The role of the invasive bivalve Ensis directus as food source for fish and birds in the Dutch coastal zone

lable at ScienceDirect

Estuarine, Coastal and Shelf Science 90 (2010) 116e128

Contents lists avai

Estuarine, Coastal and Shelf Science

journal homepage: www.elsevier .com/locate/ecss

The role of the invasive bivalve Ensis directus as food source for fish and birds inthe Dutch coastal zone

Ingrid Tulp a,*, Johan Craeymeersch b, Mardik Leopold c, Cindy van Damme a, Frouke Fey c, Hans Verdaat c

a Institute for Marine Resources and Ecosystem Studies, P.O. Box 68, 1970 AB IJmuiden, The Netherlandsb Institute for Marine Resources and Ecosystem Studies, P.O. Box 77, 4400 AB Yerseke, The Netherlandsc Institute for Marine Resources and Ecosystem Studies, P.O. Box 167, 1790 AD Den Burg, The Netherlands

a r t i c l e i n f o

Article history:Received 25 March 2010Accepted 17 July 2010Available online 23 July 2010

Keywords:invasive specieseidercommon scotermarine birdsshellfishcoastal zonediets

* Corresponding author.E-mail address: [email protected] (I. Tulp).

0272-7714/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.ecss.2010.07.008

a b s t r a c t

The razor clam Ensis directus was introduced to Europe presumably as larvae in ballast water around1978. Starting in the German Bight it spread northward and southward along the continental coastline.Currently it is the most common shellfish species in the Dutch coastal zone, where it mainly occurs in theVoordelta and off the Wadden Sea islands. The mean density of E. directus in the Dutch coastal zoneincreased from around 2e5 individuals m�2 in the late ‘90’s to around 12e19 individuals m�2 from 2002onwards. Diet studies show that E. directusmakes up a significant proportion in the current diet of plaice,sole, dab, flounder and dragonet and in the diet of eider and common scoter. In recent years E. directuscontributed 20e100% of the total wet weight in fish stomachs. The proportion E. directus in the dietincreases with fish length. Based on stomach contents of oiled and beached birds and of faeces samplesthe recent frequency of occurrence is 85e90% in eider and 26% in common scoter. Also waders, gulls andcorvids prey on E. directus but the contribution to the diet is still unquantified. Because of its greatburying depth the species is not easily accessible. Fish either profit from massive die-offs that regularlyoccur, or they extract (probably only the smaller) individuals from the sediment. Sea ducks can extract E.directus from the sediment, while shorebirds and gulls feed on dying E. directus washing up on the shore.E. directus is possibly an important food item for fish and seabirds when they occur in high densities andin the right size classes. Since the availability depends greatly on massive die-offs, shell size, buryingdepth and water depth, it is probably not a very reliable food source. Judging from the role E. directuscurrently plays for the higher trophic levels, its introduction must have caused a major change in thefood relations in its distribution area.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

The razor clam Ensis directus is native to North America and isassumed to have been introduced into Europe as larvae in ballastwater of a ship crossing the Atlantic around 1978 (von Cosel, 1982;Essink et al., 2005, VLIZ Alien Species Consortium, 2008; Daisie,2009). The first strong year class occurred in the German Bight in1979 (vonCosel,1982) and rapidly spreadnorthwardand southwardalong the continental coastline. It now occurs abundantly from theWest coast of Sweden (Gürs et al.,1993) to Northwest France as wellas along the British North Sea coast (Howlett, 1990; Luczak et al.,1993; Armonies and Reise, 1998; Severijns, 2001; Palmer, 2003).Within 30 years, it has become the most abundant (in terms ofbiomass) shellfish in several areas along the continental coast. In the

All rights reserved.

Dutch coastal zone, its distribution has been closely monitored inyearly shellfish surveys since 1995 (Smaal et al., 2001).

