Drymoluber dichrous (Peters, 1863)

Costa, Henrique Caldeira, Moura, Mário Ribeiro & Feio, Renato Neves, 2013, Taxonomic revision of Drymoluber Amaral, 1930 (Serpentes: Colubridae), Zootaxa 3716 (3), pp. 349-394 : 364-374

Drymoluber dichrous (Peters, 1863)

Herpetodryas dichroa Peters, 1863 . Monatsberichte der königlich Akademie der Wissenschaften zu Berlin, 29, p. 284. Syntypes: ZMB 1661, ZMB 1662, ZMB 2603.

Herpetodryas occipitalis Günther, 1868 . Annals and Magazine of Natural History, Fourth Series, 1, p. 420. Holotype: BMNH 1946.1.14.61, formerly 1867.9.17.28.

Spilotes piceus Cope, 1868 . Proceedings of the Academy of Natural Sciences of Philadelphia, 20, p. 105–106. Holotype: ANSP 3920.

Coluber dichrous— Boulenger, 1894. Catalogue of the Snakes in the British Museum (Natural History). Volume II. British Museum of Natural History, London, p. 30–31.

Elaphe dichrous— Gomes, 1918. Memórias do Instituto Butantan, 1, p. 67.

Drymoluber dichrous— Amaral, 1930. Memórias do Instituto Butantan, 4, p. 337.

Lectotype (here designated in accordance with Article 74 of the International Code of Zoological Nomenclature): Museum für Naturkunde Berlin ZMB 1661, adult of undetermined sex (probably a male), SVL 585 mm, TL 242 mm, collected in Brazil during the first half of the 19th century by Georg Wilhelm Freyreiss. Specimen examined by photographs. Although we do not know specifically where specimen ZMB 1661 was collected, the itinerary of its collector is known, and this information leads us to designate it as the lectotype, instead of the syntype ZMB 2603, which was listed as “probably from Suriname ” or ZMB 1662 which has a broken tail.

Paralectotypes: Museum für Naturkunde Berlin ZMB 1662, adult of undetermined sex, SVL 568 mm, TL 237+N mm (broken tail), collected in Brazil during the first half of the 19th century by Georg Wilhelm Freyreiss; ZMB 2603, adult of undetermined sex (probably a female), SVL 623 mm, TL 225+N mm (broken tail), supposedly bought in Suriname. This same information about the collection site of ZMB 2603 is recorded in the catalogue of the Museum für Naturkunde Berlin and on the oldest label of the specimen. However, a newer label (with the name Drymoluber dichrous ) indicates “ Brasilien Becker ”. The reason and source for the adjusted locality is unknown even to the current curator (M. O. Rödel, pers. com.). We examined the specimens from photographs.

About the type locality: There is no information about the dates of collection, shipment to Europe or arrival at the Museum für Naturkunde Berlin of the type series of D. dichrous (M. O. Rödel, pers. com.). Thus, the type locality of D. dichrous was reported as “ Brazil and Surinam ” (e.g. Peters & Orejas-Miranda 1970). Since the lectotype designated above (ZMB 1661) is the new name-bearing type of D. dichrous , its place of collection is the type locality. Despite the lack of detailed information as to where the specimen was collected, we consider the type locality to be the area traveled by its collector, G.W. Freyreiss, in Brazil (Fig. 13 and text below).

