| The most basic concepts, terminology and ecological philosophy used in this presentation and interpretation of vegetation was Clementsian (including modifications of the Clementsian model by Clements’ students and “disciples”). This was not primarily because the author is a dyed-in-the wool-Clementsian (though he unabashedly is), but because this was the ecological language and perspective taken by the founders of the Anglo-American Tradition of Ecology. This was detailed in the review of the historic development of common units of vegetation. It was shown that range and forest cover types were largely adaptations of the association traceable to Alexander von Humboldt, yes, but specifically as used by Frederic Edward Clements and his close friend Arthur G. Tansley. The most likely origin, at least the source from the written record, of range and forest cover types is Plant Indicators: The Relation of Plant Communities to Process and Practice in which grazing types were introduced and various kinds of forest types were reviewed by Clements (1920, ps. 270-345). |
| Reliance upon Clementsian doctrine (some would prefer “dogma”) was not only universal, or nearly so, among rangemen like Arthur W. Sampson and E.J. Dyksterhuis and grassland ecologists like John E. Weaver, but also among prominent forest ecologists like E. Lucy Braun. Braun’s forest regions and associations of the North American Eastern Deciduous Forest Formation were the “pure milk” of Clementsianism, and her Clementsian vegetation units still serve as the basis for ecological consideration of the eastern deciduous forests. See for example the most recent edition of North American Terrestrial Vegetation (Delcourt and Delcourt in Barbour and Billings, 2000, ps. 365-378). |
| This is not to deny the obvious fact that other prominent ecologists, including those of the Zurich-Montpellier and Northern Traditions, had some influence on the Anglo-American School and, for that matter, on the Clements-Tansley-Weaver-Braun lineage. Nor can the impact of Henry Allen Gleason be denied. For instance, Gleason’s continuum or individualistic theory of plant communities was adapted or fashioned by Robert H. Whittaker into the climax pattern theory to bring closer together the monoclimax and polyclimax theories. This is the proverbial exception that proves the rule. The rule— the historic fact —in this case being that Clements had more influence, and continuing influence, on Plant Ecology in America than any other ecologist. Whittaker’s use of Gleason’s individualistic theory modified it as much as Clement’s monoclimax or Tansley’s polyclimax theory. Most conspicuously, the Whittaker synthesis was still a climax theory and it was Clements who coined the term climax. |
| Finally, this application of a principle of history bears making at the onset as well as in the body of the report: the important point is not which, if any, theory, tradition, ecologist, etc. was “right” or “wrong”, but rather which had the most influence on views of vegetation science (units of vegetation in this specific instance) as used today. Beyond any doubt the answer with regard to range and forest cover types, forest and range succession, and natural vegetation mapping is the Clementsian paradigm. In fact, it (and the precedential views of August Grisebach, J.E.B. Warming, Oscar Drude, and Henry Chandler Cowles) is the only model consistent with designation and delineation of types of vegetation at large spatial and long temporal scale. |
| Vegetation can be mapped according to all three traditions. Plant communities can be mapped, recognized, named, etc. according to the continuum or individualistic view. But commonly recurring units of natural (define it however pleases you) vegetation that occur consistently, routinely, enough to be classified in a hierarchial arrangement of units that includes a broad or general level of the same repeating type is most consistent with the Clementsian paradigm. The recognition of some unit at the level of cover type (a general community defined by dominant species and physiogonomy) and based on development of vegetation (ie. changes in plant communities over time) is perhaps inconsistent with any model except that of the Anglo-American Tradition. |
| A comparison with the Gleasonian view, which is “completely antithetical” to Clementsian climax theory (Barbour et al., 1999, p. 23), illustrates the above point. Gleason argued, convincingly to many, that “plant communities are not real, natural units; they are merely human abstractions” (Barbour et al., 1999, p. 23). Carried to it’s logical end, the continuum hypothesis argues against the very concept of forest and rangeland cover types. If Gleason was correct then cover types are counterfeit. They are a product of human imagination and nothing more. They are based on faith and creative fantasy and not on vegetation science. If plant communities are individualistic assemblages of plants occurring along a continuum then there is no consistently repeating plant community (except by chance) and therefore there cannot be cover (= dominance) types. Right or wrong, true or false, correct or incorrect: individualistic theory and a conceptually valid basis for cover types of vegetation are “completely antithetical” (Rosiere, here and now). |
