Gopherus evgoodei Edwards, Karl, Vaughn, Rosen, Melendez Torres & Murphy, 2016

Edwards, Taylor, Karl, Alice E., Vaughn, Mercy, Rosen, Philip C., Torres, Cristina Melendez & Murphy, Robert W., 2016, The desert tortoise trichotomy: Mexico hosts a third, new sister-species of tortoise in the Gopherus morafkai - G. agassizii group, ZooKeys 562, pp. 131-158 : 138-147

Gopherus evgoodei Edwards, Karl, Vaughn, Rosen, Melendez Torres & Murphy sp. n.

Figs 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14

Xerobates agassizii Cooper, 1861 (partim)

Gopherus agassizii (Cooper, 1861) (partim). Generic reassignment by Stejneger (1893)

Scaptochelys agassizii ( Cooper 1861) (partim). Generic reassignment by Bramble (1982)

Xerobates lepidocephalus (ex errore) Ottley and Velázques Solis 1989. In error by Crumly and Grismer (1994)

Gopherus morafkai Murphy, Berry, Edwards, Leviton, Lathrop & Riedle, 2011 (partim)

Holotype.

AMNH (American Museum of Natural History) R64160; adult male from Alamos (approximate location 27°02'N, 108°55'W, elevation 433 m), Sonora, Mexico, collected on 27 August-2 September 1942 by Charles M. Bogert and preserved in ethanol (Figs 6 - 14 ).

Paratypes.

AMNH R64157, an adult male; AMNH R64158, an adult female; and ROM (Royal Ontario Museum) 53301 (formerly AMNH R64159), an adult female; all with same collecting data as the holotype and all preserved in ethanol.

Referred specimens.

ASU (Arizona State University, Tempe) 6427, ASU 6543-44, ASU 6605-06, ASU 6620-22, ASU 6702-03, ASU 6769, ASU 8534-39, CAS (California Academy of Sciences) 142243, CM (Carnegie Museum) Herps:62200, CNAR ( Colección Nacional de Anfibios y Reptiles)-4002, LACM (Los Angeles County Museum) 105338, LSUMZ (Louisiana State UniversityMuseum of Zoology, Baton Rouge) 34925, MSB (Museum of Southwestern Biology) MSB 41497-99, MVZ (Museum of Vertebrate Zoology) 129943, SMNS (Staatliches Museum fuer Naturkunde, Stuttgart) 7367-68, SMNS 7515, TNHC (Texas Memorial Museum) 60607, UAZ (University of Arizona) 28105, UAZ 35405, UAZ 36875-76, UAZ 56589-PSV, UAZ 56607-PSV, and UIMNH (University of IllinoisMuseum of Natural History) 85836.

Diagnosis.

Molecular data can readily diagnose all species of Gopherus and their hybrids ( Murphy et al. 2011; Edwards et al. 2016). Morphologically, G. evgoodei , G. agassizii and G. morafkai (the Gopherus agassizii group) can be separated generally from both G. flavomarginatus Legler and G. polyphemus (Daudin) in having relatively smaller front feet. Whereas the distance from the bases of the first to fourth claws is the same on all feet in the Gopherus agassizii group, in the latter two species the distance from the bases of the first and third claws on the forelimb is about the same as the distance between the bases of the first and fourth claws on the hindlimb ( Auffenberg and Franz 1978). Living captive specimens of the Gopherus agassizii group and G. berlandieri cannot all be distinguished morphologically because of extensive hybridization ( Edwards et al. 2010) and developmental abnormalities in shell, head and limb integument from poor nutrition ( Donoghue 2006). However, in native non-hybrid individuals, G. berlandieri can be separated from the Gopherus agassizii group in having a wedge-shaped snout when viewed from above in contrast to a rounded snout (Fig. 12 ) ( Auffenberg and Franz 1978). Further, the gular projections of G. berlandieri often diverge in large males and the species often exhibits paired axillary scales preceding each bridge. In contrast, the gular projections do not normally diverge in the Gopherus agassizii group and there is a single axillary scale. Morphological characters among the Gopherus agassizii group exhibit overlap ( Germano 1993; McLuckie et al. 1999) and characters like coloration in desert tortoises can be highly variable ( Legler and Vogt 2013). However, G. evgoodei differs from G. morafkai and G. agassizii (Table 1 ). Gopherus evgoodei is flatter in shell profile (Fig. 2 ). It has rounded foot pads, multiple enlarged spurs on the radial-humeral joint (Fig. 3 ). The new species has short tails (Fig. 4 ), orange tones in the integument (skin) and shell (Fig. 5 ), and a distinctly shallower concavity on the plastron of males.

