Ecological pattern and ecosystem theory

Ecological pattern and ecosystem theory

Ecological Modelling 158 (2002) 181 /200 Ecological pattern and ecosystem theory C.S. Reynolds * CEH Windermere La...

356KB Sizes 1 Downloads 17 Views

Ecological Modelling 158 (2002) 181 /200

Ecological pattern and ecosystem theory C.S. Reynolds * CEH Windermere Laboratory, Algal Modelling Unit, Centre for Ecology and Hydrology, The Ferry House, Ambleside GB-LA22 0LP, Cumbria, UK

Abstract This paper considers the central tenet of the ‘Econet Initiative’ and the developing view of the structure and function of ecosystems and their essential emergence from ecological interactions at the level of individuals and populations. Most of what we consider to be community ecology is concerned with how individuals and species-specific populations respond to, smooth, or fail to accommodate, the consequences of inhabiting continuously variable environments. The essential principle is that individuals, populations, communities, systems, each build towards a definable, contemporaneous, habitat-specific supportive capacity, set by the resource base and the opportunities to process them. While both are amenable, interspecific competition is weak and many species may have the chance to raise their recruitment potential and build biomass, with the dynamic advantage going to the most efficient converters at every trophic level (r - vs. K -selection). As a result of the biomass expansion or of the independent diminution in opportunity or resource base, the special adaptations of disturbance-(R) or stress-(S) tolerant species allow them to glean survival under increasingly competitive strains. Persistent variability and renewal of growth opportunities make exclusion difficult and extinction is more a local than a regional consequence. Poorly competitive species survive through an ability to find new habitat patches ‘released’ by variability. Integration to larger geographical patterns gives the linkage to higher-order ecosystem behaviour. Some examples of contemporary ecological issues (latitude gradients, optimality, threshold behaviour and the maintenance of biodiversity) are reviewed in the context of systems behaviour. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Emergence; Community ecology; Stress; Disturbance; Selection; Extinction; Biodiversity

1. Introduction This paper has been prepared in response to an invitation from the editor, Professor Sven-Erik Jørgensen to contribute to his ‘Econet Initiative’. Its aim is to compound and promote the emerging, unifying view of the structure, function and

* Tel.: /44-1539-44-2468; fax: /44-1539-44-6914 E-mail address: [email protected] (C.S. Reynolds).

thermodynamics of ecosystems. The attainment of a nascent but reasonably robust ecosystem theory has, for long been an implicit, if rather vague, goal in ecology generally. The bar to its achievement has sometimes been attributed to problems of scaling, both within and among systemic hierarchies (Allen and Starr, 1982). Considerable difficulties are encountered in linking, on the one hand, the biologies of organisms and the ecologies of populations to the fluxes of material and energy quantifiable at the level of ecosystems

0304-3800/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 3 8 0 0 ( 0 2 ) 0 0 2 3 0 - 2


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

and biomes, on the other (Lawton, 1999). In truth, it now seems more likely that the sheer complexity of the ramifying hierarchical linkages within natural systems provides the most daunting barrier to understanding. The interspecific interactions that have, for long interested ecologists most */ competition, predation, co-existence */have not been made any easier to explain by expressing them in the terms of energy flux (Paine, 1980), so investigation of the higher levels of organisation have tended to be deferred. Filling in the relevant vertical connectivities has continued to present a considerable philosophical challenge. An important advance came with the attempts to quantify the dynamism of ecosystems, based upon the emergence, organisation and regulation of working structures (‘ascendancy’: Ulanowicz, 1986). These have been developed into a substantial theoretical base of guiding principles and rules for the assembly and organisation of ecosystems (Jørgensen, 1992, 1997; PahlWostl, 1995; Belyea and Lancaster, 1999; Brown, 1999; Strasˇkraba et al., 1999). So far as providing an empirical base for linking the organisation of ecosystems to the ecologies of the component individual organisms is concerned, the clearest large-scale statistical patterns of species distribution, abundance and richness have appeared in the literature on pelagic systems (Reynolds, 1997a, 1998, 2001; Jørgensen and Marques, 2001). The point that Jørgensen and Marques (2001) make strongly is that we now have an outline theory of ecosystems that is widely applicable in ecology generally and not to hydrobiology alone. Of course, the theory has to be refined and its applicability must be improved. For this to happen, ecologists need to be encouraged to contribute to its development. They may do this by adding to the theoretical base but the greatest need is to establish the extent to which the relevant observations and experimental results appearing in the ecological literature actually conform to and support the developing rules of system assembly and to identify where adjustments are necessary. In order to fulfil the detail of Professor Jørgensen’s invitation, this paper will consider some recent contributions to the ecological literature,

and attempt to show that the specific observations and scientific deductions discussed might also have contributed to the construction of a broadly acceptable view of ecosystem theory. Before that, however, some brief review of the network of principles and basic rules for the assembly and operation of ecosystems that have emerged from the study of aquatic systems is appropriate.

2. Principles of ecosystem elaboration Individual living organisms are dynamic, selfmaintaining systems, engaged in well-known characteristic vital activities */nutrition, growth, respiration, excretion*/directed towards their own self-replication (reproduction). From a reductionist standpoint, organisms are collections of atoms of some 20 elements, arranged into molecules that comprise structures (organelles, organs, tissues) or that constitute a store of potential chemical energy needed to fuel the characteristic activities. The activities themselves are regulated through another set of molecular structures */the genes. These direct the organism’s functions and, to a great extent, determine the limits to the organism’s suite of activities. It is self-evident that none of these organisms is able to function without exchanges of molecules and matter with its surroundings (its environment) or, ultimately, without impinging upon the activities and environments of other organisms. Such interactions are, of course, the basis of ecology. Organisms differ in their sources of energy potential. The absorption of short-wave solar radiation and its storage in high-energy carbon bonds by photoautotrophic plants (photosynthesis) is of special relevance to ecosystem function. Gathering the other components, transporting and assembling them into biomass are variously achieved by transformation or oxidation of the primary, reduced-carbon product. Others rely on being able to acquire and to oxidise primary or recombined product, either as consumers (phagotrophs) or decomposers. This arrangement introduces interactions among the organisms, embracing the supply of resources and energy, and the manner of energy transfer among the

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

interacting components. These include the close relationships that are mutually beneficial (as in symbioses) as well as the more familiar consequences of prey /predator dynamics, electivity and defence, comparative dynamics and competitive efficiencies. There is a wider, community effect of these interactions in the sense that they modulate and damp the pulsed nature of the primary energy income, to bring about something closer to a more even and steady state, in which certain outcomes become predictable. The higher levels of organisation are set by the consequences of interaction among the functional components. Besides the primary producers (which manufacture the materials of biomass and, to varying degrees, the habitat architecture), the network of herbivorous and carnivorous consumers that make up the food web and the decomposers that exploit the residual energy in the cadavers and metabolic waste products, the systemic consequences are also dependent upon the resource base (taken here in the widest sense-to include light, water, nutrients and carbon) and the anabolic climate (temperature fluctuation, toxic gases). Evolved community function capitalises carbon flow into maximum sustainable structure, favours material re-use and conservation, and biases towards the most energyefficient biomass yields. There is debate about the network structure required to deliver functional efficiency (see later). For the moment, it is sufficient to point out that only the key roles need to be fulfilled (primary producer, phagotrophic consumer, heterotrophic recycler) in order for a network to be closed and to operate at the capacity of the effective resource base or the modulated energy flux. The activities of all but the most efficient participants in each role become redundant and weaker species become excluded by the stronger ones. On the other hand, simplified structures are poorly equipped to meet all contingent variability and pulsation in the abiotic components. As with the highway network, alternative routes offer other ways for the transport of essential goods when the primary route becomes blocked. The community equivalent is that though there may be dominant producers, herbivores, carnivores and decomposers, other


species are often present or sufficiently abundant locally to be able to assume the functional role at each trophic level. Again, functional feedbacks have the effect of damping the potential pulsation in the flow of energy. Following this logic, the hierarchy progresses smoothly to the ecosystemic level. Here the focus is rather more on the aggregate of species function. Rather than precisely who does what, it is the overall allocation and dissipation of the energy income that is relevant. Of the original photosynthetic investment (supposing a yield of 1 kg of carbohydrate per 15 MJ of energy captured, roughly 470 kJ mol 1 of carbon, thus, reduced), estimable fractions are allocated to protein synthesis and increase in biomass, and to meeting the often-considerable energy requirements to drive the actual foraging and assembly processes. As almost every energy-consuming process requires the controlled oxidation of organic carbon bonds, the ecosystem is comparable to a market economy, using organic carbon (Corg) as its trading currency. At each link in the network, more of the original investment in Corg is spent (Oxidised), until the final heterotrophic decomposers eke out the last of it. The ecosystem is a network of residual organic carbon being gradually oxidised away as it is processed through the prey /predator network. The reductionist view of the ecosystem is that it is a network of regulated energy dissipation. Before seeking more explicit connections between ecological and ecosystemic patterns, due emphasis must be given to the nature of the vertical linkages. One is that there is no internal controlling direction given to ecosystems other than at the genetic level. Organisms, indeed, are the largest self-regulating units of assembly with any kind of tangible, regulatory and reproducible set of genomic instructions. The behaviour of an ecosystem is often said to be self-organising or endowed with a ‘goal function’ (Strasˇkraba, 1980); in truth, it is entirely emergent, consequent upon the sum of the component processes attributable to the activities of individual organisms. The functional responses of ecosystems to a fluctuating abiotic environment are, accordingly, shaped from the bottom upwards. However, it has to be appreciated that the organismic interrelationships


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

are exceedingly complex: systems are beset with dynamic feedback mechanisms such that many ecological effects are interpretable in terms of a top /down modulation. However, when it becomes important to establish the broader contribution of these mesoscale feedbacks to the general patterns of ecosystemic function, it is helpful to consider both in the context of the gradients and units of material or energy flow.

