Testudo scabra Linnaeus 1758

Rhodin, Anders G. J. & Carr, John L., 2009, A quarter millenium of uses and misuses of the turtle name Testudo scabra: Identification of the type specimens of T. scabra Linnaeus 1758 (= Rhinoclemmys punctularia) and T. scripta Thunberg in Schoepff 1792 (= Trachemys scripta scripta), Zootaxa 2226, pp. 1-18 : 8-12

Testudo scabra Linnaeus 1758

The original type specimen (holotype) of T. scabra L. 1758 was donated to Carl Linnaeus for his personal collection by Jonas Alströmer in 1749 [Alströmer was a Swedish industrialist and co-founder with Linnaeus of the Swedish Academy of Sciences] and became part of the combined Alströmer/Linnaeus collections later donated by Linnaeus to the Museum in Uppsala ( Linné and Thunberg 1780; Thunberg 1785 –1817, 1787a, 1808–1815; Lönnberg 1896; Holm 1957; Wallin 1992). After Linnaeus died in 1778, the specimen was catalogued in 1780 by his son and curatorial successor, Carl von Linné, Jr., as being in the Linnaean collection as their single specimen of “ Testudo scabra ” ( Linné and Thunberg 1780), and that it was at that time stored in alcohol. After Linné died in 1784, this specimen was catalogued as “ T. scabra α Α L. ” by his curatorial successor, Carl Petter Thunberg (Thunberg 1808 –1815) indicating that the specimen had been donated by Alströmer, and distinguishing it from two specimens, “ T. scabra β Th. " and “ T. scabra γ Th. ," donated to the collection in 1785 by Thunberg himself ( Thunberg 1785 –1817). Later, the original Alströmer specimen was labeled “ Testudo scabra α Linn. Mus .” by Thunberg, reaffirming that it had come specifically from the Alströmer/Linnaeus donation ( Wallin 1992). Thunberg labeled the two subsequent specimens of what he considered to be T. scabra as “ T. scabra β Mus. Thunb. ” and “ T. scabra γ Mus. Thunb. ” Sometime later, but prior to 1817, Thunberg donated another specimen that he catalogued as “ T. scabra δ Mus. Thunb. ” ( Thunberg 1785 –1817). The same catalogues ( Thunberg 1785 –1817, 1808–1815) also list the donation by Thunberg of a single specimen of “ T. scripta Th. ”, the type of Testudo scripta Schoepff 1792 (see below).

Sometime during the 1800s the holotype was removed from alcohol storage and dried out. Lönnberg (1896:34) examined the type specimen of T. scabra L. and noted that it was “quite young, dried and in a very bad condition.” He noted further that Boulenger (1889:121) had synonymized T. scabra L. with a question mark under Nicoria trijuga (presently Melanochelys trijuga ), but disagreed with this, noting that “it is possible that [the type] is a Nicoria , but it is not Nicoria trijuga .” However, most subsequent authors followed Boulenger’s lead (without examining the specimen) and continued to place T. scabra L. in the synonymy of trijuga .

FIGURE 5. Type specimen (holotype) of Testudo scabra Linnaeus 1758:198 , donated by Jonas Alströmer in 1749 to Carl Linnaeus and then later by Linnaeus to the Uppsala University Museum of Zoology (UUZM Linné Collection 129), dried hatchling, ca. 44 mm straight CL. The original tag by Carl Petter Thunberg from ca. 1785 reads “ Testudo scabra α Mus . Linn.” This specimen represents a Rhinoclemmys punctularia .

Holm (1957) noted that the holotype was still present in the Uppsala museum and had by then been catalogued as Linné Collection no. 129. The specimen (with its original Thunberg tag) is still extant in the collection, and is badly preserved as noted by Lönnberg (1896).

Examination of the holotype of Testudo scabra Linnaeus 1758 (UUZM Linné Collection no. 129), illustrated here for the first time (Fig. 5), indicates that it is not referable to eight of the nine nominal species with which the name scabra has previously been synonymized, i.e., Melanochelys trijuga , Pelomedusa subrufa , Emys orbicularis , Trachemys scripta , Glyptemys insculpta , Cylindraspis vosmaeri , C. peltastes , or C. indica . We have instead identified the specimen as belonging to the ninth species with which it has previously been synonymized, Rhinoclemmys punctularia ( Daudin 1801) , based on comparison with all nine species of Rhinoclemmys and 13 genera of Asian Geoemydidae ( Carr 1991; Vetter and van Dijk 2006; Rueda- Almonacid et al. 2007).

