One of the foremost objectives in re-circulating aquaculture is to maintain system stability. This is not wholly unreasonable in light of the fact that 1) our closed systems can be inherently prone to sudden and devastating disturbances (i.e., “crashes”), and 2) a great many aquatic organisms depend upon a certain degree of environmental stability for their survival and reproduction (Riddle, 2009). One may further suppose that long-term stability, in allowing for full ecological succession, promotes species richness; however, evidence from a growing body of ecological field investigation continues to suggest that this is not always so.

Until the early 1960’s, most experimental evidence of interspecific competition was derived from the observation of laboratory animal populations under more or less static laboratory conditions. Where the indirect influence of competitive interaction among natural communities was not self-evident, the role of competition in community organization was customarily overlooked. By the early 1970’s, numerous researchers in the rapidly expanding field of community ecology (which emphasizes structural characteristics and relies heavily upon interpretation of field data) were eager to demonstrate that biological communities are not merely isolated clusters of populations, but highly dynamic, interconnected systems with multiple levels of order.

A characteristically dense aggregation of sea palms (Postelsia palmaeformis) on a newly exposed rock face. Photo courtesy of Mendocino Coast Botanical Gardens.

A considerable breakthrough was made on their behalf with the publication of Paul K. Dayton’s highly influential demonstration (1971) of the competitive interactions that occur among sessile marine organisms following major disturbance events. This important work served, and continues to serve, as the basis for numerous other researchers’ efforts to develop a general theory of disturbance (Sebens 1984).

In 1) reviewing a selection of studies that were influential in the development of ecological disturbance theory, and 2) describing marine hard-bottom community structures in the context in which they served as experimental and theoretical subjects, this article seeks to call attention to the implied (albeit strong) relevance of disturbance theory in the culture of marine ornamental species. Though the investigations herein described were conducted on natural populations, the models are intended to be universally applicable; hence, they are theoretically analogous to the competitive interactions of sessile organisms in all disturbed hard-bottom communities–whether between macroalgae and juvenile farm clams on a freshly-scoured net pen, between tropical sponges and corals after a sustained influx of sediment over a reef, or between microorganisms on the human tooth after a good flossing.

1. Introduction

To avoid wide dispersal, mature sea palm sporophytes such as these only release spores during low tide emersion; spores simply drip from deep grooves in their blades onto the immediate vicinity. Photo by Lyudmila Zinkova.

Temperate rocky intertidal habitats are among the earliest and most thoroughly studied biological communities by field ecologists. It remains firmly established that the lower intertidal zone is typically dominated by flora with a distribution determined most directly by biotic factors (e.g., competition for light, resistance to herbivory), while the upper intertidal zone is typically dominated by sessile fauna with a distribution determined most directly by physical factors (e.g., heat, desiccation, wave exposure, battering by drift logs) (Connell 1961; Dayton 1971; Nielson 2006). Evidence from earlier studies (Connell 1961) convinced Dayton (1971) that disturbance indirectly reduces the intensity of competitive interactions in the upper intertidal zone, and that species richness is greatest in environments subject to intermediate levels of disturbance. In so doing it resulted in the development of a mechanism (Dayton 1971) that seemingly contradicts presumptions (McNaughton and Wolf 1970) that competitive dominance is typified by abundance, that dominance is strongly correlated by broad niches, and that species attain membership in communities by increasing carrying capacity or narrowing niches (or both).

As of late, ecological research in this area usually 1) describes patterns of abundance/distribution of sessile organisms, 2) explains how these bands of distribution arise from the various community members’ responses to various physical pressures, and 3) experimentally demonstrates how abrupt temporal (including seasonal) and spatial discontinuities in distribution arises from interspecific competition (Dayton 1971; Sousa 1985; Blanchette 1996).

Sea stars (Pisaster ochraceus) taking refuge in the shade during a low tide. Evidence of sea star predation can be observed in the nearby mussel bed (Mytilus californicus). Photo by Kenneth Wingerter.

