I get Ecolog posts in digest form, so usually when I am tempted to comment it is already a day late, and so I don't. But this time I'll make an exception in defense of Henry Gleason (and also in defense of Clements) though probably not in time to forestall Dr. Hamilton's "trashing" of Gleason to his class.
It is very much worthwhile to consider not only the state of ecology back in 1916 and in the 1930s, but also to consider the backgrounds of both Clements and Gleason. I would argue that based on their experience both men would have no choice to come up with their respective interpretations of succession and that if you flipflopped their experience, both men would have developed the others' theories. Clements primarily worked mostly in temperate and montane environments in the USA (Neb, Min, Col, Calif, Ariz). Gleason began his research in Ill, following up on work by Cowles, and continued in Michigan and at the NY Botanical Garden. Most importantly for this post, Gleason spent time putting together a botanical survey for Puerto Rico and worked in the Asian tropics. He had already begun to question the association model of Clements by 1927, but his tropical trips resulted in a more complete criticism. From PR, Gleason published on plant ecology and noted that multiple transects or plots would seldom result in similar species compositions. (Gleason, H.A. and M.T. Cook, Plant ecology of Porto Rico. Scientific Survey of Porto Rico and the Virgin Islands, 1927. 7(1-2): p. 1-173.) Granted, a series of plots or transects may only be descriptive, but PR at the time that Gleason was here was already a highly fragmented habitat, severely affected by land use history. The Spanish had been here for >400 years. Agriculture was already well on its way to deforesting 96% of the island (a level reached within 25 yr of Gleason's visit) but some marginal lands and former crown lands were already developing secondary forests. We have aerial photos of the island from 1936 that show clear patchworks of forest fragments of various ages witihin an agricultural matrix. Gleason was a bright fellow, was well trained by Cowles and in Clementsian succession before arriving in PR, and would have recognized the value of sampling in what we would now recognize as a chronosequence. I would argue that chronosequence sampling is in fact experimental, but that is probably another topic for Ecolog. Finally, I think if you read more of Gleason's work, you'll find that he is cut more from a modeler's mold. If you blindfold yourself and run through a mature forest in Michigan, it's pretty sure that you'll run into a majority of beeches and maples, or oaks and hickories, etc. If you do the same in mature forests in Puerto Rico, you won't run into a majority of anything, except trees, and our flora is "depauperate" compared to continental tropical tree floras. In fact, we do refer to "tabonuco" or "colorado" forests, but these aren't dominant species in the temperate sense, they are better understood as species that one might frequently find in mature lowland or lower montane (respectively) wet forests. They are indicators of a forest type, rather than a successional association. If you go into a mature (>80 year old) forest in the tropics and predict the species of tree next to the one you are standing under, you'll need a long list to be correct--much longer than in most temperate/alpine regions. On the other hand, Clements did not work in the tropics, as far as I know. As such, he would have a shorter list of species to work with and a set of "climax" forests that were very predictable in dominant species--even in species from earlier successional stages, as Marks so well illustrated in Pennsylvania. A most reasonable explanation of these patterns would be plant associations. When forced to explain "exceptions" due to waterlogged soil, sandy patches, etc. Clements defined smaller associations. Perhaps the tropical forests represent ever more smaller associations due to edaphic and other factors, but it is very hard not to believe that our forests are more individualistic and our species are more interchangeable. Gleason began his career in Illinois using a Clementsian approach and found cracks in it, even from his descriptive work. Perhaps Clements' calls for more experimentation were a smokescreen to diminish the attention paid to his critics. Perhaps it was genuine. Regardless, in hindsight, it was ahead of the curve to call for manipulative support for successional concepts. On the other hand, 10 years after the publication of Clements' seminar work, Gleason would have had ample experience with very controlled observations in much different systems to both support his individualistic hypothesis and criticize superorganisms. So there is my 2c. and as it's after midnight, this post will probably miss yet another day in the life of ecolog digests, but thanks for letting me provide a postscript. Skip J. Van Bloem, PhD Dept. of Agronomy and Soils University of Puerto Rico at Mayaguez > Date: Sat, 17 Feb 2007 13:51:50 -0600 > From: Robert Hamilton <[EMAIL PROTECTED]> > Subject: Re: 1916 Clements reference in electronic version? > > The problem I see with Gleason is that his argument is purely > rhetorical. He does use examples, but no experimental analyses of any > sort. A key point, again JMHO, is that Gleason talks of SPECIES while > Clements talks more of life forms. Clements does not predict rigid > SPECIES compositions, however one of the problems with a lot of Clements > work is the attempt to define smaller and smaller scale associations of > life forms. > > I could rant on almost indefinitely! Reading Gleason reveals a person > who is consumed with description only; there is no attempt at any sort > of experimental analysis. Clements continually insists on experimental > analysis. The 1916 paper, for example, includes a lot of data. One needs > to remember where Ecology was in 1916. We had no Evolutionary Synthesis, > but rather Darwinists vs Mendelists. We had no concept of any sort of > Functional Ecology outside Clements and his group. Clements challenged > people who just wanted to describe; Clements wanted experimental > analysis.