A new bombyliid species Marleyimyia xylocopae Marshall & Evenhuis, sp. n., an apparent mimic of the carpenter bee Xylocopa flavicollis (De Geer), is described from South Africa on the basis of photographs only. The pros and cons of species descriptions in the absence of preserved type specimens are discussed.

South Africa, bee fly, mimicry, Xylocopa , type specimens, photography, taxonomy

A recent paper in Science by Minteer et al. (2014) ignited a controversy by suggesting that specimens are no longer the ‘gold standard’ for species description, and that alternative methods such as high-resolution photography should be considered instead. The quotes above, from the Minteer et al. (2014) paper and a response letter by Krell and Wheeler (2014), give some indication of the apparent polarization of the community on this issue. We take the opportunity here to contend that the apparently antipodal positions summarized in those quotes are both correct, and provide a description of a new species based only on photographs to illustrate and confirm our points.

First of all, let us reiterate the obvious reasons that collecting specimens is highly desirable, and briefly consider and reject arguments against collecting. Specimens are indeed the ‘gold standard’ for species descriptions. Not only do they allow for consideration of a full suite of characters including internal morphology, microscopic and genetic characters, they preserve data for future access with future technologies and future questions. Specimen collections are our greatest treasure trove of biodiversity information and continued collection development must remain a priority. Arguments against specimen collections usually pivot on the potential impact of collecting on fragile populations. Such arguments are weak, since there are vanishingly few examples of scientific collecting having a detrimental effect at the population or species levels and there are very few circumstances under which the removal of a few individuals from a population might seriously harm the ultimate survival of the species. Most, by orders of magnitude, of the animal species awaiting description are invertebrates, and it is especially difficult to make a case against invertebrate collecting on the basis of conservation biology. As invertebrate taxonomists, we therefore agree with the letter of Krell’s (2014) statement that “Describing a new species without depositing a holotype when a specimen can be preserved borders on taxonomic malpractice” (p. 815), but at issue here is the caveat of when a specimen can be preserved. We do not accept opposition to killing and preserving invertebrate specimens on moral or conservation biology grounds, but there are circumstances under which a type cannot be preserved. Furthermore, there is no doubt that collection of potential type specimens will become more and more difficult as restrictions on collecting and transporting specimens continue to increase. These difficulties, along with the rapidly increasing numbers of skilled ‘digital collectors’ who are building collections of images instead of specimens, will inevitably force the biodiversity community to adapt to growing numbers of new taxa recognized without benefit of dead, preserved type specimens. Fortunately, as Minteer et al. (2014) put it, “collecting specimens is no longer required to describe a species.... (p. 260)”. Collecting specimens is highly desirable, but it is indeed no longer required.

As explained by Wakeham-Dawson et al. (2002) and Polaszek et al. (2005), although Article 16.4 of the ICZN Code (ICZN 1999) requires all holotypes that are “extant” to be deposited in a collection, Article 73.1.4 allows for the description of new taxa without preserving a collected specimen by the following statement: “Designation of an illustration of a single specimen as a holotype is to be treated as designation of the specimen illustrated; the fact that the specimen no longer exists or can be traced does not of itself invalidate the designation”. Additionally, we interpret the wording of Article 16.4 to allow for description of a new species on the basis of a lost or escaped holotype, where the term “extant” means a physically “existing” specimen. Thus a lost, escaped, or purposefully released specimen is not “extant”.

Even in the absence of a collected type specimen, current technologies such as high-resolution photography can often provide enough information for a proper description resulting in a readily recognizable and unequivocally distinct newly named species, and in some cases can provide more information (such as colour, soft parts, delicate structures, posture, behaviour) than could be extracted from a preserved specimen. The few previous descriptions of extant new species without a type (or part thereof) have for the most part been restricted to large vertebrates, for example primate species known from only small populations (Jones et al. 2005, Li et al. 2015, Mendes Pontes et al. 2006). We provide the first example of a new insect species described and named solely on the basis of field photographs of the type specimen.

In the example provided here, a highly distinctive fly species belonging to an extremely rare genus (only three other known specimens of two species) was photographed on two occasions and then collected in the field, but the captured individual escaped before it could be preserved as a traditional dead type specimen. Our description, based on photographs of two different living flies, is complete and adequate to identify this species and adds an interesting and easily recognized species to the literature. It not only increases our knowledge of the biodiversity of the area in which it was collected and of the genus in which it is placed but, as we explain below, also provides some interesting ecological and biological information.

