O

Ms

Pa

b

c

a

ARRA

KCMBSDP

1

itWsnfimH

OH

h1

Ecological Indicators 77 (2017) 276–285

Contents lists available at ScienceDirect

Ecological Indicators

jo ur nal ho me page: www.elsev ier .com/ locate / ecol ind

riginal Article

ucous sheet production in Porites: an effective bioindicator ofediment related pressures

ia Bessell-Browne a,b,c,∗, Rebecca Fisher a,c, Alan Duckworth a,c, Ross Jones a,c

Australian Institute of Marine Science, Townsville, QLD and Perth, WA, AustraliaThe Oceans Institute and The Centre for Microscopy, Characterisation and Analysis, The University of Western Australia, Perth, WA, AustraliaWestern Australian Marine Science Institution, Perth, WA, Australia

r t i c l e i n f o

rticle history:eceived 19 October 2016eceived in revised form 11 February 2017ccepted 15 February 2017

eywords:oralucus

ioindicatoredimentredgingorites

a b s t r a c t

Some coral species of the genus Porites can produce thick mucous sheets that partially or completelyenvelope the colony’s surface. This phenomenon has been reported many times, but the cause and eco-logical significance remains unclear. In this study, sheet production was examined in response to elevatedsuspended sediment concentrations associated with a large-scale, extended dredging project on a coralreef. Approximately 400 corals at 16 locations situated from 0.2–33 km from the excavation area wereexamined at fortnightly intervals over the 1.5 year dredging campaign. Mucous sheets were observed on447 occasions (from 10,600 observations), with average mucous prevalence ranging from 0–10%. Overall74 ± 5% of the colonies <1.5 km from the dredging produced one or more sheets. High levels of mucouscoverage (≥95% of the colony surface) was observed on 68 occasions, and 82% of these occurred at sitesclose to the dredging. Approximately 50% of colonies produced ≥3 sheets over the monitoring period, and90% of these were located close to the dredging. In contrast, at distantly located reference sites (>20 kmaway), mean mucous sheet prevalence was very low (0.2% ± 0.1), no colonies produced more than 1sheet, and only 1 colony was observed with high mucous coverage. In a laboratory-based experiment,explants of Porites spp. exposed to fine silt also produced mucous sheets (105 sheets recorded in 1100observations), with nearly 30% of the fragments exposed to repeated sediment deposition events of 10

−2 −1

and 20 mg cm d producing 2 new sheets over the 28 day exposure period. These multiple lines ofevidence suggest a close association between mucous sheet formation and sediment load, and that sheetformation and sloughing are an additional mechanism used by massive Porites spp. to clear their surfaceswhen sediment loads become too high. These results suggest that mucous sheet formation is an effectivebioindicator of sediment exposure.

© 2017 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license

. Introduction

Hard corals continually synthesise and secrete mucus, maintain-ng a near continuous mucosal barrier of several hundred-micronhickness over their surfaces (Brown and Bythell, 2005; Bythell and

ild, 2011; Jatkar et al., 2010b; Lewis, 1973). The majority of thisecreted mucus dissolves in the water column and is generally nototiceable (Wild et al., 2004), although some mucus forms visible

laments and strings that eventually detach from corals producingucus flocs in the water column (see images in Wild et al. (2004),uettel et al. (2006) and Bythell and Wild (2011)). A more conspic-

∗ Corresponding author at: Australian Institute of Marine Science, Level 3 Indiancean Marine Research Centre, The University of Western Australia M097, Stirlingighway, Crawley, Western Australia, Australia.

E-mail address: piabessellbrowne@gmail.com (P. Bessell-Browne).

ttp://dx.doi.org/10.1016/j.ecolind.2017.02.023470-160X/© 2017 The Author(s). Published by Elsevier Ltd. This is an open access article

(http://creativecommons.org/licenses/by/4.0/).

uous form of mucus production occurs in corals from the genusPorites, resulting in complete covering of a thick mucosal layer (seeVeron (2000), page 278). This phenomenon has been reported manytimes, but the cause and ecological significance remains unclear.

The production of a thick mucous layer in Porites was firstdescribed by Duerden (1906), who referred to it as a mucousfelt, and since then many other terms have been used, includ-ing: sheet (Johannes, 1967), envelope (Lewis, 1973), web (Ducklowand Mitchell, 1979), sheath (Thompson, 1980), layer and cover-ing (Marcus and Thorhaug, 1981), coat (Duerden, 1906), tunic(Edmunds and Davies, 1986), film (Coffroth, 1988), and mat(Stafford-Smith and Ormond, 1992). The term sheet is used here-after.

These sheets initially appear as clear mucous films, tightlycovering the coral surface (Lewis, 1973), but once formed themucus undergoes a physicochemical transformation from a fluid-

under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

P. Bessell-Browne et al. / Ecological Indicators 77 (2017) 276–285 277

F tern Aa ning bd

l1Smdasap2Bae

pna2flutc1

i

ig. 1. (a) Location of the study area near Barrow Island off the Pilbara coast of Wesnalysis, their position and proximity to the material offloading facility (MOF), turredged.

ike (Ducklow and Mitchell, 1979), to gel-like consistency (Coffroth,985, 1988; Duerden, 1906; Stafford-Smith and Ormond, 1992).EM imaging of sheets from Porites compressa indicate that they areade of multiple layers, which act as a physical barrier over the epi-

ermis, but still allows the diffusion of dissolved material (Johnstonnd Rohwer, 2007). The biochemical composition of Porites mucousheets has been described by Lewis (1973), and Coffroth (1990);nd is likely to be composed of mucins, i.e. large glycoproteins thatossess different properties of elasticity and viscosity (Jatkar et al.,010a). The term mucus is used hereafter, and in a broad sense (seeythell and Wild (2011)), as the sheets are likely to contain mucinslong with a variety of dissolved and/or particulate organic matterxcreted by the coral.

Once formed, mucous sheets can become fouled with algae,rotozoa, zooplankton (Mayer and Wild, 2010), bacteria, ciliates,auplii, crustacea (Coffroth, 1990), fecal pellets, other unidentifi-ble debris and sediment (Hartnoll, 1974; Johnston and Rohwer,007; Krupp, 1984; Lewis, 1973; Mayer and Wild, 2010). The trans-ormation of viscous mucus to a sheet like formation has beeninked to fouling by sediment (Ducklow and Mitchell, 1979). Sheetssually remain covering colonies from days to months, but even-ually are shed (commonly referred to as ‘sloughed’) by water

urrents, leaving the colony free of fouling material (Duerden,906; Lewis, 1973).

