Thursday, 17 November 2016

Analysing photographic records of hoverflies

When I started extracting photographic records I will admit that I was highly sceptical that they were of much use. In the intervening years I have been convinced that there is an important place for photographic recording, but it needs to be embraced with a degree of caution. The data extracted from photographs now comprises well over 60,000 records and can be used to paint a clear picture of the strengths and weaknesses of this approach to biological recording as it applies to one family of flies. Rather than place the analysis in inaccessible academic literature, I have started to produce annual reports that explore different aspects of the possible use of these data. the second report can be downloaded from

The lessons learned do not necessarily apply across all taxa, but I think there are good grounds for encouraging photographic recorders at the same time as continuing to emphasise the importance of traditional recording from specimens.

On the positive side, photographic recording is valuable because:
  • It reaches out to a much bigger constituency of natural historians and also to people whose main interest is photography rather than natural history.
  • It allows the generation of a large body of data for those taxa that are regularly encountered by photographic recorders. These data largely comprise the most visible and readily recorded species and provide a valuable window into the live of animals that many specialist might otherwise ignore. In the case of hoverflies photographic records are at the point where they exceed the data arriving from traditional recorders.
  • It helps to generate a new body of individuals with basic (or more advanced) taxonomic knowledge. Some of these people may eventually more on to taking specimens and making more detailed studies, and may eventually take on some of the essential administration of biological recording.
  • Photography can be active or passive. Some of the most comprehensive datasets are compiled by people with limited mobility but plenty of patience. Sitting, watching and photographing over several hours can help to amass a really nice dataset and is something that can be done by disabled people who still want to connect with nature.
 There are some drawbacks, however:
  • Not all animals and plants can be identified from photographs. In the case of hoverflies, around 60% are doable on at least a proportion of occasions. That leaves 40% that cannot be recorded and unless there are traditional recorders who retain specimens this element of the fauna will be under-recorded.
  • Identification from photographs is an art (science) that depends upon many years of experience and a comprehensive knowledge of the taxonomic group in question. It depends upon a mixture of experience and caution. I have been accused of being too timid in providing identifications, but it seems to me that caution is an essential component of sound recording. If data are based on a gun-ho approach then they will be vulnerable to challenge and the results of analysis will be open to criticism.
  • There is a very small body of specialists who are willing to engage and to provide identifications. There are even fewer who are willing to extract data. This is always going to be a limiting factor and will inevitably act as a brake on what can be done.
  • The composition of the data differs markedly from the data that is assembled by traditional net, pooter and microscope. I have been looking at this in some detail and suspect that it will start to be necessary to split datasets into two - one for photographic records and the other for data compiled by recognised specialists. Both datasets will tell important stories but the messages may not completely agree with each other.

No comments:

Post a Comment