Extant turtle diversity is represented by only 335 species (Turtle Taxonomy Working Group [TTWG] 2014), and the taxonomy of this major amniote lineage is characterized by a long and complex history that is full of nomenclatural confusion (e.g., Schneider 1783, 1787, 1789; Schoepff 1792−1801; Schweigger 1812a, 1812b; Fitzinger 1835; Gray 1844; Siebenrock 1909; Mertens and Wermuth 1955; Wermuth and Mertens 1996; Fritz and Havaš 2007; TTWG 2014). A well-supported taxonomy that reflects phylogeny and well-defined species boundaries with unambiguous nomenclature, however, is the very foundation for any meaningful biological research, ranging from basic to applied science and eventually including conservation work. As a result of the plethora of available names, changes in turtle taxonomy resulting from modern and integrative revisions frequently are reflected by synonymizations and revalidations. However, previously unrecognized taxa are still being discovered and formally described (e.g., Petzold et al. 2014; Thomson et al. 2015; Poulakakis et al. 2015; Edwards et al. 2016; Thomson and Georges 2016).

During early 2016, a previously unrecognized species was described within the Southeast-Asian snail-eating turtles. The genus Malayemys Lindholm 1931 (p. 30) until rather recently was considered to be monotypic, represented solely by M. subtrijuga (Schlegel and Müller 1845, p. 30). This semiaquatic species was assumed to be distributed all over Southeast Asia, from Myanmar through Thailand, Laos, and Cambodia, to Malaysia and Vietnam. Brophy (2004) resurrected M. macrocephala (Gray 1859, p. 479) and hypothesized it to represent a western form of snail-eating turtle, restricting M. subtrijuga to the eastern areas of this wide distribution range. Unexpectedly, 2 studies almost simultaneously hypothesized a third species, which appears to be restricted to the Khorat Basin in northeastern Thailand and Laos (Ihlow et al. 2016; Sumontha et al. 2016). Based on the evidence presented in these publications it seems unambiguous that both refer to the same previously unrecognized taxon, which raises the question of its valid name. The aim of the present study was to resolve this issue via a comprehensive bibliographical and nomenclatural analysis of both publications.

MALAYEMYS KHORATENSIS IHLOW ET AL. 2016

The study by Ihlow and colleagues was submitted to the journal PLoS ONE (ISSN 1932-6203) on 23 December 2015. It was accepted for publication on 22 March 2016 and was published electronically on 6 April 2016. The authors registered their new name M. khoratensis on ZooBank The Official Registry of Zoological Nomenclature (www.zoobank.org) and assigned it to an LSID (urn:lsid:zoobank.org:act:B88DB370-C3D0-4E64-A79A-3A77DDE7BCD8). PLoS ONE is properly registered at ZooBank, complying with all regulations for the electronic publication of nomenclatural acts according to the corresponding amendment of the International Code of Zoological Nomenclature (ICZN 1999, 2012). The name M. khoratensis hence has to be considered available for the purpose of nomenclature and was, as issued, published on 6 April 2016.

The name-bearing type specimen (holotype) of M. khoratensis (THNHM 25816, Natural History Museum, Pathum Thani, Thailand, young adult female) was collected during July 2014 by Flora Ihlow and Michael Cota at Udon Thani (Udon Thani Province, Thailand, 17.36555°N, 102.81427°E, WGS 1984).

MALAYEMYS ISAN SUMONTHA ET AL. 2016

The study by Sumontha and colleagues was submitted to the journal Taprobanica (ISSN 1800-427X) on 21 January 2016. It was accepted for publication on 18 February 2016 and was distributed electronically on 26 March 2016. The authors registered their new name M. isan on ZooBank and assigned it to an LSID (urn:lsid:zoobank.org:act:6D1A31B9-8960-42CE-97D5-0F48AF46FD7A). Taprobanica is registered with ZooBank; however, the entry does not provide reference to a digital archive other than that of the publisher, which renders the electronic edition of the journal unavailable for nomenclatural purposes because it does not constitute published work in the sense of the Code (ICZN 1999, 2012, Article 8.5.3.1). The publisher, according to the ZooBank entry, is Taprobanica Private Limited, Tangerang, Indonesia, whereas the actual Web site of the journal designates the Research Center for Climate Change at the University of Indonesia in Depok as its sole current publisher. Regardless of this inconsistency, it was necessary to establish the actual date the printed edition of Taprobanica 8(1) appeared, because this is the version of record and must be considered the date of publication for the name M. isan.

As of 6 April 2016, the Web site of Taprobanica listed the libraries of the University of Indonesia in Depok, the Natural History Museum in London (England), the National University of Singapore, and the Museum of Comparative Zoology at Harvard University (USA) as the 4 official repositories that receive archival copies of the printed journal free of charge. These libraries were contacted and asked whether the printed edition of Taprobanica 8(1) had already been received at this time. All of them except for the University of Indonesia replied and stated that it had not yet been received. Feedback from Harvard University on 6 April 2016 stated that “[t]hey send us the deposit copies in a delayed fashion and without notice”. The Natural History Museum in London stated on 8 April 2016: “We have not received a print copy since Vol. 7(1), which we received 23 Feb[ruary] 2015. We have missed Vol. 5(2) and Vols. 6(1)&(2) and have not yet received Vol. 7(2) and Vol. 8(1)”. On 11 April 2016 the National University of Singapore also had not yet received a printed copy of Volume 8(1).