3. Rules of ecosystem elaboration The vast detail of bottom-up processes, with positive as well as negative ecological feedbacks, nevertheless, unfolds within boundaries that are constrained. As Lawton (1999) points out, to be able to define these constraints by ecological laws or principles is still difficult, as we have remained for some time at the stage of setting out the general patterns rather than the defining the governing rules. Of course, biological systems necessarily conform to universal principles and properties of matter */the laws of thermodynamics, the rules of atomic structure and stoichiometry, as well as the principles governing the diffusion of solutes, the movement of particles through fluids, the behaviour of gases, and so on. Lawton (1999) also makes the point, already emphasised here, that the general pattern of biological systems are clearest at the grand scale-in macroecology and developing ecosystem theory. Can we detect the emergent patterns of the robust principles through the analysis of the patterns of contingent responses to variability? I tried to devise a very simple model of ecosystem growth based on minimalist components and to build up its components progressively

(Reynolds, 1997a). Invoking the bottom-up principle, I used morphological and physiological data on the alga, Chlorella , to establish the metabolic limits of its assembly. Against a finite and modest resource base, the maximum supportable biomass (B ) is set by the least available resource (K ) relative to the lowest adequate stoichiometric quota (q ) required for the assembly of complete cells (B /K/q ). If the nutrients are made nonlimiting, the ceiling quickly moves to processing limitations set by the flux of photons and the rate of supply of carbon dioxide. If the latter are assumed to be delivered at the highest rates possible (the solar constant, or at least, in terms of the photosynthetically active wavelengths, corrected for atmospheric absorption and the selfshading by the assembling biomass: atmospheric CO2 invades the water surface under the most favourable concentration gradient), then a maximum biomass carrying-capacity may be derived. Precisely these calculations were used to construct Fig. 1a, which compares the ability of the assembling producer biomass to intercept incoming energy with the respiratory costs of its maintenance. Energy handling increases with photosynthetic biomass but with a decreasing efficiency as light harvesting centres increasingly interfere with (‘self-shade’) each other. The respiration and repair make over-provided biomass too expensive to maintain-balance sets the upper capacity limit. Below this level, energy harvesting capacity is shown to exceed contemporary costs, with a flux margin that may be invested into the production of new biomass. In Fig. 1a, the excess is represented by the segment above the plot of biomass-dependent maintenance. Conceptually, this model differs little from income /cost representations for individual organ-

Fig. 1. (a) Comparison of the light-harvesting potential of a hypothetical pelagic vegetation, based on the physiological characters of a cultured Chlorella , with the fixed (mainly respiration) costs of its maintenance. The bow shaped area between the curves represents the potential exergy of light harvesting that is available to invest in new standing biomass. (b) Most autotrophs have to function where nutrient resources fluctuate within subsaturating limits and the light income is continuously variable. (c) Fluctuations, as here, in energy income necessitate alternating growth and readjustment responses in the supportable primary biomass; as the main food source of consumers, fluctuations in the availability of primary biomass precipitate responses in the consumer biomass. Habitats may vary between being typically (d) resource-deficient, and so able to support predominantly stress-tolerant (S) species, or (e) being short of processing opportunities, and so able to support predominantly tolerant ruderal or explerent R species. Based on Figures in Reynolds (1997a).

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200


Fig. 1

isms and populations, including the stock-recruitment models used by fishery scientists (Ricker, 1954; Gulland, 1983), where increase in biomass depends upon a net positive balance of growth and

recruitment over losses and exploitation. Viewed in terms of energy flux, the balance is the opposite of entropy: it has been called ‘negative entropy’ or ‘negentropy’ but the correct (and preferable) name


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

is exergy (Mejer and Jørgensen, 1979). A positive exergy flux is essential to the net production (P ) of new biomass (B ). It is a short step in logic to see that those that increase specific biomass fastest (high P /B ) are the most efficient retainers of exergy and, given initial equality, the species most likely to dominate the assembling communal biomass. As Jørgensen and Marques (2001) put it, prominence of the life forms invoking the most ordered structure and the greatest distance from equilibrium is favoured. So far, the predictable outcome is a rapid ascent to the theoretical limit of the capacity. This is rarely attained, for two reasons. One is that the nutrient base (K ) may be inadequate to sustain the maximum live biomass (B ) anyway. The other is that the speed of processing resources and their assembly may be constrained by the supply of reduced carbon, either because the supply of inorganic carbon dioxide is strained (often a real problem in water, in spite of the solubility of the gas) or because the light is not just submaximal but is sometimes insufficient even to meet the energy costs of biomass maintenance. In the latter case, the biomass becomes literally unsustainable, and must be shed, either as a cyclical or a disturbance response. These important symptoms of environmental variability, to which system development must conform, are sketched in Fig. 1b. The resultant inconsistency in the energy flux is the main source of variation in the biomass capacity (Fig. 1c). The model applies with little material difference to the system that includes heterotrophs. If we introduce a herbivore (in this instance, the filter-feeding Daphnia ), where the alga becomes the source of resources (recombinable amino-acids and lipids) and exergy (all reduced organic carbon, including carbohydrates), the mass of the animal can increase within the capacity of the food. Correspondingly, it must soon decline if the supply fails the maintenance requirement and/or the harvesting threshold (the minimum food concentration), accepting that this may be the consequence of the rate of consumption exceeding that of the recruitment of new food organisms. The final part of the assembly model concerns how the functional specialisms of the main organ-

isms are selected. The longheld presumption that the best-fitted species will be favoured is giving way to a view that communities often comprise much more random assemblages of species that have arrived in the particular habitat and have found the conditions tolerable (Keddy, 1992; Weiher and Keddy, 1995; Kelt et al., 1995; Rojo et al., 2000). The principle that, of those present, the most efficient users of available exergy are most likely to dominate is not violated but, so long as there is variability in the habitat, the resource supply and the processing opportunities, many other, functionally similar species may co-exist. This is not just a restatement of Connell’s (1978) hypothesis to account for high species diversity in habitats disturbed at moderate frequencies in relation to the generation times of the main species but it is also a recognition that exclusion of competitively inferior species is harder to achieve while all those present are able to satisfy their respective requirements and so secure their survival. However, it is not difficult to anticipate events in less benign environments, or less accommodating circumstances in erstwhile benign environments, where the satisfaction of growth and survival criteria becomes the subject of competition. In the case of chronically resource-deficient environments (a limnologist considers a nutrientpoor, oligotrophic lake but analogies run to a fertile but rainfall-deficient semi-desert), the environment exerts a severe and rarely relieved stress that tests the abilities of plants, animals and decomposers to fulfil their standard functions. The conditions act like a filter, allowing only the fittest, best-adapted species to survive. Moreover, the more severe the stress, the finer the filter becomes; the less is the available resource, the greater is the competitive value to individual organisms to sequester and conserve it. Truly, only the fittest will survive, until even they are prevented from functioning. In communal terms, biomass is likely to be maintained near the resource-determined capacity and to be dominated by species maintaining a high exergy flux (Fig. 1d). The structure is rarely forced into energy deficit and is, therefore, difficult to disturb in the Connellian sense: progress towards low diversity and