Freyreiss was born in Frankfurt on 12 July 1789. In 1813 he left St. Petersburg bound for Brazil, starting his expeditions in June 1814. He departed from the province (currently state) of Rio de Janeiro and travelled south to Minas Gerais, along the Caminho do Proença, a path of the ancient royal road to Vila Rica (currently the municipality of Ouro Preto) (Freyreiss 1907; Papavero 1971) (Fig. 13, localities 1–12). In September 1814, Freyreiss explored the region around the Abaeté, Indaiá and São Francisco Rivers, later returning to Vila Rica (Freyreiss 1907; Papavero 1971) (Fig. 13, localities 12–18–12). In December 1814 he began a second trip to study Indian tribes, travelling from Vila Rica to the vicinities of Presídio São João Batista (currently the municipality of Visconde do Rio Branco) (Fig. 13, localities 12–23). Later he returned to Vila Rica, and then (January 1815) to Rio de Janeiro, leaving no records of his path in this part of the voyage (Freyreiss 1907; Papavero 1971). In July 1815, Maximilian Alexander Philipp, prince of Wied-Neuwied (also known just as Wied) arrived in Brazil, and in August began a natural history trip together with Freyreiss and Friedrich Sellow (Wied 1989; Papavero 1971). From the city of Rio de Janeiro they headed north to Espírito Santo and Bahia provinces (Wied 1989; Papavero 1971). In February 1816, the naturalists were in the Mucuri River, southern Bahia (Fig. 13, localities 1, 24–37). There, Freyreiss decided to return to Espírito Santo, while Wied continued his travel northward (Wied 1989, p. 186). In Espírito Santo, Freyreiss visited again the localities of São Mateus and Linhares (Fig. 13, localities 35 and 36), and in May 1816 he travelled back to Mucuri (Fig. 13, locality 37) to visit Wied (Wied 1989, p. 162, 170, 201). While Freyreiss and Sellow stayed in Mucuri, Wied continued travelling toward the north. In Ocotber 1816, Wied returned to Mucuri to visit Freyreiss and Sellow, with whom he spent three weeks before heading north again (Wied 1989, p. 273). There is no itinerary of the voyages of Freyreiss after that, but it is known that in Bahia he also visited the localities of Caravelas, Canavieiras (Wied 1989, p. 330) and Salvador (Fig. 13, localities 39, 40 and 41), and contributed to the foundation of a German colony, Colônia Leopoldina , currently part of the municipality of Nova Viçosa (Papavero 1971) (Fig. 13, locality 38). There are some inconsistencies related to the place of death of Freyreiss, in 1825. Löfgren (1902) and Papavero (1971) wrote that the German naturalist died in Colônia Leopoldina , while Rocha (1972 “1973”) stated that he died in Europe.

Wied described the species collected during his travels, and his collections are now in the American Museum of Natural History. Even though the type series of D. dichrous is in Germany, we believe that if Freyreiss collected the specimens of D. dichrous during the trip from Rio de Janeiro to Bahia, it probably was done at times when he was not with Wied’s expedition. It is known that from his trips through Espírito Santo and Bahia, Freyreiss sent three shipments of collected specimens to Europe (Papavero, 1971), but, as we have written above, there is no information about the dates when the type series of D. dichrous was collected, shipped to, or arrived in Europe. Additionally, we have record of a single shipment of specimens sent to Europe from his trip through Minas Gerais, when he stated that his collections were sent to Rio de Janeiro on 30 July 1814 (Freyreiss 1907, p. 167). As his travels continued, he certainly made other shipments, of which we have no information.

Diagnosis: Drymoluber dichrous is distinguished from D. brazili and D. apurimacensis by the following combination of characters: a) 15-15-15 dorsal scale rows with two apical pits; b) 157–173 ventrals in males, 160– 180 in females; c) 87–110 subcaudals in males, 86–109 in females; d) 19–26 maxillary teeth. See Table 5.

Comparisons: Drymoluber brazili has 17-17-15 dorsal scales rows, and D. apurimacensis has 13-13-13. Apical pits are absent in D. apurimacensis . Drymoluber brazili has 182–200 ventrals in males and 185–202 in females, 109–127 subcaudals in males and 109–126 in females. Drymoluber apurimacensis is not distinguishable from D. dichrous based on ventrals and subcaudals counts, having 158–164 ventrals in males and 166–182 in females, 84–93 subcaudals in males and 87–91 in females. Drymoluber apurimacensis has 14–16 maxillary teeth.

Small specimens of D. dichrous have dark crossbands 1.5–7 scales wide (mean 3.6) and light interspaces 0.5– 2.5 scales wide (mean 0.8), while in D. apurimacensis the dark crossbands are 1–2 scales wide, and the pale interspaces are 2–3 scales wide. Juvenile specimens of D. brazili have dark crossbands of similar width to those of D. dichrous (2–6 scales; mean 3.6), but the pale interspaces are wider (0.5–5 scales; mean 1.6).

The hemipenes of D. dichrous tend to have more calyces than D. brazili , smaller spinulated flounces, and no spines in the lobular region. The walls of the sulcus spermaticus tend to have more ornamentation, at least in the lobular region, with small jagged papillae. The spines of the asulcate face are generally larger than those of D. brazili , especially those most proximal. The hemipenial morphology of D. dichrous and D. apurimacensis is similar and of little value in differentiating these species.

