Harmful Algae 5 (2006) 565–585
                                                       www.elsevier.com/locate/hal




          Harmful algae on tropical coral reefs: Bottom-up
            eutrophication and top-down herbivory
            Mark M. Littler a,*, Diane S. Littler a,b, Barrett L. Brooks a
             a
               Department of Botany, MRC 166, PO Box 37012, National Museum of Natural History,
                        Smithsonian Institution, Washington, DC 20013, USA
                  b
                   Division of Marine Science, Harbor Branch Oceanographic Institution,
                         5600 US 1 North, Ft. Pierce, FL 34946, USA
            Received 13 July 2005; received in revised form 25 October 2005; accepted 10 November 2005


Abstract
  A conceptual paradigm, the ‘‘Relative Dominance Model’’, provides the perspective to assess the interactive external forcing-
mechanisms controlling phase shifts among the dominant benthic functional groups on tropical coral reefs [i.e., microalgal turfs and
frondose macroalgae (often harmful) versus reef-building corals and calcareous coralline algae (mostly beneficial due to accretion
of calcareous reef framework)]. Manipulative experiments, analyses of existing communities and bioassays tested hypotheses that
the relative dominances of these functional groups are mediated by two principal controlling factors: nutrients (i.e., bottom-up
control) and herbivory (i.e., top-down control). The results show that reduced nutrients alone do not preclude fleshy algal growth
when herbivory is low, and high herbivory alone does not prevent fleshy algal growth when nutrients are elevated. However, reduced
nutrients in combination with high herbivory virtually eliminate all forms of fleshy micro- and macro-algae. The findings reveal
considerable complexity in that increases in bottom-up nutrient controls and their interactions stimulate harmful fleshy algal blooms
(that can alter the abundance patterns among functional groups, even under intense herbivory); conversely, elevated nutrients inhibit
the growth of ecologically beneficial reef-building corals. The results show even further complexity in that nutrients also act directly
as either limiting factors (e.g., physiological stresses) or as stimulatory mechanisms (e.g., growth enhancing factors), as well as
functioning indirectly by influencing competitive outcomes. Herbivory directly reduces fleshy-algal biomass, which indirectly (via
competitive release) favors the expansion of grazer-resistant reef-building corals and coralline algae. Because of the sensitive nature
of direct/indirect and stimulating/limiting interacting factors, coral reefs are particularly vulnerable to anthropogenic reversal
effects that decrease top-down controls and, concomitantly, increase bottom-up controls, dramatically altering ecosystem
resiliencies.
Published by Elsevier B.V.

Keywords: Algae; Nutrients; Herbivory; Corals; Coral reefs




1. Introduction                             and destructive fishing. It would appear that in regard to
                                     nutrients (NH4+, NO3À, NO2À and PO43À), the fewer
  Coral-reef ecosystems are adapted to conditions far          the better; with the opposite being the case for
removed from human influences, such as eutrophication           herbivores (parrotfishes, surgeonfishes, rudderfishes),
                                     where more are usually better. Under such conditions,
                                     coral reefs have evolved impressive levels of biological
 * Corresponding author. Tel.: +1 202 633 0956;
                                     diversity, including many uniquely specialized photo-
fax: +1 202 786 2563.                          synthetic symbionts and benthic algae. Four major
  E-mail address: littlerm@si.edu (M.M. Littler).           functional groups of benthic photosynthetic organisms

1568-9883/$ – see front matter. Published by Elsevier B.V.
doi:10.1016/j.hal.2005.11.003
566                      M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

are responsible for the bulk of coral-reef primary              proposed by Littler and Littler, 1984a) predicts that the
production: microalgal turfs (defined here as fleshy              competitive outcomes determining the relative abun-
filamentous and prostrate forms <2 cm high), frondose             dances of corals, crustose coralline algae, microalgal
macroalgae, calcareous crustose coralline algae and              turfs and frondose macroalgae on coral reefs are most
reef-building corals (containing symbiotic algae). Of             often controlled by the complex interactions of
these, cnidarian corals and coralline algae are the most           environmental factors (bottom-up controls such as
desirable due to their accretion of the CaCO3 matrix that           nutrient levels) and biological factors (top-down
comprises the reef framework, which is responsible for            controls such as grazing).
the spatial heterogeneity/complexity that supports the              Before any model can be useful, its predictions must
remarkable diversity of associated biota.                   accurately reflect the biological relationships in the target
  The concepts ‘‘top-down’’ and ‘‘bottom-up’’ con-              ecosystems. The previous evidence relevant to the RDM
trols have long been used (e.g., Atkinson and Grigg,             consists of several short-term experiments (e.g., Miller
1984; Carpenter et al., 1985) to describe mechanisms             et al., 1999; Thacker et al., 2001; Belliveau and Paul,
where either the actions of predators or resource               2002), in the case of bottom-up versus top-down effects,
availability regulate the structure of aquatic commu-             as well as considerable circumstantial evidence (e.g.,
nities; these opposing concepts can be particularly              Hallock et al., 1993; Hughes, 1994) and correlative
useful in understanding complex coral-reef ecosystems.            biogeographic surveys (Littler et al., 1991; Verheij,
The Relative Dominance Model (RDM, Fig. 1, first                1993). Using a longer-term manipulative approach on an




Fig. 1. The Relative Dominance Model. All of the four sessile functional groups depicted occur under the conditions in every compartment of the
model; however, the RDM predicts which groups will be predominant under the complex interacting vectors of eutrophication and declining
herbivory (most often anthropogenically derived). Crustose coralline algae are posited to be competitively inferior and dominate mainly by default;
where frondose algae are removed by herbivores and corals are inhibited by nutrients. The dotted lines represent tipping points where the external
forcing functions of increasing nutrients and declining herbivory reach critical levels that reduce resiliency to phase shifts. Light to dark shading
indicates declining desirability of each functional group from a management perspective. Hypothetically, one vector can partially offset the other
(e.g., high herbivory may delay the impact of elevated nutrients, or low nutrients may offset the impact of reduced herbivory). We further posit that
such latent trajectories can be activated or accelerated by large-scale stochastic disturbances such as tropical storms, cold fronts, warming events,
diseases and predator outbreaks; events from which coral reefs have recovered for millions of years in the absence of humans.
                    M.M. Littler et al. / Harmful Algae 5 (2006) 565–585               567

appropriately oligotrophic coral-dominated reef, Smith       non-calcifying macroalgae (Birkeland, 1987; Done,
et al. (2001) provide the most relevant experimental        1992; Lapointe et al., 1993, 1997; Lapointe, 1997;
evidence in support of the RDM to date.               NRC, 2000; Bellwood et al., 2004). Growth and
  Top-down control by abundant populations of large        reproduction of macroalgae are nutrient limited in
mobile herbivores has been shown repeatedly since the        oligotrophic coral-reef waters (Lapointe, 1987, 1997,
time of Stephenson and Searles (1960) for coral reefs.       1999; Larned and Stimson, 1996; Schaffelke and
As noteworthy examples, Carpenter (1986), Lewis           Klumpp, 1998; Lapointe et al., 2004) where low-
(1986), Morrisson (1988) and many other workers           nutrient concentrations and high herbivory favor the
(reviewed in Steneck, 1989; McCook, 1999; Bellwood         dominance of calcareous, hermatypic corals (Adey,
et al., 2004) have unanimously reported that lowering        1998; McConnaughey et al., 2000). Case studies in
herbivory without changing nutrient inputs often results      Kaneohe Bay, Hawaii, USA (Banner, 1974; Smith
in rapid increases in fleshy algae on coral reefs.          et al., 1981) and, more recently, the Negril Marine Park,
However, in most of the few studies that manipulated        Jamaica (Goreau, 1992; Lapointe and Thacker, 2002)
both herbivores and nutrients (e.g., Thacker et al., 2001;     have demonstrated the pivotal role of low-level nutrient
McClanahan et al., 2002; Belliveau and Paul, 2002), the       enrichment to the development of excessive macroalgal
duration was too short and adequate nutrient data were       biomass (ECOHAB, 1997) on coral reefs. Macroalgae
lacking, or ambient nutrient background concentrations       can inhibit the survival of coral recruits (Birkeland,
already exceeded levels limiting to macroalgal growth        1977; Sammarco, 1980, 1982) and because of enhanced
(e.g., Miller et al., 1999).                    growth and reproduction in the presence of elevated
  Despite many advocates, herbivory patterns alone do       nutrients, they can quickly overgrow the slower-
not consistently explain the distributions and abun-        growing hermatypic corals (NRC, 1995).
dances of benthic algae on coral reefs (Adey et al.,          Spatial and temporal patterns of nutrients also have
1977; Hay, 1981; Hatcher and Larkum, 1983; Hatcher,         been shown (Adey et al., 1977; Hatcher and Hatcher,
1983; Carpenter, 1986). For example, several studies        1981; Hatcher and Larkum, 1983) to co-vary with algal
(e.g., Hatcher, 1981; Schmitt, 1997; Lirman and Biber,       biomass. The decrease in coral cover (Pollock, 1928),
2000) found no significant correlation between grazing        relative to frondose algae (Doty, 1971) and coralline
intensity and frondose algal biomass. A dramatic          algae (Littler, 1971), on the reef flat at Waikiki, Hawaii
increase in fleshy algal biomass due to eutrophication        was the first phase shift from coral to macroalgal
was reported (Fishelson, 1973) without any concomi-         domination that was postulated (Littler, 1973) as due to
tant reduction in herbivore populations. As noted by        increases in eutrophication (bottom-up control). Shifts
Lewis (1986), frondose macroalgae occur in healthy         from coral dominance to algal dominance that suggest
reef areas of low herbivory (see also Littler et al., 1986);    linkages with chronic nutrient loading are exemplified
many such areas generate increased current accelera-        by case studies in Hawaii (Littler, 1973; Banner, 1974;
tion, like the reef crest and tops of patch-reef rocks,       Smith et al., 1981), Venezuela (Weiss and Goddard,
implicating higher nutrient fluxes (e.g., see Atkinson        1977), the Red Sea (Mergener, 1981), Barbados
and Bilger, 1992; Bilger and Atkinson, 1995). Further        (Tomascik and Sander, 1985, 1987a,b), Reunion Island
considerations are the widespread abundance of           (Cuet et al., 1988), Bermuda (Lapointe and O’Connell,
nitrogen-fixing Cyanobacteria and the now-ubiquitous         1989), the Great Barrier Reef (Bell, 1992), mainland
presence of substantial anthropogenic nitrogen sources       southeast Florida (Lapointe et al., 2005a,b), the Florida
(from burning fossil fuels) in rainfall worldwide          Keys (Lapointe et al., 1994), Martinique (Littler et al.,
(Vitousek et al., 1997)—making the terms ‘‘pristine’’        1993) and Jamaica (Goreau et al., 1997; Lapointe et al.,
or ‘‘nutrient-limited’’ relative, at best.             1997). The very low nutrient levels involved in limiting
  Coral reef ecosystems have evolved in the most          macroalgal growth (tipping points are the critical
oligotrophic of warm ocean waters and are sensitive to       nutrient levels that reduce resiliency to phase shifts),
low level increases in the concentrations of dissolved       either natural or anthropogenic, have been proposed
inorganic nitrogen (DIN = NH4+ + NO3À + NO2À) and          (Bell, 1992; Lapointe et al., 1997) regarding the
soluble reactive phosphorus (SRP = PO43À) associated        enabling of undesirable transitions from coral dom-
with human eutrophication (Johannes, 1975; Tomascik         inance toward algal stable states. Therefore, under-
and Sander, 1987a,b; Bell, 1992; NRC, 1995; Dubinsky        standing both the processes of productivity (bottom-up)
and Stambler, 1996). Nutrient enrichment of coral reefs       and those of disturbance (top-down) are critical to the
has many direct and indirect effects that, over time, can      elucidation of mechanisms that mediate algal/herbivore
result in alternative stable states dominated by fleshy,       interactions.
568                      M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

  The present 24-month investigation combines in situ            used to assess the herbivore resistances of predominant
experiments with field bioassays and descriptive               functional groups, including the massive reef-building
surveys to provide predictive information regarding             corals, as an independent test of the RDM’s efficacy.
the relative importance of bottom-up versus top-down             The controlled manipulative experiments examine the
controls on the dominant benthic functional groups on            importance of nutrient regime on long-term recruit-
coral reefs. The study includes: (1) characterization of           ment, colonization and competition patterns that
environmental parameters (i.e., nutrient analyses,              influence coral-reef community structure in habitats
herbivory assays and nutrient-limitation bioassays);             with contrasting levels of herbivory. Transplant studies
(2) distribution and abundance patterns of indicator-            test the growth/inhibition responses of reef-building
groups and their palatability to herbivores; and (3)             corals to elevated nutrients under natural levels of high
controlled manipulations of nutrient concentrations in            herbivory.
areas of both high and low herbivory. We believe that              In healthy tropical reefs, nutrient concentrations are
the strongest approach is to test multiple hypotheses            extremely low and attachment space is pre-empted by a
using multifaceted experiments. Both environmental              broad diversity of sessile benthic organisms. Given
and bioassay data are essential to characterize the             these conditions, competition between attached organ-
ambient nutrient/herbivory environments and antece-             isms should be severe. We posit that competition for
dent nutrient history of the two Study Sites (A and B,            space and light is not only important in determining the
Fig. 2). The nutrient-limitation bioassays provide              relative abundances of major functional groups, but also
physiological tests of the assumption that both Study            that the outcome of competition for these resources on
Sites A and B have had an oligotrophic history. This             coral reefs is often controlled by differential nutrient
type of assay furnishes a powerful index to the long-            and grazing effects. Controlled nutrient-enrichment
term integration of the ambient nutrient concentrations           experiments, utilized in conjunction with closely
by the naturally occurring functional producer groups            juxtaposed habitats of high versus low herbivores, test
prior to and following experimental enrichment. In the            hypotheses concerning the colonization and competi-
palatability assays, natural populations of reef fishes are          tive interactions of harmful blooms of microalgal turfs