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200


Table 1 Proposed rules of community assembly, modified after Reynolds et al. (2000), Reynolds (2001) (1) Provided that species are present in substantial numbers (i.e. they constitute ‘viable inocula’) and the conditions obtaining at a location are adequate to meet their minimal requirements to effect a net increase in biomass, then they will grow wherever and whenever the opportunities are fulfilled (2) Of the species present, the most prominent will be those bringing structure that is best ordered and furthest from energetic equilibrium (i.e. they store the most exergy) (3) In environments where there is unfilled resource and processing-flux capacity, assembly is biased towards adaptive traits favouring rapid acquisition and transformation; i.e. r -selection predominates in early succession (Odum, 1969) (4) Pending constraints on resource acquisition and processing, system growth raises the aggregate ability of the developing assemblage to harvest and store exergy. More species are able to operate in benign environments; the richer is the species representation, the more complex is the network of energy flow and the greater is the information content (5) Ascendant development brings in its wake, consequent, or ‘self-imposed’, constraints upon resource availability or the opportunities to process them. There is a progressive change in the environmental conditions and these alter the organismic traits that are beneficial to the maximisation of growth. Species-specific tolerance of resource stress (Grime’s S-strategists) or disturbance (R-strategists) become selectively decisive (6) Habitat deficiencies bias in favour of particular species traits. Organismic preadaptations and facultative adaptabilities of individual species determine their relative competitiveness and, thus, influence the species survival and the functional operation of the residual structure (7) The more severe the constraint, the more selective is its impact and the more robust is the direction of assemblage ascendancy. However, the succession of events and their emergent outcome is increasingly predictable from the attributes and performance limits of the species available: the fittest survive (8) Emergent outcomes are always subject to overriding, environmental resetting and structural reorganisation through the intervention of external forcing, loss of unsustainable biomass and a sharp return towards equilibrium conditions. Habitat opportunities are opened to post-disturbance exploitation

competitive exclusion may be slow but the direction is assured (Hardin, 1960). The contrasted case is that of severe constraints upon the opportunities to build biomass. ‘Explerent’ life-history adaptations permit plants to exploit transient environment (desert rains, recently opened forest glade, shifting flood-plain meanders) or of animals ‘to boom and bust’ within temporarily available habitats, when the missing processing limit is fulfilled. Relative to the generation times of phytoplankton, periodic access to only short bursts of daylight, as a consequence of entrainment through extreme mixed depths or severe turbidity, demands an analogous capability to maximise every growth opportunity. Now, resources are rarely limiting and the supportable biomass is often quite insensitive to modest fluctuations in the supply but biomass building is

contingent on the ability to make any capital at all from the truncated exergy fluxes (Fig. 1c). These are not new revelations but simply an alternative explanation for the importance of stress and disturbance as organising drivers in ecology (Grime, 1979). Thermodynamic arguements expose the same C-, S- and R-strategies for maximising developmental opportunities and building specific biomass. Jørgensen and Marques (2001) (see also Jørgensen, 1997) explain the whole process in terms of genetic information content and the work that is able to control. A background in the ecology of short-lived pelagic organisms persuades me that intraconversion among the units of carbon reduction and oxidation fluxes and the work exchanged are adequate to illustrate the middle orders of community ecology. Whichever is the preferred medium of empiricism, the


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

Fig. 2. Log /log representation of aquatic habitats in terms of resource-supportable biomass (which, it is assumed can be constrained stoichiometrically by phosphorus or nitrogen availability) and processing fluxes, which may be set by light income (PAR flux), carbon delivery or the oxidation flux demanded by the organic carbon delivered.

word-models describing the assembly and organisation of organisms into emergent structures are harmonious. The lay out in Table 1 follows the examples of Reynolds et al. (2000), Reynolds (2001) but the statements are not at odds with other attempts to devise the rules of community assembly. To conclude this section, reference may be made to the use of templates to represent the possible courses and likely outcomes of ecosystem elaboration. Derived in a series of steps from the original concept of Southwood (1977), Fig. 2 is the culmination of a series of attempts to show the influences of resources and of processing upon the theoretical state of the ecosystem that it is possible to assemble. Its development and its use are explained in a companion paper (Reynolds,

2002) but the relevance to the present argument is, self-evidently, that to represent habitat against axes showing how much system exergy can be harvested, stored and efficiently dissipated provides a conceptual medium for linking ecosystem processes, from the foundation of opportunity, species strategies, community assembly and ecosystem function.

4. Ecosystem elaboration and ecological issues In pursuance of Professor Jørgensen’s invitation and, indeed, the objectives set by Jørgensen and Marques (2001), this section will consider some of the topical issues raised in recent contributions in the ecological literature and the extent to which

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

they generate, and, thereby, conform to, the patterns of ecosystem emergence. Many of the current issues are related to biodiversity, the structure of functional communities and the role of competition in shaping them. Biodiversity is an issue because extinctions of species as a result of habitat destruction through human activities are occurring at a fast and accelerating rate. The genetic loss is alarming enough but ecologists are divided in how they view the consequences of these depletions. They have occurred in the past, without break down of global functions, but we have little idea on their temporal and spatial scaling. There is still little agreement about the mechanisms that uphold the richness of species or of those that gave rise to it in the first instance. Without this knowledge it is difficult to judge whether it matters to ecosystem function or to come up with the recommendations about how to reverse the trend. Vertical connectivities in ecological processes augur for the importance of competition for resources and space, subject to major interactions with other trophic levels in the food web, in how well individual species will fare or whether they will fail and die out. The relevant assembly rules favouring diversification, 3 and 4, give place to 5, 6 and 7 under conditions of competition for limiting resources or restrictive processing constraints, when species tolerances become more severely pressed. Diminution of species numbers can occur through constriction of available habitat as well as through competition for what is left of it. What is needed is a view of ecosystem capacity and a means of recognition of how filled it is (Harris, 1999). The converse of the same deduction is that we need a means to judge the extent to which biological communities are saturated or whether they are constrained by competition for limiting resources or fluxes. In the language of ecosystem theorists, the question might become ‘how much of an exergy buffer can the system raise?’ 4.1. Latitudinal gradients As it is, the main ecological paradigms relating to patterns of species richness invoke the structur-


ing driver of competition fought along large-scale latitudinal gradients in the species pool size. The long-recognised tendency for species richness to peak in equatorial regions but to diminish towards the poles (Fischer, 1960; Rohde, 1992) seems real enough (at least when applied to vascular plants and vertebrate animals; it is not clearly so for aquatic micro-organisms; see Kalff and Watson, 1986; Finlay et al., 1998; Finlay and Clarke, 1999), but it has not been adequately explained (Huston, 1979, 1994), The delineation of another latitudinal trend, that the geographical ranges of species tend to be smaller towards the equator (‘Rapoport’s rule’; Stevens, 1989), has been widely seen to contribute part of the explanation: simply, with higher conditions of warmth, moisture and material turnover, more species are able to fit into the space available. Recent attempts to simulate these effects in a model (Taylor and Gaines, 1999) revealed several surprises. By itself, the application of the Rapoport rule generated a latitudinal gradient of species richness opposite to that observed. However, the introduction of the effects of interspecific competitive interactions, equalised across all latitudes, brought about a trend consistent with the observed species-richness gradient. However, this outcome applied only at the regional level and was dependent upon all points on the Earth being filled to a competitively defined level of species saturation. The findings have many implications for assessing biodiversity. If the differing physical scales of range and interspecific competition among organisms of very different sizes and activities are not addressed, differing impressions of species richness obtain (Lyons and Willig, 1999; Loreau, 2000). Dynamic changes within small sample patches, as a result of migration, recruitment or death, may lead to almost stochastic changes in representation, whereas, the same populations seem persistent at larger spatial scales scale (Donalson and Nisbet, 1999; Godfray and Lawton, 2001). These occurrences are not well predicted by traditional ecological state formulations (e.g. the Lotka / Volterra equation), although they do remain relevant at the larger scale. Whereas, small-scale extinctions become frequent, if not inevitable, differences among patches affect competition and


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

the identity of species at risk. The system that is differentiated into patches, each described by variable limiting capacities and constraints (there is a high b diversity, in the sense of Schoener, 1986; see also Lawton, 2000) and lending a strong. TypeI proportionality between regional (g-) and local species richness (a-diversity) provides a platform wherein many species will find survival refuges. The crucial vertical connectivity is the betweenpatch variability in species ‘filtration’, sensu Weiher and Keddy (1995). If local patches are incompletely filled, species performances may be more saturated than constrained, with the more rselected species exploiting any selective strategic advantage. Once the competition for resources or processing ability intensifies to the point where the use of a molecule of nutrient or a photon of light, or the consumption of food and the use of oxidant, is such to deny it to another species, then outcomes become weighted in favour of the efficient, conservative, patient organisms-those stress-tolerant species (of Grime, 1979) or the patient species (of Ramenskiy, in Romanovsky, 1985) that can still manage to maintain the highest positive exergy balance. Theory predicts that as the criteria of filtration become finer, progressive exclusion of the less tolerant species must follow and local species richness (a-diversity) will tend to fall. Moreover, the more stressful are the regional conditions, then the greater is the number of patches in which the same limiting constraints apply (b-diversity falls), perhaps biassing towards aType-II regional richness, where the local richness is constrained above a certain level of regional richness. Interestingly, the behaviour of planktic systems is helpful in distinguishing large-scale pattern. In local phytoplankton assemblages enduring severe constraints upon processing opportunities (the establishment of ‘Planktotricheta’ is a classic case; see Reynolds et al., 2002), low adiversity dominance is achieved quickly and independently of regional richness (thus, quite unmistakably indicative of Type-II systems). This behaviour differs fundamentally from that of nearcapacity, resource-limited populations where local exclusion rates are so slow that a Type-I relationship is maintained for long periods (Reynolds and Elliott, 2002); instead of investment in standing

