Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Changing Patterns of Microhabitat Utilization by the Threespot Damselfish, Stegastes planifrons, on Caribbean Reefs

  • William F. Precht,

    Affiliation National Oceanic and Atmospheric Administration (NOAA), Florida Keys National Marine Sanctuary, Key Largo, Florida, United States of America

  • Richard B. Aronson ,

    Affiliation Department of Biological Sciences, Florida Institute of Technology, Melbourne, Florida, United States of America

  • Ryan M. Moody,

    Affiliation Dauphin Island Sea Lab, Dauphin Island, Alabama, United States of America

  • Les Kaufman

    Affiliation Boston University Marine Program, Department of Biology, Boston University, Boston, Massachusetts, United States of America



The threespot damselfish, Stegastes planifrons (Cuvier), is important in mediating interactions among corals, algae, and herbivores on Caribbean coral reefs. The preferred microhabitat of S. planifrons is thickets of the branching staghorn coral Acropora cervicornis. Within the past few decades, mass mortality of A. cervicornis from white-band disease and other factors has rendered this coral a minor ecological component throughout most of its range.

Methodology/Principal Findings

Survey data from Jamaica (heavily fished), Florida and the Bahamas (moderately fished), the Cayman Islands (lightly to moderately fished), and Belize (lightly fished) indicate that distributional patterns of S. planifrons are positively correlated with live coral cover and topographic complexity. Our results suggest that species-specific microhabitat preferences and the availability of topographically complex microhabitats are more important than the abundance of predatory fish as proximal controls on S. planifrons distribution and abundance.


The loss of the primary microhabitat of S. planifronsA. cervicornis—has forced a shift in the distribution and recruitment of these damselfish onto remaining high-structured corals, especially the Montastraea annularis species complex, affecting coral mortality and algal dynamics throughout the Caribbean.


Caribbean coral reefs have changed dramatically over the past few decades [1]. Until the late 1970s, Caribbean reefs displayed a generalized zonation dominated by three common taxa of scleractinian corals, which were the primary builders of reef framework: the branching elkorn coral Acropora palmata, the branching staghorn coral A. cervicornis, and the massive corals of the Montastraea annularis species complex [2], [3]. Since that time coral cover has declined [4], and the pattern of zonation has essentially vanished [5]. The most conspicuous change has been the near-elimination of acroporid corals across the entire region [6]. Several factors have been responsible for the mass mortality of Acropora, with white-band disease, predation, and hurricanes ranking as the most significant [1], [6]. Corals of the M. annularis complex have also declined on some reefs [7], [8], but the causes have been different than for the Acropora species [9]. In this paper, we examine the effects of these shifts in coral assemblage structure on microhabitat utilization by the ecologically significant threespot damselfish, Stegastes planifrons (Cuvier). This species of herbivorous fish is important in reef communities of the Caribbean, because it mediates interactions among corals, algae, and other herbivores [10].

Although S. planifrons are capable of occupying a number of microhabitats [10][14], they prefer thickets of A. cervicornis [10], [15], [16]. Before 1980, S. planifrons were common residents of shallow and intermediate depths (<30 m) on fore-reef terraces throughout the Caribbean [10]. Because suitable microhabitat was abundant on most Caribbean reefs at the time, it was thought that S. planifrons were not at carrying capacity and were, therefore, recruitment-limited [11]. This view may be changing, as live coral cover, especially that of the Acropora species, has plummeted in recent decades, greatly reducing the overall habitat available for S. planifrons.

Clarke [17] noted that in the Bahamas S. planifrons were 20 times more abundant in structurally complex coral thickets, especially thickets of A. cervicornis, than in any other microhabitat type. He suggested that S. planifrons utilized structurally complex microhabitats generated by A. cervicornis to avoid predation. Williams [15] indicated that predation on S. planifrons occupying A. cervicornis patches was very low. In the absence of A. cervicornis threespots exhibited a preference for structurally complex massive corals [12], [18], which presumably also provided some refuge from predators. These secondary, suboptimal microhabitats offered fewer crevices and hiding areas than A. cervicornis, and as a result resident damselfish suffered higher mortality [16], [19]. Once territories were established on massive corals, however, S. planifrons showed strong site fidelity and long-term survivorship [20][23].

S. planifrons are highly territorial and actively kill scleractinian corals by biting the living tissue and cultivating dense algal lawns on the coral skeletons [10]. Many reef fishes tend algal gardens, but S. planifrons are the only Caribbean damselfish so strongly tied to stands of living coral (Figure 1), aggressively defending and guarding their territories against other herbivores. In this respect they represent the extreme expression of a behavior manifested in a host of reef-dwelling pomacentrid species in the genera Stegastes, Dischistodus, Hemiglyphidodon, Plectroglyphidodon, Pomacentrus, Microspathodon, and Chrysiptera [24]. Although S. planifrons do not appear to kill corals for direct food value [19], the fish clearly benefit from causing coral mortality [10], [25].

Figure 1. A threespot damselfish, Stegastes planifrons, and its territory on a colony of Monastraea faveolata.

Note the bite-induced lesions of living coral tissue along the margin of the territory. From Carysfort Reef, Florida Keys National Marine Sanctuary; summer 2003; 10 m depth. Photo by WFP.