Fig. 2. Location of the two main reef-flat Zones and the two Study Sites (A = low herbivory, B = high herbivory) and diffuser arrays (open = reduced
nutrients, closed = elevated nutrients) at CBC.
                   M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                 569

and frondose macroalgae versus beneficial reef-building       tract (James et al., 1976; Burke, 1982; personal
corals and crustose coralline algae on a healthy barrier-     observations). The community composition and zona-
reef system. The RDM (Fig. 1) provides the perspective       tional patterns of the CBC region are also representative
for advancing hypotheses and is examined by the          of much of the entire barrier reef platform (Littler et al.,
following four central predictions: In the high-          1989, 1995). Furthermore, distinct similarities exist
herbivory Study Site B (Fig. 2): (1) reduced nutrients       between the Belize Barrier Reef’s biological/geological
should favor the development of calcareous coralline        zonation and the barrier reefs of the north coast of
algae and corals relative to frondose macroalgae and        Jamaica (Goreau, 1959; Goreau and Land, 1974), the
microalgal turfs; and (2) elevated nutrients should result     north coast of Haiti (Burke, 1982), the southeastern
in high coverage of coralline algae; whereas in the low-      coast of Alarcran (Burke, 1982) and the offshore reefs
herbivory Study Site A; (3) elevated nutrients should       of the Bahamas, Puerto Rico, the Lesser Antilles,
lead to the dominance of frondose macroalgae; and (4)       Panama’s San Blas Islands, Mexico’s Yucatan Peninsula
reduced nutrients should lead to an abundance of turf       and the Bay Islands of Honduras (Littler and Littler,
microalgae.                            2000, personal observations).
                                   The bottom characteristics exhibit a shoreward (i.e.,
2. Materials and methods                      westward, downstream) transition from the smooth flat
                                  pavement zone adjacent to the crest to a rubble-
2.1. Study areas                          pavement zone (Fig. 2). These are followed by a thin
                                  overlaying veneer zone of rubble and gravel-sized
  The Belize Barrier Reef complex is the largest coral-      fragments (Littler et al., 1987b; Macintyre et al., 1987),
reef tract in the western hemisphere (over 250 km in        finally grading to an epilithic Thalassia-bed. The
length and from 10 to 32 km wide), consisting of an        Thalassia plants on this reef flat are firmly anchored
almost unbroken barrier reef containing hundreds of        directly to the pavement and secondarily entrap a thin
patch reefs and mangrove islands. Within back-reef         layer of gravel and coarse sand.
habitats, such as the one studied here (Fig. 2),           The back-reef pavement zone and rubble-pavement
assemblages of framework-building corals and calcar-        zone (Fig. 2) contain numerous coral colonies (Lewis,
eous algae have the same general taxonomic composi-        1986; Littler et al., 1989) and are characterized by high
tion along the entire barrier reef (Burke, 1982, personal     densities of transient herbivorous fishes (Hay, 1981;
observations). Carrie Bow Cay (CBC) reef habitats and       Lewis and Wainwright, 1985). Sea urchins and
surrounding environs comprise a well-developed,          territorial damselfishes are uncommon in the CBC
representative, barrier-reef system remote from major       back-reef areas studied (Lewis, 1986; personal observa-
human influences. Offshore Secchi disc depths in excess       tions). The most common herbivorous fish species in the
of 43 m are typical, indicating Jerlov Type I oceanic       outer Study Site B are: the surgeonfishes Acanthurus
waters. Most importantly, nutrient levels above the        bahianus and A. coeruleus, and the parrotfishes Scarus
tipping-point concentrations noted (Bell, 1992) to         inserti, Sparisoma chrysopterum, Sparisoma viride and
potentially enable macroalgal overgrowth (i.e.,          Sparisoma rupripinne. Repeated censuses from April
>0.1 mM phosphorus and >1.0 mM nitrogen) have           1982 to March 1983 (see Table 2 of Lewis, 1986)
seldom been recorded (Lapointe et al., 1987, 1993)         indicated reasonably stable herbivorous fish populations
from coral reefs of this system.                  and this pattern has continued to the present.
  The topography, geology and general biology of
CBC are well known due to over a quarter century of        2.2. Environmental data
study (see Ruetzler and Macintyre, 1982). Herbivory
has been extensively investigated for many of the CBC         To characterize the nutrient environment of CBC,
reef habitats (Hay, 1981; Littler et al., 1983b, 1986,       water samples were collected from each of the two Study
1987a, 1989, 1995; Lewis and Wainwright, 1985;           Sites (designated A and B, Fig. 2) in 100 ml acid-washed
Lewis, 1986; Lewis et al., 1987; Macintyre et al., 1987;      polyethylene bottles. Each sample was taken as three
Reinthal and Macintyre, 1994), including the sites         separate replicates (to increase coverage) and pooled (to
studied here. The two experimental Study Sites (A and       reduce analytical costs). Samples were obtained once
B, Fig. 2), located directly shoreward of the intertidal      yearly from 3 cm above the surface (i.e., top) of
and spatially complex reef crest on the northeast side of     individual clay-pot diffusers (see description below) 3
CBC (168480 N, 888050 W), are typical of the back-reef       weeks following the addition of fertilizer (N = 12
systems found throughout much of the Belizean barrier       separate samples of three pooled replicates each) in
570                  M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

each Study Site during midday. At the same time, an         harvested and, conversely, attract them in protected (no-
additional 12 concurrent samples were taken from 3 cm        fishing) reserves. Percent eaten was determined by re-
above non-enriched (control) diffusers to compare both       measuring the algal segments and the results were
natural and enriched levels of nutrients. The samples        analyzed using one-way ANOVA followed by the
were immediately filtered through combusted Gelman          Bonferroni (Dunn) t-test (SAS, 2003). Herbivorous fish
0.45 mm GF/F filters, placed in a cooler of ice and frozen      abundances were enumerated by counting numbers of
in the laboratory until analysis. Dissolved inorganic        individuals (by species), from mid-morning to mid-day
nitrogen (DIN = NH4+ + NO3À + NO2À) and soluble           throughout a typical spring day, 1 m on either side of 15,
reactive phosphorus (SRP = PO43À) concentrations were        10-m long, north-south, transect lines. Historical values
determined by the Nutrient Analytical Services Labora-       from previous literature (Hay, 1981; Lewis and Wain-
tory, Chesapeake Biological Laboratory, Solomons, MD.        wright, 1985) in the same locations were also re-
SRP and NO3À were measured with a Technicon             examined and tabulated with the current data set.
Autoanalyzer II. NH4+ and NO2À were measured using
a Technicon TRAACS 800. The detection limits for          2.4. Biotic distribution patterns
NH4+, NO3À plus NO2À and SRP were 0.21, 0.01 and
0.02 mM, respectively.                         A cluster analysis of the coral and macrophyte cover
  The current speeds at both sites were measured          was used to test the hypothesis that grazing intensity and
sporadically under typical non-storm wind and wave         algal characteristics that resist herbivory (e.g., calcifi-
conditions on 12 separate days during the 24-month         cation) are related to the natural distribution patterns of
study by fluorescent dye injected next to the nutrient        the dominant functional groups. A single transect on
diffusers on the bottom and timing the movement over a       compass heading 908 magnetic was established begin-
horizontal distance of 2.0 m. To further characterize        ning next to shore on the CBC reef flat in 0.2 m of water
water quality (light penetration), Secchi disc depths        and extending eastward to the reef crest at a distance of
were determined just to the east of the study areas in the     111 m. Quantitative samples were obtained by photo-
deeper waters bathing the reef flat, between 1000 and        graphing (perpendicular to the substrate) 0.15-m2
1100 h on 10 separate occasions.                  quadrats centered at every third meter mark from 0
                                  to 100, and at every meter mark thereafter. Due to the
2.3. Herbivory assays                        patchy nature of the biota, uniformly spaced quadrat
                                  arrays produced a more representative sampling than
  Natural levels of herbivory close to the experimental      would patchy (i.e., randomized) hit-or-miss arrays (see
arrays at the eastern transitional margin of Study Site A      discussion in Littler and Littler, 1985). Simultaneously,
(Fig. 2, relatively remote from structural shelter) and       voucher specimens of dominant macrophytes and turf
Study Site B (relatively closer to the shelter of the crest     microalgae were taken for taxonomic purposes. In the
structure, see diffuser locations in Fig. 2) were assayed      laboratory, the images were scored using a randomized
using the palatable test alga, Acanthophora spicifera.       grid of 100 dots (see Littler and Littler, 1985).
This ubiquitous red alga is a highly preferred food item        To describe the natural species assemblages along
by both parrotfishes and surgeonfishes (Lewis and           the transect in an unbiased manner, the cover data of
Wainwright, 1985), as well as by sea urchins (Littler        each species for all quadrats (those without organisms
et al., 1983b). The alga was cut into 7.0-cm lengths and      were excluded) were subjected to hierarchical cluster
attached to $3 Â 10-cm dead coral-rubble fragments by        analysis (flexible sorting, unweighted pair-group
thin (1-mm thick  5-cm long), dull-beige, rubber          method) using the Bray and Curtis (1957) coefficient
bands. Fifteen replicates were placed haphazardly in        of similarity. The resultant dendrogram of similar
each Study Site for 3 h. Additionally, 15 replicates of       quadrat groupings was based on the dominant biota and
the seagrass Thalassia testudinum were placed (using        environmental affinities and used to characterize zones
the above methods) in Study Site A to augment the data       that were predicted (a priori) to correlate with herbivory
that Hay (1981) collected only for Study Site B. This        levels.
technique avoided both pseudo-replication (non-inde-
pendence) and novelty effects (i.e., artifactual con-        2.5. Nutrient-enrichment assays
spicuousness) that could bias grazing patterns and rates.
We have personally observed that gaudy markers, or           Nutrient-enrichment bioassays tested the hypothesis
devices such as colored rope and surveyor’s tape, alarm       that both Study Sites had an oligotrophic antecedent
herbivorous fishes in areas where they are intensively        history. This procedure assayed the light-saturated net
                   M.M. Littler et al. / Harmful Algae 5 (2006) 565–585               571

photosynthetic rates (Pmax) of the most widespread        cm2 clumps were attached to independent rubble
macroalga (Dictyota pulchella) in the CBC study area.       fragments by thin dull-beige rubber bands and deployed
The Pmax response to DIN and SRP enrichment            at $0.5 m intervals in a randomized pattern (12
(detailed in Littler and Littler, 1990) was used as an      replicate clumps per each of the 10 species). Surgeon-
index to its long-term integration of the ambient         fishes and parrotfishes showed no wariness and began
nutrient concentrations prior to the experimental         feeding immediately, moving from clump to clump and
enrichment manipulations. Factorial experiments (6        feeding persistently as they located a particularly
replicateÁplants treatmentÀ1) included overnight (dark)      palatable species. The clumps were photographed
pulsing with DIN (as NH4+, 16.0 mM), SRP (as PO43À,        immediately after deployment and 6.0 h later. Quanti-
1.6 mM), both DIN + SRP and a control (no nutrients        fication of losses was determined digitally from the
added). The above concentrations were chosen to          photographs. Published values from a similar study near
saturate the uptake rates (see Lapointe, 1987) in the       the same location (Littler et al., 1983b) also were re-
small volumes used during nutrient pulsing (4-l freezer      examined and graphically included to augment the
bags). These concentrations represent realistic levels      present data set.
encountered in eutrophic reef environments (e.g., near
bird islands, Lapointe et al., 1993), and are an order of     2.7. Top-down versus bottom-up experiments
magnitude below levels characteristic of reef inter-
stitial pore waters used by rhizophytic macrophytes         To test the RDM, two sites [Study Sites A (72–77 m)
(i.e., 120–200 mM, Williams and Fisher, 1985). The        and B (92–97 m), Fig. 2] were established in the same
bioassays were performed at 12 month intervals in 1.0 l      structureless rubble-pavement Zone 2 but differing
incubation jars containing ambient seawater under         primarily in the levels of herbivorous fish activity; as
natural saturating irradiance levels (between 1000 and      determined by the patterns of biotic cover (see Fig. 3,
1300 h, 1400–2200 mmol photonsÁmÀ2ÁsÀ1, 27–29 8C         Table 3) and palatability (Fig. 4), as well as by
water temperatures), while vigorously mixed by water-       herbivorous fish densities and assays of herbivory
driven magnetic turbines to eliminate diffusion          (Table 2). Nutrients were manipulated in these same
boundary layers.                         environments using 4-l clay diffusers. Data were
                                 assessed within functional groups (i.e., relative abun-
2.6. Palatability experiment                   dances) as well as at the community level (i.e., relative
                                 dominances). The goal of these manipulative experi-
  Natural populations of reef fishes were used to assess     ments was to provide direct experimental tests of the
the herbivore resistances of eight predominant macro-       nutrient mediated interactions posited from the RDM.
phytes representing five morphological form groups as       Proximity to seaward reef-crest shelters (Fig. 2) also
well as two species of massive corals (to test the        provided a high level of fish herbivory that was further
following prediction and document how the herbivory        manipulated for 24 months with nutrient diffusers
component of the model works). Sea urchins are no         containing coral transplants (see below).
longer common in the CBC environs. If corals and           The low-herbivory Study Site A is not regularly
members of the calcified-crustose and jointed-calcar-       frequented by herbivorous fishes [because of the lack of
eous algal forms have evolved anti-herbivore defenses       both large- and small-scale structural shelter from
(e.g., toughness, structural inhibition, low calorific       carnivorous fishes (e.g., barracudas, sharks, jacks,
content or toxicity), then they should show the greatest     snappers) and birds (e.g., ospreys, herons, cormorants,
resistance to herbivory by generalist fish grazers with a     pelicans), which forage daily on the back reef (personal
gradient of increasing palatability toward the more        observations)]. Proximity to shelter has been long
fleshy thick-leathery, coarsely-branched and sheet-like      recognized (Randall, 1965; Ogden et al., 1973) as an
algal form groups (see Littler et al. (1983a) for         important factor determining herbivorous fish foraging
morphological characterization).                 ranges. Study Site B, established 15-m seaward (92–
  Experiments were run in the rubble-pavement zone        97 m) in the same rubble-pavement zone but closer to
(Study Site B, 95 and 100 m) of high herbivory (Fig. 2,      the shelter of the reef crest, is characterized by
Reinthal and Macintyre, 1994) just shoreward of the        exceptionally high fish herbivory (Macintyre et al.,
reef crest. The algae and corals were collected while       1987; Littler et al., 1989; Reinthal and Macintyre,
submerged and separated into approximately 10-cm2         1994). Because of the close juxtaposition of the two
clumps to avoid bias arising from a size-based          Study Sites, and otherwise physical/chemical/geo-
differential attractiveness to visual feeders. The 10-      morphic similarity (see Table 1, Lewis, 1986), the
572                      M.M. Littler et al. / Harmful Algae 5 (2006) 565–585




