INTRODUCTION

To be effective, conservation efforts should protect biological and ecological diversity at multiple levels of organization (Trombulak et al. 2004), from very small (genetic) to very large (ecosystems and landscapes). Biodiversity is most effectively maintained by protecting larger areas (Soulé 1985; Noss and Cooperrider 1994), especially in cases where heterogeneous landscapes feature high gamma (regional scale) diversity and include sites of high alpha (within-habitat) diversity (spatial scales of diversity sensu Whittaker 1960, 1972). Strategies for long-term protection of biodiversity under the changing conditions of climate change have remained remarkably consistent over the past three decades (Noss 2001; Heller and Zavaleta 2009; McLaughlin et al. 2022).

Conservation potential is strengthened when protected landscapes provide gradients of environmental factors such as elevation, precipitation, and temperature, as such conditions allow for shifts in species ranges. Similarly, connectivity with continental-scale corridors (such as north–south trending mountain ranges) optimizes opportunities for population shifts when physical conditions change (Hunter et al. 1988; Ranius et al. 2023). Connectivity and climate corridors—pathways that connect warmer to cooler areas less impacted by human activity—offer efficient opportunities for ecological adaptation to climate change (Lawler 2009; Lawler et al. 2020). Conservation biologists have called for identifying “climate change refugia”—areas that are sufficiently buffered from climate change as to enable persistence of species and other valued resources, as during climate upheavals in the paleoecological past (Morelli et al. 2016). A recent analysis of this approach (Saunders et al. 2023) suggests that the southwestern United States, and particularly the Rocky Mountains and the Sierra Madre Occidental of México, represent potential high-priority climate change refugia.

In the American Southwest (especially Arizona), biological attention has been focused largely on the canyon country of the Colorado Plateau and the especially well-studied Sonoran Desert. This has been facilitated, in part, by the largest research universities and government agencies (e.g., USGS, USFS, and cooperative parks study units at Northern Arizona University in Flagstaff, Arizona State University in Tempe, University of Arizona in Tucson, and the University of Utah in Salt Lake City). The continental-scale ecotonal transition zone that connects the Colorado Plateau and the Sonoran Desert, much of it along the 200-mile-long escarpment known as the Mogollon Rim, has accommodated biological studies, often addressing autecological or ethological questions (e.g., Barber et al. 1998; Dobbs and Martin 1998; Martin and Martin 2001), but has remained less known than the Plateau country to the north and the low deserts to the south. In an analysis of herbarium records for the western United States, for example, all counties of this transitional zone were “under-collected” (Taylor 2014). This transition zone has tended to be viewed as a line—the boundary between two other places—and has been overlooked as a distinct region worthy of consideration at the landscape level.

We have identified this area as the Mogollon Highlands Ecoregion (MHE) (Fleischner et al. 2017). Here, we make the case that the MHE deserves recognition as a distinct and important ecological region in its own right, not just as a transitional zone between two long-recognized regions (Colorado Plateau and Sonoran Desert). Given its location and the evidence of prior research, we argue that the MHE represents one of the most biodiverse regions in North America —an ecological treasure that the scientific community has largely failed to recognize. As a result, this region has been undervalued in the realm of conservation strategy.

The MHE represents an ecologically fascinating North American transition zone of continental importance. This dramatic landscape of escarpments, canyons, mesas, deserts, and high conifer forests is a land of high biological and ecological diversity. It is where the Sonoran Desert of the Basin and Range Province meets the redrock canyon country of the Colorado Plateau and the Southern Rocky Mountains, and where the northern limits of some species coexist with the southern limits of others. In the Mogollon Highlands, the mega-diversity of Meso-America, and the Sierra Madre in particular (DeBano et al. 1995), has direct access into North America. As Felger and Wilson (1995) pointed out more than two decades ago, this Apachian/Madrean region is a “neglected center of biodiversity.” More recently, it has been referred to as a “biodiversity hotspot” in reference to herpetology (Bezy and Cole 2014). Davis et al. (1997) highlighted the region as a notable center of endemism in North America.

