integrating scales of seagrass monitoring to meet conservation

1 23

Estuaries and CoastsJournal of the Coastal and EstuarineResearch Federation ISSN 1559-2723Volume 35Number 1 Estuaries and Coasts (2012) 35:23-46DOI 10.1007/s12237-011-9410-x

Integrating Scales of Seagrass Monitoring toMeet Conservation Needs

Hilary A. Neckles, Blaine S. Kopp,Bradley J. Peterson & Penelope S. Pooler

1 23

Your article is protected by copyright and all

rights are held exclusively by Coastal and

Estuarine Research Federation (outside the

USA). This e-offprint is for personal use only

and shall not be self-archived in electronic

repositories. If you wish to self-archive your

work, please use the accepted author’s

version for posting to your own website or

your institution’s repository. You may further

deposit the accepted author’s version on

a funder’s repository at a funder’s request,

provided it is not made publicly available until

12 months after publication.

Integrating Scales of Seagrass Monitoringto Meet Conservation Needs

Hilary A. Neckles & Blaine S. Kopp &

Bradley J. Peterson & Penelope S. Pooler

Received: 29 November 2010 /Revised: 11 March 2011 /Accepted: 25 April 2011 /Published online: 10 May 2011# Coastal and Estuarine Research Federation (outside the USA) 2011

Abstract We evaluated a hierarchical framework forseagrass monitoring in two estuaries in the northeasternUSA: Little Pleasant Bay, Massachusetts, and Great SouthBay/Moriches Bay, New York. This approach includesthree tiers of monitoring that are integrated across spatialscales and sampling intensities. We identified monitoringattributes for determining attainment of conservationobjectives to protect seagrass ecosystems from estuarinenutrient enrichment. Existing mapping programs providedlarge-scale information on seagrass distribution and bedsizes (tier 1 monitoring). We supplemented this with bay-wide, quadrat-based assessments of seagrass percent coverand canopy height at permanent sampling stations follow-ing a spatially distributed random design (tier 2 monitor-ing). Resampling simulations showed that four observationsper station were sufficient to minimize bias in estimatingmean percent cover on a bay-wide scale, and sample sizes

of 55 stations in a 624-ha system and 198 stations in a9,220-ha system were sufficient to detect absolute temporalincreases in seagrass abundance from 25% to 49% coverand from 4% to 12% cover, respectively. We made high-resolution measurements of seagrass condition (percentcover, canopy height, total and reproductive shoot density,biomass, and seagrass depth limit) at a representative indexsite in each system (tier 3 monitoring). Tier 3 data helpedexplain system-wide changes. Our results suggest tieredmonitoring as an efficient and feasible way to detect andpredict changes in seagrass systems relative to multi-scaleconservation objectives.

Keywords Seagrass .Monitoring .Multi-scale . Eelgrass .

Measurable attributes . Sampling design

Introduction

Seagrass meadows are noted for their extremely highproductivity and contribution of valuable ecosystem func-tions and services to the coastal zone (Hemminga andDuarte 2000). Worldwide seagrass declines have beencaused by anthropogenic impacts, including direct physicaldisturbance and indirect effects of watershed developmentand consequent water quality degradation (Orth et al.2006a; Waycott et al. 2009). It is currently estimated thatat least 29% of the global seagrass habitat has disappearedsince the late 1800s, and rates of decline and the total areaof seagrass lost have risen dramatically in recent decades:the median rate of seagrass loss worldwide increasedfrom <2% per year before 1960 to 7.7% per year in the1990s (Waycott et al. 2009). Recognition of theseextensive losses has led to a concomitant proliferation ofseagrass monitoring efforts around the world (Duarte et al.

H. A. Neckles (*) :B. S. KoppUS Geological Survey, Patuxent Wildlife Research Center,196 Whitten Road,Augusta, ME 04330 USAe-mail: [email protected]

B. J. PetersonSchool of Marine and Atmospheric Sciences,Stony Brook University,Stony Brook, NY 11794 USA

P. S. PoolerNational Park Service, Northeast Coastal and Barrier Network,University of Rhode Island Coastal Institute in Kingston,1 Greenhouse Rd., Rm 105,Kingston, RI 02881 USA

Present Address:B. S. KoppKimball Union Academy,P.O. Box 188, Meriden, NH 03770 USA

Estuaries and Coasts (2012) 35:23–46DOI 10.1007/s12237-011-9410-x

Author's personal copy

2004). Paradoxically, however, seagrass monitoring pro-grams often fail to detect declines until substantial losseshave occurred, at which point management interventionsmay be ineffective or too costly to implement (Duarte2002). Therefore, the need for improved monitoringapproaches is paramount to seagrass conservation.

Given increasing demands for seagrass conservation andlimited resources for monitoring, monitoring programsmust be as efficient as possible. Several characteristics ofenvironmental monitoring are associated with improvedefficiency and effectiveness. First, monitoring programsshould be guided by models of system responses to naturalforces and management actions, with monitoring effortstargeting the precise information needed to make conser-vation decisions (Nichols and Williams 2006). The resultsof such targeted monitoring will help discriminate amonghypotheses about stresses on ecosystem integrity, evaluatethe effectiveness of management remedies, and identifyfuture management needs (Nichols and Williams 2006).Second, monitoring should be capable of detecting changein the resource with a high degree of power (Urquhart et al.1998; Elzinga et al. 2001). At a minimum, monitoring mustdetect resource declines before losses are irreversible;preferably, monitoring results should predict changes thatcan be prevented by management actions (Dale and Beyeler2001). Finally, the monitoring design must be feasible toexecute in terms of methods of data collection, samplinglogistics, and monetary costs (Dale and Beyeler 2001;Kurtz et al. 2001).

The core of efficient and effective environmentalmonitoring is a set of clear conservation objectives andthe attributes measured to evaluate whether these objectivesare achieved (Dale and Beyeler 2001; Keeney and Gregory2005). Seagrass conservation goals may be broadly definedin terms of protecting ecosystem integrity, biodiversity, orcritical habitat (ASMFC 1997; Pulich 1999; Borum et al.2004; Kenworthy et al. 2006), or they may focus onspecific seagrass restoration targets (Short et al. 2000;Virnstein and Morris 1996; Orth et al. 2002a), protectivewater quality criteria (Dennison et al. 1993; Johansson andGreening 2000; Wazniak et al. 2007), or minimizing certainanthropogenic impacts (e.g., dredging: Long et al. 1996;Sheridan 2004; recreational boating: Sargent et al. 1995;Fletcher et al. 2009; commercial fishing: Stephan et al.2000; Orth et al. 2002b). Explicitly or implicitly, seagrassconservation and management objectives frequently spanspatial scales, such as maintaining seagrass areal extent,shoot density, and primary productivity (Virnstein 2000). Amajor challenge associated with seagrass monitoring thuslies in identifying a suite of monitoring attributes thattogether are sufficient for evaluating progress towardachieving conservation goals at multiple scales yet arefeasible to measure.

Existing seagrass monitoring programs around the worldaddress questions at different scales. Periodic mappingbased on remotely sensed data is used widely to provideinformation on the location, sizes, and structure of seagrassbeds over broad geographic areas (e.g., Short and Burdick1996; Kendrick et al. 2000; Kurz et al. 2000; Handley et al.2007; Orth et al. 2010), often adhering to standardprocedures for the acquisition and interpretation of aerialphotographs (Finkbeiner et al. 2001). Higher resolutionmonitoring of seagrass condition has been implemented atthe scale of entire bays and estuaries (e.g., Durako et al.2002; Fourqurean et al. 2002; Dowty et al. 2005; Virnsteinand Morris 1996; Avery 2000; Corbett et al. 2005) andindividual seagrass beds (e.g., Neverauskas 1987; Deis2000; McKenzie et al. 2003; Morris and Virnstein 2004;Short et al. 2006a). Although survey methods depend onprogram goals and site characteristics, most seagrassmonitoring at this resolution incorporates quadrat- ortransect-based measurements using direct observations orunderwater videography, with sampling locations guided byprobability designs to allow valid inferences regarding theentire area of interest. Monitoring such characteristics aspercent ground cover, shoot density, and areal biomass atthis level provides information relevant to population-scalemanagement objectives.

A hierarchical approach to monitoring can accommodateconservation objectives at a range of scales while maxi-mizing efficiency and forecasting capabilities. A three-tierframework has been proposed for coordinated monitoringof environmental resources at multiple spatiotemporalscales (NSTC 1997; Bricker and Ruggiero 1998). Con-ceived originally to facilitate integrated monitoring andassessments on a national scale, the framework is equallyadaptable to individual ecosystems. Conceptually, the threetiers of monitoring are represented by layers of a triangle,with the spatial extent of monitoring decreasing from thebase to the vertex. Tier 1 monitoring, at the base of thetriangle, characterizes a few ecosystem properties simulta-neously over very large spatial scales, typically usingairborne or satellite remote sensing methods. Monitoringat this scale is useful to quantify resource extent anddistribution over large regions. Tier 2 monitoring addressesspecific environmental issues or ecosystem properties at ahigher resolution over large geographic areas, generallyusing ground-based approaches. Monitoring of a limitednumber of characteristics can occur at a large number ofsites. Data collected at this tier can be used to quantifystressor/response relationships and produce estimates of theecological condition of resources over broad areas, or thequality of the system as a function of physical, chemical,or biological parameters (cf. Brooks et al. 2004).However, tier 2 data are generally insufficient fordeveloping predictive capabilities. Tier 3 monitoring, at

24 Estuaries and Coasts (2012) 35:23–46


the vertex of the triangle, addresses a greater number ofproperties at a much smaller number of locations or indexsites. Intensive monitoring of drivers of change, ecosystemresponses, and ecological processes at tier 3 focuses ondiagnosing causal relationships. Critical to this frameworkis the integration across tiers of monitoring through nestedsite selection, nested attribute sets, modeling, and re-search. In this way, changes in resource extent orcondition detected at tiers 1 and 2 can drive process-based investigations at tier 3, and statistical and explan-atory models built on tier 3 and tier 2 data can be used tointerpret and predict resource patterns and condition atlarger scales (Bricker and Ruggiero 1998).

Hierarchical monitoring offers an efficient means ofdetecting and predicting change in seagrass ecosystemswithin the context of multi-scale conservation goals. Recentadvances in monitoring and assessment of inland wetlandshave focused on a hierarchy of methods varying in scale ofapplication and intensity of effort (Brooks et al. 2004;Kentula 2007; Reiss and Brown 2007). In contrast,although the need for multi-scale approaches to seagrassmonitoring has been promoted (Lee Long et al. 1996;Virnstein 2000) and individual seagrass monitoring effortshave spanned spatial scales (e.g., Lee Long et al. 1996;Pergent-Martini and Pergent 1996; Robbins and Bell 2000;Fourqurean et al. 2001; McDonald et al. 2006; Foden2007), tiered approaches to seagrass monitoring have notbeen formalized. Within the hierarchical framework describedby Bricker and Ruggiero (1998), monitoring at tiers 1 and 2can provide information on the extent, distribution, andsystem-wide condition of seagrass resources. Tier 3 moni-toring at sites nested within tiers 1 and 2 can provide detailedinformation on submeter- and shoot-scale processes andresponses to a range of potential stressors. This intensivemonitoring is needed to understand the mechanisms andcauses of change detected at larger scales. For example,through combined scales of monitoring, Fourqurean et al.(2001) were able to link seagrass distribution with seasonalpatterns of water temperature and extrapolate seagrassstanding crop to a broad region.

We tested a hierarchical approach to monitoring as alogistically and economically feasible means of evaluatingprogress toward seagrass conservation objectives withintwo lagoonal estuaries in the northeastern USA—PleasantBay, Massachusetts, and the Long Island South ShoreEstuary, New York—that differ considerably in size andcomplexity. The focus of our study was seagrass bedsassociated with National Park units in these systems, CapeCod National Seashore in Massachusetts and Fire IslandNational Seashore in New York. The densest humanpopulation in the USA is found in this northeastern region,where the most pervasive threat to estuarine resources isreduced water quality stemming from watershed develop-

ment (Roman et al. 2000; Short and Short 2003). Much ofthe watershed area of National Park coastal ecosystems liesoutside protective park boundaries and is subject to intensepressures from residential, agricultural, and urban expan-sion. Land clearing, fertilizer production and application,discharge of sewage and septic systems, and fossil fuelcombustion associated with human population growth haveaccelerated nitrogen and phosphorus loading to the world’scoastal zone since the 1950s (Nixon 1995; Cloern 2001).Consequently, the fundamental conservation goal forNational Park estuaries in this densely populated region isto protect estuarine resources from excessive nutrient loadsresulting from watershed development (Kopp and Neckles2009). Given worldwide evidence relating estuarine nutri-ent enrichment to seagrass declines (Duarte 1995; Borum1996), this broad conservation goal includes specificconservation objectives for seagrasses: (1) to maximizeseagrass distribution and (2) to maximize seagrass growthand production. These seagrass conservation objectives areexplicitly multi-scale in nature.

Seagrass mapping is conducted in both study estuariesby state programs at approximately 5-year intervals(MassDEP 2008; NYS Seagrass Task Force 2009). Theseexisting efforts provide tier 1 monitoring information onlarge-scale changes in seagrass distribution. Our aim was todevelop an approach for integrating monitoring of seagrasscondition throughout entire systems (tier 2) and at high-resolution index sites (tier 3) in order to improve abilitiesfor evaluating and anticipating threats to seagrass resourcesassociated with estuarine nutrient enrichment. Importantly,to ensure long-term implementation, the approach must beexecutable within the limited staffing and funding capacityof park management; thus, our focus was on developing anapproach that was both scientifically robust and economi-cally practicable.