Fig. 3. Dendrogram display showing differential cluster analysis (using the Bray and Curtis (1957) Similarity Coefficient) for the percent cover of
dominant taxa, including all quadrats (labeled by distance from CBC shore) except those devoid of biota. Two major zones are indicated (see Table 3,
Fig. 2).

degree of fish herbivory is the overriding ecological             with cage artifacts (e.g., shading, alteration of current
variable (supported by the herbivory assays, extensive            flow, etc.) and the necessity for cage controls.
nighttime/daytime observations over a 25-year period             Furthermore, the exclusion of fish grazers by cages
and the biotic zonational patterns, see Tables 2 and 3            has been shown to promote fouling and also shelter
and Figs. 3 and 4). Both of these experimental sites are           benthic invertebrates from predation. Such potential
in the structurally homogeneous rubble-pavement zone             artifactual increases in the density of mesograzers and
that does not support damselfish or other potentially             fouling organisms (Dayton and Oliver, 1980) would
confounding organisms. Based on earlier work (Dayton             have been undesirable during the 2-year experiment.
and Oliver, 1980; Littler et al., 1989), cages were not             Within each of the two Study Sites (A and B, Fig. 2),
used as a method of choice due to well-known problems            eight, independent, terra-cotta, clay-pot, nutrient diffu-




Fig. 4. Susceptibility to fish grazing for species representing five macroalgal form groups (N = 12ÁspeciesÀ1). A = sheet-like, B = coarsely-
branched, C = thick-leathery, D = jointed-calcareous, E = calcified-crustose. The corals (=F) Siderastrea radians and Porites astreoides showed zero
losses during this experiment. Mean data (N = speciesÁgroupÀ1) on CBC form-group palatability values included from Littler et al. (1983b) are
indicated by asterisks. All form-group differences are significant (P < 0.05, Duncan’s Multiple Range Test), vertical lines = Æ1S.E.
                      M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                  573

Table 1
Environmental data for the Study Sites on the CBC back-reef flat (means Æ 1 S.D., N = 24 (12ÁyearÀ1))
Sites   Current        Depth range    Natural DIN    Enriched DIN    Natural SRP   Enriched SRP  Distance from
      speed (cmÁsÀ1)     (m)        levels (mM)    levels (mM)    levels (mM)   levels (mM)   shore (m)
Site B   $3.0–4.7        0.4–0.6      UD to 0.51    1.9–7.1      UD to 0.07    0.18–0.76    92–97
      (mean = 3.6 Æ 0.5)            (0.37 Æ 0.06)   (3.8 Æ 0.62)    (0.03 Æ 0.02)  (0.39 Æ 0.03)
Site A   $3.0–5.7        0.3–0.4      UD to 0.61    1.9–5.7      UD to 0.06    0.18–0.88    72–77
      (mean = 4.9 Æ 0.8)            (0.44 Æ 0.03)   (3.8 Æ 0.86)    (0.03 Æ 0.02)  (0.39 Æ 0.06)
DIN = dissolved inorganic nitrogen, SRP = soluble reactive phosphorous, UD = undetectable (not used in means).


sers (4-l volume, 15.5-cm high, 22-cm mouth diameter)          study period. To test the null hypothesis that the percent
were cemented upside down to the reef substrate at            cover differences of functional groups under elevated
>1.5 m distances from each other using marine epoxy           versus reduced nutrients were not statistically different
cement to completely seal the rims. These porous clay          (at alpha = P > 0.05), we used one-way ANOVA
diffusers had 1235 cm2 of total surface area, but only the        followed by Bonferroni (Dunn), a posteriori, multiple
220-cm2 flat top was sampled. Osmocote (Sierra              classification analysis (SAS, 2003). All percent cover
Chemical Co., California, USA) slow-release (9 months)          data were arcsine transformed prior to analysis. The
fertilizer containing 18% N (as ammonium nitrate and           same statistics were used separately to compare patterns
ammonium phosphate) and 6% P (as ammonium                between the two different Study Sites.
phosphate and calcium phosphate) was poured into four
elevated-nutrient diffusers (randomly selected for treat-        2.8. Coral transplant experiment
ment) from each of the two Study Sites until each diffuser
was completely full, and the hole was then stoppered. The          We concurrently conducted long-term (24-month
fertilizer was replenished at $3-month intervals to assure        duration) transplant studies (N = 8) of the two massive
ample delivery. The remaining four low-nutrient             coral species, Siderastrea radians and Porites
diffusers (ambient controls) in each Study Site were           astreoides, to assess their performances in the high-
filled with seawater and stoppered. Consequently, the           herbivory Study Site B under the two levels of nutrients
eight diffusers (four reduced nutrients and four elevated        used in the colonization/competition experiments.
nutrients in each Study Site) provided two experimental         Specimens were cut underwater into approximately
arrays that included randomly selected independent            2-cm2 ‘‘nubbins’’. Individual 2-cm2 samples of each
nutrient treatments exposed in two closely juxtaposed          coral species were transplanted (12 cm apart) using
Study Sites chosen for their extremes of herbivory. This         marine epoxy cement onto the tops of an additional 16
design yielded the following four combinations (N = 4)          haphazardly arrayed nutrient diffusers (>1.5 m separa-
of experimental conditions: (1) reduced nutrients/high          tion), all in the Study Site B rubble-pavement zone of
herbivory and (2) elevated nutrients/high herbivory in          high herbivory (Lewis, 1986). Eight diffusers were
Study Site B, in addition to (3) reduced nutrients/low          randomly selected to remain nutrient-free, while the
herbivory and (4) elevated nutrients/low herbivory in          interspersed remaining eight were filled with slow-
Study Site A.                              release Osmocote fertilizer that was replenished every 3
  Abundances of each colonizing group were deter-            months. The transplanted nubbins were initially
mined 24 months following initial set-up by making            photographed and then re-photographed after 24
detailed field estimates through magnifying lenses            months from the same distance and orientation so that
followed by taking macro-images of the top (center            changes in two-dimensional area could be scored and
108-cm2, 9 cm  12 cm framer) of each diffuser. The           compared between the treatments (one-way ANOVA,
images were scored for percent cover of predominant           Bonferroni).
taxa (see details in Littler and Littler, 1985). The high
magnification afforded by macro-photography of the            3. Results
108-cm2 plots enhanced the resolution and, in
conjunction with the field notes, facilitated discrimina-         3.1. Environmental data
tion of microscopic turf species and crusts. Compar-
isons were made between treatments to detect changes            The DIN and SRP concentrations next to the non-
in the relative abundances of the benthic groups that          enriched diffusers (Table 1) are barely detectable in both
recruited, colonized and persisted over the 24-month           Study Sites (i.e., Study Site B, means = 0.37 Æ 0.06 S.D.
574                     M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

mM DIN and 0.03 Æ 0.02 mM SRP; Study Site A,               Zone 1) and 15 m nearer the reef crest at 92 and 97 m
means = 0.44 Æ 0.03 mM DIN and 0.03 Æ 0.02 mM               (Study Site B) on the CBC back-reef flat. The assay using
SRP), indicating oligotrophic conditions. Conversely           the palatable seaweed A. spicifera shows grazing rates
in both Study Sites B and A, the nutrient diffusers filled         that are 17 times greater in Study Site B than in Study Site
with slow-release fertilizer show nearly identical results        A. In agreement, Study Site B on the outer reef flat
(Table 1), significantly increasing DIN by 10-fold to           contains 145-fold more surgeonfish and 148-fold more
means of 3.80 Æ 0.62 and 3.80 Æ 0.86 mM and SRP by            parrotfish than Study Site A (Table 2). In support of these
13-fold to means of 0.39 Æ 0.03 and 0.39 Æ 0.06 mM at           differences, experimental assays of herbivory by others
about 3 cm above the experimental substrates. These            (Table 2) prior to this study show 15 times greater loss of
enriched values (Table 1) exceed the kinetic levels            A. spicifera in Study Site B than in Study Site A
(tipping points), noted by Bell (1992) and Lapointe et al.        (P < 0.01, Kruskal–Wallis), and 62 times greater loss of
(1993) for releasing inhibition of algal growth on coral         the palatable seagrass T. testudinum (P < 0.01, Table 2).
reefs, by approximately 3- to 4-fold.
  Predominant current speeds are reasonably constant in         3.3. Biotic distribution patterns
a northwesterly direction (3408 magnetic, Table 1),
ranging from 3.0 to 5.7 cmÁsÀ1 (mean = 4.9 Æ 0.8 S.D.)            Cluster analysis of the percent cover transects
in Study Site A and 3.0 to 4.7 cmÁsÀ1 (mean = 3.6 Æ 0.5          (Fig. 3) establishes the existence of two major biotic
S.D.) in Study Site B. These currents are driven by the          zones on the back-reef flat at CBC (Table 3), dominated
pumping action of offshore waves breaking over the reef          by benthic indicator groups corresponding to gradients
crest and slowly flowing westward through the Study            in herbivory and prey palatability (Table 2, Fig. 4). The
Sites and Zones (Fig. 2), exiting around the northern tip of       landward Zone 1 (Fig. 3) on the back-reef flat between 0
the island. Secchi disc depths seaward of CBC average           and 72 m from the shoreline, remote from herbivorous
43 Æ 3 S.D. m (N = 10, range = 38–47 m), indicating            fish activity (see Macintyre et al., 1987; Reinthal and
exceptionally clear, Type I (Jerlov, 1976), oceanic waters        Macintyre, 1994) and extending over a carbonate
consistently cascading onshore over the study area.            pavement substrate (thinly covered by sand and gravel)
                                     from 0.2 to 0.5 m in water depth (mean = 0.3 m),
3.2. Herbivory assays                           includes a discrete grouping of quadrats with a high
                                     level of similarity (Bray-Curtis Index). Total plant cover
  Large and significant (P < 0.0001, F = 53.28,             averages 70.3% and the palatable macrophyte T.
d.f. = 14, Bonferroni) differences in herbivory are            testudinum (Table 2) is dominant (Table 3, average
present (Table 2) within Zone 2 between 72 and 77 m            cover 60%, with maxima >100% due to layering).
(Study Site A, just beyond the outer transitional edges of        Sediments in this shallow grass-bed system (i.e.,


Table 2
Herbivorous fish densities and grazing intensity for the Carrie Bow Cay outer Thalassia zone (Study Site A, 72–77 m) and rubble-pavement zone
(Study Site B, 92–97 m) Study Sites (see Fig. 2)
Study and taxa              Study Site A                        Study Site B
                        À2                À1
                     NÁm          Percentage lossÁh          NÁmÀ2         Percentage lossÁhÀ1
Present investigation
 Acanthophora spicifera         –           1.0                 –           16.5
 Scaridae                0.001         –                  0.180         –
 Acanthuridae              0.001         –                  0.164         –
Lewis and Wainwright (1985)
 Acanthophora spicifera         –           0.7                 –           10.2
 Scaridae                0.001         –                  0.115         –
 Acanthuridae              0.001         –                  0.126         –
Hay (1981)
 Thalassia testudinum          –           0.5a                –           30.9
All percentage loss values (N = 15) between Study Sites are significantly different (P < 0.01 for previous studies, Kruskal–Wallis; P = 0.0001,
F = 53.28 for the present study, Bonferroni).
a
  Additional data from the present investigation.
                      M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                     575

