Spatial Distributions of Grazing Activity and MicrophytobenthosReveal Scale-Dependent Relationships Across a SedimentaryGradient

Daniel R. Pratt & Conrad A. Pilditch &

Andrew M. Lohrer & Simon F. Thrush & Casper Kraan

Received: 14 February 2014 /Revised: 29 June 2014 /Accepted: 8 July 2014# Coastal and Estuarine Research Federation 2014

Abstract The density, spatial structure and functional roles ofmacrofaunal and microphytobenthic (MPB) communitieschange across sedimentary gradients. Grazing by macrofaunacan impose considerable top-down control on MPB biomassat the scale of the animal’s feeding ambit (cm scale), yet howrelationships between deposit feeders and MPB scale upacross such transitional environments (10’s m scale) is poorlyunderstood.We determined the relationship between sedimentchlorophyll-a concentration (a proxy of MPB biomass), dis-tance to feeding traces (a proxy of recent deposit feedingactivity) made by the tellinid bivalve Macomona liliana (atcm scale) and macrofaunal densities (at 10’s m scale) across asediment mud content gradient. Correlative relationships, es-timated by generalised least-squares regression, between re-cent deposit feeding activity and MPB biomass were scaledependent and significant only at the site (10’s m) scale. MPBbiomass declined by 28 % as coverage of feeding tracesincreased from 2 to 28 %, with feeding trace area contributingsignificantly to variation in chlorophyll-a (std. coefficient=−0.24, p=0.01). However, the interaction term between the

density of the suspension-feeding clam Austrovenusstutchburyi and sediment mud content explained a largeramount of the variability (std. coefficient=0.72, p<0.001).Our results demonstrate that significant effects on MPB bio-mass can emerge across large, spatially heterogeneous areas oftidal flat, despite appearing stochastic at small scales. Theyalso highlight the need to consider interactions between MPBand macrofauna across abiotic gradients and the potentialroles of non-deposit feeding taxa.

Keywords Deposit feeders . Feeding trace .Macomonaliliana . Sediment mud content . Spatial variability

Introduction

In shallow coastal and estuarine systems, microphytobenthos(MPBs) contribute up to 50 % of the system-wide primaryproduction (Underwood and Kromkamp 1999) and thus con-stitute an important source of labile organic material for ben-thic foodwebs (Middleburg et al. 2000; Kang et al. 2003). Thedistribution of MPB biomass is affected by tidal position,nutrient availability and sediment properties (Guarini et al.1998; Light and Beardall 1998; Jesus et al. 2009; Grinhamet al. 2011) and biological interactions with macrofauna(Chapman et al. 2010), meiofauna (Pinckney and Sandulli1990) and heterotrophic microorganisms (Danovaro et al.2001). These factors operate at different scales creating spa-tially distinct patterns inMPB biomass (Saburova et al. 1995).Moreover, ecological patterns are often generated by process-es operating across different scales. For example, Weermanet al. (2010) demonstrated that interactions between small-scale mucilage production by microbial biofilms and large-scale hydrodynamic processes affect MPB growth. MPB notonly constitute an important food source but influence sedi-ment stability (Van de Koppel et al. 2001) and nutrient fluxes

Communicated by Richard W. Osman

D. R. Pratt :C. A. PilditchDepartment of Biological Sciences, University of Waikato,Private Bag 3105, Hamilton, New Zealand

D. R. Pratt (*) :A. M. Lohrer : S. F. Thrush :C. KraanNational Institute of Water and Atmospheric Research,P.O. Box 11-115, Hamilton, New Zealande-mail: dan.prtt@gmail.com

S. F. ThrushInstitute of Marine Science, University of Auckland,Private Bag 92019, Auckland, New Zealand

C. KraanBiometry and Environmental System Analysis,University of Freiburg, 79106 Freiburg, Germany

Estuaries and CoastsDOI 10.1007/s12237-014-9857-7

(Sundbäck et al. 2000) and play a pivotal role in maintainingfunctional resilience of benthic sediments (Thrush et al. 2012).Thus, identifying environmental factors contributing toMPB biomass distributions will facilitate an understand-ing of processes underlying changes in transitionalenvironments.

MPB are grazed directly at the sediment surface by surface-deposit feeders and by suspension feeders when resuspended(Sauriau and Kang 2000). Deposit feeding by macrofauna canimpose a significant top-down control on MPB biomass(Bianchi and Levinton 1981; Smith et al. 1996; Hagertheyet al. 2002; Lelieveld et al. 2004). In addition to altering MPBbiomass, grazing by benthic macrofauna are also thought toplay an important role in regulating microalgal spatial vari-ability at the scale of the area over which an individual animalgrazes (Hillebrand 2008). Sommer (2000) demonstrated thatbulldozing hydrobiid snails produce biomass-poor graz-ing tracks relative to non-grazed areas of biofilm. Thesespatial patterns are also dependent on the timescale ofthe underlying processes. Whilst deposit feeders caneffectively reduce MPB biomass, MPB have rapid turn-over rates (0.5–2 day−1; Admiraal and Peletier 1980;Smith and Underwood 2000); therefore, a significantgrazing effect on MPB biomass requires that consump-tion is higher than the rate of MPB biomass generation.The cumulative effects of individual deposit feeder-MPBinteractions could have implications for structural prop-erties at scales of several metres. As such, increases inthe populations of macrofaunal grazers have been linkedto reduced microalgal populations and the destabilisingof sediments over areas large enough to affect landscapeformation (De Brouwer et al. 2000; Weerman et al.2011).

Sediment grain size, particularly mud content, has a stronginfluence onMPB biomass (MacIntyre et al. 1996; Jesus et al.2009) and macrofauna community composition (Thrush et al.2003; Anderson 2008) and are linked with numerous othervariables that structure soft-sediment communities and influ-ence their function (Needham et al. 2011; Jones et al. 2011;Pratt et al. 2013). Mud can accumulate in sediments viabiostabilisation processes associated with MPB biomass(Van de Koppel et al. 2001). A small quantity of mud insediments is also potentially favourable to the abundance ofsome deposit and suspension-feeding species, although toomuch can lead to a significant decline (Thrush et al. 2003). Inturn, consumption of MPB and bioturbation associated withforaging activity can destabilise sediments reducing bothMPB and mud content (de Deckere et al. 2001; Ciutatet al. 2007). Despite the potential significance of thesefeedbacks to the transformation of benthic habitats, in-formation on the relationships between deposit-feedingactivity and MPB biomass across sedimentary gradientsis scarce.

Strong, estuary-wide responses in the abundance ofsurface-deposit-feeding macrofauna to changes in MPB andsediment properties have been observed using remote sensingcombined with field sampling techniques (van de Wal et al.2008). However, these relationships are more likely to reflectpatterns in the settling or migration of macrofauna in relationto sediment patches abundant in MPB than grazing effects perse. Moreover, positive effects associated with ecosystem-engineering species (e.g. bioturbation and nutrient excretion)can override the effects of grazing making relationships be-tween MPB and macrofaunal abundance complicated (Lohreret al. 2004). Furthermore, most studies investigating specifi-cally the grazing effects of macrofauna on MPB biomass areconducted in the laboratory (Sommer 2000) or studied inrelatively small areas (<5 m2) in the field (Plante et al.2011). Thus, there is a critical knowledge gap as to howdeposit feeder-MPB relationships scale up from small areaswith limited variability in sediment grain size and macrofau-nal abundance to larger areas that encompass much morevariability with respect to these factors.

