[Taxacom] Who uses biodiversity data and why? GBIF Responses

Donat Agosti agosti at amnh.org
Wed Dec 6 04:04:14 CST 2006

"Your post reminds me of a request I once received from a major telecom
company wanting to know which side of the road they should lay their cable
from Sydney to Melbourne to affect the least number of threatened taxa.

It will be a long time before we can answer such questions if ever."

Sometimes in 1996 I got involved in a committee between NASA and NGO/Museums
community to explore possible collaborations.
At that time, the request was to get one or the other landsat scene for
free, which was expensive at that time, and it was thought that this would
help our work.
In the meantime, NASA commissioned to make two complete Landsat mosaics for
the 1990s and 2000, but even more, they actually released all this data (all
in all >20,000 scenes) to the public through the Global Landcover Facility
(http://glcf.umiacs.umd.edu/index.shtml). At the same time, the global
mosaic became one starting point of Google Earth, which is now updated with
increasingly high resolution data, even for remote corners of the world. The
power of having this data accessible (mainly for non-biological issues) is
so dramatic that Google is buying ever more high res data.

But we don't seem to jump onto this bandwagon and produce data which can use
it to the max. 

Brian Fishers work in Madagascar shows what could be done. Here for example
a specimen record of an Aphaenogaster specimen, Brian confirmed that it is
on the right side of the street


Plus there are over 120 sites where he collected using standard techniques
yielding in hundreds of species for which this kind of data is available. In
some cases, where he is using transects, he has one point, but since the
point is a GPS record, it will fall into the proper (fragmented) habitat

I am aware, this is difficult to achieve with old data - but then there is a

Again, if we want to do a monitoring program (eg in the framework of
countdown2010) then we need to make use of what new satellite imagery allows
us to do. Data can always reduced in resolution so to fit the needs of
distribution model inputs, but low res can never made high res. Having hi
res will allow the land use mananger to take care of the very bit of habitat
where specimens have been sighted.

In this connect, I would also recommend that however does now, or plans to
do fieldwork, keeps a bit of a satellite image, if possible as close to the
date of the collection, projects the gps record on it to assure, the point
is in the habitat she was working in. If this is properly archived with the
necessary metadata, colleague in the future could not only get a point, but
also a piece of landcover and thus do some much more advanced studies.

If we and the conservationists, such as the RedLister at IUCN, their SSC's
global voluntary network of 6,000+ specialists or the World Commission on
Protected Area , could switch into using the highest resolution data, rather
then kind of polygons, then there would be a tremendous. GBIF would be
ideally positioned to bring these so far rather disjunct sources of data

This seems the only way we can have an impact in optimizing the protection
of biodiversity, and at the same time get a better idea about the pattern
and process of the distribution of species. It seems to me, that we arrived
at a point, where not technical issues are the limit, but our mental
barriers to live up to this chance.


-----Original Message-----
From: taxacom-bounces at mailman.nhm.ku.edu
[mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of
taxacom2 at achapman.org
Sent: Wednesday, December 06, 2006 1:31 AM
To: taxacom at mailman.nhm.ku.edu
Subject: Re: [Taxacom] Who uses biodiversity data and why? GBIF Responses

Work I have done, Donat, would suggest that retrospectively georeferenced
data is generally in the range of uncertainty of 3-10km (with some worse and
occasionally far worse)

There are examples where data from GBIF are, and will, be usefull in
conservation planning.  To explain fully here is very difficult, but already
such data is being used in Papua New Guinea where the majority of
collections from there are not held locally, but are spread throughout
various institutions in Australia (especially the West Australian Museum,
Royal Botanic Gardens in Sydney and the Australian National Herbarium and
Insect Collections in Canberra) and elsewhere around the World, including
Kew and the Natural History Museum in London as well as a number of US
institutions.  Without those data, it has proved very difficult to make
robust conservation decisions in that country.  Tools are now being used to
integrate those data to look at priority areas.  Such data are extremely
valuable at the meso scale, but local and on-ground decisions will still be
needed at the micro level to determnin the boundaries of such conservation
actions.  The global and regional/local m
 ust go hand in hand.

Another example of using global data such as available through GBIF is in
the determination of areas for conservation of the Wild Relatives of Crop
Plants.  With global change (including climate change), such conservation
actions are essential and can only be done on a global scale (or at least
large regional scale) for many taxa. Work currently going on involves,
Potatoes, Wheat, Rice, Apples, Walnuts, Coffee, Pistachios and many, many
other taxa. 

There are many other examples where the taxa will become increasingly useful
- both observational data as well as specimen data as mentioned by Meredith
in her lengthy and excellent response.

Indeed, with some of the taxa Bob is using, global data is likely to be of
little value, but that doesn't mean that it is not valuable, if not
essential, for other taxonomic groups.

Conservation is importanty for more than rare and threatened taxa -
especially in a world of changing climate regimes.

