Patterns of Species Richness

The highest species richness (172–203 species) was found at elevations of 1,000–1,700 m, mainly along the Atlantic slope (Figure 2(a)). Overall, species richness decreased as altitude increased. Low richness values, of 74 to 90 species, were recorded in two montane systems with elevations >3,500 m—the Sierra Madre de Chiapas (Mexico and Guatemala) and the Sierra de los Cuchumatanes (Guatemala), as well as in central Honduras, where richness was low despite a lack of high elevation areas. Many isolated mountains with small areas (<10 km²) had the lowest species richness values (30–73 species).

Figure 2.

Maps of species richness of the resident landbirds of the NCAH. (a) Total species richness, (b) richness of endemic species, and (c) richness of endangered species.

Despite having low overall species richness, high-elevation montane systems had the highest richness of endemic species (Figure 2(b)). The Sierra Madre de Chiapas, Cordillera de los Cuchumatanes, and the Sierra de las Minas had the highest concentration, with 18 to 21 endemic species. On the other hand, isolated high-elevation sites, such as northwestern Honduras, the volcanoes of El Salvador (except the Santa Ana Volcano), and northwest of the Central Chiapas Depression had the lowest values of endemic species richness (0–3 species). For endangered species (Figure 2(c)), the highest richness (8–11 species) was in the mountains of eastern Chiapas, the Sierra de las Minas, and the lowlands of the Sierra Madre of Chiapas and Guatemala. The lowest numbers of endangered species (0–3 species) were located in central Honduras and El Salvador.

Parsimony Analysis of Endemism

The PAE generated 360 equally parsimonious trees (length = 4,516, consistency index = 6, retention index = 84), which we used to obtain a strict consensus cladogram to develop a regionalization, which showed that the NCAH is divided into two main groups: Pacific slope and Atlantic slope (Figure 3(a)). Both groups are supported by two synapormophic species (Nyctidromus albicollis and Herpetotheres cachinnans). In the Pacific slope clade, there are four mountain systems in a successively nested pattern: center, southwestern, and eastern Guatemala, and El Salvador to western Nicaragua (Figure 3(b)). The Atlantic slope clade is divided into two groups, the lower altitude and higher altitude areas (Figure 3(c)). In Figure 3(d), the cordilleras of northern and eastern Honduras are grouped with western Nicaragua. Lastly (Figure 3(e)), there is a polytomy between the mountains that surround the Central Chiapas Depression and the mountains of eastern Chiapas, while central Honduras and eastern Guatemala form a clade together with southern Honduras, supported by the synapormophic species Atlapetes albinucha.

Figure 3.

Cladograms and geographic representation of the Parsimony Analysis of Endemism. (a–e) Representation of the process of grouping of the regions. The altitude and both Pacific and Atlantic slopes play an important role in the formation of the areas of endemism. The colors in the cladograms represent the grids grouped in the clade.

Cluster Analysis

The regionalization based on the phenogram obtained from the Cluster Analysis showed an initial separation of two main groups (with a similitude of 0.35) that represent the driest areas of the NCAH (Figure 4(a)). Next (Figure 4(b)), the low-altitude areas of the Atlantic slope (green) diverged from all other montane systems and high plains (blue). In Figure 4(c), the central zone, characterized by having high plains, was further divided into three groups, then the montane systems diverged into two groups. In Figure 4(d), the cordilleras of the Pacific slope are divided into three groups: northwest central Chiapas, central Chiapas to western El Salvador, and the Tacaná Volcano to the Tajumulco Volcano. On the other hand, the Atlantic slope is grouped as follows: eastern Guatemala to western Honduras, eastern Chiapas, and eastern Chiapas to central Guatemala. Subsequently, southern Honduras diverged, as shown in Figure 4(e). The phenogram concluded with the division of the Atlantic slope into six groups (Figure 4(f)).

Figure 4.

Phenograms and geographic representation of the Cluster Analysis. (a–f) Representation of the process of grouping of the regions. (f) The final groups are very similar to the physiographic aspect of the montane systems. The colors in the phenograms represent the grid grouped in this group.