The core of this region is what the World Wildlife Fund (WWF) named the “Arizona Mountain Forest” (Ricketts et al. 1999). WWF concluded this region had regionally outstanding biological distinctiveness, due to relatively high species richness and endemism. The Mogollon Highlands largely coincides with The Nature Conservancy's “Arizona-New Mexico Mountains” ecoregion (Marshall et al. 2006), although, as defined here, the Mogollon Highlands extends beyond these montane forests to include parts of adjacent, interwoven communities. As we recognize it, the Mogollon Highlands coincides with the northern portion of the “Apache Highlands” of The Nature Conservancy; however the Apache Highlands extends far south, including the Madrean “Sky Islands” and into the Sierra Madre Occidental of Sonora and Chihuahua, Mexico (Turner et al. 2005).

Portions of the Mogollon Highlands have been referred to, inconsistently, by many names (e.g., “Arizona Central Highlands,” “Arizona Transition Zone”), yet the area remains relatively ill-defined relative to its southern counterpart, the Sonoran Desert, and its northern neighbor, the Colorado Plateau. This region roughly aligns with the interface of two great physiographic provinces of the American West: the Basin and Range and Colorado Plateau (Hunt 1967). Due to great geologic diversity (Nations and Stump 1996), the Mogollon Highlands present dramatic topographic diversity, varying several thousand feet in elevation, and including a series of deep canyon systems that drain off the Colorado Plateau and emerge into the low Sonoran Desert. The MHE contains the highest density of springs in Arizona (Stevens 2018).

The region's positioning at a continental-scale biogeographic crossroads contributes to its high ecological diversity. The southern extent of the Rocky Mountains intersects the eastern portion of the Mogollon Highlands. All four of North America's deserts connect directly with the region: the Great Basin Desert to the north, the Mojave to the west, and the Sonoran and Chihuahuan to the south. Some species (e.g., Juniperus osteosperma and J. scopulorum) reach their southern extent here, while Sierra Madrean species (e.g., alligator juniper [Juniperus deppeana], Emory oak [Quercus emoryi], red-faced warbler [Cardellina rubifrons]) reach their northern boundaries. Additionally, some Great Plains species (e.g., blue grama [Bouteloua gracilis], thirteen-lined ground squirrel [Spermophilus tridecemlineatus], gray catbird [Dumetella carolinensis]) reach their western extent here while the eastern extent of some shrubs (e.g., Rhus ovata) lies in this region.

This rich merging of biotas owes much to larger-scale climatic patterns. The Southwestern Climate Pattern—typified by winter and summer precipitation, separated by spring and fall droughts —is influenced by both mid-latitude and subtropical atmospheric circulation regimes (Sheppard et al. 2002; Sellers 2008). One specific consequence of this atmospheric confluence is a bi-seasonal precipitation pattern (Lowe 1964), which, in turn, leads to higher botanical diversity as plant phenologies can be oriented around one or both seasons. As a result, floristic affinities extend both north and south (McLaughlin 2008).

The Mogollon Highlands includes five of the North American life zones described by Merriam (Lowe 1964; Phillips et al. 1989). It represents an interfingering of 11 of the 26 biotic communities in the southwestern United States and northwestern Mexico (southern Utah to northern Sinaloa, Pacific Coast to New Mexico), as described by Brown (1994). Because of this confluence of habitats, the MHE encompasses much of the plant diversity of Arizona, which has the third highest plant species richness of any state (Stein et al. 2000). The regional diversity is amplified even more due to punctuation by linear ribbons of riparian forest—one of the highest productivity habitats in North America. These lush green corridors concentrate wildlife, and include some of the highest biodiversity (alpha diversity) sites in North America (Johnson et al. 1977; Ohmart and Anderson 1982; Fleischner 1999). (Because of their limited geographic extent, however, riparian zones do not appear as distinct “communities” in our analysis.) Based on these observations, the MHE accords with criteria for high priority conservation consideration. In a report by the National Biological Service, two community types in the Mogollon Highlands region were identified as “endangered ecosystems,” defined as those in 85–95% decline: old-growth ponderosa pine (Pinus ponderosa) forest, and Southwest riparian forests (Noss et al. 1995).

The ongoing ecological health of the Mogollon Highlands is especially important in this age of rapid climate change. This area of dramatic elevational gradients, at a continental-scale biogeographic crossroads, can yield understanding of the capacity for species and ecological communities to adapt to global climate change. Based on these factors, we posit that the MHE is a center of biological and ecological diversity, worthy of recognition and protection.

Accurate mapping of species diversity is essential for the identification of areas of high conservation priority (Andermann et al. 2022). Having established the boundaries of the MHE (see Methods), we commenced biological analysis of portions of the flora and fauna within it. There are significant challenges in clarifying species diversity in several important taxa—for example, vascular plants—due to chaotic taxonomic revisions that render comparisons between data sets problematic. Highly mobile animals, such as birds, present in the Mogollon Highlands in various seasonal patterns, can confuse geographic analysis.