Historical Observations

In the 1970s, prior to the acroporid die-off, Kaufman [10] noted that up to 40% of the surface of the fore-reef terrace at Discovery Bay, Jamaica was covered by algal gardens of S. planifrons. At any given time more than 20% of the reef surface contained living coral under attack by S. planifrons [26]. Because the life-history strategy of M. annularis complex includes slower growth and a more massive skeletal structure than Acropora spp., the effects of S. planifrons gardens on knobs and pillars of living M. annularis complex are more devastating (Figure 1); Kaufman [10] suggested that gardening by S. planifrons could actually increase the spatial coverage of healthy stands of A. cervicornis by dampening competition from massive corals and by promoting branch fragmentation. Thresher [27] likewise suggested that the activity of S. planifrons permitted A. cervicornis to dominate, because the branching corals recovered rapidly from their injuries while more massive corals did not. Wellington [28] noted a similar relationship in the eastern Pacific, where the activity of the damselfish S. acapulcoensis facilitated the establishment of monospecific stands of branching pocilloporid corals at the expense of massive pavonid colonies.

In 1980, Hurricane Allen struck the north coast of Jamaica, drastically altering reef community structure by breaking and killing most of the branching corals [29]. S. planifrons, their territories, and the supporting thickets of A. cervicornis were almost entirely eliminated to a depth of ∼12 m on the fore reef at Discovery Bay [26]. After the storm, the density of S. planifrons increased in deeper water as the fish relocated and recruited to areas that were still relatively coral-rich. Immigration of mature S. planifrons into deeper water was reflected in patterns of coral mortality and microhabitat utilization [26]. S. planifrons that recruited to deeper water after the storm generally did not return to shallow water, a direct consequence of the disappearance of physical structure caused by the loss of the branching acroporids. New S. planifrons territories established on surviving A. cervicornis colonies in shallow water caused significant secondary mortality of the corals [30]. Knowlton et al. [30] noted that when A. cervicornis was abundant, predators such as S. planifrons generally did not have a detrimental effect; however, when A. cervicornis was rare, predation effects could devastate the surviving colonies. Roberts [31] suggested that the continuing negative effect of S. planifrons on remnant colonies of A. cervicornis may be keeping the coral rare, threatening the long-term prospects for its persistence regionally.

More than 90% of large, robust colonies of the M. annularis complex survived Hurricane Allen on the fore reef at Discovery Bay [32]. Columnar growths of Montastraea protruded upward from fields of broken and flattened A. cervicornis branches, and many surviving Montastraea colonies were subsequently colonized by the S. planifrons that had lost their territories in A. cervicornis thickets (WFP, LK personal observations). This switch caused significant collateral mortality in the remaining population of M. annularis complex, as well as subsequent algal overgrowth of the dead colonies [1]. Liddell et al. [33] were the first to suggest that the shift in microhabitat use by S. planifrons following the loss of A. cervicornis might represent an important new source of mortality for M. annularis complex and ultimately affect the production of reef framework.

Were the densities of S. planifrons artificially inflated by decades of overfishing, which released the damselfish from predation? To answer this question, Kaufman [34], [35] sampled fossil A. cervicornis branches from the Pleistocene Falmouth Formation (∼125 ka) at Rio Bueno, Jamaica. He observed abundant skeletal galls on fossil branches of A. cervicornis that had resulted from the bites of S. planifrons, suggesting that high densities of threespot territories have long been an attribute of Caribbean reef ecology. The discovery of a living amphipod species found only in S. planifrons algal gardens on A. cervicornis [36] is further circumstantial evidence that the relationship between S. planifrons and A. cervicornis is historically rooted and not a recent artifact of Caribbean reef ecology.

Recent Observations

The recent paucity of A. cervicornis throughout the Caribbean has apparently caused a shift in S. planifrons from its preferred microhabitat to secondary microhabitats. S. planifrons are familiar occupants of any microhabitat that is structurally complex with abundant vertical fissures. These include columnar morphologies of the M. annularis complex [14], [37], [38]. In Florida, Eakin [39] observed that, in the absence of live branching corals, juvenile S. planifrons preferentially recruited to living Montastraea colonies. These observations strongly suggest that M. annularis complex has now become the primary microhabitat of S. planifrons on fore-reef terraces throughout the Caribbean.

Reef-fish assemblages have changed concomitantly with coral assemblages. Predators, especially groupers (Serranidae) and snappers (Lutjanidae), have declined in recent decades due to overfishing and habitat loss [40][42]. A number of authors have asserted a causal chain leading from overfishing, to reduced densities of predators, to enhanced damselfish densities, to increased coral mortality, and thence to increased algal cover. Vicente [43] and Hernandez-Delgado [37], for example, attributed algal overgrowth of corals in Puerto Rico to overfishing, which in their view released S. planifrons from predation and allowed them to kill the corals. Ogden [44], citing results from the Caribbean Coastal Marine Productivity (CARICOMP) Program, implicated overfishing as causing enhanced abundance of damselfishes throughout the region. He suggested that increases in damselfish densities have overwhelmed the capacity of the corals to counteract their impacts, resulting in the decline of coral populations and the smothering of reefscapes with vast algal lawns. Others have voiced similar opinions in the scientific literature and the popular news media [45][47].

A major factor confounding the overfishing hypothesis is the loss of reef fish, including the predators of damselfishes and the damselfish themselves, caused by mass coral mortality and the consequent loss of reef structure [48][50]. Although it stands to reason that fewer predators could result in higher numbers of algal-gardening damselfish [51] or alterations in territorial dynamics [52], these expectations are predicated on the assumption that S. planifrons were/are in fact predator-limited. An alternative hypothesis is that S. planifrons populations are limited ultimately by predators but proximally by the availability of microhabitat: they have evolved to evade predation by remaining closely associated with appropriate structural refugia. If this alternative hypothesis is correct, reducing predator abundance could result in increased survival of non-territorial or peripheral individuals, but the density of coral-killing algal gardeners should remain approximately the same over a broad range of predation intensities. There has never been a formal test of the hypothesis that the density of territorial S. planifrons is predator-limited when sufficient preferred habitat is available.