Mucous sheets have been reported in Porites astreoides, P. divar-cata, P. furcata, and P. porites in Florida and the Caribbean (Bak

ustralia. (b) Sites used for analysis and their location. (c) The impact sites used forasin and tanker access channel, and the reef structure of the area, all which were

and Elgershuizen, 1976; Coffroth, 1985; Ducklow and Mitchell,1979; Edmunds and Davies, 1986; Glasl et al., 2016; Lewis, 1973;Marcus and Thorhaug, 1981), Porites cylindrica from Japan (Kato,1987), Porites compressa from Hawaii (Krupp, 1984; Marcus andThorhaug, 1981), along with P. lutea, P. lobata, P. australiensis andP. murrayensis on the Great Barrier Reef of Australia (Coffroth,1988; Coffroth, 1990; Stafford-Smith and Ormond, 1992). Similarsheet like structures have also been reported in the corals Diploas-trea heliopora, Pachyseris speciosa, Mycedium elephantotus, Pectinialactuca, P. paeonia, Turbinaria peltata and T. mesenterina (Sofoniaand Anthony, 2008; Stafford-Smith and Ormond, 1992), Helioporacoerulea (Lewis, 1973), and in other groups such as gorgonians,alcyonaceans and zoanthids (Coffroth, 1988; Rublee et al., 1980).

Fluid mucus production by corals aids feeding and self-cleaningby muco-ciliary transport processes (Duerden, 1906; Hubbard andPocock, 1972; Lewis and Price, 1975; Stafford-Smith and Ormond,1992; Stafford-Smith, 1993). A host of other functions for fluidmucus production have also been suggested, including defenceagainst pathogens and resistance to desiccation, ultra violet radi-ation, pollutants and physical damage (see reviews by Brown andBythell (2005) and Bythell and Wild (2011)). Whilst fluidic mucusproduction plays many important roles in corals, it is not clear

why Porites spp. produce the gel-like sheets. Some studies with P.compressa, P. porites and P. cylindrica have observed a relationshipbetween temperature, salinity and mucous sheet production (Kato,1987; Marcus and Thorhaug, 1981). Xenobiotics such as copper and

278 P. Bessell-Browne et al. / Ecological Indicators 77 (2017) 276–285

Table 1Summary statistics of site location, distance from dredging, depth and water quality characteristics over the dredging period, the number of colonies examined, the numberof observations made and summary statistics for mucous sheet formation of Category 3 events or higher (i.e. >5–35% mucus covering) including the total number of events,and the number of new mucous sheet events.

A Site name (see Fig. 1)B Distance (km) from the closest location of dredgingC Median water depth (m)D 95th percentile (P95) of all 14 d running mean averages (NTU)E 5th percentile (P5) of all 14 d daily light integral averages (DLI, mol photons m2 d−1)F Number of surveys conducted in the dredging phaseG Number of colonies at each siteH � individual observations of a single colony (i.e. colonies × surveys × baseline and dredging period)I � individual observations of a single colony (i.e. colonies × surveys × dredging period)J � number of times a colony was observed with ≥5% mucus cover (category 3) over the dredging periodK Total number of ‘new’ mucus events (≥5% mucus cover, category 3) over dredging period a

L % of colonies at each site observed with ≥5% mucus cover (category 3) at any stage over the dredging periodM Mean ± SE (Column N) incidence of mucous sheet formation (≥5% mucus cover, category 3) over the dredging period

A B C D E F G H I J K L M NSite km m NTUb DLIc n n n n n n % % %

MOFA 0.4 5 26.5 0.9 39 23 879 811 48 43 73.9 6.1 1.2MOFB 1.0 8 7.8 0.3 40 22 843 778 76 59 95.5 9.6 1.6MOF1 0.8 6 14.6 0.4 40 19 660 604 58 30 84.2 10.0 1.8LNG0 0.2 9 19.6 0.1 17 32 428 359 37 29 56.3 8.7 2.1LNGA 0.3 11 18.8 0.2 23 31 446 384 38 33 54.8 9.7 2.8LNG1 0.5 9 15.0 0.1 34 31 788 687 45 41 71.0 6.9 1.4LNG2 1.0 7 8.4 0.4 36 27 922 837 41 41 81.5 4.9 1.4LNGC 1.4 11 12.2 0.3 35 21 516 455 29 29 71.4 6.9 1.7LNG3 4 6 9.5 1.1 36 25 898 823 33 33 64.0 4.0 0.9TR 5 5 10.7 0.4 39 14 512 454 14 12 57.1 2.9 1.1DUG 9 9 9.6 1.3 35 27 997 913 16 16 40.7 1.8 0.6BAT 15 4 3.7 1.6 32 28 858 744 5 5 17.9 0.6 0.4SBS 24 5 3.2 2.1 32 25 763 690 2 2 8.0 0.3 0.3REFS 28 5 3.2 0.9 30 30 793 704 2 2 6.7 0.2 0.2AHC 30 7 2.7 1.3 30 23 662 591 3 3 13.0 0.5 0.3REFN 33 7 3.7 2.2 27 34 833 766 0 0 0.0 0.0 0.0Mean – – – – – – – – – 24 50 – –min 0.2 4 2.7 0.1 17 14 428 359 0 0 0 0.0 –max 33 11 26.5 2.2 40 34 997 913 76 59 95 10.0 –SUM – – – – 525 412 11,798 10,600 447 378 796 – –

sible i

cae

mrbsmbfawKfhabcs

CmJ(a1s

a A new mucous sheet was classified as one where there was no mucous sheet vib Nephelometric Turbidity Units.c Daily Light Integral (DLI, mol photons m2 day−1).

rude oil can increase fluid mucus production in corals (Mitchellnd Chet, 1975), and sheets have been observed in Porites lichenxposed to cyanide (Jones and Steven, 1997).

A number of studies have explored the relationship betweenucous sheet formation and sediment, but with contradictory

esults. Bak and Elgershuizen (1976) noted that mucous sheets cane induced in the laboratory in P. astreoides on contact with fineediment. While Coffroth (1985), also found a relationship betweenucous sheet formation and sediment deposition in laboratory-

ased studies with P. astreoides and P. furcata. However, sheetormation also occurred following a reduction of water movement,nd was only observed in response to sediment when combinedith reduced water flow (Coffroth, 1985). Coffroth (1991), andato (1987) could not find any relationship between mucous sheet

ormation and sedimentation in field studies. In the most compre-ensive study of the response of corals to sediments, Stafford-Smithnd Ormond (1992) could not induce mucous sheet formationy fine sediment in both laboratory and in-situ manipulations,oncluding that sheet formation had no apparent relationship toediment flux.