crop, planktic production in oligotrophic surface waters is quickly turned over by a modest but rapidly renewing biomass (Reynolds, 2001; see also below). Examples of behaviour in other systems that are consistent with these large scale patterns have appeared in the recent ecological literature. Variable performance (recruitment, growth, survival) in oysters in relation to differences in exposure to hydrodynamic variability affects the rate of habitat occupancy, before the habitat itself begins to become critical to the limitation of net production (Lemhan, 1999). Interactions among ‘competing’ species of protists affect local dominance (Morin, 1999), but the outcomes may be influenced by the frequency of disturbance, as is true of the red seaweeds studied by Dudgeon et al. (1999) Exclusion is as relatively slow when biomass productivity is kept low by the stress of modest resource availability (Morin, 1999; Wilsey and Potvin, 2000) as it is in the plankton. Such trends superimposed upon the latitudinal variability in habitat quality rather support the tenancy of the longstanding richness rules (latitudinal species richness gradients. Rapoport’s rule) but that their mechanisms are underpinned with wide-ranging macroecological principles governing the capacities of habitats and the opportunities to exploit them. 4.2. Optimality Optimality theories are based on an assumption that natural selection must favour successful behavioural and reproductive strategies (Morris and Davidson, 2000) Were it otherwise, species would quickly suffer extinction. Behaviour is said to be optimal when a marginal increase in fitness benefits is equal to the marginal increase in costs. This ideal is closely akin to the concept of raising maximum exergy. Morris and Davidson (op.cit.) investigated the foraging of white-necked mice (Peromyscus leucopus ), given varying levels of food and of habitat cover. They were able to confirm convincingly three optimality predictions: (i) that individual mice would quit from foraging at high food availability sooner in open, edge habitat, with its predation higher risks, than within its preferred woodland sites; (ii) that mice would

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

quit foraging in rich patches with little cover sooner than in rich patches located under cover; and (iii) that the difference in quitting rates between covered and open patches would be less than the difference in quitting rates between woodland and the riskier. Moreover, the differences in resource usage under the contrasted opportunities were reflected in such reproductive indicators as litter success and maternal survival. With good harvesting rates, maximal cover and minimal mortalities, the local population of mice is set to increase. There is, thus, a strong similarity with the criteria of photoautotrophic recruitment, represented in Fig. 1a and c. Within the limits of their own adaptability, individuals will modify behaviour to match the variability of habitats; once exceeded, however, less adaptable species will increasingly fail to those with greater adaptability and tolerance. This is implicit in the experimental results of phytoplankton adaptation to grow under low light doses (Reynolds, 1989) and is explicitly demonstrated in the mixed-depth manipulations of Huisman (1999) Density-dependent territoriality offers another means of reproductive optimisation variable habitat fitness. Growth of brown trout (Salmo trutta ) in streams may be sufficiently resource dependent for the self regulation of population densities to provide the best means of recruiting strong, healthy fish to the population (Jenkins et al., 1999 The long-term study of the breeding and recruitment success of trout in a small English stream (of Elliott et al., 1997) showed particularly well the accommodation and limits to interannual variability in habitat quality, as well as the time scale required to read to and recover from years with poor breeding success, mainly attributable to seasonal drought. Optimality is observed among flowering plants in the plasticity of biomass allocation to roots as opposed to shoots, or to the proportion invested in leaf surface, in response to variable water availability or shading (Grime, 1979). McConnaughay and Coleman (1999) were able to show plasticity in biomass elaboration in annuals that responded to nutrient and light availability. Grime (1987) used his CSR model of alternative strategies to repre-


sent the role of behavioural and regenerative optimisation to plot the course of vegetation development and successional pathways in benign, stressed and disturbed habitats, in what is one of the clearest illustrations of the vertical linkage between organismic properties and the emergence of high-order ecosystemic structure. 4.3. Structural thresholds The crucial question (how do the processing requirements determine the broad-scale nature of the emergent community structure and the functions to be fulfilled by its key components?) becomes a central issue in linking ecology and ecosystem science. For instance, as Deutschman et al. (1999) assume in trying to explain forest succession from initiation through to climax, there is throughout a universal requirement among the photautotrophs for light. We may make precise predictions for mass ‘capacity’, based on aggregate light income, and ‘processing rate’, based on photoperiod and intensity, but the switches from herbs to shrubs to forest canopy are only remotely sensitive. There is a clear need to incorporate the change thresholds in optimality of resource gathering. In short, what determines the province boundaries in Fig. 2? A good example of threshold relationships can he seen in the roles played by the means of harvesting primary products (the resource capacity) and by the functional limitations to trophic efficiency (the processing ability) in structuring the zooplankton. The long-held paradigm that zooplankton grazes phytoplankton to low levels unless it is simultaneously predated by fish is flawed and must be rejected. In fact, the various components of the zooplankton-from cladoccrans, copepods, rotifers and protists-do not share identical food resources, they forage in contrasted ways and they invest the exergy gain differently. Filter feeding is an efficient way of gathering food, subject to there being an adequate concentration of appropriately-sized and appropriately-nutritious particles. The dynamics of Daphnia spp., obligately filter-feeding cladocera, are conspicuously subject to temperature, hydraulic exchanges and predation (e.g. Saunders et al., 1999) but they


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

are especially influenced by the adequacy of the food supply (Burns, 1968, 1969; Lampert, 1977). ‘Adequacy’ is quantifiable in so far as food must be particulate (size generally, 0.5 /50 m) and must be able to sustain a concentration in excess of / 0.1 mg C l 1. At temperatures around 20 8C, the maximum growth requirement is saturated at around 0.5 mg C l 1. It does not matter that the food may be live phytoplankton: bacteria, detritus or some mixture of these (although, clearly, they are not equal in delivering a balance of all essential amino-acids and micronutrients). However, it is of general interest that the range 0.1 /0.5 mg C l1 as phytoplankton corresponds to a chlorophyll-a concentration of about 2/10 g l1 and to a comparable range of biologically available phosphorus (Reynolds and Davies, 2001). Except that the range of palatable particle sizes is smaller, these concentrations seem to apply also to filterfeeding rotifers (such as Keratella , Brachionus ). According to the size-efficiency principle (Brooks and Dodson, 1965), they are likely to be less successful than Daphniids except at high food concentrations or under selective predation, although their shorter generation times confer advantages in terms of being able to exploit opportunities in frequently or rapidly changing conditions (including of rivers: Reynolds and Descy, 1996). At concentrations substantially below 0.1 mg C l 1, food must be actively sought out, something that calanoid copepods do rather well. However, where the supportive capacities are chronically below 0.1 mg C l1, high biomass can never be maintained but photosynthetic production can still proceed within the limits set by the fluxes of light and carbon. As is now well known, fixed carbon is cycled principally through ‘the microbial loop’ (Azam et al., 1983), based upon bacteria mopping up the unused algal photosynthate, and their consumption at the same viscous scale heterotrophic flagellates and ciliates. The latter are large enough to reward the foraging activities of calanoids, at concentrations at least of order of magnitude less than that that is needed to satisfy filter feeders (Hart, 1996). The biomass constraint affects all trophic levels (everything is restricted to low aggregate abundance by low resource avail-

ability) but oligotrophic systems are able to process the resources quickly, theoretically to the delivery limits of the carbon or light fluxes. In this way, we see two quite different ways of driving the bottom-up, carbon harvesting of pelagic systems. The eutrophic way, involving forage by filtration, is dependent upon the producer biomass (B ) that can be maintained; the oligotrophic way depends upon the productive turnover (P ) by a meagre biomass through its ability to capture and concentrate the available exergy. Indeed, the power delivered in oligotrophic lakes (2 /4 MJ m 1 per year) is disproportionate to the producer biomass that can be supported, compared with eutrophic lakes (2 /10 MJ m 1 per year: Salmonsen, 1992). This fundamental state change in the basic functioning of ecosystems hinges upon the relative narrow boundary (0.1 / 0.5 mg C l1) separating their carrying capacities (Reynolds, 2001). Where, in fact, zooplankton is the only relevant food for fish (in the open waters of large lakes and seas), the resource is demonstrably a poor one. The biomass of filter-feeders that it is possible to sustain is approximately that which will not deplete the water of primary producers faster than they can recruit themselves. Thus, supposing a phytoplankton able to complete a cell-doubling per day, zooplankton must not remove algae from more than 500 ml l11 per day. This is within the capability of ten large or 100 small Daphnia per l of water or, using the conversions of Bottrell et al. (1976), something in the order of B/1 mg l 1 dry weight. To a planktivorous fish, this represents a food source worth not more than 20 J l 1 (from Cummins and Wuychek, 1971). Against the growth-saturating requirements of brown trout (Salmo trutta ; Elliott and Hurley, 1999: 330 J g1 freshweight for a 250 g fish; 570 for an 11 g juvenile), however, the smaller fish would have to harvest every Daphnia in 300 l of water, the larger fish would need to do the same in 4 m3. Of course, the minimum rations needed to keep the fish alive would be as much as order of magnitude lower and to feed on calanoids might be more nutritious but the calculations illustrate how unpromising is planktivory as a principal energy source. In order to gain enough nutrition at zooplankton concen-