Description of the lectotype ( Fig. 14 ): Snout-vent length 585 mm, and tail length 242 mm; head distinct from the body, 24.6 mm length (4.2% of the SVL); greatest width of head 12.9 mm (52% of its length); width of head at the supraoculars 9.3 mm; internasal distance 5.1 mm; eye diameter 4.45 mm; eye-nostril distance 4.5 mm. The morphometric measurements were taken by Christoph Kucharzewski, Museum für Naturkunde Berlin. Smooth dorsal scales in 15-15-15 rows, with two apical pits; 161 ventrals and 1 preventral (sensu Peters 1964); cloacal shield entire; tail intact, with 96 divided subcaudals and one terminal spine; rostral wider than high, visible from above; internasals and prefrontals slightly wider than long; each prefrontal contacting the frontal, supraocular, internasals, posterior nasal, preocular and loreal; frontal about 1.5 times longer than wide; supraoculars longer than wide; parietals about 1.5 times longer than wide; nasal divided above and below the naris, mainly in contact with the first supralabial, but also with the second; loreal slightly longer than high, contacting the second and third supralabials; one preocular; two subequal postoculars; three anterior temporals (one upper and two lower) and two posterior temporals (one upper and one lower) on the right side (1/2+1/1); four anterior temporals (two uppers and two lowers) and two posterior temporals (one upper and one lower) on the left side (2/2+1/1); eight supralabials, the fourth and the fifth contacting the eye; mental triangular, wider than long; nine infralabials, the first pair in contact behind the mental; first to fifth infralabials in contact with the first pair of chinshields; fifth and sixth infralabials in contact with the second pair of chinshields; sixth to ninth infralabials contacting the gulars; first pair of chinshields about the half of the length of the second.

FIGURE 13. Itinerary of the travels made by Georg Wilhelm Freyreiss in Brazil, in the 19th century, when the lectotype (ZMB 1661) and one paralectotype (ZMB 1662) of Drymoluber dichrous were collected. The inset map shows South America, highlighting in gray the current states of Rio de Janeiro (RJ), Espírito Santo (ES), Minas Gerais (MG) and Bahia (BA), visited by Freyreiss. For easy viewing, only the main localities are represented in the main map. For complete lists of localities, see Freyreiss (1907) and Wied (1989), Bokermann (1957) and Papavero (1971). When the old and current names of a toponym are different, the later is written inside brackets. 1 = Praia dos Mineiros , Rio de Janeiro; 2 = Porto Estrela (Magé); 3 = Fazenda Mandioca (Magé); 4 = Fazenda do Padre Correia ( Corrêas, Petrópolis ); 5 = Fazenda das Sebollas (Inconfidência, Paraíba do Sul); 6 = Rio Paraibuna; 7 = Matias Barbosa; 8 = Juiz de Fora; 9 = Chapéu D’Uvas (Juiz de Fora); 10 = Barbacena; 11 = Congonhas do Campo (Congonhas); 12 = Villa Rica (Ouro Preto); 13 = Ponte das Almoreiras (Ponte das Almorreimas, Brumadinho); 14 = Fazenda São Joanico (Maravilhas); 15 = Pompeu; 16 = Rio São Francisco; 17 = Fazenda do Comandante de Indaiá (Quartel Geral); 18 = Quartel do Assunção (Córrego dos Tiros , Tiros); 19 = Mariana; 20 = Santana dos Ferros (Guaraciaba) ; 21 = Santa Rita (Viçosa); 22 = Serra de S. Beralde (Serra de São Geraldo, São Geraldo); 23 = Presídio de São João Batista (Visconde do Rio Branco); 24 = S. Gonzalves (São Gonçalo); 25 = Freguesia de Maricá (Maricá); 26 = Araruama; 27 = Cabo Frio; 28 = Villa de S. João de Macahé (Macaé); 29 = Villa de S. Salvador dos Campos dos Goytacazes (Campos dos Goytacazes ); 30 = Itapemerim; 31 = Povoação de Piuma (Piúma); 32 = Nossa Senhora da Victoria (Vitória); 33 = Vila Nova do Almeida (Nova Almeida, Serra); 34 = Quartel do Riacho (Riacho, Aracruz); 35 = Linhares; 36 = Barra do São Mateus (São Mateus); 37 = Rio Mucuri (Mucuri); 38 = Colônia Leopoldina (Nova Viçosa); 39 = Caravelas; 40 = Canavieiras; 41 = Salvador. Dashed line = Travel from Rio de Janeiro to Quartel do Assunção (Tiros), in 1814 (Localities 1–12, 12–11, 11–18); Dotted line = Travel from Villa Rica (Ouro Preto) to the vicinities of Presídio São João Batista (Visconde do Rio Branco), from December 1814 to January 1815 (Localities 12–23); Dashed-dotted line = Travel from Rio de Janeiro to Mucuri, from July 1815 to 1816 (Localities 1, 24–38). Localities 39–41 are not linked by lines because of the lack of information about the itinerary took by Freyreiss when visiting them. Tropical and subtropical moist forests; Tropical and subtropical savannas; Deserts and xeric formations; Mangroves; Tropical and subtropical dry forests. Habitat types follow Olson et al. (2001).

