Genetic differentiation and secondary contact zone in the seagrass Cymodocea nodosa across the Mediterranean–Atlantic transition region

Aim A central question in evolutionary ecology is the nature of environmental barriers that can limit gene flow and induce population genetic divergence, a first step towards speciation. Here we study the geographical barrier constituted by the transition zone between the Atlantic Ocean and the Mediterranean Sea, using as our model Cymodocea nodosa, a seagrass distributed throughout the Mediterranean and in the Atlantic, from central Portugal to Mauritania. We also test predictions about the genetic footprints of Pleistocene glaciations. Location The Atlantic–Mediterranean transition region and adjacent areas in the Atlantic (Mauritania to south‐west Portugal) and the Mediterranean. Methods We used eight microsatellite markers to compare 20 seagrass meadows in the Atlantic and 27 meadows in the Mediterranean, focusing on the transition between these basins. Results Populations from these two regions form coherent groups containing several unique, high‐frequency alleles for the Atlantic and for the Mediterranean, with some admixture west of the Almeria–Oran Front (Portugal, south‐west Spain and Morocco). These are populations where only one or a few genotypes were found, for all but Cadiz, but remarkably still show the footprint of a contact zone. This extremely low genotypic richness at the Atlantic northern edge contrasts with the high values (low clonality) at the Atlantic southern edge and in most of the Mediterranean. The most divergent populations are those at the higher temperature range limits: the southernmost Atlantic populations and the easternmost Mediterranean, both potential footprints of vicariance. Main conclusions A biogeographical transition region occurs close to the Almeria–Oran front. A secondary contact zone in Atlantic Iberia and Morocco results from two distinct dispersal sources: the Mediterranean and southernmost Atlantic populations, possibly during warmer interglacial or post‐glacial periods. The presence of high‐frequency diagnostic alleles in present‐day disjunct populations from the southernmost Atlantic region indicates that their separation from all remaining populations is ancient, and suggests an old, stable rear edge.     

Combining projected changes in species richness and composition reveals climate change impacts on coastal Mediterranean fish assemblages

Species Temporal Turnover (STT) is one of the most familiar metrics to assess changes in assemblage composition as a consequence of climate change. However, STT mixes two components in one metric, changes in assemblage composition caused by a process of species loss or gain (i.e. the nestedness component) and changes in assemblage composition caused by a process of species replacement (i.e. the species replacement component). Drawing on previous studies investigating spatial patterns of beta diversity, we propose measures of STT that allow analysing each component (species replacement vs. nestedness), separately. We also present a mapping strategy to simultaneously visualize changes in species richness and assemblage composition. To illustrate our approach, we used the Mediterranean coastal fish fauna as a case study. Using Bioclimatic Envelope Models (BEMs) we first projected the potential future climatic niches of 288 coastal Mediterranean fish species based on a global warming scenario. We then aggregated geographically the species‐level projections to analyse the projected changes in species richness and composition. Our results show that projected changes in assemblage composition are caused by different processes (species replacement vs. nestedness) in several areas of the Mediterranean Sea. In addition, our mapping strategy highlights that the coastal fish fauna in several regions of the Mediterranean Sea could experience a ‘cul‐de‐sac’ effect if exposed to climate warming. Overall, the joint exploration of changes in species richness and composition coupled with the distinction between species replacement and nestedness bears important information for understanding the nature of climate change impacts on biodiversity. These methodological advances should help decision‐makers in prioritizing action in the areas facing the greatest vulnerability to climate.     

Community structure of amphipods on shallow <Emphasis Type="Italic">Posidonia oceanica</Emphasis> meadows off Tunisian coasts

The structure, diversity and spatial distribution of the amphipod fauna associated with Posidonia oceanica meadows were studied along the Tunisian coasts in 2007. Samples were collected in nine different meadows at 2 m depth. A total of 44 species belonging to 12 families were collected. The most common species in terms of abundance were Ampithoe helleri, Hyale camptonyx and Ericthonius punctatus. The highest values of abundance and species richness and the lowest values of diversity and equitability were found in meadows with high epiphyte biomass. Multivariate analyses of data indicated that epiphyte biomass and geographical position were major determinants of the distribution and composition of amphipod assemblages along Tunisian coasts. The presence of two lessepsian amphipod species in one of the southern Posidonia oceanica meadows modified the structure of assemblage.     