Table 3
Mean percent cover (Æstandard error) of the dominant macrophyte and massive coral taxa in reef-flat habitats of low (Zone 1) and high (Zone 2)
herbivory (see Table 2)
Dominant taxa                                   Zones
                                         1 (N = 25)                  2 (N = 28)
                                         0–72a                     73–111a
                                         0.2–0.4b                   0.1–0.8b
Massive corals                                   0.03 Æ 0.03                 16.60 Æ 2.75
Corallines and other calcified algae (totals)                    8.86                     46.70
 Porolithon pachydermum (Fosl.) Fosl.                       1.27 Æ 0.81                 19.95 Æ 3.99
 Hydrolithon boergesenii (Fosl.) Fosl.                      0.20 Æ 0.08                 16.77 Æ 2.93
 Amphiroa rigida var. antillana Børg.                       6.55 Æ 1.64                  0.00
 Halimeda opuntia (L.) Lamour.                          0.33 Æ 0.26                  4.31 Æ 1.02
 Jania capillacea Harvey                             0.10 Æ 0.10                  2.94 Æ 0.84
 Jania adhaerens Lamour.                             0.25 Æ 0.07                  2.68 Æ 1.01
 Peyssonnelia sp.                                 0.01 Æ 0.01                  0.05 Æ 0.05
 Neogoniolithon strictum (Fosl.) Setch. et Mason                 0.15 Æ 0.10                  0.00
Fleshy macrophytes (totals)                            61.43                     0.59
 Thalassia testudinum Banks ex Koenig                      60.10 Æ 5.50                  0.00
 Caulerpa racemosa (Forssk.) J. Ag.                        0.60 Æ 0.31                  0.46 Æ 0.21
 Dictyota pulchella Lamour.                            0.42 Æ 0.09                  0.00
 Lobophora variegata (Lamour.) Womersley                     0.15 Æ 0.15                  0.00
 Dictyota sp.                                   0.14 Æ 0.10                  0.10 Æ 0.10
 Neomeris annulata Dickie                             0.02 Æ 0.01                  0.03 Æ 0.01
a
   Distance from shore (m).
b
   Depth range (m).

Landward Zone 1, Table 3) are, hypothetically, more an          with comparable assays of other species on the Belize
effect of T. testudinum abundance, rather than a cause,          Barrier Reef (Lapointe et al., 1987), suggest severe
since the rhizomes are anchored directly to reef             antecedent nutrient limitation in the two CBC Study
pavement. Massive corals average only 0.03% cover             Sites. In the first year, SRP was most limiting with
in the Landward Zone 1.                          significant (P < 0.05) positive interactions of both
  Seaward Zone 2, between 73 and 111 m (Fig. 1, depth          nutrients combined, whereas during the second year,
range 0.1–0.8 m, mean = 0.3 m), includes the rubble-           DIN was most limiting.
pavement zone (containing Study Sites A and B), the
deeper (0.8 m) pavement zone and the inner slope of the          3.5. Palatability experiment
reef crest and is dominated by grazer-resistant calcareous
macroalgae and reef-building corals. Total plant cover            The palatability assay reveals a consistent pattern
averages 47.3% (almost all calcareous forms, Table 3)           (also noted by Littler et al., 1983a,b) regarding grazer
with the primary species being the crustose corallines          resistances of coral and algal form groups as follows
Porolithon pachydermum (19.9%) and Hydrolithon              (Fig. 4). The algal sheet forms are less resistant (100%
boergesenii (16.8%). The grazer-resistant, calcareous,          lostÁ6 hÀ1) than the coarsely-branched forms (76%
green alga Halimeda opuntia (4.3%) is conspicuous in           lost), the thick-leathery forms (30%), the jointed-
patches on the shallow leeward (inner) slope of the reef-         calcareous forms (6%), the calcified-crustose forms
crest area. Also abundant in Seaward Zone 2 are the            (3%) and massive corals (0%). All of the differences
massive corals (16.6% cover).                       between form groups are significant (P < 0.05, Dun-
                                     can’s Multiple Range Test).
3.4. Nutrient-enrichment assays
                                     3.6. Bottom-up manipulations in sites of high and
  The Pmax of the common reef-flat macroalga D.              low herbivory
pulchella (Fig. 5) shows significant (P < 0.05, Bonfer-
roni) effects of DIN and SRP enrichment during two              In these experiments, algal recruitment and sub-
assays conducted 12 months apart, with a greater overall         sequent encroachment interactions were rapid, with
effect in the second year. These results, in conjunction         multi-layered cover values approaching or exceeding
576                      M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

                                      senii and P. pachydermum). However, elevated nutrients
                                      partially offset the effects of high herbivory, with 16%
                                      more frondose macroalgal cover and 22% more
                                      microalgal turf cover (both significant at P = 0.0001)
                                      than in the reduced nutrient treatment. All three of the
                                      above abundance peaks (Table 4) were statistically
                                      greater (P < 0.05, Bonferroni) under the conditions
                                      inferred by the RDM. Colonization on the reduced-
                                      nutrient treatments in the high-herbivory Study Site B
                                      consisted of only trace patches of microalgal turfs and
                                      coralline crusts, with frondose macroalgae being only
                                      slightly more conspicuous (all values significantly
Fig. 5. Productivity assays (net photosynthesis g organic dry       lower than in other treatments, Table 4). Corals did not
weightÀ1 hÀ1, N = 6) of the predominant macrophyte Dictyota pul-
chella as a function of dissolved inorganic nitrogen (DIN) and soluble
                                      colonize any of the diffusers during the study, but were
reactive phosphorus (SRP) enrichment initially (unshaded histograms)    investigated by the separate coral transplant experiment
and 12 months later (shaded histograms). Differences significant from    below.
non-enriched controls (C, P < 0.05, Bonferroni) are indicated by        All of the cover maxima within each functional
asterisks, vertical lines = Æ1S.E.                     group were greater under the conditions predicted by
                                      the model (Table 4, A–D). In terms of relative
                                      dominance between groups, there was only a single
100% under most treatment combinations (three out of            case that was contrary to the predicted RDM patterns
four; Table 4, A–D). Under elevated nutrients in the low          (see Fig. 6 and Table 4); i.e., crustose corallines were
herbivory Study Site A (Table 4), the frondose               slightly more prevalent (insignificant at P > 0.05,
macroalgae (mostly D. pulchella, along with Gelidiop-           Table 4, column A) than algal turfs under low nutrients
sis spp., Coelothrix irregularis, Padina jamaicensis,           in the low-herbivory Study Site A.
Turbinaria turbinata and Laurencia papillosa) with
64% cover became predominant (significant at                3.7. Coral transplant experiment
P = 0.004, F = 5.14, d.f. = 3). Turf microalgae (mostly
Cyanobacteria, small Digenea simplex, Jania capilla-             The S. radians and P. astreoides transplanted to 16
cea, J. adhaerens, Vaughaniella-stage of Padina,              additional independent nutrient diffusers (in the rubble-
Centroceras clavulatum and Heterosiphonia spp.)              pavement zone between 90 and 97 m, where high grazing
attained their cover maxima (37% cover) following             restricted fleshy algae to trace quantities) showed
24 months of reduced nutrient concentrations in the low          significantly reduced (P = 0.0001, F = 49.09, d.f. = 7
herbivory Study Site A, although crustose corallines            and P = 0.0001, F = 11.68, d.f. = 7, respectively) cover
were slightly more abundant at 41% cover. Conversely,           increases in the elevated-nutrient treatments versus the
in the high-herbivory Study Site B, the elevated-nutrient         reduced-nutrient treatments (Fig. 6), consistent with the
conditions resulted in dominant cover values for              model. Interestingly, S. radians experienced significant
crustose coralline algae with 72% cover (significant            inhibition (=net losses, P = 0.0001) under elevated
at P = 0.0001, F = 89.74, d.f. = 3; mostly H. boerge-           nutrients.


Table 4
Mean percent cover (Æstandard error) of benthic functional groups colonizing clay diffusers following 24 months under reduced and elevated
nutrients in low- and high-herbivory Study Sites (N = 4)
Functional groups        Study Site A (low herbivory)        Study Site B (high herbivory)     Significant differences
                 Nutrients                  Nutrients               (P < 0.05)
                 Reduced        Elevated       Reduced         Elevated
                 A           B           C            D
Crustose corallines       41.2 Æ 4.6       1.8 Æ 1.8      <0.1          71.7 Æ 3.0   D > A > B, C
Frondose macroalgae       20.8 Æ 4.3      63.7 Æ 8.2       0.6 Æ 0.3       16.9 Æ 4.1   B > A, D > C
Algal turfs           37.1 Æ 3.9      14.5 Æ 4.7      <0.1          22.1 Æ 2.9   A>D>B>C
Predicted dominants       Turfs         Macroalgae      Corals         Corallines
                        M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                577

                                      inhibited. Orthophosphate is known (Simkiss, 1964) to
                                      inhibit CaCO3 crystal formation at concentrations
                                      above 0.01 mM and can block deposition of external
                                      skeletal materials in some marine animals. The 50%
                                      suppression of community calcification and stimulation
                                      of algal overgrowth (Kinsey and Domm, 1974; Kinsey
                                      and Davies, 1979) subsequent to the experimental
                                      fertilization of a patch reef at One Tree Island on the
                                      Great Barrier Reef, Australia is partly attributable to
                                      phosphate poisoning. A sophisticated experiment on a
                                      larger and more carefully controlled scale (Larkum and
Fig. 6. Increases and decreases (cm2) in living surface area of nubbins   Koop, 1997; ENCORE Program) did not produce
(N = 8) of Siderastrea radians and Porites astreoides, transplanted     supporting results because: (1) ambient nutrient levels
onto clay diffusers under reduced (unshaded) and elevated (shaded)     within the lagoon at One Tree Island are now well above
levels of nutrient enrichment, after 24 months between 90 and 97 m in    tipping-point concentrations that are inhibitory to some
the rubble-pavement zone fully exposed to high herbivory. Differences
within species are significant (P = 0.0001, Bonferroni), vertical line-
                                      corals, while being more than sufficient to support the
s = Æ1S.E.                                 existing luxuriant frondose macroalgal community
                                      (Bell, 1992; Larkum and Koop, 1997) and (2) the
                                      experimental organisms were isolated on raised grids,
4. Discussion—functional groups as indicators of              precluding natural encroachment, overgrowth or other
reef health                                 key competitive interactions crucial to testing the RDM.
                                      The challenge now is to rigorously conduct this type of
4.1. Reef-building corals                          large-scale manipulation in an extreme oligotrophic
                                      coral-reef setting (e.g., Smith et al., 2001), in
  Cnidarian corals, the architects of structural dimen-          conjunction with staged competitive bouts among the
sionality, while preyed upon by a few omnivorous fishes           major functional groups, to determine how herbivore/
and specialist invertebrates (e.g., crown-of-thorns sea           nutrient interactions affect relative dominances over a
star, corallivorous gastropods), generally achieve             long time scale.
dominance under the control of intense herbivory
(Lewis, 1986) and extremely low nutrient concentra-             4.2. Crustose coralline algae
tions (Bell, 1992; Lapointe et al., 1993). Massive corals
consistently prove to be the most resistant to grazing at           In contrast to the corals (and fleshy algae), crustose
the highest levels of herbivory (Figs. 4 and 6). S. radians         coralline algae tend to be slow-growing competitively
and P. astreoides, hard mound-shaped forms, show little           inferior understory taxa that are abundant in most coral-
colony mortality under high grazing pressure (Fig. 6),           reef systems (Littler, 1972); although the group includes
even though occasionally rasped by parrotfishes (see             forms ranging from thin early-successional flat sheets to
also Littler et al., 1989). Contrastingly, some delicately         long-lived massive branched heads. Their critical roles
branched corals such as Porites porites are quite              in coral reefs is to form the protective algal ridge/reef
palatable and readily eaten by parrotfishes (e.g., S.            crest, cement the dead coral and other carbonate
viride, Littler et al., 1989; Miller and Hay, 1998).            fragments into a stable framework and, by sloughing
However, many hermatypic corals are inhibited by              (Littler and Littler, 1997), prevent propagules of fouling
increases in nitrate (e.g., Montastrea annularis and P.           organisms from colonizing. Crustose corallines,
porites, Marubini and Davies, 1996), ammonia (e.g.,             because of their slow growth rates, tolerate reduced
Pocillopora damicornis, Stambler et al., 1991; Muller-           nutrient levels and generally are conspicuous, but not
Parker et al., 1994) and orthophosphate (e.g., Porites           dominant, at low concentrations of nutrients and high
compressa, Townsley cited in Doty, 1969; P. damicornis           levels of herbivory (Littler et al., 1991). Accordingly,
and Stylophora pistillata, Hoegh-Guldberg et al., 1997).          they do well under both low and elevated nutrients (i.e.,
Nutrient inhibition of coral larval settlement also is           most are not inhibited by nutrient stress and many are
known for Acropora longicyathis (Ward and Harrison,             maintained competitor-free by surface cell-layer shed-
1997). Nutrient poisoning is probably the case for S.            ding (Johnson and Mann, 1986; Keats et al., 1994), even
radians in this study where growth inhibition is              at lower levels of grazing (Littler and Littler, 1997).
apparent (Fig. 6); whereas, P. astreoides was severely           Their ability to dominate is largely controlled indirectly
578                  M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

