Mastigodryas boddaerti (Sentzen)

Myers, C. W. & Donnelly, M. A., 2008, The Summit Herpetofauna Of Auyantepui, Venezuela: Report From The Robert G. Goelet American Museum-Terramar Expedition, Bulletin of the American Museum of Natural History 2008 (308), pp. 1-147 : 130-137

Mastigodryas boddaerti (Sentzen) Figure 79

Coluber boddaerti Sentzen, 1796: 59 . Holotype and type locality unknown.

Eudryas b. boddaertii: Stuart, 1933: 2 (the double - ii termination used by various authors is an incorrect subsequent spelling).

Dryadophis boddaerti: Stuart, 1941: 66 (in generic revision).

Dryadophis boddaertii: Roze, 1958a: 264 ; 1966: 120 (Auyantepui specimens).

Mastigodryas boddaerti: Peters and Orejas-Miranda, 1970: 192 (new generic combination).

MATERIAL: Auyantepui Plateau, 2200 m (7218 ft.): AMNH R-61018, from the 1937– 1938 AMNH –Phelps Venezuelan Expedition. Camp 2, 1750 m: AMNH R-155716 (shed skin), from the 1994 AMNH –TERRAMAR Expedition to Auyantepui.

The one whole summit specimen (fig. 79) examined is an adult male 1036 mm total length, 286 mm tail length (27.6 % of total length); dorsal scales with paired apical pits, smooth, in 17-17-15 rows. Ventrals 183, subcaudals 105, paired; anal plate divided. Supralabials 9, 2nd–3rd touching loreal, 4th– 6th in eye; infralabials 10, first five touching anterior genials. Temporals 2/1 + 2. Maxil- lary teeth 17 + 2 on right. Color gray, presumably brown in life, with a faint remnant of a pale lateral stripe on scale row 4. (The juvenile pattern of M. boddaerti is very different, consisting of squarish dorsal blotches alternating with lateral ones.)

The shed skin from Camp 2 is fragmentary but identifiable. It has smooth scales with paired apical pits, in 17 dorsal scale rows anteriorly. Shapes of the dorsal head plates correspond to those of the whole summit specimen and another specimen ( AMNH R- 61035) from the base of Auyantepui, and the configuration of 2/1 + 2 temporals (on both sides of the head) is identical in all three.

REMARKS

NOMENCLATURAL COMMENT: The available name boddaerti Sentzen (1796) is affixed

to this species with complete arbitrariness— under a philosophy ‘‘often at a considerable stretch of the imagination … to use as many old names as possible’’ (Stuart, 1933: 1). The name is practically a nomen dubium and could well have been discarded as such. The fate of the original specimen is unknown, the type locality is unknown (the type locality cited in Kornacker [1999: 115] is that of a presumed synonym), and the original ‘‘description’’ could apply to many snakes (some of which are listed in the long synonymy in Boulenger, 1894: 11–12). The only data provided by Sentzen (1796: 59) were a count of ventrals and caudals (‘‘Scut. 181 – 108 5 289’’) and total length and tail measurements in old German feet (‘‘Ganze Länge 2 Fuss, 8 Zoll; Schwanzlänge 7 Zoll, 7 Lin.’’), for which there was more than one system.

Stuart (1941: 66) restricted the name to this species simply because Merrem (1820: facing pages 110) had early used Boddaertische and Boddaerti for a second specimen whose habitat was ‘‘ vorzüglich [especially] Surinam’’, which, in Stuart’s imagination, suggested ‘‘ Dutch Guiana as the type locality’’. In its present sense, the specific name boddaerti has become well established through the influence especially of Stuart (1933, 1941), Roze (1966), and Peters and Orejas-Miranda (1970). A neotype should not be designated until there is actual need for defining the taxon more objectively.

AUYANTEPUI RECORDS: The snake known as Mastigodryas boddaerti has a broad geographic range and appears to be an invader from the lowlands with an established presence on the Auyán plateau. Records are few, but they span many years.