In this study, we focus on the effects of deposit feedingactivity by a tellinid bivalve Macomona liliana on MPBbiomass and its spatial variability. This species occurs 1–10 cm below the sediment surface, can form dense beds overlarge areas (Pridmore et al. 1991; Hewitt et al. 1996) and iscommon in intertidal, sandy sediment ecosystems in NewZealand’s North Island. Tellinid bivalve species are also com-mon in many coastal ecosystems around the world. Duringdeposit feeding, M. liliana consume MPB and destabilisesediments at the surface through the movement of their inhal-ant siphon, leaving radial, branching traces (Lelieveld et al.2004). These feeding traces are washed away on the ebb andflood tides, making them useful proxy indicators of recentfeeding activity. Specifically, we aimed to (1) quantify theimpact of deposit-feeding activity (i.e. consumption and sed-iment disturbance) onMPB biomass at a local scale relative tothe deposit feeder’s grazing ambit and (2) determine howdeposit feeding over larger, heterogeneous areas affectsMPB biomass relative to mud content and other sedimentparameters. Additionally, the role of deposit-feedingM. liliana in spatially structuring MPB was contrasted withthat of a suspension feeder, the New Zealand clamAustrovenus stutchburyi. Orvain et al. (2012) speculate thatsuspension feeders may only have a limited effect on thespatial distribution of MPB since it is only available forconsumption following resuspension, assuming that effectsare driven by consumption during grazing and less soby disturbance of sediments or enhancement of nutrientregeneration. However, as organisms rarely perform onefunction in isolation, the bioturbation and destabilisingof sediments by suspension-feeding bivalves (Ciutatet al. 2007) may disturb MPB biomass and affect theirdistributions on the sandflat.

Estuaries and Coasts

Materials and Methods

Study Site

Manukau Harbour is a tidally dominated (mean tidal range=2.8 m) system entering the Tasman Sea on thewest coast of theNorth Island of New Zealand. The estuary covers an area of366 km2, of which 61 % is intertidal. The study site (60×100 m), situated at the mouth of Pukaki Creek adjacent toWairoa Island (Fig. 1), features shellfish beds dominated byM. liliana that exhibit centimetre to metre-scale variation indensity (Hewitt et al. 1996). Field sampling was conductedover 3 days (22nd, 23rd and 25th February 2013 during sunnyand calm weather conditions) in order to minimise the influ-ence of variation in climatic factors. Gradients in sedimentgrain size parameters and variation in feeding trace density andmacrofaunal community structure were evident within a rela-tively small spatial extent (Table 1), thus providing an idealsetting for studying the effects of feeding traces as a proxy ofrecent deposit-feeding activity on the distribution of MPBbiomass relative to other abiotic and biological variables.

Field Sampling and Data Processing

We sampled 55 plots with a 35×35 cm gridded quadrat (gridcell size=25 cm2). Plots were sampled randomly within areas

that were selected to maximise feeding trace density.Sediments were sampled for chlorophyll-a, a proxy of liveMPB biomass, phaeopigment (chlorophyll-a degradationproduct and a proxy of grazed MPB fraction; Cartaxana et al.2003) and a suite of environmental variables including feedingtrace density, macrofaunal abundance and sediment grain size(Table 1). To avoid temporal bias, sampling was alternatedbetween different areas of the site, and geographical coordi-nates for each plot were logged. HOBO data loggers weredeployed at three locations within the site to quantify dailyaverage temperature (°C) and light intensity (lux) to ensure thatsampling had taken place in comparable climate conditions.

Photographs of each plot were taken using a frame-mounteddigital camera to record the feeding traces and other biogenicfeatures on the sediment surface prior to sampling. The cameraframe was fitted with a light tent to provide a diffuse illumi-nation at the sediment surface. In each plot, chlorophyll-a (chl-a, μg cm−2) and phaeopigment (phaeo, μg cm−2) concentra-tions were determined from 16 subsamples extracted to a depthof 1 cm using small cut-off syringe cores (1.4-cm dia.). Thisdepth was chosen because the bulk of active chl-a in sandysediments is contained in the top centimetre (Mitbavkar andAnil 2004; Du et al. 2010) and because it integrates the sub-surface region potentially affected by deposit-feeding bivalves(Volkenborn et al. 2012). Subsamples were taken in all plotsfrom the same randomly predetermined positions (Fig. 2) to

Fig. 1 Location of the study siteadjacent to Wiroa Island (centre)inManukau Harbour (right), NewZealand, and positions of the plotswithin (bottom)

Estuaries and Coasts

make the positions of sample cores relative to feeding traceseasily identifiable. All sediment samples were kept in the dark,transported on ice and stored in the freezer at −18 °C untilanalysis. Sediment chl-a was extracted in 90 % acetone. Chl-asamples were measured on a Turner Designs 10-AU fluorom-eter before and after an acidification step to differentiate be-tween living chl-a biomass from the refractory/degradedphaeopigments (Arar and Collins 1997).

Subsequent to pigment sampling, four randomly positionedsurface sediment cores (2.5-cm dia., 1-cm depth) were sam-pled from each plot and amalgamated for analysis of sedimentgrain size properties (median grain size (MGS) (μm), mudcontent (% <63 μm), organic matter (OM) content (%) andwater content). Sediment grain size was measured on aMalvern Mastersizer-S from sediment samples prepared in a10 % hydrogen peroxide solution to remove organic material.From 32 grain size classes (within the total measured range of0.06–2,000 μm), Shannon-Weiner diversity (GS H′) was cal-culated as an index of grain size diversity (Leduc et al. 2012).OMwas determined as the percentage loss on ignition of driedsediments (24 h at 60 °C) following combustion in a furnace(550 °C) for 5 h (Singer et al. 1988). Percentage water contentwas determined from wet and dry sediment weights.

Macrofaunal community structure was determined fromone core (13-cm dia., 15-cm depth) collected from each plot.Of the 55 macrofauna cores collected, a series of alternatecores (n=24) were analysed to characterise the communitystructure of the study site: All organisms retained on a500-μm sieve were preserved in 70 % isopropyl alcohol andstained with 2 % rose bengal for sorting and identification.Macrofauna were identified to the lowest possible taxonomiclevel. Their abundance values were used to derivemacrofaunaldiversity measures (community abundance, n core−1; taxo-nomic richness, n core−1; Shannon-Weiner diversity, H′core−1). We considered separately the density of the two dom-inant bivalve species M. liliana (MLn) and A. stutchburyi(ASn). All plots were excavated and sieved on a 2-mm meshto derive density and size measurements (shell length, mm) foradultM. liliana andA. stutchburyi. We additionally consideredthe biomass ofM. liliana (MLb) and A. stutchburyi (ASb) sincemetabolic activity and energy requirements are to a largeextent a function of body size (Banse 1982). MLb and ASbper plot were estimated from the relationship between shelllength; L (mm) and ash-free dry weight (AFDW, g):

MLb ¼ aeb⋅L

ASb ¼ a ⋅Lb

where a=0.0023 and b=0.164 for M. liliana (R2=0.82,n=902) and a=6×10−7 and b=3.788 for A. stutchburyi(R2=0.93, n=140) (C. Pilditch, unpublished data).