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

trations ten or more times dilute, fish must be specially adapted to cover large distances and to strain the zooplankton from upwards of 40 m3 water passed across the gill rakers each day (as, indeed, pelagic clupeoids and scombroids are). Otherwise, supposing the option is open to them, most fish will opt to feed inshore, mainly on benthic macroinvertebrates. The cost to the 250 g fish of harvesting its 80 kJ or 4 g dry weight, requirement is much more readily covered by browsing on Gammarus or larval insects than by scooping up zooplankton. In this way, the feeding behaviour of many species of fish in small-to-medium lakes is directed principally to benthivory in the littoral and sublittoral margins. The behaviour represents a further case of optimality. This is not to say that planktivory is unimportant in large lakes, where zooplankton may be the only substantial food resource as indicated, and where species of salmonid, coregonid and stolothrissid exploit it successfully. Moreover, the larval stages of many species of fish are exclusively planktivorous. Hatched in the order of a few mm, they feed preferentially on zooplankton. Generally, they are recruited in large numbers to the pelagic, often coinciding with the maximum recruitment phases of zooplankton. Survivorship is poor but individual growth is rapid: in temperate lakes, the rates of growth and mortality experienced by the young-of-theyear recruits most underpin the seasonality and peak development of planktivory and of over exploitation of the zooplankton (Mills and Forney, 1983; Scheffer et al., 2000). As the resource is depleted, an elective switch to other, larger prey, in the littoral and sublittoral benthos is signalled. The threshold for this switch in foraging is likely triggered somewhere in the range 5 /10 J l1, perhaps representing some 25/50 small Daphnia per l. It is structurally significant in both directions, in that the concentrations of zooplankton required to control fast-growing phytoplankton are sustainable only if planktivorous fish are scarce or absent (say B/10 g wet weight m 3) or if the zooplankton is protected from predation, in the kinds of littoral refugia provided by macrophytes and where larger prey organisms will occupy the predators’ attention (Irvine et al.,


1990, 1991; Søndergaard and Moss, 1998). In all cases, the sustainability of cladoceran filter feeders remains dependent upon the simultaneity of the minimal threshold of filterable algal, bacterial and detrital particles of ]/ 0.1 g C m 3. This may be fulfilled frequently (Kamjunke et al., 1999) but, away from shallow margins, the scaling difficulties of striking and holding a steady state are strongly apparent. It is clear that among smaller and more productive lakes, threshold behaviour determines a greater role of benthivory in the overall energy flow through the system as a whole and for the ecology of the entire lake to be dominated by the behaviour of its large, motile fish species that may select where and how they will browse. These effects have been extensively investigated and debated by limnetic ecologists but hardly at all at the ecosystemic level. They do impinge on some principles germane to the basis of ecosystem elaboration and organisation. First, there is a general pattern to the types of fish that are likely to be encountered in particular lakes, determined by their optimal specialisms, and the capacity of the habitat to satisfy them. Hydraulics, depth, site metabolism, deep-water redox levels set more important boundaries than trophic state to the survival of fish species and their potential dominance, although special conditions applying to the habitat suitability for egg laying and development, as well as larval survivorship has a more conspicuous interaction with productivity. Nevertheless, representation in fish communities of lakes, by species and, certainly, by their adaptive abilities, is strongly related to the habitat conditions obtaining in lakes and rivers (Lamouroux et al., 1999; James et al., 2000; Scheffer et al., 2000; Irvine et al., 2001). The second point is that there are few well-studied fishless systems; most of our understanding of plankton dynamics has been based on studies on communities from which the influence of predation has not been satisfactorily excluded. Yet the striking impacts of fish involvement in plankton dynamics tend to be short-term, oscillating between lurching phases of relative overexploitation and underexploitation (Scheffer et al., 2000): filterable phytoplankton must exceed threshold concentrations to sustain filter-feeding zooplankton but may collapse under sufficiently


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

intense aggregate zooplankton filtration; filterfeeding zooplankton increase subject to the satisfaction of nutritive thresholds but populations are liable to collapse at threshold levels of fish predation. When and whether these interactions result in phases of clear water or high algal turbidity is related to the seasonally varying processing rates and the top-down effects of the reproductive cycles of fish (Scheffer and Rinaldi, 2000). Smoothing and stabilisation of the dynamic components requires the intervention of environmental patchiness, provided in large lakes by dynamic heterogeneity and in smaller ones by distinct types of littoral refuge. 4.4. Diversity and ecosystem function The outstanding topic of concern, not just to ecologists but also to the global-scale political decision makers, is the importance of ecological diversity to ecosystem function. While nobody seems opposed to resisting species extinctions and to protecting the gene pools they represent from being irreversibly lost, there has been extensive debate on how crucial is diversity to the ways in which ecosystems work. Views continue to range between two extremes. One approximates to the ‘rivet model’ of Ehrlich and Ehrlich (1981) where everything contributes to the function of the whole and that, like the rivots holding the skin of an aircraft to its frame, every one lost weakens the ability of the whole to keep flying. At the opposite pole, provided that the essential functional rules are fulfilled (primary producers linked to a food web, including decomposers, with resources of water, energy and nutrients, including a carbon source), it is perfectly possible for very species poor systems to operate efficiently and continuously. Most ecologists accept that ecosystem function is not equally distributed among the components, some (‘key stone species’, ‘engineers’) being rather more important than others, where there is considerable functional redundancy (Walker, 1992; Lawton and Brown, 1993). Thus, the more generally preferred model adopts a looser view, suggesting that the basic function is modified by changes in the richness of species composition but in ways which are unpredictable because the

contributions of individual species to system function are unequal and idiosyncratic (Lawton, 1994). It is abundantly clear, from many examples of shallow, weakly-flushed, nutrient-enriched bodies of freshwater that they do not have to have many species to make them function, very efficiently and in a ‘depressingly’ sustainable manner (Reynolds, 2000). These experience high levels of organic fermentation (principally of leaf fall inputs), intense in situ photosynthetic primary production is contributed largely by copious populations of Planktothrix agardhii . Microbial breakdown of the organic matter mediates intense nutrient renewal and recycling, while the food web may comprise little beyond the larvae of Chironomus plumosus or C. anthracinus and benthic-foraging carp, such as Cyprinus carpio. Humans may consider this unsavoury and unhealthy (there is little evidence that other organisms, a crowning flock of Branta canadensis notwithstanding, find this an attractive habitat either). It is difficult to manage the site away from this kind of species structure, without applying some fairly drastic environmental engineering. Such examples may persuade us that enrichment is bad for diversity. Actually, their fault is that function is extremely constrained by processing potential (light, oxidative breakdown of organic matter) and, so long as the extremes persist, the tendency for the species able to maintain the highest exergy will, progressively, exclude the weaker competitors as they do so. In much the same way, systems suffering intensifying stress (say, from chronic, unrelieved nutrient deficiency or from increasing acidity) also tend towards low species representation (Johnson, 1975; Lane and Burris, 1981). So, is diversity dependent upon being somewhere between the extremes? Certainly, this is an idea that has gained support among plant ecologists. According to recent reviews (Schlapfer and Schmid, 1999; Waide et al., 1999), plant productivity is generally beneficial to maintaining high species richness with good resource use and retention, although precise effects are difficult to predict without reference to specific local variables or to trophic impacts. Some idea of the contribution that species richness and/or diversity to

