The number of species on islands

by Greg Mayer

[The following is a trivial, and speculative, discussion about a small part of an important paper in the history of ecology and evolutionary biology.]

There are fewer species of any given group of plants or animals on an island than on an equivalent area of the mainland; and the larger the island, the more species there will be. These two general rules of natural history have been known since at least the 19th century, and are known under the rubrics that island biota are depauperate, and the species-area relationship, respectively. There have been many, not always mutually exclusive, explanations for these phenomena, and in the early 1960s they were on the minds of at least several biologists.

The most important resulting paper was Robert MacArthur and Ed Wilson‘s classic “An equilibrium theory of insular zoogeography.” This paper, together with the expansion of the ideas contained within it into a book-length monograph, The Theory of Island Biogeography (1967), were extremely influential in setting out the questions to be asked, and how to go about answering them, across many areas of ecology and evolutionary biology, not just the phenomena of island life.

MacArthur and Wilson (1963).

In the paper, MacArthur and Wilson proposed (among other things) that the number of species on an island resulted from a dynamic equilibrium between ongoing immigration and extinction of species living on the island, and that the relation of these demographic processes to various physical properties of islands led to the species-area relationship. They illustrated this relationship in two figures showing the relationship between the area of an island or island group and the number of species of land and freshwater birds occurring on that island or island group. One figure was for islands in and near the Sunda group, the other for the Moluccas, Melanesia, and Polynesia. Both figures are interesting, but the second one concerns us today.

Figure 2 of MacArthur and Wilson (1963).

For reasons relating to some research projects I’m pursuing during the pandemic, I was led to look closely at this figure, including the sources of the data as indicated in the legend of the figure: three papers by Ernst Mayr, and a book by James C. Greenway. I have a copy of Greenway’s book (both editions, actually), and Mayr’s 1943 paper is readily available online to anyone from SORA (a wonderful free repository containing vast swaths of the ornithological literature).

Neither of the two earlier papers are readily available, but the University of Wisconsin, Madison, Library scanned a copy of the earliest, and sent me a pdf (kudos to the staff of the Library there for working very hard during the pandemic to keep the scholarly literature available); and I happened to have a reprint of the other, a paper from the modestly obscure Proceedings of the Sixth Pacific Science Congress. Mayr published two papers in the proceedings, and I have reprints of both; the second is the one cited by MacArthur and Wilson.

Mayr (1940a).


Mayr (1940b).

The provenance of my copies is of interest. By looking at the stamps and annotations, you can see that the copies were originally in the library of the Bird Department of the Museum of Comparative Zoology, which stamped them, and added the author and date notation at upper left. (Departments of major museums usually maintain libraries physically separate from the main library of the institution.) Reprints were the principal way in which scientific literature circulated in the 1940s, and it is most likely that the Bird Department obtained the copies shortly after publication, not when Mayr moved to the MCZ in 1953. As duplicates (and Mayr’s arrival may be why the Department had extra copies), they then passed into the possession of my friend and colleague Bob O’Hara; we were graduate students together in the 1980s, and he worked closely with Mayr. At some point, also in the 1980s, Bob gave them to me, and I penciled my name and a correction to the date of publication on them. (See Note on the date of publication below.)

So now we come to the matter at hand. I was trying to track down the actual numbers that went into MacArthur and Wilson’s Figure 2, and some of the data came from Mayr (1940b), the second of the proceedings papers. It’s a 20 page paper, with no table of species numbers by islands, so I was reading it carefully to find what numerical data I could. It’s all on page 202– species numbers for 11 islands or island groups that were included in Figure 2. And a little later, while copying the numbers on to a data sheet, I noticed a pencil mark next to the paragraph with the data. (It is the only mark in the paper, aside from those on the first page.)

A pencil mark highlighting species number data in Mayr (1940b).

The mark was not mine; and it would not be Bob O’Hara’s, who would not have been interested in the particular questions addressed by these data. So who would have been using the MCZ Bird Department Library’s copy prior to the 1980s to highlight data included in MacArthur and Wilson (1963)? It occurred to me, could it have been Ed Wilson himself?