Table 1 Study site average (and range) values for chl-a and phaeoconcentrations, feeding traces, sediment properties and macrofauna com-munity measures

Variables Units Mean Rangemin–max

Pigment concentration

Chlorophyll-a (chl-a) μg cm−2 26.5 13.2–46.0

Phaeopigment (phaeo) μg cm−2 11.0 2.9–26.2

Biogenic features

FT density (Ftn) n plot−1 30.5 4–53

FT area cover (FTa) % 12.5 1.58–28.5

Burrow density (Bn) n plot−1 40.9 10–95

Burrow area cover (Ba) % 0.67 0.13–2.23

Macrofauna

M. liliana density n plot−1 33.6 20–46

M. liliana biomass g AFDW plot−1 5.63 3.09–10.5

A. stutchburyi density n plot−1 40.2 1–130

A. stutchburyi biomass g AFDW plot−1 1.12 0–3.84

Community abundance† n core−1 129.8 55–262

Taxonomic richness† n core−1 12.6 7–20

Shannon-Weiner diversity† H′ core−1 1.4 0.47–1.95

Sediment characteristics

Median grain size (MGS) μm 207 171–253

Mud content (mud) % <63 μm 9.41 1.9–22.6

Organic matter content (OM) % 1.14 0.59–2.02

Water content % 23.8 19.4–27.8

Data from all plots combined (n=55) with exception of † macrofaunacommunity data which were derived from 24 macrofauna cores

Fig. 2 Layout of the gridded quadrat depicting feeding traces and thepositions of 16 syringe cores. FT+ and FT− denote examples of sedimentcores where feeding traces are, respectively, present and absent. Here,feeding traces comprise 28 % of the area cover

Estuaries and Coasts

Digital Image Analysis

Density and percentage area cover of feeding traces andburrows were determined from the digital images of each plotusing ImageJ (Rasband 2012). Features within the quadratwere counted to derive the density of feeding traces (FTn, nplot−1) and burrows (Bn, n plot−1). Polychaete faecal castswere only observed in four plots and therefore not analysed.The perimeter of the image area occupied by each feedingtrace or burrow was outlined using an oval shape and the areacalculated and summed to give the percentage area cover offeeding traces (FTa, %) and burrows (Ba, %) in each plot. Themeasurement scale was set from the number of pixels relativeto the 35-cm length of the quadrat.

Data Analysis

The local scale (cm) effect of recent deposit feeding on chl-aand phaeo concentrations was determined by assessing differ-ences between subsample cores where feeding traces werepresent (FT+) and in sediments that had not recently beengrazed (FT−). During the image analysis, the position ofsampling cores (1.5 cm2) was superimposed onto the imageof the quadrat grid, and the presence or absence of feedingtraces was noted for each of the 16 subsample cores (Fig. 2).To reduce the effect of spatial heterogeneity in our FT+ andFT− comparisons (chl-a biomass can be patchy at the scale ofa few centimetres; Spilmont et al. 2011), FT+ cores werepaired with the nearest neighbouring FT− core (4–12 cmapart) and compared using a Wilcoxon paired-samples test.

The variability in chl-a and phaeo biomass was quantifiedwithin each plot (<35 cm) and between plots (5–100 m).Within-plot variability in pigment biomass was determinedfrom both the coefficient of variation (CV) and the ratiobetween maximum and minimum biomass (rb, Spilmontet al. 2011), calculated from the 16 subsamples in each plot.Variation in chl-a and phaeo between plots was determinedfrom the mean pigment concentrations of each plot. Thenumber of samples required to accurately estimate the meanchl-a biomass at the quadrat scale was assessed using randomresampling methods by bootstrapping (Grinham et al. 2007).The mean, minimum and maximum and standard errors forthe replicate set were calculated. Standard errors decreasedasymptotically with increasing sample size and indicated thatour sample size was adequate (Appendix 1). Scatterplots wereexamined to determine relationships between predictor vari-ables and response variables: chl-a and phaeo and their CVand rb coefficients.

To account for spatial autocorrelation, we used generalisedleast squares (GLS) regression to incorporate spatial covari-ance between sampling units (see Rangel et al. 2010) todetermine the best predictor variables contributing to variationin MPB biomass between plots. Preliminary analyses, based

on the lowest Akaike information criterion value (AIC), indi-cated that the most significant and parsimonious model wouldbe obtained by fitting a Gaussian autocorrelation function(with optimal nugget, sill and range parameters). Spatial inde-pendence of residual variation was checked usingcorrelograms. Duringmodel selection, we trialled all predictorvariables, their quadratic functions and two-way interactionterms. Sets of competing predictor variables were rankedbased on AIC values after assessing regression diagnosticsfor overall variance inflation (VIF) associated withmulticollinearity. Data were square root (ASb, FTa and Ba),log10 (MLb) and ln (ASn) transformed to improve normalityand reduce skewness of data distributions.

Results

Mean estimates of chl-a and phaeo between plots (site scale)were highly variable; we observed a 3.5-and 9-fold variationbetween minimum and maximum chl-a and phaeo, respec-tively (Table 1). Light intensity within the three samplingperiods (mean=56,260±28,770 lux) varied because of cloudcover; temperature conditions were, however, similar duringthe 3 days of low-tide sampling (mean=30.2±2 °C). LiDARraster data surveyed by the Auckland Regional Council (ARC,http://aucklandcouncil.govt.nz) in 2008 revealed littlevariation in surface elevation across the study site (±0.58 m);thus, it is likely that tidal elevation and emersion period had alimited effect on our measures ofMPB biomass. Our samplingarea incorporated differences in both mud content anddensities in key species M. liliana and A. stutchburyi (seeTable 1), which were found in all plots. Analysingcommunity composition from the macrofauna coresconfirmed that M. liliana (mean length >24 mm) and A.stutchburyi (mean length>15 mm) were the dominantbivalves in terms of size (and biomass) in all plots. Thepolychaete worm Aonides trifida was the most numerousspecies (Appendix 2), comprising >63 % of overallcommunity abundance, but consisted of relatively smallindividuals (<1-mm width).

Local-Scale Effects of Deposit Feeding on MPB Biomass

Differences in chl-a and phaeo were compared in grazed andnon-grazed sediments (n=150 spatially paired cores). Chl-afrom cores with FT+ and FT− conditions were similar(25.2±8.9 and 26.2±9.7 μg cm−2, respectively) and not sta-tistically significant (Wilcoxon, p=0.20). Grazed and non-grazed sediments also contained similar phaeo concentrations(FT+=10.5±7.0 μg cm−2; FT−=10.2±5.7 μg cm−2).

There was a large range in variability for both chl-a(CV=0.08–0.57, rb=1.35–8.51) and phaeo distributions

Estuaries and Coasts

(CV=0.17–0.61, rb=1.77–10.11). However, we found no ev-idence that the observed differences in the variability of chl-awas related to the measured predictor variables by inspectionof scatterplots. In contrast, phaeo distributions appeared to berelated to A. stutchburyi biomass, remaining low and relative-ly homogeneous in plots with higher A. stutchburyi biomass.Specifically, rb values were rarely above 4.5 when ASbexceeded 0.5 g or when plot density was greater than 23individuals (Fig. 3).

Factors Affecting MPB Biomass at the Site Scale

GLS models were run to identify the best predictor variablesexplaining site-scale variation in chl-a and phaeo. The mostparsimonious models explained 79 and 66 % of the variationin chl-a and phaeo, respectively. The interaction term ASn×mud content explained the largest proportion of variability inchl-a (std. coefficient=0.72, p<0.001), indicating that chl-awas higher in sediments that contained both higher levels ofmud and A. stutchburyi density. MGS, water content (posi-tively correlated) and FTa (negatively correlated) were alsoretained in the most parsimonious model (Table 2). Thus, asignificant relationship was observed between deposit feedingand chl-a at the site scale despite being undetected at thewithin-plot scale (FT+ and FT− comparisons). Water contentwas the least important predictor variable retained in the GLSmodel. GLS models identified ASb as the most importantpredictor of phaeo (std. coefficient=0.48, p=0.007); Ba, alsoretained in the model, was a comparatively weak and non-significant predictor (Table 2).

The site-scale spatial distribution of MPB biomass relativeto the best predictor variables identified in the GLS modelswas visually interpreted using spatially interpolated maps(Fig. 4) and bivariate scatterplots (Fig. 5). Given the spatialpatterns observed at the study site, we accounted for autocor-relation when determining the significance values for the

Pearson’s r coefficients, using the Dutilleul (1993) method toestimate the number of degrees of freedom. The outcome wasthat most of these Pearson’s r coefficients between predictorand explanatory variables were non-significant, highlightingthe importance of accounting for autocorrelation. Both chl-aand phaeo concentrations exhibited an along-shore gradientthat followed a similar distribution to the predictor variablesmud, water content and ASn (Figs. 4a, b, f and 5a, b, f). Therewas a high degree of spatial overlap between mud content andpredictor variables ASn (Pearson’s r=0.64, p=0.21), watercontent (r=0.60, p=0.12) and, in particular, sediment grainsize diversity (GS H′) which increased with mud content (r=0.90, p=0.04). Chl-a and phaeo appeared to be higher insediments with a greater GS H′, although these relationshipswere also non-significant after accounting for spatial autocor-relation (r=0.71 and 0.73, p>0.15 for chl-a and phaeo, re-spectively) and were not retained in the most parsimoniousGLS models. Although spatial distributions in chl-a biomasswere dissimilar to those of MLb, a negative relationship be-tween deposit feeding activity (FTa) and chl-a biomass wasobserved at the site scale (Figs. 4c, d and 5c, d). We observedlinear increases in FTn relative to MLn (Pearson’s r=0.56,p<0.001), and FTa was also correlated with MLb (Pearson’sr=0.54, p=0.02). Despite higher chl-a concentrations in

Fig. 3 Maximal variability of phaeopigment concentration (rb) relativeto A. stutchburyi density (n plot−1)

Table 2 Ordinary least squares (OLS) and generalised least squares(GLS) model results between environmental predictor variables andchlorophyll-a and phaeopigment concentrations

OLS GLS Stdcoef.

Stderror

t p value

Chl-a

ln (ASn)×Mud 2.10 2.16 0.72 0.38 5.63 <0.001

MGS 2.25 2.43 0.60 0.45 5.40 <0.001

sqrt (FTa) −21.01 −17.54 −0.24 6.58 −2.67 0.01

% water −7.47 −8.43 −0.20 3.93 −2.15 0.04

Phaeo

sqrt (ASb) 50.66 40.1 0.48 14.26 2.81 0.007

sqrt (Ba) 49.03 25.9 0.14 18.98 1.36 0.18

Total variance explained for chl-a, OLS R2 =0.76, GLS R2 =0.79; phaeo,OLS R2 =0.59, GLS R2 =0.66. Data from all plots combined (n=55)

OLS, GLS and Std coef. are ordinary least squares, generalised leastsquares and standardised slope coefficients, respectively, ASnA. stutchburyi abundance (n plot−1 ), ASb A. stutchburyi biomass (gAFDW plot−1 ), Ba burrow area cover (%), FTa feeding trace area cover(%), MGS median grain size (μm), Mud mud content (% <63 μm), %water water content

�Fig. 4 Spatial distributions chl-a (μg cm−2) (a, c, e) and phaeo (μg cm−2)(b, d, f) relative to amud content (%), b A. stutchburyi density (n plot−1),c M. liliana density (n plot−1), d feeding trace area cover (%), e mediangrain size (μm) and f % water content. Chl-a and phaeo were spatiallyinterpolated using ordinary kriging fitted with a Gaussian semivariogrammodel in ArcGIS; predictor variables are superimposed (sphere size isrelative to predictor variable range)

Estuaries and Coasts

Estuaries and Coasts

sediments with higher mud content, FTa appeared to be lowerin these sediments (Pearson’s r=0.50, p=0.14).

Discussion

Our results suggest that non-trophic interactions of macrofau-na may play a greater role in determining spatial distributions

in MPB biomass than the direct effects of deposit feeding. Wehave demonstrated that the feeding traces of M. liliana areunrelated to MPB biomass at the local (centimetre) scale butthat they show a negative relationship at the site (10’s m) scalethat incorporates sedimentary gradients. Considering the al-ternate functional roles of M. liliana (deposit feed directly onMPB) and A. stutchburyi (a water column suspension feeder),we expected a stronger influence of deposit feeding in

Fig. 5 Estimates of chl-a (open) and phaeo (filled) relative to a mudcontent (%), b A. stutchburyi density (n plot−1), c M. liliana density (nplot−1), d feeding trace area cover (%), e median grain size (μm) and f%

water content. Pearson’s r coefficients and significance (p) termscorrected for spatial autocorrelation are displayed

Estuaries and Coasts

determining the spatial structure in MPB biomass. By con-trast, our estimate of deposit-feeding activity at the site scalewas less important than the interaction between mud contentand A. stutchburyi density, which both exhibited along-shore gradients that were similar to MPB biomass.These relationships do not necessarily imply causation,but our results highlight functional links that providevaluable insights for determining the processes underly-ing MPB distributions.

Whilst deposit feeding significantly contributed to variabil-ity in chlorophyll-a at the site scale, it was a secondary factorto sediment properties and A. stutchburyi density. Sedimentdisturbance associated with foraging activity (i.e. bioturba-tion) by infauna could predominate over the effects of con-sumption on MPB biomass (Pillay et al. 2009). Given theirdensity, size and near-surface position in the sediment,A. stutchburyi have a higher bioturbation potential thanM. liliana, and the reduction in the variability in phaeo (CVand rb) with increasing A. stutchburyi biomass suggests thatsediments with high clam densities were well mixed.Bioturbation by infaunal suspension-feeding bivalves is gen-erally considered to have a destabilising effect on sedimentsby remobilising and resuspending fine fractions at the surface(Ciutat et al. 2007; Montserrat et al. 2009). In addition to theloss of MPB via sediment resuspension, microalgae are themain source of food for clams (Kang et al. 1999); therefore,we expected lower pigment concentrations in areas with highclam density. Conversely, we observed a two-fold increase inchlorophyll-a in conjunction with an increase inA. stutchburyidensity from 1 to 130 individuals per plot. Furthermore, therewas a strong spatial overlap between A. stutchburyi and mudcontent, and the interaction between these two factors was thestrongest predictor of chlorophyll-a in our models. Clams maybenefit MPB by enhancing nutrient availability through ex-cretion of NH4

+ (Sandwell et al. 2009; Jones et al. 2011) andindirectly via remineralisation of biodeposits (Newell et al.2002; Giles and Pilditch 2006). They may also increase sed-iment stability through biodeposits of silt and organic material(Widdows et al. 2004) and perhaps by physically armouringsediments against hydrodynamic erosion and trappingfine particles (proposed by Donadi et al. 2013), whichwill be of benefit to MPB. The effects of suspensionfeeding and biodeposition on particle fluxes to the seabed can be equally or more important than abioticsedimentat ion processes (Ahn 1993; Graf andRosenberg 1997). Here, the elevated concentrations ofphaeopigments (potentially originating from both phyto-plankton and resuspended MPB) in sediments containinghigh numbers of A. stutchburyi could be indicative ofsuch biodeposition effects.

Spatial relationships between MPB biomass and meio- andmacrofauna abundance are theoretically linked by the re-source dependence of consumers and the disturbance effects

of grazing activities, but significant correlations between thesefactors are oinconsistencies between animal density andftenlacking (e.g. Decho and Fleeger 1988; Pinckney and Sandulli1990; Chapman et al. 2010). Such inconsistencies betweenanimal density and the effects of grazing on MPB biomassmay arise from varying feeding rates or alternate modes offeeding behaviour among different sediment types and animaldensities (Olafsson 1986; Woodin et al. 2012). Here, therelationship between chlorophyll-a concentration andM. liliana density differed from previous experimental stud-ies. Negative effects of M. liliana density on chlorophyll-aoccurred when the manipulated density was three times higherthan ambient levels (176 individuals per 0.25m2) in sedimentscontaining 9–24 % mud (Lelieveld et al. 2004). Conversely,positive effects have also been observed at much lowerM. liliana densities (~18 individuals per 0.25 m2) in coarsersediments (mud content <5.3 %, Thrush et al. 2006). In ourstudy, M. liliana density (4–52 individuals per 0.25 m2) wasnot an important explanatory variable ofMPB biomass. Thesedifferent outcomes could be attributed to differences in spatialscale (previous studies were confined to small experimentalplots within a <300-m2 study area), organism densities andsediment properties (lower versus higher mud content).Together, these results also imply that animal densities alonedo not provide a sufficient indicator of the effects ofM. liliana-grazing pressure on the standing stock of MPB.

Patterns between grazer abundances and MPB biomassmay be confounded by temporal lags between sampling andthe interaction between the deposit feeder andMPB (Pinckneyand Sandulli 1990). MPB can turn over very quickly, poten-tially doubling their biomass within a day (Admiraal andPeletier 1980; Smith and Underwood 2000), and are able torapidly migrate and recolonise grazed sediments within a fewhours (Plante et al. 2010). Our approach of quantifying recentdeposit feeding by means of ephemeral feeding traces thatindicate that feeding had occurred just a few hours prior tosampling (after site emergence and prior to tidal inundation)aimed to minimise such complications by identifyingdirectly the action of deposit feeding, thus limiting timelag effects. As such, we found relatively weak butsignificant correlations between MPB biomass and feed-ing traces across our study site despite the fact thatchlorophyll-a concentration was not significantly relatedto the density or biomass of M. liliana.

We predicted that deposit feeding would have a largereffect at local scales than at the site scale because of theincreased variability in abiotic factors that occurs at largerscales (sensu stricto Saburova et al. 1995). Surprisingly, wefound the opposite pattern, that is, deposit feeder-MPB bio-mass relationships emerged at the site scale despite appearingstochastic at local scales. Our results differ from a laboratorystudy conducted by Sommer (2000) that showed how depositfeeding by hydrobiid snails increased spatial heterogeneity of

Estuaries and Coasts

MPB by locally reducing biomass in grazing tracks. Theabsence of local effects of deposit feeding on MPBbiomass and within-plot variability (determined fromCV and rb values) is potentially the result of overridingmicroscale processes, such as competitive interactionsbetween microalgae (Saburova et al. 1995) and grazingby meiofauna (Sandulli and Pinckney 1999) and othermacrofauna species.

The scale dependence of deposit feeder-MPB biomassrelationships may be explained by differences in their interac-tions between different habitat types (sandy and muddy sandsediments). Feeding trace area cover (FTa) was lower inmuddier sediments, whereas MPB biomass was higher inmuddier sediments. If lower area cover of feeding traces inmuddier sediments was resulting from lower foraging effortrequired to consume MPB, we would expect to observe equalor higher M. liliana biomass living in muddier sedimentscompared with sand, but this was not the case. The recoveryrates of MPB are strongly dependent on environmental con-ditions. MPB biomass turnover in sandy sediments can beapproximately seven times higher compared to silty sands(Huettel and Rusch 2000; Middleburg et al. 2000).Potentially, lower MPB recovery rates in muddier sedimentsmay have weakened the local-scale differences in MPB bio-mass between grazed and non-grazed sediments (if grazinghistory was retained longer in sediments outside the feedingtrace areas) but may strengthen overall relationships betweendeposit-feeding activity and MPB biomass between plots.Thus, encompassing a broader sediment range in our analysisas we move from the plot to the site scale may be important inexplaining the scale dependence of deposit feeder-MPBrelationships.

Sediment grain size properties have been described asprimary factors in explaining large-scale MPB distribu-tions (e.g. Delgado 1989; Brotas et al. 1995; Orvainet al. 2012). Here, mud content was also a significantpredictor variable, which we expected given the rela-tively stable environment provided by muddier sedi-ments and positive feedbacks involving the accumula-tion of mud and MPB via biostabilising processes (vande Koppel et al. 2001). Additionally, MPB biomass wasalso positively related to median grain size and particlesize diversity. Mud flocs adhered to coarse sand grainscan provide optimal habitat conditions for MPB biomassaccumulation (de Jonge 1985); therefore, sediment het-erogeneity could also be an important variable inexplaining MPB biomass distributions.

In conclusion, our study demonstrates that biological inter-actions affecting MPB biomass distribution, such as depositfeeding, may emerge at the site scale despite appearing sto-chastic at local scales. The muddier sediments contained highdensities of suspension feeders that feed on phytoplankton andresuspended MPB in the water column, rather than directly

from the sediment surface. A. stutchburyi often comprise alarge proportion of secondary-producer biomass in NewZealand intertidal sandflats, and their significance in modify-ing sediment properties and macrofaunal community structureand facilitating primary production has been demonstrated inmultiple studies (Thrush et al. 2006; Sandwell et al.2009; Jones et al. 2011; Pratt et al. 2013). Our com-bined understanding of interactive processes derivedfrom these experimental studies and the distributionalpatterns observed in this study highlight the potentialimportance of biological interactions integrated withinsedimentary gradients to the functioning of soft-sediment ecosystems. Biological factors are often ig-nored at larger scales because the framework of scalarhierarchy maintains that biological factors play a minorrole at the landscape scale compared with abiotic factorssuch as sediment properties (e.g. Saburova et al. 1995).In the light of our results, we suggest that role ofmacrofauna-sediment interactions in contributing tolarge-scale variability in biological communities shouldbe considered more carefully in future studies.

Acknowledgments We would like to thank Lisa McCartain, RebeccaGladstone-Gallagher, Raymond Tana, Dudley Bell and Andrew Jones forfieldwork and Barry Greenfield, Sarah Hailes, Kelly Carter and BarryO’Brien for technical assistance. We also thank the editor and twoanonymous reviewers for their useful comments. This project was fundedby National Institute of Water and Atmospheric Research (NIWA) underCoasts and Oceans Research Programme 3 (2013/14 SCI) and progenitorprogrammes, a NIWA PhD scholarship (Foundation for Research, Sci-ence and Technology (FRST) Project No. C01X0501) and the Universityof Waikato.

Appendix 1

Fig. 6 Bootstrap-generated standard error (mean, minimum and maxi-mum) values for chlorophyll-a concentrations across a range of subsam-ple sizes (n) across ten randomly selected plots

Estuaries and Coasts

Appendix 2

References

Admiraal, W., and H. Peletier. 1980. Distribution of diatom species on anestuarine mud flat and experimental analysis of the selective effectof stress. Journal of Experimental Marine Biology and Ecology 46:157–175.

Ahn, I.Y. 1993. Enhanced particle flux through the biodeposition by theAntarctic suspension-feeding bivalve Laternula elliptica in MarianCove, KingGeorge Island. Journal of Experimental Marine Biologyand Ecology 171: 75–90.

Anderson, M.J. 2008. Animal-sediment relationships re-visited:characterising species’ distributions along an environmental gradi-ent using canonical analysis and quantile regression splines. Journalof Experimental Marine Biology and Ecology 366: 16–27.

Arar, E.J. and G.B. Collins. 1997. Method 445.0: in vitro determinationof chlorophyll a and pheophytin a. Marine and freshwater algae byfluorescence. US Environmental Protection Agency, Office ofResearch and Development, National Exposure ResearchLaboratory.

Banse, K. 1982. Mass-scaled rates of respiration and intrinsic growth invery small invertebrates.Marine Ecology Progress Series 9: 281–97.

Bianchi, T., and J. Levinton. 1981. Nutrition and food limitation ofdeposit-feeders II. Differential effects of Hydrobia totteni andIlyanassa obsoleta on the microbial community. Journal ofMarine Research 39: 547–556.

Brotas, V., T. Cabrita, A. Portugal, J. Serôdio, and F. Catarino. 1995.Spatio-temporal distribution of the microphytobenthic biomass inintertidal flats of Tagus Estuary (Portugal).Hydrobiologia 300–301:93–104.

Cartaxana, P., B. Jesus, and V. Brotas. 2003. Pheophorbide andpheophytin a-like pigments as useful markers for intertidalmicrophytobenthos grazing by Hydrobia ulvae. Estuarine, Coastaland Shelf Science 58: 293–297.

Chapman, M.G., T.J. Tolhurst, R.J. Murphy, and A.J. Underwood. 2010.Complex and inconsistent patterns of variation in benthos, micro-algae and sediment over multiple spatial scales. Marine EcologyProgress Series 398: 33–47.

Ciutat, A., J. Widdows, and J.W. Readman. 2007. Influence of cockleCerastoderma edule bioturbation and tidal-current cycles on resus-pension of sediment and polycyclic aromatic hydrocarbons.MarineEcology Progress Series 328: 51–64.

Danovaro, R., M. Armeni, A. Dell'Anno, M. Fabiano, E. Manini, D.Marrale, A. Pusceddu, and S. Vanucci. 2001. Small-scale distribu-tion of bacteria, enzymatic activities, and organic matter in coastalsediments. Microbial Ecology 42: 177–185.

De Brouwer, J.F.C., S. Bjelic, E.M.G.T. de Deckere, and L.J. Stal. 2000.Interplay between biology and sedimentology in a mudflat(Biezelingse Ham, Westerschelde, The Netherlands). ContinentalShelf Research 20: 1159–1177.

De Deckere, E., T. Tolhurst, and J. De Brouwer. 2001. Destabilization ofcohesive intertidal sediments by infauna. Estuarine, Coastal andShelf Science 53: 665–669.

De Jonge, V. 1985. The occurrence of ‘epipsammic’ diatom populations:a result of interaction between physical sorting of sediment andcertain properties of diatom species. Estuarine, Coastal and ShelfScience 21: 607–622.

Decho, A.W., and J.W. Fleeger. 1988. Microscale dispersion ofmeiobenthic copepods in response to food-resource patchiness.Journal of Experimental Marine Biology and Ecology 118: 229–243.

Table 3 Macrofaunal species, mean abundances, size and rate of occurrence

Phylum Species Occurrence (% cores) Mean abundance (n core−1) Mean size (mm)

Mollusca Macomona liliana

Total 1.00 9.04 14.9

Adults (>5 mm) 1.00 5.00 24.1

Austrovenus stutchburyi

Total 1.00 6.92 10.0

Adults (>5 mm) 0.71 6.47 15.1

Paphies australis 0.33 6.63 4.3

Nucula hartvigiana 0.50 5.83 5.3

Zeacumantus lutulentus 0.75 2.78 11.6

Cnidaria Anthopleura aureoradiata 0.42 3.20 –

Crustacea Halicarcinus whitei 0.42 1.60 –

Colorostylis lemurum 0.46 1.45 –

Annelida Aonides trifida 1.00 82.0 –

Prionospio aucklandica 0.50 4.67 –

Heteromastus filiformis 0.63 3.27 –

Magelona dakini 0.58 3.00 –

Nicon aestuariensis 0.92 5.09 –

Sphaerosyllis semiverrucosa 0.54 7.69 –

Nemertea 0.46 2.00 –

Commonly occurring species from our study site (occurrence rate >0.4), estimated from macrofauna cores (n=24)

Estuaries and Coasts

Delgado, M. 1989. Abundance and distribution of microphytobenthos inthe bays of Ebro Delta (Spain). Estuarine, Coastal and Shelf Science29: 183–194.

Donadi, S., J.Westra, E.Weerman, T. Heide, E. Zee, J. Koppel, H. Olff, T.Piersma, H. Veer, and B. Eriksson. 2013. Non-trophic interactionscontrol benthic producers on intertidal flats. Ecosystems 16: 1325–1335.

Du, G.Y., M. Son, S. An, and I.K. Chung. 2010. Temporal variation in thevertical distribution of microphytobenthos in intertidal flats of theNakdongRiver estuary, Korea.Estuarine, Coastal and Shelf Science86: 62–70.

Dutilleul, P. 1993. Modifying the t test for assessing the correlationbetween two spatial processes. Biometrics 49: 305–314.

Giles, H., and C.A. Pilditch. 2006. Effects of mussel (Perna canaliculus)biodeposit decomposition on benthic respiration and nutrient fluxes.Marine Biology 150: 261–271.

Graf, G., and R. Rosenberg. 1997. Bioresuspension and biodeposition: areview. Journal of Marine Systems 11: 269–278.

Grinham, A.R., T.J. Carruthers, P.L. Fisher, J.W. Udy, and W.C.Dennison. 2007. Accurately measuring the abundance of benthicmicroalgae in spatially variable habitats. Limnology andOceanography: Methods 5: 119–125.

Grinham, A.R., D. Gale, and J.W. Udy. 2011. Impact of sediment type,light and nutrient availability on benthic diatom communities of alarge estuarine bay: Moreton Bay, Australia. Journal ofPaleolimnology 46: 511–523.

Guarini, J.M., G. Blanchard, C. Bacher, P. Gros, P. Riera, P. Richard, D.Gouleau, R. Galois, J. Prou, and P.G. Sauriau. 1998. Dynamics ofspatial patterns of microphytobenthic biomass: inferences from ageostatistical analysis of two comprehensive surveys in Marennes-Oléron Bay (France). Marine Ecology Progress Series 166: 131–141.

Hagerthey, S.E., E.C. Defew, and D.M. Paterson. 2002. Influence ofCorophium volutator and Hydrobia ulvae on intertidal benthic dia-tom assemblages under different nutrient and temperature regimes.Marine Ecology Progress Series 245: 47–59.

Hewitt, J.E., S.F. Thrush, V.J. Cummings, and R.D. Pridmore. 1996.Matching patterns with processes: predicting the effect of size andmobility on the spatial distributions of the bivalves Macomonaliliana and Austrovenus stutchburyi. Marine Ecology ProgressSeries 135: 57–67.

Hillebrand, H. 2008. Grazing regulates the spatial variability of periphy-ton biomass. Ecology 89(1): 165–173.

Huettel, M., and A. Rusch. 2000. Transport and degradation of phyto-plankton in permeable sediment. Limnology and Oceanography 45:534–549.

Jesus, B., V. Brotas, L. Ribeiro, C.R. Mendes, P. Cartaxana, and D.M.Paterson. 2009. Adaptations of microphytobenthos assemblages tosediment type and tidal position. Continental Shelf Research 29:1624–1634.

Jones, H.F.E., C.A. Pilditch, D.A. Bruesewitz, and A.M. Lohrer. 2011.Sedimentary environment influences the effect of an infaunal sus-pension feeding bivalve on estuarine ecosystem function. PLoSONE 6. doi:10.1371/journal.pone.0027065.

Kang, C.K., P.G. Sauriau, P. Richard, and G.F. Blanchard. 1999. Foodsources of the infaunal suspension-feeding bivalve Cerastodermaedule in a muddy sandflat of Marennes-Oléron Bay, as determinedby analyses of carbon and nitrogen stable isotopes.Marine EcologyProgress Series 187: 147–158.

Kang, C.K., J.B. Kim, K.S. Lee, J.B. Kim, P.Y. Lee, and J.S. Hong. 2003.Trophic importance of benthic microalgae to macrozoobenthos incoastal bay systems in Korea: dual stable C and N isotope analyses.Marine Ecology Progress Series 259: 79–92.

Leduc, D., A.A. Rowden, K. Probert, C.A. Pilditch, S.D. Nodder, A.Vanreusel, G.C.A. Duineveld, and R. Witbaard. 2012. Further evi-dence for the effect of particle-size diversity on deep-sea benthic

biodiversity. Deep Sea Research Part I: Oceanographic ResearchPapers 63: 164–169.

Lelieveld, S.D., C.A. Pilditch, and M.O. Green. 2004. Effects of deposit-feeding bivalve (Macomona liliana) density on intertidal sedimentstability. New Zealand Journal of Marine and Freshwater Research38: 115–128.

Light, B.R., and J. Beardall. 1998. Distribution and spatial variation ofbenthic microalgal biomass in a temperate, shallow-water marinesystem. Aquatic Botany 61: 39–54.

Lohrer, A.M., S.F. Thrush, and M.M. Gibbs. 2004. Bioturbators enhanceecosystem function through complex biogeochemical interactions.Nature 431: 1092–1095.

MacIntyre, H.L., R.J. Geider, and D.C.Miller. 1996. Microphytobenthos:the ecological role of the “secret garden” of unvegetated, shallow-water marine habitats. I. Distribution, abundance and primary pro-duction. Estuaries 19: 186–201.

Middleburg, J.J., C. Barranguet, H.T.S. Boschker, P.M.J. Herman, T.Moens, and C.H.R. Heip. 2000. The fate of intertidalmicrophytobenthos carbon: an in situ 13C-labeling study.Limnology and Oceanography 45: 1224–1234.

Mitbavkar, S., and A. Anil. 2004. Vertical migratory rhythms of benthicdiatoms in a tropical intertidal sand flat: influence of irradiance andtides. Marine Biology 145: 9–20.

Montserrat, F., C. Van Colen, P. Provoost, M.Milla, M. Ponti, K. Van denMeersche, T. Ysebaert, and P.M.J. Herman. 2009. Sediment segre-gation by biodiffusing bivalves. Estuarine, Coastal and ShelfScience 83: 379–391.

Needham, H.R., C.A. Pilditch, A.M. Lohrer, and S.F. Thrush. 2011.Context-specific bioturbation mediates changes to ecosystem func-tioning. Ecosystems 14: 1096–1109.

Newell, R.I., J.C. Cornwell, and M.S. Owens. 2002. Influence of simu-lated bivalve biodeposition and microphytobenthos on sedimentnitrogen dynamics: a laboratory study. Limnology andOceanography 47: 1367–1379.

Olafsson, E.B. 1986. Density dependence in suspension-feeding anddeposit-feeding populations of the bivalveMacoma balthica: a fieldexperiment. Journal of Animal Ecology 55: 517–526.

Orvain, F., S. Lefebvre, J. Montepini, M. Sébire, A. Gangnery, and B.Sylvand. 2012. Spatial and temporal interaction between sedimentand microphytobenthos in a temperate estuarine macro-intertidalbay. Marine Ecology Progress Series 458: 53–68.