The null hypothesis is that there is no proximal effect of piscivorous fishes on the abundance of S. planifrons. If on the other hand predation by piscivorous fishes controls the abundance and distribution of S. planifrons, reefs with higher fishing pressure should have higher densities of S. planifrons than reefs with lower fishing pressure. Under this scenario S. planifrons territories should have spread into all available microhabitats on overfished reefs, causing coral mortality and the massive proliferation of algal gardens. If microhabitat availability controls the abundance of S. planifrons, the loss of A. cervicornis should have caused a shift in microhabitat use without necessarily increasing the overall densities of these damselfish on overfished reefs. The shift to secondary, suboptimal microhabitats that were previously devoid of S. planifrons should also have resulted in coral mortality and proliferation of algal gardens. A third alternative is the combined action of the two processes: both predatory release and microhabitat availability controlling the distribution and abundance of S. planifrons.


Study Areas

During the period 1998–2001, we compared sites in Jamaica (heavily fished), Florida and the Bahamas (moderately fished), the Cayman Islands (lightly to moderately fished), and Belize (lightly fished) to test the alternative hypotheses (Table 1). We selected study sites based on the following criteria: (a) sites were chosen along a gradient of fishing pressure; (b) all sites were located in fore-reef habitats at 10–15 m depth; (c) all sites were known to have had abundant stands of A. cervicornis in the recent past; and (d) A. cervicornis was either rare or absent at each site during the study period. Differences in fishing pressure were ascertained from the published literature [40], [41], [53][57], as well as from interviews with fisherman, dive operators, reef scientists, and site managers from these locations.

Table 1. Descriptive information for the ten sites used in the study.

One to three study sites were established at each survey location. At each site, six 25-m surveyor's tapes were laid haphazardly. A diver swam along each transect, identifying and counting fishes within 1 m on either side of the tape. Fish species were categorized as: (a) S. planifrons; (b) pomacentrid species other than S. planifrons; (c) herbivores other than damselfish; or (d) piscivores. The diver then swam back over the transect line, recording the sessile organism or substratum type beneath each 10-cm mark on the tape. Finally, the diver swam the transect a third time, recording all regular echinoids within 1 m on either side.

Structural complexity, or topographic heterogeneity, was measured by conforming a 5-m length of brass chain (links 17 mm long) to the substratum along the central portion of each 25-m tape, beginning 10 m from the start of the transect. An index of structural complexity was calculated as C = 1−d/l, where d is the horizontal distance covered by the chain when conformed to the substratum and l is the length of the chain when fully extended [58].

Statistical Analysis

We used principal components analysis (PCA) on mean values from the six 25-m transects at each study site to determine the proportion of among-site variance in the abundance of S. planifrons attributable to: (a) the cover of M. annularis complex (which consisted exclusively of M. annularis sensu stricto and M. faveolata); (b) the cover of A. cervicornis; (c) the cover of living hard corals other than M. annularis complex and A. cervicornis (‘other hard corals’); (d) the structural complexity of the benthos; (e) the water depth; (f) the abundance of piscivorous fish; (g) the abundance of non-pomacentrid herbivorous fish (‘other herbivorous fishes’); and (h) the abundance of pomacentrids other than S. planifrons (‘other damselfishes’). Regular echinoids were exceedingly rare in the transects and were not included in the PCA. Significant eigenvector loadings were identified by performing Pearson correlation analyses between the original independent variables and their corresponding eigenvectors for each PC [59].

PCA requires that all samples for each independent variable be drawn from a normal distribution. To meet this requirement, point-counts of A. cervicornis were transformed using the function {Y = −1/x}. The remaining seven variables did not require data transformation.

We also performed separate linear-regression analyses of S. planifrons abundance versus: (a) the abundance of piscivores; (b) the cover of M. annularis complex; (c) the structural complexity of the substratum; and (d) the total cover of living hard corals (including M. annularis complex and A. cervicornis). Total hard-coral cover was calculated as the sum of point-counts for M. annularis complex, A. cervicornis, and all other species of hard corals. The regressions were run to determine whether the variables of primary interest had significant predictive power in explaining the abundance of S. planifrons among sites. S. planifrons abundance and all independent variables were log10-transformed prior to regression analysis.


Three principal components (PCs) extracted from the correlation matrix explained 87.3% of the variability among sites (Figure 2, Table 2). PC1 and PC2 explained 73.4% of the variability, and PC4 contributed 13.9%. The remaining five PCs explained 12.7% of the total variance and were not considered further.

Figure 2. Scatterplot of site-scores on the first and second principal components.

These two axes represent 73.4% of the total variation in the correlation matrix. Independent variables listed on each axis indicate variables with significant eigenvector loadings. Abbreviations for variables are listed in Table 2. Mean S. planifrons densities are given for each site (n = 6 transects per site).

Table 2. Eigenvectors and eigenvalues for the principal component analysis of eight independent variables collected at the 12 sites.

PC1 accounted for 38.9% of the variability in the data set, yielding significantly negative eigenvector loadings for point-counts of M. annularis complex, structural complexity, and depth (Table 2). Sites with lower PC1 scores had higher point-counts of M. annularis complex, higher structural complexities, and slightly deeper depths. M. annularis complex was the most abundant coral taxon at each site, with an overall mean proportional contribution to total living-coral point-counts of 0.53±0.04 SE (range 0.32–0.78).