Coloration of the lectotype: Peters (1863) described the coloration of Herpetodryas dichroa as: “Dorsum olive-brown, the sides, including the lateral edges of ventral plates; subcaudals olive-green; the whole venter to the tip of the tail of yellow color; a black stripe bordered by the yellow (darker than that of the belly) supralabials on both sides of the head, becoming suddenly wider behind the eyes.”. After almost 200 years of preservation, the dorsum is uniformly brownish-blue. Supralabials, gular region and venter are uniformly creamish colored, the lateral edges of ventrals and subcaudals are the same color as the dorsum. The upper margins of some supralabials, especially the last two, have the color of the dorsum. The dark lateral stripe of the head is indistinct.

Coloration of preserved adults: The dorsal coloration of adult specimens after fixation is usually darker than the lectotype, with bluish-gray or dark-blue colors. Faint crossbands that formed the juvenile color pattern are visible in smaller specimens.

The venter of most specimens is immaculate cream, with lateral edges of ventral plates having the same color as the dorsal scales (n=203; 95%). In some cases (n=3; 1.5%) the dark dorsal coloration continues into the ventral region from the lateral edges of ventral plates. There are specimens with a creamish venter with small black marks (n=3; 1.5%). The venter of some specimens is the same color as dorsal scales or a little paler (n=3; 1.5%). One specimen (0.05%) has a yellow venter, with the lateral and the posterior edges of ventral plates darkened.

The ventral part of the tail usually is creamish with lateral edges of subcaudals darkened (n=207; 97%), with little variation. The lateral edges may not be darkened (n=1; 0.05%), the subcaudals may have small black dots, in larger numbers in the posterior region (n=1; 0.05%), or be completely black along the whole tail (n=4; 2%).

The gular region is pale and immaculate in most specimens (n=153; 72%). Some of them, however, have dark marks in the posterior edges of infralabials, and sometimes on the chinshields (n=60; 28%). The lateral edges of supralabials (especially the last ones) in most specimens are dark colored (n=166; 78%), sometimes (n=14; 6.5%), with strong and thick marks. Supralabials can also be completely dark (n=15; 7%) or totally cream, without marks, or with inconspicuous marks (n=18; 8.5%).

Coloration of adults in life: Based on descriptions in the literature (e.g. Cunha & Nascimento 1978; Martins & Oliveira 1998; Bartlett & Bartlett 2003; Argôlo 2004a) and some photographs ( Fig. 15 ), we noted some variation in the color pattern of adult specimens of D. dichrous in life—but unrelated to the geographic distribution. The dorsum varies between brown, olive-brown, green, dark-green and gray-bluish. In some snakes the dorsal color changes posterior to the first third or the half of the body. The dorsum of head is sometimes a little paler than the body. A barely distinguishable black stripe from the preocular to the end of the posterior temporal may be present. Supralabials, gular region and the venter vary between yellow and white, with the lateral edges of ventrals and subcaudals of the same color or little paler than the dorsals. The supralabials also may be dark along their upper and lateral edges.