In terms of biomass and abundance, E. directus appears to havetaken the place of the trough shell Spisula subtruncata in Dutchnearshore waters, which was the most abundant bivalve species inthe Dutch Coastal zone from the early 1980s (Holtmann et al., 1996;Goudswaard et al., 2008; Baptist and Leopold, 2009). S. subtruncataoccurred in dense banks at depths of 5e20m. Predators of differenttaxa used Spisula banks to feed on, in particular, the starfish Asteriasrubens, the flatfish species plaice Pleuronectes platessa, dab Limandalimanda and flounder Platichthys flesus, and several species of seaduck were known to extensively use S. subtruncata as a food source(Todd, 1915; Hagmeier, 1930; Braber and Groot, 1973; Pihl andRosenberg, 1984; Offringa, 1991; Durinck et al., 1993; Leopoldet al., 1995, Himmelman et al., 2005). Since the start of the 21stcentury densities have decreased rapidly because of failedrecruitment (Craeymeersch and Wijsman, 2006) and, locally,overfishing (Camphuysen et al., 2002; Baptist and Leopold, 2009).

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128 117

After the decrease of S. subtruncata numbers of sea duck winteringin Dutch waters have decreased dramatically (Leopold et al., 1995,ICES, 2005, Arts and Berrevoets, 2007).

It is unclear whether or not E. directus has become an equallyimportant food source for higher trophic levels in the Dutchcoastal zone, as S. subtruncata used to be. Ensis directus has not yetbeen noted as food for fish in European coastal waters in theliterature. In 1998 (Amara et al., 2001) observed that E. directuswas available for flatfish but none were found in their stomachs.However, siphons and feet of Ensis spp. where found in thestomachs of plaice, dab and sole Solea solea before the introduc-tion of E. directus (Braber and Groot, 1973; Ortega-Salas, 1980;Gwyther and Grove, 1981; Marshall and Elliott, 1997; Rijnsdorpand Vingerhoed, 2001). In the Irish Sea Ensis ensis has beenfound to be a major prey of plaice and dab (Jones, 1952; Pentreathand Lovett, 1978; Basimi and Grove, 1985; Carter et al., 1991).

In the intertidal of the Dutch Wadden Sea, common eidersSomateria mollissima (Leopold, 2002; Nehls and Ketzenberg, 2002;Cadée, 2006; Ens et al., 2006), oystercatchers Haematopus ostrale-gus (Swennen et al., 1985) and herring gulls Larus argentatus(Noordhuis and Spaans, 1992; Cadée and Cadée-Coenen, 1994) nowexploit E. directus as a major food source, as do several avianscavengers along Dutch and Belgian sandy beaches where many E.directus wash up (photographic evidence for several waders, gullsand corvids available on www.birdpix.nl). In the eastern North Seaboth common scotersMelanitta nigra and eiders have been noted tofeed on E. directus (Swennen and Duiven,1989; Ens et al., 2002; Fox,2003; Leopold and Wolf, 2003; Wolf and Meininger, 2004;Freudendahl and Jensen, 2006; Leopold et al., 2007; Skov et al.,2008a). It remains unclear however to what extent shellfish pred-ators use E. directus as food source.

Ensis spp. are peculiar shellfish, with elongate and rather fragileshells with valves gaping at both ends. Ensis directus has a high fleshto shell weight ratio and this would seem tomake it profitable prey.Many bivalves invest in strong armor (thick shells) and matchingadductor muscles to keep the valves shut when the animal is underattack. Ensis directus does not have these defences but can rely onits shape, its speed and its behaviour to avoid predation. Its elon-gate shape makes it hard for predators to swallow it whole. Ensisdirectus generally lives buried in a vertical position in the substrate.It has a huge and powerful foot that allows it to retract extremelyfast into the sediment down to a depth of 50 cm in case of danger.Evenwhen exposed at the sediment’s surface, it can react to dangerby jumping away, using its powerful foot for propulsion (Alexander,1979; Swennen et al., 1985). In the intertidal, it will move to thesurface during high tide, exposing its siphons to filter the water forfood and oxygen. In deeper waters, it shows a diurnal rhythm, beingmore abundant at the surface at night (Gollasch et al., 1999). Thismay help to avoid predation from predators that rely largely onvision.

The aim of this study is to describe the role of E. directus as a foodsource for demersal fish and seabirds in Dutch coastal waters. Webring together all (sometimes fragmentary) information collectedfor a suit of purposes in the recent past. We carried out yearlyshellfish surveys from 1995 to 2008 and an extensive fish stomachsampling programme in the Voordelta (SW Netherlands) in 2005and 2007 and compare the results with available historical recordsof stomach contents of demersal fish. We have compiled all avail-able information on sea duck diets off the Dutch coast, both fromthe pre- and post-colonisation period of E. directus.

