How to deal with destroyed type material? The case of Embrik Strand (Arachnida: Araneae)

Abstract. When the museums of Lübeck, Stuttgart, Tübingen and partly of Wiesbaden were destroyed during World War II between 1942 and 1945, also all or parts of their type material were destroyed, among them types from spider species described by Embrik Strand between 1906 and 1917. He did not illustrate type material from 181 species and one subspecies and described them only in an insufficient manner. These species were never recollected during more than 110 years and no additional taxonomically relevant information was published in the arachnological literature. It is impossible to recognize them, so we declare these 181 species here as nomina dubia. Four of these species belong to monotypic genera, two of them to a ditypic genus described by Strand in the context of the mentioned species descriptions. Consequently, without including valid species, the five genera Carteroniella Strand, 1907, Eurypelmella Strand, 1907, Theumella Strand, 1906, Thianella Strand, 1907 and Tmeticides Strand, 1907 are here also declared as nomina dubia. Palystes modificus minor Strand, 1906 is a junior synonym of P. superciliosus L. Koch, 1875 syn. nov.

Species descriptions on the basis of type material is one of the cornerstones of taxonomy. Type material is stored in safe places, usually museums and comparable collections, so it is always possible to compare the described species with the original material. But what if this type material gets destroyed? During the last centuries, wars, fires and other catastrophes as well as inappropriate or lacking maintenance have already destroyed a lot of type material, creating a difficult situation for taxonomists.
During World War II, in the night of 28./29. Mar. 1942, the city of Lübeck was the first German city that was considerably destroyed by allied air forces. The museum in the centre of the city, that contained the natural history collection, was completely destroyed and the collection material was burnt. This also concerned the spider collection with all type material (Füting pers. comm.). The natural history museum of Stuttgart, initially the "Kurfürstliches Naturalien-Cabinett", was hit when the city was bombed in September 1944. The building burnt out completely and the collections were destroyed (Roewer 1955, 1959, 1960, Renner 1988. In the city of Tübingen 800 buildings were damaged or destroyed by air bombing. The spider collection in a small museum at the University was (probably together with other collection material) stored in wooden boxes that survived the end of the war. After the war, during cleaning-up the buil-dings in 1945, the boxes were "dumped to gain free space", i.e. they were destroyed (Weber pers. comm.). Parts of the ethanol collection of the Wiesbaden Museum were stored during the war in a gardener's house in the spa gardens due to their flammability. During an air attack in the night of 2./3. Feb. 1945, this material was entirely destroyed. The museum building including other collections was, however, left without damage (Heineck 1950). On the other hand, two very large natural history museums in Germany survived World War II with less damage: the Senckenberg Museum in Frankfurt/ Main and the Natural History Museum in Berlin had enough time to evacuate their collections into safe areas, so that they survived.
From 1906 to 1917, Embrik Strand described 381 spider species and one subspecies from collection material stored in the museums of Lübeck, Stuttgart, Tübingen and Wiesbaden. By analysing his publications (Strand 1906a(Strand , 1906b(Strand , 1906c(Strand , 1907a(Strand , 1907b(Strand , 1907c(Strand , 1907d(Strand , 1907e, 1907f, 1908a(Strand , 1908b(Strand , 1908c(Strand , 1915a(Strand , 1915b(Strand , 1917, we learnt which species he described and where the type material was deposited because in these publications he always stated to which museum the material belonged. This is in remarkable contrast to his usual habit of not indicating where he deposited the type material (Nentwig et al. 2019). In this publication, we deal with the species described by Strand and type material deposited in museums that were later destroyed. Strand also described a series of subspecies, varieties and forms, we investigated earlier (see Nentwig et al. 2019).
From these 382 taxa, a subset of 64 species has meanwhile been declared as nomina dubia and 57 species were synonymized with other species. For 79 species, Strand provided illustrations, so that the spiders could (at least theoretically) be recognized, other authors illustrated them, and/or type or non-type material was found in other museums (all information from the World Spider Catalog 2019).
For the remaining 182 spider taxa, we state here that:  Strand deposited the type material in the museums of Lübeck (58 species), Stuttgart (103 species), Tübingen (11 species) and Wiesbaden (9 species and 1 subspecies) and all type material became verifiably destroyed (references in Tab. 1).  There is no other type or non-type material known.  Since their description by Strand, during a period of more than 110 years, none of these species has ever been recollected.  None of these species has been mentioned in other publications, thus there is no further scientific literature about them. This statement does not include taxonomic revisions (such as Dalmas 1921: 292 or Laurent 1946 or catalogue works (such as Bonnet 1955-1959, 1961and Roewer 1955, 1959, 1960 that only listed the names or repeated Strand's descriptions without further investigations.  Strand described these species in his characteristic manner, this means, he sometimes delivered long descriptions but useful diagnostic aspects and comparative discussions regarding other (similar) species were often missing (see Nentwig et al. 2019). Illustrations were not provided for any of them. Considering the magnitude of the world's biodiversity known nowadays, it is impossible to recognize these species.  In 19 cases, Strand's description was based on juvenile specimens, in 153 cases, he had only one sex (41 males, 111 females), and in only 11 cases he had both sexes.
We conclude that without type material or scientific illustrations and useful descriptions, it is impossible to recognize these species and subspecies. Therefore, we declare them here as nomina dubia. This concerns the 181 species listed in Tab. 1.