Coloration of preserved juveniles: The number of dark crossbands along the body varies between 31 and 52 (mean 40; SD=4.3; n=49; 63%). Specimens with indistinct bands on the posterior third of the body are common (n=29; 37%), and in seven small specimens (SVL 285–391 mm), the banded coloration has disappeared completely. The tail bands are rarely visible (only in some specimens with SVL <385 mm). Dorsal crossbands vary from de 1.5–7 vertebral/paravertebral scales wide (mean 3.6; SD=0.87; n=306 crossbands).

The last crossband anterior to the cloacal shield tends to be the narrowest (1.5–5 scales; mean 2.9; SD=0.79; n=48) and the fifteenth after the head the widest (2–7 scales; mean 3.6; SD=0.75; n=71). The light interspaces between dark crossbands vary between 0.5–2.5 scales (mean 0.8; SD=0.43; n=306 interspaces). The interspaces anterior to the last crossband and the fifth crossband anterior to the cloacal shield are the narrowest, with 0.5–1 scales (mean 0.5; SD=0.1; n=48; and mean 0.5; SD=0.09; n=49). The interspace posterior to the first crossband and the interspace anterior to the fifth crossband are the widest, with 0.5–2.5 scales (mean 1.17; SD=0.49; n=67; and mean 0.96; SD=0.46; n=71).

The venter of juveniles usually has a creamish coloration as in adults (n=66; 85 %). It also may have black marks spreading ventromedially from the lateral edges of ventral plates (n=9; 11%), or even be completely black along the whole body (n=1; 1.5%), or only on the posterior half (n=2; 2.5%).

The subcaudals in juveniles are the same pattern as those in adults (cream color with darkened lateral edges) (n=73; 93.5%), although in some specimens these plates are completely black (n=5; 6.5%).

The head pattern consists of light internasals; light prefrontals with darkened posterior edge; and dark frontal and supraoculars with light anterior edge. A light stripe in the parietal region may be present and immaculate (n=19; 26.1%), maculate (n=5; 6.8%), or absent (n=49; 67.1%) (Fig. 13).

Like in the adults, the gular region of the juveniles usually is cream colored (n=48; 61.5%), but black marks may occur on the infralabials and chinshields (n=30; 38.5%). The supralabials have distinct dark marks on their lateral edges (n=51; 65.4%), but sometimes these marks are pale, as in adults (n=8; 10.25%), or even barely visible (n=19; 24.35%).

Coloration of juveniles in life: The light colored regions of preserved juveniles may vary from shades of white, cream, light-brown and orange in life. The dark colored regions of preserved specimens vary in shades of brown, orange-brown, reddish-brown and grayish-brown ( Fig. 16 ).

Drymoluber apurimacensis Drymoluber brazili Drymoluber dichrous

Geographic distribution South of the Apurimác and Cerrado, Caatinga, Atlantic Forests of the lower Doce Eastern Andes, Amazonia, Guiana Shield, Pampas Rivers, in the Peruvian River, and Atlantic Forest–Cerrado transitional areas. Brazilian Atlantic Forest, Atlantic Forest– Serranía Esteparia ecoregion. Caatinga and Atlantic Forest–Cerrado transitional areas.

Maximum snout-vent length 670 mm (MHNSM 18647) 1178 mm (IBSP 34369) 1050 mm (MPEG 17235)

Number of dorsal scale rows 13–13–13 17–17–15 15–15–15 (n=289); 15–15–17 (n=1); 15–15–13 (n=1)

Dorsal apical pits 0 2 2

Cloacal shield Entire Entire (n=82); Divided (n=1) Entire (n=290); Divided (n=1)

Number of pre-ventrals 1 (n=1); 2 (n=2) 1 (n=74); 2 (n=8); 3 (n=1) 0 (n=12); 1 (n=277); 2 (n=2)

Number of ventrals in males 158–164* (n=2) 182–200 (mean 189; SD=3.61; n=46) 157–173 (mean 165; SD=3.35; n=172)

Number of ventrals in females 166–182 (n=2) 185–202 (mean 193; SD=3,68; n=37) 160–180 (mean 172; SD=3.72; n=114)

Number of subcaudals in males 84–93 (mean 88; SD=4.58; n=3) 109–127 (mean 118; SD=4.37; n=29) 87–110 (mean 97; SD=4.20; n=122)

Number of subcaudals in females 87–91 (mean 89; SD=2.08; n=3) 109–126 (mean 119; SD=4.45; n=26) 86–109 (mean 98; SD=4.79; n=89)