Roze (1958a: 265; 1966: 122) reported four specimens from Auyantepui, including AMNH R-61018 described above, and AMNH R-61035. The AMNH catalog shows the first as collected by W. H. Phelps on the summit at 2200 m elevation, in December 1937 or January 1938 ; the second was taken at at 460 m (base of Auyantepui) in February 1938. A third specimen ( MBUCV 3073 ) was taken by P. Trebbau at 1020 m (Guayaraca) on the southern slope in April 1956. Two additional localities were given by Roze with identical catalog numbers ( MBUCV 3072 ) and identical data except for one listed as ‘‘ Auyantepui’ ’ and one as ‘‘ El Oso , Auyantepui’ ’, both said to have been collected at 1800 m by P. Trebbau on April 24, 1956. Presumably these two entries are a duplication and apply to a single specimen. Aside from being on the southern end of Auyantepui , the location of Trebbau’s 1800 m ‘‘ El Oso’ ’ locality is uncertain. Dunsterville (1965: 170, foldout map) mapped a ‘‘ Bosque del Oso’ ’ but gave an elevation of about 2050 m and equated it roughly with the ‘‘2200-meter camp’’ of the 1937–1938 expedition .

In May 1989, Fabián Michelangeli photographed a cazadora identifiable as Mastigodryas boddaerti on the northern end of the western branch of Auyantepui (inaccurately cited as ‘‘Sima Aonda’’ in A. Michelangeli, 2005: 135, fig. 7 [F. Michelangeli, personal commun.]). During the 1994 expedition, we found only the shed skin at Camp 2, on the eastern branch of the Auyán plateau.

The above records, from 1937 to 1994, suggest that Mastigodryas boddaerti maintains a permanent presence on the summit of Auyantepui.

FAMILY CROTALIDAE Bothrops atrox (Linnaeus) Figure 80

Coluber atrox Linnaeus, 1758: 222 . ‘‘ Habitat in Asia’’ in error.

MATERIAL: West side of Auyantepui summit, 5197 ft. [1584 m], at 5 ° 55956.90N, 62 ° 39932.20W fide R. Guerrero (personal communication). Specimen not preserved, photographed March 22, 2006 by José Miguel Pérez G. (see fig. 80).

The photograph provides the only record of a fer-de-lance from the summit of any tepui.

SUMMARY DISCUSSION AND CONCLUSIONS

Much knowledge has accrued since Hoogmoed (1979b) provided a substantive review of the 408 species of amphibians and reptiles then known from the Guayana (or Guiana) Shield. Among later studies, Duellman (1999: appendix 5.1) increased the known amphibian fauna, and Gorzula and Señaris (1999) reported on the herpetofauna of the Venezuelan Guayana. Recent checklists tabulated 564 species of amphibians and reptiles for the Guayana Shield (Señaris and MacCulloch, 2005; Ávila-Pires, 2005). See McDiarmid and Donnelly (2005) for a review of the highland herpetofauna of Guayana, with specific emphasis on the tepuis.

Myers (2005) mentioned that over 100 species of frogs, lizards, and snakes have been found on sandstone tepuis and certain neighboring granitic mountains. McDiarmid and Donnelly (2005: 159) more specifically enumerated 159 species from these mountains, with a conservative prediction that further exploration will increase the number to over 200 species. Our own attempts at collecting representative samples (see p. 7; also Myers and Donnelly, 2001: 79) from a number of summits reinforce several general conclusions about the composition of tepui herpetofaunas:

(1) Tepuis have relatively depauperate herpetofaunas; (2) neighboring tepuis are likely to have significantly different faunas; (3) tepui endemics outnumber widespread highland species; (4) some endemic species have widespread counterparts, the remnants perhaps of a once widespread tepui fauna; and (5) lowland species find their way onto tepuis in an irregular, unpredictable manner. (Myers and Donnelly, 2001: 79)

Consideration of the Auyán summit fauna allows some elaboration and clarification of these points.

1. Although summit herpetofaunas are depauperate relative to lowland faunas in general, they also are impoverished relative to one another according to a complex of factors that include elevation, area, and habitat. Generally, species diversity is negatively correlated with increasing elevation and positively correlated with increasing area and habitat diversity. The fewest species are to be expected on high (. 2000 m), cold summits that are covered mainly by bare rock and pioneer vegetation (e.g., Cerro Roraima, 34 km 2 with four or five species). Increased habitat diversity can increase faunal size even on high summits of small area (e.g., Cerro Yaví,, 6 km 2 with seven known species), whereas large summits with varying elevation and diverse habitats harbor the largest herpetofaunas (especially Cerro de la Neblina, 235 km 2 with 26 known species).

