Sunday 13 October 2013

Data from Photographs

In the course of the next six months I intend to prepare a preliminary analysis of the data gathered from websites over the last five years and to publish somewhere in the peer-reviewed literature. Odd snippets have emerged already but so far (in my article in British Wildlife and also a paper in the journal of the BENHS) but  I have not undertaken a comprehensive review. I'm not aware of any other such analyses but if anybody has come across them then do let me know.

In the past two month there has been quite a recognisable surge in data via photographers as a result of the establishment of the UK Hoverflies Facebook page. I have been amazed by its impact. There is now quite a recognisable community of photographic recorders and a steady stream of new records comes from this source. Other sources such as Ispot and Flickr remain very important, however. The ability of participants in Flickr to communicate with one another has meant that several Flickr users have now become very active recorders of hoverflies. This is fantastic.

The following graphs represent early analysis of the data but present a flavour of the nature of what is emerging from all photographic sources. So far, around 15,500 records have been assembled.

Figure 1. Numbers of identified photographs over the period 2002 to 2013. Data for 2013 are incomplete because the graph represents data to 12 October 2013. It can be expected that data for 2013 will increase substantially by the end of the year and will increase further over coming years.

The data tell an important story about the level of photographic activity. Relatively few people regularly photograph hoverflies. Eleven people (0.5% of recorders) have contributed around 20% (3,120) of all records and ~50% (7,600) of the records have been contributed by 100 people (4.2% of recorders). The data also tell us a great deal more and can be interrogated in many ways. We can say with some confidence that there is no specific tipping point where photographic recording came into its own. There has been a continuing and consistent growth in recording as shown in Figure 1. However, the development of ISpot has certainly helped (since 2009). The data collected since this point are arguably robust enough to show how this type of recording might be used to develop citizen science monitoring programmes. As a taster, Figure 2 shows how photographic activity mirrors levels of insect abundance in response to weather patterns.

Figure 2. The percentage of yearly records occurring each month over the period 2009 to 2013. This illustrates how periods of unfavourable weather affects recording effort and possibly also hoverfly abundance. For example, low numbers in July 2013 reflect a major decline in photographs posted during a period of extremely hot weather.

These notes should be seen as a highly preliminary analysis and should not be used as firm results. More work is required to make sure that duplication has been weeded out and to make sure that 2013 has been properly documented. However, they do indicate the potential of photographic data, whilst also highlighting the possible level of effort required to extract such data.