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

communal function has come from experiment. Wilsey and Potvin (2000) reduced the number of dominant plants from old-field communities, though without reducing the overall richness, and found that biomass increased in proportion to the evenness, independently of which had been dominant previously. In the experiments of Wardle et al. (1999) plants of the most aggressive of the grass species, Lolium perenne , were removed altogether from a perennial meadow in New Zealand. There followed an increase in biomass of the species remaining and richness was increased by germlings of invading species. For a time, at least, a broadly similar function was maintained in the grassland. When they removed all plants, however, there were immediate repercussions in other community components, most notably, among the nematode consumers of plant material and their (predominantly) nematode predators. There was little impact on the activity of the longer-term decomposers. Wardle et al. deduced that deliberate exclusion of a (high-exergy) dominant stimulates a recovery in which the function of the dominant is assumed by the nextfittest component(s) available from the same functional group (a precisely analogous response among phytoplankton has been modelled by Elliott et al., 2001), but if the role went unfulfilled, the ecosystem function quickly breaks down as the impact is experienced by functional dependants elsewhere in the trophic network. Further evidence of the functional flexibility and alternative energy-flow routings that species diversity confers in non-extreme systems has been given by Dodson et al. (2000) From a literature survey of 33 well-studied lakes, they were able to demonstrate a pattern of species richness corresponding to a flattish but unimodal function of overall areaspecific primary production, peaking in the range 30 /300 g C m 2 per year (which range covers moderately oligothrophic to moderately eutrophic lakes). When they came to look at the impacts of a series of experimental enrichments, there was little evidence of such a pattern: varied and unpredictable outcomes persisted depending upon the time lapsed since initiation. Evidently, time is needed for quasi-steady states in the relationship of


species richness to productivity to establish at the community level. It remains to be confirmed, but it is probable that net productivity, potential biomass increase and a non-excluded species structure are all indicative of current underuse of ecosystem capacity, much as anticipated in the models of ecosystem expansion included in Fig. 1. To fill the present carrying capacity is to set-up reciprocal resource stress or to drain one of the processing fluxes to the limit of its supply and, thus, progressively to squeeze the diversity towards competitive exclusion. Then, diversity might be seen as the consequence of environmental variability and the springboard of positive biomass responses at such times when capacity is underfilled and at least main respondents are not exposed to destructive competition for resources or for processing opportunities. Biodiversity is maintained because a combination of variable forces-disturbances, gaps, mobility between patches-compound a tendency for the rarer, subdominant species not to be excluded (Levin, 2000). As pointed out earlier (Section 4.1), diversity and species richness are scale sensitive, it being easier to extinguish species at the habitat scale than at the regional scale. The issue of fragmentation and the increasing separation of favourable patches is crucial to mesoscale and regional survival as the opportunities for isolated populations to re-establish in rejuvenated patches are countered by rarity and/or distance. The risk of abrupt extinctions of the species at these larger scales is increased as the combination of forces is weakened, though the responses may be significantly delayed by local within-patch persistence (see Casagrandi and Gatto, 1999). If the functional roles are fulfilled, the ecosystem as a whole may continue to function adequately, at the highest levels of exergy that replacement species can maintain. However, loss of local species richness in increasing isolation from re-invasions from other localities must be supposed to lead eventually to impoverished function with rather low exergy. At least the mechanism of fragmentation species losses is described from quantitative measurements of species richness in diminishing and increasingly isolated areas of North American


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

tallgrass prairie (Bascompte and Rodriguez, 2001). The anthropogenically-determined replacement communities presumably function as intended but not necessarily so well as the natural vegetation and with little prospect of a return, once the regional landscape diversity has been structurally modified to a point where its structure can never be regained.

5. Conclusions The intent of this dissertation has been to compound the case that a robust theory of ecosystem structure, function and thermodynamics is clearly now developing and that its relationships with the established ecological approaches are becoming strongly identified. Ecological patterns and ecosystem theory are surely merging together and to the benefit of ecological understanding in either direction. The trend is probably not altogether new, as it has had several strong advocates for a number of years. Many ecologists may be now willing to endorse a view expressed by Harris (1999): that a really quite small number of physiological and functional properties are required to explain the emergence and basic working of a ‘very complex world’ ecosystem. His paper makes explicit reference to the idea of environmental capacity, to its filling, and to the part played by environmental perturbations in promoting biodiversity and the value of viewing it as a series of functional groups. The developing overview accepts and accommodates the underpinning theories of competition and competitive exclusion. The real depressant to species richness is the functional specialism required to deal with spatial and temporal monotony of a characteristic stress (say, water shortage or nitrogen deficiency) or high frequency interruptions of processing opportunities (temporary habitats, soil disturbance, truncated photoperiods in conditions of turbidity). Intermediate resource levels and intermediate processing opportunities permit more species to assemble communities and to diversify the community structure. Successful components must satisfy certain criteria of presence and suitability, beyond which the habitat is

likely, to a greater or lesser extent and a with variable temporal intensity to ‘filter’ the fittest survivors. At times, filtration may intensify, in favour of species most able to cope with the stress or the physical disturbance, or conceivably, the challenge provided by persistent renewal of ideal conditions. In the terminology of Grime (1979), we recognise the interplay of primary evolutionary specialisms (adaptive strategies) shown among various organisms to exploit otherwise adverse habitats */what he labelled, (respectively) as S (stress-tolerant), R (ruderal) and C (for ‘competitor’). Conversely, much ecology is prosecuted in habitats subject to sufficient cyclical or stochastic environmental variability and to be endowed with an attractive species richness that the subject is really the study of how heterogeneity foils specialist organisms. Variability is exploited at every scale, by differential life forms, physiological flexibilities and modified living plans, that together represent the diversification of roles, niche preferences and life histories of the majority of the species living on the planet, now or in the past. Species opting for the more exploitative r-selected life histories are ‘doomed to lose in competition on small spatial scales [to more acquisitive K -selected species] . . .but survive on larger spatio-temporal scales by their ability to find new patches of vacated habitat more effectively than the dominant competitors’ (Levin, 2000). The linkages to higher-order ecosystem behaviour arise from the responses of organisms acquiring resources from finite patches of space that themselves integrate into larger geographical patterns and how the requirements and gleaning abilities of species permit the habitat space to be shared (Ritchie and Olff, 1999). Raising the scale of focus one stage further, the broad function of ecosystems appears, to be further stabilised and averaged out by considerable internal dynamism and feedback. This is consistent with the ecosystem characteristic of resilient, multiply-mediated homeostatic capability (Jørgensen, 1997). Given sufficient weakening or with the application of sufficient force, however, this capacity is liable to exceedence, leaving the system vulnerable to disruption and organisational

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

change. Indeed, once the resilience is overcome, structural changes can be both severe and abrupt, the system ‘flipping’ to a more minimalist, more primitive, yet more sustainable organisation, upon which a renewed expansion might be based. These flip-points, famously represented in May (1977) ball-and-cup analogy, have been quantified and partly verified for pelagic communities (Reynolds, 1997b). At every level of ecological function, the sameprinciples of resource processing into structures fuelled by the controlled dissipation of energy and moving further from thermodynamic equilibrium is poised against the chaos of a continuously variable non-living environment. It is the strength of this pattern, repeated at every hierarchical scale from organelle to organism to organisation, that does most to consolidate the emergence of analogous patterns in the disciplines spanning physiological ecology through to ecosystem science. The penultimate of work that I cite to illustrate the extent of the philosophical linkages among environment, ecological function and self-organisation at the ecosystem level is the recent study of Kutsch et al. (2001). Using field observations made on two adjacent but contrasted terrestrial systems (beech forest, crop field), they devised a series of ecological indicators (areal biomass, fractioned among plant organs, microbes, animals) and of extrabiotic carbon. These they compared with the balances of carbon flux through the systems, in order to establish quantitative measures of ecosystem efficiency, accumulation and productive yield from the incoming solar energy. The approach adopts a model of a nested organisation balancing the influences of capacity and metabolic constraints on the structure of the system. It is perhaps premature to judge how well this ecological integration works or is applicable elsewhere but the attempt is laudable and, in the context of this essay, warmly welcomed for its confirmation of the duality of the linkage: ecosystems are the sum of component ecological functions; ecology functions exclusively within ecosystemic principles. Moreover, it exemplifies the kind of useful interfaces with landscape, biogeochemistry, site history and spatial-temporal


heterogeneity that ecosystemic approaches can bring to ecology. The final point is that we have a basis for judging function by what the main players do rather than simply what they are. Their actual identities are influenced by how well their capabilities fit with what the environmental fluxes and the resource base will allow. In the end, they will be distinguished by their accumulated biomass and their turnover of the carbon harvest. The balance that they strike returns us to the concept of exergy, to their position on the plot in Fig. 1a and to the scope for an ascendant response or for a contraction to achieve an ecologically sustainable position with respect to stochastic environmental variability. The persistent need for adjustment pervades every ecological scale because, at every level, there is the persistent force of changeability. Truly, change is nature’s only constant (I.P.Hosein, see below). I am proud to append the poem by my late friend, Ismet Pasha Hosein, from which this clause is drawn. Provocative and deeply moving, his words capture eloquently the paradox of life and death. He wrote them just before he died, by wretched irony, at the very time I was writing this article. I reproduce the entire poem by with the generous permission of Ismet’s wife and daughters. ‘‘One moment, single, solitary, then it is gone Years of hardship, years of toil; no more! A universe ended, once begun. Life is but transient. Whither goest? Should we but live, surely death as darkness does befall Would come. What does nature crave that, within her breast to take, Must make Man live? And, still unsatisfied, must make Man die? Is there a choice in such a choice? Is there a choice in such a choice? End? No end! For end is but a change. Change.