The primary mechanism linking increased nutrientloading to seagrass declines is attenuation of light throughthe water column and at leaf surfaces by fast-growingphytoplankton, epiphytic microalgae, and ephemeral mac-roalgae (Sand-Jensen and Borum 1991; Duarte 1995). Thedepth limit of seagrass meadows is inversely related towater column nitrogen concentration (Sand-Jensen andBorum 1991; Borum 1996) and light attenuation (Duarte1991; Dennison et al. 1993), and seagrass responses toaltered light availability associated with reduced waterquality will be most evident at the deepest locations ofexisting beds (Dennison and Alberte 1982). Therefore, weexpected that high-resolution monitoring of seagrass attrib-utes along a depth gradient (tier 3) would be helpful inelucidating monitoring data on seagrass condition through-out entire bays (tier 2) and forecasting system-widechanges. Given the prevalence of water quality degradationas a primary threat to global seagrass survival (Short and

Estuaries and Coasts (2012) 35:23–46 25


Wyllie-Echeverria 1996; Waycott et al. 2009), this approachto seagrass monitoring should be broadly applicable.

Study Location

Pleasant Bay, Massachusetts

The Pleasant Bay estuary is the largest coastal embaymenton Cape Cod, Massachusetts (Fig. 1). The estuary receivesfreshwater from northern and western shore tributaries anddirect groundwater discharge. A narrow barrier beach formsthe eastern shore of Pleasant Bay, and the estuary isconnected to the Atlantic Ocean through two tidal inlets atits southern end (Fig. 1). The northernmost inlet formed inApril 2007, during the course of our investigation, and hashad substantial impact on the hydrodynamics in the estuary.The Pleasant Bay system encompasses approximately2800 ha with an average depth of 1.8 m below mean low

water (MLW) and a mean tidal range of approximately1.5 m. The estuarine system is characterized by broadintertidal and subtidal flats, a narrow channel, and deeperbasins in the lower bay and some tributary ponds.Sediments are primarily medium to very fine sand (Psutyand Silveira 2009).

The greatest extent of seagrass in the Pleasant Bayestuary is found in Little Pleasant Bay (LPB), whichoccupies the estuary’s shallow upper basin (Fig. 1).Eelgrass (Zostera marina L.) is the only seagrass in thissystem. The eastern half of the eelgrass meadow fallswithin the boundary of Cape Cod National Seashore and isof primary management interest. The main sources ofnitrogen loading to the Pleasant Bay estuary are wastewater,fertilizer, runoff from impervious surfaces, and direct atmo-spheric deposition (Howes et al. 2006). Although nitrogenloading rates are lower than those of other estuaries in thenortheastern USA (Carmichael et al. 2004), they haveincreased substantially over the past few decades due to

Fig. 1 Location of study sitesin Little Pleasant Bay,Massachusetts, showing tier 1data on seagrass coverage in2001 (MassDEP 2008), the tier2 sampling frame, and the tier 3monitoring location (star).“Zones” refer to distance zonesused in the tier 2 samplingdesign



changing land use patterns in the watershed (Howes et al.2006). Eelgrass coverage in the Pleasant Bay systemdeclined by 24% from the 1950s to the present (Howes etal. 2006). Retraction of vegetation from the upper reaches oftributaries during this time period is consistent withconcurrent patterns of nutrient enrichment in the system(Howes et al. 2006). Our sampling frame for tier 2monitoring of seagrass condition included all of LPB andthe upper tributaries and was 624 ha in size. In 2001, therewere 301 ha of eelgrass within the boundaries of oursampling frame (MassDEP 2008). We established a tier 3monitoring site in the eastern portion of LPB off the shore ofHog Island in an eelgrass bed that was easily accessible; thatspanned the depth range of eelgrass in LPB; and was typicalfor the system in terms of eelgrass density, morphology, andsubstrate characteristics (Fig. 1).

Long Island South Shore Estuary, New York

The Long Island South Shore Estuary is a series ofinterconnected bays bounded by the southern coast ofLong Island, New York, and outer barrier islands. Ourstudy was within the Great South Bay (GSB) and MorichesBay (MB) portion of South Shore Estuary bordered by FireIsland, from Fire Island Inlet on the west to Moriches Inleton the east (Fig. 2). The GSB/MB system stretchesapproximately 50 km between the two inlets. Freshwaterinput is primarily from Long Island tributaries andgroundwater seepage. The average depth of GSB is 1.3 m(MLW) and of MB is 0.9 m (Wilson et al. 1991); both

basins have tidal channels associated with the inlets, butmost of the area is <1 m deep (MLW). The tidal range at theFire Island Inlet is approximately 1.25 m, but it decreases to0.45 m at a point equidistant from Fire Island Inlet andMoriches Bay (Wong 1981), and sediments are primarilymedium to very fine sand (Psuty and Silveira 2009).

The boundary of Fire Island National Seashore extendsfrom the bayside shoreline of the island about 1.2 km intothe estuary. The majority of the seagrass in GSB/MB isfound within the park boundary (Fig. 2) and includes botheelgrass and widgeon grass (Ruppia maritima). The mostsignificant sources of nitrogen to the system are runoff fromimpervious surfaces and wastewater (NYS DOS 2001).There is a gradient of decreasing water quality withdistance from the inlets (Weiss et al. 2007). Eelgrass inGSB declined in coverage by 23% from 1967 to 1977,presumably in response to reduced light availabilityassociated with eutrophication (Dennison et al. 1989).Although changing land use practices have resulted indeclining water column nitrogen concentrations since themid-1970s (Gobler et al. 2005; Hinga 2005), harmfulbrown tides (Aureococcus anophagefferens) have recurredsince 1985 (Gobler et al. 2005). Dennison et al. (1989)estimated that the shading effect of brown tide blooms from1985 to 1988 caused an additional 40% reduction ineelgrass coverage in GSB from pre-bloom conditions. Oursampling frame for tier 2 monitoring extended 1 km outsidethe seashore boundary (Fig. 2) and was 9,220 ha in size. In2002 there were 2,760 ha of seagrasses within the limits ofour sampling frame (NYS Seagrass Task Force 2009;

Fig. 2 Location of study sites inGreat South Bay and MorichesBay, New York, showing tier 1data on seagrass coverage in2002 (NYS Seagrass Task Force2009), the tier 2 sampling frame,and the tier 3 monitoringlocation (star). “Zones” refer todistance zones used in the tier 2sampling design



Fig. 2). We established a tier 3 monitoring station in easternGSB off of Fire Island, in the portion of the bay with themost continuous mapped seagrass distribution (Fig. 2). Asin LPB, the exact location of the tier 3 station was selectedto be accessible for monitoring and representative of theseagrass beds throughout GSB/MB in terms of depth rangeand plant and substrate characteristics.

Materials and Methods

Measurable Attributes

To ensure that monitoring programs target the informationneeded to make conservation decisions, monitoring attrib-utes must be linked explicitly to conservation objectives(Nichols and Williams 2006). We identified attributes of theseagrass ecosystem for evaluating the attainment ofconservation objectives associated with seagrass distribu-tion and seagrass condition (Table 1). The attributesspanned scales of monitoring and incorporated indicatorsof ecological structure, composition, and function (cf. Daleand Beyeler 2001; Evans and Short 2005) that are sensitiveto light limitation and thus relevant to managementconcerns. For example, the correlation between increasednutrient loading and declines in seagrass extent (Sand-Jensen and Borum 1991; Duarte 1995; Latimer and Rego2010) and population parameters such as shoot density,percent cover, and biomass (Neckles et al. 1993; Short et al.1995; Taylor et al. 1995; Nixon et al. 2001; Hauxwell et al.2003) are well documented. Collectively, the variables weselected are among the most commonly measured charac-teristics in seagrass monitoring programs (Duarte et al.2004; Krause-Jensen et al. 2004).

Information on seagrass bed size is already gathered byexisting tier 1 mapping programs based on aerial photo-

graphs acquired at a scale of 1:20,000; our study focused ontesting attributes at tiers 2 and 3. The time needed forseagrass sampling depends greatly on whether in situobservations and plant harvesting are required, withsampling time reduced by avoiding need for observationsby SCUBA and for processing large quantities of plantmaterial (cf. Jackson and Nemeth 2007). Therefore, tominimize sampling time and associated costs, attributesmonitored at tier 2 were limited to variables that could bemeasured either from a boat or by using snorkel. Methodsfor sampling at tier 2 and tier 3 scales of monitoring aredescribed separately below.

Tier 2 Monitoring

Sampling Design

We used a restricted random sampling design (Elzinga et al.2001) to select sampling stations in LPB and GSB/MB. Ingeneral, restricted random designs are characterized by firstdividing the area to be monitored into equal sizedcontiguous segments and then randomly positioning asingle sampling location within each segment. This ap-proach ensures good dispersion of sampling points throughthe entire sampled population while incorporating random-ization in sample placement. We used a randomized-tessellation stratified design (Stevens 1997) in which a gridof tessellated hexagons serves as the basis for locating randomsampling stations. An alternative method for maintaining aspatially well-balanced random sample is the generalizedrandom-tessellation stratified design (Stevens and Olsen2004), which has more flexibility for incorporating stratawith different sampling intensities.

The LPB and GSB/MB sampling frames were parti-tioned into tessellated hexagonal cells (Figs. 3 and 4) and arandom sampling point (set of latitude and longitudecoordinates) was selected within each cell. The hexagonalcells in LPB were each 500 m wide and the cells in GSB/MB 925 m wide. A very small fraction of the original cellscould not be monitored because they fell entirely on land.The sample sizes were thus defined by the number ofattainable cells in each system: 55 in LPB and 198 in GSB/MB. Sampling points were categorized within equal-widthdistance zones in each bay that represented varyingdistances from flushing sources. Distance zones in LPB(n=3) were 1.5-km bands of decreasing distance from themouth of the bay (Fig. 1) and in GSB/MB (n=4) were 11.5-km bands along the axis of Fire Island between Fire IslandInlet to the west and Moriches Inlet to the east (Fig. 2).Sampling points were permanent to maximize the ability todetect changes in condition over time (Elzinga et al. 2001).

We sampled tier 2 attributes (percent cover, canopyheight; Table 1) in midsummer at the time of maximum

Table 1 Seagrass attributes measured at each tier of monitoring

Conservation objective Measurableattribute

Tier1

Tier2

Tier3

Maximize seagrassareal distribution

Bed size x

Depth limit x

Maximize seagrassgrowth and reproduction

Percent cover x x

Canopy height x x

Total shoot density x

Reproductive shoot density x

Biomass x

Tier 1 monitoring is provided by existing seagrass mapping programs;tier 2 and tier 3 attributes were tested in this study



eelgrass leaf biomass in the region (Wilson and Brenowitz1966; Roman and Able 1988). Sampling was conductedduring three consecutive years in LPB (2006–2008) and2 years in GSB/MB (2007, 2009). During sampling, wenavigated to sampling points using GPS with ≤4-mpositional accuracy. A sampling station was defined as a10-m diameter circle around the GPS point, determined bythe length of the boat plus the potential error in reaching theprecise coordinates. At each station, we collected foursubsample observations within quadrats located haphazard-ly off the four quarters of the boat. In addition, during the2007 only sampling at LPB, the number of subsamples atsix randomly selected test stations was increased from 4 to12. The minimum distance between the 12 subsamples ateach test station was 3 m. These data were used for

simulations to test the subsampling intensity necessary tocharacterize the 10-m diameter sampling station.

Field Measurements

We measured tier 2 attributes once per subsample siteduring each sampling interval. We estimated seagrasspercent cover by species within a 0.25-m2 quadrat byvertical observation. Cover measurements in LPB weremade exclusively from the boat using a variety of methods,depending on water depth and surface conditions; at a givenstation, we used the easiest method that allowed a clearview of the substrate. At shallow sites (approx. <60 cmdeep), the substrate was clearly visible directly through thewater, and at mid-depth sites (approx. 60 cm–1.25 m), weused a view scope to penetrate the water surface; in theseinstances, the sampling quadrat was a stainless steel frameattached to a rope. At deeper sites (maximum of about 6 m),we viewed the substrate using an underwater video cameramounted directly above an aluminum sampling frame; theframe was lowered to the substrate on a rope. In GSB/MB,we used the same boat-based methods where visibilityallowed, but low water clarity demanded snorkeling foraccurate measurements at about 40% of the stations. Coverestimates were standardized based on the photographicreference guide published in Short et al. (2006b) and inter-observer training and calibration (Krause-Jensen et al.2004). Cover from 5% to 100% was recorded in 5%increments, and the presence of very few plants (<5%cover) was assigned a value of 1% cover. Canopy height(defined as the height of 80% of the leaf material of thedominant species, ignoring the tallest 20% of the leaves;Duarte and Kirkman 2001) was measured on samples ofthree large shoots collected with a rake on a telescopinghandle (LPB) or by hand while snorkeling (majority ofstations at GSB/MB). Water depth and time were recordedat each station and corrected to MLW using predicted tideheights.