PC2 accounted for 32.9% of the total variance in the data set and yielded significantly positive eigenvector loadings for mean counts of piscivores; non-pomacentrid, other herbivores; other damselfishes; and point-counts of other corals. Positive eigenvector loadings were generated for all functional groups of fish, indicating that piscivores did not have a negative impact on either other pomacentrids or other herbivorous fishes. Sites with higher PC2 scores exhibited higher abundances of piscivores, other herbivorous fishes, other damselfishes, and other hard corals. PC3 explained 15.8% of the variance in the data set and represented the abundance of A. cervicornis at each location.

The piscivores detected in the transects at all sites were small to intermediate in size, at 20–50 cm standard length. Belt transects are poor estimators of the abundance of large, wide-ranging, predatory fishes, as compared to smaller, site-attached fishes [60], [61]. We noted the presence/absence of larger piscivores, including sharks, barracuda, groupers, snappers, and jacks, adjacent to our transects; at all study sites these fishes were uncommon to rare and, therefore, assumed not to be of primary importance to our analysis.We focused on small- to intermediate-sized piscivores, which included the smaller serranids and lutjanids, because they are the primary predators of adult and juvenile S. planifrons [47], [62]. These smaller predators could themselves have been released by the overfishing of larger predators; thus, fishing pressure could actually be expected to result in fewer, rather than more, damselfishes. For example, Stallings [63] found that the harvesting and depletion of Nassau grouper, Epinephelus striatus, allowed two smaller-bodied, intermediate predators (coney and graysby groupers, Cephalopholis fulva and C. cruentata) to proliferate, which in turn had strong negative effects on their prey. Our data, however, do not support such a hypothesis.

Sites with low PC1 loadings—high point-counts of M. annularis complex, high levels of structural complexity, and deeper depths—exhibited high densities of S. planifrons; these sites were located in Belize, the Bahamas, and the Cayman Islands. Sites with high PC1 loadings exhibited low and intermediate densities of S. planifrons; these sites were located in Jamaica and the Florida Keys (Figure 2). Sites exhibiting intermediate densities of S. planifrons, all of which were located in Florida, exhibited PC1 loadings >1.

Positive PC2 loadings, denoting high abundances of piscivores, other herbivorous fishes, other damselfishes, other corals, were associated with intermediate to high abundances of S. planifrons. We interpret this to mean that the site groupings reflect differences in overall biotic composition among locations rather than the impact of one particular PC2 variable on the abundance of S. planifrons. Thus, counts of S. planifrons did not decline monotonically as abundances of piscivores and other damselfishes increased but instead tracked the availability of coral-generated habitat complexity.

In agreement with our interpretation of PC1, regression analysis did not reveal a significant relationship between the abundance of S. planifrons and piscivores among our study sites (r2<0.01; p = 0.435; n = 12; Figure 3A). We did, however, detect a strong relationship between the abundances of S. planifrons and M. annularis complex (r2 = 0.71; p<0.001; n = 12; Figure 3B), structural complexity (r2 = 0.90; p<0.001; n = 12; Figure 3C), and total coral counts (all hard corals, including M. annularis complex and A. cervicornis: r2 = 0.68; p<0.001; n = 12; Figure 3D). These findings are consistent with the high PC1 loadings for each of these variables and consistent with the fact that M. annularis complex is both a dominant member of the coral assemblage at each site and currently the preferred microhabitat of S. planifrons at these depths.

Figure 3. Relationships between key parameters and the abundance of S. planifrons.

The abundance of S. planifrons regressed on: (A) piscivore abundance; (B) proportional cover of the Montastraea annularis species complex; (C) structural complexity; and (D) proportional cover of total hard corals. The coordinates of each point are the log-transformed means of the transects within a site.


Our data strongly suggest that the availability of appropriate microhabitat is the primary determinant of the population density of S. planifrons. The negative correlation between the abundance of (small- to intermediate-sized) piscivores and the abundance of S. planifrons, expected under the hypothesis of predatory control, was negligible and non-significant. The evolution of microhabitat preference by S. planifrons is likely to have been at least in part a consequence of predation pressure [64]; however, the survey data do not support the hypothesis that current densities of piscivorous fishes determine current densities of S. planifrons, a conclusion borne out by other studies in Curaçao and the Florida Keys [65].

In a long-term monitoring program in the U.S. Virgin Islands, smaller serranids and lutjanids increased over an 18-year interval, while larger piscivores remained rare [66]. Population densities of S. planifrons increased during the same period. The increase in small- to intermediate-sized predators, however, should have resulted in fewer damselfish. Clearly the abundances of piscivores and S. planifrons were decoupled. Likewise, whether or not fishing pressure enhanced the abundances of small- to intermediate-sized piscivores at any of our sites is moot from the standpoint of the abundance of threespot damselfish.

Williams [67] performed exclusion experiments in Jamaica to test the effect of predation on adult S. planifrons. She found no significant change in numbers of S. planifrons in the absence of predation. Likewise, She noted that when A. cervicornis was abundant predators did not have a strongly detrimental effect on S. planifrons populations [15], [67]. On the Great Barrier Reef, artificial reefs constructed from the high-complexity coral Pocillopora damicornis supported the same numbers of juvenile damselfish when predators were abundant as when predators were absent [68].

