The Role of Museums in Safeguarding Biodiversity: A Happy End of the Century-Long Saga of Tritogenia Zuluensis (Beddard, 1907) (Oligochaeta: Tritogeniidae)

ABSTRACT Tritogenia zuluensis (Beddard, 1907) is re-described and the taxonomic position of this species is verified. Limited knowledge of South African earthworms at the time of the original species description and a misleading specimen illustration led to difficulty in the establishment of its true identity. Thanks to protection of the type material at the Natural History Museum, University of Oslo, Norway, this species can be redescribed and illustrated, and its lectotype and paralectotype are designated. Types of Microchaetus zulu Michaelsen, 1907, a junior synonym of Tritogenia zuluensis, safely stored for more than a hundred years in the Göteborg Natural History Museum, Sweden, was most helpful in the process of re-description. The significant role played by museums and similar organizations in protecting past and present natural resource treasures for the future, is highlighted. The crucial function of natural history collections in supporting traditional taxonomy, the key to understanding biodiversity, is clearly demonstrated.


INTRODUCTION
The description of Tritogenia zuluensis (Beddard, 1907) in the genus Microchaetus Rapp, 1849, was published on 5 March 1907 on the basis of two fragments of specimens collected in the 'locality' indicated broadly as 'Zululand', which is a vast area in northern and north-eastern KwaZulu-Natal. Less than two months later (24 August 1907) Michaelsen described Microchaetus zulu from 'Zulu-Land, Umpolozi'. Two similar descrip tions of two species collected in the area 'Zululand' suggested possible species sy nonymy, and motivated Michaelsen (1913) to re-examine specimens and to accept the species priority of Beddard's zuluensis. However, during Michaelsen's re-examination of the zuluensis and zulu type material, some of the 'immature and several mature specimens collected by E. Warren, June 1903in Eastern Zululand, and by T.W. Jones, September 1911in Mfongosi, Zululand' (Michaelsen 1913 were included, and a composite de scription of Microchaetus zuluensis Beddard, 1907 was set up. Similar data, enlarged with added information on calciferous glands, were restated by Michaelsen (1918). Subsequently, Plisko (1992) examined the material kept in the KwaZulu-Natal Museum collection, labelled as ' NMSA/Olig.00349, Zululand, Mfongosi, 19 March 1913, W.E. Jones leg.', and found out that the initials 'T.W' for Jones, a collector of the mate rial studied by Michaelsen (1913), was a printing error for W.E. Jones. She believed that the earlier material studied by Michaelsen for the zuluensis re-description and the spe cimens collected by the same person (W.E. Jones) in 'Mfongosi, Zululand', may be accredited to the same species. After extended examination of the material, Plisko noted some of the characters used earlier by Michaelsen as being characteristic for zuluensis and accredited the new material to this species. Considering some of the exclusive features of the genus Tritogenia (location of gizzard in segment 6-7, septum 6/7 adherent to gizzard, meroic nephridial excretory system), zuluensis (as defined by Michaelsen (1913)) was transferred from Microchaetus to Tritogenia Kinberg (1867) (see Plisko 1992: 373). The presented photograph of tubercula pubertatis and description of the material were based on individuals other than type material, and referred to 'composite zuluensis', as determined earlier by Michaelsen (1913). Later, however, Plisko (2008) become aware of this error and suggested a revision based on the type material in order to establish the real characters of this species. Closer evaluation of the earlier descriptions (Beddard 1907;Michaelsen 1907Michaelsen , 1913Michaelsen , 1918Plisko 1992) confirmed the necessity of revision based on type material. The features used by authors since the original descriptions of zuluensis, comprising a complex species portrayal (Table 1), validated the decision. Furthermore, it was noted that zuluensis and zulu had been described in Microchae tus, as was commonly practised at that time for the majority of the other South African earth worm species, and continued to be kept in this genus, although Tritogenia generic characters were indicated by Kinberg in 1867, and also in the description of Brachydrilus Benham (1888). Supposedly, the erroneous assemblage of generic data for Tritogenia and Brachydrilus produced by Michaelsen (1900) were the basis for incorrect species placement. The exclusive Tritogenia generic characters recognized later by Michaelsen (1918) allowed him to synonymize Brachydrilus with Tritogenia, and to transfer all the species known at the time (sulcata, howickiana, benhami, morosa