Multiauthorship, an Indicator of the Trend Toward Team Research in Ecology

Older scientists are aware that before World War II most research was published by single authors, whereas since then the trend has been toward team research and multiauthored papers. We have investigated this trend in ecology by counting the number of research articles with 1 author, and with 2, 3, 4, 5, and more than 5 authors, in the British Journal of Ecology and the American Ecology since the latter began in 1920. They exhibit similar patterns, which we show decade by decade. The number of articles published (total over nine decades was 15,740, 71% in Ecology) rose exponentially from 508 in the decade of the 1920s to 4139 in the decade 2000–2010 (Fig. 1). The pattern of authorship changed greatly from the beginning to the end of this period (Fig. 2). In the 1920s, 85–87% of articles were written by single authors in the two journals, and by the 2000s only 10–16%, with the median number of authors ~2.2–2.3. Dual authorship dominated in the Journal of Ecology by the 1980s, but was below 45%, and in Ecology by the 1990s, but was below 40%. Dual authorship exceeded 2% already in the 1920s, and triple authorship by the 1930s (Ecology) and 1940s (Journal of Ecology). It took five and four decades, until the 1980s, respectively, for articles with 4 authors to rise above 2%, but only successive single decades to reach 5 and 6 authors. It will be interesting to see whether this rapid trend continues. In this connection, our colleague Clarence Lehman points out that over the last couple of decades the unparalleled ease of internet communication allows even far-distant team members to communicate with one another or in groups, sharing computer screens and jointly constructing and editing documents. The increase of total publications in Ecology and the Journal of Ecology from the decade 1920–1929 to the decade 2000–2010. The decade of World War II is excluded from the equation; log10N (number of articles) = 2.44 + 0.0117D (from 20 for the 1920s to 100 for the 2000s; R2 = 0.98). Contrasting author numbers from the decades 1920–1929 and 2000–2010. The shift toward multiauthorship is shown in Fig. 3. Single authorship declines slowly in both journals to 70–79% through the 1950s, and then declines rapidly to 8–13% in the 2000s. Dual authorship rises from 11–12% in the 1920s to around 40% in the 1990s, and then declines to 26–28% in the 2000s. Articles with 3, 4, and 5 authors begin at 0–2% in the 1920s. Those with 3 authors differ somewhat in the two journals. In Ecology there is a gradual rise from 2% in the 1920s to 13% in the 1980s, and then a rapid rise to 24% in the 2000s. The Journal of Ecology remains at 1–2% through the 1950s and then rises to 24% in the 2000s. Those with 4 and 5 authors remain low until the 1970s and 1980s, respectively, and rise in the 2000s to 17% and 9–11%. Changing patterns of authorship between the 1920s and the 2000s. Solid lines, Ecology; dashed lines, Journal of Ecology. (A) Single and dual authorship, (B) three, four, and five authors. The exponential rise in ecological publications shown in Fig. 1 is notable, due presumably to increasing government funding after World War II that reflected increasing environmental awareness in society. When EG began research in 1946, ecologists were located in biology (or botany or zoology) departments—not all of which found them necessary. In the 1960s they began to form separate departments that gradually increased in size. The Ecological Society of America, which had only about 600–700 members from the 1920s to the 1950s, then began to grow rapidly, membership reaching 6000 in 1977 (Burgess 1977) and 10,000 in 2014. Increased funding allowed the setting up of teams, often interdisciplinary in nature, and scientists gradually got used to—and to expect—working in them. Since World War II a great many journals with a connection to ecology have appeared. Entering “ekolojinet.com/journals.html” on Google provides a list of 627 titles with some relevance to ecology. On that list, titles with “eco” or “eko” in them number 125, of which 110 titles begin that way. Titles with “environment” or “environmental” number 65. The decline in single authorship has not been uniform across fields. Comparing 1981 with 2012, the social sciences and economics showed declines of 72% to 38% and 69% to 27%, respectively, whereas pharmacology and immunology showed declines of 13% to 4% and 10% to 2% (Voosen 2013). Environmental and ecological sciences were intermediate, declining from 35% to 5%. Over all fields the average number of authors has risen between 1981 and 2012 from 2.5 to just above 5 (King 2013b). In some fields—notably physics