The rocky shores of the Northeast Pacific Ocean provide an ideal setting to study ecological succession and the effects of disturbance regimes; not only are they reasonably accessible, but they are experimentally amenable in that they are geographically unbroken with flora and fauna that have overlapping ranges that are confined to the intertidal zone. These communities are open-ended, as most primary production and decomposition takes place elsewhere (Dayton 1971). Furthermore, a great majority of interspecific competitive interaction involves the struggle to claim space, which is relatively simple to analyze in this environment, owing to a two-dimensional substrate and clearly delineated bands of vertical distribution (Dayton 1971).

Garden-variety models of competition predict that, under static environmental conditions, a single species would effectively dominate in each intertidal zone. That being said, field evidence indeed suggests that certain stochastic disturbance events (particularly those that result in the localized clearing of habitable space) actually serve to promote species richness (Dayton 1971; Blanchette 1996). Identifying factors that influence community succession may be critical in describing the path to dominance under circumstances where habitable space is likely to be a limiting resource (e.g., high-density monoculture). Processes that drive community organization cannot, in these cases, be fully explained without evaluating the disturbance regimes that relieve pressure for space (i.e., allow coexistence).

In Northeastern Pacific littoral habitats, monopolization of space is prevented by both biological and physical disturbance. Biological disturbance results mainly from the activity of limpets, gastropods, and starfish on barnacles, mussels, and kelps. Physical disturbance results mainly from wave action and impacts with floating logs (Dayton 1971). The most important ecological function of these periodic disturbances is the clearing of primary substrata, which may immediately be recolonized. Within this setting, competitive dominants ultimately claim all available space at low disturbance rates, whereas superior colonizers claim all available space at high disturbance rates; coexistence occurs at intermediate disturbance rates. (Sebens 1984).

As theoretical ecologists have moved away from approaches that are preoccupied with equilibrium, the role of spatial heterogeneity in models of population dynamics has received greater attention (Nielson 2006; Sebens 1984). In these models, patches of primary substrata are treated as fundamental structural units; patterns of spatio-temporal heterogeneity emerge with increasing regional scale. Patches are taken to be a “holes” in an otherwise homogeneous reference background. Though bounded and discrete, they are not viewed as closed systems, but rather as individual components of a larger spatio-temporal mosaic. The random colonization and founder effects that follow natural succession regimes guarantee the ongoing integrity of this larger system. At least certain species in these communities depends upon certain levels of local disturbance (Levin and Paine 1974).

2. Typical Field Composition

A markedly different community can be observed only a couple meters away in calmer, landward-facing tidepools; this community is dominated by sea anemones (Anthopleura xanthogrammica), serpulid tube worms, and red algae. Photo by Kenneth Wingerter.

Ecological succession usually leads to dominance by mussels (Mytilus spp.), which tend to displace barnacles (Balanus spp. and Cthamalus spp.) and have a negative influence on the recruitment of sea palms (Postelsia palmaeformis). Mussels quickly claim primary substrata. Subsequent mussel recruitment results in multilayered colonies that become increasingly threatened with disruption by wave action as they grow (Dayton 1973; Sousa 1985). Attacks by starfish (Pisaster ochraceus) and beatings by drift logs disrupt the continuity of mussel beds, thereby increasing their susceptibility to damage from wave shock. Small clearings become enlarged as newly exposed mussels are ripped away at the edges (Dayton 1971; Dayton 1973).

Significant changes in community structure are usually not brought about by either periodic or stochastic episodes alone; it indeed seems that a combination of these effects are necessary (Liddel 2001). Although it is nearly impossible to differentiate between patches cleared by starfish from those cleared by wave shock, recovery is usually more rapid in the lower intertidal zone (where mussels are more susceptible to starfish) than in the upper intertidal zone (where they are more susceptible to drift log impact). Damage from logs tends to be seasonally punctuated by winter storms, while damage from starfish predation occurs steadily.

The extent of direct wave exposure of the site also seems to influence the initial patch size, helping to vary the size and degree of disturbance from patch to patch. Expressed as percent of original cleared substrata, the mean increase in size of initial patch size has been found to range from 24-4,884% (Dayton 1971; Dayton 1973). In recent models, attempts have been made (Liddel 2001) to account for spatio-temporal patch variability on species richness.