The highest densities of piscivores and damselfishes other than S. planifrons were found in the Florida Keys and could have been related to the local protection afforded no-take reserves. S. planifrons occurred at intermediate abundances in Florida, despite the negative effects of predation and competition that might have been expected (contra 69).

Prior to the demise of the acroporids, Bohnsack [70] tested the predation hypothesis in the Florida Keys. In contrast to our results, he found that reefs with high fishing pressure did have significantly higher numbers of S. planifrons than reference reefs with lower fishing pressure and higher numbers of piscivores. The results, however, were confounded by differences in habitat type between his study reefs: the protected sites with higher numbers of S. planifrons were dominated by A. cervicornis, whereas the sites with lower numbers of S. planifrons were dominated by an assemblage of head corals. Bohnsack [70] recognized this problem and was careful to note, “Stating that piscivorous predation is an important factor controlling community structure of reef fishes based on present evidence would be premature.”

Perhaps the most persuasive evidence against monolithic predator-limitation of S. planifrons is that even on the north and west coasts of Jamaica, an extreme example of a chronically overfished situation [71], S. planifrons exhibited high microhabitat specificity. Population and territory spillage onto hemispheric or horizontal-foliose corals occurred only where these non-preferred microhabitats were immediately adjacent to patches of preferred microhabitat [10], [72]. When acroporids virtually disappeared, threespots moved onto very specific secondary and tertiaty microhabitats (M. annularis complex and Porites porites; [26]). Where only low-relief fields of coral rubble remained, the density of adult S. planifrons was drastically reduced [see also 73]. The greatly expanded small-coral and coral-rubble microhabitats were heavily colonized by two other damslefish species that became more prevalent: Stegastes diencaeus, which is a less active gardener than S. planifrons, and S. partitus, which is a planktivore and not an algal gardener (LK personal observation). Gladfelter et al. [74] also noted a decrease in numbers of S. planifrons in St. Croix after Hurricane Hugo as a direct consequence of the physical loss of microhabitat.

On a regional level, the loss of structural complexity caused by the Caribbean-wide mass mortality of A. cervicornis in the late 1970s and 1980s [50] reduced the total amount of shelter available to S. planifrons. The damselfish relocated or recruited to remaining high-structured living corals, especially M. annularis complex. S. planifrons predation on the living coral tissue of these secondary microhabitats has been chronic and intense, resulting in extensive coral mortality and proliferation of algal gardens (Figure 1).

Wellington [75] demonstrated that loss of structurally complex branching corals from disturbance resulted in the relocation of S. acapulcoensis to secondary microhabitats, with lethal consequences for massive corals. Monospecific stands of Pocillopora damicornis, the microhabitat preferred by the damselfish [28], were killed in the Gulf of Panama by the 1982–1983 El Niño event. As structural complexity was reduced by bioerosion in the years following this disturbance, S. acapulcoensis colonized the massive coral Gardineroseris planulata. Colonies of Gardineroseris that had been monitored for 14 years were free of S. acapulcoensis before the El Niño event. These colonies subsequently suffered substantial mortality from S. acapulcoensis [7], [76].

These examples show that when the preferred microhabitats of territorial damselfish are abundant, there is little collateral mortality or algal overgrowth on secondary microhabitats. When the preferred microhabitats are eliminated by mortality of the engineer species, which is to say branching corals, the impact on secondary microhabitats can be dramatic and intense. Because massive corals grow more slowly than branching corals, episodes of mass mortality of branching corals inhabited by algal gardeners may leave an enduring imprint on community structure, continuing long after the branching corals reestablish themselves and the damselfishes move back into them and away from their suboptimal microhabitats.

Observations from Pleistocene coral assemblages confirm that S. planifrons had been abundant and A. cervicornis had been their preferred microhabitat in the tropical Atlantic for a long time prior to any human interference [34], [35]. Other paleoecological evidence indicates the recent mass mortality of Caribbean acroporids to be a novel event in the late Holocene [77]. Although some populations of Acropora spp. have been extirpated locally, the two species are surviving regionally and may yet recover to their former large population sizes [78]. Caribbean-wide regeneration and recovery of the S. planifronsAcropora relationship could take decades or centuries. Locally, however, acroporid restoration could yield improved survivorship for massive corals by allowing the S. planifrons to relocate to their preferred microhabitat; this idea is currently being tested by two of us (LSK and WFP) in the Florida Keys. In the meantime, further community disintegration should be expected as S. planifrons continue their turf wars on slow-growing, long-lived, massive corals.


We acknowledge with gratitude the many individuals who have influenced our thinking about damselfish and their role in the ecology of Caribbean reefs, including Jamie Bechtel, Jim Bohnsack, John Ebersole, Bill Fitt, Peter Glynn, Ann Houston Williams, Jeremy Jackson, Brian Keller, Phil Lobel, Ian Macintyre, Steven Miller, John Ogden, Jim Porter, Enric Sala, Alina Szmant, and John Valentine. Special thanks are due Peter Gayle, Michael Haley, Brian Keller, Norman Quinn, George Warner, and Jeremy Woodley for providing facilities at the Discovery Bay Marine Laboratory (DBML) in Jamaica over the three decades of our research there. Jennifer Hobbs, Jelani Reynolds, Burton Shank, and Lauren Toth provided technical assistance. Comments from two anonymous reviewers significantly improved this manuscript. This is Contribution No. 882 from the Smithsonian Institution's Caribbean Coral Reef Ecosystems Program, DBML Contribution No. 735, and Contribution No. 35 from the Institute for Research on Global Climate Change at the Florida Institute of Technology. Field work was carried out under permits from the Belize Department of Fisheries and the Florida Keys National Marine Sanctuary. The content of this manuscript does not reflect any position of the U.S. Government or the National Oceanic and Atmospheric Administration unless otherwise specified. We dedicate this paper to the memory of our friend and colleague Dr. Brian Keller. He brought transcendent wisdom and blessed calm to an academic reef of small territories defended by aggressive beings.