Highland Birds and Protected Areas

Figure 5 depicts the distribution of species richness maps compared with the protected areas system. Although all the species analyzed are present in at least one protected area, the protected areas of Nuclear Central America are mainly located in tropical lowland ecosystems. A few regions of the montane systems have a notably high number of protected areas, like the Sierra Madre de Chiapas, the Sierra de las Minas, the volcanoes of El Salvador, and the Honduras summits. However, there are large gaps along the Guatemalan high plateaus, the mountains of eastern Chiapas, the Cuchumatanes, and the lower areas of Honduras mountains. The protected areas located in the Sierra de las Minas and Sierra Madre de Chiapas are notable also for simultaneously protecting an important species richness of resident, endemic, and endangered species (Figure 5).

Figure 5.

Protected areas on maps of species richness of the resident landbirds of NCAH and complementarity analysis. (a) Total species richness, (b) richness of endemic species, (c) richness of endangered species, (d) complementary grids, and (e) species accumulation curve of NCAH.

Complementarity Analysis

The complementarity analysis shows that 6 of the 360 cells, distributed along the whole area mainly in middle elevations (Figure 5(d)), hold the total number of 282 species of resident landbirds. All the species of conservation concern are concentrated in the two hexagonal cells with more complementary species, and the endemic species are all present in the combination of the first four hexagonal cells with more complementary species (Table 3). The species accumulation curve evidences that all the species are represented in a few hexagonal cells, reaching the asymptote with just one sixth of the cells (Figure 5(e)).

Table 3.

Number of Complementary Species per Grid of the Total Species Richness, Richness of Endemic Species and Richness of Endangered Species.

Patterns of Species Richness and Endemism

The highest concentration of species richness is located in middle elevations, between 1,000 and 1,700 m, where tropical and temperate ecosystems coexist (Mid-domain effect; Colwell & Hurtt, 1994; Colwell & Less, 2000; Colwell et al., 2004; Zapata, Gaston, & Chown, 2005). This area includes the cloud forest and its great biodiversity (Gual-Díaz & Rendón-Correa, 2014; Rzedowski, 2006). The high species richness found on the Atlantic slope is an extension of a pattern that starts in eastern Mexico south to South America, which is an effect of the great diversity of Neotropical species present in southeastern Mexico (Navarro-Sigüenza et al., 2014).

Our data also show a difference of 173 species between the sites with the lowest and those with the highest species richness. The low species richness found on the small isolated mountains may be due to ecological differentiation from the other montane systems caused by the Massenerhebung effect (Flenley, 1995), in combination with geographical barriers such as valleys and depressions which cause abrupt changes in humidity and temperature (Lim, Rahman, Lim, Moyle, & Sheldon, 2011; Robin, Vishnudas, Gupta, & Ramakrishnan, 2015). These phenomena shape variation in humidity, influencing ecological factors important for birds, such as altitudinal distribution of vegetation and structure among different types of mountains (Bruijnzeel et al., 1993; Terborgh, 1977).

The decrease in species richness at higher altitudes corresponds to a general pattern that has been documented in many taxa (Campbell & Vannini, 1989; Schuster & Cano, 2005; Terborgh, 1977). In the case of birds, the gradual decrease in the biodiversity from lowlands to summits corresponds with the simplification of the vegetation structure and the reduction of available resources (e.g., food; Terborgh, 1977). However, there are exceptions to this overall pattern in individual taxa; for example, our results show that the resident species of the families Emberizidae, Corvidae, Fringillidae, and Turdidae have higher species richness at higher altitudes (Stotz et al., 1996).