by the factors influencing the abundances of the other       however, contrary to several literature citations, no
groups, primarily corals and fleshy algae. In this study,      significant increases occurred in any of the major
crustose corallines were shown to predominate mainly        macroalgal species such as Turbinaria turbinata and
by default (i.e., under conditions of minimal competi-       Halimeda spp.
tion), where either corals were inhibited by elevated
nutrients (Fig. 6) or fleshy algae were removed by         4.4. Frondose macroalgae
intense herbivory (Table 2, Fig. 4). In independent
corroborations of the RDM, a gradient from frondose-          Terrestrial plant abundances and evolutionary stra-
to turf- to coralline-algal groups correlated closely with     tegies theoretically (Grime, 1979) are controlled by
escalating herbivory (Steneck, 1989); and increased sea      physiological stresses (external factors that limit
urchin populations under elevated nutrients (Lapointe       production) coupled with disturbances (factors that
et al., 1997) at Discovery Bay, Jamaica resulted in a       physically remove biomass); a concept expanded to
dramatic frondose macroalgal to crustose coralline algal      apply to marine macroalgae (Littler and Littler, 1984b;
phase shift (Aronson and Precht, 2000).              Steneck and Dethier, 1994). In the RDM (Fig. 1),
                                  nutrients (bottom-up) control production and grazing
4.3. Turf microalgae                        (top-down) physically reduces biomass of undesirable
                                  fleshy algal overgrowth. We demonstrate experimen-
  Low-stature turf microalgae tend to become domi-        tally that distributions and abundances of functional
nant under minimal inhibitory top-down and stimula-        groups on tropical coral reefs result from bottom-up
tory bottom-up controls (Table 4). Their relatively small     forces that affect metabolic production and growth (i.e.,
size and rapid regeneration/perennation results in         nutrients—mainly SRP and DIN); however, as shown
moderate losses to herbivory at low grazing pressures.       (Table 2), patterns vary between habitats (i.e., Study
Turf microalgae have opportunistic life-history char-       Site B) having beneficial counterbalancing top-down
acteristics, including the ability to maintain substantial     forces that limit or remove detrimental algal biomass
nutrient uptake and growth rates under low-nutrient        (i.e., herbivores—mainly Scaridae, Acanthuridae and
conditions (Rosenberg and Ramus, 1984). Convincing         Kyphosidae).
evidence also is afforded by large-scale mesocosms           Most importantly, we found a complexity of
with controlled low-herbivory and reduced water-          stimulation/inhibition interactions acting either directly
column nutrient regimes (McConnaughey et al.,           or indirectly (see also McCook, 1999). For example, our
2000), where turf algae invariably dominate due to         data reveal that insufficient nutrients act directly to
the inclusion of low-lying, microscopic, nitrogen-fixing      inhibit (limit) fleshy-algal domination (via physiologi-
Cyanobacteria as a source of within-turf nutrients. In       cal stress, Fig. 5); conversely, abundant nutrients
agreement, algal turfs have been shown to be favored        stimulate (enhance) fleshy-algal growth, with the
under reduced nutrient-loading rates (Fong et al., 1987)      opposite effect on reef-building corals (via toxic
or episodic nutrient pulses (Fujita et al., 1988) and this     inhibition? (Fig. 6), see Marubini and Davies, 1996).
can lead to extensive, two-dimensional, horizontal         Furthermore, the effects of controls also can be indirect
mats. Numerous other studies have shown the expan-         by influencing competition. Even this seemingly
sion of algal turfs, not macroalgae, resulting from the      indirect control can have further levels of complexity
removal of fish or echinoid grazers in a wide variety of      because competition between algae and corals can be
oligotrophic sites worldwide, including the Red Sea        direct (e.g., overgrowth) or indirect (e.g., pre-emption
(Vine, 1974), Fiji (Littler and Littler, 1997), Belize       of substrate). Low nutrients and high herbivory (via
(Lewis, 1986), the Great Barrier Reef (Sammarco,          physical removal) also act indirectly on fleshy algae
1983; Hatcher and Larkum, 1983; Klumpp et al., 1987)        through reduced competitive abilities; whereas, lowered
and Saint Croix (Carpenter, 1986). In the study of Lewis      herbivory and elevated nutrients also indirectly inhibit
(1986) on the same reef flat studied here, increases in an     (limit) corals (e.g., Banner, 1974; Birkeland, 1977) and
algal turf form with its upright Padina blades, not        coralline algae (e.g., Littler and Doty, 1975; Wanders,
blooms of mixed macroalgae, followed short-term (11-        1976) by directly stimulating (enhancing) fleshy-algal
wk) reductions of herbivorous fish grazing under          competition. With an increase in nutrients, the growth of
conditions of low nutrient levels. Lewis’ (1986) Table       fleshy algae is favored over the slower-growing corals
4 shows statistically significant, although relatively       (Table 4, Genin et al., 1995; Miller and Hay, 1996;
small, increases (28%) in algal turfs such as the above      Lapointe et al., 1997) and the latter can become
Vaughaniella-stage and its frondose form Padina;          inhibited by either poisoning (direct effect, Fig. 6) or, as
                    M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                579

mentioned, by competition for space and light (indirect       (Precht and Miller, in press), diseases (e.g., Littler and
effect, Jompa and McCook, 2002). Other ecologically         Littler, 1997; Santavy and Peters, 1997; Nugues et al.,
important factors, such as light regime, abrasion,         2004) and predator outbreaks (e.g., Cameron, 1977),
allelopathy and sediment smothering (e.g., Littler         which typically trigger or accelerate such low-resilience
et al., 1983c; Ruyter van Steveninck, 1984; Chadwick,        reef systems (Scheffer et al., 2001; Bellwood et al.,
1988; Coen, 1988; Coles, 1988; Keats et al., 1997;         2004) toward the long-term externally-mediated phase-
Littler and Littler, 1997), also can indirectly influence      shifts postulated in the RDM. For completeness, we also
further outcomes of competition.                  point out the obvious devastating effects of sedimenta-
  On healthy oligotrophic coral-reefs, even very low        tion (land-based and dredging), toxic spills, carbonate
nutrient increases (Tables 1 and 4) may exceed critical       mining and landfill. Such catastrophic events selectively
tipping-point levels that can shift relative dominances       eliminate the longer-lived organisms in favor of fast-
by releasing macroalgal production from nutrient          growing early-successional macroalgae (Littler and
limitation. Birkeland (1977) also noted that filamentous       Littler, 1984b), which can prevent settlement of coral
and frondose algae can outcompete corals, some of          planulae and become competitively superior (Birke-
which are inhibited under elevated nutrient levels         land, 1977; Lewis, 1986) to persist as alternative stable
(reviewed in Marubini and Davies, 1996, Fig. 6). Fast-       states.
growing algae are not just opportunists that depend on         The macroalgal overgrowth recorded under the
disturbances to release space resources from established      elevated-nutrient treatments in the low herbivory Study
longer-lived populations (cf. McCook, 1999), but,          Site A (Table 4) demonstrates that the tipping-point
hypothetically, become the superior competitors when        nutrient concentrations needed to support substantial
provided with abundant nutrients. Macroalgae, such as        primary production are quite low, but comparable to
Halimeda, also gain competitive advantage by serving        those reported for other tropical marine algae. For
as carriers of coral disease (Nugues et al., 2004).         example, several controlled, high-flux, continuous-
Potential competitive dominance of fast-growing           culture laboratory experiments and detailed field studies
macroalgae is inferred from their overshadowing           have demonstrated the physiological basis for low-
canopy heights, as well as from inverse correlations        nutrient tipping-points (i.e., $0.5–1.0 mM DIN) leading
in abundances between algae and the other benthic          to macroalgal blooms. The tropical rhodophytes
functional groups (Lewis, 1986; Bellwood et al., 2004),       Gracilaria foliifera and Neoagardhiella baileyi
particularly at the higher nutrient concentrations (e.g.,      (DeBoer et al., 1978) and the chlorophyte Ulva fasciata
Littler et al., 1993; Lapointe et al., 1997). Turbulent       (Lapointe and Tenore, 1981) all achieved maximal
water motion driven by wind and wave action can be         growth rates in continuous cultures at DIN levels of
sufficient to reduce oligotrophic boundary-layer diffu-       approximately 0.5–0.8 mM. Comparable low nutrient
sion gradients and increase delivery rates to support        levels (i.e., $0.10 mM SRP, $1.0 mM DIN) have been
considerable macroalgal biomass (e.g., Atkinson and         correlated with macroalgal blooms and the subsequent
Bilger, 1992), but the abundant herbivores may mask         decline of coral reefs from eutrophication at Kaneohe
these effects. The fleshy algal form groups (both micro-       Bay in Hawaii, fringing reefs of Barbados and inshore
and macro-) are particularly vulnerable to herbivory        reefs within the Great Barrier Reef lagoon (Bell, 1992),
(Table 4, see also Hay, 1981; Littler et al., 1983a,b) and,     as well as the macroalgal-dominated reefs of the
in accordance with the predictions of the RDM, only         Houtman Abrolhos Islands off Western Australia
become abundant in habitats where grazing is low. Such       (Crossland et al., 1984). These low nutrient concentra-
over-compensation by herbivory may explain some of         tions were also experimentally corroborated (Lapointe
the reported cases (e.g., Crossland et al., 1984; Szmant,      et al., 1993) for macroalgal overgrowth of seagrass and
1997; Glynn and Ault, 2000) of specific corals surviving       coral reef communities along natural nutrient gradients
in high-nutrient reef environments.                 on the Belize Barrier Reef. We recognize that coral reef
  The complex interactions of herbivory and nutrients       organisms can tolerate higher levels of DIN and SRP;
can change gradually with no apparent effects to induce       however, these nutritional levels represent tipping-point
subtle declines in resiliencies of coral/coralline-domi-      concentrations that reduce resiliency to a point at which
nated reef systems (Scheffer et al., 2001). These          coral-reef ecosystems can potentially shift towards
systems then become vulnerable to catastrophic impacts       dominance by fleshy algae.
by large-scale stochastic disturbances such as tropical        Tropical reefs in different geological systems have
storms (e.g., Done, 1992), warming events (e.g.,          contrasting patterns of photosynthetic nutrient limita-
Macintyre and Glynn, 1990; Lough, 1994), cold fronts        tion in regard to nitrogen and phosphorus availability
580                  M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

(Littler et al., 1991; Lapointe et al., 1992). Such patterns    induced shifts (long predicted by the RDM—see also
of nutrient limitation have been correlated (Littler et al.,    Fig. 2a in Bellwood et al., 2004 and anticipated by
1991) with the biogeographic distributions of major         Nixon, 1995) will expand geographically at an
groups of epilithic photosynthetic organisms that were       accelerated pace.
consistent with the RDM (see also Verheij, 1993).
Long-term ecological studies coincide in reporting         5. Conclusions
general worldwide declines in live coral cover and
concomitant increases in macroalgal abundances             By simultaneously using multifaceted descriptive
(Ginsberg, 1993; Birkeland, 1997). These kinds of          and experimental approaches, conducted for a sufficient
biotic phase shifts have been steadfastly attributed        duration on a healthy coral-dominated reef, this study
solely to over-fishing and diseases of herbivore stocks       provides the critically-needed long-term data to begin to
(e.g., see Hughes, 1994 on trends in Jamaican reefs over      close the historically-polarized intellectual rift invol-
the past 20 years, Hughes et al., 1999); however, such       ving the importance of eutrophification versus herbivore
shifts more-often-than-not occur in concert with          overfishing in causing coral to algal phase shifts. We
cultural eutrophication (Goreau et al., 1997; Lapointe       found (Table 4), as have others, that reduced nutrients
et al., 1997, 2005a,b). The spatial/temporal changes in       alone do not prohibit fleshy algal growth when
the patterns of algal dominance with regard to local        herbivory is low, and that high herbivory alone does
nutrient inputs in Jamaica and reefs around the world        not prevent fleshy algal growth when nutrients are
(Goreau, 1992, 2003; Goreau and Thacker, 1994)           elevated. However, reduced nutrients in combination
undermine the sea-urchin demise and overfishing           with high herbivory virtually eliminate all forms of
explanations, while lending further empirical support        harmful micro- and macro-algae. It is our opinion that
to the RDM.                             on the few remaining undisturbed, oligotrophic, coral-
  It is encouraging that the critical role of excess       reef systems, the effects of top-down inhibitory controls
nutrients on coral reefs has begun to receive appropriate      via intense herbivory prevail; whereas, bottom-up
recognition in recent review papers (Scheffer et al.,        stimulatory controls are less prevalent, due to the lack
2001; Hughes et al., 2003; Bellwood et al., 2004;          of nutrient availability and over-compensatory con-
Pandolfi et al., 2005). Although, some scientists (e.g.,       sumption by grazers. However, eutrophic systems may
Precht and Miller, in press) continue to downplay          lose their resiliency to inundation by macroalgae, with
human-induced declining resiliency issues, instead         herbivores becoming swamped by bottom-up (nutrient-
invoking unmanageable stochastic factors like upwel-        induced) harmful algal blooms. We show that the
lings, hurricanes and cold fronts (see Fig. 1 caption);       growth of reef-building corals can be inhibited under
events from which coral reefs have recovered for          elevated nutrients relative to low nutrients (Fig. 6), even
millions of years. Also, nutrient/herbivory models         though herbivory remains high.
similar to the RDM are receiving considerable attention        Changes in bottom-up controls and their interactions
(cf. Fig. 1 this paper and Fig. 2a in Bellwood et al.,       not only alter the dominance patterns of the major
2004). The coral-reef community needs a broader           benthic functional groups on coral reefs, but, hypothe-
biological perspective to further the recognition of the      tically, could have profound long-term consequences
role played by chronic nutrient enrichment in the coral       mediated through structural transformations and che-
reef health/resilience paradigm. Hopefully, the well-        mical modifications to reef systems and their herbivor-
intended plea (Pandolfi et al., 2005) for scientists to . . .    ous fish populations. In other words, excessive nutrient
‘‘stop arguing about the relative importance of different      enrichment not only increases the productivity and
causes of coral reef decline’’ . . ., will not discourage      biomass of weedy macroalgae via bottom-up controls
much-needed insightful research on nutrification.          that alter patterns of competitive dominance (Littler
  Unfortunately, the recurrent role of modern human-       et al., 1993), but, over the long term, may lead to coral
kind on coral reefs will continue to be to elevate         habitat degradation through: (1) reduced spatial
nutrients via sewage and agricultural eutrophication        heterogeneity by overgrowth (Johannes, 1975; Pastorok
(i.e., increasing bottom-up controls, Littler et al., 1991,     and Bilyard, 1985; Szmant, 1997); and (2) nighttime
1993; Goreau et al., 1997; Lapointe et al., 1997), while      anoxic conditions (tolerated by macroalgae, but not by
simultaneously decreasing herbivorous fishes (Littler        coral competitors and herbivorous predators, Lapointe
et al., 1991, 1993; Hughes, 1994) through trapping,         and Matzie, 1996) that could indirectly reduce top-
netting, poisoning and blasting (i.e., reducing top-down      down grazer effects. Furthermore, eutrophication-
controls). Unless curbed, such anthropogenically          induced macroalgal blooms decrease the growth and
                        M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                       581

