Sunday, 9 October 2016

Identification Caution

Following on from the last post I made regarding the identification of the new species of mammal down the mill I have an update.

Originally I believed the organism found to be a Water Shrew, I based this on the size and rough colouration of the pelt. Unsure I put out a plea for assistance on the Mammal Society Facebook page. Over the week the consensus came in that the image was in fact of a Common Shrew (Sorex sp.)

At the same time this was going on a paper was published in Nature called 'Species identification by experts and non-experts:comparing images from field guides' by researchers from the University of Kent. This research looked at how accurately experts and non-experts could identify images of UK Bumblebees.

They discovered that between experts and non-experts overall accuracy was 56%. This is a remarkably low figure especially given the increase in citizen science used in research these days, many of which I take part in myself such as Snapshot Serengeti. The study adds to other research conducted over the years looking at the efficacy of identification,

Obviously, experts have a higher strike rate but the fact is that much of community-sourced data uses a user base of with a wide range of skills from the novice to the professional. There are also an amazing array of very high-quality field guides that can help everyone improve their skills.
Nevertheless these types of papers ensure that we do not always take information at face value and that at some point each of us as wildlife enthusiasts must take responsibility for our own identifications.

As a younger birder I was convinced I had seen a flock of Twite and even included it in my notes, when questioned by the County Recorder it became apparent to me that I did not have the evidence to support my claim. It was a moment that made me think long and hard about what I record and with what confidence.

It is the responsibility of every wildlifer to  take care in their identifications and today with a plethora or guides, message boards, and forums it is much easier to crowdsource an ID. It is also the responsibility of the scientists to balance the error rate with the information they are collecting.
What is needed is balance in this field, citizen science allows the collection of far greater datasets than would ever be achieved with the use of experts alone something we cannot ignore and I don't think anyone can deny the value such projects bring to the field of ecology.

Use such studies as a warning to over interpretation and use each new identification as a learning point. I now know to check the teeth of shrews, the hairs on the hindfoot and the covering of the underside of the tail.