Auyantepui falls in the last category, with a summit area of 667 km 2 and 24 known species.38 This fauna is relatively large (for a tepui) owing in part to intensity of collecting on the enormous, habitat-rich summit. We document 24 species on the summit—13 anurans, 5 lizards, and 6 snakes—of which 16 species were collected during the 1994 AMNH –TERRAMAR Expedition. Although that dry-season expedition obtained only 67 % of the presently known fauna, it contributed perhaps five or six species that to our knowledge had not been previously collected on the Auyán summit.

2. Myers and Donnelly (2001: 79–80) perceived an emerging generalization that neighboring tepuis may have significantly different faunas, which, although complicat- ed by habitat differences, seemed evident in comparing the faunas of the northwestern tepuis (Yaví, Yutajé-Corocoro, Guanay). Myers (1997) had earlier commented on the striking difference between the herpetofauna of Auyantepui and that of the Chimantá massif. This deserves further comment, inasmuch as the Chimantá massif is highly dissected but nonetheless has a relatively unified fauna.

Auyantepui and Chimantá are massifs of comparable size and similar origin—remnants of the possibly Cretaceous ‘‘Auyántepui planation surface’’ ( Briceño et al., 1990). They are less than 50 km apart at the 2000 m contour and are continuous at the 1000 m contour. The Auyantepui massif is for practical purposes a single large mountain (see p. 14). The Chimantá massif, on the other hand, is so highly eroded that its components comprise 10 adjacent tepuis (11 according to Briceño et al., 1990), whose combined summit area (615 km 2) is smaller

38 Additional to Cerro de la Neblina and Auyantepui, the only other tepui with a known summit fauna exceeding 20 species is Cerro Guaiquinima (1096 km 2 with 22 species at last count [McDiarmid and Donnelly, 1999: 533]). However, Cerro Guaiquinima is not quite comparable. Although there is a peak at about 1680 m, this is an exceptionally low tepui in average height. It is inclined downward from escarpments of about 1300 m elevation to 700 m on the southern side. There are a few highland endemics, but not surprisingly, the herpetofauna of Cerro Guaiquinima consists predominately of lowland species that comprise well over 50 % of the total.

than that of Auyantepui proper (667 km 2) according to data provided by Huber (1995a: 60, table 1–3).

A total of 16 species are known from nine of the 10 or 11 adjacent mountains in the Chimantá massif. The endemic Thamnodynastes chimantá and Stefania ginesi are found on 6–7 of the Chimantá tepuis, which do appear to share a similar herpetofauna that can be treated as a unit, as was done originally by Gorzula (1992) and subsequently by Myers (1997).39 The high headwater and other valleys separating some of the Chimantá tepuis are perhaps not deep enough in some cases, or not old enough in others, to have created significant vicariance within the general herpetofauna. Huber (1992: 26) observed that ‘‘A pesar de su complejidad fisiográfica, el macizo del Chimantá se presenta como un conjunto montañoso bastante homogéneo. [Despite its physiographic complexity, the Chimantá massif is a mountainous ensemble that is quite homogeneous]’’.

Treated as a unit, the herpetofauna of the Chimantá massif also appears homogeneous and very different from that of Auyantepui,

39 Gorzula (1992) listed his specimens as coming from Chimantá camps I–XIX, which were established in the period 1983–1986, as described and mapped by Huber (1992: 31–33, fig. 1–3), who gave an elevational range of 1800–2600 m for the 19 camps. Gorzula and Señaris (1999: 255–256) provided a few summary details and some individual tepui names in a gazetteer, and added Chimantá Camp XXII (2100 m) from 1988. But a bit of confusion was introduced in that coordinates and some elevations for camps I– IV do not quite match the data in Huber (1992); coordinates for Camp I are the same as given for Camp XIV.

McDiarmid and Donnelly (2005) perpetuated some of this confusion. Following the above duplication in Gorzula and Señaris, coordinates (5 ° 169N, 62 ° 099W) belonging to Camp XIV were assigned also to ‘‘Chimantá proper’’ (Camp I); Huber (1992) had described and mapped Camp XIV as situated at 2100 m in a broad valley between the northeast sector of Toronó-tepui and the southeast sector of Chimantá (herein shown as Chi/To in table 10). Coordinates (5 ° 229N, 62 ° 089W) belonging to Camp XIX also were assigned by McDiarmid and Donnelly (2005: 527) to Chimantá-tepui, but, in another place (pp. 526–527), correctly noted as lying between Apacará-tepui and Murey-tepui (herein shown as Ap/Mu in table 10).