The specimen itself is a small, dry, somewhat shriveled hatchling currently measuring approximately 44 mm straight CL (Fig. 5). The carapace dorsum is brown, with the undersides of the marginal scutes and plastron periphery a yellow color, with a large black figure in the center of the plastron. Soft parts of the limbs, head and neck are shriveled and appear nearly black, except that the dorsal and lateral portions of the head exhibit lighter-colored (yellowish) stripes. Diagnostically important parts of the head color pattern and bridge region are somewhat obscured by the poorly-preserved state of the specimen; however, through a careful and thorough examination of many externally visible characteristics of the specimen we were able to restrict the identification to a single species of Rhinoclemmys .

The following combination of characters identify the holotype as belonging to the testudinoid family Geoemydidae rather than the Emydidae or Testudinidae : 1) contact between marginal scutes and the pectoral and abdominal scutes on the bridge; 2) hatchling size> 40 mm; 3) presence of a nuchal scute; 4) presence of a vertebral keel; 5) presence of two axillary scutes; 6) plastral coloration a dark, centrally located, longitudinal band that extends from the gular scutes to the posterior end of the plastron; and 7) presence of five separate digits of the manus.

Now restricted to the Geoemydidae , and acknowledging that no single characteristic is diagnostic for the genus, we can further identify the specimen as belonging to Rhinoclemmys based on the following combination of characteristics: 1) lateral keel present in the form of a discontinuous, longitudinally oriented keratinous ridge on the costal scute areolae; 2) relatively weak degree of serration of the posterior marginal scutes; 3) presence of an anal notch; 4) dark central plastral coloration not extending across the anterior portion of the gular scutes, but extending around the anal notch; 5) presence of two relatively small axillary scutes; 6) contact between the humeral scute and posterior axillary scute; 7) ventral portions of marginal scutes 4, 5, and 6 contact the plastral scutes (pectoral and/or abdominal) on the bridge (and marginal 7 would apparently contact the abdominal); and 8) presence of a narrow zygomatic arch (“excavated dorsally and ventrally” as described by Feuer 1970).

Features of the specimen that indicate probable identification as R. punctularia are [comparison to other nominal species of Rhinoclemmys in brackets]: 1) the CL of approximately 44–49 mm is close to the size range known for hatchlings of R. punctularia , i.e., smallest = 47 mm and average = 56–58 mm ( Ewert 1979; Pritchard and Trebbau 1984) [other known hatchling sizes are: 35–51 mm ( R. pulcherrima ), 46 mm ( R. diademata ), 52 mm (mean for R. rubida ), 39–59 mm ( R. melanosterna ), 55 mm (mean for R. areolata ), 63 mm (mean for R. funerea ), and 64 mm (mean for R. annulata ) (mostly Ewert 1979)]; 2) the maxillary tomium is notched in the midline with a cusp on each side [characteristic of all but R. annulata and R. rubida ]; 3) the dark central plastron color does not extend across the anterior portion of the gular scutes, but does extend around the anal notch, with a yellow margin laterally and anteriorly [typical of most of the genus, but it is typically narrower in R. pulcherrima and R. rubida ; a different, blotched pattern is found in R. nasuta ]; 4) the relative lack of dark pigmentation on the ventral surfaces of the marginal scutes [more extensive in R. annulata , R. diademata , R. funerea , R. nasuta , R. rubida ; also dark, or with ocelli in R. pulcherrima ]; 5) presence of dark blotches on the lateral pectoral and abdominal scutes and adjacent marginals [typical of several species, but R. annulata , R. pulcherrima and R. rubida have the bridge portions of the pectoral and abdominal scutes nearly uniformly dark in coloration; and dark coloration is often absent from the area in R. areolata and R. melanosterna ]; 6) axillary scutes contact marginal scutes 2, 3, and 4 [this is the most common character state only in R. melanosterna and punctularia ; more common in all other species is contact with marginals 3 and 4 only]; 7) interhumeral seam length is less than the intergular seam length [characteristic of all species except R. annulata ]; 8) interanal seam length is greater than the interfemoral seam length [also characteristic of R. diademata and R. rubida ; all other species are the converse]; 9) seam B contact (intercostal seam between C1 and C2) ( Tinkle 1962) on anterior marginal 5 [the dominant character state in all species except R. annulata and R. nasuta ]; and 10) seam C contact (intercostal seam between C2 and C3) ( Tinkle 1962) on anterior marginal 7 [the most common character state in all species except R. funerea , R. melanosterna , and R. nasuta ].