Biogeographical patterns of marine larval trematode parasites in two intermediate snail hosts in Europe

Aim We used published inventories of trematodes in Littorina littorea (L.) and Hydrobia ulvae (Pennant) in European seas to search for two basic biogeographical patterns in the spatial occurrence of various trematode species: (1) do parasite distribution and richness patterns in the two host snails overlap with known ecoregions of free‐living organisms; and (2) does trematode species richness in the snails follow latitudinal or longitudinal gradients? Location North East Atlantic. Methods We used multidimensional scaling (MDS), analysis of similarity (ANOSIM) and analysis of variance (ANOVA) to test whether there were overlaps of parasite distribution and richness with known ecoregions of free‐living organisms. In addition, we used linear regression analyses to test whether trematode richness in snails (corrected for sampling effort) was correlated with the latitude or longitude of the sampling sites. Results When corrected for sampling effort, mean trematode species richness per site did not differ among the different ecoregions in L. littorea. In contrast, in H. ulvae, mean species richness was much lower for sites from the Celtic Sea compared with sites from the Baltic Sea and the North Sea. Based on the results of MDS analyses, trematode species composition was distinct among ecoregions; in particular, communities from the Baltic Sea differed markedly from communities in the Celtic Sea, for both snail species. Latitude and longitude were not significantly correlated with parasite species richness in either snail species. Most trematode species had restricted distributions, and only three species in L. littorea and five species in H. ulvae occurred at more than 50% of the sites. Main conclusions There is more structure in the large‐scale distribution of trematodes in gastropods than one would expect from the large‐scale dispersal capabilities of their bird and fish final hosts. We propose mechanisms based both on limited dispersal via fish and bird final hosts and on gradients in environmental factors to explain the observed patterns.     

Community structure and biogeography of shore fishes in the Gulf of Aqaba,Red Sea

Shore fish community structure off the Jordanian Red Sea coast was determined on fringing coral reefs and in a seagrass-dominated bay at 6 m and 12 m depths. A total of 198 fish species belonging to 121 genera and 43 families was recorded. Labridae and Pomacentridae dominated the ichthyofauna in terms of species richness and Pomacentridae were most abundant. Neither diversity nor species richness was correlated to depth. The abundance of fishes was higher at the deep reef slope, due to schooling planktivorous fishes. At 12 m depth abundance of fishes at the seagrass-dominated site was higher than on the coral reefs. Multivariate analysis demonstrated a strong influence on the fish assemblages by depth and benthic habitat. Fish species richness was positively correlated with hard substrate cover and habitat diversity. Abundance of corallivores was positively linked with live hard coral cover. The assemblages of fishes were different on the shallow reef slope, deep reef slope and seagrass meadows. An analysis of the fish fauna showed that the Gulf of Aqaba harbours a higher species richness than previously reported. The comparison with fish communities on other reefs around the Arabian Peninsula and Indian Ocean supported the recognition of an Arabian subprovince within the Indian Ocean. The affinity of the Arabian Gulf ichthyofauna to the Red Sea is not clear. Received in revised form: 2 November 2001 Electronic Publication     

Fish otoliths from the pre-evaporitic (Early Messinian) sediments of northern Italy: their stratigraphic and palaeobiogeographic significance

The study of otolith assemblages from the pre-evaporitic Messinian deposits allows the reconstruction of a fauna of 79 taxa of which 35 could be identified at the specific level. Three of these are new: Diaphus rubus, Myctophum coppa, and Uranoscopus ciabatta. The assemblages reflect mainly a neritic environment influenced by the oceanic realm. Analysis of the global present-day geographic distribution of 42 of the recognised Messinian genera indicates that 88% of these are still living in the Mediterranean, 98% in the Atlantic and 78% in the Indo-Pacific realm. These results are in good agreement with the evolutionary trends documented for the Oligocene and Miocene teleost fauna, specifically an increase in percentage of genera inhabiting the modern Mediterranean, a very high percentage of Atlantic and Indo-Pacific genera, and a slight fall of the importance of present-day Indo-Pacific genera from the Rupelian up to the Late Miocene. Analysing the composition of the Early Messinian fauna at the level of nominal species indicates that about 53% of the species represented in the assemblages are still living in the Recent Mediterranean, and that a significant number of these were already present in the Tortonian. It is interesting that these species are mainly neritic. This seems to confirm that the close affinity of the fossil assemblage with the present-day Mediterranean neritic fauna, which was already recorded at the genus level for the Rupelian fauna, persists during the Neogene and continues until the Pleistocene.     