Description of holotype

(parallels that of G. morafkai by Murphy et al. 2011). An adult male, with carapace length at the midline (MCL) = 209 mm; curved carapace length from nuchal scute to supracaudal scute = 254 mm; plastron from tip of gular horn to tip of anal scutes = 219 mm; plastron from gular notch to anal notch = 202 mm; maximum height of shell at 3rd vertebral scute = 83 mm; width at 3rd/4th marginal scute seam = 137 mm; width at 6th marginal scute = 140 mm; greatest width at mid-8th marginal scute = 158 mm; plastron concavity depth = 10.1 mm; head length = 51.3 mm; and tail = 8 mm. Eleven marginal scutes present on both right and left edges of the carapace. Five toenails present on each forelimb and four toenails on each hind limb. The third nail of each hind limb slightly longer than the others. Multiple enlarged, raised scales present on the anterior ventral surface of each foreleg. No scale “spikiness” on the posterior femoral surface of the rear legs. Scales on head smooth and asymmetrical, with two large pre-frontal scales and smaller scales in the temporal area. Shape of head prefrontal profile rounded/bulging. Shell profile/shape of the carapace appearing nearly flat. Shape of ventral surface of rear feet rounded and lacking projecting, enlarged scales on the posterior plantar surface. Areolae and>17 growth laminae present on all carapacial scutes, although areola are diminishing, especially on the anterior scutes. In alcohol, the color of areolae dark, fading to dark brown with orange hue in outer portion of carapacial scutes. Color of areolae on the plastron dark brown and rest of the plastron medium orange brown. Head and neck tan to dark tan with an orange hue. Skin in the axillary and inguinal areas lighter in coloration; light tan fading to medium tan toward axillary. Nails dark brown, lighter brown at the tips.

Coloration of the species in life.

Gopherus evgoodei may exhibit orange or yellow mottling or spotting on the shell and integument. Because color constitutes a diagnostic feature, these data are given above.

Variation.

As with all species of Gopherus , substantial variability exists among individuals for most morphological features ( Germano 1993; McLuckie et al. 1999). Bogert and Oliver (1945) first recognized the distinct morphology of tortoises at Alamos, but they were unable to quantify it due to small sample sizes. Shell profile is generally flat but may also appear domed in some individuals. Spikiness of scales on forelimbs can vary widely and the shape of the plantar surface of the rear feet, while generally rounded, can be difficult to classify in some cases.

Distribution.