2. Methods

Ensis directus tends to live in shallower waters than the nativeEnsis species (von Cosel, 1982) and is found from the intertidal

down to water depths of about 20e30 m (von Cosel, 1982). OtherEnsis species rarely occur inwaters shallower than 20 m (von Cosel,1982). Therefore, we consider all Ensis found in nearshore Dutchwaters, i.e. waters shallower than 20 m to be E. directus. This iscorroborated by observations on mass strandings of E. directus onDutch beaches in recent years (www.anemoon.org) and by obser-vations during yearly benthos surveys in Dutch coastal waters(Daan and Mulder, 2006). Hereafter Ensis will be used when Ensisdirectus is meant.

2.1. Benthos sampling

The Dutch coastal area (up to 12 km offshore) has beenmonitored for bivalve distribution since 1995. The area sampledcomprises the entire Dutch North Sea coastal zone (Fig. 1). Thesurveys are mainly carried out as statutory tasks related to Dutchlegislation in fisheries management (Smaal et al., 2005; Perdonand Goudswaard, 2007). Following a stratified approach800e1000 samples are taken in the coastal area each year inspring (Fig. 1). In the coastal zone most of the stations weresampled either with a trawled dredge or with a modified hydraulicdredge. Samples were sieved through a 5 mm sieve, sorted andcounted. Measurements on shell length and weight were, becauseof the damage on the shells, not possible for the majority ofspecimens. In the density calculations the efficiency of thedifferent sampling gears was included. Based on 400 stations,sampled both with a box-corer and a trawled dredge in 2004 and2005 (unpublished data IMARES and Netherlands Institute ofEcology) we estimated the efficiency of the trawled dredge to be50%. As the penetration of the hydraulic dredge and the grab areabout the same, we used this figure for all stations in the calcu-lations of standing stock. This figure is of the same order ofmagnitude as found by Beukema (1974) for a van Veen grab forrazor shells (E. ensis) of about 6 cm (60%). More detailed infor-mation on the yearly stock assessments can be found inGoudswaard et al. (2008).

2.2. Diet demersal fish

To describe historic fish diet data we used published studiesfrom the period 1916 till 2004. Only data from the Dutch andBelgium coastal areas were used, however for the non-commercialspecies only data from other North Sea or Northeastern Atlanticcoastal areas are available. For the description of the historic dietthe same species were selected as in the sampling program.

Information on recent diet of demersal fish species wascollected during four surveys in spring and early autumn 2005 (23May-3 June and 29 August-9 September) and 2007 (14 May-1 Juneand 20 August-7 September) in the Voordelta (Fig. 1). Fish werecollected on board a research vessel that fished with two 6 mshrimp trawls, with a 20 mm (mesh size) cod end and a towingspeed of 2.5 knots. In 2005 larger fish were sampled on boarda commercial vessel with two 4.5 m beam trawls with 80mmmeshsize and a towing speed of 4 knots.

A total of ca 103 hauls were taken in each period in 2005 and 53hauls in each period in 2007. Fish used for stomach analyses werecollected from several hauls within each area. Stomachs (withoesophagus and intestines, this combination is from now onreferred to as stomachs) of lesser weever Echiichthys vipera, drag-onet Callionymus lyra, gobies (Pomatoschistus minutus and P. loz-anoi), scaldfish Arnoglossus laterna, plaice P. platessa, flounder P.flesus, dab L. limanda, sole S. solea and solenette Buglossidiumluteum were collected and stored in 4% buffered formaldehyde.These species were selected because they are the most commonspecies found in the area, not because of their preferred diet.

Fig. 1. Map showing the sampling locations of benthos (dots), bird faeces and finding locations of dead birds mentioned in the text. In the Voordelta area the subareas used in Fig. 5are shown.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128118