We were unable to order printed copies of the journal via the Web site, advertised as “printed on high quality mat art paper” with color illustrations for USD 60 per current issue. Thus, the editor-in-chief was contacted by email on 11 April 2016 2354 hrs Central European Time (CET) with the intention to purchase a printed copy of Taprobanica 8(1). The request was responded to on 12 April 2016 0149 hrs CET and, surprisingly, 3 different editions of the journal were offered for sale: black and white normal paper (USD 30), color normal paper (USD 40), and color art paper (USD 50), the last of which—apart from the price—seemed to correspond to the only edition that was advertised online. One copy each of the 3 editions was ordered and payment was transferred the same day directly to the editor-in-chief via Western Union. An acknowledgment of payment was received together with a shipping confirmation on 14 April 2016 0942 hrs CET.

Upon request, the editor explained this unusual publication model with 3 different editions on 12 April 2016 0847 hrs CET and stated that at the same time the online version appears they “publish black & white normal paper copies (around 50 copies) for local and international archives”. The editor stated that, in addition, “around” 25 colored copies on normal paper and “around” 25 copies on art paper are printed for individual purchases. In case of a higher demand, additional issues would be printed as needed.

The official information on the Web site regarding the standard “mat art paper” edition was in strong contrast with the editor's initial statement as quoted above, which indicated an entirely black and white edition on normal paper as the standard edition, while the colored editions were only intended for special collectors. After the editor was confronted with the discrepancy in terms of the journal's “standard edition”, he responded on 12 April 2016 0953 hrs CET and stated that the colored mat art paper edition was “the standard format”. This was in agreement with the information on the Web site, but contradicted his previous statement. This was also the first time the editor mentioned that the black and white edition of Volume 8(1) was not entirely black and white after all, but that “the first article [Sumontha et al. 2016] [was] printed in color by request of the authors”. The editor felt obliged to state “that all three copies are identical except the printing material (paper, ink)”.

During the afternoon (CET) of 12 April 2016, after the above-quoted correspondence with the editor, the Web site of Taprobanica was recognized to have been altered, most notably in that no official repositories holding archival copies of the journal were mentioned by name anymore. Furthermore, the explanatory statement “[t]he journal is archived (free of charge) in four (4) international archives” had been changed to “[t]he journal is archived (free of charge) in local and international archives”. The WayBackMachine of the Internet Archive (https://archive.org) provides access to preserved versions of the journal's “About us” page (http://taprobanica.org/about/about-us/) dating 14 March 2016 and 23 October 2016 that document both states of the Web site. Currently (continuously since at least 27 February 2017), the entire Web site of Taprobanica is not functional anymore. Access to the electronic version of articles published in Taprobanica is only possible via these preserved snapshots at the Internet Archive. The continued functionality of such preserved Web sites cannot be guaranteed and automated archiving cannot be considered in the intention of the Code, which enhances the importance of the printed archival copies of Taprobanica in providing a public and permanent scientific record.

The libraries in Singapore, London, and Cambridge eventually each received its printed copy of Taprobanica 8(1) along with several other back issues that had not been received until then, with Singapore being the first on 27 April 2016. The shipping labels on the envelopes received by all 3 institutions consistently document an identical date of dispatch: 13 April 2016, all registered to the Indonesian postal system within a window of < 10 min between 11:35:00 AM and 11:42:59 AM local time (Fig. 1). Note that this coincidentally was around noon the day after the above-quoted correspondence concerning the unusual publication model of Taprobanica took place, and 1 d after the information on archiving repositories had been removed from the journal's Web site. The dispatcher in all 3 cases was the editor-in-chief in person.

View Figure

• Download as Powerpoint
• Download Figure

Published work in the sense of the Code must be issued for the purpose of providing a public and permanent scientific record (Article 8.1.1), which for printed publications is primarily achieved by institutional libraries. According to Article 8.1.2, published work furthermore must be obtainable, when first issued, free of charge or by purchase. Libraries usually record the date of arrival of a journal (see also Recommendation 21D), which indicates when a publication actually became accessible to the public. In the present case, the first publicly accessible copy of Taprobanica 8(1) can be documented for the library of the National University of Singapore on 27 April 2016. London and Cambridge received their copies shortly thereafter. Nonetheless, the journal must be considered having been obtainable already prior to its accessibility at the libraries. Upon our request, the archiving libraries recorded the date of dispatch for Taprobanica 8(1), which consistently was 13 April 2016 (Fig. 1). This is the first day a copy of this volume objectively can be demonstrated to having been in existence and obtainable, although the theoretical possibility of an earlier printing remains.