Wilson was at the MCZ with Mayr, and would have had access to this copy. In their paper, he and MacArthur thank Mayr (and a few others) for “material aid and advice during the course of the study.” There is thus no question that they consulted Mayr, and used data from a number of Mayr’s papers and books (in Figure 1, as well as Figure 2). If they looked up the data themselves (as opposed to Mayr directly telling them the species numbers), then this particular copy, now in my possession, is a likely copy for them to have used, and Wilson is the likely person to have made the mark highlighting the data that was used. This then, is my speculative (and trivial) suggestion: that Wilson used this copy in the preparation of his and Robert Mac Arthur’s classic and influential paper.

MacArthur and Wilson’s (1963) Figure 2 plots data for 26 islands or island groups. In the cited references, I can find data for only 21 of them; I do not know where they got the data for Tonga, Kei, Tanimbar, Buru, and Ceram. Jurgen Haffer (2007:163), Mayr’s biographer (and himself an accomplished contributor to ornithological science) records the following interesting tidbit:

Mayr (pers. comm.) had copious data on island sizes, distances from mainlands or other islands, number of species, etc. When he tried to determine relations among all these figures he got into mathematical problems and turned this material over to a graduate student with mathematical abilities. However this student got sidetracked into other problems and this material was never exploited.

This account was recalled to Haffer by Mayr decades later, and what data was compiled, when it was compiled, and who the student was, isn’t known.

Haffer (2007: 170) also states that in the 1933 and 1940b papers Mayr “clearly discussed what became later known as the equilibrium theory of insular biogeography”, a claim that has been echoed by other authors. This is not quite correct. Mayr discussed a number of relevant factors contributing to the characteristics of island faunas in 1940b, and the paper is well worth reading and studying today. But he did not formulate in any clear way, even verbally, MacArthur and Wilson’s later theory.  (My very limited German does not allow me to properly assess the 1933 paper, but Mayr’s two 1940 papers in English seem to parallel closely the earlier paper in German.)

For example, Mayr (1940b:215) does note that on a smaller island a species will have a smaller “effective breeding population” and thus be vulnerable to extinction; this is a striking formulation, obviously influenced by theoretical population genetics (likely gotten from Theodosius Dobzhansky). But in discussing extinction on New Caledonia he is clearly discussing evolutionary events stretching over much of the Tertiary (i.e. tens of millions of years), and not the turnover in ecological time of insular populations contemplated by MacArthur and Wilson’s equilibrium theory. The latter has been demonstrated to occur over annual and decadal scales on, for example, smaller islands in the British Isles.

While not excluding the possibility of a MacArthur and Wilson style ecological equilibrium (because he does not consider the situation), Mayr is clearly discussing the origin and persistence (or not) of endemic forms (species, genera, even families) over evolutionary time. Mayr himself (quoted in Haffer, 2007: 163) offers a much more nearly accurate appreciation of his views, stating that his own ideas embraced the “basic thesis” of MacArthur and Wilson; he did not “clearly discuss” the equilibrium theory.

Note on the date of publication. The Sixth Pacific Science Congress was held in 1939, but the Proceedings were published later, from 1940-1943.  Greenway (1958), who knew Mayr well, cited them as 1941, and this may have been the source of my handwritten correction, but I cannot now recall. Haffer’s (2007) definitive list, based on Mayr’s own lists, cites them as 1940.

Greenway, J.C. 1958. Extinct and Vanishing Birds of the World. American Committee for International Wild Life Protection, New York.

Haffer, J. 2007. Ornithology, Evolution, and Philosophy: The Life and Science of Ernst Mayr 1904–2005. Springer, Berlin.

MacArthur, R.H. and E.O. Wilson. 1963. An equilibrium theory of insular zoogeography. Evolution 17:373-387.  pdf

MacArthur, R.H. and E.O. Wilson. 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, N.J.