Fig. 3 Grid of tessellated hexagons used as the basis for locatingsampling stations in Little Pleasant Bay. One random point wasselected as a sampling station within each hexagonal cell

Fig. 4 Grid of tessellatedhexagons used as the basis forlocating sampling stations inGreat South Bay and MorichesBay. One random point wasselected as a sampling stationwithin each hexagonal cell



Statistical Analysis of Status and Trends

We used the inverse distance-weighted (IDW) method ofspatial interpolation to reveal bay-wide patterns in seagrassstatus from measurements of percent cover (Watson andPhilip 1985). This is a simple method of deterministicinterpolation that estimates the value of unsampled pointsas the weighted average of values from a given number ofsurrounding points, giving more weight to closest points.IDW assumes a functional form for the weights based onthe inverse of the Euclidean distance, but does not assumedata characteristics that are required for probabilisticinterpolation methods (e.g., kriging) such as a regionallyconstant mean and a specified spatial covariance structure(Davis 2002). Given the highly irregular shoreline of LPB,we used an impermeable boundary during IDW processingto control inappropriate influences of data points acrossseparate branches of the system. The impermeable bound-ary was created as a 100-m buffer around the shoreline,with manual adjustments to further isolate kettle hole saltponds. Interpolated values were categorized within Braun-Blanquet cover classes for display (Elzinga et al. 2001),modified by combining all values ≤5% cover into a singleclass. Spatial interpolation was done using Spatial Analystin ArcGIS 9.1.

Spatial interpolation of cover data yields snapshots ofseagrass status, but does not provide statistical comparisonsof condition between years. Therefore, we used generallinear models to make quantitative comparisons over time.We used a repeated measures analysis of variance to test thenull hypothesis that percent cover did not change over timeand that patterns of response did not differ among distancezones, treating year and distance from flushing (distancezone) as fixed effects, and treating station location as arandom effect (Green 1993; Schabenberger and Pierce2002). Depth was included as a fixed effect in LPB (lessthan or greater than 1 m MLW), but was not included in theGSB/MB model due to the fairly homogeneous depths inthat system. We used various methods to ensure that thedata met the assumptions of the analysis. We examined thepotential effect of spatial autocorrelation by developingvariograms for within-year data sets and evaluating the fitof candidate spatial covariance structures to the data(Cressie 1993). We then compared the likelihood ofwithin-year models specifying these spatial covariancestructures (exponential, Gaussian, linear, spherical, andpower) to that of a null model (i.e., with no spatialcovariance structure; Cressie 1993; Littell et al. 2006). Ineach case, the inclusion of a spatial covariance structurefailed to improve the fit of the model. We used Mauchley’scriterion to verify the compound symmetry of the covari-ance structure among repeated measures (Potvin et al.1990) and residual analysis to confirm the overall aptness

of the model (Kutner et al. 2005). Variograms weredeveloped using the VARIOGRAM procedure, and dataanalysis models were developed using the MIXED proce-dure of the SAS statistical software package (SAS 9.2).Sample means were reported to the nearest 5% cover tomatch the degree of precision of the original data.

Sample Size Simulations

Using percent cover data collected in 2007 from the sixLPB test stations with 12 subsamples each, we performedresampling simulations to determine the within-stationsampling intensity (i.e., the number of subsample observa-tions) necessary to achieve high accuracy in estimating thestation mean. For each of the six test stations, we drew setsof 10,000 random samples of different sizes (n=3–11subsamples) from the pool of 12 subsamples. We usedthese simulated samples of different sizes to examine theabsolute bias associated with estimates of mean percentcover within a test station, or the difference between thesample mean and the true test station mean. The probabilityof achieving certain levels of bias at individual test stationsfor each within-station number of subsamples (n=3–11)was determined from the frequency distribution of themeans of 10,000 samples of size n around the true teststation mean (n=12). The probability of achieving certainlevels of overall bias for each within-station number ofsubsamples was determined from the frequency distributionof 10,000 overall means (where the overall mean = themean of six station means derived from n=3–11 subsam-ples) around the grand mean (n=72, based on 12 subsampleobservations at each of six test stations).

We examined the influence of the bay-wide sample sizesin LPB and GSB/MB on detecting change in percent coverby simulating sample sets using increased hexagon cellsize, resulting in corresponding decreases in the number ofstations sampled throughout the bay. We simulatedincreases of two, three, and four times the original hexagonsize by combining adjacent hexagons within distance zonesinto larger areas and then randomly selecting one of theoriginal groups of subsample observations (i.e., foursubsamples from a single station) to represent the datafrom the simulated larger hexagon. We created multiplesimulated data sets for each “new” hexagon size by usingdifferent groups of subsamples. For example, for hexagonstwice the original size, we randomly divided the set oforiginal subsample groups into two simulated data sets,each using one half of the sample stations. Similarly, wecreated three simulated data sets for hexagons three timesthe original size and four simulated data sets for hexagonsfour times the original size. Simulated data sets weregenerated for LPB percent cover data in 2006 and 2007 andGSB/MB in 2007 and 2009. We then applied the same



repeated measures analysis of variance model to thesimulated data sets that was applied to the full data setwithin each system and compared the results generatedusing the decreased sample sizes to those generated usingthe original sample size.

Tier 3 Monitoring

Sampling Design

We established intensive monitoring sites in areas of fairlyhomogeneous seagrass cover that were broadly representa-tive of conditions in LPB and GSB/MB in terms of seagrassspecies composition, depth gradient, and sediment charac-teristics. For tier 3 monitoring, we followed the spatialdesign of the global seagrass monitoring program Sea-grassNet (Short et al. 2006b), although the attributesreported here are a subset of those included in theSeagrassNet protocol. By adopting the SeagrassNet design,our tier 3 data form part of a global monitoring network andcan contribute to worldwide assessments of seagrasschange. As prescribed by SeagrassNet, each tier 3 siteconsisted of three permanent 50-m transects establishedparallel to the shoreline: one transect was near the shallowedge of the seagrass bed, one at a moderate depth, and oneat a deep location of the bed (Table 2). We embeddedpermanent helical screw anchors in the substrate to markthe ends and the middle of each transect (0-, 25-, and 50-mfixed reference points). Twelve permanent 0.25-m2 quadratswere positioned at randomly selected distances along eachtransect. Sampling was conducted in midsummer duringfive consecutive years in LPB (2005–2009) and during2 years in GSB/MB (2007, 2009). At each samplinginterval, the precise quadrat locations were determined bystretching temporary fiber glass measuring tapes betweenthe fixed transect markers.

Field Measurements

We tested all tier 3 monitoring attributes (Table 1) at LPBand all but biomass at GSB/MB. We estimated percent cover,measured canopy height, and counted the number ofreproductive shoots by seagrass species within the entire

0.25-m2 quadrats by snorkeling (shallow transects) orSCUBA (deep transects). We determined eelgrass shootdensity by direct counts of all shoots rooted within the entirequadrat if percent cover was ≤25% or shoot distribution washighly clumped; if percent cover was >25% and shoots werehomogeneously distributed, all shoots within a 0.0625-m2

subquadrat were counted. Low visibility in the mixed-species seagrass bed at GSB/MB prohibited accuratecounts of the slender, branched widgeon grass stems, sodensity measurements at GSB/MB were confined toeelgrass.

At LPB, we used an indirect method to determineaboveground and belowground biomass of eelgrass in thepermanent quadrats. Biomass samples were harvested usinga 0.0035 m2 (6.7-cm diameter) core tube, the size specifiedby the SeagrassNet protocol. Samples were collected froman area 0.5–1.0 m away from each permanent quadrat withvegetation of the same percent cover and canopy height asin the quadrat. Samples were rinsed with fresh water andstored under refrigeration for processing within 48 h.During processing, only plants with intact meristems wereretained. Shoots were cleaned of debris, dead plantmaterial, and epiphytes and sorted into living leaf (includ-ing sheaths) and root + rhizome fractions. The number ofplants with intact meristems in the biomass sample wasrecorded and plant material was then dried to constantweight at 60°C and weighed. The average leaf and root +rhizome weight per shoot was determined for each biomasssample and then was applied to shoot density in therespective quadrat to derive the permanent quadrat biomassper unit area (Duarte and Kirkman 2001).

We compared this indirect method of determiningbiomass in the permanent quadrats with direct measurementby collecting paired test samples at LPB in 2006 (12 pairedsamples along the mid-depth transect only) and 2007 (fiveto six paired samples along each of the three transects). Weidentified sampling locations using a 1-m2 frame dividedinto 16 quadrats (0.0625 m2). For each sample pair, weplaced the frame 1 m away from the transect and randomlyselected one 0.0625-m2 test quadrat for sampling. Wecollected a 0.0035-m2 core from the middle of each 0.0625-m2 test quadrat and then collected all remaining plantmaterial in the quadrat by cutting around the inside of thequadrat to below the root zone. Plant material within bothsample fractions (in the central core and remaining in thequadrat) was processed as described previously. Biomass ofthe 0.0625-m2 test quadrat was then determined indirectlyby applying the average weight per shoot from the centralcore sample to the shoot density of the entire test quadrat(the number of shoots in the core + the number of shootsremaining in the quadrat) and directly (the measuredbiomass of shoots in the core + the measured biomass ofshoots remaining in the quadrat).

Table 2 Depth of tier 3 monitoring sites in LPB and GSB/MB

Transect Depth (meters below MLW)

LPB GSB/MB

Shallow 0.64 0.76

Mid-depth 0.84 0.97

Deep 1.35 1.34



We recorded the position of the deep edge of thecontinuous seagrass meadow as the distance seaward fromfixed reference points of the deep transect. We usedunderwater video with GPS overlay to locate the deep edgeof the seagrass bed. The edge of the seagrass bed wasdefined operationally as the farthest point where shootswere ≤1 m apart (Short et al. 2006b). An underwater videocamera was towed behind a boat along offshore trajectoriesperpendicular to the deep transect beginning at the 0-, 25-,and 50-m transect marks. In LPB, the continuous seafloorimage was recorded on a digital videotape stamped withGPS coordinates, depth of the camera, date, time, speed,and heading, and the deep edge of the seagrass bed wasidentified by viewing the video transect images on acomputer monitor. In GSB/MB, the edge of the bed wasidentified on the video monitor in the field and verified bysnorkeling.

Statistical Analysis

Annual quadrat data from PB and GSB/MB were analyzedusing a repeated measures analysis of variance with fixedeffects of year and depth and random effects of quadrat;this model tested the null hypotheses that seagrassattributes did not change over time or depth and thatpatterns of response were similar across the depthgradient. We ensured the appropriateness of the statisticalmodels as described for tier 2 trend data, and data weretransformed where necessary to correct for non-constancyof error variance. The relationship between indirect anddirect methods of biomass determination tested at LPBwas quantified with simple linear regression, and apredictive relationship between seagrass biomass, percentcover, and canopy height within the permanent quadratswas developed with multiple regression analysis. Residualplots were used to verify the aptness of all regressionmodels (Kutner et al. 2005).

Results

Tier 2 Monitoring

Little Pleasant Bay

The tier 2 survey of LPB was completed annually in 2–3 days by two to three people. System-wide monitoring ofeelgrass condition revealed a change in percent coverfollowing formation of the new inlet. In 2006, the uppertributaries were unvegetated or sparsely vegetated, andeelgrass in the lower bay ranged mostly from 6% to 50%cover (Fig. 5). In 2007 and 2008, the upper tributariesremained sparsely vegetated, but eelgrass in lower LPB

ranged mostly from 6% to 75% cover (Fig. 5); thus, afterthe new inlet formed in April 2007, there was a shift indistribution of observations toward higher cover classes.Trend analysis showed a significant increase in meanpercent cover from 2006 to 2007 within the southernmostportion of LPB (Tables 3 and 4, distance zone 3; p<0.0001).There were no changes in eelgrass cover from 2007 to 2008(Table 4; within-zone p ranged from 0.44 to 1.0). Depthcategory had no effect on cover (p=0.52) and was removedfrom the model. Average eelgrass canopy height ranged fromabout 40 cm in areas <1 m deep to 75 cm in deeper areasand was consistent among years.

Great South Bay/Moriches Bay

It required 10–12 days for three people to complete the tier2 survey in GSB/MB. In 2007, eelgrass was densest in thewestern end of GSB near Fire Island inlet, whereaswidgeon grass occurred nearly exclusively in the easternend of GSB (Fig. 6). In 2009, eelgrass had expanded withineastern GSB (Fig. 6), and both species showed a shift tohigher cover classes. The mean percent cover of bothspecies increased significantly within the eastern section ofGSB between years (Tables 5 and 6, distance zone 3;eelgrass, p=0.02; widgeon grass, p<0.0001), and the meanpercent cover of eelgrass also increased significantly withinthe westernmost section of the bay (Tables 5 and 6, distancezone 1; p<0.0001). There were no other changes inseagrass cover between years (Table 6; within-zone pranged from 0.18 to 0.85). Average canopy height ofeelgrass-dominated beds was between about 15 and 25 cmin GSB and was 35 cm in MB, whereas average canopyheight where widgeon grass dominated was <10 cm;patterns were consistent between years.

Sample Size Simulations

The number of subsamples, or observations per station,required to achieve a low bias in estimating mean percentcover at a single station varied across the six LPB teststations. At stations with fairly low within-station variabil-ity, three to four subsamples yielded a high probability(>80%) that the expected value of the estimator for within-station mean percent cover was within 10 percentage pointsof the true mean (true mean percent cover ±10 points),regardless of whether mean percent cover was high (Table 7,test stations 2 and 5) or low (Table 7, test station 3). Atindividual test stations with greater within-station hetero-geneity, six to eight subsamples were required to achieve a

Fig. 5 Percent cover of eelgrass (Z. marina) in Little Pleasant Bay,2006–2008. Seagrass cover was measured at 55 sampling stationsusing tier 2 monitoring methods and was estimated at unsampledlocations using inverse distance weighted spatial interpolation

�



similar absolute bias (Table 7, test stations 1, 4, and 6).However, a relatively low subsampling effort was stillsufficient to achieve a low overall bias in estimating themean percent cover across all test stations: four subsamplesyielded an 80% probability that the expected value of theestimator for overall mean percent cover was within 5percentage points of the true mean (true mean percent cover±5 points) and a >99% probability that the expected valueof the estimator was within 10 percentage points of the truemean (true mean percent cover ±10 points).