Lithyphantes lenzi
Steatoda lenzi (Strand, 1907)  Palystes modificus minor was described by Strand (1906c: 41) as "var. minor". This must have been overlooked by other taxonomists because none of the arachnological catalogues had ever listed this taxon. The male type was deposited in the museum Wiesbaden and was destroyed ( Jäger 1998). Following the logic of this publication, we should also declare it a nomen dubium. From the description it becomes clear that Strand actually considered it conspecific with another species he described (Palystes modificus): "Ein zweites Exemplar, das wohl derselben Art angehört…" [A second specimen that belongs apparently to the same species…]. Moreover, he admitted that the difference in body size he recognized between the nominal form and his variety also occurs in P. superciliosus L. Koch, 1875, the species which was recognised by Croeser (1996: 56) as senior synonym of P. modificus. Therefore, we propose Palystes modificus minor Strand, 1906 as a junior synonym of P. superciliosus L. Koch, 1875 syn. nov. Five genera become now also nomina dubia because they do not include valid species any more (see Tab. 1: comments (8) and (9)). Four of them are monotypic (Carteroniella, Strand 1907a: 543;Eurypelmella, Strand 1907e: 51;Thianella, Strand 1907e: 448;Tmeticides, Strand 1907c: 729) and one is ditypic (Theumella, Strand 1906a: 608).

Discussion
A declaration as a nomen dubium on the basis of the above listed argumentation is in line with the decision of many other researchers who came in the past to the same conclusion with respect to Strand's species. Concerning Eurypelmella masculina, Gabriel (2016: 86) opposed a transfer to Schizopelma, as proposed by Raven (1985: 153), because Strand's insufficient description would not allow this. For Mallinus defectus, Haddad et al. (2019: 157) wrote "The description of the species by Strand … does not provide a single clue as to the real identity" and Jäger (2014: 184) concluded for Torania/Barylestis manni that "the identity of this species remains unsolved". All three examples concern species listed in Table 1.
Generally, other arachnologists used the fact that the type material was lost or destroyed as a very strong argument to declare a species as a nomen dubium, often in combination with the weak descriptions of Strand and the unavailability of further specimens. For example, Benoit (1968: 141) mentioned that Selenops modestellus Strand, 1907 was "détruite durant la guerre … et décrite de façon insuffisante" [destroyed during the war … and described in an insufficient manner]; Blandin (1975: 384) wrote for Perenethis brevipes (Strand, 1906) that "la description originale ne permet pas de se faire une idée correcte de cette espèce" [the original description does not allow to identify this species correctly]; Russell-Smith (1982: 91) similarly concluded for Trabaeops bidentigera Strand, 1906 that "the type … was destroyed … and Strand's description is insufficient to identify the species". "The female syntypes are lost and the type locality ('Central Australia') is unknown" wrote Rix et al. (2017: 585) for Can-tuarides exsiccatus Strand, 1907 andconcluded on nomen dubium. FitzPatrick (2007: 98) came to a similar conclusion for nine Zelotes species described by Strand because "type material [was] lost or destroyed and the taxa are not recognisable from the original descriptions", and Fukushima & Bertani (2017: 129) simply argued "type considered lost" and declared three species and one subspecies of Avicularia as nomina dubia. When Raven (1985: 113) synonymized the genus Nossibea with Tigidia, he only argued "The type of Nossibea is lost and was left unplaced by Benoit (1965). Strand (1907a) gave no characters by which Nossibea differs from Tigidia. " We can conclude that arachnologists frequently argued in comparable situations in favour of a nomen dubium statement and, meanwhile, such a statement is a kind of routine, also accepted by the World Spider Catalog (2019). It is necessary to underline that such a declaration needs strong guidelines. Unfortunately, the International Code of Zoological Nomenclature (ICZN 2012) does not provide such rules. The term "nomen dubium" does not even occur in the Code, it is only mentioned in its glossary where a nomen dubium is explained as "a name of unknown or doubtful application". It should also be kept in mind that a nomen dubium declaration is a reversible process, just as the synonymization of two species that can be reversed, for example, within a comprehensive genus revision. The reality, however, shows that this is an extremely rare event, indicating, that the parameters used so far hold. We presented here a set of strong arguments (type material lost, insufficient description, no illustration, never recollected) and recommend to be as critical as we were here. We are convinced that there are more doubtful species listed in the World Spider Catalog (2019) and such "cleaning-up" may help to release us from taxonomic ballast.

Recommendation
This unsatisfactory resolution of many taxa described by Embrik Strand provides a few lessons that are still valuable today and allow us to express three recommendations.
(1) Type material has to be described as completely and carefully as possible. Fortunately, standards today are much higher than at Strand's time. He delivered sometimes long descriptions but often descriptions of the differences to similar species are missing. Illustrations and/or photos are today state of the art. (2) Even today, a museum can burn down, as two recent incidents from Brazil showed us. To prevent similar total losses in future, we recommend that authors provide sound descriptions of both sexes based on all the available specimens, and, whenever possible, to deposit the type material in more than one institution (ideally in four to five), and never in private collections, as these unfortunately are often not curated after the collector's death and therefore are even more prone to loss or ruin.