AnotheOne such example where such data are beginning to prove valuable is in
managing the spread and control of invasive species.  By using collections
in native ranges, climate profiles can be built up which can then aid in
determining areas of most effective control.  This is an issue of
conservation that is working at the global scale, and will do so
increasingly. An example is in the spread of Cactoblastus cactoprum in
Mexico and southern USA, where data from Australia has been used to
determine areas of likely spread.

Amazing what one can do while sitting at an airport awaiting a delayed plane
- thank goodness for airport lounges!

Arthur D. Chapman

>From Donat Agosti <agosti at amnh.org> on 5 Dec 2006:

> The failure of our community to build up specimen observation based
> monitoring system (eg within the target2010 initiative to stem loss of
> biological diversity) is probably the best case to demonstrate the
> disconnect between systematics data and conservation.
> It is not just enough to collect old data, but also to assure that
> collect
> new data as well. GBIF is certainly indispensable in this to show what
> is
> out there.
> In 1996, Norm Johnson run a project to collect all the data for one
> genus, a
> hymenopteran Pelecinus with three species. He included 7859 records from
> 100
> collections, 2,310 distinct localities in 18 countries. Still few
> countries
> are missing, but it gives a very interesting insight into various
> biases,
> such as using data of only one or few data.
> http://rd.mailshell.com/iris.biosci.ohio-state.edu/projects/tpp/
> GBIF is certainly at best slowly getting into a position to address such
> problems. There are tools to measure some of the deficiency of the data.
> At
> the same time it seems to be simple task to figure out distribution
> ranges
> for species accessible through GBIF.
> But even in the case, where new data has to be colleted it ought be made
> accessible and, hopefully in ten years, it could be used to measure and
> monitor conservation management.
> Of course, this depends on using sort of standardized protocol to
> collect
> data in the field. And that is of course one of the impediment: A
> taxonomist
> want's to get the biggest number of species, a conservationist ought to
> have
> a representative sample. 
> If conservation and systematists are really feeling they need each
> other,
> then both ought to make a step towards such as collaboration, and since
> these are large projects, GBIF ought to be included so data can be
> accessed
> in the easiest possible way.
> What we do now is to declare biodiversity hotspots, new parks down to
> management plans. But most of these projects do not have an element
> allowing
> to measure change in species over time besides using proxys.
> Google-Earth contribution to this is to show how very weak most of the
> historic data is. Whilst we can now increasinglgy soom into few meters
> anywhere in the world, and thus could see one which side of a stream a
> specimen has been collected, the geographic error of a posterior
> georeferenced data is probably at about 20 to 50km.
> The old data though allows still well informed sample strategies, but
> only
> if we have not too much sampling artifacts, which brings us back to
> Johnsons' work and the need that we need all to contribute - may be with
> some focus on taxa where further analysis and usage is going on
> Donat
> -----Original Message-----
> From: taxacom-bounces at mailman.nhm.ku.edu
> [mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of Rob Guralnick
> Sent: Tuesday, December 05, 2006 2:55 AM
> To: Bob Mesibov
> Subject: Re: [Taxacom] Who uses biodiversity data and why? GBIF
> Responses
> >
> > I am also surprised that you seem to see so much potential in
> 110,588,578 
> > occurrence records, i.e. the bigger the number, the greater the
> usefulness. 
> > Every biodiversity conservation effort I've been involved with has
> dealt 
> > with rare and geographically restricted taxa. Prior to these efforts,
> there 
> > were no or very few occurrence records for the taxa concerned, because
> a 
> > majority of occurrence records are for common and/or widespread
> species.
> In 
> > order to plan successful conservation, large numbers of _new_
> occurrence 
> > records had to be obtained by diligent field work. By the time these
> records 
> > entered local museum databases, the conservation plan was already in
> place.
> >   
>     Bob and Taxacomers ---
>     This is an important topic to hash out, and I worry that no response
> to Bob's comments will be taken as implicit agreement rather than 
> "argument exhaustion".   I truly disagree with Bob's views here, and for
> me it comes down to a simple point:   If one is interested in 
> conservation or biodiversity at local levels, there is absolutely no 
> compelling reason why one shouldn't determine the past sampling in the 
> local region of interest by accessing GBIF data.  To do so is 
> effectively zero-effort.   How many of you would spend the effort, 
> money, gas, and wear and tear on the environment doing needless 
> fieldwork when the data to do good research and make good management 
> decisions already exists?  The rapidity at which we are losing 
> biodiversity requires us to be more efficient with our time and 
> resources.  In cases where more fieldwork is required, lets use the 
> existing occurrence data to make good choices about how to effectively 
> choose new sites.  Lets do what makes sense given what we already 
> know.   And lets not presume or make blanket statements about the sample
> quality or usefulness of our global biodiversity registry - lets go out 
> there and test what we can legitimately do with it and how it will help 
> us more quickly make a wise policy of conservation and development. 

=== message truncated ===

Taxacom mailing list
Taxacom at mailman.nhm.ku.edu

More information about the Taxacom mailing list