Research networks in the Amazon, climate change and conservation

Below are a couple of interesting network structure analyses on the topics "Amazon and climate change" and "Amazon and conservation". They were conducted by Richard Ladle of the Federal University of Alagoas and Oxford University. Research was based on 1063 papers from Web of Science with "Amazon/climate change" or "Amazon/conservation" in title, abstract or keywords. Richards only mapped "the most productive scientists ("the silverbacks") as these are the guys that drive the networks".

Lines represent co-authorship connections (thicker lines=more joint  papers) and balls represent number of papers. The proximity of the balls is an indication (somehow) of how closely people work together. Centrality in the diagram indicates to what extent the person is a key node linking different regions of the network.

The analysis needs three programs: web of science/scopus (to get the data), bibexcel (which converts the bibliographic file and does the network statistics) and Pajek (used for the visualization). The software is freely available, and there is a powerpoint that shows you how to use bibexcel.

Beyond the obvious chance for self-aggrandisement ("how big is your ball?") there are a number of  features of interest.

The first diagram is for "Amazon" and "climate change". It shows a number of interesting features. There are three countries (or four if Laurance is taken as Australian): the UK, the USA and Brazil. In the centre is a closely collaborating UK cluster (Malhi, Meir, Philips and formerly Aragao) which has links across the network, and which focusses on field monitoring and experiments. Also close to the centre are two key Brazilian meteorologists (Nobre/Marengo); Nobre is particular was a central figure in the influential LBA programme in Brazil.. On bottom right is a cluster of climate and biosphere modellers (Betts, Sitch, Jones, Cox, Huntingford) with tight links to each other and fairly strong interactions with the field-focused centre. To the left is a US/Brazil cluster (Nepstad/Costa/Coe/Soares/
Moutinho) with strong links within the cluster and multiple but weaker links to rest of the network . Top left is another prolific figure (Laurance) with links to multiple strands of the network, and Barlow, with some links to other parts of the network. Above centre are two strongly collaborating figures (Bush/Silman) with a strong focus on the Andes and western Amazonia, and on paleoecology. A number of other researchers (Fearnside, Clark, Cerri) are prolific but with only weaker links to the rest of the network. It should be emphasised that the network only shows connectivity under this specific search - two researchers may be collaborating on various topics but if they have not published on Amazon and climate change they would not show up as linked in this diagram.

The second is for "Amazon" and "conservation". Some familiar names from the previous diagram, but also some key new ones, notably Peres and associates, and some changes in relative size of publication volume. The most striking feature is that two discrete networks exist, at least for this search. Why is this? I can safely vouch that it is not competition or animosity, simply happenstance that key connections have not been made in under this research question (incidently, there is a new project on human-modified tropical forests in Brazil, ECOFOR, that involves Barlow, Gardner, Malhi and Phillips, so we can expect these two networks to connect in a few years).

All fascinating, but what do we learn that is new? To an outsider it is certainly a useful overview of key actors and their collaborative relationships, and can also reveal some interesting bulk national features such as the surprisingly strong role of UK scientists, or the lack of visible high publication scientists in South America  outside of Brazil, or in Europe outside of the UK (in the English language, admittedly, which is a significant filter). And how does visibility as a scientist map on to influence in key debates?