Numbers of flounder in 2007 were too low to perform stomachanalyses. In the lab stomachs were dissected and the contents wereexamined under a dissecting microscope. Prey items were identi-fied to species level where possible, counted and weighed (to thenearest 0.001 g). The diet per fish was expressed as % of the weightof the total stomach content and averaged over all individual fishper species. Taxonomic groups were summarized in the groups:Actinaria, Amphipoda, Anomura, Brachyura, Crangonidae, Cuma-cea, Decapoda, Ensis, Mysidacea, Nemertea, Pelecypoda, Platy-helminthes, Polychaeta, unidentified organic material, fish andother. The category “other” consists of prey items that singlycontributed less than 1% to the total wet mass. The gobies consistedof two species but were not analysed separately. Comparisonbetween fish diet and local benthos was possible for 2005 becauseof the combination with a benthos sampling program. For this aimthe study area was divided in three subareas (Fig. 1). These areasoverlapped with those used for the benthos sampling within thesame monitoring program (Steenbergen and Escaravage, 2006).The processing of benthos sampling was carried out as describedabove and in Steenbergen and Escaravage (2006).

The percentage Ensis in diet in relation to fish predator size wasanalysed using a logistic regression. Separate analyses were carried

out for spring and autumn. Because the data were overdispersed,the dispersion parameter was estimated.

2.3. Diet of sea ducks

The historic diet of sea ducks in the Netherlands is relativelywell-known and reviewed (Swennen, 1976, 1991; Leopold et al.,1995, 2001; Camphuysen et al., 2002; Fox, 2003; Kats, 2007), andis only summarized here.

Recent diets of sea ducks were studied by examining fragments ofcrushed shells in the stomachs and guts of common scoters and infaecesof eiders.Eiders comeashoreperiodically to rest. Faecal samplesof eiders were collected on resting sites on several occasions andlocations. Scoters remain at sea at all times in winter and defecate inthe water, rendering faeces studies impossible. Instead, we usedstomachandgut contentsof oil victims,washeduponvariousbeachesin the Netherlands. Ensis hinges and larger shell fragments werecollected from these samples. The size of retrieved Ensis hinges instomachs, guts or faeces were used to estimate ingested Ensis shelllengths, using the regressions given in Leopold et al. (2007). Shellthickness was measured in other shell fragments and these valueswere used to estimate ingested Ensis lengths (Leopold et al., 2007).

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128 119

Hingesweremeasured using a Zeiss StereoMicroscope SV6, equippedwith a Zeiss AxioCam MRc digital camera, and Axiovision4 software(http://www.zeiss.de/c12567be0045acf1/Contents-frame/cbe917247da02a1cc1256e0000491172). Several samples of ducks and faeceswere studied:

During a large die-off of eiders in the Wadden Sea and adjacentNorth Sea inwinter 2001/2002, carcasses were collected from April2000 to March 2002 (Ens et al., 2002). A total of 205 birds werecollected in the Wadden Sea and 166 birds off North Sea beaches.Stomachs and guts were opened and food remains described(presence/absence). No quantitative measurements were taken.

Eider faeces were collected on four occasions in the period2001e2007. In December 2001 and in February 2003, eider faeceswere collected at the ‘Razende Bol’, a sandbank off Texel (Fig. 1).Here several thousands of eiders were resting during high tide. Atotal of 47 (2001) and 45 (2003) individual droppings werecollected and the different species of prey present in the faeceswere identified from remaining hard fragments. Size distribution ofEnsis available to the eiders roosting on the Razende Bol wasdetermined in the same winters by using a Van Veen bottom grab(Leopold et al., 2007). In early April 2007, eider faeces werecollected on the ‘Bollen van de Ooster’ a sandbank in the Voordelta(Fig. 1). In May 2007, eider droppings were collected at a harbourpier on SE Texel. Several dozens of eiders were resting here, whichfed on the adjacent intertidal mudflats. All hard prey remainspresent in the eider faeces collectedwere identified to species level.

When over 100.000 common scoters wintered in theNetherlands in the early 1990’s, small numbers in the main flockresiding on a Spisula bank off the Wadden Sea island Terschellingdied from oil contamination and were collected for stomach anal-ysis in February and March 1993 (n ¼ 31). Twenty-three oiledcommon scoters were collected on the Terschelling beach inJanuary 1995. Since 2000 numbers of common scoters havedecreased dramatically in Dutch coastal waters and there havebeen no major oil spills in the Netherlands. As a result, numbers ofdead birds beached have been low and only 19 common scoterscould be collected from 2001-2007 on Wadden Sea islands.

3. Results

3.1. Recent developments in Ensis

Ensis occurred from the start of the monitoring series in 1995 inthe Voordelta and off the DutchWadden Sea islands. (Figs. 2 and 3).