| Perhaps rangeland and forest cover types are but “figments of the imagination”. Maybe they are not real, but instead counterfeit, units of vegetation. If this is the case why did we invent them? A valid— at least necessary —assumption would seem to be that there are range cover types or else we would not bother to name, number, and describe them. |
| Allowing for some degree of abstraction (even some imagination), of taxonomic versus ecological definitions, and of “lumpers” versus “splitters”, a unit called species is either real or it is not. If species “are not real, natural units” but “merely human abstractions” taxonomists could not logically designate such a unit nor identify or recognize individuals as belonging to that unit. If each “species” is just a happenstance collection of genes and genotypes from all over then there is no such thing as species. |
| Ditto on vegetation. If each plant community is just a happenstance collection of plants there is no “taxonomic” unit, there is no plant community type but only plant communities (just chance combinations of kinds of plants). Alternatively, if there are kinds of vegetation and if these kinds of vegetation are repeatable or recurring units with different kinds at different spatial scales these kinds of plant communities could be identified and described as recognizable units. In other words there would then be a hierarchial arrangement of kinds of vegetation similar to taxonomic systems for organisms. The alternative just described allows classification and mapping of existing or current plant communities (ie. vegetation as it lives at present). It generally describes the basic outline of the Zurich-Montpellier Phytosociological School most closely associated with Josias Braun-Blanquet. |
| If the alternative view of vegetation, that plant communities are real and there are identifiable units of vegetation at various sizes, is more true than not then classification of recurrent units of vegetation is based on valid assumptions. If such a system of vegetation classification is to include the time dimension (temporal scale) such that the vegetation scientist or manager could predict what kind of future vegetation can or will come to occupy a given space (ie. the potential natural vegetation) another ecological philosophy or vegetation model is needed. A system and a perspective of vegetation that goes beyond existing plant communities is required. There is one, only one. It is the English Tradition or the Anglo-American School. It is the Clementsian paradigm. |
| Yes, both Society of American Foresters and the Society for Range Management stated specifically that their forest and rangeland cover types were of “existing tree cover´, “forests as they are today”, (Eyre, 1980, p.1) and the “existing” or “present” vegetation and not the potential natural plant communities but those that are “the result of human influence” (Shiflet, 1994, p.xi-xii). These criteria were stated several times herein “for the record”. Close study of the forest and rangeland cover types, however, showed them to be, with some notable exceptions, fairly close to if not identical with what are generally accepted as climax or potential natural vegetation units of such authors as Clements (1920), Weaver and Clements (1938), Braun (1950), Kuchler (1966), Franklin and Dyrness (1973), Dick-Peddie (1993), Kuchler (1977), and Barbour and Billings (2000). In fact, detailed comparison of the vegetation units of these various authors among themselves and with them and the SAF and SRM cover types revealed more discrepancies among the authors than between the consensus of those authors and the SAF and SRM units. Other than for a few obvious exceptions (eg. seral cover types such as alder and birch communities, brush patches on go-back land, the California annual grassland, and Douglas-fir maintained silviculturally as single-species stands) most SAF and SRM cover type descripitions match descriptions of climax vegetation. A relatively high proportion of these exceptions were regarded by the cited authors as man-made climaxes that are the potential naturalized vegetation (eg. California annual type). It was noted that the Society for Range Management did not recognize cheatgrass-dominated communities or vegetation on go-back land as any rangeland cover type. |
| SRM and SAF rangeland and forest cover types are obviously late seral to climax vegetation. They are relatively stable plant communities with community change most likely to be downward or back to lower seral stages. They are benchmark units of vegetation within which change in species composition, dominance, structure, etc. can be related to all manner of disturbances and human action (management). |
| All perspectives (certainly the three main traditions) of vegetation realize that vegetation is dynamic (ever changing), but to map or classify existing vegetation is to adopt a view of vegetation as if it was static. At very least it is to describe vegetation at or for an instant in time. And it is to do so with no reference to this existing state of vegetation, this present plant community, in relation to whether any community change is more productive, stable, or advanced in maturity. The Anglo-American model of vegetation holds that there is a final plant community that is relatively stable (in dynamic equilibrium) for each area of land (or water). This is the potential natural vegetation and it is relatiely stable, sort of static. The student of vegetation cannot know if the existing plant community is this potential or not without knowing how the vegetation on that area of land will change (ie. without knowing the pattern and process by which the vegetation will develop). |
| In Plant Sociology (Braun-Blanquet, 1932) detailed the factors influencing the plant community, but when he came to discuss the development of plant communities his chapter began with a departure from the “Clements-Weaver system”. While Braun-Blanquet (1932, ps. 305-306) had serious and valid criticisms of the Clementsian model he had no alternative for treatment of potential vegetation (Clements’ climax). Things have advanced some in the last two-thirds of a century but, as shown in the review below, the major vegetation classification system being proposed today by The Nature Conservancy and the “big name” organizations is the Zurich-Montpellier Phytosociological Method (or a slightly modified form thereof). |
| To some people the potential natural plant community or climax (in some instances the former may not have as its potential the latter) is not important. Apparently it is not critical to The Nature Conservancy and Ecological Society of America as shown in the review below. To rangemen and foresters, however, knowing or being able to estimate (if not predict) the potential plant community is essential because it serves as a benchmark or point of departure by which their management and the vegetation managed can be appraised. Said another way, the most useful— if not the only useful and relevant— state of “present” or “existing” vegetation that can be used as a rangeland or forest type is the one with which changes in vegetation can be related to human manipulation (management). By professional definition these are managed plant communities The only state or status of vegetation by which man can judge his managerial inputs is the end point, the terminus, of vegetation development (ie. climax or potential natural vegetation). Even if a given cover type was deemed as the desired plant community based on non-arbitrary criteria such as maximum (or economically optimum) production of board foot of lumber, animal turnoff, or maximum biological diversity this point of maximization or optimizaton cannot be known until the final endpoint (and its wood production, herbage yield, or species diversity) is known. It cannot be known until all potential plant communities can be compared. This requires completion of plant succession. |
| This terminus of plant succession (climax) may or may not meet the management goal(s) but this cannot be known until the array of vegetational development (the span of plant communities along a sere) is known, and that certainly includes the successional endpoint (Clements’ mature vegetational organism). Adult size (height, mass, etc.) and mature development of any organism cannot be known until adult size and maturity is reached and capable of measurement. |
| This does not mean (as shown in the review below, it never did mean) that “mature” (climax or potential natural) vegetation is the goal of management. It merely means that climax (at least later seral) vegetation is the “high water mark” from which to measure movement away from the final plant community that is capable of development on a given habitat. If maximum biodiversity, herbage yield, soil fertility, water yield, etc. occurs at some stage below climax or natural potential (eg. subclimax) then the rangeman, forester, or wildlifer knows that (this may well mean that climax is known not to be the management objective), but it is known. “Present” or “existing” vegetation does not provide that knowledge unless the present or existing plant community is climax or can be compared with climax. |
| If a subclimax seral stage of vegetation is more valuable for, say, lumber or deer browse production it is rational to select that as the forest or rangeland cover type to describe and distinguish as the desired forest or rangeland plant community. If a single-species stand of trees that is subclimax (or comprised of a single climax associate) produces more high-grade lumber on a sustained yield basis than a climax forest community made up of several tree species the single species can legitimately be taken as the forest cover type for which to manage. It is impossible to know this unless climax is known (or there is a reasonable “guesstimate” of it). |
| The basic Clementsian paradigm (which is most certainly not the exclusive, the original, or the final work of F.E. Clements) with its many modifications is the only model of vegetation that satisfies the above conditions. |
| While Clements system “was overloaded with hypothetical assumptions” (Braun-Blanquet, 1932, p. 361) there has been no system as comprehensive— and as workable —with which to replace it. Approaches such as “rangeland health” (Committee on Rangeland Classification, 1994) and critiques of rangeland condition (Lauenroth and Laycock, 1989), while well-written, indicative of much thought, and useful for other purposes, offered no actual alternative to the Anglo-American Tradition or Clementsian model of “dynamic plant ecology” or “vegetation development”. At least they did not in this author’s opinion after he studied closely both of these publications. Maybe it was not their objectives to develop an alternative or replacement system but rather to find methods complementary to the existing approaches and perspectives. |
| Likewise, such innovative and useful models as those of Westoby et al. (1989) and Friedel (1991) are not of the comprehensive nature as to be an alternative to the traditional Anglo-American paradigm. They are complementary to it, however, which increases their usefulness. The Clementsian model was built on the concept of plant succession ending at climax that was, as later described, at “dynamic equilibrium”. The Westoby et al. (1989) model was for natural vegetation not at equilibrium. Clements also had that condition covered with his successional unit of disturbance climax or disclimax, “a community with the general character of the original climax “ which “… may be indefinitely maintained” (Weaver and Clements, 1938, ps. 86, 88). |
| The concept of “global stabiliy” of plant communities was adapted from Physics and applied to Ecology so as to appear in undergraduate Ecology texts by the early 1970s, complete with the novel pictorial device of a ball stuck in a trough (eg. Krebs, 1972, ps. 543-545). Global stability was the term used for the capacity of a community, ecosystem, etc. to recover to its original state when subjected to a major disturbance (= perturbation). Local stability was the term for return following minor disturbance. Within 15-20 years this concept had been applied to range plant communites as, for example, by Westoby et al. (1989). |
| For a summary explanation of community stability and the distinct but related terms of resilience and resistance readers are refered to Begon et al. (1990, ps. 793-798). The study of patch-dynamics is a related subject that is commonly used in Landscape Ecology. All of these are much newer areas of study than plant succession, climax, vegetation classification and mapping, and vegetation analysis. They are largely outgrowths from these older areas of vegetation research and management. |
| Application of the stability model to plant communities at disequilibrium is obviously compatable with the disclimax communities of the Clementsian model that appeared in Plant Ecology (Weaver and Clements, 1938). In the first edition of Plant Ecology Weaver and Clements (1929) labeled as subclimax several kinds vegetation currently known as range communities at disequilibrium or range not at equilibrium (Behnke et al., 1993). Thus Weaver and Clements (1929, p. 468) described the California annual grassland as being “of such permanence as to simulate a subclimax”. On the same page they remarked that the cheatgrass ranges of the Intermountain Region resembled the California annual grassland subclimax. They wrote that the shrubs of the Rio Grande plains had “so increased by overgrazing as to simulate a scrub climax” (Weaver and Clements, 1929, p. 466). By the second edition of Plant Ecology subclimax had been changed to disclimax while the other words remained the same (Weaver and Clements, 1938, ps. 86-88, 525-527). |
| The disequilibrium concept was the basis for the vegetation state and transition model(s). (See the pioneer work of Fridel [1988,1990, 1991], Westoby et al. [1989] and Fridel et al., [1993]). Disequilibrium and the state and transition model are asides to the concept and appliction of range cover (= dominance) types. They were introduced here to illustrate the germane point. |
| The pertinent point is that this disequilibrium concept is not, contrary to assetions by some who work with it, an “alternative theory” (Behnke et al., 1993, ps. 1-3) to the traditional range succession-retrogression model. It is a special case of it. The sub-climax maintained by grazing which Behnke et al. (1993, p. 3) wrote of was the verbatim idea of Weaver and Clements (1929, p. 76-77). Even the same term, subclimax, was used. This was textbook knowledge decades ago. It was as if workers like Behnke et al. (1993) never read Plant Ecology, but instead independently discovered the same ecological principle (subclimax is often more valuable for uses like grazing than climax) 60 years after it was textbook knowledge. Then they wrote as if they had developed an “alternative theory”. |
| The key or pivotal point about the disequilibrium model(s) discussed by range workers like Behnke et al. (1993) is that they are not just plant succession models or ecological models period, but rather human cultural models superimposed on the standard Anglo-American, the Clementsian, model of vegetation development. The principal question raised by studies like those reported in Range Ecology at Eisequilibrium (Behnke et al. 1993) is not whether the basic, seminal Clementsian paradigm of plant succession or the ball-and-stick, state and transition modification of the Clementsian paradigm is the more appropriate model for range vegetation. Instead the central issue is a dichotomy of management alternatives. One alternative is management of communally grazed range ecosystems (grazing commons) in developing nations (those in Africa in Behnke et al., 1993) at or for dynamic equilibrium (ie. potential natural or climax vegetation or, at least, higher seral stages of range vegetation) for the primary purposes of food, fiber, draft, etc. production from range livestock. The other alternative is management of grazing commons ecosystems and their livestock for traditional tribal values. In this second alternative, tribal commons range would be grazed at stocking rates that exceed grazing capacity (ie. ranges at disequilibrium) for maximization of livestock numbers in order to perpetuate native values and goods and services consistent with the nomadic way of life. This is management for high numbers of low-producing animals for social-cultural traditions like bride price-bridewealth (lobolo) or more total milk from more head of lower-producing animals (see especially pages 93, 99, 113, 118, 119, 135, 153, 173, 182-184). |
| Again, this (or these for versions of the general model) is a range model of policy and not of biology, ecology, or even economics. Authors in Behnke et al. (1993, ps. 3-8, 67-76, 89-103,122-130, 153, 225) more or less accepted at face-value the traditional concepts of carrying capacity, climax, dynamic equilibrium, economic optimization, diminishing returns, and so on (ie. they did not challenge these ecological and economic theories or paradigms). They merely questioned the application and relevance of these ecological-economic models when— to be blunt about it— tribal herdsmen are going to overgraze communal ranges in order to maintain high numbers of animals for traditional pastoral-subsistence values and not for industrial-capitalistic values. Authors in Behnke et al. 1993, especially ps. 89, 153, 176, 194, 225) were explicit that ecological-economic models based on assumptions of equilibrium and sustained yield (“environmental carrying capacity”, “ecological carrying capacity”, maximum carrying capacity”, etc.) were relevant for the management-production goals of commercial ranching operations in capitalistic economies where greater individual animal performance is necessary for production of commodities like “high quality meat”. |
| In industrialized democracies the policy was settled long ago in favor of scientific management for sustainable production of affordable food and fiber and not for status symbols (also the established policy of one wife to one husband). In these more or less competitive, capitalistic economies the standard ecological-economic equilibrium models are generally valid. This is not the case for subsistence, barter-and-exchange economies so models of ranges “at disequilibrium” are valid and relevant. Scoones in Behnke et al. (1993, p. 75) concluded that stocking rates for “’economic carrying capacity” for commercial beef production …” were “irrelevant for communal area situations”. “[W]hile the primary aims of cattle-keeping are security, savings and subsistence, it will remain logical for pastoralists to maintain high stocking densities. The logic will change if the production goals change” (Abel in Behnke et al. 1998, p. 194). |
| The thrust of arguments advanced by range scientists like those in Behnke et al. (1993) support the fundamental Clementsian model of plant succession that presents the retrograde progression down the sere from climax as retrogression (ie. range deterioration as departure from climax or dynamic equilibrium). Contrary to what the casual reader might prematurely conclude from reading a title like Range Ecology at Disequilibrium, careful reading shows the thesis of ranges at disequilibrium is verification of— not evidence against— the Clementsian doctrine of “vegetation development” (ie. plant succession). Range at disequilibrium is range disclimax. |
| For several academic generations students have been taught that subclimax or disclimax California annual grassland and not the pre-Columbian perennial bunchgrass prairie was the grazing type to be managed. They learned (at least were taught) that subclimax wheatgrass range was superior for livestock grazing. The desirability under certain conditions of managing range vegetation for other than climax was well-established by the first third of the last century (eg. Weaver and Clements, 1929, ps. 76-77). This lesson came from Plant Ecology more than it did from Range and Pasture Management (Sampson, 1923), the first actual textbook devoted to Range Management. |
| The next point that follows from the one just made is that the Clementsian paradigm is compatable with the disequilibrium models, which even if developed without prior knowledge of the Clementsian subclimax or disturbance climax, are but one condition or case of a decades-older textbook model. The only real difference (again, even the words like subclimax are the same) is that while disclimaxes were the exceptions in the Clementsian model they are the majority cases in the disequilibrium model. Different aspects of the same model (vegetation development; dynamic plant communities) are emphasized. |
| Ditto on the relativlely recent concept mentioned above that is known variously as patch dynamics or gap theory and the view of shifting-mosaic steady state. This general model developed along with and is a central part of Landscape Ecology. Briefly stated, patch or gap dynamics looks at succession, often secondary, on small patches of vegetation (eg. forest gaps) caused by local disturbances and from the perspective of landscape-scale. At heart, this is application of “vegetation dynamics” or “vegetation development” at local-scale within the overall larger plant community at scales up to regional size. A good summary paper on role of scale in landscapes as related to landscape equilibrium, disturbance, and stability is that of Turner et al. (1993). This is, of course, just another view or prespective on the same basic philosophy of vegetation as “developed” by the monoclimax-polyclimax-climax pattern progression or the Clements-Tansley-Dyksterhuis-Whittaker lineage of the Anglo-American tradition. In other words, the basic Celemntsian paradigm. Even the key word “dynamics” says this. Turner et al. (1993, ps. 216, 217, 218, 221, 223) spoke of “landscape dynamics”. All this is pure Clementsian “dynamic plant ecology”, at least that is the point from which these later perspectives begin. Patch is a unit in Landscape Ecology so this is essentially the application to Landscape Ecology of Clementsian ecology, specifically the processes of plant succession or what Egler (1954) later termed relay floristics and Connell and Slatyer (1977) called facilitation. See Barbour (1999, ps. 293-300, 301-302 and related ps. 22-23, 182-183, 294, 295). |
| Walker and Chapin (1987) felt that our understanding of vegetation to date was inadequate to formulate a “unified theory” of plant succession, but Connell et al. (1987) expressed the view that succession models “based upon incomplete knowledge” should be welcomed as useful because they stimulated thought and suggestions for further research. |
| The disequilibrium concept and the state and transition model is just such a case in point. The fact that some vegetation, ecosystems, and landscapes are at disequilibrium, disclimax, subclimax, etc. and that state and transition model(s) offer some explanation makes them consistent and complementary with the basic original Clementsian model of “dynamic vegetation” which emphasized stability or dynamic equilibrium. This more recent emphasis on disequilibrium along with the state-transition model can be and is being used with the older, overall model of plant succession or vegetation development. That this can be done in ways applied to on-the-ground-management was evident when in the National Range and Pasture Handbook the Natural Resources Conservation Service (1997) integrated the traditional range succession-retrogression model of Clements-Weaver-Dyksterhuis with the vegetation state and transition model. The result was a combined model having the option of evaluating range by the traditional ecological approach or by assessment for “rangeland health” (Natural Resources Conservation Service, 1997, Chapter 3, Section 1, p. 1-5 and exhibits 3.1-3 and 3.1-4). |
| The final point following from above is that Clementsian “dynamic plant ecology” is not only still relevant as historic background but also as the basis, the very foundation, for contemporary work with native vegetation. This includes the most current method used to evaluate the ecological status of range plant communities. In the National Range and Pasture Handbook the climax plant community was the “first vegetation state” and the state of departure (= retrogression) for all the other vegetation states. Climax is still the benchmark from which to assess other states and transitions. It is the terminus from which “a ‘road map to other states can be developed” (Natural Resource Conservation Service, 1997, p. 3.1-3). This does not mean that climax vegetation is the management goal. Climax is the ecological state— the termination of succession or the mature stage in the development of a sere— from which vegetation change is traced, mapped, evaluated, or whatever. It was just shown that this was exactly the case presented in textbooks written by Weaver and Clements over 70 years ago. |
| The latest method of range analysis by a federal agency even retained the key words of the Clementsian paradigm: “plant community development and dynamics”, “climax”, “plant succession”, “retrogression” (Natural Resources Conservation Service, 1997, p. 3.1-3). Some of those key terms and concepts trace back even beyond Clements but he was the one, first in The Development and Structure of Vegetation (Clements,1904) and from thereafter in Plant Succession- An Analysis of the Development of Vegetation (Clements, 1916), that forged them in the collective memory of vegetation students and in the lexicon of Range Management. Cowles (1899, p. 95) in the second paragraph of the first published study of plant succession in the United States described ecology as “a study in dynamics”. In the first paragraph of Development and Structure Clements told the world that change constitutes the “fundamental phenomena of vegetation” and, in the next sentence, that “vegetation is essentially dynamic” (Clements, 1904, p. 6). |
| It was shown in the review of the history of the vegetation cover type which follows that this unit, which became fundamental to Range Management and Forestry, was also the offspring of the Clementsian paradigm of vegetation development. |
| If that makes the range type an organism make the best of it. |