In addition, the color pattern of the head apparently includes both supratemporal and postorbital stripes, both of which are light-colored stripes on the dark ground color that is typical of all species of Rhinoclemmys . The dorsum of the specimen’s head appears to be covered with an extensive, light-colored blotch with the supratemporal stripe extending from the dorsolateral corner. We suspect that the extent of this “blotch” may be an artifact of drying, and we note the similarity to the “lunata” form of head marking found in some specimens of R. punctularia ( Fretey et al. 1977) . Occipital spots that would be diagnostic for R. punctularia may be represented by the two, oval, posterior extensions of the “blotch” nearest the midline (Fig. 5).

We can unequivocally exclude T. scabra L. from the synonymy of Melanochelys trijuga based on the continuous lateral keel present from hatching (as evidenced in the Seba illustration reproduced here as Fig. 1 ) and the great extent of dark pigmentation on the plastron found in that taxon. Similarly, T. scabra L. is obviously not a testudinid, and may be excluded from the synonymy of Cylindraspis spp. based on the presence of a nuchal scute and separate 12th marginals, as opposed to a single supracaudal scute as is characteristic of most Testudinidae . Among the features that characterize Trachemys scripta that are not present in the holotype of T. scabra are: 1) vertebral keel absent or evidenced as a narrow ridge; 2) no trace of lateral keels; 3) dark plastral coloration is not in the form of a broad central figure; 4) a single relatively large axillary scute prevents contact between the 4th marginal and pectoral scutes; and 5) a rather broad zygomatic arch. The other two emydids with which T. scabra L. has been synonymized, Glyptemys insculpta and Emys orbicularis , differ in the absence of humeral-axillary contact, only one axillary scute is present, and the zygomatic arch is relatively broad. Two obvious features of T. scabra L. that indicate it is not a specimen of Pelomedusa subrufa are its lack of an intergular scute and the presence of a nuchal scute.

Our preliminary impression of the holotype had been that it appeared to possibly represent R. annulata ( Gray 1860) , but we are now convinced that it is referable to punctularia rather than annulata . However, genetic testing of the specimen with comparison to all Rhinoclemmys species would be helpful to absolutely confirm its identity.

FIGURE 6. One of three specimens of “ Testudo scabra ” at the University of Uppsala Museum of Zoology donated by Carl Petter Thunberg between 1785 and 1815 (UUZM s/n, juvenile shell, 86 mm CL). Original Thunberg tag reads “ Testudo scabra β Mus . Thunb.” and newer tag reads “ Chelopus punctularius Mus. Thunb. ” This specimen represents Rhinoclemmys punctularia , but has no status as a type.

The type locality for T. scabra Linnaeus (1758:198) was originally given as “Indiis” (= West Indies or South America), but emended by Linnaeus (1766:351) to “ India orientali, Carolina .” Since the name T. scabra represents a specimen of R. punctularia , we hereby restrict the type locality of T. scabra L. to “Cayenne, French Guiana ”, the same type locality as for Testudo punctularia Daudin (1801:252) (= R. punctularia ).

Thunberg’s three specimens of “ T. scabra ” donated to the Uppsala collections betwen 1785 and 1817 ( Thunberg 1785 –1817) are also still extant. The first, T. scabra β (Fig. 6) is a dried shell of Rhinoclemmys punctularia measuring 86 mm CL. The second, T. scabra γ (not figured), has been re-identified as an Emys orbicularis . The third, T. scabra δ (not figured), is a juvenile Rhinoclemmys punctularia measuring 51 mm CL. None of these specimens have any standing as Linnaean types, nor were they described by Thunberg.

Figures

FIGURE 1. Illustrations from Seba (1734: pl. 79, figs. 1 and 2), showing his Testudo terrestris amboinensis major, later synonymized by Linnaeus (1766) under his concept of Testudo scabra. This appears to show a very young specimen of Melanochelys trijuga trijuga or Melanochelys trijuga thermalis.