Mussels prefer to settle onto barnacle encrustations (as opposed to bare substrata), and may eventually smother a barnacle colony; however, barnacles can invade invader, later resettling onto mussel shell parts. Photo by Darkone.

The race to invade disturbed marine hard-bottom habitat is intense. Following a major disturbance, the activity of limpets foraging on the newly cleared patch creates a minor disturbance by dislodging recently recruited mussels. Thus, by feeding on limpets as a secondary prey item, starfish have a direct positive effect on the growth of its preferred prey. In the total absence of starfish predation, mussels (owing to their higher growth rate) eventually dominate other sessile community members for space (Dayton 1971). Mussels are not only slower recruiters than barnacles, but are less successful at settling on bare rock, preferring instead to settle onto–and eventually outgrow–barnacle encrustations. Barnacles thusly often get first crack at freshly cleared substrata, with Balanus dominating Chtamalus in all but the uppermost intertidal zones, where it succumbs to physical stress. Further barnacle invasion is accelerated as additional recruits are stimulated to settle near conspecifics already attached to the surface; intraspecific competition is rather minimal (Connell 1961).

Field investigation has suggested that sea palms, while preferring to settle on bare rock, are not only weaker dispersers than are barnacles, but have a narrow reproductive season (notably during winter storm season) (Paine 1988). As an annual kelp, Postelsia is a particularly weak disperser (Dayton 1973; Paine 1988; Nielson 2006). Postelsia has a conspicuously aggregated distribution; its growth habit results in structures that are rather clumped, with sporophytes growing on the holdfasts of older individuals (Dayton 1973; Kusumo 2006). Genetic relatedness between individual sea palms within a single cluster has been shown to be strikingly high (Coyer 1997). As with barnacles, intraspecific competition among sea palms appears to be minimal; indeed, the larger a population size is in a given area during year x, the greater are its odds of persistence into year x+1 (Paine 1988).

3. Discussion of Models

Implementing a periodic disturbance regime may help to promote diversity (i.e., discourage dominance) within sessile invertebrate populations, as in this crowded Oceanário de Lisboa coral reef exhibit. Photo by Csörföly D.

All relevant variables of disturbance (including frequency, magnitude, and size) should be accounted for in any disturbance theory that proposes a testable hypothesis and can consistently make useful predictions. Any such predictions should relate directly to the disturbance variables and response parameters (Dayton 1971; Pickett 1985). Response parameters that are most vital to community structure can be determined by experimental field manipulation (removal of predators, etc.) after rates and patterns of succession are identified (Dayton 1971). Due to the uncertainty inherent in the very process of stochastic disturbance itself, the greatest difficulty in developing a general disturbance theory lies in accurately accounting for particular factors that modify responses in certain contexts. The special character of a single disturbance event may be defined by factors such as landscape shape/composition, system structure (i.e., degree of connectedness in relation to the substrate), resource base, life histories of the component populations, and site-specific competitive dynamics (Pickett 1985).

The model presented by Paine and Levin (1974) was significant not only in that it was a direct answer to Dayton’s influential mechanism (1971), but also in that it was held to be applicable to systems outside the realm of marine hard-bottom biotypes; assemblages of patches are treated as one would treat an assemblage of cells, with each distinguished by its age and size. With the development of this input-output model, Paine and Levin sought to relate the distributional properties (particularly with regard to age and size) of disturbed patches to the biological properties of the invading species.

Polyculture of sessile marine invertebrates may demand some understanding of disturbance-facilitated coexistence. Photo by Daben 2000.

Armstrong (1976) pointed out that the Paine-Levin model placed an undue emphasis on patch structure, while trivializing differential abilities of the invaders to settle on and compete for primary space. The “fugitive species model” he developed expressed these differences in an effort to better describe patterns of biodiversity as they emerge in disturbance-facilitated community organization.