The distribution of endemic species richness occurs in a pattern that is opposite to the distribution of total species richness. The highest concentration of endemic species is found in the highest areas of Guatemala and southern Chiapas, and the values decline as altitude decreases. The high endemic species richness in the Sierra Madre de Chiapas, the Cordillera de los Cuchumatanes, and the Sierra de las Minas is likely a consequence of isolation due to the presence of large mountain chains surrounded by plains (Lim et al., 2011; Robin et al., 2015). In these areas, speciation is the product of multiple vicariant events (Rovito et al., 2012). High geological activity created mountains and plains, such as the Motagua fault, the Isthmus of Tehuantepec, and the Nicaraguan Depression, which eventually served as biogeographical barriers to biotic dispersal (Castoe et al., 2009; Husley & López-Fernández, 2011; Rovito et al., 2012). These biogeographical barriers also played an important role in the dynamics of species formation in the NCAH, as has been evidenced in birds and other taxa, such as pit vipers (Crotalinae; Castoe et al., 2009), freshwater fishes (Husley & López-Fernández, 2011), and salamanders (Rovito et al., 2012). Sánchez-González, Morrone, and Navarro-Sigüenza (2008) suggested that the high number of endemic species is evidence of long isolation of the region and asynchronous colonizations. On the other hand, Weir et al. (2009) mentioned that geological dynamics at different times during the formation of Central America (GABI) also allowed the interchange of avifaunas that had been previously isolated in North and South America.

The number of bird endemics in the NCAH led to its recognition as an Endemic Bird Area, in addition to its biological importance in some other aspects like taxonomic diversity (Stattersfield et al., 1998). We found slight discrepancies in the number of endemic species. For example, BirdLife International records only 20 endemic species (BirdLife International, 2015;  http://www.birdlife.org/) compared to the 26 reported in this article. This might be the result of using different taxonomic approaches; for instance, the approach used here (following Gill & Donsker, 2013) recognizes more species because it considers evolutionary allopatric units to be separate species where other traditional species concepts would recognize only one species (see Navarro-Sigüenza & Peterson, 2004; Peterson & Navarro-Sigüenza, 2009) that have proved to give a finer resolution to biogeographic (Bertelli et al., 2017) and conservation analyses (Peterson & Navarro, 1999). Despite these differences, the general patterns of altitudinal and regional variation of species richness and endemism, as they represent different species assemblages, have implications in the understanding of the biogeography of the region and especially in the conservation perspective, given that conservation not only involves the care of areas with pristine vegetation and high species richness, but also represent biotic interactions and evolutionary processes (Blois, Zarnetzke, Fitzpatrick, & Finnegan, 2013; Fjeldsa, 1994; Margules & Sarkar, 2009).

Regionalization of Nuclear Central American Highlands

PAE suggests a complex pattern of biogeographic structuring resulting also from a complex history, also giving avifaunistic identity to each of the groupings formed in the analysis. PAE showed that the NCAH is divided into two main clades—the Pacific slope and the Atlantic slope. This pattern of distribution is largely influenced by the presence of species that are mostly associated with lowlands that are at the upper limits of their altitudinal range in the region. These species are found mainly along the tropical lowlands of the Gulf of Mexico, Yucatan Peninsula, lowlands of the Caribbean coast, the lower parts of the NCAH, and some along the Pacific lowlands of the region. Unlike the lowland species, the distribution of the highland-associated species does not strictly follow the pattern associated with slopes. Rather, some of these species, like those associated with less-humid vegetation such as drier pine and oak forests, are distributed either only in the NCAH or along both slopes of the mountain chains of Mesoamerica. Species associated with more humid highland forests, such as cloud forests, also show a particular pattern in the NCAH. Sánchez-González et al. (2008) analyzed the biogeography of the avifauna of the Neotropical cloud forests and obtained a clade conformed of cloud forests from eastern Mexico to north-central Nicaragua. This is interesting because bird communities of cloud forests that grouped the NCAH with eastern Mexico also have endemic species that give biogeographical identity to the NCAH. Another example includes the volcanoes of El Salvador and Cadena de los Maribios that are characterized by being isolated and surrounded by lowlands and dry tropical forests. In this case, the regionalization of birds of dry tropical forests by Ríos-Muñoz and Navarro-Sigüenza (2012), this mountainous region forms a clade called Inner Central America.