Loreal longer than high longer than high longer than high (n=193); as high as long

(n=92); higher than long (n=5)

Preoculars 1 1 1 (n=572 sides); 2 (n=10 sides)

Postoculars 2 2 (n=159 sides); 3 (n=5 sides) 1 (n=1 side); 2 (n=551 sides); 3 (n=28) sides

Temporal formula 1+1/1 1+1/1 (n=9 sides), 1+2/3 (n=1 side), 1/1+1 (n=1 side), 1+1 (1 side), 1+1/1 (n=38 sides), 1/1+1 (n=2 1/1+1/1 (n=90 sides), 1/1+1/2 (n=8 sides), 1/1+2/1 (n=10 sides), 1/1+1/1 (n=480 sides), 1/1+1/2 (n=7 sides), 1/1+2/2 (n=1 side), 1/1+3/1 (n=1 side), 1/2+1/1 sides), 1/1+2/1 (n=8 sides), 1/1+2/2 (n=1 (n=19 sides), 1/2+1/2 (n=5 sides), 1/2+2/1 (n=5 sides), side), 1/2+1/1 (n=20 sides), 2/1+1/1 (n= 14 1/2 +2/1/2 (n=1 side), 1/2+3/1 (n=1 side), 1/2+4/1 (n=1 sides), 2/1+1/2 (n=1 side) or 2/2+1/1 (n=4 side), 2/1+1/1 (n=5 sides), 2/1+1/2 (n=3 sides), 2/2+2/2 sides)

(n=1 side), 2/2+4/2 (n=1 side), 2/3+1/2 (n=1 side)

……continued on the next page iv–vi (n=34 sides) vi–vii (n=2 sides) Infralabials contacting the gulars v–vii (n=2 sides); iii–viii (n=1 side); iv–viii (n=1 side); v–viii (n=4 sides); v–ix (n=7 sides);

v–viii (n=4 sides) ** v–viii (n=4 sides); v–ix (n=143 sides); vi–viii (n=75 sides); vi–ix (n=485 sides);

v–x (n=6 sides); vi–ix (n=4 sides); vi–x (n=2 sides); vii–x (n=6 sides)

vi–x (n=4 sides)

Maxillary teeth 14–16 (mean =15; SD=1) 19–25 (mean =23; SD=1.05) 19–26 (mean =23; SD=1.21) Lehr et al. (2004) cited 182 ventrals for MTKD 44669 (male), but this specimen is in poor condition ( Fig. 23 B), with at least three regions where the scale count is impaired; thus, we do not consider the countings of Lehr et al. (2004) in the current work.

We have no information concerning the contact between infralabials, chinshields and gulars for MHNSM 18647, which, according to Lehr et al. (2004) has nine infralabials.

Hemipenial morphology (n=17) ( Fig. 17 ): Hemipenis single, subcylindrical, not capitate. Sulcus spermaticus single and centrolinear. Lobe about half of the hemipenis length, with papillate calyces (5–10 triangular papillae per calyx). Proximally, the calyces are gradually replaced by spinulate flounces and spines. Body covered by spines arranged in more or less transverse rows (about 60–70 spines in total). Walls of the sulcus spermaticus ornamented at least in the lobular region by jagged papillae (sometimes in small number), and some spinules. These walls are also bordered on both sides by a longitudinal row of 6–10 spines (mean 7; SD=1.22) increasing in size toward the proximal region. The spines from the left side may increase in size up to the middle of the row, and then decrease in size toward the proximal region. A hook is present at the end of each row of spines bordering the sulcus. Both left (n=11) and right (n=2) hooks can be located more proximally than the other hook, or reach the same position in the hemipenis base (n=4). None (n=8), or three to four small spines (n=7) or spinules (n=2) may be present between the left hook and the wall of the sulcus spermaticus.

One (n=10) or two (n=2) lateral spines, larger than the hooks, may be present to the left of the sulcus spermaticus, or a single large lateral spine may be present on both sides of the sulcus (n=2). In some cases (n=3), lateral spines are smaller than the hooks. The asulcate face is formed by spines arranged in 5–7 more or less transverse rows (counted from proximal to distal region), with largest spines in the median rows. The base of the hemipenis is smooth, with some grooves and several sparse spinules.

