Study area

We selected the state of Hidalgo, in central Mexico (Fig 2) as study area because of its high biological diversity, which results from the geologic, climatic and biogeographical heterogeneity. The state covers 20,905 km² and an elevation gradient that ranges from 46 to 3,358 m a.s.l. [27]. For this area Morrone [28] proposed four biogeographical provinces: the Mexican Plateau Province (also known as the Altiplano), which belongs to the Nearctic region, the Gulf of Mexico Province that belongs to the Neotropical region, and two provinces of the Halffter’s Mexican Transition Zone sensu [29]: the Sierra Madre Oriental, and the Trans-Mexican Volcanic Belt.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Study area.

Spatial distribution of the species records used in this work for terrestrial vertebrates in the ecoregions located in the state of Hidalgo, central Mexico.

https://doi.org/10.1371/journal.pone.0160438.g002

We based the analysis of beta diversity in the biotic composition of the ecoregions that occur in the state of Hidalgo. Ecoregions are areas that comprise distinct assemblages of species, with limits that approach the original extent of natural communities before main land use changes [24]. We used the ecoregions delimited by Olson & Dinerstein [25], which correspond to those promoted by the WWF [30] and are used worldwide to evaluate species diversity, endemicity and vulnerability, as well as to identify priority areas for conservation [25, 26].

The ecoregions located in Hidalgo State are (Fig 2): Veracruz Moist Forests (shortened in this work as Moist Forests), Veracruz Montane Forests (Montane Forests), Trans-Mexican Volcanic Belt Pine-Oak Forests (Volcanic Belt), Meseta Central Matorral (Meseta Central), Central Mexican Matorral (Mexican Matorral) and Sierra Madre Oriental Pine-Oak Forests (Sierra Madre).

Databases

We used the databases of terrestrial vertebrates (amphibians, reptiles, birds, and mammals) from the project Fomix Conacyt-Hidalgo “Biological diversity of the state of Hidalgo” first and second stages (http://citnova.gob.mx/diversidad-biologica-del-estado-de-hidalgo-segunda-fase/; S1 File). We updated those databases for taxonomy and nomenclature up to 2014. The included information results from extensive field work from 2006 to 2010 carried out by specialists in each biological group from the UAEH, and can also be found in other publications [31, 32, 33]. We geo-referenced all records and overlapped them with the ecoregions using the ArcGis 10.1 program [34] to get species lists per ecoregion. We used these lists for all analyses in this work.

We used different sources to determine the taxonomic structure of each vertebrate group and the taxonomic levels used for calculating taxonomic beta diversity. For amphibians and reptiles, we used the catalogs of taxonomical authorities of CONABIO [35, 36]; for mammals we followed the classification of Ramírez-Pulido et al. [37]; and for birds we used the information available in the AOU [38] web page and the phylogenetic arrangement of Stefan Hintsche [39].

Ethics Statement

Databases include records of field samplings done on public and private lands, with the corresponding permission of owners. Field sampling was authorized by the Secretaría del Medio Ambiente y Recursos Naturales (SEMARNAT, Mexican Council for the Environment and Natural Resources), which legislates scientific field samplings in Mexico, through permissions FAUT-0052 and SGPA/DGVS/02726/10 for amphibians and reptiles, FAUT-0221 and FAUT-0232 for mammals. Birds were not collected. We did not perform any other activities that required specific permissions. For field sampling in Mexico, the approval by an Institutional Animal Care and Use Committee (IACUC) or equivalent animal ethics committee is not required.

Data analysis

To assess inventories’ completeness for each group of vertebrates at each ecoregion, we calculated the sample coverage [40], considering the total number of records per species, and the number of species with one and two records.

We partitioned beta diversity following the procedure of Carvalho et al. [8, 9], which is based on the approach of Baselga [5]. According to this method, total dissimilarity (βcc) is 1 minus the similarity coefficient of Jaccard. This total dissimilarity is divided into two components: the dissimilarity due to turnover (β.3) and the dissimilarity due to richness differences (βrich). We did this partitioning both for dissimilarity in species composition, and for dissimilarity in taxonomic structure considering the composition of higher taxa. For this last case, we used the method of Bacaro et al. [16], so that total taxonomic dissimilarity, herein βccT (1-Δ_T sensu Bacaro et al. [16]) equals the dissimilarity of Jaccard coefficient but taking into account higher taxa. Taxonomic dissimilarity is measured as: βccT = 1-(Ta/Ta+Tb+Tc), where Ta is the total number of taxa shared between two communities, Tb is the number of taxa present only in the first community but absent in the second, and Tc is the number of taxa present exclusively in the second community (Fig 1). Values of βccT range from zero when the taxonomic structure of both communities is identical, to 1 when the taxonomic structure is totally different [6]. Taxonomic dissimilarity measures the proportion represented by not-shared taxa from the total number of taxa in the two communities. Therefore, partitioning of βccT with the procedure of Carvalho et al. [8] gives one component of dissimilarity due to taxon turnover (β.3T) and one component of dissimilarity due to difference in taxa richness (βrichT).