Mayr, E. 1933. Die Vogelwelt Polynesiens. Mitteilungen aus dem Zoologischen Museum in Berlin 19:306-323.

Mayr, E. 1940a. Borders and subdivisions of the Polynesian Region as based on our knowledge of the distribution of birds. Proceedings of the Sixth Pacific Science Congress 4:191-195.

Mayr, E. 1940b. The origin and history of the bird fauna of Polynesia. Proceedings of the Sixth Pacific Science Congress 4:197-216. (This paper was reprinted in Mayr (1976) but Mayr sometimes updated papers in that collection, and for my purposes I needed to see the original.)

Mayr, E. 1976. Evolution and the Diversity of Life. Harvard University Press, Cambridge, Mass.

Mayr, E. 1943. The zoogeographic position of the Hawaiian Islands. Condor 45:45-48. pdf


  1. Posted July 27, 2020 at 1:09 pm | Permalink

    Wonderful post Greg, and just to be sure, in case of interest, you know of Jim Brown’s paper (in Ecology I think) on an obscure butterfly paper that did indeed presage Wilson and MacArthur’s theory…

    • Posted July 27, 2020 at 2:53 pm | Permalink

      Thanks, Harry! And thanks for the mention of the “true” anticipation of M&W by Eugene Munroe (who did his work at Cornell!). Interestingly, Munroe published his equilibrium theory in the Proceedings of the Seventh Pacific Science Congress— the Congresses seem to have been a graveyard for good ideas! I was thinking a little a bit about Munroe (who published in the 1950s), but even more so about Dick Levins and Hal Heatwole’s paper in the Caribbean Journal of Science, published at nearly the same time as M&W. (See some discussion about this earlier here at WEIT.)


  2. Jeff Dennis Campbell
    Posted July 27, 2020 at 1:15 pm | Permalink

    Fascinating glimpse into the actual culture and workings of science in the pre-digital era!

  3. BobTerrace
    Posted July 27, 2020 at 1:38 pm | Permalink

    That’s some amazing detective work, Mr M!

  4. DrBrydon
    Posted July 27, 2020 at 1:44 pm | Permalink

    I thought I was going to be the only weird one, wondering about the provenance of the reprints.

  5. Ken Kukec
    Posted July 27, 2020 at 2:21 pm | Permalink


  6. rickflick
    Posted July 27, 2020 at 2:25 pm | Permalink

    Interesting to dig out the history. The equilibrium theory sounds very reasonable to my novice ears. The thing about an equal area on the mainland is, it has no boundaries. Thus, it is part of a much wider area of interactions between populations. It is subject to vastly greater dynamic influences. More movement. There is more genetic mixing and diversity, for one thing. A smaller island population may easily go extinct, while a comparable but much larger population on the mainland would be less likely to do so. If I understand it correctly.

    • Torbjörn Larsson
      Posted August 2, 2020 at 12:57 pm | Permalink

      This seems to be a vast research area – see comments in the thread – but regarding different time scales of ecology and evolution as well as insularity I was reminded of a recent paper on Fijian archipelago trap-jaw ants [ ]. They seem to have evolved from a common ancestor that early split in lowland and upland species, where the latter behaves more insular. It looks like their mountain habitats are islands within the islands of the archipelago. They spread slower (and are now generally threatened by deforestation).

  7. Posted July 27, 2020 at 3:38 pm | Permalink

    Thanks for the interesting read. These sort of island vs continent questions fascinate me. I always imagined that these kinds of effects were simply the result of scaling along many dimensions, starting with the area of the land mass of course. If we consider the limit case of a very small island, we can easily imagine that it is limited to few species for which the individuals are small as the opposite would be so crazy. We’re never going to find a naturally occuring Elephant Island consisting of a single species of large mammal. Smaller land size implies few ecological niches implies fewer species. Ecological niches harbor multiple species whose population sizes are inter-dependent. If we reduce a critical resource, everything else scales too. These relationships aren’t necessarily linear, of course.