Variation: Largest male with SVL 1050 mm, TL 340+N mm (MPEG 17235); largest female with SVL 801 mm, TL 89+N mm (IBSP 2198). Seventy seven (77) examined specimens (26.5%) had broken tail. The tail of 54 of those specimens (18.5% of the total examined) was healed, suggesting that the breakage did not occur during collection or preservation, and is evidence of the presence of pseudoautotomy in this species, a defensive behavior already recorded in several other Neotropical Colubrinae , including Dendrophidion , Drymobius , Mastigodryas and Scaphiodontophis (Mendelson III 1992; Slowinski & Savage 1995; Prudente et al. 2007; Leite et al. 2009). Tail length of specimens with complete tail is 25.9–48.4% of the SVL in males (mean 40.6%; SD=3.18; n=122) and 35.4–65.3% in females (mean 41%; SD=3.75; n=89). For variation in meristic characters, see Table 5.

Geographic distribution: Drymoluber dichrous inhabits the eastern portion of the Andean mountain range ( Peru and Ecuador), the Amazonia and Guiana Shield ( Bolivia, Peru, Ecuador, Colombia, Venezuela, Brazil, Guyana, Suriname and French Guiana), and the Atlantic Forest and its transitional areas with the Caatinga (brejos nordestinos) and Cerrado domains ( Fig. 12 ; Appendices II and III). The Andes of Colombia and the Venezuelan Llanos may limit the northward expansion of D. dichrous , while the transition from tropical to subtropical climate below the Tropic of Capricorn may act as a southern barrier. The Andes may also limit its western distribution, while in the east the species reaches the coast. The elevational distribution of D. dichrous is extremely wide, varying from sea level to about 3500 meters.

The absence of morphological characters distinguishing the Amazonian and Atlantic Forest populations of Drymoluber dichrous could be due to genetic continuity maintained until the Pleistocene, when both forested regions were in contact; today this past area of contact is located in the Caatinga (Vanzolini 1981; Rodrigues 1990; Costa 2003) and part of the Cerrado (Ledru 1993; Costa 2003). Future phylogenetic analyses can help solving the issue concerning where was the bridge between the Amazonia and the Atlanctic Forest that maintained the genetic flux of D. dichrous in the past.

Natural history: Drymoluber dichrous inhabits mainly forests with additional records from natural savannas and environments with a relative degree of anthropization ( Cunha & Nascimento 1978; Martins & Oliveira 1998; Argôlo 2004a; França et al. 2006). Drymoluber dichrous is diurnal, forages on the ground and rests on the vegetation at night ( Cunha & Nascimento 1978; Martins & Oliveira 1998). Its diet seems to be represented mainly by terrestrial anurans and lizards (including their eggs), but also snakes (even cospecifics). The following taxa were recorded as prey of D. dichrous : Adelophryne maranguapensis , Alobates aff. marchesianus, Anomaloglossus (= Colostethus ) sp., Ischnocnema cf. ramagii , Leptodactylus sp., L. didymus and Physalaemus gr. cuvieri (Anurans); Ameiva ameiva, Anolis fuscoauratus , Arthrosaura reticulata , Cercosaura eigenmanni , Ce. ocellata, Coleodactylus meridionalis , Colobosauroides cearensis , Gonatodes hasemani , Go. humeralis , Iphisa elegans , Kentropyx calcarata , Leposoma percarinatum , Le. baturitensis, Neusticurus ecpleopus and Placosoma sp. (Lizards); Oxybelis sp. and Thamnodynastes hypoconia (Snakes) ( Cunha & Nascimento 1978; Duellman 1978, 2005; Dixon & Soini 1986; Avila-Pires 1995; Martins & Oliveira 1998; Vitt et al. 2000; Borges-Nojosa & Lima 2001; Pinto 2006; Bernarde & Abe 2010; Veríssimo et al. 2012). During the present study, three newborns of an opossum species ( Didelphidae ) and the remains of an amphisbaenian ( Amphisbaenidae ) where found in the stomach of a previously dissected specimen (MPEG 2670, male, 920 mm SVL). These food items constitute new registers for the species’ diet.