reproductive capacity of the more structurally complex           Atkinson, M.J., Grigg, R.W., 1984. Model of a coral reef ecosystem.
reef-building corals (Tanner, 1995; Miller and Hay,               II. Gross and net benthic primary production at French Frigate
                                        Shoals, Hawaii. Coral Reefs 3, 13–22.
1996), as well as inhibit coral larval recruitment             Banner, A.H., 1974. Kaneohe Bay, Hawaii: urban pollution and a coral
(Birkeland, 1977; Tomascik, 1991; Ward and Harrison,              reef ecosystem. In: Proceedings of the Second International Coral
1997) and survival (Lewis, 1986; Hughes et al., 1987;              Reef Symposium 2. pp. 685–702.
Hughes, 1989; Wittenberg and Hunte, 1992). Such               Bell, P., 1992. Eutrophication and coral reefs: some examples in the
                                        Great Barrier Reef Lagoon. Wat. Res. 26, 555–568.
complicated feedback loops following eutrophication
                                      Belliveau, S.A., Paul, V.J., 2002. Effects of herbivory and nutrients on
are known to occur in seagrass meadows (Sand-Jensen               the early colonization of crustose coralline and fleshy algae. Mar.
and Borum, 1991; Duarte, 1995) and could also explain              Ecol. Prog. Ser. 232, 105–114.
decreases in fish populations on coral reefs with long-           Bellwood, D.R., Hughes, T.P., Folke, C., Nystrom, M., 2004. Con-
term histories of eutrophication (Johannes, 1975).               fronting the coral reef crisis. Nature 429, 827–833.
  Healthy coral-reef ecosystems are composed of              Bilger, R.W., Atkinson, M.J., 1995. Effects of nutrient loading on mass
                                        transfer rates to a coral-reef community. Limnol. Oceanog. 40,
diverse and highly productive communities. In natural              279–289.
reef ecosystems, much of the overall diversity at the            Birkeland, C., 1977. The importance of rate of biomass accumulation
primary-producer functional-group level is afforded by             in early successional stages of benthic communities to the survival
the intricate interaction of opposing herbivory/nutrient            of coral recruits. In: Proceedings of the Third International Coral
stimulating/limiting controls with the local physical/             Reef Symposium 1. pp. 15–21.
                                      Birkeland, C., 1987. Nutrient availability as a major determinant of
spatial variability. This results in a mosaic of                differences among coastal hard-substratum communities in dif-
continually changing environmental conditions in close             ferent regions of the tropics. In: Birkeland, C. (Ed.), Comparison
proximity. Because of the sensitive nature of direct/              Between Atlantic and Pacific Tropical Marine Coastal Ecosys-
indirect and stimulating/limiting interacting factors,             tems: Community Structure, Ecological Processes, and Produc-
                                        tivity. UNESCO Reports in Marine Science 46, pp. 45–97.
coral reefs are particularly vulnerable to anthropogenic
                                      Birkeland, C., 1997. Life and Death of Coral Reefs. Chapman and
reversal effects that decrease top-down controls and,              Hall, New York, 536 pp.
concomitantly, increase bottom-up controls, dramati-            Bray, J.R., Curtis, J.T., 1957. An ordination of the upland forest
cally altering ecosystem resiliencies.                     communities of southern Wisconsin. Ecol. Monog. 27, 325–349.
                                      Burke, R.B., 1982. Reconnaissance study of the geomorphology and
Acknowledgments                                 benthic communities of the outer barrier reef platform, Belize, In:
                                        Ruetzler, R., Macintyre, I.G. (Eds.), The Atlantic barrier reef
                                        ecosystems at Carrie Bow Cay, Belize, I. Structure and commu-
  We are sincerely grateful to Gene Rosenberg who               nities. Smithsonian Contributions to Marine Science No. 12,
critically read the manuscript and offered useful                Washington, DC.
suggestions. Thanks are extended to Randall Glaholt             Cameron, A.M., 1977. Acanthaster and coral reefs; population out-
and Joseph Smith for refilling the nutrient diffusers in             breaks of a rare and specialized carnivore in a complex high-
                                        diversity system. In: Proceedings of the Third International Coral
our absence. Support for this study came from the
                                        Reef Symposium 1. pp. 193–200.
Caribbean Coral Reef Ecosystems Program, ably                Carpenter, R.C., 1986. Partitioning herbivory and its effects on coral
administered by Klaus Reutzler (CCRE Contribution                reef algal communities. Ecol. Monog. 56, 345–363.
No. 735), and the Smithsonian Marine Station at Ft.             Carpenter, S.R., Kitchell, J.F., Hodgson, J.R., 1985. Cascading trophic
Pierce, Valerie Paul, Head Scientist (SMSFP Contribu-              interactions and lake productivity. Bioscience 35, 634–639.
                                      Chadwick, N.E., 1988. Competition and locomotion in a free-living
tion No. 626) of the National Museum of Natural
                                        fungiid coral. J. Exp. Mar. Biol. Ecol. 123, 189–200.
History, Smithsonian Institution. Contribution No. 1614           Coen, L.D., 1988. Herbivory by crabs and the control of algal
of the Harbor Branch Oceanographic Institution.[SE]               epibionts on Caribbean host corals. Oecologia 75, 198–203.
                                      Coles, S.L., 1988. Limitations of coral reef development in the
References                                   Arabian Gulf: temperature or algal competition. In: Proceedings
                                        of the Sixth International Coral Reef Symposium 1, Australia, pp.
                                        211–216.
Adey, W.H., 1998. Coral reefs: algal structured and mediated ecosys-
  tems in shallow, turbulent, alkaline waters. J. Phycol. 34, 393–406.   Crossland, C.J., Hatcher, B.G., Atkinson, M.J., Smith, S.V., 1984.
Adey, W.H., Adey, P., Burke, R., Kaufman, L., 1977. The Holocene        Dissolved nutrients of a high-latitude coral reef, Houtman Abrol-
  reef systems of eastern Martinique, French West Indies. Atoll Res.      hos Islands, Western Australia. Mar. Ecol. Prog. Ser. 14, 159–163.
                                      Cuet, P., Naim, O., Faure, G., Conan, J.Y., 1988. Nutrient-rich
  Bull. (218), 40.
Aronson, R.B., Precht, W.F., 2000. Herbivory and algal dynamics on       groundwater impact on benthic communities of la Saline fringing
  the coral reef at Discovery Bay, Jamaica. Limnol. Oceanogr. 45,       reef (Reunion Island, Indian Ocean): preliminary results. In:
  251–255.                                   Proceedings of the Sixth International Coral Reef Symposium
                                        2. pp. 1335–1340.
Atkinson, M.J., Bilger, R.W., 1992. Effects of water velocity on
  phosphate uptake in coral reef-flat communities. Limnol. Ocea-      Dayton, P.K., Oliver, J.S., 1980. An evaluation of experimental
  nog. 37, 273–279.                              analyses of population and community patterns in marine benthic
582                       M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

