The idea of a single climatic climax, which is defined in relation to regional climate, originated with Frederic Clements in the early 1900s. The first analysis of succession as leading to something like a climax was written by Henry Cowles in 1899, but it was Clements who used the term "climax" to describe the idealized endpoint of succession.
Clements's theory sought to define a single climax-type for each area. Arthur Tansley developed this idea with the "polyclimax" -- multiple steady-state end-points, determined by edaphic factors, in a given climatic zone. Clements had called these end-points other tems, not climaxes, and had thought they were not stable, because by definition climax vegetation is best-adapted to the climate of a given area. Henry Gleason's early challenges to Clements's organism simile, and other of his strategies for describing vegetation, were largely disregarded for several decades until substantially vindicated by research in the 1950s and 1960s (below). Meanwhile, climax theory was deeply incorporated in both theoretical ecology and in vegetation management. Clements's terms such as pre-climax, post-climax, plagioclimax and disclimax continued to be used to describe the many communities which persist in states that diverge from the climax ideal for a particular area.
Though the views are sometimes attributed to him, Clements never argued that climax communities must always occur, or that the dominant cause of vegetation is climate, or that the different species in an ecological community are tightly integrated physiologically, or that plant communities have sharp boundaries in time or space. Rather, he employed the idea of a climax community--of the form of vegetation best adapted to some idealized set of environmental conditions--as a conceptual starting point for describing the vegetation in a given area. There are good reasons to believe that the species best adapted to some conditions might appear there, when those conditions occur. But much of Clements's work was devoted to characterizing what happens when those ideal conditions do not occur. In those circumstances, vegetation other than the ideal climax will often occur instead. But those different kinds of vegetation can still be described as deviations from the climax ideal. Therefore Clements developed a very large vocabulary of theoretical terms describing the various possible causes of vegetation, and various non-climax states vegetation adopts as a consequence. His method of dealing with ecological complexity was to define an ideal form of vegetation--the climax community--and describe other forms of vegetation as deviations from that ideal.
Although Clements recognized that vegetation follows gradients rather than being tightly bound, his rhetorical comparisons of ecological communities to organisms fostered the impression that communities, including the climax, have distinct edges in space and time. Yet Robert Whittaker's research demonstrated plant species distribute themselves along nutrient and other environmental gradients. Many ecologists saw this as a major reason to stop using the climax concept.
More recent palynological studies showed that modern species assemblages are ephemeral; vegetation in eastern North America since the last glacial maximum has consisted of several different species assemblages, many of which have no analogues in modern "climax" communities. That would mean, at least, that the climax types for those areas could not be stable to the degree Clements believed they were.
Ultimately, even if succession tends towards a steady state, the time required to achieve this state is unrealistically long; in most cases, external disturbances and environmental change occur so frequently that the realization of a climax community is unlikely, and therefore it has come to be regarded as a less useful concept. Long-term vegetation dynamics are now more often characterized as resulting from the action of stochastic factors.'''