Biogeographical patterns of algal communities in the Mediterranean Sea: Cystoseira crinita‐dominated assemblages as a case study

Aim The aim of this study was to describe the composition, community structure and biogeographical variation of subtidal algal assemblages dominated by the brown alga Cystoseira crinita across the Mediterranean Sea. Location The Mediterranean coast, from Spain (1°25′ E) to Turkey (30°26′ E). Methods Data on the species composition and structure of assemblages dominated by the species C. crinita were collected from 101 sites in nine regions across the Mediterranean Sea. Multivariate and univariate statistical tools were used to investigate patterns of variation in the composition of the assemblages among sites and regions, and to compare these with previously defined biogeographical regions. Linear regressions of species richness versus longitude and versus latitude were also carried out to test previously formulated hypotheses of biodiversity gradients in the Mediterranean Sea. Results The main features characterizing C. crinita‐dominated assemblages across the Mediterranean included a similar total cover of species, a similar cover of C. crinita, and consistency in the presence of the epiphyte Haliptilon virgatum. Biogeographical variation was detected as shifts in relative abundances of species among regions, partly coinciding with previously described biogeographical sectors. A significant positive correlation was found between species richness and latitude, while no significant correlation was detected between species richness and longitude. Main conclusions The patterns of variation in community structure detected among the studied regions reflected their geographical positions quite well. However, latitude seemed to contribute more to the explanation of biological patterns of diversity than did geographical distances or boundaries, which classically have been used to delimit biogeographical sectors. Moreover, the positive correlation between species richness and latitude reinforced the idea that latitude, and possibly temperature as a related environmental factor, plays a primary role in structuring biogeographical patterns in the Mediterranean Sea. The lack of correlation between species richness and longitude contradicts the notion that there is a decrease in species richness from west to east in the Mediterranean, following the direction of species colonization from the Atlantic.     

Detection of environmental impacts by bottom trawling onPosidonia oceanica (L.) Delile meadows: sensitivity of fish and macroinvertebrate communities

Along the Mediterranean coast,Posidonia oceanica (L.) Delile meadows have a great ecological and economical importance. However, there is a general regression of these meadows due to human activities such as illegal bottom trawling, may be affecting to overall ecosystem health. We examined changes in the community structure of mobile fauna associated withP. oceanica meadows at different spatial scales and taxonomic levels. The aim of this paper was to identify the most efficient taxonomic level to use in environmental impact studies of bottom trawling. At the macroscale level (10 to 100 m), there were significant differences between sites in the densities of some fish species and also the total fish assemblage structure, at both family and species taxonomical levels. At the microscale (0.1 to 1 m), some species of amphipods and isopods showed significant differences in their population densities. In the overall analysis of community structure, the coarse taxonomical levels, such as phyla and class, did not show significant differences, however amphipods and isopods showed significant differences at family and species levels. From these results, both study scales are required to detect changes onPosidonia meadows' fauna. Monitoring of some fish species such asDiplodus annularis (Linnaeus, 1758) and the overall fish assemblage as well as the structure of the amphipod and isopod communities appears to be the most efficient tool in the assessment of environmental impacts by bottom trawling onP. oceanica meadows.     

Projected impacts of climate warming on the functional and phylogenetic components of coastal Mediterranean fish biodiversity

Climate warming affects biodiversity distribution across all ecosystems. However, beyond changes in species richness, impacts on other biodiversity components are still overlooked, particularly in the marine realm. Here we forecasted the potential effect of climate warming on the phylogenetic and functional components of coastal Mediterranean fish biodiversity. To do so, we used species distribution models to project the potential distribution of 230 coastal fish species by the end of the 21st century based on the IPCC A2 scenario implemented with the Mediterranean climatic model NEMOMED8. From these projections, we assessed the changes in phylogenetic (PD) and functional diversity (FD) of fish assemblages at multiple spatial scales using a dated molecular phylogeny and an extensive functional trait database. At the scale of the entire Mediterranean Sea, the projected extinctions of 40 coastal fish species would lead to a concomitant erosion of PD and FD (13.6 and 3%, respectively). However, a null model revealed that species loss at this scale would not lead to a disproportionate erosion of PD and FD. Similar results were found when considering fish assemblages at the grid cell scale. Indeed, at this scale, the projected changes in species richness would lead to unexpected losses of PD and FD for localized and small areas only. A disproportionate erosion of PD under climate warming was only forecasted when analysing fish assemblages at an intermediate spatial scale, namely the Mediterranean marine ecoregions. Overall, our results emphasize the importance of considering multiple spatial scales when assessing potential impacts of climate warming on the multiple components marine biodiversity.     