Regular mucous sheet formation has been observed inaribbean Porites spp. (Lewis, 1973), and cyclical mucous sheet for-ation linked to lunar cycles has been recorded in P. cylindrica from

apan (Kato, 1987), and P. furcata and P. astreoides from Panama

Coffroth, 1991). In Porites from Panama, the sheet formationppeared to be preceded by periods of polyp retraction (Coffroth,991). This observation, coupled with the response of Porites toediments in laboratory experiments described previously, led

n the previous survey and therefore had to have formed between surveys.

Coffroth (1988) to suggest sheet formation is an epiphenomenonof some other coral activity pattern, and possibly the expansion,contraction behaviour of the polyps related to suspended sedimentconcentrations (SSCs).

While evidence is currently equivocal, a potential link betweensediment related stresses and mucous sheet production suggeststhat the presence of these sheets on Porites colonies may pro-vide a potential environmental indicator of sub-lethal stress oncorals. In this study, we use information from observations duringa large-scale, capital dredging project to examine the relationshipbetween mucous sheet formation in massive/sub-massive Poritesspp., and dredge related sediment pressures. Specifically, we testwhether temporal patterns and spatial distributions of mucoussheet formation can be explained by timing and proximity to dredg-ing, and whether sheet formation is influenced or driven by otherenvironmental factors such as lunar periodicity and temperature.Subsequently, we test whether mucous sheet formation can beinduced in Porites spp. in a controlled, laboratory-based manip-ulative study, involving exposure to a range of high suspendedsediment concentrations and sediment deposition rates.

2. Materials and methods

2.1. Field study

The field study was conducted before and during a large-scale capital dredging project on the reefs of Barrow Island,located ∼50 km offshore of the Pilbara coast of north-west West-

gical Indicators 77 (2017) 276–285 279

efa(oapwae

ainmapteiq4(Sdscw

ssssfsw

PMifAsimcc6pwam

i(dtw

2

o(Pl

Fig. 2. Stylised representation of tank design and sediment delivery system used forexperimental investigations. The system consisted of 10 experimental tanks whichwere supported by 2 sediment stock tanks and sediment delivery loops. Live feed-backs between turbidity sensor and sediment dosing valve, PAR sensor and customLED lighting; and Deposition sensor, water movement and circulation pumps were

P. Bessell-Browne et al. / Ecolo

rn Australia (Fig. 1). Dredging primarily occurred at 2 locations,or a turning basin near a future material offloading facility (MOF),nd for an entrance channel associated with a liquefied natural gasLNG) jetty (Fig. 1). Dredging was undertaken with a combinationf trailing suction hopper, cutter suction and back hoe dredges,nd typically occurred on a 24 h basis. The material dredged wasredominantly unconsolidated, undisturbed carbonate sediments,hich ranged from rubble to gravelly sand mixed with fine silts

nd clays. For further detailed description of the project see Fishert al. (2015), and Jones et al. (2015).

As part of the state and federal regulatory approval conditionsssociated with the dredging, water quality and coral health mon-toring was conducted at numerous sites, at different distancesorth and south of the primary dredging area (Fig. 1). Water qualityonitoring included measurements of turbidity, light, temperature

nd water depth at ∼10 min intervals, and included an extendedre-dredging baseline period. For further detailed description ofhe water quality monitoring plan see Fisher et al. (2015), and Jonest al. (2015). Coral health monitoring involved regular photograph-ng of individually marked (tagged) corals at each of the wateruality monitoring sites. The sites, which ranged in depth from–11 m, were installed from June 2009 onwards in a pre-dredgingbaseline period), and before dredging began on the 20 May 2010.urveys were undertaken typically every 14 d for the duration ofredging (until 11 November 2011) (Table 1). For most sites, 27–40urveys were undertaken over the dredging period, but for the siteslosest to the dredging (LNG0 and LNGA) only 17 and 23 surveysere conducted respectively (Table 1).

During the dredging there was a near continuous unidirectional,outherly movement of the sediment plumes through the 1.5 ytudy (Evans et al., 2012; Fisher et al., 2015). Analysis of mucousheet formation and water quality was therefore restricted to sitesouth of the MOF and LNG excavation areas, ranging in distancerom 0.2–33.8 km from dredging activity (Fig. 1). Two referenceites (AHC and REFN), located >28 km to the north of the dredgingere also examined (Fig. 1).

Photographs of tagged colonies of massive and sub-massiveorites spp. colonies were examined for the presence of mucus.assive Porites spp. colonies can grow to more than several meters

n diameter, are common throughout the Indo-Pacific, and areound from offshore clear water reefs to coastal, turbid regions.s Porites spp. are difficult to identify underwater due to theirmall and variable corallites (Veron, 2000), the species most likelyncluded a mix of P. lobata and P. lutea, which are the most com-

on massive Porites spp. in the region (Jones, 2016). Porites spp.olonies were assigned to one of seven categories based on mucusoverage, where 1 = 0%, 2 = 1–5%, 3 = 5–35%, 4 = 35–65%, 5 = 65–95%,

= 95–99, and 7 = 100% coverage. The number of mucous sheetsroduced per colony over the dredging phase was also determinedhere a mucous sheet event was defined as any colony classified

s category 3 or greater (>5–35% mucus coverage), and where noucous sheet was observed in the previous survey.

Water quality data were examined using the methods describedn Fisher et al. (2015). Turbidity data were averaged on a daily basis30 min). Photosynthetically active radiation (400–750 nm, PAR)ata was used to calculate a daily light integrals (DLI) as mol pho-ons m−2 d−1 and the 14 d average running mean of these variablesas subsequently calculated.

.2. Laboratory study

All laboratory testing was conducted at the Australian Institute

f Marine Science (AIMS) Sea Simulator (SeaSim) at Cape ClevelandQueensland, Australia), using 50 mm diameter cores of massiveorites spp. colonies. Cores were collected from parent coloniesocated at 3–10 m depth at Davies Reef, a clear-water, mid-shelf

controlled by a programmable logical control (PLC) system.

reef in the central Great Barrier Reef region. The species collectedincluded P. lobata and P. lutea, which were the two most dominantspecies at Davies Reef, and match those present at Barrow Island(Jones, 2016; Veron, 2000). Cores were extracted using a pneu-matic drill, glued onto numbered aragonite coral plugs, and leftfor a 6 week recovery-period in 200 L flow-through holding tanksat the SeaSim before experimentation. Sediment used in the studywas predominantly biogenic, calcium carbonate, collected from theDavies Reef lagoon. The sediment was dried and ground using arod mill grinder until the mean grain size was ∼30 �m (range:0.5–140 �m), as measured using laser diffraction techniques (Mas-tersizer 2000, Malvern instruments Ltd, UK). Total organic contentof the sediment was 0.25%. All coral and sediment were collectedunder GBRMPA permit G13/35758.1.