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

Nature’s only constant.’’ -I.P.Hosein (1930 /2001)

References Allen, T.F.H., Starr, T.B., 1982. Hierarchy: Perspectives for Ecological Complexity. University of Chicago Press, Chicago. Azam, F., Fenchel, T., Field, J.G., Gray, J.S., Meyer-Reil, A., Thingstad, F., 1983. The ecological role of water-column microbes in the sea. Marine Ecology-Progress Series 10, 257 /263. Bascompte, J., Rodriguez, M.A., 2001. Habitat patchiness and plant species richness. Ecology Letters 4, 417 /420. Belyea, L.R., Lancaster, J., 1999. Assembly rule, within a contingent ecology. Oikos 86, 402 /416. Bottrell, H.H., Duncan, A., Gliwiez, Z.M., Grygierk, E., Herzig, A., Hillbricht-Ilkowska, A., Kurusawa, H., Larsson, P., Weglenska, T., 1976. A review of some problems in zooplankton production studies. Norwegian Journal of Zoology 24, 419 /456. Brooks, J.L., Dodson, S.I., 1965. Predation, body size and composition of the plankton. Science 150, 28 /35. Brown, J.H., 1999. Macroecology: progress and prospect. Oikos 87, 3 /14. Burns, C.W., 1968. The relationship between body size of filterfeeding Cladocera and the maximum size of particle ingested. Limnology and Oceanography 13, 675 /678. Burns, C.W., 1969. Relation between filtering rate, temperature and body size in four species of Daphnia . Limnology and Oceanography 14, 693 /700. Casagrandi, R., Gatto, M., 1999. A mesoscale approach to extinction risk in fragmented habitats. Nature 400, 560 / 562. Connell, J.H., 1978. Diversity n tropical rain forests and coral reefs. Science 199, 1302 /3310. Cummins, K.W., Wuychek, J.C., 1971. Caloric equivalents for investigations in ecological energetics Mitteilungen der internationale Uereinigung fu¨r theoretische und angewandte. Limnologie 18, 1 /158. Deutschman, D.H., Lenin, S.A., Pacala, S.W., 1999. Error propagation in a forest succession model: the role of finescale heterogeneity in light. Ecology 80, 1927 /1943. Dodson, S.I., Arnott, S.E., Cottingham, K.L., 2000. The relationship in lake communities between primary productivity and species richness. Ecology 81, 2662 /2679. Donalson, D.D., Nisbet, R.M., 1999. Population dynamics and spatial scale: effects of system size on population persistence. Ecology 80, 2492 /2507. Dudgeon, S.R., Steneck, R.S., Davison, I.R., Vadas, R.L., 1999. Coexistence of similar species in a space-limited intertidal zone. Ecological Monographs 69, 331 /352.

Ehrlich, P.R., Ehrlich, A.H., 1981. Extinction. The causes and consequences of the disappearance of species. Random House Publishers, New York. Elliott, J.M., Hurley, M.A., 1999. A new energetics model for brown trout, Salmo trutta . Freshwater Biology 42, 235 / 246. Elliott, J.M., Hurley, M.A., Elliott, J.A., 1997. Variable effects of droughts on the density of a sea-trout Salmo trutta population over 30 years. Journal of applied Biology 34, 1229 /1238. Elliott, J.A., Reynolds, C.S., Irish, A.E., 2001. An investigation of dominance in phytoplankton using the PROTECH model. Freshwater Biology 46, 99 /108. Finlay, B.J., Clarke, K.J., 1999. Ubiquitous dispersal of microbial species. Nature 400, 828. Finlay, B.J., Esteban, G.F., Fenchel, T., 1998. Protozoan diversity: converging estimates of the global number of free-living ciliate species. Protist 149, 29 /37. Fischer, A.G., 1960. Latitudinal variation in organic diversity. Evolution. 14, 64 /81. Godfray, H.C.J., Lawton, J.H., 2001. Scale and species number. Trends in Ecology and Evolution 16, 400 /403. Grime, J.P., 1979. Plant Strategies and Vegetation Processes. Wiley, Chichester. Grime, J.P., 1987. Dominant and subordinate components of plant communities: implications for succession, stability and diversity. In: Gray, A., Edwards, P., Crawley, M. (Eds.), Colonization, Succession and Stability. Bladtwell Scientific Publications, Oxford, pp. 413 /428. Gulland, J.A., 1983. Fish Stock Assessment. Wiley, Chichester. Hardin, G., 1960. The competitive exclusion hypothesis. Science 131, 1292 /1297. Harris, G., 1999. This is not the end of limnology (or of science): the world may be a lot simpler than we think. Freshwater Biology 42, 689 /706. Hart, R.C., 1996. Naupliar and copepodite growth and survival of two freshwater calanoids at various food levels: demographic contrasts, similarities and food needs. Limnology and Oceanography 41, 648 /658. Huisman, J., 1999. Population dynamics of light-limited phytoplankton: microcosm experiments. Ecology 80, 202 / 210. Huston, M., 1979. A general hypothesis of species diversity. American Naturalist 113, 81 /101. Huston, M., 1994. Biological Diversity: the Coexistence of Species on Changing Landscapes. Cambridge University Press, Cambridge. Irvine, K., Moss, B., Stansfield, J.H., 1990. The potential of artificial refugia for maintaining a community of largebodied Cladocera against fish predation in a shallow eutrophic lake. Hydrobiologia 200/201, 379 /389. Irvine, K., Stansfield, J.H., Moss, B., 1991. The use of enclosures to demonstrate the enhancement of Daphnia populations when isolated from fish predation in a shallow eutrophic lake. Memorie dell Istituto italiano di Idrobiologia 48, 325 /344.

C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200 Irvine, K., Patterson, G., Allison, E.H., Thompson, A.B., Menz, A., 2001. The pelagic ecosystem of Lake Malawi, Africa: trophic structure and current threats. In: Munawar, M., Hecky, R.E. (Eds.), The Great Lakes of the World (GLOW): Food Web, Health and Integrity. Backhuys, Leiden, pp. 3 /30. James, M.R., Hawes, I., Weatherhead, M., Stanger, C., Gibbs, M., 2000. Carbon flow in the littoral food web of an oigotrophic lake. Hydrobiologia 441, 93 /106. Jenkins, T.M., Dichl, S., Kratz, K.W., Cooper, S.D., 1999. Effects of population density on individual growth of brown trout in streams. Ecology 80, 941 /956. Johnson, L., 1975. Physical and chemical characteristics of Great Bear Lake. Journal of the Fisheries Research Board of Canada 32, 1971 /1987. Jørgensen, S.E., 1992. Integration of Ecosystem Theory: a Pattern. Kluwer, Dordrecht. Jørgensen, S.E., 1997. Integration of Ecosystem Theory: a Pattern, second ed.. Kluwer Academic Publishers, Dordrecht. Jørgensen, S.E., Marques, J.C., 2001. Thermodynamics and ecosystem theory: case studies from hydrobiology. Hydrobiologia 445, 1 /10. Kalff, J., Watson, S., 1986. Phytoplankton and its dynamics, in two tropical lakes: a tropical and temperate zone comparison. Hydrobiologia 138, 161 /176. Kamjunke, N., Benndorf, A., Wilbert, C., Opitz, M., Kranich, J., Bollenbach, M., Benndorf, J., 1999. Bacteria ingestion by Daphnia galeata in a biomanipulated reservoir: a mechanism stabilizing biomanipulation. Hydrobiologia 403, 109 / 121. Keddy, P.A., 1992. Assembly and response rules: two goals for predictive community ecology. Journal of Vegetation Science 3, 157 /164. Kelt, D.A., Taper, M.L., Meserve, P.L., 1995. Assessing the impact of competition in community assembly. Ecology 76, 1283 /1296. Kutsch, W.L., Steinburn, W., Herbst, M., Baumann, R., Barkmann, J., Kappen, L., 2001. Environmental indication: a field test of an ecosystem approach to quantify biological self-organization. Ecosystems 4, 49 /66. Lamouroux, N., Olivier, J.M., Persat, H., Pouilly, M., Souchon, Y., Slatzner, B., 1999. Predicting community characteristics from habitat conditions: fluvial fish and hydraulics. Freshwater Biology 42, 275 /299. Lampert, W., 1977. Studies on the carbon balance of Daphnia pulex de Geer as related to environmental conditions. IV. Determination of the threshold concentration as a factor controlling the abundance of zooplankton species. Archiv fu¨r Hydrobiologie (Supplementband) 48, 361 /368. Lane, A.E., Burris, J.H., 1981. Effects of environmental pH on the internal pH of Chlorella pyrenoidosa, Scenedesmus quadricauda and Euglena mutabilis. Plant Physiology 68, 439 /442. Lawton, J.H., 1994. What do species do in ecosystems. Oikos 71, 367 /374.