Mainly because of locality mix-up, McDiarmid and Donnelly (2005: 527) listed 11 species for Chimantátepui proper, whereas only six are documented (‘‘Chi’’ in table 10).

TABLE 10 Comparisons of the Herpetofaunas of Auyantepui and Chimantá Massif (* 5 endemic species)

less than 50 km to the north-northwest. Comparisons are shown in table 10, in which the taxonomy and intramassif distribution of Chimantá taxa are updated from the basic accounts in Roze (1958b), Barreat et al. (1986), Gorzula (1992), and Gorzula and Señaris (1999). McDiarmid and Donnelly (2005) added several taxa to the faunas of Murey-tepui and especially Tirepón-tepui (which were visited by McDiarmid and Brewer-Carias in 1978).

There are 36 species in the combined faunas of Auyantepui (24 spp.) and the Chimantá massif (16 spp.), of which only four (11 %) wide-ranging species are known to be shared. Auyantepui has 11 endemic species (including endemics not confined to the summit, see below); the Chimantá massif has at least nine endemic species.40 There are 25 genera in total, of which 11 (44 %) are shared, usually by different species. Eight (62 %) of 13 families are shared. Differences in shared families and genera probably could be somewhat diminished by increased collecting on Chimantá (it is difficult to believe that there are no centrolenid frogs there). Nonetheless, the family Tropiduridae (usually conspicuous lizards) is perhaps absent from Chimantá, and the microhylid genus Otophryne (with distinctive call and tadpoles) seems unlikely to have escaped detection on the summit of Auyantepui.

3. Our generalization that endemics in a tepui fauna will outnumber widespread highland species applies to Auyantepui, which has nine species possibly endemic to the summit, three species endemic to the immediate area, and five more broadly distributed highland species, of which only one is a ‘‘tepui species’’. Excluding eight ‘‘invaders’’ from lower elevations (see point 5 below), the Auyán summit harbors a mainly highland fauna, whose species may be grouped as follows:

‘‘WIDESPREAD’’ GUAYANA HIGHLAND ENDEMICS: Hyalinobatrachium crurifasciatum , Hypsiboas sibleszi , Eleutherodactylus pulvinatus , Tepuihyla edelcae , Liophis trebbaui . The first three frogs are upland species found occasionally on tepuis, but Tepuihyla edelcae is known only from tepuis. The snake Liophis trebbaui , once thought to be endemic to Auyantepui, appears to be widely distributed in uplands east of the Río Caroní, from Auyantepui to the Brazilian frontier.

ENDEMIC TO AUYANTEPUI REGION: Anomaloglossus tepuyensis , Hypsiboas jimenezi . These occur on the Auyán slopes and in adjacent lowlands as well as on the summit.

40 Omitted from these considerations are skins of a colubrid snake from Auyantepui and of a gymnophthalmid lizard from Tirepón-tepui in the Chimantá massif. These were included in the species counts in McDiarmid and Donnelly (2005: 511, 523, 529, table 18A.1), but there was no claim that the skins do not represent one of the five colubrids or three gymnophthalmids already scored from Auyantepui and the Chimantá massif, respectively.

POSSIBLE AUYÁN SUMMIT ENDEMICS: Oreophrynella cryptica , Centrolene gorzulai , Hypsiboas angelicus , Stefania schuberti , Eleutherodactylus auricarens , Anadia sp. , Arthrosaura montigena , Tropidurus bogerti , Atractus guerreroi . These nine species are known at present only from the Auyán summit, but some (e.g., C. gorzulai , E. auricarens ) seem likely to inhabit forest at lower elevations and might prove to be endemic to the Auyán massif broadly.41

4. Various authors have noted that a few genera of amphibians and reptiles are represented by obviously related species on different tepuis, a fact of obvious biogeographic significance. However, biogeographic scenarios for the highland herpetofauna are uncomfortably speculative, given the virtual lack of testable phylogenetic hypotheses and absence of molecular studies. Myers and Donnelly (2001: 80) simply observed that ‘‘tepui endemics are likely to have counterparts’’ on other tepuis—defining ‘‘counterparts’’ as a purposely vague term used for congeneric species judged to be phenotypically and ecologically similar, with phylogenetic closeness being probable.