An envelope, denoting a date of dispatch of 15 April 2016 and containing the 3 privately purchased copies of Taprobanica 8(1), arrived in Bonn on 9 May 2016 (Fig. 2). With the exception of a glossy copy, which is assumed to represent the “standard edition” of the journal originally denoted as being printed on “mat art paper”, this satisfyingly constituted what was ordered.

View Figure

• Download as Powerpoint
• Download Figure

Along with the order, enclosed within the full-color, normal-paper edition, 2 documents relevant to the reconstruction of the publication process were delivered: a handwritten general manufacturing order form from the Indonesian printing shop, dated 13 April 2016, and an accompanying more detailed sheet specifying the colored printouts (Fig. 3). The general manufacturing order form identifies the customer as the editor-in-chief and unambiguously denotes the order as “Taprobanica”. According to these forms, the order consisted of 3 different elements: 5 copies each of the full-color art paper and full-color normal paper edition, as well as an illegible number of copies of the black and white edition, all of which were commissioned and produced on 13 April 2016.

View Figure

• Download as Powerpoint
• Download Figure

Aside from the general aspects regarding the publication procedures of Taprobanica described above, it furthermore should be noted that character (3) of Sumontha et al. (2016, p. 2), an uninterrupted supraorbital stripe, which is considered to be diagnostic for “M. isan”, should not be considered diagnostic for M. khoratensis because the supraorbital stripe can be interrupted in specimens that genetically can be assigned to the new species in question (cf. Ihlow et al. 2016, fig. 7). The holotype of “M. isan” (THNHM 25609, adult female) was collected at Ban Na Klang (Nong Bua Lamphu Province, Thailand) on an unknown date.

CONCLUSIONS

No objective evidence allows us to document the physical existence of any of the archival copies of Taprobanica 8(1) prior to 13 April 2016. The manufacturing order forms found retained in one of the privately purchased copies are undeniable and unambiguous evidence that several copies of the journal were printed that very day. No other copy is known to us. In conclusion, given that all known copies of Taprobanica 8(1) somehow can be dated back to the same day strongly suggests that all of them in fact derived from the same print run on 13 April 2016.

The time stamps on the envelopes received by the archiving libraries indicate a registry with the Indonesian postal authorities at around noon local time on 13 April 2016. This would easily have allowed a waiting customer (as our documentation in Fig. 3 indeed proves to have been the case) to get the copies printed earlier that morning and to ship them afterward, especially with the printing shop being located so close to the editor's working place. We provide solid evidence that several copies of this journal were printed that day, in commission by the editor-in-chief, even though he communicated matters 1 d earlier as if the journal had been produced with a substantial print run already a couple of weeks before that. The production of several copies already on Saturday, 26 March 2016, however, seems very implausible, given that until the “reiterated” printing on 13 April 2016 not even the archival copies had been sent out. The coincidental removal of information concerning the deposition of the archival copies from the journal's Web site, immediately after our correspondence with the editor, is consistent with the notified general irregularities in the publication process of Taprobanica and raises further reservations about this outlet and the reliability of the information provided. The shipments documented for all known archival copies (lacking only feedback from the University of Indonesia, the journal's formal publisher), point directly to the day we are able to demonstrate that copies of Taprobanica 8(1) evidently were printed. Therefore, we consider the indicated date of publication (26 March 2016) to be incorrect, and in agreement with Article 21.4 hereby correct it to 13 April 2016, because this is the first day a printed copy of the journal—being the nomenclaturally relevant edition of this outlet—can be demonstrated to have been in existence.

Our inquiry to purchase a printed copy of the journal already started on 11 April 2016 and the correspondence with the editor continued throughout 12 April 2016, placing the date of publication of Taprobanica 8(1) after these events. By simple definition, this indicates it was being operated under a print-on-demand publishing model that seems to have been triggered by our order and our general interest in the journal's publication process. Until we placed our order, no physical copy of the journal can be demonstrated to having been issued for the purpose of providing a public and permanent scientific record, especially not in an edition containing simultaneously obtainable copies by a method that assures numerous identical and durable copies. It remains speculative whether the printing would have been undertaken without our initiative at all, but the complaints about the overdue supply of printed back issues, as consistently noted by the different archiving libraries, raise certain doubts. The electronic version of Taprobanica 8(1) posted online on 26 March 2016 does not meet the requirements to constitute published work (Article 8.5.3.1), given that the journal is not properly registered at ZooBank and especially not adequately digitally archived. In conclusion, we consider this a fundamental violation of Article 8.1 that renders Taprobanica 8(1), in agreement with Article 9.12, not being published for the purpose of nomenclature. Following Article 11.1, the name “M. isan” consequently cannot be considered to be available for the purpose of zoological nomenclature either.

Malayemys khoratensis, on the other hand, fulfills all requirements of nomenclatural availability and can be reliably dated to 6 April 2016: 1 wk before the first physical printed copy of Taprobanica 8(1) can be demonstrated to have been in existence (13 April 2016). This makes M. khoratensis by all means the senior synonym and valid name for the recently described species of snail-eating turtle from Southeast Asia. Malayemys khoratensis consequently has to be adopted in any publications mentioning the Khorat snail-eating turtle.