Hastings (1980) attempted to build upon earlier models by accounting more realistically for the effect that the number of competing species and their relative abundances have on coexistence. Biodiversity is here simply calculated as a function of percent coverage of substrata. The type of competitive interaction this model is meant to best represent is that for space on a hard surface where space is the limiting resource. It predicts that the type of competition and rates of colonization are far more consequential in the development of community structure than are any specialized interactions between component species. While this model proved to be more objective and, through its treatment of competitive hierarchy, offered more detailed predictions, it was (admittedly) over-simplified; not only did it lack accountability for recruiter priority effect, but it failed to demonstrate how a single competitor can dominate at levels of intermediate disturbance. Furthermore, the effect of disturbance (whether physical or biological) is administered at a fixed rate, acting upon populations of all species equally–a scenario that is unrealistic in both natural and even captive conditions.

Allowing for variable outcomes was a principle concern of Sebens (1984) in the development of an “indeterminate coexistence” model. Here it is acknowledged that there is a certain level of unpredictability in the outcome of competitive interaction between individuals of two different species–that is, species identity alone cannot guarantee the outcome of interspecific competition between individuals. In some competitive encounters, size or stage of settlement may convey a greater advantage than species identity. Sufficient numbers of these “reversals” are capable of locally disrupting competitive hierarchies. These reversals are now known to be rather common during competitive interactions for space in marine hard-bottom habitats (Sebens 1984). In one example, sea palms can gain a foothold in seral communities by settling on and displacing young mussel beds (where they settle on the shells of their competitor, sufficiently increasing the drag of wave action on their unwilling hosts, eventually causing them to be ripped from the rock (Dayton 1971; Dayton 1973).

In this model, non-competitive mortality is the result of physical or biological disturbance where a certain fraction of individuals is removed, while intrinsic mortality rates are considered to be insignificant. N₁, B₁, and D₁ respectively are per patch population size (number), recruitment (birth) rates, and disturbance (patch death) rates for Species 1, while K is a per patch carrying capacity. In a system with two competitors, mortality of individuals of Species 2 from competitive interaction with Species 1 is equal to the recruitment rate of Species 1 (B₁N₁/K per patch) multiplied by the population size of Species 2 (Sebens 1984).

dN₂/dt = B₂N₂ (K-N₁-N₂/K)-D₂N₂-B₁ (N₁N₂/K).

The range of disturbance rates at which coexistence may occur increases with larger difference of recruitment rates among competitors; such is the case where competitive hierarchies are rigid and well defined. This model utilizes additional terms for cases involving encounters where competitive success is indeterminate, probability f_ij, where the term (1-f_ij)N₁ is the approximate number of spaces j does not take from i per unit time (Sebens 1984).

(dN_i/dt)(1/N_i) = B_i(K-N_i-((1-f_ji)N_j)/K)-D_i-(B_jN_j/K)f_ij.

This can be simplified and given as the equation for each competitor.

(dN₁/dt)(1/N₁) = (B₁/K)(K-N₁-N₂+f₂₁N₂)-D₁-(B₂/K)N₂f₁₂.

(dN₂/dt)(1/N₂) = (B₂/K)(K-N₁-N₂+f₁₂N₁)-D₂-(B₁/K)N₁f₂₁.

If D = D₁ = D₂, the resulting two inequalities below define the upper and lower boundaries of D. Coexistence of two competitors is possible within this range.

Only the superior recruiter persists above some maximum difference in growth rates by dominating the other in some fraction of encounters. This advantage is magnified by the larger recruitment rate. For some value of B₁, let B_2max be the maximum value at which coexistence is possible at any rate of disturbance. Then where D_max=0, B_2max=B₁(f₂₁/f₁₂) . This can be interpreted to mean that domination is reversed between competitors with increasing disturbance rates. In the absence of this occurrence, competitors may coexist at any disturbance rate greater than D_min so long as it remains lower than the recruitment rate of the competitive dominant. Where competitors have similar recruitment rates, coexistence is only possible if fluctuation of disturbance rates is minimal. Such reversals of competitive success are credited for delaying competitive exclusion, and may allow the prolonged coexistence of competitors in systems that are subjected to highly stochastic disturbance regimes (particularly where there are large differences in recruitment rates). In applications that evaluate the competitive dynamics of multiple species, a single species may come to dominate a patch at intermediate disturbance levels; this was not a possible outcome in earlier models, though it clearly occurs in nature (as observed in patches dominated by Postelsia) and apparently in aquaria (e.g., Aiptasia plagues). In these cases, the competitive dominant may not necessarily be at the top of the competitive hierarchy, but is the best competitor that can persist under the current disturbance regime (Sebens 1984).