Information about reproduction in D. dichrous is scarce and restricted to Amazonian specimens. Drymoluber dichrous is oviparous and there are reports of 2– 6 eggs per clutch (Fitch 1970; Martins & Oliveira 1998). Three females examined in this study had four (MPEG 10832, MPEG 16551, MPEG 19007) and one (MPEG 20330) had five eggs. This information, together with other literature records, indicates an average of four eggs per clutch. The reproductive season in Amazonia seems to be prolonged throughout the year, according data from Fitch (1970) and Martins & Oliveira (1998).

Drymoluber dichrous has several defensive behaviors. When handled, individuals will rotate their body, vibrate their tail and occasionally bite (Martins & Oliveira 1998). Martins et al. (2008) observed three types of defensive behaviors directed to visually oriented predators: crypsis, head elevation and neck S-coil. Brodie III & Brodie Jr. (2004) suggested that young cross-banded specimens might mimic coralsnakes ( Elapidae ).

As mentioned above, the long tail of D. dichrous (mean 40.6% of the SVL in males and 41% in females), with a tail breakage ratio of 18.5% (considering only those specimens where breakage occurred before collection) is evidence of the defensive behavior known as pseudoautotomy (Slowinski & Savage 1995). This defensive behavior may explain the long tail in this primarily terrestrial species.

The presence of sexual dimorphism in D. dichrous , with males larger than females, was already observed by Fitch (1981). The longer tail of males of D. dichrous (mean 277 mm) related to females (mean 253 mm) is a common character in snakes (Shine et al. 1999; Pizzatto et al. 2007). Future research on the reproductive biology of the species would provide more information and a better understanding of these topics.

Etymology: Peters (1863) did not comment on the reasons that led him to choose the name dichroa for the species he described. The words dichroa / dichrous comes from the ancient Greek, meaning “two-colored” or “two-skins”, and we suggest that the name may be an allusion to the contrasting color of adult specimens, which dorsum is dark colored, while the venter is light colored.

Figures

FIGURE 12. Geographical distribution of Drymoluber Amaral, 1930. Black symbols represent localities with specimens examined, and white symbols represent literature records. Circles = Drymoluber dichrous (Peters, 1863); Triangles = Drymoluber brazili (Gomes, 1918); Lozenges = Drymoluber apurimacensis Lehr, Carrillo & Hocking, 2004. Tropical and subtropical moist forests; Tropical and subtropical savannas; Deserts and xeric formations; Mangroves; Floodplains;

FIGURE 14. Lectotype of Drymoluber dichrous (Peters, 1863) (ZMB 1661). A) dorsal view of body; B) ventral view of body; C) right side of the head; D) left side of the head; E) dorsum of head; F) gular region. Photos: Mark-Oliver Rödel.

FIGURE 15. Color in life of some adult specimens of Drymoluber dichrous (Peters, 1863). A) Barra do Choça, Bahia, Brazil; B) Reserva Extrativista Riozinho da Liberdade, Tarauacá, Acre, Brazil; C) Reserva Ducke, Manaus, Amazonas, Brazil; D) Nouragues Station, French Guiana; E) Parque Estadual Guajará - Mirim, Rondônia, Brazil; F) Marabá, Pará, Brazil. Photos: Marco Antônio de Freitas (A), Paulo Sérgio Bernarde (B), Rafael de Fraga (C), Maël Dewynter (D), Laurie J. Vitt (E) and Pedro L. V. Peloso (F).

FIGURE 16. Color in life of some juvenile specimens of Drymoluber dichrous (Peters, 1863). A) Reserva Ducke, Manaus, Amazonas, Brazil; B) Reserva Ducke, Manaus, Amazonas, Brazil; C) Barra do Choça, Bahia, Brazil; D) Ubajara, Ceará, Brazil. Photos: Rafael de Fraga (A), William Quatman (B), Marco Antônio de Freitas (C) and Daniel Loebmann (D).

FIGURE 17. Hemipenis of Drymoluber dichrous (FMNH 40206, La Pampa, Madre de Dios, Peru). A) Sulcate face; B) Asulcate face. Scale bar = 1 cm. Photos: Henrique C. Costa.

FIGURE 23. Four of the five paratypes of Drymoluber apurimacensis Lehr, Carrillo & Hocking, 2004. A) FMNH 81542; B) MTKD 44669; C) MTKD 45192; D) MTKD 45193. Photos: Henrique C. Costa (A) and Barbara Bastian (B – D).