Mucous sheet formation in response to sediment-exposure wasexamined in 10 fibreglass tanks, each holding 1000 L of water.The tanks received a continuous flow of 0.4 �m filtered seawater(27 ◦C seawater, 33 PSU), at a rate of 400 mL min−1, resulting in0.6 turnovers of water each day. The control system for the exper-iment (custom control logic on Siemens S7-1500 PLC) providedthe integrated management of sediment delivery, light intensity,turbidity and deposition rates (Fig. 2). Sediment was delivered toindividual tanks according to scheduled simulated plumes, con-trolled by real time feedback from turbidity sensors (TurbimaxCUS31, Endress and Hauser) (Fig. 2). Stock sediments were mixedin 500L tanks and kept in suspension by an air diaphragm pump(SandPiper S1F), while being delivered to tanks by a high velocityloop (3 m s−1), and pivoting solenoid valves (Burket, 0331), con-trolled by the PLC (Fig. 2). Sediment was kept in suspension by acentrifuge pump (Iwaki MX400), providing upwelling flow, and awater-movement pump (Hydrowizard ECM63, Panta Rhei), gen-erating trains of static waves (Fig. 2). Sediment deposition eventsof 0, 2, 5, 10 and 20 mg cm−2 d−1 were created by generating ele-vated SSCs before turning off water movement pumps to allow the

sediment to fall out of suspension according to sediment specificparticle settlement velocities. Sediment deposition was measuredin real time with a sediment deposition sensor (Ridd et al., 2001),

280 P. Bessell-Browne et al. / Ecological Indicators 77 (2017) 276–285

Fig. 3. (a) Satellite imagery of the sediment plume on August 29, 2010 with coral health monitoring sites. (b) Mucous sheet production in a Porites spp. colony in the fieldshowing a heavily sediment-laden, opaque mucous sheet, that is peeling away from the tissue surface (see arrow). (c) A colony with an opaque sediment-laden mucous sheetthat is beginning to slough off the surface by water movement (see arrow). (d) A colony with a semi-transparent sediment-laden mucous sheet that is beginning to sloughoff the surface. (e, f) 50 mm (diameter) fragments of Porites spp. in the laboratory, (e) that has begun to shed a sediment laden sheet by peeling of the mucous sheet, and (f)by disintegration of the mucous sheet. Note the extended coral polyps in image (e), as opposed to the retracted state in image (f). (g–k) A colony 0.3 km from the dredgingactivity displaying repeat mucous sheet formation and sloughing. Satellite imagery from the Japan Aerospace Exploration Agency (JAXA) Advanced Land Observing Satellite(

adsfiascbtvtC30eimp

ALOS) Advanced Visible and Near Infrared Radiometer type 2 (AVNIR-2).

nd using sediment pods (SedPods) (Field et al., 2012). Every feways the SedPods were capped and any trapped or accumulatedediment was determined gravimetrically. Sediment samples wereltered through pre-weighed 0.4 �m, 47 mm diameter polycarbon-te filters, incubated at 60 ◦C for ≥24 h, and weighed to determineediment mass. Lighting was provided by a custom designed multi-hip LED (SeaSim) controlled by the PLC, on the feedback providedy the PAR quantum sensor (Skye Instruments, Wales, UK). The con-rol system processed the turbidity along with the time of day andirtual experiment depth (5 m), to provide light exposures similaro those experienced during a dredging project (Jones et al., 2015).orals were exposed to a 12 h Light:Dark period which included a

h period of gradually increasing light levels in the morning (from6:00–09:00 h), and a 3 h period of gradually decreasing light lev-ls in the afternoon (from 15:00 − 18:00 h). The maximum lightntensity in the control (sediment-free) tanks was 400 �mol quanta

−2 s−1. At 12:00 each day for 28 d the Porites spp. cores werehotographed and inspected for the presence of mucous sheets.

2.3. Data analysis

Both in situ and experimental data was analysed using R (ver-sion 3.2.3, R Core Team (2015)) using a complete-subsets modellingapproach where all sensible model combinations were fitted usingthe appropriate statistical methods and subsequently comparedusing Akaike Information Criterion (AICc), AICc weight values (�i)and R2 (Burnham and Anderson, 2002).

For the field study, models were fitted using generalised addi-tive mixed models (GAMM) using the gamm4 package, and eachmucus event was modelled with a binomial distribution (Woodand Scheipl, 2013). We used GAMM rather than linear mixed mod-els to allow for potential non-linear relationships between theresponse variable and the various continuous environmental pre-dictors. Smoothing terms were fit with a cubic spline (Wood, 2006),with ‘k’ argument limited to 5 (to reduce over-fitting and ensure

monotonic relationships that would be ecologically interpretable).A suite of fixed factors was examined to explain mucous sheetpresence (ESM Table S1). Site, colony ID and quarter of the year

P. Bessell-Browne et al. / Ecological Indicators 77 (2017) 276–285 281

F of Porn

wfiismtmcit

uml(imsfitu

3

3

omPttm(r

oD1

ig. 4. The relationship between distance from dredging and (a) the surface area

umbers above bars indicate the total number of sheets produced.

ere included as random factors. Two models sets were fitted,rstly an observational level model investigated the parameters

nfluencing mucous sheet production through time and secondlyite wide metrics were assessed to identify broad scale drivers ofucus. Assumptions were evaluated using residual plots and found

o be adequately met. Following standard convention, the simplestodel within 2 AICc values of the model with the lowest AICc, was

onsidered to be the “best” or optimal model. A null model includ-ng only an intercept and the random factors was also included toest if any of the included factors were indeed useful predictors.

For the laboratory study, a similar full subsets approach wassed to determine the most influential parameters affectingucous sheet production in the experimental setting. Due to the

imited number of unique values for the continuous predictorsonly 5 levels of sedimentation were used), GAMM was deemednappropriate. Instead, models were fitted with generalised linear

ixed effect models (GLMM), where time (week number since thetart of the experiment), and deposition levels were included asxed factors. Both colony ID and tank were included as random fac-

ors. For the full-subsets comparison (based on AIC) models were fitsing the glmer function from the lme4 library (Bates et al., 2015).