and biomedicine—articles with >100 authors are observed (King 2013a). It may be that many single-author articles nowadays are reviews, which can be important generators of citations. Of the 20 most cited among EG's 160 cited papers, 8 were reviews, accounting for 72% of citations. A minor factor contributing to the decline in single authorship may be changing standards for entitlement, so that co-authorship may now include some (e.g., technicians) who in the distant past might have been credited in the Acknowledgments. Authorship is discussed in a broader context in Chapter 4 of an upcoming publication by the British ecologist Clymo (2014). On reading a draft of this article, he reminded us that the prevalence of multiauthored citations is accentuated by the tendency of authors to focus heavily on very recent articles (Gorham 1968), which are more and more likely to be multiauthored. The great increase in funding for ecological research has allowed the lone researcher of earlier times to expand his/her reach by obtaining sufficient support to bring together a team with varied talents and backgrounds. But does the present culture of team research militate against a lone researcher obtaining grant support, perhaps because team research is regarded as a more efficient use of resources? Or is the framing of research questions different—perhaps broader—because of the possibilities offered by a team approach, in which the leader chooses among ideas put forward by team members as well as her/himself? Does the need for very substantial team funding lead to collectivistic framing in terms of a practical mission, instead of individualistic curiosity about a fascinating problem? And does a focus on mission inhibit following up such curiosity, which is often a result of chance and serendipity that lead one's research in diverse directions (Gorham 2012)? Might research proposals by individuals be likely to stray farther from the beaten path than those by teams? In the same vein, are individuals or teams more likely to challenge the “dominant paradigm” in their field? As to the quality of multiauthored vs. single-authored articles, Bridgstock (1991) viewed the question as unresolved. More recent studies have found greater citation rates for papers in Oecologia with more than four authors (Leimu and Koricheva 2005a), but the benefits of collaboration appear relatively minor. The authors point out that citation rates measure only utility to other scientists, and not, for instance, originality and methodological quality. Furthermore, results of their earlier study of factors affecting citation rates (Leimu and Koricheva 2005b) “cast doubt on the validity of using citation counts as an objective and unbiased tool for academic evaluation in ecology.” Originality is a primary aim for researchers, but original discoveries vary greatly in their importance and utility to other scientists. They are, therefore, bound to vary greatly in their citation rates. Examination of EG's citation record in Google Scholar since 1975 (missing prior citations of papers published earlier) reveals that his early truly original studies—of lake acidification by acid rain generated from far-distant urban/industrial air pollution (Gorham 1955, 1958)—rank 13th and 14th on his list at 155 citations each. A review arising from those studies (Gorham 1961) was apparently of greater utility, attracting 263 citations and ranking fifth. Of far greater utility were later reviews on two very different topics (Bray and Gorham 1964, Gorham 1991) that gathered 2284 and 1224 citations, respectively, ranking first and second. Whereas the main acid rain studies were both original and useful, others were clearly original but of lesser utility. Among them was the discovery that in British cities, acid rain was predominantly due to hydrochloric acid (Gorham 1958), cited 46 times and ranking 62nd. Apparently of little interest was the discovery that bronchitis mortality in Britain was related to the acidity of precipitation (Gorham 1958), cited only 16 times and ranked 108th. Even less frequently cited were the relationships of lung cancer mortality to tar (Gorham 1959a) and pneumonia mortality to sulfate (Gorham 1959b) in precipitation, after partial correlation and regression that included bronchitis data as well. (At this level self-citation becomes important.) It appears, therefore, that utility trumped originality in generating citations. Nevertheless, with extremely important discoveries, such as that of the double helix, the two might well coincide. We thank students Eva Thomas and Amanda Wanous for tabulating our data. The data table on which this article is based is available from the first author.