Pillay, D., G.M. Branch, and A. Steyn. 2009. Complex effects of thegastropod Assiminea globulus on benthic community structure in amarine-dominated lagoon. Journal of Experimental Marine Biologyand Ecology 380: 47–52.

Pinckney, J., and R. Sandulli. 1990. Spatial autocorrelation analysis ofmeiofaunal and microalgal populations on an intertidal sandflat:scale linkage between consumers and resources. Estuarine,Coastal and Shelf Science 30: 341–353.

Plante, C.J., S. Feipel, and J.L. Wilkie. 2010. Disturbance effects ofdeposit feeding onmicroalgal community structure andmechanismsof recolonisation. Journal of Phycology 46: 907–916.

Plante, C.J., E. Frank, and P. Roth. 2011. Interacting effects of depositfeeding and tidal resuspension on benthic microalgal communitystructure and spatial patterns. Marine Ecology Progress Series 440:53–65.

Pratt, D.R., A.M. Lohrer, C.A. Pilditch, and S.F. Thrush. 2013. Changesin ecosystem function across sedimentary gradients in estuaries.Ecosystems 17: 182–194.

Pridmore, R.D., S.F. Thrush, R.J.Wilcock, T.J. Smith, J.E. Hewitt, and V.J.Cummings. 1991. Effect of the organochlorine pesticide technicalchlordane on the population structure of suspension and depositfeeding bivalves. Marine Ecology Progress Series 76: 261–271.

Rangel, T.F., J.A.F. Diniz-Filho, and L.M. Bini. 2010. SAM: a compre-hensive application for spatial analysis in macroecology. Ecography33: 46–50.

Estuaries and Coasts

Rasband, W.S. 2012. ImageJ: image processing and analysis in Java.Astrophysics Source Code Library 1: 06013.

Saburova, M.A., I.G. Polikarpov, and I.V. Burkovsky. 1995. Spatialstructure of an intertidal sandflat microphytobenthic community asrelated to different spatial scales. Marine Ecology Progress Series129(1): 229–239.

Sandulli, R., and J. Pinckney. 1999. Patch sizes and spatial patterns ofmeiobenthic copepods and benthic microalgae in sandy sediments: amicroscale approach. Journal of Sea Research 41: 179–187.

Sandwell, D.R., C.A. Pilditch, andA.M. Lohrer. 2009. Density dependenteffects of an infaunal suspension-feeding bivalve (Austrovenusstutchburyi) on sandflat nutrient fluxes and microphytobenthic pro-ductivity. Journal of Experimental Marine Biology and Ecology373: 16–25.

Sauriau, P.G., and C.K. Kang. 2000. Stable isotope evidence of benthicmicroalgae-based growth and secondary production in the suspen-sion feeder Cerastoderma edule (Mollusca, Bivalvia) in theMarennes-Oléron Bay. Hydrobiologia 440: 317–329.

Singer, J., J. Anderson, M. Ledbetter, I. McCave, K. Jones, and R.Wright. 1988. An assessment of analytical techniques for the sizeanalysis of fine-grained sediments. Journal of SedimentaryPetrology A 58: 534–543.

Smith, D.J., and G.J.C. Underwood. 2000. The production of extracellularcarbohydrates by estuarine benthic diatoms: the effects of growth phaseand light and dark treatment. Journal of Phycology 36: 321–333.

Smith, D.J., R. Hughes, and E. Cox. 1996. Predation of epipelic diatomsby the amphipod Corophium volutator and the polychaete Nereisdiversicolor. Marine Ecology Progress Series 145: 53–61.

Sommer, U. 2000. Benthic microalgal diversity enhanced by spatialheterogeneity of grazing. Oecologia 122: 284–287.

Spilmont, N., L. Seuront, T. Meziane, and D.T. Welsh. 2011. There’smore to the picture than meets the eye: sampling microphytobenthosin a heterogeneous environment. Estuarine, Coastal and ShelfScience 95: 470–476.

Sundbäck, K., A. Miles, and E. Goeransson. 2000. Nitrogen fluxes,denitrification and the role of microphytobenthos in microtidalshallow-water sediments: an annual study. Marine EcologyProgress Series 200: 59–76.

Thrush, S.F., J.E. Hewitt, A. Norkko, P.E. Nicholls, G.A. Funnell, and J.I.Ellis. 2003. Habitat change in estuaries: predicting broad-scale

responses of intertidal macrofauna to sediment mud content.Marine Ecology Progress Series 263: 101–112.

Thrush, S.F., J.E. Hewitt, M. Gibbs, C. Lundquist, and A. Norkko. 2006.Functional role of large organisms in intertidal communities: com-munity effects and ecosystem function. Ecosystems 9: 1029–1040.

Thrush, S.F., J.E. Hewitt, and A.M. Lohrer. 2012. Interaction networks incoastal soft-sediments highlight the potential for change in ecolog-ical resilience. Ecological Applications 22: 1213–1223.

Underwood, G., and J. Kromkamp. 1999. Primary production by phyto-plankton and microphytobenthos in estuaries. Advances inEcological Research 29: 93–153.

Van de Koppel, J., P.M. Herman, P. Thoolen, and C.H.R. Heip. 2001. Doalternate stable states occur in natural ecosystems? Evidence from atidal flat. Ecology 82: 3449–3461.

Van de Wal, D., P.M. Herman, R. Forster, T. Ysebaert, F. Rossi, E.Knaeps, Y. Plancke, and S. Ides. 2008. Distribution and dynamicsof intertidal macrobenthos predicted from remote sensing: responseto microphytobenthos and environment. Marine Ecology ProgressSeries 367: 57–72.

Volkenborn, N., C. Meile, L. Polerecky, C.A. Pilditch, A. Norkko, J.Norkko, J.E. Hewitt, S.F. Thrush, D.S. Wethey, and S.A. Woodin.2012. Intermittent bioirrigation and oxygen dynamics in permeablesediments: an experimental and modeling study of three tellinidbivalves. Journal of Marine Research 70: 794–823.

Weerman, E.J., J. Van de Koppel, M.B. Eppinga, F. Montserrat, Q.X. Liu,and P.M. Herman. 2010. Spatial self-organization on intertidal mud-flats through biophysical stress divergence. The AmericanNaturalist 176: E15–E32.

Weerman, E.J., P.M. Herman, and J. Van de Koppel. 2011. Top-downcontrol inhibits spatial self-organization of a patterned landscape.Ecology 92: 487–495.

Widdows, J., A. Blauw, C.H.R. Heip, P.M. Herman, C.H. Lucas, J.J.Middelburg, S. Schmidt, M.D. Brinsley, F. Twisk, and H. Verbeek.2004. Role of physical and biological processes in sediment dynam-ics of a tidal flat in Westerschelde Estuary, SW Netherlands.MarineEcology Progress Series 274: 41–56.

Woodin, S.A., D.S. Wethey, J.E. Hewitt, and S.F. Thrush. 2012. Smallscale terrestrial clay deposits on intertidal sandflats: behavioralchanges and productivity reduction. Journal of ExperimentalMarine Biology and Ecology 413: 184–191.

Estuaries and Coasts