4. Conclusion

It is now widely accepted among theoretical ecologists that periodic, severe, localized environmental disruption can increase biological diversity. However, many of the processes by which this takes place are not yet fully understood. Ongoing research in this area will continue to be of particular benefit to government fisheries resource managers as well as producers of sessile marine fauna (e.g., coral farmers). A comprehensive understanding of how ecological disturbance influences biological community structure and diversity could aid considerably in carrying out stock assessments (as in efforts to minimize the ecological impact of wild harvest for the aquarium trade) as well as operational planning for intensive re-circulating aquaculture operations.

References

1. Armstrong, Robert A. 1976. Fugitive species: Experiments with fungi and some theoretical consideration. Ecology vol. 57, no. 5: 953-963.
2. Blanchette, Carol Anne. 1996. Seasonal patterns of disturbance influence recruitment of the sea palm, Postelsia palmaeformis. Journal of Experimental Marine Biology and Ecology 197: 1-14.
3. Connell, Joseph H. 1961. The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology vol. 42, no. 4: 710-723.
4. Coyer, James A., Jeanine L. Olsen and Wytze T. Stam. 1997. Genetic variability and spatial separation in the sea palm kelp Postelsia palmaeformis (Phaeophyceae) as assessed with M13 fingerprints and APDS. Journal of Phycology 33: 561-568.
5. Dayton, Paul K. 1971. Competition, disturbance, and community organization: The provision and subsequent utilization of space in a rocky intertidal community. Ecological Monographs vol. 41, no. 4: 351-389.
6. Dayton, Paul K. 1973. Dispersion, dispersal, and persistence of the annual intertidal alga, Postelsia palmaeformis Ruprecht. Ecology vol. 54, no. 2: 433-438.
7. Hastings, Alan. 1980. Disturbance, coexistence, history, and competition for space. Theoretical Population Biology 18: 363-373.
8. Kusumo, Handojo T., Catherine A. Pfister and J. Timothy Wootton. 2005. Small-scale genetic structure in the sea palm Postelsia palmaeformis Ruprect (Pheaophyceae). Marine Biology 149: 731-742.
9. Levin, Simon A. and R. T. Paine. 1974. Disturbance, patch formation, and community structure. Proc. Nat. Acad. Sci. USA vol. 71, no. 7: 2744-2747.
10. Liddel, Michael. 2001. A simple space competition model using stochastic and episodic disturbance. Ecological modelling vol. 143, issues 1-2: 33-41.
11. McNaughton, S. J. and L.L. Wolf. 1970. Dominance and the niche in ecological systems. Science 167: 131-142.
12. Nielsen, Karina J., Carol A. Blanchette, Bruce Menge and Jane Lubchenco. 2006. Physiological snapshots reflect ecological performance of the sea palm, Postelsia palmaeformis (Phaeophyceae) across intertidal elevation and exposure gradients. Journal of Phycology 42: 548-559.
13. Paine, R. T. 1988. Habitat suitibility and local population persistence of the sea palm Postelsia palmaeformis. Ecology vol. 69, no. 6: 1787-1794.
14. Pickett, S. T. A. and P. S. White. 1985. The Ecology of Natural Disturbance and Patch Dynamics. Patch dynamics: A synthesis. Chap. 21: 371-384.
15. Riddle, Dana. 2009. Environmental Stability: A Comparison of a Natural Reef and an Aquarium. Advanced Aquarist’s Online Magazine, Volume 8, Book 1: 2009 Edition. Pomacanthus Publications, Inc.
16. Sebens, Kenneth P. 1984. Competition for space: Effects of disturbance and indeterminate competitive success. Theoretical Population Biology 32: 430-441.
17. Sousa, Wayne P. 1985. The Ecology of Natural Disturbance and Patch Dynamics. Disturbance and patch dynamics on rocky intertidal shores. chap. 7: 101-124.