As a primary example, we focus here on snake diversity in the MHE. In a recent comprehensive review, Holycross and Mitchell (2020) thoroughly documented the snake fauna of Arizona. Their work demonstrates that biotic communities in Arizona each have their own relatively distinct herpetofauna (see also Holycross et al. 2022). Snake species distributions are limited by a number of environmental factors such as precipitation, surface temperature, soil type, or elevation (Schall and Pianka 1978) and are often constrained by landscape features such as canyons, mountains, or rivers (Brown et al. 2020). Because of the high-quality data available from this ongoing research, this taxon offers the ideal starting point for corroborating the biological importance of the MHE. Given the ecological diversity of available habitats in this region, directly resulting from its ecotonal character, we predict higher snake diversity in the MHE compared with other regions of comparable size in the contiguous United States.

METHODS

RESULTS

Our refined map of the MHE is represented in Figure 1. The MHE covers 70,490 km² of highly diverse terrain spanning elevations above 1061 m (3500 feet) in a south-east trending arc through central Arizona and western New Mexico (Figure 1). The diverse landscapes of this Ecoregion support 20 communities. Vegetation and habitat characteristics of these 20 communities are summarized in Appendix 1.

The most extensive communities in the MHE are three types of conifer forests, five types of woodlands primarily dominated by conifers, oaks and other perennial shrubs (interior chaparral), six types of grasslands and tablelands, and six high deserts and basins; winding through each of these is riparian vegetation. The region overlaps small upland areas of Mohave, Sonoran, and Chihuahua desert origin.

We identified 2223 vouchered observations representing 39 species of snakes in the Mogollon Highlands (Appendix 2). Three Thamnophis records and one Salvadora record were not assigned to species. Both of these genera are represented by more than one species with the Mogollon Highlands and cannot be assigned to a species without examining the specimens. The number of recorded snake species varied considerably across the Mogollon Highlands communities with records lacking for six communities (Table 1). Variability among communities in the number of records is likely due to several factors: areal representation of the community within the Mogollon Highlands, rugged terrain inhibiting travel, lack of public access (e.g., on some tribal lands), and remoteness from population centers. Clearly, areal representation (km²) in the Mogollon Highlands is a critical factor; the six communities lacking records altogether were among the eight smallest communities. Interestingly, the Low Mountains and Bajadas community was represented by only 95.4 km² and yielded only 7 snake records, but was among the best sampled and highest diversity communities in metrics controlling for community area (records/km² and species/km²). After controlling for community size (area in km²), the communities capturing the highest snake diversity were (in descending order) (1) Arizona Upland/Eastern Sonoran Basins, (2) Low Mountains and Bajadas, (3) Chihuahuan Deserts and Grasslands, (4) Lower Mogollon Transition, (5) Montane and Subalpine Grasslands, and (6) Madrean and Lower Montane Woodlands. Several of these communities (1, 2, and 5) rank highly in this metric because although relatively few species were documented within them, they are also represented by relatively few km² in the MHE. The majority of the communities contributing the most to absolute diversity are relatively low-elevation habitats along the southern boundary of the MHE, as expected when assaying an ectothermic taxon.

Figure 1.

The Mogollon Highlands Ecoregion, showing 20 component communities.

Snakes Endemic to the Mogollon Highlands

Of particular note are two species of snakes that are endemic, or nearly endemic, to the Mogollon Highlands: the Mogollon narrow-headed gartersnake (Thamnophis rufipunctatus) and the Arizona black rattlesnake (Crotalus cerberus). The Mogollon narrow-headed gartersnake (T. rufipunctatus) is an aquatic specialist and restricted to the upper watersheds of the Gila River in the Salt, Verde, and San Francisco rivers in Arizona and the San Francisco and Upper Gila rivers in New Mexico. (Holycross et al. 2020b). It is found in Conifer Woodlands and Savannas, Lower Madrean Woodlands, Lower Mogollon Transition, Madrean Lower Montane Woodlands, Mogollon Transition Conifer Forests, and Montane Conifer Forests where it is restricted to perennial streams at elevations ranging from ∼700 to 2430 m (Rossman et al. 1996). Thamnophis rufipunctatus is listed as threatened under the U.S. Endangered Species Act (USFWS 2014). Precipitous decline has been recorded in this species in recent years, likely because of its high degree of habitat specialization and its diet—which is almost exclusively fish (Hibbitts et al. 2009).