Lawton, J.H., 1999. Are there general laws in ecology. Oikos 84, 177 /192. Lawton, J.H., 2000. Community Ecology in a Changing World. Ecology Institute, Oldendorf. Lawton, J.H., Brown, V.K., 1993. Redundancy in ecosystems. In: Schulze, E.D., Mooney, H.A. (Eds.), Biodiversity and Ecosystem Function. Springer, Berlin, pp. 155 /270. Lemhan, H.S., 1999. Physical /biological coupling on oyster reefs: how habital structure influences individual performance. Ecological Monographs 69, 251 /275. Levin, S.A., 2000. Multiple scales and the maintenance of biodiversity. Ecosystems 3, 498 /506. Loreau, M., 2000. Are communities saturated? On the relationship between a, b and g diversity. Ecology Letters 3, 73 /76. Lyons, S.K., Willig, M.R., 1999. A hemispheric assessment of scale dependence in latitudinal gradients of species richness. Ecology 80, 2483 /2491. May, R.M., 1977. Thresholds and breakpoints in ecosystems with a multiplicity of stable states. Nature 269, 471 /477. McConnaughay, K.D.M., Coleman, J.S., 1999. Biomass allocation in plants: ontogeny or optimality? A test along three resource gradients. Ecology 80, 2581 /2593. Mejer, H., Jørgensen, S.E., 1979. Exergy and ecological buffer capacity. In: Jørgensen, S.E. (Ed.), State-of-the-Art in Ecological Modelling, vol. 7. International Society for ecological Modelling, København, pp. 829 /846. Mills, E.L., Forney, J.L., 1983. Impact on Daphnia pulex of predation by young yellow perch, Perca flavescens , in Oneida Lake, New York. Transactions of the American Fisheries Society 112, 154 /161. Morin, P., 1999. Productivity, intraguild predation and population dynamics in experimental food webs. Ecology 80, 752 / 760. Morris, D.W., Davidson, D.L., 2000. Optimally foraging mice match patch use with habitat differences in fitness. Ecology 81, 2061 /2066. Odum, E.P., 1969. The strategy of ecosystem development. Science 164, 262 /270. Pahl-Wostl, C., 1995. The Dynamic Nature of Ecosystems. Wiley, Chichester. Paine, R.T., 1980. Food webs: linkage, interaction strength and community infrastructure. Journal of Animal Ecology 49, 667 /685. Reynolds, C.S., 1989. Physical determinants of phytoplankton succession. In: Sommer, U. (Ed.), Plankton Ecology. Brock /Spring, Madison, pp. 9 /56. Reynolds, C.S., 1997. Vegetation Processes in the Pelagic: A Model For Ecosystem Theory (Excellence in Ecology Series, 9). Ecology Institute, Oldendorf. Reynolds, C.S., 1997. Successional development, energetics and diversity in planktonic communities. In: Abe, T., Levin, S.R., Higashi, M. (Eds.), Ecological Perspectives of Biodiversity. Springer, New York, pp. 167 /202. Reynolds, C.S., 1998. The state of freshwater ecology. Freshwater Biology 39, 741 /753. Reynolds, C.S., 2000. Defining sustainability in aquatic systems: a thermodynamic approach. Verhandlungen der


C.S. Reynolds / Ecological Modelling 158 (2002) 181 /200

internationale Vereinigung fur theoretische und angewandte Limnologie 27, 107 /117. Reynolds, C.S., 2001. Emergence in pelagic communities. Scientia Marina 65 (Suppl. 2), 5 /30. Reynolds, C.S., 2002 Planktic community assembly in flowing water and the ecosystem health of rivers. To be published, in Ecological Modelling (‘Namur’). Reynolds, C.S., Desey, J.P., 1996. The production, biomass and structure of phytoplankton in large rivers. Archiv fu¨r Hydrobiologie (Supplementband) 113, 161 /187. Reynolds, C.S., Davies, P.S., 2001. Sources and bioavailability of phosphorus fractions in freshwaters: a British perspective. Biological Reviews 76, 27 /64. Reynolds, C.S., Elliott, J.A., 2002. Phytoplankton diversity: discontinuous assembly responses to environmental forcing. Verhandlungen der internationale Vereinigung fur theoretische und angewandte Limnologie 28, 336 /344. Reynolds, C.S., Dokulil, M., Padisak, J., 2000. Understanding the assembly of phytoplankton in relation to the trophic spectrum; where are we now. Hydrobiologia 424, 147 /152. Reynolds, C.S., Huszar, V.L.M., Kruk, C., Naselli-Flores, L., de Melo, S., 2002. Towards a functional classification of the freshwater phytoplankton. Journal of Plankton Research 24, 417 /428. Ricker, W.E., 1954. Stock and recruitment. Journal of the Fisheries Research Board of Canada 20, 257 /284. Ritchie, M.E., Olff, H., 1999. Spatial scaling laws yield a synthetic theory of biodiversity. Nature 400, 557 /560. Rohde, K., 1992. The latitudinal gradient in species diversity; the search for the primary cause. Oikos 65, 5l4 /527. Rojo, C., Ortega-Mayagoitia, E., Alvarez Cobelas, M., 2000. Lack of pattern among phytoplankton assemblages or what does the exception to the rule mean. Hydrobiologia 424, 133 /140. Romanovsky, Y.H., 1985. Food limitation and life-history strategies in cladoceran crustaceans. Ergebnisse der Limnologie 2l, 363 /372. Salmonsen, J., 1992. Examination of the properties of exergy, power and ascendancy along a eutrophication gradient. Ecological Modelling 62, 171 /181. Saunders, P.A., Porter, K.G., Taylor, B.E., 1999. Population dynamics of Daphnia spp. and implications for trophic interactions in a small, monomictic lake. Journal of Plankton Research 21, 1823 /1845. Scheffer, N., Rinaldi, S., 2000. Minimal models of top /down control of phytoplankton. Freshwater Biology 45, 265 /283.

Scheffer, N., Rinaldi, S., Kuznetsov, Y.A., 2000. Effects of fish on plankton dynamics: a theoretical analysis. Candian Journal of Fisheries and Aquatic Sciences 57, 1208 /1219. Schlapfer, F., Schmid, B., 1999. Ecosystem effects of biodiversity: a classification of hypotheses and exploration of empirical results. Ecological Applications 9, 893 /912. Schoener, T.W., 1986. Overview: kinds of ecological communities /ecology becomes pluralistic. In: Diamond, J., Case, T.J. (Eds.), Community Ecology. Harper and Row, New York, pp. 556 /586. Søndergaard, M., Moss, B., 1998. Impact of submerged macrophytes on phytoplankton in shallow lakes. In: Jeppesen, E., Søndergaard, M., Christoffersen, K. (Eds.), The Structuring Role of Submerged Macrophytes in Lakes. Springer, New York, pp. 115 /132. Southwood, T.R.E., 1977. Habitat, the templet for ecological strategies. Journal of Animal Ecology 46, 337 /365. Stevens, G.C., 1989. The latitudinal gradient in geographical range: how so many species coexist in the tropics. American Naturalist 133, 240 /256. Strasˇkraba, M., 1980. Cybernetic categories of ecosystem dynamics. Journal of the International Society for Ecological Modelling 2, 81 /96. Strasˇkraba, M., Jørgensen, S.E., Patten, B.C., 1999. Ecosystems emerging:2. Dissipation. Ecological Modelling 117, 3 /39. Taylor, P.H., Gaines, S.D., 1999. Can Rapoport’s rule be rescued? Modelling causes of the latitudinal gradient in species richness. Ecology 80, 2474 /2482. Ulanowicz, R.E., 1986. Growth and Development */Ecosystems Phenomenonology. Springer, New York. Waide, R.B., Willig, M.R., Steiner, C.F., Mittelbach, G., Gough, L., Dodson, S.L., Juday, G.P., Parmenter, R., 1999. The relationship between productivity and species richness. Annual Reviews in Ecology and Systematics 30, 257 /300. Walker, B.H., 1992. Biodiversity and ecological redundancy. Biological Conservation 6, 18 /23. Wardle, D.A., Bonner, K.I., Barker, G.M., Yeates, G.W., Nicholson, K.S., Bardgett, R.D., Watson, R.N., Ghani, A., 1999. Plant removals in perennial grassland: vegetation dynamics, decomposers, soil biodiversity and ecosystem properties. Ecological Monographs 69, 535 /568. Weiher, E., Keddy, P.A., 1995. Assembly rules, null models and trait dispersion: new questions from old patterns. Oikos 74, 159 /164. Wilsey, B.J., Potvin, C., 2000. Biodiversity and ecosystem functioning importance of species evenness in an old field. Ecology 81, 887 /892.