and crassa) from Microchaetus to Tritogenia, although the composite species zuluensis was still left in Microchaetus. Plisko & Zicsi (1991) emended Michaelsen's (1900emended Michaelsen's ( , 1918emended Michaelsen's ( , 1928b) generic characters given for Tritogenia, and added a few new species to this genus. Plisko (1992Plisko ( , 1997Plisko ( , 2003Plisko ( , 2005Plisko ( , 2006Plisko ( , 2008 included zuluensis in Tritogenia, described a number of new species, and provided more information concerning the taxonomic position of the genus. During the recent separation of the Tritogenia species from Microchaetidae (Plisko 2013) and their re-location to the family Tritogeniidae Plisko, 2013, a merged de scription of Tritogenia zuluensis was again noted, and it was discovered that no holotype or paratype was indicated amongst the original material or in the description. There fore, a species revision was suggested. The aim of the present paper is to fulfil this necessity. The species re-description is founded upon examination of the type material and specimens that had been described by Michaelsen (1907) as the taxon zulu, and supplemented with data taken from the original species description. Designation of a lectotype and paralectotype of Tritogenia zuluensis (Beddard, 1907) is also included.

MATERIAL AND METHODS
The following acronyms and abbreviations are used in this paper: GNM -Natural History Museum, Göteborg, Sweden; NMSA -KwaZulu-Natal Museum, Pie ter maritzburg, South Africa; NHMU -Natural History Museum & University, Oslo, Nor way; NMSA/Olig. -NMSA Oligochaeta Collection, followed by accession numbers.
The type material of Microchaetus zuluensis Beddard, 1907, housed for more than a hundred years at the NHMU, containing two anterior parts of the body with small portions of posterior segments. A clitellate specimen illustrated by Beddard (1907: 279, text-fig. 85) (Fig. 1) and described as 'more fully mature than the other', abscised behind the clitellum, with a total segment number reaching ca 86, presently designated as a lectotype of Tritogenia zuluensis (Beddard, 1907) (NHMU Oslo C5726, Figs 2, 3). The other spe cimen, abscised at ca the 109 th segment, with slightly evident tubercula pubertatis and genital papillae, but with no clitellum, numbered, designated as the paralectotype (NHMU Oslo C5727, Fig. 4). Material loaned by courtesy of the museum staff engineer Mrs Ann-Helén Rønning and Dr Philip D. Harris, helminth collection curator of the NHMU, who also provided kind advice regarding the accession numbers.
The type material of Microchaetus zulu Michaelsen, 1907, accepted by Michaelsen (1913 as a synonym of zuluensis, and declared by Reynolds and Cook (1976) as 'typus amissus'. Thanks to its long-term protection at the GNM, it was rediscovered and loaned to me by courtesy of Prof. Kennet Lundin, senior curator of marine and lim nic invertebrates. The dissected specimen, although slightly decomposed, is still in good con dition but with no internal organs in segments 9-16, and is labelled 'GNM Oli gochaeta 16' (Figs 5-7).
Male pores: Not detected externally and internally. Spermathecal pores: Not observed, although Michaelsen (1907)  Internal characters: Septa (Figs 3, 6): 4/5 slightly thickened, 5/6-8/9 markedly so, si milar in appearance; 9/10 very thin, partly aborted; other septa in preclitellar segments thin; in posterior segments somewhat thicker. Gizzard (Fig. 6): Oesophageal, large, muscular, commencing in 6, extends and occupies whole of 7, terminating abruptly at septum 7/8. Calciferous glands (difficult to see in type material; not present in comparative mate rial, having been removed by Michaelsen (1918: 331-332) for a study with its particu lar de scription): Half-globular in 9-10; closely connected to oesophagus; dorsally and vent rally separated. Intestine (Fig. 3): Commences in 12 in lectotype (it is difficult to see where the origin of the intestine is due to desiccation). Typhlosole: Commencement not detected. Dorsal blood vessel: Double in 5-10 and also when crossing septa 4/5-10/11; in 12 and the segments that follow, broader and simple. Paired dorso-ventral commissural vessels: In 4-8, slender tubes; in 9-11, thick moniliform 'hearts'. Excretory system: Meroic; minute mero-nephridia difficult to detect between fragile, decomposing internal tissues, so their exact number and position were not established; two tiny pairs possibly occur in each segment, as was found in a few postclitellar segments of type material. (Obvious meroic pairs observed in anterior segments in material described by Michaelsen (1907) were mistakenly interpreted: 'Nephridialsystem meganephridisch'  Beddard, 1907 (text- fig. 85), the anterior part of the body, with misleading indication of the simple segmentation.