To calculate total taxonomic dissimilarity and its components we used as many taxonomic levels as possible. For amphibians: order, family, subfamily, genus and species; and for reptiles: subclass, order, suborder, infraorder, superfamily, family, subfamily, genus and species [35, 36]. For birds: subclass, infraclass, parvoclass, superdivision, division, subdivision, infradivision, superorder, order, suborder, superfamily, family, subfamily, genus and species [38, 39]. Finally, for mammals: order, suborder, infraorder, superfamily, family, subfamily, tribe, genus and species [37].

We did all the analysis of dissimilarity partitioning in the R program [41], using the script of Carvalho et al. [9]. First, to assess if the two components of total species and taxonomic dissimilarity have equal contributions, we used multiple-site measures, sensu Diserud & Ødegaard [42], including all the ecoregions that occur in Hidalgo. Second, to detect which ecoregions have the highest values of species and taxonomic dissimilarity for each biological group, we used pairwise comparisons indices between ecoregions, as described above (βcc, βrich, β.3, βccT, βrichT, and β.3T). Then, to spatially represent the values of these indices we used multivariate ordinations of non-metric multidimensional scaling (NMDS). Finally, we did Pearson correlations between species and taxonomic dissimilarity, as well as their components, for each biological group. The NMDS and correlations were done in the program Past 3.07 [43].

Contribution of dissimilarity components

Not surprisingly, turnover has a more important contribution to total dissimilarity than richness differences, and birds had the highest turnover value while amphibians had the lowest one. This prominence of turnover coincides with beta diversity patterns of vertebrates in the Isthmus of Tehuantepec [44], amphibians in Brazilian Mata Atlantica [45], reproductive birds and lizards in flooded islands of East China [11], and with temporal beta diversity of birds in agricultural fields in France [14]. The spatial scale may have an influence in all these results because at small scales environmental filters exclude species that lack physiological tolerance to climatic or environmental local conditions [46]. In contrast, when the study area is broader, biogeographic processes influence species turnover and nestedness due to richness differences [23]. This occurs for example in global patterns of amphibian distributions, where the relative importance of beta diversity components changes with latitude: turnover is more important below parallel 37, while above this richness differences become more important [47]. Also, in America nestedness due to richness differences for amphibians increases as we approach to the poles, but it decreases for birds and mammals [48]. However, in the Mexican Transition Zone, biogeographic processes may influence turnover even in small areas.

Differences in species and taxa composition between ecoregions for reptiles, birds and mammals may be more influenced by environmental heterogeneity (differences in ecological conditions and vegetation types that lead to turnover) and historical factors than by differences in the number of species. But for amphibians, richness differences are more important in determining beta diversity than turnover between ecoregions. Other studies with vertebrates have explored possible causes. For example, in a global study with amphibians Baselga et al. [47] found that turnover was related to variation in elevation and temperature, while richness differences were related to evapotranspiration, annual precipitation and mean temperature. Also, Melo et al. [46] found that bird and mammal beta diversity in America is related to environmental heterogeneity, measured as elevation range, and this variable is responsible for up to 52% of bird beta diversity and 21% of mammal beta diversity. Moreover, for mammals in Europe beta diversity is related to topographic heterogeneity, mean annual temperature, as well as real and potential evapotranspiration [49]. Similarly, for terrestrial mammals in Mexico the highest beta diversity occurs in the Transvolcanic Belt, which is the most heterogeneous region of the country, while the lowest beta diversity corresponds to the most homogeneous region, i.e. the Yucatan Peninsula [50]. This pattern was later confirmed for terrestrial vertebrates [51], supporting the idea that environmental heterogeneity is a key factor determining beta diversity [46, 49]. Thus, although we do not have data on environmental heterogeneity at the scale of ecoregions, we assume that this factor may be crucial in our study area given the high proportion of turnover found in our results, and how turnover is related with heterogeneity as described in these studies.