With one notable exception, the Auyán endemics do have endemic congeneric counterparts on other tepuis (or, in the case of Centrolene , at least in other parts of the Guayana highlands), although degree of relatedness is unknown. The exception is the common lizard Tropidurus bogerti , whose phylogenetically closest relatives may occur south of the Amazon. Interestingly, T. bogerti does have a phenotypically and ecologically similar ‘‘counterpart’’ in the form of Plica lumaria on Cerro Guaiquinima.

Among the Auyán summit taxa, at least Oreophrynella (and perhaps Stefania and Tepuihyla ) may represent part of a relatively old and widespread highland fauna in Pantepui. Certain other elements postulated as part

41 There may be an overemphasis here and elsewhere on summit endemics, especially for those species that seem disposed to forest habitats. Lower tepui slopes are extensive (for estimated areas, see Huber, 1995a: table 1–3), and are sometimes forested and well watered. But the terrain tends to be difficult and has not often been collected except by botanists and ornithologists (most serious herpetological collecting commenced after the advent of helicopters).

of this fauna are absent from Auyantepui. The missing taxa include the genus Riolama , species of the Phenacosaurus neblininus complex, and species of the dwarf Thamnodynastes complex, although members of the last two groups occur in the neighboring Chimantá massif. Such speculation assumes (1) that the involved species comprise monophyletic highland groups rather than separate invasions from lower elevations, and (2) that spotty distribution patterns are the result of random extinction events as well incomplete collecting.

5. Lowland species of amphibians and reptiles ‘‘doubtless invade all tepuis, although they may not necessarily reach the highest summits … or, if they do, they may not establish permanent populations there [and although] lowland invaders are expected they can seldom be successfully predicted’’ (Myers and Donnelly, 2001: 81). The summit of Auyantepui has been invaded by at least the following eight species from lower elevations: Hyalinobatrachium taylori , Leptodactylus rugosus , Neusticurus rudis , Anolis chrysolepis planiceps , Chironius fuscus , Leptodeira annulata ashmeadii , Mastigodryas boddaerti , Bothrops atrox .

Of these only the Anolis seems to be a regular invader of tepuis, although there are tepui records for a few of the other genera and species as well (for records, see McDiarmid and Donnelly, 2005; Señaris and Ayarzagüena, 2005: 230, for H. taylori ).

Venomous snakes have rarely been found on tepuis. The present report of Bothrops atrox is only the second record of a viper from a tepui summit, the other being the holotype of Bothrops lichenosa (5 Bothriopsis taeniata ) collected in 1955 on the summit of Chimantá Tepui. Coral snakes occur as rarely; Myers (2005) pictured a specimen of Micrurus psyches from the lower part (1400 m) of the summit of Cerro Guanay.

Whether lowland invaders can maintain permanency on tepui summits is information hard to come by. However, documentation provided herein for Mastigodryas boddaerti indicates that this lowland racer has maintained a presence on the Auyán summit for nearly 60 years at least. And, judged from the number of records, the lowland snake Leptodeira annulata may be a permanent resident of the Chimantá massif, although only a single specimen is known from Auyantepui.

CONCLUDING STATEMENT

Auyantepui has had an interesting history of exploration and is among the best known of the Guayana table mountains, with a summit herpetofauna matched in size only on a few other tepuis having large summit areas. Although every tepui appears to have a distinctive herpetofauna not precisely matched on any other, some generalities have emerged concerning faunal composition. Tepuis even in the same region may have very different herpetofaunas. An exception is provided by the fragmented Chimantá massif, whose adjacent and incompletely isolated tepuis at least appear to share a common herpetofauna. However, although the Chimantá massif and Auyantepui are less than 50 km apart and of the same geological origin, they share only four (11 %) of a combined total of 36 species.

On a given tepui, endemic species are expected to outnumber widespread highland species and, except on the very highest tepuis, a part of the fauna will consist of an unpredictable variety of invaders from lower elevations. Some lowland species may attain relative permanency, as in the case of the snake Mastigodryas boddaerti , which has been known on Auyantepui for nearly 60 years.

Some elements of the hypothesized old Pantepui herpetofauna can probably be identified. However, biogeographic scenarios lack corroboration, which eventually might be provided by revisionary studies with explicit phylogenies based on morphological and, in the best of possible worlds, molecular data. More collecting on more tepuis, with routine tissue sampling, will be necessary before the herpetofauna can be treated to the kind of analyses being provided for botanical taxa in Pantepui (e.g., see Givnish et al., 1997, 2000).