[excretory system meganephridial], which means a holoic system.) Male funnels: Two pairs of funnels, indicating holandric nature of this species, and not proandric as is stated by Beddard (1907: 281) and Michaelsen (1907: 8); the first pair much larger than second, both closely connected with seminal vesicle. Vasa deferentia: Not detected due to internal dryness and slight decomposition of specimens. Seminal vesicles: Two pairs, second being much smaller than the anterior pair, in 10 and 11 respectively, both linked with testis sacs. Spermathecae: Spermathecal ampullae were not observed in type material by Beddard (1907), possibly because of their small size, and perhaps emptiness; being uninseminated, these structures were difficult to trace amongst the thick, slightly decomposed body tissues. Some tiny remnants of the ampullae that were seen at 350× magnification in segment 13, suggest their possible presence in other segments as well. In a specimen collected by I. Trägårdh in Umfolozi, close to the type locality, two to eleven spermathecal ampullae are present, near intersegmental furrows 11/12-15/16. Ovaries: In 13; one funnel-like, near septa 13/14. Genital glands: Variable in size and shape, associated with genital papillae. Genital setae: Noted once in genital gland of 22 nd segment. Type material and locality: Lectotype (NHMU Oslo C5726) and paralectotype (NHMU Oslo C5727) originated from 'Zulu-Land', the area north of the Thukela River in northeastern KwaZulu-Natal, South Africa. Unfortunately, no additional data concerning the collection site were given. It was probably in the vicinity of where Proandri cus colletti (Beddard, 1907) (Plisko 2000) was collected by Knut Dahl during his hunting expedition to Zululand in 1893-1894 (Dickison 1951;Pethon 2009), and likely the same as what was known at that time as Umfolozi Game Reserve, which was visited by various hunters and researchers.
Other material was collected by Swedish researcher Ivor Trägårdh on 6 June 1905, in 'Zulu-Land, Umpolozi' (Michaelsen 1907), probably close to the type locality of zu luensis but described by Michaelsen as 'Microchaetus zulu', which he later accepted as a synonym of Microchaetus zuluensis. The present study indicates that this type specimen (GNM Oligochaeta 16), should now be known as Tritogenia zuluensis (Beddard, 1907). Distribution: The species is so far known from the Hluhluwe-iMfolozi Nature Reserve in the KwaZulu-Natal Province of South Africa. Remarks: The type material of T. zuluensis was probably collected in the southern part of the Hluhluwe-iMfolozi Park which, before its proclamation in 1895 as a National Park, was a hunting area named Umfolozi. The area covers over 50,000 ha of magnifi cent foothills of the first escarpment rising from the coastal plain, with the two main rivers, the Black and White Mfolozi, joined by a number of streams. Erosion over millions of years have resulted in a variety of soils with many endemic plant and animal species, and some earthworm material was sporadically collected in the region.