0

2

4

6

8

10

12

14

16

18

20

1995

1996

1997

1998

1999

2000

2001

den

sity (in

d / m

2)

Fig. 2. Time series of average density (n/m2)

The mean density of Ensis in the Dutch coastal zone increased fromaround 2e5 individuals/m�2 in the late ‘90’s to around 12e19 indi-viduals/m�2 from2002 onwards (Fig. 2). In 2001 averygood spat falloccurred (see also Daan and Mulder, 2006) which resulted in theenormous increase in Ensis in 2002 (Fig. 2). Although the stockremained very high from 2001 onward, there was a strong year toyear variation.Main concentrationswere found in theVoordelta, butrecently also north of the Wadden Sea Islands (Ameland) (Fig. 3).

3.2. Diet of demersal fish in 2005 and 2007 compared to historicdiet descriptions

In total 1423and1445 stomachswere analysed, ofwhich252 and289 were empty in 2005 and 2007 respectively. In 2005 for five ofthenine species (plaice, sole, dab,flounder, dragonet) sampled in theVoordelta, Ensiswas an important prey contributing 20e100% of thediet based on wet mass (Fig. 4) both in spring and in autumn. Inspring stomachs of lesser weever and gobies and in autumn stom-achs of scaldfish also contained small quantities of Ensis (<15% wetmass). Ensis was completely absent in solenette stomachs in bothseasons. Most species ate a variety of benthos species (Fig. 4). Apartfrom Ensis, polychaetes were often found and for some speciescrangonids (scaldfish) and flatworms (solenette and dragonet)wereimportant. In most species the diet did not show large variationbetween spring and autumn. Prey sizes could not be measured butwedidnotice that unlike in autumn the Ensis in the stomachs didnotcontain any shell fragments in spring. In autumn the Ensis eatenweremainly young, small Ensis including their shellswhile in springthey were larger, older specimens, which were ingested withouttheir shells. In spring 2007 only for dab Ensiswas an important preysource, while the contribution of Ensis to the stomach contents ofplaice, dragonet and lesser weever varied between 5 and 16%. Inautumn 2007 Ensis only occurred in the diet of plaice, dragonet andgobies. The large share of Ensis in thefish’ diet in 2005was not takenover by a single other prey group, but the diet of nearly all specieswas more diverse in 2007 compared to 2005 (Fig. 4).

In 2005 a positive relation was found between the presence ofEnsis in the fish stomachs and local Ensis densities (Fig. 5). A size-specific analysis of the diet of the five species for which Ensiscontributes a significant part of the stomach contents, shows that inmost species the contribution to the diet of Ensis increasedwith fishlength. This pattern is significant for plaice, sole and dab in spring2005 (F1,96 ¼ 53.73, p < 0.001; F1,69 ¼ 12.69, p < 0.001, F1,81 ¼12.37,p < 0.001), for plaice, dab and dragonet in autumn 2005

2002

2003

2004

2005

2006

2007

2008

of Ensis spp. in the Dutch coastal waters.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128120

(F1,108 ¼ 17.44, p < 0.001; F1,70 ¼ 9.00, p ¼ 0.004; F1,74 ¼ 8.35,p ¼ 0.005) and only for plaice in autumn 2007 (F1,165 ¼ 40.66,p < 0.001) (Fig. 6). Surprisingly the proportion of Ensis in the dietdecreased with flounder size in spring 2005 (F1,44 ¼ 9.73,p ¼ 0.003).

In conclusion Ensis plays an important role in the current diet ofthe larger flatfish species and dragonet at least in some years. Forlesser weever, gobies, scaldfish and solenette Ensis was not animportant prey item (contributing less than 10% to the diet). In thepre-Ensis era the diet of dragonet consisted mostly of echino-dermata, polychaetes and crustaceans (Creutzberg andWitte,1989;Van der Veer et al., 1990; Klimpel et al., 2003) (Table 1). The diet ofplaice consisted of polychaetes and molluscs (Blegvad, 1916; Jones,1952; Groot, 1964; Rijnsdorp and Vingerhoed, 2001), with smallerplaice taking a larger proportion of polychaetes and shrimps (Beystet al., 1999; Amara et al., 2001). The larger plaice gradually switched