The ability to detect temporal changes in seagrass coverwithin entire sampling frames with decreasing numbers ofstations depended on the magnitude of change thatoccurred. For eelgrass in LPB, analysis of the twosimulated data sets including half the original number ofsampling stations (n=26 vs. original n=55) yielded thesame pattern of response as did analysis of the original dataset: percent cover increased significantly from 2006 to 2007within the southernmost distance class (individual simu-lations: p=0.001, p=0.008). The results were inconsistent,however, as the sample size was reduced further. At onethird the original number of sampling stations (n=16), asignificant increase in percent cover in the southernmostdistance class was detected in two out of three simulations(p=0.003, p=0.003, p=0.20), and at one fourth the originalnumber (n=13), this increase was detected in two out offour simulations (p<0.0001, p=0.004, p=0.18, p=0.59). In

Table 4 Mean (SE) percent cover of eelgrass (Z. marina) withindistance zones of Little Pleasant Bay measured during tier 2monitoring, 2006–2008

Distance zone Year

2006 2007 2008

1 0 (0)a 0 (0)a 0 (0)a

2 10 (6)a 10 (6)a 10 (6)a

3 25 (6)a 50 (9)b 45 (8)b

Distance zones are 1.5-km-wide horizontal bands numbered fromnorth to south. Letters refer to comparisons among years withindistance zones; values with like letters within a row are notsignificantly different (p>0.05). Following statistical analysis, samplemeans were rounded to the nearest 5% cover to match the degree ofprecision of the original data

Table 3 Repeated measures analysis of variance of eelgrass (Z.marina) percent cover measured during tier 2 monitoring in LittlePleasant Bay, 2006–2008

Fixed effect df F p

Year 2, 104 3.67 0.029

Distance 2, 52 14.40 <0.001

Year × Distance 4, 104 5.07 0.001



GSB/MB, the ability to detect temporal changes withreduced sample sizes varied between species. For eelgrass,a significant increase in percent cover from 2007 to 2009

was observed in the westernmost section of GSB (distancezone 1) for both simulations using half the original numberof sampling stations (n=98 vs. original n=198; individual

Fig. 6 Percent cover of eelgrass(Z. marina) and widgeon grass(R. maritima) in Great SouthBay and Moriches Bay, 2007and 2009. Seagrass cover wasmeasured at 198 samplingstations using tier 2 monitoringmethods and was estimated atunsampled locations usinginverse distance weightedspatial interpolation



simulations: p=0.0005, p=0.001) and all three using onethird the original number (n=65; p=0.0009, p=0.003,p=0.04). At one fourth the original number of samplingstations (n=48), this increase was detected in three out offour simulations (p=0.0003, p=0.002, p=0.006, p=0.75).In the eastern section of GSB, where the absolute increasein eelgrass cover between years was not as great (Table 6,distance zone 3), the change from 2007 to 2009 wasmarginally to non-significant in simulations using half theoriginal sampling stations (p=0.06, p=0.17). A significantincrease was detected in one out of three simulations usingone third of the original number of sampling stations(p=0.02, p=0.14, p=0.91) and in one out of foursimulations using one fourth the original number (p=0.03,p=0.10, p=0.41, p=0.92). The effect of reducing samplesize on the ability to detect an increase in widgeon grassbetween years in the eastern section of GSB (distance zone3) varied across simulations, but the significance of the testwas always <0.10 (simulations at one half the originalsample size: p=0.0006, p=0.09; at one third the originalsample size: p=0.01, p=0.02, p=0.09; and at one fourth theoriginal sample size: p=0.02, p=0.03, p=0.09, p=0.09).

Tier 3 Monitoring

Little Pleasant Bay

Annual monitoring of permanent quadrats at the LPB tier 3site was completed in 1 day by four people. We observedconsiderable spatial and temporal variability in eelgrasscharacteristics at this scale (Fig. 7 and Table 8). Within theeelgrass bed, canopy height increased consistently from theshallow to the deep transect (p<0.0001; Fig. 7b), but thepatterns of other attributes across the depth gradient variedamong sampling intervals. Therefore, our analysis ofmonitoring results focused primarily on temporal ratherthan spatial differences in tier 3 attributes. Results of testingthe null hypothesis of no temporal changes in eelgrassattributes within transects are expressed at a statisticalthreshold of p=0.05 (Table 8). We also performed all testsusing a critical value of p=0.10 and found the interpreta-tions of the results to be consistent with the moreconservative criterion.

Along the shallow transect, all eelgrass attributesdeclined dramatically between 2005 and 2006 as a resultof a severe storm in early summer 2006, as evidenced bydirect observation of deep piles of uprooted plants on theshoreline immediately following the storm. By 2007, shootdensity had already rebounded significantly through tre-mendous production of lateral shoots, and by 2009 themean shoot density along the shallow transect was 57% ofthe high value recorded in 2005 (Fig. 7a). The meanproportion of reproductive shoots was consistently very low(<0.3% of total density). Percent cover showed a patternsimilar to density, with a significant increase over 2006 by2008 (Fig. 7c and Table 8). Canopy height and totalbiomass along the shallow transect recovered from thestorm impact more slowly; it was not until 2009 that themean values for these attributes increased significantly over2006 (Fig. 7b, d and Table 8). The ratio of aboveground tobelowground biomass was low and stable over time (meanof 2.3±0.2 SE).

Eelgrass attributes along the mid-depth transect alsodeclined in 2006, but the decrease from 2005 was less thanthat observed along the shallow transect and the subsequentrecovery from the storm impact was more rapid (Fig. 7 andTable 8). By 2007, shoot density, canopy height, percentcover, and biomass had already increased significantly from2006 levels, and these increases were generally maintainedthrough 2009 (Fig. 7a–d and Table 8). The mean proportionof reproductive shoots along the mid-depth transect wasalways <0.6% and the aboveground-to-belowground bio-mass ratio was stable over time (mean of 4.3±0.6 SE).

In contrast to eelgrass along the shallow and mid-depthtransects, eelgrass attributes along the deep transect werefairly stable over the course of the study. From 2005 to

Table 6 Mean (SE) percent cover of eelgrass and widgeon grasswithin distance zones of Great South Bay/Moriches Bay measuredduring tier 2 monitoring, 2007 and 2009

Distance zone Eelgrass Widgeon grass

2007 2009 2007 2009

1 15 (3)a 30 (5)b 1 (1)a 1 (1)a

2 1 (1)a 5 (3)a 5 (2)a 5 (2)a

3 5 (2)a 10 (4)b 10 (3)a 20 (4)b

4 5 (3)a 5 (3)a 1 (1)a 0 (0)a

Distance zones are 11.5-km bands along the axis of Fire Islandbetween Fire Island Inlet and Moriches Inlet numbered from west toeast. Letters refer to comparisons between years for each specieswithin distance zones; values of the same species with like letterswithin a row are not significantly different (p>0.05). To match thedegree of precision of the original data, following statistical analysissample means >2.5% cover were rounded to the nearest 5% andpositive means <2.5% were reported as 1

Table 5 Repeated measures analysis of variance of eelgrass andwidgeon grass percent cover measured during tier 2 monitoring inGreat South Bay/Moriches Bay, 2007 and 2009

Fixed effect df Eelgrass Widgeon grass

F p F p

Year 1, 194 21.67 0.000 2.48 0.117

Distance 3, 194 10.93 0.000 15.66 0.000

Year × Distance 3, 194 4.51 0.004 5.97 0.001



2007, there were relatively small annual differences inshoot density, canopy height, and percent cover, but thesedid not translate to any biomass differences (Fig. 7d andTable 8). The mean proportion of reproductive shoots was<0.4% during this interval. Canopy height and percentcover increased substantially in 2008, leading to a similarlylarge increase in biomass (Fig. 7d and Table 8) with asignificantly greater investment in aboveground tissue(mean 2008 aboveground-to-belowground biomass ratioof 15.3±3.4 SE vs. annual means from 3.2 to 9.0 in otheryears). The mean proportion of reproductive shoots alongthe deep transect was 1.5% (±1.0 SE) in 2008 and 8.8%(±2.9 SE) in 2009. Shoot density, canopy height, percentcover, and biomass decreased on the deep transect from2008 to 2009 (Fig. 7a–d and Table 8) so that at the end ofthe study, eelgrass biomass was similar between the shallowand deep transects (p=0.55; Fig. 7d).

Tier 3 monitoring revealed a shift in the location of theouter edge of the eelgrass bed over time (Table 9). In 2004and 2005, along all three offshore trajectories, the eelgrassbed was limited by the central channel in LPB. Distance-to-edge data were not acquired in 2006. In 2007, the eelgrassalong the trajectories beginning at the 0- and 25-m transectmarkers extended continuously through the channel to theopposite shore of LPB, resulting in a near doubling of thedistance from the fixed reference points to the edge of theseagrass bed. Along the trajectory beginning at the 50-mtransect marker, however, the eelgrass bed remained limitedby the channel, which was slightly deeper in that location.This pattern persisted through 2009.

Great South Bay/Moriches Bay

Monitoring of the permanent quadrats at the GSB/MB tier 3site was completed in 1 day by four people. An arcsinetransformation was applied to percent cover data to stabilizesample variances. The shallow and mid-depth transectswere dominated by widgeon grass (Fig. 8); vegetated

quadrats contained either widgeon grass exclusively or amixed community with a small proportion of eelgrass.Percent cover and density of eelgrass were very low alongthese transects, with little to no change from 2007 to 2009(shallow transect: cover, p=0.05; density, p=0.22; mid-depth transect: cover, p=0.38; density, p=0.65). During thisinterval, widgeon grass cover increased substantially alongthe shallow transect (p<0.0001) to a mean of about 90%cover and was stable along the mid-depth transect (p>0.27). Widgeon grass reproductive shoot density decreasedfrom a mean of 35 shoots/m2 (±10 SE) in 2007 to a mean of17 shoots/m2 (±3 SE) in 2009 (p=0.008). There was anaverage of nine widgeon grass reproductive shoots persquare meter (±4 SE) along the mid-depth transect, with nodifference between years (p=0.41). The seagrass communityalong the deep transect was dominated by eelgrass. Eelgrasspercent cover (p<0.0001) and seagrass canopy height(p=0.04) declined along the deep transect between years(Fig. 8), while widgeon grass cover declined to zero (Fig. 8).There were no reproductive shoots of either species. Averageeelgrass density along the deep transect declined from101 shoots/m2 (±18 SE) in 2007 to 47 shoots/m2 (±13 SE)in 2009 (p<0.001). However, the seagrass bed expandedseaward between 21 and 32 m along trajectories from thedeep transect fixed reference points (Table 9).

Biomass Predictions

There was a strong relationship between direct and indirectmethods of measuring eelgrass biomass in the LPB testquadrats. Total biomass harvested from 0.0625-m2 testquadrats in LPB was predicted quite accurately frombiomass estimated from shoot density in the test quadratsand the average total weight per shoot within the central0.0035-m2 core (R2=0.90, p<0.0001; Fig. 9). The rela-tionship was stronger for aboveground (R2=0.91, p<0.0001) than belowground (R2=0.71, p<0.0001) biomasscomponents.

Table 7 Probability of achieving an absolute bias <10 using different numbers of subsamples to estimate mean percent eelgrass cover at tier 2 teststations in Little Pleasant Bay

Station ID True mean (n=12) Standard deviation Probability

No. of subsamples

3 4 5 6 7 8 9 10 11

1 52.5 23.8 0.617 0.661 0.766 0.840 0.918 0.956 0.990 1 1

2 82.5 17.8 0.705 0.807 0.912 0.938 0.987 0.996 1 1 1

3 4.2 6.3 1 1 1 1 1 1 1 1 1

4 73.8 32.8 0.335 0.455 0.587 0.656 0.738 0.865 0.948 0.984 1

5 100 0 1 1 1 1 1 1 1 1 1

6 76.3 31.8 0.343 0.511 0.643 0.722 0.724 0.903 0.952 0.985 1



Multiple linear regression analysis showed that above-ground eelgrass biomass in the LPB tier 3 permanentquadrats depended usually on both percent cover andcanopy height (Table 10). In 2006, the canopy height ofthe vegetation persisting following the severe stormdisturbance ranged from 14 to 67 cm and did not explainany variation in aboveground biomass among quadratsbeyond that already explained by percent cover (p=0.69;Table 10). However, as eelgrass recovered in 2007 and2008, the range in canopy height in the permanent quadratsincreased substantially (8 to 91 cm in 2007, 9 to 110 cm in2008), and the unique contribution of canopy height to

explaining the variation in aboveground biomass washighly significant (p<0.01; Table 10).

Discussion

Feasibility of Monitoring

Environmental monitoring is recognized globally as criticalto environmental decision making (Griffith 1998; OECD2003). The success of monitoring in influencing conserva-tion depends on many facets of program development andexecution, including planning, statistical design, selectionof attributes to measure, data collection and synthesis, andinformation transfer (Elzinga et al. 2001; Reid 2001; Fancyet al. 2009). Ultimately, the overall feasibility of imple-mentation is related to each of these monitoring elements.In this study, we examined whether tiered seagrassmonitoring using time-saving sampling methods enhancesfeasibility while maintaining rigor. Our use of permanentsampling stations within a hierarchical framework maxi-mized monitoring efficiency and improved the potential toproduce statistically meaningful results.

Fig. 7 Shoot density (a), canopy height (b), percent cover (c), andtotal biomass (d) of eelgrass (Z. marina) at tier 3 monitoring site inLittle Pleasant Bay, 2005–2009 (mean ± SE)

Table 8 Differences in tier 3 eelgrass attributes over time in LittlePleasant Bay as determined by repeated measures analysis of variance,2005–2009

Attribute Year

2005 2006 2007 2008 2009

Densitya

Shallow a b c d e

Mid-depth a b ac d c

Deep a bc ab c a

Canopy height

Shallow a b b bc c

Mid-depth a b c cd d

Deep a bc c d ab

Percent coverb

Shallow a b bc c d

Mid-depth ab c b ab a

Deep a a b c d

Total biomass

Shallow a b b bc ac

Mid-depth a b a c c

Deep a a a b a

Multiple comparisons are across years within depths. Like letterswithin a row indicate lack of difference between mean attributes (p>0.05)a Analysis of square root-transformed datab Analysis of arcsin transformed data



Sampling effort is determined by the number and typesof attributes incorporated in a monitoring plan, the methodsused to characterize each attribute, and the number ofsamples collected. We limited tier 2 sampling to only twoindicators of seagrass condition (Table 1), each of which iseasily measured. Resampling simulations at LPB showed

that four subsamples of percent cover per station weresufficient to avoid effects of within-station bias (differencebetween the mean of the subsamples and the true stationmean) on detecting bay-wide change, which served to limitthe time on station. The complexity of our analytical modelfor detecting system-wide trends, incorporating bothrepeated measures and distance zones, prohibited anaccurate calculation of the model’s statistical power orthe probability of detecting a true change in percentcover of seagrass (cf. Heidelbaugh and Nelson 1996).However, resampling simulations using reduced samplesizes showed indirectly that our sample size in LPB (55stations in a 624-ha sampling frame) yielded more thansufficient power to detect an absolute change in meancover from 25% to 49% (distance zone 3). Within thelarger, more complex seagrass system in GSB/MB, oursample size (198 stations in a 9,220-ha sampling frame)was more than sufficient to detect an absolute change inmean cover from 14% to 31% (eelgrass in distance zone1), and even sampling only half the number of stations, we

Fig. 9 Relationship between direct (y-axis) and indirect (x-axis)methods of eelgrass (Z. marina) biomass determination in LittlePleasant Bay test quadrats. Points represent total biomass

Fig. 8 Percent cover of eelgrass (Z. marina) (a), percent cover ofwidgeon grass (R. maritima) (b), and canopy height of the seagrasscommunity (c) at tier 3 monitoring site in Great South Bay/MorichesBay, 2007 and 2009 (mean ± SE)

Site Transect reference point (m) Distance (m)

2004 2005 2007 2008 2009

LPB 0 466 502 1,000 1,000 1,000

25 494 655 993 993 993

50 207 192 203 nd 215

GSB/MB 0 – – 129 – 152

25 – – 138 – 159

50 – – 133 – 165

Table 9 Distance seaward(meters) from fixed referencepoints on the deep transect ofthe tier 3 monitoring stations inLPB and GSB/MB to the edgeof the seagrass bed

Reference points are the begin-ning, middle, and end of the50-m transect. Monitoring atGSB/MB occurred during 2007and 2009 only

nd no data



would still expect to detect an absolute change in coverfrom 4% to 12% (eelgrass in distance zone 3) and from10% to 19% (widgeon grass in distance zone 3) at least50% of the time. Thus, we infer that our sampling design,with an average of 500 m (LPB) to 925 m (GSB/MB)between stations, was appropriate and sufficient in thesesystems. Our sampling design and attributes measured fortier 2 monitoring are very similar to those of the long-termFisheries Habitat Assessment Program in Florida Bay,USA (Durako et al. 2002; Hall et al. 2007), with theaddition of permanent sampling locations for trenddetection. The efficiency afforded by repeated measure-ments at permanent locations minimized the number ofsamples, and the consequent field effort, required to detecta certain magnitude of change in percent cover on a bay-wide scale (cf. Elzinga et al. 2001).

We based our hexagon sizes and associated sample sizesfor tier 2 monitoring on logistical and statistical consid-erations. We initially determined the maximum number ofstations we could visit in LPB and GSB/MB within areasonable time frame and then compared the relative areathat would be sampled in each location (0.0009% of thetotal area in LPB, 0.0002% of total area in GSB/MB) toexisting seagrass assessments with similar spatial designs(approx. 1×10−5% to 9×10−5% of total area was sampled

in South Florida seagrass studies; Fourqurean et al. 2001,2003). Although the use of power analysis to plan samplesizes is complicated by the combined incorporation of bothpermanent sampling stations and distance strata in ananalytical model, our study may offer general guidancefor future seagrass monitoring efforts using permanent,spatially distributed sampling stations. When permanentsampling units are being compared over two time periodsusing an unstratified sampling design, it is fairly simple tocalculate the number of samples required to detect aspecified level of change with a known degree of certainty(statistical power) and probability of falsely concluding achange occurred (probability of committing a type I error,or α). This calculation depends on an estimate of thestandard deviation of the differences in observations atindividual permanent stations between years (sdiff; Elzingaet al. 2001). Within the distance zones in LPB and GSB/MB where the most eelgrass occurred (Table 4, LPB zone3; Table 6, GSB/MB zone 1), sdiff was remarkablyconsistent: LPB zone 3 2006–2007, sdiff=31; LPB zone 32007–2008, sdiff=29; GSB/MB zone 1 2007–2009, sdiff=31. Given a mean eelgrass cover of 50% and a sdiff of 30,from 55 to 94 samples would be required to detect a 20%change in cover using permanent quadrats, depending onthe statistical power (0.80 or 0.90) and α (0.05 or 0.10)

Table 10 Regression coefficients (SE) relating eelgrass aboveground biomass to population parameters measured during tier 3 monitoring inLittle Pleasant Bay

Variable 2006 2007 2008

Constant 1.497 (1.078) 0.201 (0.573) 0.858 (0.763)

Percent cover 0.092* (0.026) 0.044** (0.010) 0.082** (0.014)

Canopy height 0.018 (0.043) 0.074* (0.019) 0.062* (0.022)

R2 0.71 0.81 0.84

No. observations 25 29 36

Dependent variable is square root of aboveground biomass measured indirectly in permanent 0.25-m2 quadrats; explanatory variables areuntransformed

*p<0.01; **p<0.001

Table 11 Number of samples required to detect a specified change in eelgrass cover over two time periods using an unstratified sampling designand permanent sampling stations

Power α Minimum detectable relative change from 50% cover (absolute change in percentage points)

10% (5) 20% (10) 30% (15) 50% (25)

0.80 0.05 282 71 31 11

0.10 221 55 25 9

0.90 0.05 378 94 42 15

0.10 307 77 34 12

Calculations assume a standard deviation of the difference in percent cover between years of 30, an initial eelgrass cover of 50%, and a nullhypothesis of no change. Power = probability of detecting a real change; α = probability of falsely concluding that a change occurred. Thespecified percent changes from 50% cover = absolute changes in cover within parentheses



desired (Table 11). The required number of samples isdirectly related to the power of the test and inversely relatedto the magnitude of difference detected and the probabilityof falsely concluding a change occurred (Table 11).Although pilot sampling is advised to determine the sdiffin a new area, the consistency of sdiff among portions ofdifferent estuaries and different years in our study suggeststhat our data may provide a springboard for initialplanning of similar monitoring programs in seagrasssystems without strong environmental gradients that mightinfluence response (i.e., without strata). If environmentalgradients do exist, however, at large sample sizes (greaterthan approx. n=30), more statistical precision and powerwill be gained through designs that reduce the variationamong stations as much as possible, e.g., by incorporatingstrata as appropriate, than by simply adding samplingstations (cf. Elzinga et al. 2001).

Our use of direct observation for tier 2 monitoringincorporated standard field methods and readily availableequipment such as quadrats and boats. However, otherapproaches for monitoring seagrass condition at meter-scaleresolution throughout entire systems exist. Low-altitudeaerial photographs taken from planes (Short and Burdick1996; Neckles et al. 2005), blimps, or kites (Krause-Jensenet al. 2004) can be used to quantify very small seagrasspatches, and various hydroacoustic techniques for seafloorcharacterization have been applied to measuring seagrasscover, canopy height, and biomass (e.g., Sabol et al. 2002;Komatsu et al. 2003; Riegl et al. 2005; Warren and Peterson2007). These methods rely on specialized equipment fordata acquisition and specialized skills for image and dataprocessing, and require additional field efforts for groundverification and obtaining calibration samples; therefore,depending on the resources available for seagrass monitor-ing, these approaches may not be as cost-effective as directobservations. However, they offer the means to acquirevery fine-scale measurements over very large areas andcould be integrated into a hierarchical seagrass monitoringprogram if resources permit.

The efficiency of our tier 2 data collection was related inlarge part to the relative ease of field sampling. Weintentionally restricted monitoring attributes to seagrasscharacteristics that could be measured from a boat or bysnorkeling: percent cover and canopy height (Table 1).Sampling these attributes was simplified by the generallyshallow depths of the lagoonal estuaries we studied. Byusing an underwater camera, we were able to view thesubstrate clearly from the boat for percent cover observa-tions at even the deepest sampling stations (about 6 m). Wewere able to collect seagrasses from our deepest vegetatedsites (about 3 m) for canopy height measurement using arake on a telescoping handle, but collection devices such asbenthic grabs or corers (e.g., Raz-Guzman and Grizzle

2001) would permit sampling in deeper water. However, invery deep or exposed seagrass systems where remotesampling for percent cover and canopy height is impracti-cable and SCUBA is required, the time and cost formonitoring these attributes will increase accordingly.

Tier 3 monitoring yields more detailed information onseagrass condition while providing insight into possiblecauses of change observed at larger scales. Seagrass shootdensity, areal biomass, and patterns of biomass allocationare sensitive to perturbation and offer valuable insight intopopulation processes and primary production associatedwith bed expansion or retraction (Durako 1994; Kirkman1996; Duarte and Kirkman 2001), but direct shoot countsand biomass harvesting are very labor-intensive. Tieredmonitoring restricted these time-consuming measurementsto a small geographic area and the use of permanentquadrats improved the ability to detect change with alimited number of samples; both of these factors limited thesampling effort required at this scale.

Employing time-saving methodology for tier 3 biomassdetermination further increased the feasibility of monitor-ing. Various nondestructive methods of estimating seagrassbiomass from easily measured variables exist (Duarte andKirkman 2001). For example, models to predict seagrassbiomass from ranked standing crop categories (Mellors1991; Mumby et al. 1997), aerial photographic coverclasses (Moore et al. 2000), and percent leaf covermeasured directly (Orth and Moore 1988) or by photo-graphic analysis (Lee 1997) have been described. Becausethese methods rely on traditional destructive samples tocalibrate biomass regressions, the time saved through theuse of nondestructive sampling is minimal until the totalnumber of samples in the survey is considerably greaterthan the number of calibration samples harvested (Mumbyet al. 1997). In addition, nondestructive methods mayrequire substantial cross-calibration among observers(Mellors 1991), and estimation of aboveground biomassin very dense beds (Lee 1997) or belowground biomass ingeneral (Mellors 1991) may be too imprecise to be useful.By harvesting very small samples (approx. 7-cm diametercores), we limited sample processing time while measuringboth aboveground and belowground shoot biomass. Subse-quent indirect determination of areal biomass from averageshoot weight and shoot density avoided inter-observer biasand was a reasonable proxy for labor-intensive harvest ofquadrats sized appropriately for seagrasses at densities foundin LPB (0.0625- to 0.25-m2 quadrats; Duarte and Kirkman2001). The small core size minimized both the effortrequired for sample processing and the impact of harveststo National Park sites. An important caveat regarding thisindirect method is the need to collect biomass cores froman area in which the plants are the same size as thoserooted in the permanent quadrat used for density



measurements. Our paired indirect and direct measurementsof biomass (Fig. 9) emanated from small (0.0625 m2) areas,which minimized the variation in eelgrass canopy heightbetween the harvested core and the larger quadrat. Duringactual implementation, biomass cores were collected fromareas 0.5–1.0 m away from the permanent quadrats. Plantsize can vary considerably within this distance, so biomasssample locations must be selected carefully. We found thatthe sampling time added by comparing canopy heightmeasurements between permanent quadrats and potentialharvest locations was minimal.

Overall, implementation of tier 2 and tier 3 monitoring atthe sampling intensities we tested required a moderateinvestment of personnel. Annual monitoring of both tierscombined required a total of 3 to 13 field days for amaximum of four people, depending on the size of thesystem, although most sampling was conducted by two tothree people. Limiting the resources required to complete asampling event increases the likelihood of long-termimplementation. Ultimately, conservation decisions are bothlocal and regional in scale, and monitoring tools capable ofevaluating broad regional trends are needed to supportlarger scale management and policy decisions (Urquhart etal. 1998). This is particularly true for managing threats toseagrasses from estuarine nutrient enrichment, whichtypically involve multiple watersheds and localities. Therelative efficiency of tiered monitoring would make itfeasible to implement throughout a regional network.

Relevance of a Tiered Approach

We found targeted monitoring within a hierarchical frame-work to have distinct advantages for understanding seagrassstatus and trends at multiple scales. Sampling at each tierwas optimized to measure seagrass attributes linked to thestressor of primary concern, estuarine nutrient enrichment,and integration of information across tiers offered acomprehensive, yet efficient, assessment of seagrass extentand ecological condition. At both LPB and GSB/MB, tier 1mapping programs provided information on long-termchanges in seagrass distribution on bay-wide scales.Nesting tier 2 and tier 3 monitoring within bays withknown seagrass distribution supplemented seagrass mapswith higher resolution information that is relevant to multi-scale conservation objectives (Table 1) while being usefulfor interpreting mechanisms of change.

The monitoring data from LPB and GSB/MB indicatedan improvement in the condition of seagrass resourcesduring the study period. The increase in percent cover ofeelgrass we observed in LPB following formation of thenew inlet (tier 2 data, Fig. 5 and Table 4) coincided with anincrease in tidal flushing (Kelley and Ramsey 2008). Tier 3monitoring showed an increase in the seagrass depth limit

following the new inlet formation, which is indicative ofenhanced light penetration through the water column(Duarte 1991; Dennison et al. 1993) and helps explain thechanges observed at tier 2. Tier 3 monitoring also revealeda very high capacity for lateral shoot production in thispopulation, which would contribute to increased cover andbed expansion following an improvement in the lightclimate. In GSB/MB, the system-wide increases in eelgrassand widgeon grass cover that occurred from 2007 to 2009(tier 2 data, Fig. 6 and Table 6) bracketed a 2008 brown tidebloom (A. anophagefferens) that was the most intense everrecorded locally (Gobler 2008). Two hypotheses weresuggested to explain the unexpected increase in seagrasscover following the bloom: (1) an increase in sexualrecruitment following a disturbance caused by extremelight limitation (cf. Duarte et al. 2006; Orth et al. 2006b) or(2) an increase in overall light availability despite thebrown tide event. Tier 3 monitoring revealed either adecrease (widgeon grass) or no change (eelgrass) insexual reproduction between 2007 and 2009; however,the considerable increase in widgeon grass cover atshallow depths and the significant bed expansion bothsuggest an increase in light availability between years.Subsequent application of the Wave Exposure Model(Malhotra and Fonseca 2007) showed a reduction in waveenergy during 2008, which could have reduced sedimentresuspension in this shallow coastal lagoon sufficiently tocompensate for increased algal light attenuation. In bothLPB and GSB/MB, tier 3 monitoring also highlightedpatch-scale variability that can contribute to system-widechanges.

The addition of water column light attenuation to thesuite of attributes measured at tier 3 would improve theability to diagnose causal relationships. The SeagrassNetprotocol includes 2-week intervals of continuous lightmeasurement at two depths during quarterly samplingevents (Short et al. 2006b). We have subsequently initiateddeployment of continuously recording light sensors withautomatic wipers at two depths for 4 weeks surroundingannual tier 3 monitoring (Kopp and Neckles 2009).

We established one tier 3 site per estuary as acompromise between information gain and monitoringfeasibility. If time and resources permit, however, monitor-ing results from additional tier 3 sites can help elucidatelarger scale trends. In particular, if tier 2 sampling isstratified by gradients in habitat characteristics, it would beuseful to locate a high-resolution index site within multipletier 2 strata if possible. Although information from oursingle tier 3 site in distance zone 3 of GSB/MB (Fig. 2)helps explain the overall changes in seagrass abundance inthat system, supplemental high-resolution data from dis-tance zone 1 are needed to interpret the mechanismsunderlying the relative increase in dominance of eelgrass



in the western end of GSB near Fire Island Inlet (Fig. 6 andTables 5 and 6).

In areas with substantial depth gradients such as LPB,eelgrass shoot morphology and bed structure vary consid-erably with depth, from dense short shoots in shallow waterto long, relatively sparser shoots in deeper areas (Fig. 7). Inareas of more uniform eelgrass morphology, biomass can bepredicted from percent cover alone with high accuracy(Orth and Moore 1988), but regressions of tier 3 data fromLPB showed that biomass in this location is generallydependent on both the percent cover and the canopy heightof the vegetation (Table 10). Because these metrics aremeasured through tier 2 monitoring, it is possible to applythe biomass models developed through intensive tier 3monitoring to system-wide data collected at tier 2 to predictaboveground biomass distribution on a bay-wide scale (cf.Fourqurean et al. 2001). Such integration of data acrossscales contributes additional information for evaluatingecosystem condition relative to conservation objectives formaximizing seagrass growth (Table 1).

The tiered framework (Bricker and Ruggiero 1998) weapplied to seagrass monitoring is similar in some ways to athree-level process for comprehensive wetland monitoringand assessment that has been recently evaluated (Brooks etal. 2004; Wardrop et al. 2007c). The wetland assessmenthierarchy includes level 1 or landscape assessments usingremote sensing data to characterize wetlands at thewatershed scale (Brooks et al. 2004; Reiss and Brown2007), level 2 or rapid assessments using semiquantitativemetrics and best professional judgment to classify wetlandcondition based on field observations (Fennessy et al.2007), and level 3 or intensive assessments of biological,physical, and chemical characteristics based on quantitativefield measurements (Reiss and Brown 2007; Wardrop et al.2007b). Estimates of wetland condition are produced ateach level by comparing assessment results to referencestandards, and as the intensity of measurement increasesfrom level 1 to level 3, so too does the confidence in theresults. Whereas the tiered framework we adopted relies onnesting of sampling areas and indicators across tiers andconsistent application of all scales of monitoring, thewetland assessment model uses unique indicators at eachlevel to produce information of varying spatial resolutionson the same location, with the number of levels dictated byproject needs; conceptually, the wetland assessment modelis more of a layer cake (with the same geographic area ateach scale of measurement) than a triangle (with geographicarea decreasing in extent from tier 1 to tier 3). For smallseagrass systems, it may be possible to expand the tier 3sampling frame to provide detailed information on a largerarea; conversely, for expansive wetland systems, it mayprove beneficial to limit intensive measurements to a subsetof the rapid assessment area. Common to both hierarchical

approaches, however, integration of information acrossscales can be used to calibrate ecosystem conditioncategories, predict condition at local or regional levels,and focus management targets (c.f. Bricker and Ruggiero1998; CRMSW 2000; Brooks et al. 2004; Reiss and Brown2007; Wardrop et al. 2007c; Whigham et al. 2007).

Monitoring results are most useful for resource managersif monitoring attributes are linked explicitly to conservationobjectives (Nichols and Williams 2006) and are highlyrelevant to management endpoints (Wardrop et al. 2007a).The monitoring attributes in this study were selected forevaluating whether specific seagrass conservation objec-tives related to goals of protecting estuarine resources fromnutrient enrichment are being achieved. The efficiency oftiered monitoring will facilitate generating information onsystem-wide seagrass condition and high-resolution mech-anisms of change to guide conservation decisions.

Acknowledgments Funding for this study was provided by theUS Geological Survey, Park Oriented Biological Support Programand the National Park Service, Northeast Coastal and BarrierNetwork. We thank Fred Short for instruction in the SeagrassNetapproach; Fred Short, Jeffrey Gaeckle, and David Rivers for helpin establishing our tier 3 monitoring site in LPB according to theSeagrassNet spatial design; Holly Bayley, Lena Curtis, SteveDwyer, Bev Johnson, Carrie Phillips, Lotte Rivers, Steve Smith,Adam Thime, and Jesse Wheeler for field assistance in LPB; andBrooke Rodgers and Jamie Brisbin for field assistance in GSB/MB. We also thank Andrew Gilbert and Dennis Skidds for expertGIS assistance and tutelage. We are grateful to Carrie Phillips andMegan Tyrrell of Cape Cod National Seashore and MichaelBilecki of Fire Island National Seashore for logistical support,and we appreciate the cooperation of Dawson Farber, Harbor-master and Shellfish Constable for the town of Orleans, MA, inmaintaining a long-term tier 3 monitoring site in LPB. Thismanuscript was greatly improved by the comments of NancyRybicki, Melisa Wong, and two anonymous reviewers. Use oftrade, product, or firm names does not imply endorsement by theUS Government.

References

ASMFC (Atlantic States Marine Fisheries Commission). 1997.Submerged aquatic vegetation policy. ASMFC Habitat ManagersSeries No. 3, Washington, DC.

Avery, W. 2000. Monitoring submerged aquatic vegetation in Hills-borough Bay, Florida. In Seagrasses: Monitoring, ecology,physiology, and management, ed. S.A. Bortone, 137–145. BocaRaton: CRC.

Borum, J. 1996. Shallow waters and land/sea boundaries. InEutrophication in coastal marine ecosystems, ed. B.B. Jorgensenand K. Richardson, 179–203. Washington: American GeophysicalUnion Coastal and Estuarine Studies.

Borum, J., C. M. Duarte, D. Krause-Jensen, and T. M Greve, eds.2004. European seagrasses: An introduction to monitoring andmanagement. EU Project Monitoring and Managing of Euro-pean Seagrasses (M&MS) EVK3-CT-2000-00044. http://www.seagrasses.org/handbook/european_seagrasses_high.pdf.Accessed 6 January 2009.



http://www.seagrasses.org/handbook/european_seagrasses_high.pdf


Bricker, O.P., and M.A. Ruggiero. 1998. Toward a national programfor monitoring environmental resources. Ecological Applications8: 326–329.

Brooks, R.P., D.H. Wardrop, and J.A. Bishop. 2004. Assessingwetland condition on a watershed basis in the Mid-Atlantic regionusing synoptic land-cover maps. Environmental Monitoring andAssessment 94: 9–22.

Carmichael, R.H., B. Annett, and I. Valiela. 2004. Nitrogen loading toPleasant Bay, Cape Cod: Application of models and stableisotopes to detect incipient nutrient enrichment of estuaries.Marine Pollution Bulletin 48: 137–143.

Cloern, J.E. 2001. Our evolving conceptual model of the coastaleutrophication problem. Marine Ecology Progress Series 210:223–253.

Corbett, C.A., P.H. Doering, K.A. Madley, J.A. Ott, and D.A.Tomasko. 2005. Using seagrass coverage as an indicator ofecosystem condition. In Estuarine indicators, ed. S.A. Bortone,229–245. Boca Raton: CRC.

Cressie, N.A.C. 1993. Statistics for spatial data, revised edition. NewYork: Wiley.

CRMSW (Coastal Research and Monitoring Strategy Workgroup).2000. Clean water action plan: Coastal research and monitoringstrategy. Washington, DC: Coastal Research and MonitoringStrategy Workgroup. http://purl.access.gpo.gov/GPO/LPS17885.Accessed 8 September 2010.

Dale, V.H., and S.C. Beyeler. 2001. Challenges in the developmentand use of ecological indicators. Ecological Indicators 1: 3–10.

Davis, J. 2002. Statistics and data analysis in geology, 3rd ed.Hoboken: Wiley.

Deis, D.R. 2000. Monitoring the effects of construction andoperation of a marina of the seagrass Halophila decipiens inFort Lauderdale, Florida. In Seagrasses: Monitoring, ecology,physiology, and management, ed. S.A. Bortone, 147–155. BocaRaton: CRC.

Dennison, W.C., and R.S. Alberte. 1982. Photosynthetic responses ofZostera marina L. (eelgrass) to in situ manipulations of lightintensity. Oecologia 55: 137–144.

Dennison, W.C., G.J. Marshall, and C. Wigand. 1989. Effect of“brown tide” shading on eelgrass (Zostera marina L.) distribu-tions. In Novel phytoplankton blooms: Causes and impacts ofrecurrent brown tides and other unusual blooms, ed. E.M.Cosper, V.M. Bricelj, and E.J. Carpenter, 675–692. New York:Springer.

Dennison, W.C., R.J. Orth, K.A. Moore, J.C. Stevenson, V. Carter, S.Kollar, P.W. Bergstrom, and R.A. Batuik. 1993. Assessing waterquality with submersed aquatic vegetation. Bioscience 43: 86–94.

Dowty, P., B. Reeves, H. Berry, S. Wyllie-Echeverria, T. Mumford, A.Sewall, P. Milos, and R. Wright. 2005. Puget Sound SubmergedVegetation Monitoring Project 2003–2004 Monitoring Report.Olympia: Washington State Department of Natural Resources,Nearshore Habitat Program.

Duarte, C.M. 1991. Seagrass depth limits. Aquatic Botany 40: 363–377.

Duarte, C.M. 1995. Submerged aquatic vegetation in relation todifferent nutrient regimes. Ophelia 41: 87–112.

Duarte, C.M. 2002. The future of seagrass meadows. EnvironmentalConservation 29: 192–206.

Duarte, C.M., and H. Kirkman. 2001. Methods for the measurement ofseagrass abundance and depth distribution. In Global seagrassresearch methods, ed. F.T. Short and R.G. Coles, 141–153.Amsterdam: Elsevier Science B.V.

Duarte, C. M., E. Alvarez, A. Grau, and D. Krause-Jensen. 2004.Which monitoring strategy should be chosen? In Europeanseagrasses: An introduction to monitoring and management,eds. J. Borum, C. M. Duarte, D. Krause-Jensen, and T. MGreve, 41–44. EU Project Monitoring and Managing of

European Seagrasses (M&MS) EVK3-CT-2000-00044. http://www.seagrasses.org/handbook/european_seagrasses_high.pdf.Accessed 6 January 2009.

Duarte, C.M., J.W. Fourqurean, D. Krause-Jensen, and B. Olesen.2006. Dynamics of seagrass stability and change. In Seagrasses:Biology, ecology and conservation, ed. A.W.D. Larkum, R.J.Orth, and C.M. Duarte, 271–294. Dordrecht: Springer.

Durako, M.J. 1994. Indicators of seagrass ecological condition: Anassessment based on spatial and temporal changes. In Changes influxes in estuaries, ed. K.R. Dyer and R.J. Orth, 261–266.Fredensborg: Olsen and Olsen.

Durako, M.J., M.O. Hall, and M. Merello. 2002. Patterns of changein the seagrass dominated Florida Bay hydroscape. In TheEverglades, Florida Bay, and coral reefs of the Florida Keys:An ecosystem sourcebook, ed. J.W. Porter and K.G. Porter, 523–537. Boca Raton: CRC.

Elzinga, C.L., D.W. Salzer, J.W. Willoughby, and J.P. Gibbs. 2001.Monitoring plant and animal populations. Malden: BlackwellScience.

Evans, N.T., and F.T. Short. 2005. Functional trajectory models forassessment of transplanted eelgrass, Zostera marina L., in theGreat Bay Estuary, New Hampshire. Estuaries 28: 936–947.

Fancy, S.G., J.E. Gross, and S.L. Carter. 2009. Monitoring thecondition of natural resources in US national parks. Environ-mental Monitoring and Assessment 1521: 161–174.

Fennessy, M.S., A.D. Jacobs, and M.E. Kentula. 2007. An evaluationof rapid methods for assessing the ecological condition ofwetlands. Wetlands 27: 543–560.

Finkbeiner, M., B. Stevenson, and R. Seaman. 2001. Guidance forbenthic habitat mapping: An aerial photographic approach.Charleston, South Carolina: US NOAA Coastal Services Center,NOAA/CSC/20117-PUB.

Fletcher, R.S., W. Pulich Jr., and B. Hardegree. 2009. A semi-automated approach for monitoring landscape changes in Texasseagrass beds from aerial photography. Journal of CoastalResearch 25: 500–506.

Foden, J. 2007. Assessment metrics for littoral seagrass under theEuropean Water Framework Directive; outcomes of UK intercal-ibration with the Netherlands. Hydrobiologia 579: 187–197.

Fourqurean, J.W., A. Willsie, C.D. Rose, and L.M. Rutten. 2001.Spatial and temporal pattern in seagrass community compositionand productivity in south Florida. Marine Biology 138: 341–354.

Fourqurean, J.W., M.J. Durako, M.O. Hall, and L.N. Hefty. 2002.Seagrass distribution in South Florida: A multi-agency coordi-nated monitoring program. In The Everglades, Florida Bay, andcoral reefs of the Florida Keys: An ecosystem sourcebook, ed. J.W. Porter and K.G. Porter, 497–522. Boca Raton: CRC.

Fourqurean, J.W., J.N. Boyer, M.J. Durako, L.N. Hefty, and B.J.Peterson. 2003. Forecasting responses of seagrass distributions tochanging water quality using monitoring data. EcologicalApplications 13: 474–489.

Gobler, C. J. 2008. Brown tide affects Long Island’s marineecosystem. In I Fish NY, Fall 2008 ed. Stony Brook, New York:New York State Department of Environmental Conservation.

Gobler, C.J., D.J. Lonsdale, and G.L. Boyer. 2005. A review of thecauses, effects, and potential management of harmful brown tideblooms caused by Aureococcus anophagefferens (Hargraves etSieburth). Estuaries 28: 726–749.

Green, R.H. 1993. Application of repeated measures designs inenvironmental impact and monitoring studies. Australian Journalof Ecology 18: 81–98.

Griffith, J.A. 1998. Connecting ecological monitoring and ecologicalindicators: A review of the literature. Journal of EnvironmentalSystems 26: 325–363.

Hall, M. O., K. Madley, M. J. Durako, J. C. Zieman, and M. B.Robblee. 2007. Florida Bay. In Seagrass status and trends in the



http://purl.access.gpo.gov/GPO/LPS17885



northern Gulf of Mexico: 1940–2002, eds. L. Handley, D.Altsman, and R. DeMay, 243–254. Reston, Virginia: USGeological Survey Scientific Investigations Report 2006-5287and US Environmental Protection Agency 855-R-04-003.

Handley, L., D. Altsman, and R. DeMay, eds. 2007. Seagrass statusand trends in the northern Gulf of Mexico: 1940–2002. Reston,Virginia: US Geological Survey Scientific Investigations Report2006-5287 and US Environmental Protection Agency 855-R-04-003.

Hauxwell, J., J. Cebrián, and I. Valiela. 2003. Eelgrass Zostera marinaloss in temperate estuaries: Relationship to land-derived nitrogenloads and effect of light limitation imposed by algae. MarineEcology Progress Series 247: 59–73.

Heidelbaugh, W.S., and W.G. Nelson. 1996. A power analysis ofmethods for assessment of change in seagrass cover. AquaticBotany 53: 227–233.

Hemminga, M.A., and C.M. Duarte. 2000. Seagrass ecology. Cam-bridge: Cambridge University Press.

Hinga, K. R. 2005. Water quality and ecology of Great South Bay(Fire Island National Seashore Science Synthesis Paper).National Park Service Technical Report NPS/NER/NRTR-2005/019, Boston, Massachusetts.

Howes, B., S.W. Kelley, J.S. Ramsey, R. Samimy, D. Schlezinger, andE. Eichner. 2006. Linked watershed-embayment model todetermine critical nitrogen loading thresholds for Pleasant Bay,Chatham, Massachusetts. Boston Massachusetts: MassachusettsDepartment of Environmental Protection, Massachusetts Estuar-ies Project.

Jackson, J.B., and D.J. Nemeth. 2007. A new method to describeseagrass habitat sampled during fisheries-independent monitor-ing. Estuaries and Coasts 30: 171–178.

Johansson, J.O.R., and H.S. Greening. 2000. Seagrass restoration inTampa Bay: A resource-based approach to estuarine manage-ment. In Seagrasses: Monitoring, ecology, physiology, andmanagement, ed. S.A. Bortone, 279–293. Boca Raton: CRC.

Keeney, R.L., and R.S. Gregory. 2005. Selecting attributes to measurethe achievement of objectives. Operations Research 53: 1–11.

Kelley, S.W., and J.S. Ramsey. 2008. Hydrodynamic model ofChatham Harbor/Pleasant Bay including 2007 North Breach.Mashpee: Applied Coastal Research and Engineering, TechnicalMemo.

Kendrick, G.A., B.J. Hegge, A. Wyllie, A. Davidson, and D.A. Lord.2000. Changes in seagrass cover on Success and Parmelia Banks,Western Australia between 1965 and 1995. Estuarine, Coastaland Shelf Science 50: 341–353.

Kentula, M.E. 2007. Foreword: Monitoring wetlands at the watershedscale. Wetlands 27: 412–415.

Kenworthy, W.J., S. Wyllie-Echeverria, R.G. Coles, G. Pergent, andC. Pergent-Martini. 2006. Seagrass conservation biology: Aninderdisciplinary science for protection of the seagrass biome. InSeagrasses: Biology, ecology and conservation, ed. A.W.D.Larkum, R.J. Orth, and C.M. Duarte, 595–623. Dordrecht:Springer.

Kirkman, H. 1996. Baseline and monitoring methods for seagrassmeadows. Journal of Environmental Management 47: 191–201.

Komatsu, T., A. Mikami, S. Sultana, K. Ishida, T. Hiraishi, and K.Tatsukawa. 2003. Hydro-acoustic methods as a practical tool forcartography of seagrass beds. Otsuchi Marine Science 28: 72–79.

Kopp, B. S. and H. A. Neckles. 2009. A protocol for monitoringestuarine nutrient enrichment in coastal parks of the NationalPark Service Northeast Region. National Park Service NaturalResource Report NPS/NCBN/NRR-2009/110, Fort Collins,Colorado.

Krause-Jensen, D., A. L. Quaresma, A. H. Cunha, and T. M. Greve.2004. How are seagrass distribution and abundance monitored?In European seagrasses: An introduction to monitoring and

management, eds. J. Borum, C. M. Duarte, D. Krause-Jensen,and T. M Greve, 45–53. EU Project Monitoring and Managing ofEuropean Seagrasses (M&MS) EVK3-CT-2000-00044. http://www.seagrasses.org/handbook/european_seagrasses_high.pdf.Accessed 6 January 2009.

Kurtz, J.C., L.E. Jackson, and W.S. Fisher. 2001. Strategies forevaluating indicators based on guidelines from the Environmen-tal Protection Agency’s Office of Research and Development.Ecological Indicators 1: 49–60.

Kurz, R.C., D.A. Tomasko, D. Burdick, T.F. Ries, K. Patterson, and R.Finck. 2000. Recent trends in seagrass distributions in southwestFlorida coastal waters. In Seagrasses: Monitoring, ecology,physiology, and management, ed. S.A. Bortone, 157–166. BocaRaton: CRC.

Kutner, M.H., C.J. Nachtsheim, J. Neter, and W. Li. 2005. Appliedlinear statistical models. New York: McGraw-Hill.

Latimer, J.S., and S.A. Rego. 2010. Empirical relationship betweeneelgrass extent and predicted watershed-derived nitrogen loadingfor shallow New England estuaries. Estuarine, Coastal and ShelfScience 90: 231–240. doi:10.1016/j.ecss.2010.09.004.

Lee, S.Y. 1997. Annual cycle of biomass of a threatened population ofthe intertidal seagrass Zostera japonica in Hong Kong. MarineBiology 129: 183–193.

Lee Long, W. J., L. J. McKenzie, M. A. Rasheed, and R. G. Coles.1996. Monitoring seagrasses in tropical ports and harbours. InSeagrass biology: Proceedings of an International Workshop,Rottnest Island, Western Australia, 25–29 January 1996, eds. J.Kuo, R. C. Phillips, D. I. Walker, and H. Kirkman, 345–350.Nedlands, Western Australia: Faculty of Sciences, The Universityof Western Australia.

Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. Wolfinger, and O.Schabenberger. 2006. SAS for mixed models, 2nd ed. Cary: SASInstitute Inc.

Long, B.G., D.M. Dennis, T.D. Skewes, and I.R. Poiner. 1996.Detecting an environmental impact of dredging on seagrass bedswith a BACIR sampling design. Aquatic Botany 53: 235–243.

Malhotra, A., and M. S. Fonseca. 2007. WEMo (Wave ExposureModel): formulation, procedures and validation. Beaufort, NorthCarolina: NOAA Technical Memorandum NOS NCCOS #65.http://www.ccfhr.noaa.gov/stressors/wemo/default.aspx.Accessed 15 April 2010.

MassDEP (Massachusetts Department of Environmental Protec-tion). 2008. Eelgrass interactive map. Mass DEP WetlandsConservancy Program Eelgrass Mapping http://www.mass.gov/dep/water/resources/maps/eelgrass/eelgrass.htm. Accessed 22April 2009.

McDonald, J.I., G.T. Coupland, and G.A. Kendrick. 2006.Underwater video as a monitoring tool to detect change inseagrass cover. Journal of Environmental Management 80:148–155.

McKenzie, L.J., S.J. Campbell, and C.A. Roder. 2003. Seagrass-Watch: Manual for mapping and monitoring seagrass resourcesby community (citizen) volunteers, 2nd ed. Cairns: State ofQueensland, Department of Primary Industries.

Mellors, J.E. 1991. An evaluation of a rapid visual technique forestimating seagrass biomass. Aquatic Botany 41: 67–73.

Moore, K.A., D.J. Wilcox, and R.J. Orth. 2000. Analysis of theabundance of submersed aquatic vegetation communities in theChesapeake Bay. Estuaries 23: 115–127.

Morris, L.J., and R.W. Virnstein. 2004. The demise and recovery ofseagrass in the northern Indian River Lagoon, Florida. Estuaries27: 915–922.

Mumby, P.J., A.J. Edwards, E.P. Green, C.W. Anderson, A.C. Ellis,and C.D. Clark. 1997. A visual assessment technique forestimating seagrass standing crop. Aquatic Conservation: Marineand Freshwater Ecosystems 7: 239–251.





http://dx.doi.org/10.1016/j.ecss.2010.09.004

http://www.ccfhr.noaa.gov/stressors/wemo/default.aspx

http://www.mass.gov/dep/water/resources/maps/eelgrass/eelgrass.htm

http://www.mass.gov/dep/water/resources/maps/eelgrass/eelgrass.htm

Neckles, H.A., R.L. Wetzel, and R.J. Orth. 1993. Relative effects ofnutrient enrichment and grazing on epiphyte–macrophyte (Zosteramarina L.) dynamics. Oecologia 93: 285–295.

Neckles, H.A., F.T. Short, S. Barker, and B.S. Kopp. 2005.Disturbance of eelgrass Zostera marina by commercial musselMytilus edulis harvesting in Maine: Dragging impacts and habitatrecovery. Marine Ecology Progress Series 285: 57–73.

Neverauskas, V.P. 1987. Monitoring seagrass beds around a sewagesludge outfall in South Australia. Marine Pollution Bulletin 18:158–164.

Nichols, J.D., and B.K. Williams. 2006. Monitoring for conservation.Trends in Ecology & Evolution 21: 668–673.

Nixon, S.W. 1995. Coastal marine eutrophication: A definition, socialcauses, and future concerns. Ophelia 41: 199–219.

Nixon, S., B. Buckley, S. Granger, and J. Bintz. 2001. Responses ofvery shallow marine ecosystems to nutrient enrichment. Human& Ecological Risk Assessment 7: 1457–1481.

NSTC (National Science and Technology Council). 1997. Integratingthe nation’s environmental monitoring and research networksand programs: A proposed framework. Washington: The Envi-ronmental Monitoring Team, Committee on Environment andNatural Resources, National Science and Technology Council.

NYS (New York State) Seagrass Task Force. 2009. Draft report of theNew York State Seagrass Task Force: Recommendations to theNew York State governor and legislature, October 21. http://www.dec.ny.gov/docs/fish_marine_pdf/seagrass.pdf. Accessed16 November 2009.

NYS DOS (New York State Department of State). 2001. Long IslandSouth Shore Estuary Reserve comprehensive managementplan. NYS DOS Division of Coastal Resources. http://www.nyswaterfronts.com/Final_Draft_HTML/Pdf/Main_Page_PDF.htm. Accessed 10 September 2010.

OECD (Organization for Economic Cooperation and Development).2003. OECD environmental indicators: development, measure-ment and use. Paris: OECD Environment Directorate. http://www.oecd.org/dataoecd/7/47/24993546.pdf. Accessed 30 June2010.

Orth, R.J., and K.A. Moore. 1988. Distribution of Zostera marina L.and Ruppia maritima L. sensu lato along depth gradients in thelower Chesapeake Bay, U.S.A. Aquatic Botany 32: 291–305.

Orth, R.J., R.A. Batiuk, P.W. Bergstrom, and K.A. Moore. 2002a. Aperspective on two decades of policies and regulations influenc-ing the protection and restoration of submerged aquatic vegeta-tion in Chesapeake Bay, USA. Bulletin of Marine Science 71:1391–1403.

Orth, R.J., J.R. Fishman, D.J. Wilcox, and K.A. Moore. 2002b.Identification and management of fishing gear impacts in arecovering seagrass system in the coastal bays of the DelmarvaPeninsula, USA. Journal of Coastal Research 37: 111–129.

Orth, R.J., T.J.B. Carruthers, W.C. Dennison, C.M. Duarte, J.W.Fourqurean, K.L. Heck Jr., A.R. Hughes, G.A. Kendrick, W.J.Kenworthy, S. Olyarnik, F.T. Short, M. Waycott, and S.L.Williams. 2006a. A global crisis for seagrass ecosystems.Bioscience 56: 987–996.

Orth, R.J., M.C. Harwell, and G.J. Inglis. 2006b. Ecology of seagrassseeds and dispersal strategies. In Seagrasses: Biology, ecologyand conservation, ed. A.W.D. Larkum, R.J. Orth, and C.M.Duarte, 111–133. Dordrecht: Springer.

Orth, R.J., M.R. Williams, S.R. Marion, D.J. Wilcox, T.J.B.Carruthers, K.A. Moore, W.M. Kemp, W.C. Dennison, N.Rybicki, P. Bergstrom, and R.A. Batiuk. 2010. Long-term trendsin submersed aquatic vegetation (SAV) in Chesapeake Bay, USA,related to water quality. Estuaries and Coasts 33: 1144–1163.

Pergent-Martini, C. and G. Pergent. 1996. Spatio-temporal dynam-ics of Posidonia oceanica beds near a sewage outfall(Mediterranean–France). In Seagrass biology: Proceedings of

an International Workshop, Rottnest Island, Western Australia,25–29 January 1996, eds. J. Kuo, R. C. Phillips, D. I. Walker,and H. Kirkman, 299–306. Nedlands, Western Australia:Faculty of Sciences, The University of Western Australia.

Potvin, C., M.J. Lechowicz, and S. Tardif. 1990. The statisticalanalysis of ecophysiological response curves obtained fromexperiments involving repeated measures. Ecology 71: 1389–1400.

Psuty, N.P., and T.M. Silveira. 2009. Geomorphological evolution ofestuaries: The dynamic basis for morpho-sedimentary units inselected estuaries in the northeastern United States. MarineFisheries Review 71: 34–45.

Pulich, Jr., W., ed. 1999. Seagrass conservation plan for Texas.Austin, Texas: Texas Parks and Wildlife Department

Raz-Guzman, A., and R.E. Grizzle. 2001. Techniques for quantitativesampling of infauna and small epifauna in seagrass. In Globalseagrass research methods, ed. F.T. Short and R.G. Coles, 237–252. Amsterdam: Elsevier Science B.V.

Reid, L.M. 2001. The epidemiology of monitoring. Journal of theAmerican Water Resources Association 37: 815–820.

Reiss, K.C., and M.T. Brown. 2007. Evaluation of Florida palustrinewetlands: application of USEPA levels 1, 2, and 3 assessmentmethods. EcoHealth 4: 206–218.

Riegl, B.M., R.P. Moyer, L.J. Morris, R.W. Virnstein, and S.J. Purkis.2005. Distribution and seasonal biomass of drift macroalgae inthe Indian River Lagoon (Florida, USA) estimated with acousticseafloor classification (QTCView, Echoplus). Journal of Exper-imental Marine Biology and Ecology 326: 89–104.

Robbins, B.D., and S.S. Bell. 2000. Dynamics of a subtidal seagrasslandscape: Seasonal and annual change in relation to water depth.Ecology 81: 1193–1205.

Roman, C.T., and K.W. Able. 1988. Production ecology of eelgrass(Zostera marina L.) in a Cape Cod salt marsh–estuarine system,Massachusetts. Aquatic Botany 32: 353–363.

Roman, C.T., N. Jaworski, F.T. Short, S. Findlay, and R.S. Warren.2000. Estuaries of the northeastern United States: Habitat andland use signatures. Estuaries 23: 743–764.

Sabol, B.M., R.E. Melton Jr., R. Chamberlain, P. Doering, and K.Haunert. 2002. Evaluation of a digital echo sounder system fordetection of submersed aquatic vegetation. Estuaries 25: 133–141.

Sand-Jensen, K., and J. Borum. 1991. Interactions among phyto-plankton, periphyton, and macrophytes in temperate freshwatersand estuaries. Aquatic Botany 41: 137–175.

Sargent, F. J., T. J. Leary, D. W. Crewz, and C. R. Kreur. 1995.Scarring of Florida’s seagrasses: Assessment and managementoptions. Florida Marine Research Institute FMRI Tech. Rept. TR-1, St. Petersburg, Florida.

Schabenberger, O., and F.J. Pierce. 2002. Contemporary statisticalmodels for the plant and soil sciences. Boca Raton: CRC.

Sheridan, P. 2004. Recovery of floral and faunal communitiesafter placement of dredged material on seagrasses in LagunaMadre, Texas. Estuarine, Coastal and Shelf Science 59: 441–458.

Short, F.T., and D.M. Burdick. 1996. Quantifying eelgrass habitat lossin relation to housing development and nitrogen loading inWaquoit Bay, Massachusetts. Estuaries 19: 730–739.

Short, F.T., and C.A. Short. 2003. The seagrasses of the western NorthAtlantic. In World atlas of seagrasses, ed. E.P. Green and F.T.Short, 207–215. Berkeley: University of California Press.

Short, F.T., and S. Wyllie-Echeverria. 1996. Natural and human-induced disturbance of seagrasses. Environmental Conservation23: 17–27.

Short, F.T., D.M. Burdick, and J.E. Kaldy III. 1995. Mesocosmexperiments quantify the effects of eutrophication on eelgrass,Zostera marina. Limnology and Oceanography 40: 740–749.



http://www.dec.ny.gov/docs/fish_marine_pdf/seagrass.pdf

http://www.dec.ny.gov/docs/fish_marine_pdf/seagrass.pdf

http://www.nyswaterfronts.com/Final_Draft_HTML/Pdf/Main_Page_PDF.htm



http://www.oecd.org/dataoecd/7/47/24993546.pdf

http://www.oecd.org/dataoecd/7/47/24993546.pdf

Short, F.T., D.M. Burdick, C.A. Short, R.C. Davis, and P.A. Morgan.2000. Developing success criteria for restored eelgrass, saltmarsh and mud flat habitats. Ecological Engineering 15: 239–252.

Short, F.T., E.W. Koch, J.C. Creed, K.M. Magalhaes, E. Fernandez,and J.L. Gaeckle. 2006a. SeagrassNet monitoring across theAmericas: Case studies of seagrass decline. Marine Ecology 27:277–289.

Short, F. T., L. J. McKenzie, R. G. Coles, K. P. Vidler, and J. L.Gaeckle. 2006b. SeagrassNet manual for scientific monitoringof seagrass habitat, worldwide edition. Durham, New Hamp-shire: University of New Hampshire Publication. http://www.SeagrassNet.org. Accessed 20 August 2010.

Stephan, C. D, R. L. Peuser, and M. S. Fonseca. 2000. Evaluatingfishing gear impacts to submerged aquatic vegetation anddetermining mitigation strategies. Washington, DC: AtlanticStates Marine Fisheries Commission, ASMFC Habitat Manage-ment Series #5.

Stevens Jr., D.L. 1997. Variable density grid-based sampling designsfor continuous spatial populations. Environmetrics 8: 167–195.

Stevens Jr., D.L., and A.R. Olsen. 2004. Spatially balanced samplingof natural resources. Journal of the American StatisticalAssociation 99: 262–278.

Taylor, D.I., S.W. Nixon, S.L. Granger, B.A. Buckley, J.P. McMahon,and H.-J. Lin. 1995. Responses of coastal lagoon plantcommunities to different forms of nutrient enrichment—Amesocosm experiment. Aquatic Botany 52: 19–34.

Urquhart, N.S., S.G. Paulsen, and D.P. Larsen. 1998. Monitoring forpolicy-relevant regional trends over time. Ecological Applications8: 246–257.

Virnstein, R.W. 2000. Seagrass management in Indian River Lagoon,Florida: Dealing with issues of scale. Pacific ConservationBiology 5: 299–305.

Virnstein, R. W. and L. J. Morris. 1996. Seagrass preservation andrestoration: A diagnostic plan for the Indian River Lagoon. St.Johns River Water Management District. Technical Memoran-dum No. 14, Palatka, Florida.

Wardrop, D.H., C. Hershner, K. Havens, K. Thornton, and D.M.Bilkovic. 2007a. Developing and communicating a taxonomy ofecological indicators: A case study from the mid-Atlantic.EcoHealth 4: 179–186.

Wardrop, D.H., M.E. Kentula, S.F. Jensen, D.L. Stevens Jr., K.C.Hychka, and R.P. Brooks. 2007b. Assessment of wetlands in the

Upper Juniata watershed in Pennsylvania, USA using thehydrogeomorphic approach. Wetlands 27: 432–445.

Wardrop, D.H., M.E. Kentula, D.L. Stevens Jr., S.F. Jensen, and R.P.Brooks. 2007c. Assessment of wetland condition: An examplefrom the Upper Juniata watershed in Pennsylvania, USA.Wetlands 27: 416–431.

Warren, J.D., and B.J. Peterson. 2007. Use of a 600-kHz AcousticDoppler Current Profiler to measure estuarine bottom type,relative abundance of submerged aquatic vegetation, and eelgrasscanopy height. Estuarine, Coastal and Shelf Science 72: 53–62.

Watson, D.F., and G.M. Philip. 1985. A refinement of inverse distanceweighted interpolation. Geoprocessing 2: 315–327.

Waycott, M., C.M. Duarte, T.J.B. Carruthers, R.J. Orth, W.C.Dennison, S. Olyarnik, A. Calladine, J.W. Fourqurean, K.L.Heck Jr., A.R. Hughes, G.A. Kendrick, W.J. Kenworthy, F.T.Short, and S.L. Williams. 2009. Accelerating loss of seagrassesacross the globe threatens coastal ecosystems. Proceedings of theNational Academy of Sciences 106: 12377–12381.

Wazniak, C.E., M.R. Hall, T.J.B. Carruthers, B. Sturgis, W.C.Dennison, and R.J. Orth. 2007. Linking water quality to livingresources in a mid-Atlantic lagoon system, USA. EcologicalApplications 17: S64–S78.

Weiss, M.B., P.B. Curran, B.J. Peterson, and C.J. Gobler. 2007. Theinfluence of plankton composition and water quality on hardclam (Mercenaria mercenaria L.) populations across LongIsland’s south shore lagoon estuaries. Journal of ExperimentalMarine Biology and Ecology 345: 12–25.

Whigham, D.F., A.D. Jacobs, D.E. Weller, T.E. Jordan, M.E. Kentula,S.F. Jensen, and D.L. Stevens Jr. 2007. Combining HGM andEMAP procedures to assess wetlands at the watershed scale—Status of flats and non-tidal riverine wetlands in the NanticokeRiver watershed, Delaware and Maryland (USA). Wetlands 27:462–478.

Wilson, R.S., and A.H. Brenowitz. 1966. A report on the ecology ofGreat South Bay and adjacent waters. Garden City: AdelphiUniversity.

Wilson, R.E., K.C. Wong, and H.H. Carter. 1991. Aspects ofcirculation and exchange in Great South Bay. In The GreatSouth Bay, ed. J.R. Schubel, T.M. Bell, and H.H. Carter, 33–42.Albany: State University of New York Press.

Wong, K. C. 1981. Subtidal volume exchange and the relationship toatmospheric forcing in Great South Bay, New York. Ph.D.Dissertation, State University of New York at Stony Brook.



http://www.SeagrassNet.org

http://www.SeagrassNet.org

integrating scales of seagrass monitoring to meet conservation

Documents