. Results

.1. Field study

In the field study, one of the most conspicuous responsesf the corals over the monitoring period was the formation ofucous sheets, occurring exclusively in massive and sub-massive

orites spp. colonies (Fig. 3b–d). The sheets initially appeared asranslucent coatings covering up to 100% the colony surface. Oncehe mucous sheets had formed they became fouled with sedi-

ents making them progressively opaque and easier to observeFig. 3b–d). Often the sheets were observed peeling off the surface,evealing relatively clean, sediment-free coral tissue (Fig. 3b–d).

In the pre-dredging baseline phase only 4 mucous sheets werebserved in 1,198 observations of individual colonies (i.e. 0.33%).uring the dredging phase, 441 mucous sheets were observed in0,600 observations (4.0%, Table 1). Of the events during the dredg-

ites spp. covered in mucus, and (b) the number of new mucous sheets produced,

ing phase, 371 (or ∼85%) were unique events, i.e. new incidences,with no sheet visible in the previous survey (Table 1).

Average mucous sheet prevalence (mean percentage of all indi-viduals with a mucous sheet at any particular time), ranged from0 to 10% over the dredging phase, with prevalence much higher incorals close to the primary excavation area, and decreasing withincreasing distance from dredging (Table 1, Fig. 4a). The highestprevalence at any one time was 48% (10 out of 21 colonies) duringDecember 2010 at the site MOFB, located 0.37 km from the dredg-ing. Over the dredging period, 74 ± 5% (mean ± SE, n = 7 sites, range55–96%) of the Porites spp. colonies <1.5 km from the dredging wereobserved with a mucous sheet (Table 1).

Of the 378 incidences of new mucous sheet formation (i.e.colonies that did not have a mucous sheet in the previous survey),43% were scored as category 3 (i.e. covering 5–30% of the colonysurface). Near complete and complete coverage of the colony (i.e.category 6 or 7, or >95% of the colony) was observed on 68 occa-sions, and of these events 82% (n = 56) occurred in the sites <1.5 kmfrom the dredging area (Table 1, Fig. 4a).

Repeated mucous sheet formation during the dredging phasewas also commonly observed, with ∼50% of the colonies producing3 or more mucous sheets, and ∼25% producing 4 or more (Table 1).Of the colonies producing 3 or more sheets, 90% were located within<1.5 km from the dredging activity (Table 1, Fig. 4b). The maximumnumber of distinct mucus events for an individual colony was 9from a colony at MOFA (located 0.4 km from the dredging).

At the four reference sites, located >20 km either north or southof the main dredging area, mean mucous sheet prevalence was0.2 ± 0.1% (mean ± SE, n = 4 sites) (Table 1, Fig. 4). Only 1 colony wasobserved with mucous coverage >95% of the colony surface area (i.e.category 6), and no colonies produced more than 1 mucous sheet(Fig. 4, ESM Table S2). At one site, REFN, no mucous sheet produc-tion was ever observed in the 34 separate surveys of the 27 coralsconducted over the pre-dredging and dredging phases (totalling766 observations).

These results are confirmed by the site level analysis, whichrevealed the strongest predictor of mucus cover was the maxi-

mum 14-day running mean turbidity (NTU), with this parameterhaving the lowest �AIC value, a weight of 0.53 and an R2 value of0.91 (Table 2b). This relationship demonstrated an upward trend

282 P. Bessell-Browne et al. / Ecological Indicators 77 (2017) 276–285

Fig. 5. The relationship between mucous sheet formation in Porites spp. (red), and the 14 day running mean NTU value (black) with distance from dredging activity. Datapresented is mean ± SE, with n = 17–40 surveys. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

F ng wes ucous

iSaps

tt(bomFfi(

ig. 6. Probability of mucus occurrence with increasing sediment deposition durihaded areas indicating ± 95% confidence limits. Points represent fragment level m

n the probability of mucus with increasing NTU values (ESM Fig.1b). There were other variables that also had elevated R2 valuesnd model weights, with the next 2 best parameters including theresence of sediment in the previous survey and the present ofediment on the colony currently (Table 2b).

The observational level analysis revealed the strongest predic-or of mucus was the presence of sediment on a colony, which hadhe lowest �AIC value, a model weight of 1, and a R2 value of 0.38Table 2a). The relationship between sediment cover and the proba-ility of mucus occurring followed an upward trend, with very lowccurrence when sediment cover is low, and high probability ofucus when sediment covers over 50% of the colony surface (ESM

ig. S1a). The remaining environmental variables provided poorts to the probability of mucus event occurrence during dredging

Table 2a).

ek 1, 2, 3 and 4 of the experimental exposure. Lines represent fitted values with sheet production.

When examining site averaged relationships between mucusprevalence and distance from dredging, a logarithmic trend isobserved with increased prevalence at those sites closest to thedredging activity (Fig. 5). It is apparent that the highest mucusprevalence is located within 1 km of the dredging activity (Fig. 5).The same pattern is observed with 14-day mean turbidity (NTU),with the same logarithmic pattern of decrease apparent (Fig. 5).When comparing mean turbidity with mucus prevalence it is clearthat they are correlated (Fig. 5).

3.2. Laboratory study

Over the 28 d laboratory exposure study, 1100 observa-tions were made of individual Porites spp. fragments (frag-ment × treatment × time), and mucous sheets were recorded on

P. Bessell-Browne et al. / Ecological Indicators 77 (2017) 276–285 283

Table 2Model fit statistics for the relationship between various environmental factors and the presence of mucus on massive Porites spp. colonies, at both an (a) observationallevel through time and (b) site level. Summaries are presented, with �Akaike information criterion (AIC) scores, model weights and R2 relationships. The most influentialparameter for each model set is bold.

Parameter (a) Observational level analysis (b) Site level analysis

�AIC �AIC weight R2 �AIC �AIC weight R2

Sediment presence 0.00 1 0.38 3.95 0.07 0.87Distance from dredging 1176.2 0 0.02 4.45 0.06 0.8614-day mean temperature 1204.6 0 0.00 35.41 0.00 0.26Sediment lag 1208.8 0 0.00 2.51 0.15 0.8914-day mean light stress 1211.6 0 0.01 7.19 0.01 0.8614-day mean NTU 1211.6 0 0.00 0.00 0.53 0.91Days since dredging 1212.9 0 0.00 34.38 0.00 0.08Depth 1214.7 0 0.01 32.15 0.00 0.19Maximum NTU 1215.6 0 0.00 2.62 0.14 0.89Moon phase 1215.9 0 0.00 N/A N/A N/ABleaching lag 1217.5 0 0.00 32.91 0.00 0.15Maximum deposition 1217.5 0 0.00 11.39 0.00 0.80

1tftst(wtentttfceehS

4

itsmtbcbPtispttmmTtd

Mean deposition 1217.5 0

Mean maximum NTU 1218.1 0

Null model 5247.5 0

05 occasions (9.2%). Mucous sheets were only observed in thoseanks with elevated sediment treatments. Sheets became quicklyouled with sediments becoming opaque, and eventually disin-egrating and sloughing off the colony surface, revealing cleanediment-free tissue underneath (Fig. 3e and f). In some instances,he polyps underlying the mucous sheets appeared fully extendedFig. 3e). Of the 105 observations of mucous sheets, 42% (n = 44)ere unique events, i.e. new incidences, with no sheet visible on

he previous day. Sheet production was initially high, with 9 of 60xplants producing a sheet in the first 24 h, decreasing to only 1ew sheet observed between days 4–8 (Fig. 6). New sheet produc-ion increased again in the second half of the study and there was aendency towards the sheets remaining on the fragments for morehan 1 d, increasing their prevalence (Fig. 6). Typically, sheets lastedor 24 h before being sloughed, and ∼72% lasted 3 days or less. Oneolony kept a mucous sheet for 13 d. Full-subsets modelling of thexperimental data suggested the best model included an additiveffect of deposition and time, having an AICc of 530.3 (2.6 AICc unitsigher than the next best model), and weight of 0.68 (ESM Table3).

. Discussion

Since Duerden’s (1906) description of mucous sheet formationn Porites there have been many studies that have investigatedhe cause and ecological significance of the phenomenon. Severaltudies have examined the association between mucous sheet for-ation and sediments, and have produced conflicting results. In

his study, there are multiple lines of convincing evidence fromoth field observations during a large-scale dredging project, andontrolled laboratory-based studies, to suggest a close associationetween sediment loads and mucous sheet presence in massiveorites colonies. Mucous sheets appear to be formed in responseo elevated sediment loads and subsequently trap sediment, caus-ng them to build up on the colony’s surface. The debris ladenheet is episodically sloughed, aided by currents and possibly byolyp expansion patterns, leaving the colony surface compara-ively sediment free. Stafford-Smith and Ormond (1992) notedhat despite having quite poor sediment-rejection capability by

uco-ciliary transport and a morphology prone to sediment accu-

ulation, massive Porites are often found in turbid inshore waters.

hey suggested that Porites must have an alternative strategy forolerating occasional high sediment loads, with mucous sheet pro-uction potentially such a strategy.

0.00 15.35 0.00 0.740.00 6.48 0.02 0.870.00 30.35 0.00 0.00

Mucus sheet prevalence in Porites spp. was examined beforeand during a large-scale, extended (1.5 y) dredging program whichresulted in a 10, 5 and 3-fold increase in the intensity durationand frequency of turbidity events respectively compared to base-line levels (Jones et al., 2015). Water quality showed a strongpower-decay relationship with distance from dredging, with effectsobserved predominantly up to 3 km from dredging (Fisher et al.,2015). Close to the dredging, mucous sheet prevalence reached ashigh as 50%, and over the duration of the dredging ∼75% of colonieswere observed with a mucous sheet at some stage. In contrast, atthe distantly located reference sites (>20 km away), mucous sheetprevalence was negligible (∼0.2%). In addition to mucous sheetpresence, there were also clear spatial patterns in the percent-age of a coral’s surface covered by mucus. High levels of coverageby mucus (i.e. ≥95% of the colony surface) was observed on 68occasions, and 82% of these occurred at sites close to the dredg-ing. Furthermore, the production of multiple mucous sheets alsoappeared correlated with proximity to dredging. For example, over50% of the colonies produced ≥3 sheets over the monitoring periodand of these 90% were located close to the dredging. In contrast, atdistantly located reference sites, no colonies produced more than1 sheet, and only 1 colony was observed with a high level of mucuscoverage. Overall, the strong association between mucous sheetcoverage and prevalence suggests this is a sub-lethal indicator ofsediment stress for corals.

Mucous sheet formation and cyclical sheet formation linked tolunar cycles have been reported previously in Porites cylindrica fromJapan (Kato, 1987), and P. furcata and P. astreoides from Panama(Coffroth, 1991). We found no evidence that mucous sheet produc-tion in the massive, hemispherical Porites spp. investigated (whichincluded Porites lobata and Porites lutea) was linked to lunar cycleor phase of the moon. No relationship between temperature andmucous sheet production was observed, which has also previouslybeen described in some Porites species (Coffroth, 1991; Kato, 1987),despite the occurrence of a marine heat wave (Feng et al., 2013)during the dredging project. The strongest driver of mucous sheetproduction in our data appears to be sediment related.

The laboratory based studies paralleled the field observations,and in response to elevated SSCs and deposition of fine silt Poritesspp. quickly developed mucous sheets, which became progres-sively fouled with sediments. As with field observations, sloughingof mucous sheets resulted in a sediment-free surface. As also

observed in the field studies, the fragments were capable of produc-ing more than 1 mucous sheets, with nearly 30% of the fragments

2 gical I

is

Oergochdnrsgs

tbAmavNcctFsbcsmlrms

oitocfip(sotmie

atdmmaffblfl

84 P. Bessell-Browne et al. / Ecolo

n the higher two sediment treatments producing two new mucousheets over the 28 d exposure period.

These results differ from those presented by Stafford-Smith andrmond (1992), who did not observe mucous sheet production inither the laboratory or field, even when exposing colonies to aain of fine, silt sized particles. The results presented here sug-est that mucous sheet production is an emergency response toverwhelming sediment concentrations, and therefore, the con-entrations used by Stafford-Smith and Ormond (1992) may notave been sufficient to result in sheet production. Our results alsoiffer from those presented by Coffroth (1985), who found no sig-ificant increase in mucous sheet production in Porites astreoides,elated to increased sedimentation, although P. furcata was found toignificantly increase mucous sheet production. These results sug-est that the production of mucous sheets in response to elevatedediment exposure may be species dependent.

Massive and sub-massive Porites spp. typically use a combina-ion of active and passive processes to remove sediments, as haseen reported in numerous coral species (Stafford-Smith, 1993).ctive processes include fluidic mucus production and ciliaryovement (i.e. muco-ciliary transport), and Porites spp. colonies

re likely to be continuously employing this mechanism to pre-ent the build-up of sediment (i.e. smothering) on their surfaces.evertheless, Porites spp. are known to be poor sediment rejecters

ompared to other species, having small calices, producing weakiliary currents, and having little capacity for tissue expansiono actively shed sediments (Stafford-Smith and Ormond, 1992).urthermore, they have a morphology that is naturally prone toediment accumulation (in surface hollows), as opposed to finelyranching species which can easily manipulate sediments to theolony edges and use passive, gravitational forces to self-clean. Con-idering this, Stafford-Smith and Ormond (1992) questioned whyassive Porites spp. can inhabit turbid inshore waters, given their

ow capacity for sediment removal. We suggest the capacity forepeated mucous sheet production and sloughing is an additional

echanism for Porites spp. to self-clean under particularly heavyediment loads.

The cue for mucous sheet formation appears to be the presencef elevated SSCs, and sediment contact with the colony surface. This

s followed or accompanied by a gradual build up and incorpora-ion of sediments into a mucous layer which becomes progressivelypaque with time (Ducklow and Mitchell, 1979). Occasionallyolonies were observed with transparent sheets, (suggesting sheetormation occurs before sediment deposition), although thesencidents were comparatively rare. It has been suggested thatolyp retraction is the ultimate cause of mucous sheet productionCoffroth, 1988), and it may also be related to a cessation in theecretion of mucus, with this hypothesis supported by the resultsf the laboratory investigation. In the laboratory based experimentshe polyps were usually in a highly contracted state under the

ucous sheet, but expansion of the polyps appeared to play a rolen the peeling of the mucous sheet from the surface, suggesting anlement of control by the corals in the mucus sloughing process.

Overall, the strong association between mucus sheet coveragend prevalence suggests this may be a strong sub-lethal indica-or of dredging related sediment stress for corals during futureredging programs or exposure to elevated sediment loads. Theassive, hemispherical Porites spp. investigated here are com-on in turbid inshore and clear-water offshore environments,

re prominent components of reef assemblages, and significantrame-work builders in the Indo-Pacific region (Veron, 2000). Sheetormation is a sub-lethal, organism-level response and exposure

io-indicator that could provide an early warning signals of bio-

ogical effects (Dodge et al., 1974; Lam and Gray, 2003). Sheetormation is rapidly identifiable by scuba divers, or by using diver-ess technologies such as remotely operated vehicles (ROVS), and

ndicators 77 (2017) 276–285

can provide near real time information without the need for sam-pling and subsequent laboratory-based analyses. False positives(Type I errors) are a significant issue for biomarkers, but mucoussheet formation in clear-water environments such as Barrow Islandis typically low (<0.5% prevalence), and can increase 10-fold inresponse to elevated sediment loads.

Further research is required to determine the spatial extent ofthis mucous sheet production, and whether it is also producedin response to runoff related sediment exposure and natural tur-bidity events. Further investigation is also required to determinewhich other species of Porites produce mucous sheets in responseto sediment, and whether different sediment composition, parti-cle size, and colour impacts upon the production of these sheets.Mucous sheet formation and sloughing are likely to be an additionalmechanism used episodically by massive Porites spp. to compli-ment known active and passive sediment rejection mechanisms,and serves to clear the colony surface when sediment loads becometoo high. As a protective reflex response, it is functionally analogousto a human sternutation. In summary, mucous sheet prevalenceappears to be a useful sub-lethal indicator of sediment relatedstress for coral communities and should be considered as a moni-toring tool during future dredging projects and periods of naturallyoccurring elevated turbidity.

Acknowledgements

This research was funded by the Western Australian MarineScience Institution (WAMSI) as part of the WAMSI Dredging Sci-ence Node, and made possible through investment from ChevronAustralia, Woodside Energy Limited, BHP Billiton as environmentaloffsets and by co-investment from the WAMSI Joint Venture part-ners. Additional funding was supplied by a University PostgraduateAward to P.B-B. The commercial entities had no role in data analy-sis, decision to publish, or preparation of the manuscript. The viewsexpressed herein are those of the authors and not necessarily thoseof WAMSI. We thank the staff at the AIMS National Sea Simulator,and in particular C. Humphrey and A. Severati for assistance withexperimental systems.

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.ecolind.2017.02.023.

References

Bak, R., Elgershuizen, J., 1976. Patterns of oil-sediment rejection in corals. Mar. Biol.37, 105–113.

Bates, D., Maechler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effectsmodels using lme4. J. Stat. Softw. 67, 1–48.

Brown, B., Bythell, J., 2005. Perspectives on mucus secretion in reef corals. Mar.Ecol. Prog. Ser. 296, 291–309.

Burnham, K.P., Anderson, D.R., 2002. Model Selection and Multimodel Inference: aPractical Information-Theoretic Approach. Springer, New York.

Bythell, J.C., Wild, C., 2011. Biology and ecology of coral mucus release. J. Exp. Mar.Biol. Ecol. 408, 88–93.

Coffroth, M.A., 1985. Mucous sheet formation on poritid corals: effects of alteredsalinity and sedimentation. In: Gabrie, C.E.A. (Ed.), 5th International Coral ReefCongress. Antenne Museam-EPHE, Moorea, French Polynesia, pp. 165–171.

Coffroth, M.A., 1988. The function and fate of mucous sheets produced by reefcoelenterates. Proceedings of the 6th International Coral Reef Symposium 2,15–20.

Coffroth, M.A., 1990. Mucous sheet formation on poritid corals an evaluation ofcoral mucus as a nutrient source on reefs. Mar. Biol. 105, 39–49.

Coffroth, M.A., 1991. Cyclical mucous sheet formation on poritid corals in the San

Bias Islands, Panama. Mar. Biol. 109, 35–40.

Dodge, R.E., Aller, R.C., Thomson, J., 1974. Coral growth related to resuspension ofbottom sediments. Nature 247, 574–576.

Ducklow, H.W., Mitchell, R., 1979. Bacterial populations and adaptations in themucus layers on living corals. Limnol. Oceanogr. 24, 11.

gical I

DE

E

F

F

F

G

H

H

H

J

J

J

J

J

J

J

K

P. Bessell-Browne et al. / Ecolo

uerden, J.E., 1906. The role of mucus in corals. Q. J. Microsc. Sci. 49, 591–614.dmunds, P.J., Davies, P.S., 1986. An energy budget for Porites porites (Scleractinia).

Mar. Biol. 92, 339–347.vans, R.D., Murray, K.L., Field, S.N., Moore, J.A., Shedrawi, G., Huntley, B.G., Fearns,

P., Broomhall, M., McKinna, L.I., Marrable, D., 2012. Digitise this! A quick andeasy remote sensing method to monitor the daily extent of dredge plumes.PLoS One 7, e51668.

eng, M., McPhaden, M.J., Xie, S.-P., Hafner, J., 2013. La Nina forces unprecedentedleeuwin current warming in 2011. Sci. Rep. 3.

ield, M.E., Chezar, H., Storlazzi, C.D., 2012. SedPods: a low-cost coral proxy formeasuring net sedimentation. Coral Reefs 32, 155–159.

isher, R., Stark, C., Ridd, P., Jones, R., 2015. Spatial patterns in water qualitychanges during dredging in tropical environments. PLoS One 10, e0143309.

lasl, B., Herndl, G., Frade, P., 2016. The microbiome of coral surface mucus has akey role in mediating holobiont health and survival upon disturbance. ISME J.10, 2280–2292.

artnoll, R.G., 1974. The annual cycle of alcyonium digitatum. Esturine CoastalMar. Sci. 3, 71–78.

ubbard, J., Pocock, Y., 1972. Sediment rejection by recent scleractinian corals: akey to palaeo-environmental reconstruction. Int. J. Earth Sci. 61, 598–626.

uettel, M., Wild, C., Gonelli, S., 2006. Mucus trap in coral reefs: formation andtemporal evolution of particle aggregates caused by coral mucus. Mar. Ecol.Prog. Ser. 307.

atkar, A., Brown, B., Bythell, J., Guppy, R., Morris, N., Pearson, J., 2010a. Coralmucus: the properties of its constituent mucins. Biomacromolecules 11,883–888.

atkar, A., Brown, B., Bythell, J., Guppy, R., Morris, N., Pearson, J., 2010b. Measuringmucus thickness in reef corals using a technique devised for vertebrateapplications. Mar. Biol. 157, 261–267.

ohannes, R.E., 1967. Ecology of organic aggregates in the vicinity of a coral reef.Limnol. Oceanogr. 12, 189–195.

ohnston, I., Rohwer, F., 2007. Microbial landscapes on the outer tissue surfaces ofthe reef-building coral Porites compressa. Coral Reefs 26, 375–383.

ones, R.J., Steven, A.L., 1997. Effects of cyanide on corals in relation to cyanidefishing on reefs. Mar. Freshwat. Res. 48, 517–522.

ones, R., Fisher, R., Stark, C., Ridd, P., 2015. Temporal patterns in seawater qualityfrom dredging in tropical environments. PLoS One 10, e0137112.

ones, R.J., 2016. Corals of the North West of Western Australia: Biogeography and

Considerations for Dredging-related Research. Report of Theme 4–Project 4.3,Prepared for the Dredging Science Node. Western Australian Marine ScienceInstitution (WAMSI), Perth, Western Australia.

ato, M., 1987. Mucus-sheet formation and discoloration in the reef-building coral,Porites cylindrica: effects of altered salinity and temperature. Galaxea 6, 1–16.

ndicators 77 (2017) 276–285 285

Krupp, D., 1984. Mucus production by corals exposed during an extreme low tide.Pac. Sci. 38, 1–11.

Lam, P.K., Gray, J.S., 2003. The use of biomarkers in environmental monitoringprogrammes. Mar. Pollut. Bull. 46, 182–186.

Lewis, J.B., Price, W., 1975. Feeding mechanisms and feeding strategies of Atlanticreef corals. J. Zool. 176, 527–544.

Lewis, J.B., 1973. The formation of mucus envelopes by hermatypic corals of thegenus Porites. Caribb. J. Sci. 13, 207–209.

Marcus, J., Thorhaug, A., 1981. Pacific versus Atlantic responses of the subtropicalhermatypic coral Porites spp. to temperature and salinity effects. Proceedingsof the 4th International Coral Reef Symposium, 15–20.

Mayer, F., Wild, C., 2010. Coral mucus release and following particle trappingcontribute to rapid nutrient recycling in a Northern Red Sea fringing reef. Mar.Freshwat. Res. 61, 1006–1014.

Mitchell, R., Chet, I., 1975. Bacterial attack of corals in polluted seawater. Microb.Ecol. 2, 227–233.

R Core Team, 2015. R: a Language and Envrionment for Statistical Computing. RFoundation for statistical computing, Vienna, Austria.

Ridd, P., Day, G., Thomas, S., Harradence, J., Fox, D., Bunt, J., Renagi, O., Jago, C.,2001. Measurement of sediment deposition rates using an optical backscattersensor. Estuar. Coast. Shelf Sci. 52, 155–163.

Rublee, P.A., Lasker, H.R., Gottfried, M., Roman, M.R., 1980. Production andbacterial colonization of mucus from the soft coral Briarium asbestinum. Bull.Mar. Sci. 30, 888–893.

Sofonia, J.J., Anthony, K.R.N., 2008. High-sediment tolerance in the reef coralTurbinaria mesenterina from the inner Great Barrier Reef lagoon (Australia).Estuarine Coastal Shelf Sci. 78, 748–752.

Stafford-Smith, M., Ormond, R., 1992. Sediment-rejection mechanisms of 42species of Australian scleractinian corals. Mar. Freshwat. Res. 43, 683–705.

Stafford-Smith, M.G., 1993. Sediment-rejection efficiency of 22 species ofAustralian scleractinian corals. Mar. Biol. 115, 229–243.

Thompson, J.H., 1980. Responses of Selected Scleractinian Corals to Drilling FluidsUsed in the Marine Environment. PhD Dissertation. Texas A & M University,College Station.

Veron, J., 2000. Corals of the World, Vol. 1–3. Australian Institute of MarineScience, Townsville, Australia.

Wild, C., Huettel, M., Klueter, A., Kremb, S.G., Rasheed, M.Y., Jørgensen, B.B., 2004.Coral mucus functions as an energy carrier and particle trap in the reef

ecosystem. Nature 428, 66–70.

Wood, S., Scheipl, F., 2013. Gamm4: Generalized Additive Mixed Models UsingMgcv and Lme4. R Package Version 0. 2-2.

Wood, S.N., 2006. Generalized Additive Models: an Introduction with R. CRC Press,Boca Raton, FL.