lack of rare species, which also depresses diversity indices. This may be a consequence of two factors: (i) some forest specialist SIEs may be at advantage over other, less adapted species; (ii) the entire populations of SIEs are by definition concentrated on a single island, without possibility for inter-island source-sink dynamics; hence all populations must have a minimum number of individuals to survive natural, often unpredictable, fluctuations. These findings are supported by higher values of the α parameter of the Gambin mode for SIEs. In contrast, AZEs and NATs had lower regression slopes, lower α but higher diversity indices, resulting from their widespread distribution over several islands. We conclude that these differences in the SAD models and diversity indices demonstrate that the study of these metrics is useful for biogeographical purposes.

Island specificities matter: cat diet differs significantly between islands of a major breeding archipelago for a vulnerable endemic seabird

Domestic cats are one of the most widespread predators on islands worldwide and are responsible for numerous reductions and extinctions of species on islands. The three main islands of the Hyeres Archipelago house one of the largest colonies of the Mediterranean endemic Yelkouan shearwater Puffinus yelkouan that has recently been up-listed by the IUCN to ‘vulnerable’. The main objectives of this study were to assess the diet of cats and to study the effect of cat predation on Yelkouan shearwater populations at the archipelago scale. The diet of cats was studied using scat analyses according to years and seasons for each island. Simultaneously, Yelkouan shearwater breeding success was monitored during a period of 8 years on Port-Cros and Porquerolles, and 3 years on Le Levant. Descriptive analyses and GLM were used to compare data gathered on each island. At the archipelago scale, cats preyed strongly upon introduced mammals and shearwaters. Surprisingly, large differences appeared in cats’ diet according to the island considered. The Yelkouan shearwater was the primary prey of cats on Le Levant, but secondary on Port-Cros and Porquerolles. Cat predation was mainly concentrated during the shearwater prospecting period, when birds arrive at the colonies and look for a mate (if they are not already paired) and a burrow before breeding. Consequently cat impact was low on shearwater breeding success. However, this study demonstrates that the cat management conducted on Port-Cros was positive for fledging success. The successful cat eradication on Port-Cros supports the need to continue working for Yelkouan shearwater conservation with Le Levant as a priority, because this is where the colonies are largest and predation on Yelkouan shearwaters is very high.

Trends of the seagrass <em>Cymodocea nodosa</em> (Magnoliophyta) in the Canary Islands: population changes in the last two decades

Seagrass meadows perform essential ecosystem functions and services. Though the meadows are globally deteriorating, numerous regressions remain unreported as a result of data fragmentation. Cymodocea nodosa is the most important seagrass in shallow coastal waters of the Canary Islands. No study has so far investigated temporal population trends at the entire archipelago scale. Using data collected in the past 23 years by local companies, public authorities and research groups, the population trends of Cymodocea nodosa were analysed over the past two decades at the scales of islands, island sectors and meadows. During this period, a prevalence of negative trends was revealed for three seagrass demographic descriptors (seagrass shoot density, coverage and leaf length) at the three scales, evidencing an overall deterioration in seagrass meadow integrity. These results suggest the need to develop correct management strategies to guarantee the conservation of this seagrass and the meadows it creates.

From sampling stations to archipelagos: investigating aspects of the assemblage of insular biota

ABSTRACTAim To investigate the formation of nestedness and species co‐occurrence patterns at the local (sampling station), the intermediate (island group), and the archipelago scale.Location The study used data on the distribution of terrestrial isopods on 20 islands of the central Aegean (Greece). These islands are assigned to two distinct subgroups (Kyklades and Eastern islands).Methods The Nestedness Temperature Calculator was used to obtain nestedness values and maximally nested matrices, the EcoSim7 software and a modified version of Sanderson (2000) method were used for the analysis of species co‐occurrences. Idiosyncratic temperatures of species and the order of species placement in the maximally nested matrices were used for further comparisons among spatial scales. The relationships of nestedness values with beta‐diversity, habitat diversity and a number of ecological factors recorded for each sampling station were also investigated.Results Significant nestedness was found at all spatial scales. Levels of nestedness were not related to beta‐diversity or habitat diversity. Nestedness values were similar among spatial scales, but they were affected by matrix size. The species that contributed most to the nested patterns within single islands were not the same as those that produce nestedness at the archipelago scale. There was significant variation in the frequency of species occurrence among islands and among spatial scales. There was no direct effect of ecological factors on the shaping of patterns of nestedness within individual islands, but habitat heterogeneity was crucial for the existence of such patterns. Positive associations among species prevailed at all scales when species per station were considered, while negative associations prevailed in the species per island matrices. All associations resulted from the habitat structure of sampling stations and from particularities of geographical distributions.Conclusions There was no clear‐cut distinction between nestedness patterns among spatial scales, even though different species, and partially different factors, contributed to the formation of these patterns in each case. There was a core of species that contributed to the formation of nested patterns at all spatial scales, while the patterns of species associations suggested that biotic interactions are not an important causal factor. The results of this study suggest that locally rare species cannot be widespread at a higher spatial scale, while locally common species can have a restricted distribution.