Monday 3 October 2016

Composition of the photographic dataset of hoverfly records

A recent thread on the NFBR Facebook page drew attention to a piece of work looking at the differences between identification by 'experts' and various (relative) novice recorders. This paper generated a range of comments and stimulated me to reflect on the advent of photographic recording. I took the view that the glass is definitely half full and probably a good deal more full than the most sceptical specialists might think!

I have written previously on the importance of modern media as a way of bringing 'new blood' into biological recording and the need to embrace this new paradigm. The Hoverfly Recording Scheme has done so for many years and I think the dividends are there for all to see: we have seen a massive increase in recorder effort but have also started to recognise that this sort of effort is unsustainable without a fundamental change in the way recording schemes are organised, taxonomic guides are written and perhaps even in the way taxonomists work to describe species.

Live animal taxonomy is a challenge, and probably cannot be taken to the refinement that is possible with preserved specimens and modern molecular analysis. BUT, it is likely to be the major source of data in the future and we really need to understand its implications for data analysis and for conversion of these analyses into land management policies. This paradigm shift has been the driving influence behind my interest in photographic recording and trying to understand and quantify what is possible.

After at least eight years of data extraction from photographs, I think we are in a strong position to analyse the potential of photographic recording for hoverflies. The sheer scale of the data arriving is obvious (figure 1), but what does it really mean in terms of the spread of species recorded? Furthermore, are the numbers of records translated into something useful for land management policy? More analysis is needed to come to any firm conclusions but the data do suggest that at least for hoverflies there is potential to use photographic recording for certain analyses. The spread of species is substantial (over 150 species in 2016) and, as can be seen from Figure 2, the tail of irregularly reported species is long. At the moment I am far from clear how this compares with traditional data and will have to access the full dataset to gather a better understanding of this relationship.
Figure 1. Monthly records of hoverflies based solely on photographs extracted from Flickr, iSpot and Facebook


The spread of species is substantial (over 150 species in 2016) and, as can be seen from Figure 2, the tail of irregularly reported species is long. At the moment I am far from clear how this compares with traditional data and will have to access the full dataset to gather a better understanding of this relationship.
Figure 2. Composition of the 2016 dataset (23,790 records)
 Figure 3 (below) also illustrates how the dataset is dominated by an ever-shifting range of species throughout the year, with some obvious surprises amongst the listings of the top 20 species for each month. There must be a health warning, however, as data from traditional sources who retain specimens have yet to be analysed. My past experience has found that there are substantial differences and that one has to be clear about the source of data before drawing any conclusions about species abundance and distribution.
Figure 3. The 20 most frequently reported species each month in 2016 organised in rank order for each month.
More analysis will be provided in due course, but I hope that these three figures alone will act as a stimulus to other recording schemes to engage with photographic recorders. There is plenty of good will out there and lots of potential for growing our knowledge of Britain's wildlife. My one word of caution is that the level of interest can be overwhelming and consequently we need to change the way in which recording schemes are run. In future we will have to look far more towards a team effort rather than one or two people working on their own. That means that over time some of the contributors must be given the opportunity to play a more active role in data management and mentoring; something that the Hoverfly Recording Scheme is currently working to address.

No comments:

Post a Comment