  environments. In: Tenore, K.R., Coull, B.C. (Eds.), Marine          the Fourth International Coral Reef Symposium 2, Manila, pp.
  Benthic Dynamic. University of South Carolina Press Columbia,        419–424.
  South Carolina, pp. 93–120.                       Hatcher, B.G., 1981. The interaction between grazing organisms and
DeBoer, J.A., Guigli, H.J., Israel, T.L., D’Elia, C.F., 1978. Nutritional    the epilithic algal community of a coral reef: a quantitative
  studies of two red algae. I. Growth rate as a function of nitrogen      assessment. In: Proceedings of the Fourth International Coral
  source and concentration. J. Phycol. 14, 261–266.              Reef Symposium 2, Manila, pp. 515–524.
Done, T.J., 1992. Phase shifts in coral reef communities and their      Hatcher, B.G., 1983. Grazing in coral reef ecosystems. In: Barnes, D.J.
  ecological significance. Hydrobiologia 247, 121–132.             (Ed.), Perspectives on Coral Reefs. Australian Institution of
Doty, M.S., 1969. The ecology of Honaunau Bay, Hawaii. Univ.           Marine Science, Townsville, Australia, pp. 164–179.
  Hawaii Bot. Sci. Paper No. 14, 221 pp.                  Hatcher, B.G., Larkum, A.W.D., 1983. An experimental analysis of
Doty, M.S., 1971. Antecedent event influence on benthic marine algal       factors controlling the standing crop of the epilithic algal com-
  standing crops in Hawaii. J. Exp. Mar. Biol. Ecol. 6, 161–166.        munity on a coral reef. J. Exp. Mar. Biol. Ecol. 69, 61–84.
Duarte, C.M., 1995. Submerged aquatic vegetation in relation to       Hay, M.E., 1981. Spatial patterns of grazing intensity on a Caribbean
  different nutrient regimes. Ophelia 41, 87–112.               barrier reef: herbivory and algal distribution. Aq. Bot. 11, 97–109.
Dubinsky, Z., Stambler, N., 1996. Marine pollution and coral reefs.     Hoegh-Guldberg, O., Takabayashi, M., Moreno, G., 1997. The impact
  Global Change Biol. 2, 511–526.                       of long-term nutrient enrichment on coral calcification and
ECOHAB, 1997. The ecology and oceanography of harmful algae           growth. In: Proceedings of the Eighth International Coral Reef
  blooms. In: Anderson, D.M. (Ed.), A National Research Agenda.        Symposium I. pp. 861–866.
  WHOI, Woods Hole, Massachusetts.                     Hughes, T.P., 1989. Community structure and the diversity of coral
Fishelson, L., 1973. Ecology of coral reefs in the Gulf of Aqaba (Red      reefs: the role of history. Ecology 78, 275–279.
  Sea) influenced by pollution. Oecologia 12, 55–67.            Hughes, T.P., 1994. Catastrophes, phase shifts, and large-scale degra-
Fong, P., Rudnicki, R., Zedler, J.B., 1987. Algal community response       dation of a Caribbean coral reef. Science 265, 1547–1551.
  to nitrogen and phosphorus loading in experimental mesocosms:      Hughes, T.P., Baird, A.H., Bellwood, D.R., Card, M., Connolly, S.R.,
  management recommendations for Southern California lagoons.         Folkes, C., Grosberg, R., Hoegh-Guldberg, O., Jackson, J.B.C.,
  Rep. Calif. St. Wat. Cont. Bd., 88 pp.                                             ¨
                                         Kleypas, J., Lough, J.M., Marshall, P., Nystrom, M., Palumbi,
Fujita, R.M., Wheeler, P.A., Edwards, R.L., 1988. Metabolic regula-       S.R., Pandolfi, J.M., Rosen, B., Roughgarden, J., 2003. Climate
  tion of ammonium uptake by Ulva rigida (Chlorophyta): a com-         change, human impacts, and the resilience of coral reefs. Science
  partmental analysis of the rate-limiting step for uptake. J. Phycol.     301, 929–933.
  24, 560–566.                               Hughes, T.P., Reed, D.C., Boyle, M., 1987. Herbivory on coral reefs:
Genin, A., Lazar, B., Brenner, S., 1995. Vertical mixing and coral        community structure following mass mortalities of sea urchins. J.
  death in the Red Sea following the eruption of Mt. Pinatubo.         Exp. Mar. Biol. Ecol. 113, 39–59.
  Nature 377, 507–510.                           Hughes, T.P., Szmant, A.M., Steneck, R., Carpenter, R., Miller, S.,
Ginsberg, R.L. (compiler), 1993. Global aspects of coral reefs. Health,     1999. Algal blooms on coral reefs: what are the causes? Limnol.
  hazards, and history. University of Miami, Miami, Florida, 420 pp.      Oceanogr. 44, 1583–1586.
Glynn, P.W., Ault, J.S., 2000. A biogeographic analysis and review of    James, N.P., Ginsburg, R.N., Marszalek, D.S., Choquette, P.W., 1976.
  the far eastern Pacific coral reef region. Coral Reefs 19, 1–23.       Facies and fabric specificity of early subsea cements in shallow
Goreau, T.F., 1959. The ecology of Jamaican coral reefs. I. Species       Belize (British Honduras) reefs. J. Sediment. Petrol. 46, 523–544.
  composition and zonation. Ecology 40, 67–90.               Jerlov, N.G., 1976. Marine Optics. Elsevier Scientific Publishing
Goreau, T.F., Land, L.S., 1974. Fore-reef morphology and deposi-         Company, 231 pp.
  tional processes, north Jamaica. In: Laporte, L.F. (Ed.), Reefs in    Johannes, R.E., 1975. Pollution and the degradation of coral reef
  Time and Space, Sp. Publ., vol. 18. Soc. Econ. Paleontolog.         communities. In: Wood, E., Johannes, R.E. (Eds.), Tropical Mar-
  Mineralolog., Tulsa, Oklahoma, USA, pp. 77–89.                ine Pollution. Elsevier, New York, pp. 13–51.
Goreau, T.J., 1992. Bleaching and reef community change in Jamaica:     Johnson, C.R., Mann, K.H., 1986. The crustose coralline alga, Phy-
  1951–1991. Am. Zool. 32, 683–695.                      matolithon Foslie, inhibits the overgrowth of seaweeds without
Goreau, T.J., 2003. Waste nutrients: impacts on coastal coral reefs and     relying on herbivores. J. Exp. Mar. Biol. Ecol. 96, 127–146.
  fisheries, and abatement via land recycling. In: Proceedings of the    Jompa, J., McCook, L.J., 2002. Effects of competition and herbivory
  UN Expert Meeting on Waste Management in small island devel-         on interactions between a hard coral and a brown alga. J. Exp. Biol.
  oping states, Havana, Cuba.                         Ecol. 271, 25–39.
Goreau, T.J., Thacker, K., 1994. Coral reefs, sewage, and water quality   Keats, D.W., Chamberlain, Y.M., Baba, M., 1997. Pneophyllum
  standards. In: Proceedings of the third annual Caribbean water and      conicum (Dawson) comb. Nov. (Rhodophyta, Corallinaceae), a
  wastewater association conference: water and wastewater needs        widespread Indo-Pacific non-geniculate coralline alga that over
  for the Caribbean: 21st century, 3. pp. 98–117.               grows and kills live coral. Bot. Mar. 40, 263–279.
Goreau, T.J., Dalay, L., Ciappara, S., Brown, J., Dourke, S., Thacker,    Keats, D.W., Wilton, P., Maneveldt, G., 1994. Ecological significance
  K., 1997. Community-based whole-watershed and coastal zone          of deep-layer sloughing in the eulittoral zone coralline alga,
  management in Jamaica. In: Proceedings of the Eighth Interna-        Spongites yendoi (Foslie) Chamberlain (Corallinaceae, Rhodo-
  tional Coral Reef Symposium 2. pp. 2093–2096.                phyta) in South Africa. J. Exp. Mar. Biol. Ecol. 175, 145–154.
Grime, J.P., 1979. Plant Strategies and Vegetation Processes. John      Kinsey, D.W., Davies, P.J., 1979. Effects of elevated nitrogen and
  Wiley & Sons, New York.                           phosphorus on coral reef growth. Limnol. Oceanogr. 24, 935–
        ¨
Hallock, P., Muller-Karger, F.E., Hallas, J.C., 1993. Coral reef decline.    940.
  Nat. Geog. Res. Exp. 9, 358–378.                     Kinsey, D.W., Domm, A., 1974. Effects of fertilization on a coral reef
Hatcher, A.I., Hatcher, B.G., 1981. Seasonal and spatial variation in      environment—primary production studies. In: Proceedings of the
  dissolved nitrogen in One Tree Reef Lagoon. In: Proceedings of        Second International Coral Reef Symposium 1. pp. 49–66.
                         M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                       583

Klumpp, D.W., McKinnon, D., Daniel, P., 1987. Damselfish terri-        Larkum, A.W.D., Koop, K., 1997. ENCORE, algal productivity and
  tories: zones of high productivity on coral reefs. Mar. Ecol. Prog.      possible paradigm shifts. In: Proceedings of the Eighth Interna-
  Ser. 40, 41–51.                                tional Coral Reef Symposium 1. pp. 881–884.
Lapointe, B.E., 1987. Phosphorus-and nitrogen-limited photosynth-       Larned, S.T., Stimson, J., 1996. Nitrogen-limited growth in the coral
  esis and growth of Gracilaria tikvahiae (Rhodophyceae) in the         reef chlorophyte Dictyospheria cavernosa, and the effect of
  Florida Keys: an experimental field study. Mar. Biol. 93, 561–568.       exposure to sediment-derived nitrogen on growth. Mar. Ecol.
Lapointe, B.E., 1997. Nutrient thresholds for bottom-up control of        Progr. Ser. 145, 95–108.
  macroalgal blooms on coral reefs in Jamaica and southeast         Lewis, S.M., 1986. The role of herbivorous fishes in the organization
  Florida. Limnol. Oceanogr. 42, 1119–1131.                   of a Caribbean reef community. Ecol. Monog. 56, 183–200.
Lapointe, B.E., 1999. Simultaneous top-down and bottom-up forces       Lewis, S.M., Norris, J.N., Searles, R.B., 1987. The regulation of
  control macroalgal blooms on coral reefs (reply to the comment by       morphological plasticity in tropical reef algae by herbivory.
  Hughes et al.) Limnol. Oceanogr. 44, 1586–1592.                Ecology 68, 636–641.
Lapointe, B.E., Barile, P.J., Littler, M.M., Littler, D.S., Bedford, B.J.,  Lewis, S.M., Wainwright, P.C., 1985. Herbivore abundance and
  Gasque, C., 2005a. Macroalgal blooms on southeast Florida coral        grazing intensity on a Caribbean coral reef. J. Exp. Mar. Biol.
  reefs: I. Nutrient stoichiometry of the invasive green alga Codium       Ecol. 87, 215–228.
  isthmocladum in the wider Caribbean indicates nutrient enrich-       Lirman, D., Biber, P., 2000. Seasonal dynamics of algal communities
  ment. Harmful Algae 4, 1092–1105.                       in the northern Florida reef tract. Bot. Mar. 43, 305–314.
Lapointe, B.E., Barile, P.J., Littler, M.M., Littler, D.S., 2005b. Macro-   Littler, D.S., Littler, M.M., 2000. Caribbean Reef Plants: An Identi-
  algal blooms on southeast Florida coral reefs. II. Cross-shelf         fication Guide to the Reef Plants of the Caribbean, Bahamas,
  discrimination of nitrogen sources indicates widespread assimila-       Florida and Gulf of Mexico. Offshore Graphics, Inc., Washington,
  tion of sewage nitrogen. Harmful Algae 4, 1106–1122.              DC.
Lapointe, B.E., Barile, P.J., Matzie, W.R., 2004. Anthropogenic        Littler, M.M., 1971. Standing stock measurements of crustose coral-
  nutrient enrichment of seagrass and coral reef communities in         line algae (Rhodophyta) and other saxicolous organisms. J. Exp.
  the lower Florida Keys: discrimination of local versus regional        Mar. Biol. Ecol. 6, 91–99.
  nitrogen sources. J. Exp. Mar. Biol. Ecol. 308, 23–58.           Littler, M.M., 1972. The crustose corallinaceae. Oceanogr. Mar. Biol.
Lapointe, B.E., Matzie, W.R., 1996. Effects of storm water nutrient        Ann. Rev. 10, 311–347.
  discharges on eutrophication processes in nearshore waters of the     Littler, M.M., 1973. The population and community structure of
  Florida Keys. Estuaries 19, 422–435.                      Hawaiian fringing-reef crustose Corallinaceae (Rhodophyta,
Lapointe, B.E., Littler, M.M., Littler, D.S., 1987. A comparison of        Cryptonemiales). J. Exp. Mar. Biol. Ecol. 11, 103–120.
  nutrient-limited productivity in macroalgae from a Caribbean        Littler, M.M., Doty, M.S., 1975. Ecological components structuring
  barrier reef and from a mangrove ecosystem. Aq. Bot. 28, 243–         the seaward edges of tropical Pacific reefs: the distribution,
  255.                                      communities and productivity of Porolithon. J. Ecol. 63, 117–129.
Lapointe, B.E., Littler, M.M., Littler, D.S., 1992. Nutrient availability   Littler, M.M., Littler, D.S., 1984a. Models of tropical reef biogenesis:
  to marine macroalgae in siliciclastic versus carbonate-rich coastal      the contribution of algae. In: Round, F.E., Chapman, D.J. (Eds.),
  waters. Estuaries 15, 76–83.                          Progress in Phycological Research, vol. 3. Biopress, Bristol, pp.
Lapointe, B.E., Littler, M.M., Littler, D.S., 1993. Modification of        323–364.
  benthic community structure by natural eutrophication: the Belize     Littler, M.M., Littler, D.S., 1984b. Relationships between macroalgal
  barrier reef. In: Proceedings of the Seventh International Coral        functional form groups and substrata stability in a subtropical
  Reef Symposium 1. pp. 323–334.                         rocky-intertidal system. J. Exp. Mar. Biol. Ecol. 74, 13–34.
Lapointe, B.E., Littler, M.M., Littler, D.S., 1997. Macroalgal over-     Littler, M.M., Littler, D.S., 1985. Non-destructive sampling. In:
  growth of fringing coral reefs at Discovery Bay, Jamaica: bottom-       Littler, M.M., Littler, D.S. (Eds.), Handbook of Phycological
  up versus top-down control. In: Proceedings of the Eighth           Methods. Ecological Field Methods: Macroalgae. Cambridge
  International Coral Reef Symposium 1. pp. 927–932.               University Press, Cambridge, pp. 161–175.
Lapointe, B.E., Matzie, W.R., Clark, M.W., 1994. Phosphorus inputs      Littler, M.M., Littler, D.S., 1990. Productivity and nutrient relation-
  and eutrophication on the Florida Reef Tract. In: Ginsberg, R.N.        ships in psammophytic versus epilithic forms of Bryopsidales
  (compiler), Proceedings of the colloquium on global aspects of         (Chlorophyta): comparisons based on a short-term physiological
  coral reefs: health, hazards and history. University of Miami, pp.       assay. Hydrobiologia 204–205, 49–55.
  106–112.                                  Littler, M.M., Littler, D.S., 1997. Disease-induced mass mortality of
Lapointe, B.E., O’Connell, J.D., 1989. Nutrient-enhanced productiv-        crustose coralline algae on coral reefs provides rationale for the
  ity of Cladophora prolifera in Harrington Sound, Bermuda:           conservation of herbivorous fish stocks. In: Proceedings of the
  eutrophication of a confined phosphorus-limited marine ecosys-         Eighth International Coral Reef Symposium 1. pp. 719–724.
  tem. Estuar. Coast. Shelf Sci. 28, 347–360.                Littler, M.M., Littler, D.S., Lapointe, B.E., 1993. Modification of
Lapointe, B.E., Tenore, K.R., 1981. Experimental outdoor studies         tropical reef community structure due to cultural eutrophication:
  with Ulva fasciata Delile. I. Interaction of light and nitrogen on       the southwest coast of Martinique. In: Proceedings of the Seventh
  nutrient uptake, growth, and biochemical composition. J. Exp.         International Coral Reef Symposium 1. pp. 335–343.
  Mar. Biol. Ecol. 53, 135–152.                       Littler, M.M., Littler, D.S., Taylor, P.R., 1983a. Evolutionary strate-
Lapointe, B.E., Thacker, K., 2002. Community-based water quality         gies in a tropical barrier reef system: functional-form groups of
  and coral reef monitoring in the Negril Marine Park, Jamaica:         marine macroalgae. J. Phycol. 19, 229–237.
  land-based nutrient inputs and their ecological consequences. In:     Littler, M.M., Littler, D.S., Taylor, P.R., 1987a. Animal-plant
  Porter, J.W., Porter, K.G. (Eds.), The Everglades, Florida Bay, and      defense associations: effects on the distribution and abundance
  Coral Reefs of the Florida Keys: An Ecosystem Sourcebook. CRC         of tropical reef macrophytes. J. Exp. Mar. Biol. Ecol. 105, 107–
  Press, Boca Raton, Florida, pp. 939–963.                    121.
584                       M.M. Littler et al. / Harmful Algae 5 (2006) 565–585