Fish Assemblages in Different Shallow Water Habitats of the Venice Lagoon

The small-sized fish assemblages of the Venice Lagoon were investigated and compared among five shallow subtidal habitats (seagrass beds, sparsely vegetated habitats, unvegetated sand bottoms, mudflats and saltmarsh creeks) in the Northern lagoon basin. Sampling was carried out seasonally (Spring, Summer and Autumn of 2002) in 4–7 stations for each habitat type, by means of a fine-mesh, small beach seine. Two-way analysis of variance was applied to assess the differences in species richness, fish diversity, density and standing stock amongst habitats, whereas fish assemblage composition was investigated by using multivariate analyses (MDS, ANOSIM, SIMPER). The analyses indicated that seagrass beds and saltmarsh creeks are relevant shallow habitats in structuring the small-sized fish assemblages of the Venice Lagoon, supporting specialized and recognizable fish assemblages. Those in seagrass beds, in particular, were characterized by higher species richness and standing stock with respect to all the others. The structuring role of these habitats was discussed in terms of both habitat complexity and degree of confinement. In contrast, sandy bottoms, mudflats and sparsely vegetated habitats were identified as “transition” habitats, with highly variable fish assemblages, influenced by the contribution of the adjacent habitats, and acting probably as both ‘buffer zones’ between the other habitats and migration routes for many fish species in the lagoon.     

A comparative measure of biodiversity based on species composition

In conservation planning, species richness and species endemism are the most often used metrics for describing the biodiversity importance of areas. However, when it comes to prioritizing regions for conservation actions these measures alone are insufficient because they do not reveal how similar or different the actual composition of species may be from one area to another. For comparative analysis an additional useful metric would be one that indicates the degree to which the species assemblage in one area is also represented in—or is distinct from—species assemblages of other areas. Here we describe a method for quantifying the compositional representativeness of species assemblages among geographic regions. The method generates asymmetric pairwise similarity coefficients that are then used to calculate separate measures for the representativeness and the distinctiveness of species assemblages in the regions being compared. We demonstrate the method by comparing fish communities among freshwater ecoregions of the Mississippi Basin, and then among smaller hydrological units within two individual freshwater ecoregions. At both scales of analysis, our measures of representativeness and distinctiveness reveal patterns of fish species composition that differ from patterns of species richness. This information can enhance conservation planning processes by ensuring that priority-setting explicitly consider the most representative and distinctive species assemblages.     

Increasing southern invasion enhances congruence between endemic and exotic Mediterranean fish fauna

Species movements in relation with global warming may increase the spatial overlap between exotic and endemic species, which is a critical issue for the conservation of biodiversity. The Mediterranean Sea, which is a receptacle for exotic species while being a hotspot for endemism, provides exceptional material for a case study. The aim of our study was to quantify (i) the increasing invasion from southern fish exotic species (Red Sea and Atlantic Ocean) that the Mediterranean biota is experiencing and (ii) the spatial overlap between exotic and endemic Mediterranean fish fauna following the northward movement of exotic species within the Mediterranean Sea in the context of global warming. The historical invasion dynamic of exotic fish species and the sea surface temperature series were reconstructed from 1810 to 2006 in order to estimate the correlation between invasion rate and climate. The geographical distributions of exotic and endemic fish richness before and after the period of global warming were used to assess the dynamic of spatial congruence. The results revealed (i) an acceleration of successful introductions from the Red Sea and (ii) the introduction of Atlantic species from lower latitudes in correlation with the increasing temperature of the Mediterranean Sea. We also showed an increasing overlap between the spatial distributions of endemic and exotic species richness. Taken together, our results suggest that endemic fish species are facing a growing number of exotic species because the Mediterranean Sea is acting as a catchment basin for southern species.     