The distribution of G. evgoodei (Fig. 1 ) occupies roughly 24,000 km2 and corresponds to habitat. The species primarily occurs in tropical deciduous forest (TDF) and relatively mesic Sinaloan thornscrub (STS) in the state of Sonora, Mexico, and its distribution extends southward into TDF and the southern part of the STS where it still remains intact in northern Sinaloa south of the Río Fuerte ( Loomis and Geest 1964; Edwards et al. 2015). It also occurs in the TDF of extreme southwestern Chihuahua ( Smith et al. 2004). Thus, G. evgoodei occupies both STS and Sinaloan TDF ( Fritts and Jennings 1994; Berry et al. 2002). The eastern limit of its known range is the foothills of the Sierra Madre Occidental at elevations of 800-1,000 m where the TDF transitions rather abruptly into oak woodlands ( Bury et al. 2002). Although the southern limit of its range remains undetermined, continuous TDF extends along the West Coast of Mexico from Sonora through Sinaloa to Nayarit (>500 km), although it only maintains an average width of 50 km ( Krizman 1972). It is unlikely that G. evgoodei occurs very much further south in Sinaloa than currently known, or as far south as Nayarit, based on an absence of records for this relatively conspicuous and readily recognizable animal, and presumably due to as yet unidentified environmental limitations ( Bury et al. 2002). The northern boundary of G. evgoodei corresponds approximately to the transition from STS to SDS ( Edwards et al. 2015). Although characteristic thornscrub maintains 100% ground cover, where it grades into desertscrub it becomes patchy ( Felger et al. 2001). The transition of TDF and thornscrub to desertscrub dominated by more xeric species often occurs at elevations between 200 and 300 m a.s.l., but with notable exceptions ( Van Devender et al. 2000). Broadly, the distribution approaches the boundary of the Sonoran Desert as defined by Brown and Lowe (1980) and Turner (1982). However, this transition zone is patchy, with a mosaic of SDS and STS. Both G. morafkai and G. evgoodei occur in the more arid, desert-like ecotone-phase of STS, where limited hybridization has been observed ( Edwards et al. 2015; Fig. 1 ). As such, we conservatively estimate the distribution of G. evgoodei by excluding sites where G. evgoodei and G. morafkai come into contact (Fig. 1 ).

Natural history.

Gopherus agassizii , G. morafkai and G. evgoodei appear to have diverged roughly 5.7-5.9 Ma from a common ancestor that was potentially widespread throughout what is now the Mojave, Colorado and Sonoran desert regions (Edwards et al. 2016). Gopherus agassizii likely diverged first via allopatric speciation when the Bouse embayment extended northward between 8-4 Ma ( Lamb et al. 1989). This waterway (now the Colorado River) created a barrier between the Sonoran and Mojave deserts. About the same time, G. morafkai and G. evgoodei began to segregate into tropical and arid ecosystems, possibly under a parapatric model of speciation (ecological isolation), although allopatric speciation owing to climatic change and ephemeral isolation can also explain the split. By the end of the Miocene (5.3 Ma) much of the Sonoran region was likely covered in tropical forests or desert thornscrub but orogenesis initiated the drying trend that lead to the formation of the current North American deserts. The changing environment would have created new arid niches in the northern portion of the ancestral range of the desert tortoise. This could have started the ecological divergence of the three species.

Microhabitat.

Ecologically, G. evgoodei occupies hills and low mountains with at least some large boulders or rock outcrops in the TDF, and the TDF-STS ecotone. Its distribution differs from G. morafkai by its strong association with TDF and STS, as well as its absence from SDS. Similar to G. morafkai , G. evgoodei often associates with slopes where rock outcrops and boulders are common. In TDF, the tortoise generally excavates burrows under already existing boulders or enters and modifies existing rock cavities. In flatter areas where boulders are not be available, it digs burrows in soil, although possibly not as extensively as its congeners. During 2012-2013 surveys in Sonora, only 9 of 44 tortoise burrows (20%) in TDF were in soil. In comparison, 56 of 87 burrows (64%) occurred in soil in STS and SDS. Local variation was not surprising. In northern Sinaloa, Vargas V (1994) reported G. evgoodei used packrat middens, dry cacti and even burrows dug by other animals (e.g. nine-banded armadillo, Dasypus novemcinctus ). Our observations of G. evgoodei , as part of an ongoing radio-telemetry study near Alamos, Sonora, suggested that Goode’s Thornscrub Tortoise uses several burrows a year and exhibits strong site-tenacity, returning to familiar dens year after year (unpublished data), just like its sister-species.

Activity.

Presumably, tortoise activity corresponds with monsoonal rains and vegetation growth ( Bury et al. 2002). Goode’s Thornscrub Tortoise is active from at least June well into November; we lack data on activity during the dry season. In Sonora, the TDF hugs the western edge of the Sierra Madre Occidental and the biome hosts extremely lush vegetation during periods of summer rainfall (July-September). During dry periods, the TDF is almost entirely leafless, but with many spectacularly blooming trees and large columnar cacti ( Krizman 1972; Van Devender et al. 2000).