The areas of endemism obtained here partially coincide with those suggested by Schuster, Cano, and Cardona (2000) using beetles (Coleoptera, Passalidae). The coinciding areas are located in the Sierra Madre, eastern Chiapas mountains, Cordillera de los Cuchumatanes, Sierra de las Minas, Sierra del Merendón, and Sierra Gallinero-Cordillera Celaque. However, our results suggest other areas of endemism in the volcanoes of El Salvador-Cordillera de los Maribios, Cordillera de Montecillos, great volcanoes of Sierra Madre, Sierra de Dipilto-Meseta de las Tablas, and Cuenca de Esquías-Cuenca de Mosquitia. In the same group of beetles, Schuster and Cano (2005) found that altitudinal distribution does not correspond with clade formation, surely due to the Massenerhebung effect (Flenley, 1995). Our data suggest the same pattern overall, in the border of the NCAH region both along the Pacific and Atlantic slopes. This can be attributed to the different elevation and isolation of the mountains but also to the presence of lowland species that arrive to the highlands along both slopes.

On the other hand, regions formed by the Cluster Analysis are congruent with the overall physiography of the montane systems. First, the volcanoes of El Salvador, Cadena de los Maribios, and part of the Central Chiapas Depression are grouped, characterized by being the most isolated and driest areas of the NCAH. Possibly, their grouping is a consequence of the presence of species that are mainly distributed in the tropical dry forest. Ríos-Muñoz and Navarro-Sigüenza (2012) presented two groupings that are similar to the previous areas, although these do not form a monophyletic group. The next group corresponds to the areas bordering the Atlantic slope, that at some sites coincide with the regions of highest species richness. This grouping is influenced by species shared among the lowlands of the Atlantic slope and the lowest elevations of the NCAH. The distribution of these (typically lowland) species only in the lowest altitudes of the NCAH generates greater similarity among sites with lower altitudes than between adjacent sites. This influence allows, for example, Monte Victoria (Belize) to be grouped with Sierra Merendón and Montaña Botaderos, despite these sites are geographically isolated from each other.

The effects of both geological and climatic changes, adaptation processes, continuous dispersion, biotic interactions, altitude, and biogeographical barriers have shaped the bird community of the NCAH (Chapin & Körner, 1996; Körner, 2004; Lim et al., 2011; Robin et al., 2015). It is noticeable that many groupings coincide among the PAE (Figure 3) and the Cluster Analysis (Figure 4), such as the volcanoes of El Salvador and Cadena de los Maribios (Nicaragua), Meseta de las Tablas and Sierra Dipilto (Nicaragua), Cuenca de Esquías and Mosquitia (Nicaragua), Cordillera de los Montecillos (Honduras), Sierra del Merendón (Honduras), Tacaná and Tajumulco volcanoes (Guatemala), Guatemalan high plains, and central Chiapas (México). Other groupings only coincide partially, such as Sierra Gallinero (Honduras), Cordillera de los Cuchumatanes, and Sierra de las Minas (Guatemala), and the volcanoes of central Guatemala (see Figure 1). Only three regions do not coincide: the western central Honduras, western Chiapas, and north-eastern Chiapas.

The patterns of distribution here described follow some general patterns outlined in previous works, but our results now provide a more detailed figure of such regions having unique combinations of bird species: the Atlantic slope, the Pacific slope, and the Highlands (Álvarez del Toro, 1964, 1971; T. R. Howell, 2010; Jolón-Morales, 2005; Komar, 1998). Moreover, regionalization presented herein partially coincides with that of Schuster et al. (2000), improved by the finding of additional areas of endemism. Thus, our study adds biogeographic data on the avifauna that were scarce (Peterson et al., 1998), improving the framework on which both further research and conservation decisions can be designed.