Diversity and seasonal dynamic of a lizard assemblage in a Neotropical semiarid habitat

We assessed the lizard assemblage from a priority conservation Caatinga area from northeastern Brazil, through a pluriannual ecological approach, to expand the understanding on biodiversity patterns of Neotropical semiarid habitats. The studied area presented one of the richest lizard faunas among Caatinga sites, being composed primarily by species typical from open landscapes. The local species composition was more similar to assemblages from adjacent Caatinga ecoregions than to those from other areas within the same ecoregion. The inventoried lizard assemblage consisted of a few common species and a majority of low abundance ones, and its overall richness and abundance did not differ between rainy and dry months. Our findings demonstrated that the composition of lizard assemblages did not match with the current proposed Caatinga ecoregions, and revealed that the studied assemblage followed a lognormal species-abundance distribution, showing no significant seasonal fluctuation in richness and abundance.     

A baseline for prioritizing the conservation of the threatened seagrass Cymodocea nodosa in the oceanic archipelago of Madeira

Seagrasses are experiencing fragmentation and regression globally; thus, protection and recovery of meadows are a preservation priority. However, conservation actions must consider inherent regional conditions, since certain coastal areas are not suitable for the settlement of extensive meadows. Likewise, small oceanic archipelagos are not always able to fulfil the habitat requirements of seagrass habitats but can harbour small patches that in turn provide unique research opportunities. In this study, we focused on the seagrass Cymodocea nodosa in the archipelago of Madeira (NE Atlantic Ocean). Here we compile historical and contemporary records of this species along with characterization of associated communities (fish and invertebrates). A bionomic map with potentially suitable areas for the establishment and settlement of this species is also included. Lastly, we highlight coastal management and restoration actions and future research directions to preserve this species in Madeira Island.     

Valley‐scale hydrogeomorphology drives river fish assemblage variation in Mongolia

River hydrogeomorphology is a major driver shaping biodiversity and community composition. Here, we examine how hydrogeomorphic heterogeneity expressed by Functional Process Zones (FPZs) in river networks is associated with fish assemblage variation. We examined this association in two distinct ecoregions in Mongolia expected to display different gradients of river network hydrogeomorphic heterogeneity. We delineated FPZs by extracting valley‐scale hydrogeomorphic variables at 10 km sample intervals in forest steppe (FS) and in grassland (G) river networks. We sampled fish assemblages and examined variation associated with changes in gradients of hydrogeomorphology as expressed by the FPZs. Thus, we examined assemblage variation as patterns of occurrence‐ and abundance‐based beta diversities for the taxonomic composition of assemblages and as functional beta diversity. Overall, we delineated 5 and 6 FPZs in river networks of the FS and G, respectively. Eight fish species were found in the FS river network and seventeen in the G, four of them common to both ecoregions. Functional richness was correspondingly higher in the G river network. Variation in the taxonomic composition of assemblages was driven by species turnover and was only significant in the G river network. Abundance‐based taxonomic variation was significant in river networks of both ecoregions, while the functional beta diversity results were inconclusive. We show that valley‐scale hydrogeomorphology is a significant driver of variation in fish assemblages at a macrosystem scale. Both changes in the composition of fish assemblages and the carrying capacity of the river network were driven by valley‐scale hydrogeomorphic variables. River network hydrogeomorphology as accounted for in the study has, therefore, the potential to inform macrosystem scale community ecology research and conservation efforts.     

Shifts in the composition and distribution of Pacific Arctic larval fish assemblages in response to rapid ecosystem change

