MammalWeb UK hits 100% classified

10-04-2019

In March, we reached a major milestone. Specifically, for the first time since the MammalWeb UK project began, we reached a point at which every image sequence submitted to the MammalWeb UK project had been classified at least once! This is a great achievement and we’d like to thank everyone who has contributed classifications, as well as Helen, whose tireless work on the web platform has made Spotting such an effortless (and, we think, addictive) process!

Although MammalWeb UK is now just one of the projects on the site, it remains the project to which most people contribute data. At the point when all sequences had been classified at least once, it included almost 80,000 sequences from over 300 sites. As you can see from the detailed map (in which the MammalWeb UK sites are shown in red, overlying blue sites from other projects), survey efforts remain dominated by sites in the North East of England, where the project began. Zooming out to the UK map, however, you can see that MammalWeb UK is continuing to expand. If you are, or know, a camera trapper elsewhere in the country, please do upload your camera trap photos/encourage others to do so. This way we can continue to expand our coverage, and consequently learn more about mammal ecology across the UK.

At the point at which all image sequences had been classified at least once, the data suggest that about 62% of the sequences contained identifiable animal life. So, what do they contain? As you can see from the graph on the left below, the most frequently sighted animal is the grey squirrel which, as many of you will know, is an invasive non-native species. Nevertheless, there are a number of native species that appear frequently, including roe deer in about 1 in every 12 sequences you view, and badger or red fox in about 1 in 18 sequences. Bear in mind that these statistics are based on what species people have said are in sequences. As we get more classifications per sequence and more conviction regarding what is pictured, this could change.

A more intriguing picture arises when we consider the number of sites at which different species have been photographed (see graph on the right below). This suggests that roe deer and red fox are actually more geographically widespread in our data than are grey squirrels. This probably arises because they are more wide-ranging: when grey squirrels are active around a camera, they are usually very busy, generating lots of image sequences; roe deer and red foxes turn up at more cameras – but don’t linger by the camera for so long.

As contributors are increasingly trapping in other parts of the country, these patterns will almost certainly change. We have recently seen muntjac in the data set from a site in Worcestershire, as well as fallow deer from a site in South Wales. Who knows what other species we might see as the project continues to expand?