The situation leading to this approach is a simple one to understand, since it pivots on an accidentally lost type specimen that might not be collected again due to its rarity. Although this description is by definition singular because it is the first of its kind (at least for Diptera), we predict that a growth in descriptions without physical type specimens is inevitable, and that this growth will result not as much from accidental loss of specimens as from increasing restrictions on collecting. Every taxonomist has been in the position of completing a revision that needs to be rounded out with species from places from which specimens are difficult or impossible to obtain, often because of laws preventing collection or export of specimens. For many of these required taxa, the solution to this problem is for the taxonomist to instead “collect” digital images that, in the case of new species, can represent type specimens. We are not arguing that this practice is generally desirable, only that it is inevitable and increasingly practical when diagnostic characters are distinct and discernable through photographs.

Another trend pushing us inexorably to a wider acceptance of species descriptions without physical type specimens is the rapid growth of extensive, high quality digital image collections dissociated from collections of physical specimens (Marshall, in press). As these image collections become curated just as dead specimens are curated today, these digital “specimens” will find their way into the work of practicing taxonomists, and they will need names. At a time when we need more than ever to identify the biodiversity of this planet before it disappears, it is unrealistic to think that distinct and diagnosable new taxa known only from good photographs and appropriate associated metadata should be organized and referred to only as “undescribed species #nnn”, when they can and should be organized and named using the existing rules of nomenclature.

In recognizing the need to name species without dead type specimens we are not arguing for a loosening of taxonomic standards. In fact, we expect that descriptions unsupported by existing physical type specimens will be subject to especially critical scrutiny by skeptical editors and responsible reviewers. We expect that such descriptions that do not render new species unequivocally recognizable will be rejected, just as they should be if they were based on dead type specimens. Once published, digital representations of type specimens will be much more widely available for use and scrutiny than physical type specimens archived in distant museums.

Dubois and Nemésio (2007) argue specifically against the use of photographs as surrogates for type specimens, suggesting that digital photographs can be faked or misinterpreted. This is, of course true, but one could also alter or misinterpret a type specimen. We see no merit in impeding the documentation of biodiversity on the baseless assumption that there will be more errors, incompetency or dishonesty in descriptions based on photographs than currently exists in descriptions based on specimens or parts thereof.

The species described below was photographed in nature using a Nikon D800 with a 105 macro lens and a hand-held flash. The holotype specimen was not captured, so the image presented serves as representation of the holotype. Morphological terminology follows Greathead and Evenhuis (2001) and wing venation follows Saigusa (2006).

The striking yellow and black vestiture pattern on the thorax and abdomen, and the body shape, are unusual in bombyliids and show a remarkable similarity to xylocopid bees. The model for this possible case of Batesian mimicry appears to be Xylocopa flavicollis (De Geer), which was also photographed at around the same time in the Ndumo area (Fig. 6) and is fairly widespread throughout eastern and southeastern Africa. It is not known if Marleyimyia xylocopae is a parasite of the bee. Other bee fly parasites of Xylocopa are also unusually large-bodied and similarly shaped (e.g., Xenox Evenhuis and Satyramoeba Sack). Previous published records of African xylocopid biologies are few and the only recorded bee fly associate is in the genus Anthrax (Watmough 1974). More biological observations on xylocopid nesting behavior will hopefully lead to better knowledge of the full suite of their possible bee fly parasites.

Figure 6. 

Xylocopa flavicollis (De Geer) from Ndumo Game Preserve, South Africa, 8 December 2014. Photo: Steve Marshall.

Figure 7. 

Map (http://www.simplemappr.net/map/4577) showing the known localities of Marleyimyia natalensis Hesse and M. xylocopae sp. n., in Kwa-Zulu Natal, South Africa.

We thank and Nigel Wyatt (BMNH) for kindly providing photos of M. goliath and the undescribed Nigerian Marleyimyia. Burgert Muller kindly checked the KwaZulu-Natal Museum for possible collected specimens of M. xylocopae. The observations that made this paper possible were made during the KwaZulu Natal Expedition 2014, organized by A.H. Kirk-Spriggs (National Museum, Bloemfontein, South Africa) and V.R. Swart (University of the Free State, Bloemfontein, South Africa).