Littler, M.M., Littler, D.S., Taylor, P.R., 1995. Selective herbivore    National Research Council, 1995. Understanding Marine Biodiversity.
  increases biomass of its prey: a chiton-coralline reef-building       Ocean Studies Board, Biology Board. National Academy Press,
  association. Ecology 76, 1666–1681.                     Washington DC.
Littler, M.M., Littler, D.S., Titlyanov, E.A., 1991. Comparisons of N-    National Research Council, 2000. Clean Coastal Waters: Understand-
  and P-limited productivity between high granitic islands versus       ing and Reducing the Effects of Nutrient Pollution. Ocean Studies
  low carbonate atolls in the Seychelles Archipelago: a test of the      Board, Water Science and Technology Board. National Academy
  relative-dominance paradigm. Coral Reefs 10, 199–209.            Press, Washington DC.
Littler, M.M., Martz, D.R., Littler, D.S., 1983c. Effects of recurrent    Nixon, S.W., 1995. Coastal marine eutrophication: a definition, social
  sand deposition on rocky intertidal organisms: importance of         causes, and future concerns. Ophelia 41, 199–219.
  substrate heterogeneity in a fluctuating environment. Mar. Ecol.     Nugues, M.M., Smith, G.W., van Hooidonk, R.J., Seabra, M.I., Bak,
  Prog. Ser. 11, 129–139.                           R.P.M., 2004. Algal contact as a trigger for coral disease. Ecol.
Littler, M.M., Taylor, P.R., Littler, D.S., 1983b. Algal resistance to      Let. 7, 919–923.
  herbivory on a Caribbean barrier reef. Coral Reefs 2, 111–118.      Ogden, C.J., Brown, R.A., Salesky, N., 1973. Grazing by the echinoid
Littler, M.M., Taylor, P.R., Littler, D.S., 1986. Plant defense associa-     Diadema antillarum Philippi: formation of halos around West
  tions in the marine environment. Coral Reefs 5, 63–71.            Indian patch reefs. Science 182, 715–717.
Littler, M.M., Taylor, P.R., Littler, D.S., 1989. Complex interactions in  Pandolfi, J.M., Jackson, J.V.C., Baron, N., Bradbury, R.H., Guzman,
  the control of coral zonation on a Caribbean reef flat. Oecologia       H.M., Hughes, T.P., Kappel, C.V., Michell, F., Ogden, J.C.,
  80, 331–340.                                 Possingham, H.P., Sala, E., 2005. Are U.S. coral reefs on the
Littler, M.M., Taylor, P.R., Littler, D.S., Sims, R.H., Norris, J.N.,      slippery slope to slime? Science 307, 1725–1726.
  1987b. Dominant macrophyte standing stocks, productivity and       Pastorok, R.A., Bilyard, G.R., 1985. Effects of sewage pollution on
  community structure on a Belizean barrier-reef. Atoll Res. Bull.       coral-reef communities. Mar. Ecol. Prog. Ser. 21, 175–189.
  No. 302, 24 pp.                             Pollock, J.B., 1928. Fringing and fossil coral reefs of Oahu. Bull.
                          ˜
Lough, J.M., 1994. Climate variation and El Nino—Southern Oscilla-        Bernice P. Bishop Mus. 55, 1–56.
  tion events on the Great Barrier Reef: 1958–1987. Coral Reefs 13,    Precht, W.F., Miller, S.L., in press. Ecological shifts along the Florida
  181–195.                                   reef tract: the past as a key to the future. In: Aronson, R.B. (Ed.),
                               ˜
Macintyre, I.G., Glynn, P.W., 1990. Upper limit of El Nino kill off.       Geological approaches to coral reef ecology. Springer-Verlag,
  Coral Reefs 9, 92.                              New York.
Macintyre, I.G., Graus, R.R., Reinthal, P.N., Littler, M.M., Littler,    Randall, J.E., 1965. Grazing effect on seagrasses by herbivorous
  D.S., 1987. The barrier reef sediment apron: Tobacco Range,         marine fishes. Ecology 46, 255–260.
  Belize. Coral Reefs 6, 1–12.                       Reinthal, P.N., Macintyre, I.G., 1994. Spatial and temporal variations
Marubini, F., Davies, P.S., 1996. Nitrate increases zooxanthellae        in grazing pressure by herbivorous fishes: Tobacco Reef, Belize.
  population density and reduces skeletogenesis in corals. Mar.        Atoll Res. Bull. 425, 1–11.
  Biol. 127, 319–328.                           Rosenberg, G., Ramus, J., 1984. Uptake of inorganic nitrogen and
McClanahan, T.R., Cokos, B.A., Sala, E., 2002. Algal growth and         seaweed surface area: volume ratios. Aq. Bot. 19, 65–72.
  species composition under experimental control of herbivory,       Ruetzler, K., Macintyre, I.G., 1982. The habitat distribution and
  phosphorus abundance in Glovers Reef, Belize. Mar. Poll. Bull.        community structure of the barrier reef complex at Carrie Bow
  44, 441–451.                                 Cay, Belize. In: Ruetzler, K., Macintyre, I.G. (Eds.), The Atlantic
McConnaughey, T.A., Adey, W.H., Small, A.M., 2000. Community           barrier reef ecosystem at Carrie Bow Cay, Belize, I. Structure and
  and environmental influences on reef coral calcification. Limnol.       communities. Smithsonian Contributions to Marine Science, No.
  Oceanogr. 45, 1667–1671.                           12, pp. 9–45.
McCook, L.J., 1999. Macroalgae, nutrients and phase shifts on coral     Ruyter van Steveninck, E.D., 1984. The composition of algal vegeta-
  reefs: scientific issues and management consequences for the         tion in and outside damselfish territories on a Florida reef. Aq. Bot.
  Great Barrier Reef. Coral Reefs 18, 357–367.                 20, 11–19.
Mergener, H., 1981. Man-made influences on and natural changes in       Sammarco, P.W., 1980. Diadema and its relationship to coral spat
  the settlement of the Aqaba reefs (Red Sea). In: Proceedings of the     mortality: grazing, competition, and biological disturbance. J.
  Fourth International Coral Reef Symposium 1. pp. 193–207.          Exp. Mar. Biol. Ecol. 45, 245–272.
Miller, M.W., Hay, M.E., 1996. Coral–seaweed–grazer–nutrient inter-     Sammarco, P.W., 1982. Effects of grazing by Diadema antillarum
  actions on temperate reefs. Ecol. Monog. 66, 323–344.            Philippi (Echinodermata: Echinoidea) on algal diversity and com-
Miller, M.W., Hay, M.E., 1998. Effects of fish predation and seaweed       munity structure. J. Exp. Mar. Biol. Ecol. 65, 83–105.
  competition on the survival and growth of corals. Oecologia 113,     Sammarco, P.W., 1983. Effects of fish grazing and damselfish terri-
  231–238.                                   toriality on coral reef algae. I. Algal community structure. Mar.
Miller, M.W., Hay, M.E., Miller, S.L., Sotka, E., Szmant, A.M., 1999.      Ecol. Prog. Ser. 13, 1–14.
  Effects of nutrients versus herbivores on reef algae: a new method    Sand-Jensen, K., Borum, J., 1991. Interactions among phytoplankton,
  for manipulating nutrients on coral reefs. Limnol. Oceanogr. 44,       periphyton, and macrophytes in temperate freshwaters and estu-
  1847–1861.                                  aries. Aq. Bot. 41, 137–175.
Morrisson, D., 1988. Comparing fish and urchin grazing in shallow       Santavy, D.L., Peters, E.C., 1997. Microbial pests: coral disease in the
  and deeper coral reef algal communities. Ecology 69, 1367–          western Atlantic. In: Proceedings of the Eighth International
  1382.                                    Coral Reef Symposium 1. pp. 607–612.
Muller-Parker, G., McCloskey, L.R., Høegh-Guldberg, O., McAuley,       SAS, 2003. SAS for Windows, v. 9.1. SAS Institute, Inc., Cary, North
  P.J., 1994. Effects of ammonium enrichment on animal and algal        Carolina, USA.
  biomass of the coral Pocillopora damicornis. Pac. Sci. 48, 273–     Schaffelke, B.D., Klumpp, W., 1998. Nutrient-limited growth of the
  283.                                     coral reef macroalga Sargassum baccularia and experimental
                        M.M. Littler et al. / Harmful Algae 5 (2006) 565–585                       585

  growth enhancement by nutrient addition in continuous flow        Tomascik, T., 1991. Settlement patterns of Caribbean scleractinian
  culture. Mar. Ecol. Progr. Ser. 164, 199–211.                corals on artificial substrata along an eutrophication gradient,
Scheffer, M.S., Carpenter, S., Foley, J.A., Folke, C., Walker, B., 2001.    Barbados, West Indies. Mar. Ecol. Prog. Ser. 77, 261–269.
  Catastrophic shifts in ecosystems. Nature 43, 591–596.          Tomascik, T., Sander, F., 1985. Effects of eutrophication on reef-
Schmitt, E.F., 1997. The influence of herbivorous fishes on coral reef      building corals. I. Growth rate of the reef-building coral Montas-
  communities with low sea urchin abundance: a study among reef        trea annularis. Mar. Biol. 87, 143–155.
  community types and seasons in the Florida Keys. Ph.D. disserta-     Tomascik, T., Sander, F., 1987a. Effects of eutrophication on reef-
  tion, University of Miami, Miami, Florida, 221 pp.              building corals: reproduction of the reef-building coral Porites
Simkiss, K., 1964. Phosphate as crystal poisons of calcification. Biol.     porites. Mar. Biol. 94, 77–94.
  Rev. 39, 487–505.                            Tomascik, T., Sander, F., 1987b. Effects of eutrophication on reef
Smith, J.E., Smith, C.M., Hunter, C.L., 2001. An experimental          building corals. II. Structure of scleractinian coral communities on
  analysis of the effects of herbivory and nutrient enrichment on       fringing reefs, Barbados, West Indies. Mar. Biol. 94, 53–75.
  benthic community dynamics on a Hawaiian reef. Coral Reefs 19,      Verheij, E., 1993. Marine plants of the Spermonde Archipelago, SW
  332–342.                                   Sulawesi Indonesia: aspects of taxonomy, floristics, and ecology.
Smith, S.V., Kimmerer, W.J., Laws, E.A., Brock, R.E., Walsh, T.W.,       Ph.D. Dissertation, State University Leiden, Rijksherbarium, Lei-
  1981. Kaneohe Bay sewage diversion experiment: perspectives on        den.
  ecosystem responses to nutritional perturbation. Pac. Sci. 35, 279–   Vine, P.J., 1974. Effects of algal grazing and aggressive behavior of
  397.                                     the fishes Pomacentrus lividus and Acanthurus sohal on coral reef
Stambler, N., Popper, N., Dubinsky, Z., Stimson, J., 1991. Effects of      ecology. Mar. Biol. 4, 131–136.
  nutrient enrichment and water motion on the coral Pocillopora      Vitousek, P.M., Aber, J., Howarth, R.W., Likens, G.E., Matson, P.A.,
  damicornis. Pac. Sci. 45, 299–307.                      Schindler, D.W., Schlesinger, W.H., Tilman, D.G., 1997. Human
Stephenson, W., Searles, R.B., 1960. Experimental studies on the        alteration of the global nitrogen cycle, sources and consequences.
  ecology of intertidal environments at Heron Island. I. Exclusion of     Ecol. Adapt. 7, 737–750.
  fish from beach rock. Aust. J. Mar. Freshwat. Res. 11, 241–267.      Wanders, J.B.W., 1976. The role of benthic algae in the shallow reef of
Steneck, R.S., 1989. Herbivory on coral reefs: a synthesis. In:         Curacao (Netherlands Antilles). I. primary production in the coral
                                           ¸
  Proceedings of the Sixth International Coral Reef Symposium         reef. Aq. Bot. 2, 235–270.
  1, Australia, 1988, pp. 37–49.                      Ward, S., Harrison, P.L., 1997. The effects of elevated nutrient levels
Steneck, R.S., Dethier, M.N., 1994. A functional group approach to       on settlement of coral larvae during the Encore experiment, Great
  the structure of algal-dominated communities. Oikos 69, 476–498.       Barrier Reef, Australia. In : Proceedings of the Eighth Interna-
Szmant, A.M., 1997. Nutrient effects on coral reefs: a hypothesis on      tional Coral Reef Symposium 1996, vol. 1, pp. 891–896.
  the importance of topographic and trophic complexity to reef       Weiss, M.P., Goddard, D.A., 1977. Man’ s impact on coastal reefs: an
  nutrient dynamics. In: Proceedings of the Eighth International        example from Venezuela. Am. Assoc. Petrol. Geolog., Stud. Geol.
  Coral Reef Symposium 2. pp. 1527–1532.                    4, 111–124.
Tanner, J.E., 1995. Competition between hard corals and macroalgae:     Williams, S.L., Fisher, T.R., 1985. Kinetics of nitrogen-15 labeled
  an experimental analysis of growth, survival, and reproduction. J.      ammonium uptake by Caulerpa cupressoides (Chlorophyta). J.
  Exp. Mar. Biol. Ecol. 190, 51–168.                      Phycol. 21, 287–296.
Thacker, R.W., Ginsburg, D.W., Paul, V.J., 2001. Effects of herbivore    Wittenberg, M., Hunte, W., 1992. Effects of eutrophication and
  exclusion and nutrient enrichment on coral reef macroalgae and        sedimentation on juvenile corals. I. abundance, mortality, and
  Cyanobacteria. Coral Reefs 19, 318–329.                   community structure. Mar. Biol. 112, 131–138.