Author Contributions

Conceived and designed the experiments: WFP RBA LK. Performed the experiments: WFP RBA. Analyzed the data: WFP RBA RMM LK. Contributed reagents/materials/analysis tools: WFP RBA. Wrote the paper: WFP RBA RMM LK.


  1. 1. Aronson RB, Precht WF (2001a) Evolutionary paleoecology of Caribbean coral reefs. In: Allmon WD, Bottjer DJ, editors. Evolutionary paleoecology: the ecological context of macroevolutionary change. New York: Columbia University Press. pp. 171–233.
  2. 2. Goreau TF (1959) The ecology of Jamaican coral reefs. I. Species composition and zonation. Ecology 40: 67–90.
  3. 3. Goreau TF, Goreau NI (1973) The ecology of Jamaican coral reefs. II. Geomorphology, zonation and sedimentary phases. Bull Mar Sci 23: 399–464.
  4. 4. Gardner TA, Côté IM, Gill JA, Grant A, Watkinson AR (2003) Long-term region-wide declines in Caribbean corals. Science 301: 958–960.
  5. 5. Jackson JBC (1992) Pleistocene perspectives of coral reef community structure. Am Zool 32: 719–731.
  6. 6. Aronson RB, Precht WF (2001b) White-band disease and the changing face of Caribbean coral reefs. Hydrobiologia 460: 25–38.
  7. 7. Edmunds PJ, Elahi R (2007) The demographics of a 15-year decline in coral cover of the Caribbean reef coral Montastaea annularis. Ecol Monogr 77: 3–18.
  8. 8. Carpenter KE, Abrar M, Aeby G, Aronson RB, Banks S, et al. (2008) One-third of reef-building corals face elevated extinction risk from climate change and local impacts. Science 321: 560–563.
  9. 9. Aronson RB, Precht WF (2006) Conservation, precaution, and Caribbean reefs. Coral Reefs 25: 441–450.
  10. 10. Kaufman LS (1977) The threespot damselfish: effects on benthic biota of Caribbean coral reefs. Proc 3rd Intl Coral Reef Symp, Miami 1: 559–564.
  11. 11. Robertson DR, Hoffman SG, Sheldon JM (1981) Availability of space for the territorial Caribbean damselfish Eupomacentrus planifrons. Ecology 62: 1162–1169.
  12. 12. Lirman D (1994) Ontogenetic shifts in habitat preferences in the three-spot damselfish, Stegastes planifrons (Cuvier), in Roatán Island, Honduras. J Exp Mar Biol Ecol 180: 71–81.
  13. 13. Lirman D (1999) Reef fish communities associated with Acropora palmata: relationships to benthic attributes. Bull Mar Sci 65: 235–252.
  14. 14. Tolimieri N (1998) Contrasting effects of microhabitat use on large-scale adult abundance in two families of Caribbean reef fishes. Mar Ecol Prog Ser 167: 227–239.
  15. 15. Williams AH (1978) Ecology of threespot damselfish: social organization, age structure, and population stability. J Exp Mar Biol Ecol 34: 197–213.
  16. 16. Itzkowitz M (1979) Territorial tactics and habitat quality. Am Nat 114: 585–614.
  17. 17. Clarke RD (1977) Habitat distribution and species diversity of chaetodontid and pomacentrid fishes near Bimini, Bahamas. Mar Biol 40: 277–289.
  18. 18. Ebersole JP (1985) Niche separation of two damselfish species by aggression and differential microhabitat utilization. Ecology 66: 14–20.
  19. 19. Williams AH (1980) The threespot damselfish: a noncarnivorous keystone species. Am Nat 116: 138–142.
  20. 20. Gutierrez L (1998) Habitat selection by recruits establishes local patterns of adult distribution in two species of damselfishes: Stegastes dorsopunicans and S. planifrons. Oecologia 115: 268–277.
  21. 21. Levin PS, Tolimieri N, Nicklin M, Sale P (2000) Integrating individual behavior and population ecology: the potential for habitat-dependent population regulation in a reef fish. Behav Ecol 11: 565–571.
  22. 22. Caldow C, Wellington GM (2003) Patterns of annual increment formation in otoliths of pomacentrids in the tropical western Atlantic: implications for population age-structure examination. Mar Ecol Prog Ser 265: 185–195.
  23. 23. Abrey CA (2005) The effect of community on the behavior of the threespot damselfish. Env Biol Fish 73: 163–170.
  24. 24. Ceccarelli DM, Jones GP, McCook LJ (2001) Territorial damselfishes as determinants of the structure of benthic communities on coral reefs. Oceanogr Mar Biol Ann Rev 39: 355–389.
  25. 25. Lobel PS (1980) Herbivory by damselfishes and their role in coral reef community ecology. Bull Mar Sci 30: 273–289.
  26. 26. Kaufman LS (1983) Effects of Hurricane Allen on reef fish assemblages near Discovery Bay, Jamaica. Coral Reefs 2: 43–47.
  27. 27. Thresher RE (1980) Reef fish: behavior and ecology on the reef and in the aquarium. St. Petersburg, FL: Palmetto Publishing.
  28. 28. Wellington GM (1982) Depth zonation of corals in the Gulf of Panama: control and facilitation by resident reef fishes. Ecol Monogr 52: 223–241.
  29. 29. Woodley JD, Chornesky EA, Clifford PA, Jackson JBC, Kaufman LS, et al. (1981) Hurricane Allen's impact on Jamaican coral reefs. Science 214: 749–755.