Little is known about daily activity patterns, reproduction, movements or forage of G. evgoodei . Like other species of Gopherus , their activity relates to forage availability and ambient temperatures. Van Devender et al. (2002) reported that scat from tortoises near Alamos, Sonora contained many species of plants not found in the Sonoran Desert, suggesting differences in foraging activity and selection, although species-availability might also account for this. We observed adults to begin seasonal activity shortly in advance of the growth of forage, usually in June at the leading edge of the monsoons, and enter winter dens by sometime in December and remain underground during the dry, cool winter season (unpublished data).

Etymology.

The new species is a patronym, a noun in the genitive case, in recognition of Eric V. Goode, a conservationist, naturalist, and founder of the Turtle Conservancy. He has contributed generously to the conservation of this species via the preservation of land in Mexico, and he actively pursues the conservation of turtles and tortoises on a global scale. Eric sets an important precedent by complementing this taxonomic description with a tangible action that contributes to the conservation of the new species in its native habitat.

Figures

Figure 1. Conservative estimate of the distribution of G. evgoodei in Mexico indicated by diagonal lines. Desert tortoise range limit, modified based on our field sampling, from Germano et al. (1994). Squares indicate museum and literature records of occurrence of G. spp. Circles are sample locations from Edwards et al. (2015) for both G. morafkai (black) and G. evgoodei (white). Localities in the Sinaloan thornscrub-Sonoran desertscrub ecotone indicated by split circles, which indicate the occurrence of both G. evgoodei and G. morafkai genotypes and / or hybrids.

Figure 2. The flat shell profile / shape of carapace generally distinguishes G. evgoodei from other species of desert tortoises. Live, wild-caught individuals from (i-iv) Rancho Las Cabras and (v-vi) Rancho La Sierrita near Alamos, Sonora, Mexico (in Tropical Deciduous Forest).

Figure 3. The rounded ventral surface of the rear feet (i-ii) and multiple enlarged, raised scales present on surface of forelegs generally (iii-iv) diagnose G. evgoodei in relation to other species of desert tortoises. i-ii same individual in Figure 2 from Rancho Las Cabras near Alamos, Sonora, Mexico (in Tropical Deciduous Forest) iii-iv two individuals from Rancho Las Cabras.

Figure 4. Gopherus evgoodei differs from other species of desert tortoises in having a very short tail. i Rancho El Chupadero east of Guaymas (in thornscrub habitat); ii Rancho Las Cabras; and iii-i Rancho La Sierrita near Alamos, Sonora, Mexico (in Tropical Deciduous Forest).

Figure 5. Gopherus evgoodei differs from other species of desert tortoises in often having yellow / orange integument (skin) and shell. i Rancho El Divisadero ii-iii Rancho Las Cabras; and iv-v Rancho La Sierrita near Alamos, Sonora, Mexico (in Tropical Deciduous Forest).

Figure 6. Dorsal view of the holotype of Gopherus evgoodei, AMNH R 64160. Scale bar 50 mm in 10 mm increments.

Figure 7. Ventral view of the holotype of Gopherus evgoodei, AMNH R 64160. Scale bar 50 mm in 10 mm increments.

Figure 8. Anterior view of the holotype of Gopherus evgoodei, AMNH R 64160.

Figure 9. Posterior view of the holotype of Gopherus evgoodei, AMNH R 64160.

Figure 10. Left lateral view of the holotype of Gopherus evgoodei, AMNH R 64160.

Figure 11. Right lateral view of the holotype of Gopherus evgoodei, AMNH R 64160.

Figure 12. Detail of head scales of the holotype of Gopherus evgoodei, AMNH R 64160.

Figure 13. Detail of the tail of the holotype of Gopherus evgoodei, AMNH R 64160.

Figure 14. Ventral surface of the right rear foot of the holotype of Gopherus evgoodei, AMNH R 64160.