The Arizona black rattlesnake (C. cerberus) is nearly endemic to the MHE. Outside of the Mogollon Highlands, populations of C. cerberus are found only in peripheral isolated mountain ranges and canyons (Holycross et al. 2020a; Nowak et al. 2020). This rattlesnake occurs most often in high-elevation woodlands, usually in riparian areas, and is associated with a variety of ecoregions, including Arizona/New Mexico Subalpine Forests, Chino/Coconino Grasslands and Shrubsteppe, Conifer Woodlands and Savannas, Low Mountains and Bajadas, Lower Madrean Woodlands, Lower Mogollon Transition, Madrean Basin Grasslands, Madrean Lower Montane Woodlands, Mogollon Transition Conifer Forests, and Montane Conifer Forests (Hulse 1973; Jones 1988; Nowak and Schofer 2006; Griffith et al. 2014; Holycross et al. 2022). It is most commonly found at elevations from 900 to 3000 m (Nowak et al. 2020).

Table 1.

Snake diversity and frequency by community type in the Mogollon Highlands Ecoregion.

DISCUSSION

The Mogollon Highlands Ecoregion merits consideration as a prominent, unique, high biodiversity transition zone of North America—not merely as the neglected edges of other provinces. Moreover, regions with significant elevational gradients, and with broad interpenetration of numerous ecological communities, represent living laboratories for adaptation to ecological and climatic change. The MHE is ideally suited for ongoing studies of adaptation to a changing Earth, and provides an important conservation opportunity.

The snake database serves as a starting point for evaluating the biogeographic affinities and patterns of distribution of a portion of the Mogollon Highlands fauna. There are four main lines of evidence that illustrate the biogeographic affinities and patterns in the MHE. First, the Mogollon Highlands exists at the nexus of several ecoregions, each with a relatively distinct herpetofauna (Holycross et al. 2022). Although largely peripheral to the Mogollon Highlands, these ecoregions—Sonoran Desertscrub, Mohave Desertscrub, Great Basin Desertscrub, Plains and Great Basin Grassland, Semidesert Grassland, Rocky Mountain Subalpine Conifer Forest, and Subalpine Grassland—contribute to the diversity of the Mogollon Highlands snakes. For example, Crotalus pyrrhus, which is associated with the xeric rocky environs of the Sonoran Desert, is included in the biota of the Highlands because portions of this desert are represented within the southern margin of the western Highlands. Likewise, Lampropeltis triangulum, which is primarily associated with Plains and Great Basin Grassland found to the north of the Highlands, is represented because pockets of this community persist within its boundaries.

Second, the proximity of the Cochise Filter Barrier contributes to diversity from east to west across the Mogollon Highlands. The Cochise Filter Barrier is a region that (very generally) follows the spine of the Peloncillo Mountains from the Sierra Madre in Mexico up to the Mogollon Highlands. This region serves or served as a barrier to gene flow for closely related organisms (including many amphibians and reptiles) distributed both east and west of it. Historically, separate subspecies of several species of snake have been recognized on either side of the Cochise Filter Barrier, although recently, many have been elevated to full species status. While some of these elevations are controversial, it is clear that the barrier contributes to diversity in the region. One example is the presence of both Crotalus molossus and Crotalus ornatus within the Mogollon Highlands. Only recently recognized as a separate species, C. ornatus is only present in the far eastern portion of the Mogollon Highlands.

Third, the steep elevational gradient present within the Mogollon Highlands contributes to diversity. Species limited in their elevational distribution can be found at the ends of this broad elevational gradient within the bounds of the Highlands. Examples among the snakes include Phyllorhynchus decurtatus (generally limited to low elevations) and Lampropeltis pyromelana (generally limited to relatively higher elevations).

Figure 2.

Bird species richness in the coterminous United States, derived from species habitat distribution models (from Gergely et al. 2019: p. 14.). Mogollon Highlands Ecoregion outlined in yellow.

Finally, the dissected nature of the highlands contributes to the diversity of the region. Large canyons carved by major rivers are lined by biotic communities associated with lower elevations, creating a unique mosaic of communities. These incursions of low-elevation habitat explain the presence of species like Crotalus atrox along the Salt River at a location very near the geographic center of the Mogollon Highlands.