Zululand, leg. Dahl, det. Beddard, kopi av kartotekkort, Lumbrici, Zululand 1893, Dahl'. The following appears on the other label ( Fig. 9): 'Microchaetus colletti sp. n., Mic rochaetus zuluensis sp. n., see P.Z.S. 1907 August'. After the colletti type had been examined, it was sent back to the NHMU, and the species was re-described by Plisko (2000). The two abscised specimens of Microchaetus zuluensis are described in the present paper. The label claiming two species names and the note 'see P.Z.S. 1907 August' was probably written by Beddard at the time of the original description of both species, as the note refers to Proceedings of the Zoological Society of London, where Beddard described colletti and zuluensis (1907: 277-281). As the specimens of both species and the labels were inserted into the same tube, it is likely that both colletti and zuluensis were collected by Knut Dahl during his excursion to Zululand in 1893-1894. Other earthworm material included in the re-description of zuluensis by Michaelsen (1913) was collected further south-west of the zuluensis type locality. The site indicated as Mfongosi is located to the south, at the Mfongosi River, a tributary of the Thukela River, and differs from the iMfolozi area as regards soil types, flora and fauna. W.E. Jones was a 'keen amateur naturalist' (to quote from Herbert and Kilburn (2004: 53)) who collected numerous litter and soil invertebrates, many of which are located at the NMSA. This non-type material, once included with zululensis, is now found to differ from the type material and should be separately revised.
It is anticipated that new material of zuluensis may provide more data confirming the present species evaluation, and modern research methods might contribute information that sheds further light on relationships between zuluensis and other native South African Tritogenia species.

DISCUSSION
At the time of its initial description, Tritogenia zuluensis (Beddard, 1907)  the studied specimen, thereby conflicting with the actual characters. This inaccuracy was later, during species re-description, partly corrected by Michaelsen (1913). How ever, in clusion of specimens other than type material, which had been collected at dis tant sites, made his re-description erroneous. The need for revision based on the type mate rial was therefore a necessity. Thanks to more than a hundred years of high-care protection of the earthworm specimens at the NHMU and GNM, this new revision in terms of as sessment of taxonomic status and species re-description, was made possible. The museums, herbariums, and other institutions mandated with safeguarding of collected specimens are sanctuaries of the past and present that enhance our know ledge of life on Earth. A significant role of such collections in the recognition, under standing, and preservation of natural resource treasures has been emphasised in many publications (e.g. Suarez & Tsutsui 2004;Pyke & Ehrlich 2010;Williams 2010; and referen ces therein). Biological diversity refers to the variety of life on Earth, and needs constant awareness and documentation. However, the concept of biodiversi ty can only be proper ly understood if viewed in the context of an ecosystem in which all forms of life on the planet require our continual protection. Special concerns about the conservation of glo bal biodiversity emerged in Rio de Janeiro during the 1992 United Nations Conference on Environment and Development as the Convention on Biologi cal Diversity. Over the past few decades, the protection of biodiversity has often come to be understood as guarding of natural ecosystems, with parallel dis regarding of traditio nal taxonomy (Mace 2004). However, it should be remembered that taxonomy is a discip line that provides a basis for understanding biodiversity (Ebach & Holdrege 2005). Only through descriptive re cognition of species richness, the global protection of Nature may be understood and im plemented. As was stated by Boero (2010), 'The study of biodiversity cannot proceed further without the contribution of in tegrative taxonomy' and this is clearly demonstrated in recent research such as that by Blakemore (2013). Taxonomy requires the dedica ted at tention of the researchers who create faunal and floral collections, identify and name taxa, so the latter can be duly recognized and used in future studies. Natural history collections are the major foundation for traditional taxonomy; they should receive greater attention in the academy and must get a higher level of support from the government (e.g. Hamer 2012). It should be remembered that taxonomy is a basis for understanding biodiver sity and the whole complexity of life on our planet Earth.
I cannot resist including a quotation (after Herbert & Kilburn 2004) taken from a letter written in 1919 by of the first Director of the Natal Museum, Dr Ernest Warren, to the Hon. Jan Smuts, who was then the Prime Minister of South Africa. The conception is still profoundly true at the present time: 'In many respects the African fauna is one of the most wonderful in the world, and should be a cause of national pride, but at the present time on one pretext or another it is being rapidly and ruthlessly destroyed in many areas, and once destroyed it can never be restored ... It is slowly being realized throughout the world that the indigenous fa una of the country is a heritage that should be passed on, unimpaired as far as possible, to the next generation, and it is not something that the present generation has the right to destroy for the sake of temporary convenience. '  TABLE 1 Summary of characters and their states used by authors for the Microchaetus zuluensis Beddard, 1907 andM. zulu Michaelsen, 1907 descriptions, compared with the present observations. Abbreviations: L -lectotype, P -paralectotype.