    If we ever find bigger planets with life, I bet we also find that they have bigger continents and each has more species, all other things being equal.

  8. Posted July 27, 2020 at 3:56 pm | Permalink

    A very interesting tale!

  9. Posted July 27, 2020 at 4:21 pm | Permalink

    How do I get wildlife / readers’ photos etch to you. Pity a somewhat young but sometimes interweb-challenged follower.

  10. ThyroidPlanet
    Posted July 27, 2020 at 4:51 pm | Permalink

    There’s something about old scientific papers that is fascinating to me. I imagine a wizard blowing dust off ancient texts, like that scene in Lord of the Rings..,.

    • Posted July 27, 2020 at 4:57 pm | Permalink

      Are you familiar with “A Canticle for Leibowitz”? It’s a sci fi novel where the maintenance of old scientific documents, by monks that haven’t a clue as to their meaning, plays a central role. Highly recommended!

      • ThyroidPlanet
        Posted July 27, 2020 at 5:13 pm | Permalink

        Interesting premise! I’ll check it out!

  11. jedijan
    Posted July 27, 2020 at 7:52 pm | Permalink

    An interesting story of directive work. Wonder if there is paper comparing the species number data in Mayr (1940b) to the current day; 80 years later. Assume there must have been a list of those numbers which identifies the species also.

  12. openidname
    Posted July 27, 2020 at 11:55 pm | Permalink

    Not to burst your bubble, but as I understand it, the copy you had could not possibly be the copy Wilson used. Because you had a .pdf, not a physical copy.

    Michael Shermer has some trenchant comments about how even scientists give way to this kind of magical thinking about objects — that they somehow take on some essence of their possessors. He gives the example of Richard Dawkins — not a man given to mysticism — getting excited about touching Einstein’s desk.

    • Posted July 28, 2020 at 5:58 am | Permalink

      Those are photographs of the physical objects, with mine and others’ handwriting and stamps on them. Why would you think they are pdfs?


  13. Posted July 28, 2020 at 2:09 am | Permalink

    Very interesting. Sadly reprints are being thrown away in most libraries I know – when I worked in the Institute of Ophthalmology Library in Moorfields we had thousands, many from the authors. Likewise in the Action on Hesring Loss (RNID) Library, I had to throw away thousands of old reprints. Most were easily available now as e-journals however. There were quite a few that were signed or had “compliments of author” & those I tried to rescue but I rather fear they will not have survived the hasty move which has just taken place with no one who has any knowledge of the collection involved. Oh well…

    • Posted July 28, 2020 at 2:40 am | Permalink

      One thing-
      There is surely a difference with oceanic islands & continental shed islands, depending on geology. For example Hawaii island chain has/ had a diverse array of endemic species presumably some picked up in the deeper past when they were nearer Asia. Anyway, oceanic islands must have a slightly different set of rules depending on their geological past in comparison? This is really fascinating!

      You can see the island paucity in the British Isles & Ireland compared with the continental European biodiversity. For example no tawny owls in Ireland, or moles, & lots of bird species found only a few miles across the channel that are not found in Britain. Clearly humans have made huge differences – we could eliminate rival species such as wolves that enabled sheep raising to be much easier. Incidentally I recall a theory that Irish wolves were quite big so they could hunt Megaloceros…

      • Posted July 29, 2020 at 7:21 am | Permalink

        Yes, there are big differences between continental (recently separated from the mainland) and oceanic (never connected to the mainland) islands. The British Isles are classically continental. As regards wolves, Irish elk, and other mammals, D.W. Yalden’s The History of British Mammals (1999) is a must-read delight for fans of Bretannic mammals. Yalden details the Late Glacial and Post Glacial extinctions and colonizations of the islands by these animals. (There’s much later literature, as well, but it’s a good read and a place to start.)


        • dom
          Posted July 30, 2020 at 2:44 am | Permalink

          Thanks! 🤓

%d bloggers like this: