The orientor approach that was briefly introduced above, describes ideal-typical trajectories of ecological properties on an integrated ecosystem level. Therefore, it follows the traditions of various concepts in ecological theory, which are related to environmental dynamics. A significant example is succession theory, describing "directional processes of colonization and extinction of species in a given site" (Dierssen, 2000). Although there are big intersections, these conceptual relationships have not become sufficiently obvious in the past, due to several reasons, which are mainly based on methodological problems and critical opinions which have been discussed eagerly after the release of
Odum's paper on the strategy of ecosystem development (1969). Which were the reasons for these controversies?
Traditional succession theory is basically oriented toward vegetation dynamics. The pioneers of succession research, Clements (1916) and Gleason (1917) were focusing mainly on vegetation. Consequently, also the succession definitions of Whittacker (1953), Egler (1954), Grime (1979), or Picket et al. (1987) are related to plant communities, while heterotrophic organisms often are neglected (e.g., Horn, 1974; Connell and Slayter, 1977). Therefore, also the conclusions of the respective investigations often have to be reduced to the development of vegetation components of ecosystems, while the orientor approach refers to the whole ensemble of organismic and abiotic subsystems and their interrelations. These conceptual distinctions for sure are preferable sources for misunderstandings.
A sufficient number of long-term data sets are not available. Therefore, as some authors state throughout the discussions of Odum's "strategy" paper (1969), the theoretical predictions of succession theory seem to be "based on untested assumptions or analogies" (e.g., Drury and Nisbet, 1973; Horn, 1974; Connell and Slayter, 1977), while there is only small empirical evidence. This situation becomes even more problematic if ecosystem data are necessary to test the theoretical hypotheses. Consequently, we will also in future have to cope with this lack of data, but we can use more and more empirical investigations, referring to the orientor principle, which have been reported in the literature (e.g., Marques et al., 2003; Müller et al., i.p.). We can hope for additional results from ecosystem analyses and Long Term Ecological Research Programs. Meanwhile validated models can be used as productive tools for the analysis of ecosystem dynamics.
The conceptual starting points differ enormously. Referring to the general objections against the maturity concept, Connell and Slayter (1977) funnel their heavy criticism about Odum's 24 ecosystem features into the questions of whether mature communities really are "internally controlled" and if "steady states really are maintained by internal feedback mechanisms". Having doubts in these facts, they state that, therefore, no characteristics can be deduced from this idea. Today, there is no doubt about the existence of self-organizing processes in all ecosystems (e.g., Jorgensen, 2002). Of course there are exterior constraints, but within the specific degrees of freedom, in fact the internal regulation processes are responsible for the development of ecosystems. Hence, the basic argument against the maturity concept has lost weight throughout the years.
Comparing successional dynamics, often different spatial and temporal scales are mixed. This point is related to the typical time scales of ecological investigations. They are most often carried out in a time span 2-4 years. Of course it is very difficult to draw conclusions over centuries from these short-term data sets. Also using paleo-ecological methods give rise to broad uncertainties, and when spatial differences are used to represent the steps of temporal developments, the questions of the site comparability introduces problems which might reduce the evidence of the findings enormously. Furthermore, there is the general problem of scale. If we transfer short-term results to long-term processes, then we cannot be sure to use the right algorithms and to take into account the correct, scale conform constraints and processes (O'Neill et al., 1986). And, looking at the spatial scale, the shifting mosaic hypothesis (Remmert, 1991) has shown that there will be huge differences if different spatial extents are taken into account, and that local instabilities can be leading to regional steady-state situations. What we can see is that there are many empirical traps we can fall into. Maybe the connection of empirical research and ecological modeling can be helpful as a "mechanism of self-control" in this context.
Due to the "ontic openness " of ecosystems, predictability in general is rather small but in many cases exceptions can be found. The resulting dilemma of a system's inherent uncertainty can be regarded as a consequence of the internal complexity of ecosystems, the non-linear character of the internal interactions and the often-unforeseeable dynamics of environmental constraints. Early on, succession researchers found the fundamentals of this argument, which are broadly accepted today. The non-deterministic potential of ecological developments has already been introduced in Tansley's (1953) polyclimax theory, which is based on the multiple environmental influences that function as constraints for the development of an ecosystem. Simberloff (1982) formulates that "the deterministic path of succession, in the strictest Clementsian mono-climax formulation, is as much an abstraction as the Newtonian particle trajectory" and Whittaker (1972) states, "the vegetation on the earth's surface is in incessant flux". Stochastic elements, complex interactions, and spatial heterogeneities take such important influences that the idea of Odum (1983) that "community changes...are predictable", must be considered in relative terms today, if detailed prognoses (e.g., on the species level) are desired. But this does not mean that general developmental tendencies can be avoided, i.e. this fact does not contradict the general sequence of growth forms as formulated in this volume. Quite the opposite: this concept realizes the fact that not all ecosystem features are optimized throughout the whole sequence, a fact that has been pointed out by Drury and Nisbet (1973) and others.
Disturbances are causes for separating theoretical prognoses from practical observations. One example for these non-deterministic events is disturbance, which plays a major role in ecosystem development (e.g., Drury and Nisbet, 1973; Sousa, 1984). Odum (1983) has postulated that succession "culminates in the establishment of as stable an ecosystem as its biologically possible on the site in question" and he notes that mature communities are able to buffer the physical environment to a greater extent than the young community. In his view stability and homoeostasis can be seen as the result (he even speaks about a purpose) of ecological succession from the evolutionary standpoint. But in between, the guiding paradigm has changed: today Holling's adaptive cycling model (1986) has become a prominent concept, and destruction is acknowledged as an important component of the continuous adaptation of ecosystems to changing environmental constraints. This idea also includes the feature of brittleness in mature states, which can support the role of disturbance as a setting of new starting points for an oriented development.
Terminology has inhibited the acceptance of acceptable ideas. The utilization of terms like "strategy", "purpose", or "goal" has led to the feeling that holistic attitudes toward ecological successions in general are loaded with a broad teleological bias. Critical colleagues argued that some of these theories are imputing ecosystems to be "intentionally" following a certain target or target state. This is not correct: the series of states is a consequence of internal feedback processes that are influenced by exterior constraints and impulses. The finally achieved attractor state thus is a result, not a cause.
Summarizing, many of the objections against the initial theoretical concepts of ecosystem development and especially against the stability paradigm have proven to be correct, and they have been modified in between. Analogies are not used anymore, and the number of empirical tests is increasing. On the other hand, the theory of self-organization has clarified many critical objections. Thus, a consensus can be reached if cooperation between theory and empiricism is enhanced in the future.
Was this article helpful?