ConservationBytes.com

—

If you’re a regular reader of CB.com, you’ll be used to my year-end summaries of the influential conservation papers of that calendar year (e.g., 2016, 2015, 2014, 2013), as somewhat subjectively assessed by F1000 Prime experts. You might also recall that I wrote a post with the slightly provocative title Essential papers you’ve probably never read back in 2015 where I talked about papers that I believe at least my own students should read and appreciate by the time they’ve finished the thesis.

But this raised a much broader question — of all the thousands of papers out there that I should have read/be reading, is there a way to limit the scope and identify the really important ones with at least a hint of objectivity? And I’m certainly not referring to the essential methods papers that you have to read and understand in order to implement their recommended analysis into your own work — these are often specific to the paper you happen to be writing at the moment.

The reason this is important is that there is absolutely no way I can keep on top of my scientific reading, and not only because there are now over 1.5 million papers published across the sciences each year. If you have even the slightest interest in working across sub-disciplines or other disciplines, the challenge becomes more insurmountable. Finding the most pertinent and relevant papers to read, especially when introducing students or young researchers to the concepts, is turning into an increasingly nightmarish task. So, how do we sift through the mountain of articles out there?

It was this question that drove the genesis of our paper that came out only today in Nature Ecology and Evolution entitled ‘100 articles every ecologist should read‘. ‘Our’ in this case means me and my very good friend and brilliant colleague, Dr Franck Courchamp of Université Paris-Sud and the CNRS, with whom I spent a 6-month sabbatical back in 2015.

So, how can one objectively identify such papers, and is such a list even relevant in today’s massive array of research papers? Franck and I tackled this problem first by contacting hundreds of ecology ‘experts’ — i.e., those people who occupied editorial positions for some of the best generalist ecology journals — and then asking them to provide 3-5 (or more) of the articles that they thought should be ‘essential’ reading in ecology.

But we didn’t stop there, because this approach merely identified the candidates preferred by individual editors. We then asked these same people to score a randomly generated sample of 20 of the more than 540 articles suggested as being in the ‘Top 10’, ‘Between 11–25’, ‘Between 26–100’, or ‘Not in the top “100”’. Each voter could score as many randomly sampled sets of 20 papers.

From this voting procedure, we generated a ranked list of all 544 papers, and then looked at the relationships between rank and various aspects, such as journal Impact Factor, citation rate, article age, and type, field, or approach. The interesting relationships that emerged were that more highly ranked papers tended to be older, and have higher citation rates (citations/year), but there was no obvious dominant type, field, or approach, nor were the most highly ranked articles in journals with the highest Impact Factors.

Some of these trends are expected, of course, but here is where it gets interesting. When we restricted the scoring to papers that the voters had actually read (cf. just knew about, but hadn’t yet read), the relationships to age and citation rate disappeared. This means that the experts were driven by the reputation of articles in addition to their actual ‘quality’ (as potentially assessed by them). This means that that defining essential-reading lists is not a futile exercise, because it highlights what even the most-experienced researchers should ideally read.

If you can access the article directly, then the 100 top-ranked list is available in a table (Box 1), and Franck Courchamp has provided the full, ranked list on his blog.

Yes, there are useful services out there like F1000 Prime that attempt to identify the hot topics of the times, but there is really no way to identify the ‘essential’ papers to read using this approach. Our retro-active exercise helps in this regard, and should provide ecologists will a sample of articles that we should probably all read (or re-read).

CJA Bradshaw