The Pacific Arctic marine ecosystem has undergone rapid changes in recent years due to ocean warming, sea ice loss, and increased northward transport of Pacific-origin waters into the Arctic. These climate-mediated changes have been linked to range shifts of juvenile and adult subarctic (boreal) and Arctic fish populations, though it is unclear whether distributional changes are also occurring during the early life stages. We analyzed larval fish abundance and distribution data sampled in late summer from 2010 to 2019 in two interconnected Pacific Arctic ecosystems: the northern Bering Sea and Chukchi Sea, to determine whether recent warming and loss of sea ice has restricted habitat for Arctic species and altered larval fish assemblage composition from Arctic- to boreal-associated taxa. Multivariate analyses revealed the presence of three distinct multi-species assemblages across all years: (1) a boreal assemblage dominated by yellowfin sole (Limanda aspera), capelin (Mallotus catervarius), and walleye pollock (Gadus chalcogrammus); (2) an Arctic assemblage composed of Arctic cod (Boreogadus saida) and other common Arctic species; and (3) a mixed assemblage composed of the dominant species from the other two assemblages. We found that the wind- and current-driven northward advection of warmer, subarctic waters and the unprecedented low-ice conditions observed in the northern Bering and Chukchi seas beginning in 2017 and persisting into 2018 and 2019 have precipitated community-wide shifts, with the boreal larval fish assemblage expanding northward and offshore and the Arctic assemblage retreating poleward. We conclude that Arctic warming is most significantly driving changes in abundance at the leading and trailing edges of the Chukchi Sea larval fish community as boreal species increase in abundance and Arctic species decline. Our analyses document how quickly larval fish assemblages respond to environmental change and reveal that the impacts of Arctic borealization on fish community composition spans multiple life stages over large spatial scales.     

Assessing lake typologies and indicator fish species for Italian natural lakes using past fish richness and assemblages

In order to establish a fish-based typology of Italian lakes and identify possible reference and indicator fish species for each lake type, we analysed historical data on fish assemblages of all Italian natural lakes >0.5 km² from the period prior to the major decline in water quality in the 1950s. General linear regression models showed the ecoregion and lake altitude being the best predictors of fish species richness. The number of species was significantly higher in the Alpine than in the Mediterranean ecoregion. Among Alpine lakes, the number of fish species increased significantly with lake volume whilst decreased with altitude. In the Mediterranean lakes, none of the selected parameters was significant. Cluster analysis of fish assemblages (presence/absence) divided the lakes of the Alpine and Mediterranean ecoregions into four and two types, respectively. Pike (Esox lucius), rudd (Scardinius erythrophthalmus) and tench (Tinca tinca) were the main indicator species for the small and mostly shallow lakes in both the Alpine (Type 1) and Mediterranean (Type 6) ecoregions, minnow (Phoxinus phoxinus) for the alpine high altitude lakes (Type 2) and landlocked shad (Alosa fallax lacustris), European whitefish (Coregonus lavaretus) and burbot (Lota lota) for the large and very deep alpine lakes (Type 4). The European whitefish was the only indicator species for the deep Mediterranean lakes (Type 5). These species and associated fish assemblages may be useful indicators in future assessments of the ecological status of Italian lakes according to the European Directives (2000/60/EC and 2008/105/EC).     

Lake depth and geographical position modify lake fish assemblages of the European 'Central Plains' ecoregion

1. Classification of European lake fish assemblages can be based on fish‐assemblage structure or morphological, geographical, physical and chemical lake attributes. However, substantial gaps in knowledge exist with respect to the correspondence between both classification approaches. 2. Here, we compiled fish assemblage data from 165 lakes situated in the European ‘Central Plains’ ecoregion. Cluster analysis of fish abundances was performed to compare fish assemblage types of the entire ecoregion with those from previous country‐specific studies. Nonparametric group comparisons, classification trees and partial canonical ordinations were used to infer the correspondence between fish assemblage types and morphology, geographical position and nutrient concentration of the lakes. 3. Three distinct fish assemblages were revealed: vendace (Coregonus albula), ruffe (Gymnocephalus cernuus) and roach (Rutilus rutilus) lake types. Both latitude and lake depth were the best determinants of lake type, but total phosphorus (TP) concentrations were also important. Vendace lakes were deep and had low TP concentrations, whereas the shallower ruffe and roach lakes had higher TP values. Roach lakes were more frequent in the north‐west area of the ecoregion, whereas ruffe lakes were more often found south of the Baltic Sea. 4. Controlling for the influence of nutrient concentration showed that lake morphology and geographical position were important determinants of fish assemblages. However, the variance explained was low (<20%), implying that biological interactions may also be important in forming the lake‐specific fish assemblages. 5. The results suggest that fish assemblages differ between deep and shallow lakes, and between the north‐west and south‐east locations within the Central Plains ecoregion. Accordingly, establishment of depth‐related lake morphotypes is needed, and the European ecoregions recommended to be used in evaluation systems according to the Water Framework Directive seem to be too coarse to reflect the subtle differences of fish species richness along geographical gradients.