Online tools reveal a lot about world science — except location. 'Geotags' can fill a knowledge gap and throw up surprises, says Nigel Pitman.
Anyone who has done a search on Google Scholar can attest to the astonishing advances we have seen in accessing, organising and analysing scientific literature.
As old card catalogues are replaced by lightning-fast digital reference tools, researchers have an increasingly clear view of which questions have been addressed, and when, why, how and by whom.
But there is something missing from that list. Online databases and reference tools still do a very poor job of telling us where science is being carried out.
Which country's rivers have been studied more intensively by ichthyologists (fish scientists)? What proportion of scientific publications about the developing world has been written by researchers actually based there? What's the best-studied field site in the Amazon?
Quick and rigorous answers to questions like these remain beyond the reach of today's database tools. But there is a simple way to solve the problem — assign a 'geotag' to every article, book and thesis in the scientific literature. A geotag is a set of geographic coordinates that shows where the authors of a publication obtained their data.
Imagine for a minute that we could do it. The result would be a map of the world capable of pinpointing every site where scientists have ever collected data, and capable of telling you, via the reference databases that already exist, what has become of the findings.
Point at a spot on the map and you would have a comprehensive list of what has been published about that particular region — and how many scientists have written about it, when, and other information.
Is this too ambitious, or too difficult? Not really. In 2009, colleagues at the University of Florida and I built just such a map to describe patterns in ecological research across the tropical Andes and the Amazon basin. 
We did it the old-fashioned way: paging through the two leading journals on tropical ecology from 1995 to 2008 in search of any article based on work in those regions, and then noting the coordinates of the sites where field work was carried out.
The result was a 'cloud' of 278 points scattered across the map of tropical South America — the field sites where data had been collected for 373 journal articles.
The dataset revealed a number of patterns no one had noticed before. Who suspected, for example, that tiny Ecuador led all South American countries in the number of field studies published per square kilometre of territory? Or that three field sites accounted for more than half of all Amazonian publications?
There were more surprises. The field site responsible for producing the most peer-reviewed science on western Amazonia turned out to be a thatch-roofed shack in southern Peru, 100 kilometres from the nearest road. And more than 85 per cent of what was written about Andean biodiversity in that period was based on field work in the northern half of the cordillera (mountain range).
Pinpointing gaps and opportunities
Given the speed and creativity with which the world's scientific literature is being organised, it is just a matter of time before scientific geotagging becomes a more widespread research tool.
But there are several reasons why science agencies in developing countries should aim to get a head start.
For example, maps like ours can help identify especially productive field sites (field stations, nature reserves, monitoring sites and others) that deserve long-term financial support but are often overlooked by funding agencies.
These maps can also identify knowledge gaps — such as regions that scientists have not yet explored adequately — and help design incentives to fill them.
Most importantly, our data indicate that most publication hotspots in South America are also hotspots for training. In other words, these maps can help science agencies locate the unsung but crucial places where young developing world scientists receive hands-on experience in the field.
It should be relatively easy to build a comprehensive map of field science in developing countries because there are much fewer field studies than in the developed world. It would be easier to map all the research sites in Guyana than in the United Kingdom, for example.
And while maps of geotagged publications will be useful for many different kinds of science, it may be that they will be most valuable for field and conservation biology. In these fields, where scientists' ability to organise and communicate information is directly related to their effectiveness in protecting the Earth's embattled wilderness, mapping the world may be one small step towards saving it.
Nigel Pitman is a research associate at the Center for Tropical Conservation at Duke University in Durham, North Carolina, United States. He is currently based in Paraná, Brazil.
 Pitman, N.C.A. et al. Volume and geographical distribution of ecological research in the Andes and the Amazon, 1995–2008. Tropical Conservation Science 4, 64–81 (2011)