Implications for Conservation

The highest concentration of endangered species coincides with sites where the total species richness is also high. The endangered endemics Oreophasis derbianus and Tangara cabanisi (EN; IUCN, 2018) are restricted to cloud forest. According to this category, habitat loss or change has a negative impact on species population size. There are five vulnerable species (VU; IUCN, 2018) that face habitat or population decrease; these species are the regional endemics Penelopina nigra, Megascops barbarus, and Cardellina versicolor, as well as the more widespread Crax rubra and Electron carinatum. Also, six species are categorized as Near Threatened (NT; IUCN, 2018), at risk because they are particularly sensitive to habitat loss due to their ecological specialization and therefore restricted distribution, like Xenotriccus callizonus and Pharomachrus mocinno (Dayer, 2010; S. N. G. Howell & Webb, 1995). On the other hand, other species listed have ampler ranges: Colinus virginianus is widely distributed in lowlands and North America, and Tinamus major, Harpyhaliaetus solitaries, and Spizaetus ornatus are distributed largely in the lowlands and South America.

The function of conservation areas is to protect samples of biota so that evolution may act on them in the future; for that reason, we should aim to have the highest representativeness of species possible in protected areas (Margules & Sarkar, 2009; Whittaker et al., 2005). Among the areas protected in the NCAH, those that stand out as important sites for regional avifauna are located in Sierra de las Minas and Sierra Madre (Chiapas and Guatemala). These areas protect a high richness of resident species, endemic species, and endangered species. However, both montane systems belong to very close groups according to the Cluster Analysis (Figure 4(f)) and do not match any area of endemism (Figure 3(e)). This leads us to deduce that they protect high species richness but of very similar bird communities. Despite their similarity, the analysis also suggests that the two sites are complementary and together they contain more than 95% of the species, 69% of endangered species, and all the endemic species. This is a common problem; therefore, it is recommended that conservation areas complement each other to protect a greater representation of species diversity (Justus & Sarkar, 2002; Margules & Sarkar, 2009; Whittaker et al., 2005).

Another important point is the inclusion of rare species, which are defined by their area of distribution, habitat specificity, or local population size (Margules & Sarkar, 2009). In the case of the high plains of Guatemala region, overall species richness is not particularly high, but it concentrates the majority of endemic species; it also forms a distinct group in the Cluster Analysis (Figure 4(f)) and is an area of endemism (Figure 3(e)). Despite its characteristics, the protected area located there is smaller than in the montane systems. Similar to the high plains of Guatemala, eastern Chiapas requires more protected areas because it concentrates high species richness, especially of endangered species, it is an important area of endemism (Figure 3(e)), and Cluster Analysis shows it to be a distinct group (Figure 4(f)). Added to this, degradation and loss forest due to social development and drug trafficking are a problem in the NCAH (McSweeney et al., 2014; Redo et al., 2012). Increasing the coverage of protected areas in these zones would benefit the complementarity and the conservation of communities and habitats that do not have an adequate representation in existing protected areas of the NCAH. Another issue is that even when protected areas conserve complementary and distinctive communities, if they are not adjacent they lead to an archipelago of disconnected protected areas (Margules & Sarkar, 2009).

Our results offer a new view on the regional understanding of bird diversity patterns and the biogeographical processes that have shaped regional richness. We believe that these maps of both species richness and regionalization can serve as a useful tool for conservation biologists and authorities in implementing strategies for the protection of natural areas in NCAH. However, more work is still needed to obtain a more complete picture of the complexities of bird diversity in the northern Neotropics. Prioritization of conservation areas must take into account the different sets of information, as the ones provided herein. A combination of areas detected by different sets of data like with high species richness (e.g., Sierra de las Minas), high endemism (e.g., Sierra Madre), and those with avifaunistic identity detected by the biogeographic regionalization analyses (e.g., eastern Honduras highlands), would be ideal. However, more exhaustive field evaluation of the representation of taxa within protected natural areas, analysis of the complementarity between these areas, possibilities of connectivity among highland areas through its connection via the lowlands (e.g., the Mesoamerican Biological Corridor, CONABIO 2018), and assessment of the vulnerability of their habitat are all necessary in order to arrive at adequate conservation strategies for this unique avifauna.