County moth datasets tend to be large and 'messy' affairs - messy in the sense that they are large aggregations of data from a variety of sources, collected using a variety of methods. Some people will run a mercury vapour trap all night long in their garden, several times a week; others will run an occasional actinic trap for a few hours on a nature reserve; others will just send in a few sightings of moths they've found by day. Is it possible to draw any overall conclusions about which moths are increasing or decreasing from this mass/mess of data?

One way I've tried to do this with the Berkshire macro-moth database is to look at the number of records for a particular species in relation to the number of records of all moths recorded each year. This takes into account the variation in total recording effort each year. Similarly I've looked at number of individuals of particular moth species compared to total number of individuals recorded each year.

Some examples of the resulting data for Berkshire is shown in the graphs on the right, using data extracted via these MapMate user queries (.txt file download). Read on for more details of how the queries work.

The abundance queries exclude all records that have no abundance data (i.e. those with a quantity of zero, which is MapMate-speak for "Present but not counted").

Data is not calculated for those years in which fewer than 2000 moth records were received (for the no. of records queries), or fewer than 5000 individuals were recorded (for the abundance queries). This is an arbitrary cut-off, the hope being that it is enough to exclude the worst effects of poorly recorded years.

There is no other filtering however, and the queries will pick up all other records, whether they are one-off sightings of day-flying individuals or full MV trap lists. The assumption is that although the data is messy, it will be more-or-less equally messy each year, and thus useable as an indicator of trends. I don't know how far this assumption is justified!

You should ensure that your filters are set to something sensible. In Berkshire I only use this for macro-moths (our micro-moth data is not comprehensive enough to analyse in this way). I'm currently only using them to look at the period 1995 to 2008, where we have a more consistent run of data.

There are two main queries in the above document:
- Abundance for <taxon> per 1000 total abundance
- Records for <taxon> per 1000 total records

The other queries in the document are the various subqueries that are needed to make the two main queries work - NB that the subqueries MUST be given the exact names that I've given them.