  30. 30. Knowlton N, Lang JC, Keller BD (1990) Case study of natural population collapse: post-hurricane predation on Jamaican staghorn corals. Smithsonian Contrib Mar Sci 31: 1–25.
  31. 31. Roberts CM (1993) Our shifting perspectives on the oceans. Oryx 37: 166–177.
  32. 32. Woodley JD (1989) The effects of Hurricane Gilbert on coral reefs at Discovery Bay. In: Bacon PR, editor. Assessment of the economic impacts of Hurricane Gilbert on coastal and marine resources in Jamaica. UNEP Regional Seas Rep Stud 110, United Nations Environment Programme, Nairobi, Appendix 9.
  33. 33. Liddell WD, Ohlhorst SL, Coates AG (1984) Modern and Ancient Carbonate Environments of Jamaica. Sedimenta X, Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, FL.
  34. 34. Kaufman L (1981) There was biological disturbance on Pleistocene coral reefs. Paleobiology 7: 527–532.
  35. 35. Kaufman LS (1986) Why is the ark sinking. In: Kaufman LS, Mallory K, editors. The last extinction. pp 1–41. Cambridge, MA: MIT Press.
  36. 36. Thomas JD, Barnard JL (1991) A review of the genus Iphimedia (Crustacea: Amphipoda) with descriptions of three new species from Australia, Papua New Guinea and Florida. Invert Taxon 5: 469–485.
  37. 37. Hernandez-Delagado EA (1999) The structure of coral reef epibenthic and fish communities through anthropogenic environmental stress and fishing pressure gradients, and the role of damselfish (Pomacentridae) territorial behavior in coral reef degradation. PhD Dissertation, Department of Biology, University of Puerto Rico, San Juan, Puerto Rico.
  38. 38. Bruckner AW, Bruckner RJ (2006) The recent decline of Montastraea annularis (complex) coral populations in western Curaçao: a cause for concern? Rev Biol Trop 54: Suppl 345–58.
  39. 39. Eakin CM (1989) Microhabitat selection in juveniles of the damselfishes Pomacentrus planifrons and P. partitus. Proc Assoc Isl Mar Labs Carib 22: 35–36.
  40. 40. Ault JS, Bohnsack JA, Meester GA (1998) A retrospective (1979–1996) multispecies assessment of coral reef fish stocks in the Florida Keys. Fish Bull 96: 395–414.
  41. 41. Chiappone M, Sluka R, Sullivan-Sealey K (2000) Groupers (Pisces: Serranidae) in fished and protected areas of the Florida Keys, Bahamas and northern Caribbean. Mar Ecol Prog Ser 198: 261–272.
  42. 42. Stallings CD (2009) Fishery-independent data reveal negative effect of human population density on Caribbean predatory fish communities. PLoS ONE 4(5): e5333.
  43. 43. Vicente VP (1994) Structural changes and vulnerability of a coral reef (Cayo Enrique) in La Parguera, Puerto Rico. In: Ginsburg RN, editor. Proceedings of the colloquium on Global Aspects Coral Reefs: Health, Hazards and History. Miami, FL: University of Miami. pp. 227–232.
  44. 44. Ogden JC (1995) Coral reef decline in the Caribbean region. Proc 2nd European Regional Meeting ISRS, Publ Serv Geol Lux XXIX 224.
  45. 45. Luoma JR (1996) Ecosystems—reef madness. Audubon Mag (Nov–Dec) 24–27.
  46. 46. Steneck RS, Sala EA (2005) Large marine carnivores: trophic cascades and top-down controls in coastal ecosystems past and present. In: Ray JC, Redford KH, Steneck R, Berger J, editors. Large carnivores and the conservation of biodiversity. Washington, DC: Island Press. pp. 110–137.
  47. 47. Harris R (2009) Tiny damselfish may destroy a Caribbean reef. National Public Radio,
  48. 48. Graham NAJ, McClanahan TR, MacNeil MA, Wilson SK, Polunin NVC, et al. (2008) Climate warming, marine protected areas and the ocean-scale integrity of coral reef ecosystems. PLoS ONE 3(8): e3039.
  49. 49. Paddack MJ, Reynolds JD, Aguilar C, Appeldoorn RS, Beets J, et al. (2009) Recent region-wide declines in Caribbean reef fish abundance. Curr Biol 19: 590–595.
  50. 50. Alvarez-Filip L, Dulvy NK, Gill JA, Côté IM, Watkinson AR (2009) Flattening of Caribbean coral reefs: region-wide declines in architectural complexity. Proc R Soc B 276: 3019–3025.
  51. 51. McClanahan TR (2005) Recovery of carnivores, trophic cascades, and diversity in coral reef marine parks. In: Ray JC, Redford KH, Steneck R, Berger J, editors. Large carnivores and the conservation of biodiversity. Washington, DC: Island Press. pp. 247–267.
  52. 52. Ceccarelli DM, Hughes TP, McCook LJ (2006) Impacts of simulated overfishing on the territoriality of coral reef damselfish. Mar Ecol Prog Ser 309: 255–262.
  53. 53. Munro JL (1983) Caribbean coral reef fishery resources. ICLARM Stud Rev 7: 1–276.
  54. 54. Koslow JA, Aiken K, Auil S, Clementson A (1994) Catch and effort analysis of the reef fisheries of Jamaica and Belize. Fish Bull 92: 737–747.
  55. 55. Williams ID, Polunin NVC (2001) Large-scale associations between macroalgal cover and grazer biomass on mid-depth reefs in the Caribbean. Coral Reefs 19: 358–366.
