Saturday, 24 August 2013
Identifications from Photographs
Unfortunately, quite a lot of hoverflies are identified using characters that are difficult to depict in photos or require manipulation of the animal to get the right orientation. Thus, it is not uncommon for people to post photos that cannot be taken much further than genus, and occasionally not even that far! Most people when told that a definite ID is not possible will accept this and move on. Occasionally the person providing advice gets a torrent of abuse that they are not living in the real world and should ‘wake up to reality’. So, I thought it might be worth discussing the issues in a bit more detail. This note is a brief canter through obvious points and must not be regarded as definitive.
The first point that emerges from the photographic record is that quite a narrow spectrum of hoverflies gets photographed regularly. These are primarily the showy ones – bright colours attract attention. Many of these are perfectly doable at least to species pairs and often directly to species. There are, however, a few genera that regularly crop up and cause problems:
Cheilosia: There are some obvious species, led by C. illustrata, but accompanied by C. chrysocoma and perhaps C. grossa and C. albipila. A few regularly get photographed: C. bergenstammi, C. pagana, C. variabilis. In some cases one can be reasonably certain but in others it is more conjectural and circumstantial. Those quoted, together with occasional specimens of C. impressa are likely to be accepted as records subject to a good photo. The surprise is C. caerulescens which is often readily recognisable and has a certain jizz about it that I cannot describe. It is not infrequently photographed in gardens but I would be at a loss to say quite how/why I would recognise this apart from the facial profile and when the wing shade is obvious – it is considerably hairier in some places but cannot be described. These apart, I fear the majority cannot be done although one might have a good idea of what something is likely to be. I guess the difference between me and some recorders is that I am immediately thinking of what happens to the record and how this affects data reliability.
Epistrophe/Eupeodes/Melangyna/Parasyrphus: A few are readily identifiable such as E. eligans and E. grossulariae. Many firm IDs rely on micro characters such as wing microtrichia or hair colours that are generally poorly depicted in photos. Thus, the degree to which an ID can be given is often directly proportional to the quality of the photo. I’ve seen some fantastic stacked shots that make ID quite possible for many species, but not all features of importance are readily depicted on a live photo because wings may cover them or they may be on the underside of the animal.
Eumerus: these are generally not possible because the critical characters on the hind femur are obscured unless the animal is at least unconscious and inverted. It has been further complicated because a new species has been added to the list - Eumerus sogdianus, which is very similar to E. strigatus and can only be separated on genital characters.
Melanostoma: At certain angles, male M. mellinum and M. scalare can be separated with modest confidence, provided they are from lowland locations. Even so, there are some where the elongate abdomen of M. scalare is not as pronounced as illustrations suggest. It is highly possible that a further species lurks here and several specialists have also suggested that there is a much bigger complex. In females, the dusting on the frons (not withstanding Species A and M. dubium) is distinctive provided one is looking directly down on the head. The problem is that extent of dusting is frightfully variable, with lots of intermediates.
Platycheirus: A few are eminently doable such as P. granditarsus, P. rosarum and perhaps P. manicatus. So too are many, but by no means all, P. albimanus. Thereafter, life becomes complicated and those males where the pits on the underside of the front tarsi are key characters are not generally doable because the animal is standing on the character concerned! In males, greater chances of ID are possible if the front legs and especially the tarsi are sharply depicted. Females are much trickier, and in general I would hesitate to tackle many of the yellow ones. Now, in reality, many of these segregate themselves because relatively few are photographed: most occur low in wetland vegetation and only a few will bask on leaves and flowers sufficiently regularly to get noticed. Also, they are small and therefore less likely to be noticed.
Pipizini: A very tricky tribe, although the odd specimen is doable. Pipiza austriaca is occasionally obvious and so too is Pipizella viduata at some profile angles where the antennae are well illustrated. These apart, there is little to go on without micro characters and genitalia examination for some.
Sphaerophoria: Lots of people put names to females, and although there are continental keys they are not reliable. From personal experience I have found them to be extremely tricky with preserved specimens that can be highly magnified, so I fear there is a minefield if one is tackling them from a blurry photo or even a reasonably sharp one. The resolution is rarely likely to be adequate. In males, apart from S. scripta, S. rueppellii and S. loewi, there are no good external characters and one is dependent upon the structures within the genital capsule. Male Sphaerophoria are bashful and hide these bits, so ID from a photo is pretty well impossible.
Syrphus: We see hundreds of shots of Syrphus and many relative novices try to put a name to specimens. It is not unusual to see names being allocated on the colour of bands or other characters that are not used in the keys. In my experience, it is rarely possible to do males, although occasionally one sees a nice shot of S. torvus in which the eye hairs are visible. In females, S. ribesii can be separated if the hind femora are fully visible (not possible from above); but only if one assumes that S. rectus does not occur in the UK. The jury is out on that one and probably we won’t know until there is DNA analysis of European and North American specimens.
There are a few other challenges, but the above is enough to get started with. The key point that emerges out of this is that the better quality the photo is, the greater the likelihood of getting a firm ID. In a recent debate it was pointed out to me that I should be providing advice on the angles that people should be taking shots from. This really is not practical – the angles depend upon a thorough knowledge of the insects in question and therefore until the photographer has gained an understanding of the genera and what they are photographing they will almost inevitably miss the key character. Thus, it seems to me that the best advice one can give the novice is to try to get to know the genera first. This is the approach we take in our training courses – pretty much the first half day ID session focuses on embedding the use of the key to get to Tribes. Thereafter greater refinement slots into place.