Dissimilarity between ecoregions

The ecoregions in Hidalgo have different origins and geological histories, which result in the current complex topography, climatic heterogeneity and variation in vegetation types [27–29]. For example, the Meseta Central ecoregion is isolated form other arid zones, and in Hidalgo it is surrounded by the Sierra Madre Oriental. This Sierra Madre ecoregion harbors conifer forests, mainly pine-oak forests, and goes from North U.S.A. to Southern Mexico, and connects with the Volcanic Belt [30]. The Volcanic Belt is located in central Mexico and has several active and inactive volcanoes, including 13 of the highest mountains in the country. This intense volcanic and orogenic activity has allowed the formation of microhabitats, which promoted taxa radiation and speciation [28–29]. The Mexican Matorral is one of the most extensive ecoregions in the country, and the largest in Hidalgo. It emerged from superficial folds of Cretaceous sediments, resulting in a region with abrupt topography that includes valleys, canyons and ravines [28]. In Hidalgo this ecoregion includes three types of scrubland. Finally, the Moist Forest and Montane Forest ecoregions are located in the northeastern portion of the state, and include tropical perennial and sub-perennial forests, as well as the northern mountain cloud forests, that encompass high plant and animal richness and are currently threatened by human activities [30].

In our results, these two ecoregions (Moist Forest and Montane Forest) resulted with similar taxonomic composition because of the Neotropical affinity of their fauna, while they are located on the other side than Meseta Central, Sierra Madre Oriental, Volcanic Belt and Mexican Matorral in the NMDS graphs (Fig 3). These four ecorregions are more related with North America and thus their fauna has Nearctic affinity [28–29]. Thus, dissimilarity is clearly influenced by the biogeographic origin of the biota.

Differences among vertebrates

Contrary to previous studies [44, 46, 48, 51–53], we found that reptiles and mammals have the highest values of dissimilarity, while amphibians and birds have the lowest, although differences are small. For example, in the Isthmus of Tehuantepec in southeastern Mexico, where there are no altitudinal differences birds had the highest values of beta diversity, while amphibians had the lowest values [44]. Also, at the scale of the American continent birds had higher values of beta diversity than mammals [46]. Other studies show that beta diversity may be higher for amphibians and reptiles than for mammals and birds [48, 51–53]. In most cases, these results seem to be related with the dispersal abilities and niche limitations, because exothermal organisms (amphibians and reptiles) had in general lower dispersal capacities and stronger niche limitations than birds and mammals [48, 52–53].

Relationship between dissimilarity types

Correlations between species and taxonomic dissimilarities indicate that the number of species that may be considered in conservation efforts is related to the number of higher taxa. This is important because it suggests that, in our study area, if we focus on the species level for conservation, we are not only protecting current biological diversity, but also the taxonomic and evolutionary diversity of each ecoregion will be preserved, which is the result of millions of years of evolution.

Izsac & Price [6] also found this relationship between beta diversity and taxonomic beta diversity for different echinoderm communities in Asia. Furthermore, we think that it would be possible to find positive relationships between taxonomic beta diversity, phylogenetic and functional beta diversities for terrestrial vertebrates, in a similar way that has been found for freshwater fish in watersheds [54], and for birds in environmental gradients in France [55]. However, sometimes, taxonomic beta diversity is not related in any way to other measures of beta diversity, as occurs for estuarine tropical fish in Mexico, where functional beta diversity is low even when taxonomic beta diversity is high, because of the dominance of few functionally similar species and the low proportion of specialists [56].

Concluding remarks

This extraordinarily vast heterogeneity and environmental complexity requires complementary approaches so current biodiversity can be understood form different perspectives. In this paper we have shown trends in species and taxonomic beta diversity and their components. However, including other aspects such as phylogenetic and functional diversity measures will give a broader overview of biodiversity. Also, incorporating information of different biological groups and spatial scales would be desirable. Our results for terrestrial vertebrates may be helpful to assess risks and implications of beta diversity loss and the consequent homogenization of biotas, especially for ecological integrity and functioning at each ecoregion. For example, Brazilian Atlantic mixed forests have higher phylobetadiversity than other forest types because of the presence of temperate and tropical taxa that make them unique, and therefore this region requires the implementation of different conservation strategies [19]. This result from the Brazilian mixed forest coincides with our results of vertebrate beta diversity in the Mexican Transition Zone, especially when we include the taxonomic structure and highlight the importance of species and higher taxa turnover among ecoregions. As a conservation strategy of vertebrates, we believe that its necessary to review in future research projects the design of the current system of natural protected areas in Hidalgo, in order to guarantee the inclusion of relevant representative areas at the ecoregion level. In particular, our results indicate that to conserve amphibians the richest sites within each ecoregion require protection, because of the high contribution of dissimilarity due to richness differences between ecoregions. But for the conservation of reptiles, birds and mammals a system of protected areas, that include most of the heterogeneity within and among ecoregions, is needed. These local protection actions will warrant regional biological diversity.