[Taxacom] Chameleons, GBIF, and the Red List
mesibov at southcom.com.au
Tue Aug 19 19:09:39 CDT 2014
The chameleons paper came out while I was in the field here in Australia with two visiting European zoologists. They were surprised by the numbers in the paper, but not because they or their colleagues rely on GBIF. They asked two rhetorical questions:
Who would use GBIF to obtain reliable records for a particular taxon of interest? (The Europeans didn't.)
Why would anyone use GBIF for multi-taxon or area-focused data?
Expanding, they said that any specialist wanting reliable records would go first to specialist-compiled data. If they didn't exist, the specialist would make a fresh start from collections, the assumption being that collections data (the stuff uploaded to GBIF) could well be wrong. The Europeans gave me alarming examples, both for taxon ID and locality information. (My 2013 audit of GBIF's Australian millipede data wasn't half so alarming.)
This is the specialist viewpoint. There is presumably a perfectly good answer to Rhetorical Question 2 from a non-specialist viewpoint, but if it involves conservation, then the dodgy, broad-brush picture you get from GBIF is probably inadequate, and the non-specialist needs to find specialists for further work: taxon specialists for taxa, and area specialists for areas.
Whether broad-brush users of GBIF data would appreciate that the picture is inadequate is just one problem. It could be partially solved with more 'Warning!' messages from GBIF, and maybe something similar from data providers.
I suspect the reality is that broad-brush users don't care whether GBIF data are OK or not. This is a harder problem. 'Fit for purpose' is a nice-sounding Band-Aid but it leaves open the questions 'How fit?' and 'For what purpose?'
Getting back to chameleons, the standard against which GBIF data were compared was specialist-compiled, as it was in my millipedes audit. Maybe the best future for GBIF isn't in aggregation, but in *disaggregation*. When GBIF reckons it has the bulk of existing records for taxon X, it could cut them out and hire specialists to audit them. The annotation would be from GBIF: 'This taxon's records need to be checked by specialists and should not be used for decision-making'
In the meantime, while funded aggregators continue to accumulate unvetted data from collections and other sources, unfunded specialists continue to quietly compile more reliable data and put it online. Another valuable GBIF annotation would be: 'A vetted alternative to this taxon's records is here [hyperlink]'.
Dr Robert Mesibov
Honorary Research Associate
Queen Victoria Museum and Art Gallery, and
School of Land and Food, University of Tasmania
PO Box 101, Penguin, Tasmania, Australia 7316
(03) 64371195; 61 3 64371195
More information about the Taxacom