  56. 56. Kramer PA (2003) Synthesis of coral reef health indicators for the western Atlantic: results from the AGGRA program (1997–2000). Atoll Res Bull 496: 1–57.
  57. 57. Pattengill-Semmens CV, Semmens BX (2003) Status of coral reefs of Little Cayman and Grand Cayman, British West Indies, in 1999 (part 2: fishes). Atoll Res Bull 496: 226–247.
  58. 58. Aronson RB, Edmunds PJ, Precht WF, Swanson DW, Levitan DR (1994) Largescale, long-term monitoring of Caribbean coral reefs: simple, quick, inexpensive techniques. Atoll Res Bull 421: 1–19.
  59. 59. Ludwig JA, Reynolds JF (1988) Statistical ecology: a primer on methods in computing. New York, NY: John Wiley and Sons.
  60. 60. Schmitt EF, Sluka RD, Sullivan-Sealy KM (2002) Evaluating the use of roving diver and transect surveys to assess the coral reef fish assemblage off southeastern Hispaniola. Coral Reefs 21: 216–223.
  61. 61. Toller W (2002) Quantitative estimates of species composition and abundance of fishes, and fish species/habitat associations in St. Croix, U.S. Virgin Islands. Final report: patterns of habitat utilization by reef fish on St. Croix, F-7-17, Study 3. Division of Fish and Wildlife, Department of Planning and Natural Resources, US Virgin Islands, St. Croix, USVI.
  62. 62. Randall JE (1967) Food habits of reef fishes of the West Indies. Stud Trop Oceanogr 5: 665–847.
  63. 63. Stallings CD (2008) Indirect effects of an exploited predator on recruitment of coral-reef fishes. Ecology 89: 2090–2095.
  64. 64. Almany GR (2004) Does increased habitat complexity reduce predation and competition in coral reef fish assemblages? Oikos 106: 275–284.
  65. 65. Neely KL (2008) Influence of substrate on coral reef fish communities. PhD Dissertation, University Program in Ecology, Duke University, Durham, NC.
  66. 66. Friedlander AM, Beets J (2008) Temporal trends in reef fish assemblages inside Virgin Islands National Park and around St. John, U.S. Virgin Islands (1988–2006). NOAA Technical Memorandum NOS NCCOS 70, Silver Spring, MD.
  67. 67. Williams AH (1981) An analysis of competitive interactions in a patchy back-reef environment. Ecology 62: 1107–1120.
  68. 68. Beukera JS, Jones GP (1997) Habitat complexity modifies the impact of piscivores on a coral reef fish population. Oecologia 114: 50–59.
  69. 69. Baker B (1999) First aid for an ailing reef—research in the Florida Keys National Marine Sanctuary. BioScience 49: 173–178.
  70. 70. Bohnsack JA (1982) Effects of piscivorous predator removal on coral reef fish community structure. In: Caillet GM, Simenstad CA, editors. Gutshop '81: fish food habits studies. pp. 258–267. Washington SeaGrant Publication, University of Washington, Seattle, WA.
  71. 71. Hardt MJ (2008) Lessons from the past: the collapse of Jamaican coral reefs. Fish Fisheries 10: 1–16.
  72. 72. Kaufman LS (1979) Damselfish disturbance on Caribbean coral reefs. PhD Dissertation, Johns Hopkins University, Baltimore, MD.
  73. 73. Wilkes AA, Cook MM, DiGirolamo AL, Eme J, Grim JM, et al. (2008) A comparison of damselfish densities on live staghorn coral (Acropora cervicornis) and coral rubble in Dry Tortugas National Park. Southeast Nat 7: 483–492.
  74. 74. Gladfelter EH, Bythell JC, Hillis ZM (1992) Changes in fish assemblage structure at Buck Island, St. Croix, US Virgin Islands from 1980–1990: an indication of predictability in coral reef fish assemblages based on known habitat changes? In: Bythell JC, Gladfelter EH, Bythell M, editors. pp. 14–29. Ecological studies of Buck Island Reef National Monument, St. Croix, US Virgin Islands: a quantitative assessment of selected components of the coral reef ecosystem and establishment of long-term monitoring sites. Part 2, US Department of the Interior, National Park Service Report. Island Resources Foundation and West Indies Laboratory, St. Croix, US Virgin Islands.
  75. 75. Wellington GM, Victor BC (1985) El Niño mass coral mortality: a test of resource limitation in a coral reef damselfish population. Oecologia 68: 15–19.
  76. 76. Glynn PW (1993) Feeding ecology of selected coral-reef macroconsumers: patterns and effects on coral community structure. In: Dubinsky Z, editor. Ecosystems of the world 25: coral reefs, Elsevier, Amsterdam, Netherlands. pp. 365–400.
  77. 77. Aronson RB, Macintyre IG, Precht WF, Murdoch TJT, Wapnick CM (2002) The expanding scale of species turnover events on coral reefs in Belize. Ecol Monogr 72: 233–249.
  78. 78. Precht WF, Aronson RB (2006) Death and resurrection of Caribbean reefs: a paleoecological perspective. In: Côté I, Reynolds J, editors. Coral reef conservation. Cambridge, UK: Cambridge University Press. pp. 40–77.