A meeting of the mapping and data accuracy group was organised in Utrecht by Walter Finsinger, between the 26th and 28th of March 2008. The following people participated: Heather Binney, Simon Brewer, Jacques-Louis de Beaulieu, Walter Finsinger, Thomas Giesecke, Ralph Fyfe, Petr Kuneš, Norbert Kühl and Michelle Leydet.
From left to right: Ralph Fyfe, Heather Binney, Simon Brewer, Thomas Giesecke, Walter Finsinger, Norbert Kühl, Petr Kuneš, Andy Lotter, Jacques-Louis de Beaulieu, Michelle Leydet.
The main points of discussion and decisions taken are given below:
1. Corrections and changes to be compiled in an Excel spreadsheet, listing problems and solutions. Simon to send round an example spreadsheet 30th March. Each person to complete for their country and return to Simon - deadline is 15th April 2008.
2. Restricted data. Heather to write standard letter. List of restricted sites to be circulated by Simon. Each person to contact authors in their countries.
3. Michelle to add 50 new sites to EPD. The chronologies are to be checked by members of the group to help Michelle. Thomas to distribute the chronologies. Agreed that original submitted data should be backed up separately from the EPD.
4. Petr discussed the use of the corrections that have already been made to sites in the Czech and Polish databases. It was agreed that we could simply transfer these changes to the EPD. Petr: can you let us know if this will be possible?
5. Protocol for correcting data: 1. In the case that one taxa count is obviously swapped for another obtain csv file from Simon correct the raw count and send it back to Simon – also send a short note documenting the change. 2. In the case that the correction is not obvious and the author is still active contact author otherwise follow below: a. In one or more samples one taxon has obviously a too low count but the supposable count can be guessed; the published diagram is available and contains the taxon in question: If so follow as under 1 if not than 2b. b. Samples go down to zero for different taxa obvious guesses can not be made or/and the diagram is not available. Mark the samples as not to be included in any analyses. 3. A sample depth occurs twice resulting in a wiggle in the taxa curves. Based on the published diagram and an obvious gap elsewhere correct depth should be assigned to samples. 4. If samples occur in the EPD that are not present in the publication and these samples lead to obvious wiggles in the curves of all taxa or a doubled sample depth can not be corrected with some certainty than these samples should be marked as not to be included in any analyses.
6. Procedure for corrections to sites: a) Produce CSV files for entity counts (30/03/08) b) Produce correction files EXCEL c) Send Simon corrected excel CSV file + comments on changes d) Corrected CSV/Excel files to be converted to Tilia (Simon) e) Send Tilia to Michelle
7. Recording changes. We discussed the need to have a record of the changes made. This could take the form of a set of extra tables in the database, with the following format: a) entity number b) Name of person who identified the error c) Date problem identified d) Description of problem e) Description of action taken to solve problem f) Date problem solved
A number of different projects were discussed as possible focus studies for the group or individual members of the group. It was agreed that the main focus of the group would be a study of biodiversity through time, looking at diversity gradients in Europe (lat/lon), and possibly long-term diversity in Mediterranean/other hotspots. The methods to be used will include eveness, rarefaction, and a functional approach based on plant traits. No deadline was set for this. I think it would good to try and meet as a group at Bonn and outline the work needed.