We documented 39 species of snakes within the boundaries of the Mogollon Highlands. For comparison, Bezy and Cole (2014) documented 35 species of snakes in the Madrean Archipelago of southern Arizona and New Mexico. While their methods of constructing species lists differ from ours (their criteria being stricter in some senses), the Madrean Archipelago is often heralded for its herpetological diversity. These data support our prediction that the snake diversity of the Mogollon Highlands Ecoregion is on par with that of that of the Madrean Archipelago and, by extension, on par with, or higher than, many other regions (Bezy and Cole 2014; Holycross and Mitchell 2020).

Our findings on snake diversity supplement previous knowledge on biodiversity of other taxa. There are more species of butterflies (Lepidoptera: Rhopalocera) in the Central Arizona Highlands—essentially, the Arizona subset of the MHE—than in most American states (McNally 2020). Recent analysis (Gergely et al. 2019) shows that the highest bird habitat diversity in the coterminous United States is centered in the MHE (Figure 2). The MHE is also an area of great botanical importance due to elevated levels of endemism and genetic exchange between species, as exemplified by the rich diversity in the gymnosperms. Included in the MHE are gymnosperms that approach the limits of their geographic distributions, for example, Juniperus deppeana, alligator juniper (MHE is northern limit, with a few fingers reaching north of the Mogollon Rim) and Pinus edulis and Juniperus osteosperma (MHE is the southern limit).

An important group of gymnosperms of the Mogollon Highlands are the cembroid pines, including many of the piñon pine species. The most widespread piñon pine in the Mogollon Highlands, known variously as Pinus edulis var. fallax (Little 1968; Felger et al. 2019), Pinus × fallax (Buck et al. 2023), or Pinus fallax (elevated to species level by Montes et al. 2019, 2022), is tolerant of higher temperatures and lower moisture than other closely related piñon species. Here, we call this species the Mogollon Highlands piñon, Pinus × fallax (after Buck et al. 2023), to call attention to its narrow distribution and unique characters. Long known for its varying needle number (typically one needled with some two-needle fascicles), it was thought that the Mogollon Highlands piñon arose from hybridization of the more northern two-needled Pinus edulis and the Great Basin one-needled P. monophylla (e.g., Lanner 1974a, 1974b). However, recent genetic studies using nuclear DNA and chloroplast origin show that this piñon is a hybrid, sharing genes of P. edulis (Colorado piñon) and P. californarium, a one-needled species that occurs today well west, on California hillsides (Buck et al. 2023). Today it occupies hotter and drier habitats than either of its parents (i.e., 12 mm less rainfall and 2–4°C higher temperatures). It is a stable morphotype, shows less heterozygosity than might be expected of hybrids (Buck et al. 2023), and has been found in packrat middens from 48,000 y ago (Cole et al. 2013). As such, novel gene combinations in this piñon make it well adapted to the transitional shrub-rich habitats of the Mogollon Highlands through its ability to tolerate aridity and slightly higher temperatures than its parental genotypes. These novel gene combinations may well be reintroduced into the parent populations by backcrossing in the future, heightening the importance of this Mogollon Highlands hybrid. The resilience of this piñon pine can, in part, be attributed to its hybrid status; between 1985 and 2000, woodlands increased in extent in the Mogollon Highlands, and have declined only slightly since (0.3–0.8%), despite drought and low levels of beetle infestations (Rodman et al. 2022). In contrast, recent climate shifts have had more severe impacts on Pinus ponderosa (ponderosa pine) forests of the Mogollon Highlands, as warming and drying conditions approach the physiological limits for this species, which now occurs as “trailing edge” forests (Huffman et al. 2020). Landscape-scale analyses have shown climate and wildfire-related declines since 2000, favoring the oak components of the forest such that a shift is seen from pine forest to the more xeric pine-oak (Rodman et al. 2022).

The Mogollon Highlands Ecoregion, then, represents a center of ecological diversity significant on regional and continental scales. This is clearly exemplified by the data presented here on exceptional snake diversity, and by the unusual convergence of conifer species, including endemism and genetic recombination not observed elsewhere. Recognition of the ecological importance of this ecoregion is long overdue. We believe that—especially in an age of rapid climate change—the MHE should be valued and protected as a remarkable center of biodiversity, a home to taxa with a unique array of adaptations, and an area of great biological potential, with so much yet to be understood.

SUMMARY