Fig. 3. Spatial distribution and maximum densities of Ensis spp

to molluscs, mostly bivalves (Basimi and Grove, 1985). Seasonalchanges in the diet occurred, with polychaetes being more impor-tant inwinter while in spring and summer consumption of bivalveswas higher (Basimi and Grove, 1985). In the Irish Sea E. ensiswas animportant prey species for plaice (Jones, 1952; Pentreath andLovett, 1978; Basimi and Grove, 1985; Carter et al., 1991). The dietof flounder consisted of polychaetes, amphipods, crustaceans andbivalves (Muus et al., 1999). The diet of dab was diverse, withpolychaetes, crustacea andmolluscs being the most important prey(Amara et al., 2001). Dab and plaice are described as competitors forfood (Muus et al., 1999), and like plaice, dab also preyed on E. ensisin the Irish Sea (Jones, 1952; Ortega-Salas, 1980; Gwyther andGrove, 1981; Carter et al., 1991). The diet of sole was dominatedby polychaetes and crustaceans (Beyst et al., 1999, Amara et al.,2001; Rijnsdorp and Vingerhoed, 2001; Hoines and Bergstad,2002). Rijnsdorp and Vingerhoed (2001) showed that bivalves

. (n/m2) the Dutch coastal zone in the period 1995e2008.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128 121

were the main prey of plaice and sole at the beginning of the 20thcentury, while in 1996 diet of both species consisted mainly ofpolychaetes.

3.3. Recent diet of eiders compared to historic diet descriptions

The main feeding area for eiders in The Netherlands is theWadden Sea. Bivalves, particularly blue mussels and edible cocklesgenerally make up the largest part of the diet of common eider inthe Dutch Wadden Sea, with a smaller share of other bivalves,crabs, starfish and periwinkles (Swennen,1976,1991; Leopold et al.,2001; Kats, 2007). Whenmussel and cockle stocks were very low inthe early 1990’s, after a series of years with low recruitment and on-going shellfish fisheries, tens of thousands of eiders went to theNorth Sea, where they joined the common scoters and fed on S.subtruncata (Leopold, 1993), but suffered a high mortality(Camphuysen et al., 2002). During the next low in Wadden Seastocks of mussels and cockles in 2005, most eiders remained in theWadden Sea and did not suffer high mortality; this time they werebelieved to have been saved by the increased stock of Ensis (Enset al., 2006). At the time there was, however, very little directevidence of eiders feeding on Ensis. Swennen and Duiven (1989)described a first winter eider that had died from ingesting fivelarge (9.4e13.4 cm long) Ensis and failed to crush these in thestomach. Leopold (2002) found one eider faeces with Ensis remainson a littoral sand flat off Texel and witnessed eiders scavengingwhole Ensis (range 7e12 cm shell length), discarded froma lugworm boat in February 2001.

For this study stomach and gut analysis was performed on 317dead eiders that died during a die-off in 2000e2002 in theWadden

Fig. 4. Diet of the most common demersal fish species: plaice, sole, dab, flounder, solenettethe Voordelta. Because of low catches, flounder was not sampled in 2007. The diet is expressampled.

Sea. Only 24 birds of the 317 studied had empty stomachs and guts,and another 38 only had non-food items such as grit or plastic intheir stomachs. The diet of the remaining 255 birds was clearlydominated by prey from theWadden Sea (mussels and cockles) andpossibly from stony dikes and breakwaters (periwinkles and greencrabs); Ensis was only found in 6 birds and Spisula in only 4(Table 2). Four of the birds with Ensis were found in the WaddenSea; the other two on the Razende Bol (Fig. 1). The Wadden Seabirds contained mostly shell fragments, and one intact Ensis of7.4 cm shell length. Stomach and gut of one bird from the RazendeBol were full of Ensis shell fragments and three intact Ensis werefound in the stomach, ranging from 12.7 to 13.5 cm. All 4 birds withSpisula were found on the Hondsbossche Zeewering, a sea-dike atthe Dutch mainland North Sea coast. Between them, 17 intact Spi-sulawere found in the stomachs, ranging in shell length from 2.0 to2.8 cm.

In the faecal samples from roughly the same period as thestomach and gut samples (2001 and 2003 at the Razende Bol.) ninedifferent prey species were found. Ensiswas present in 87 out of 92droppings (94.7%) and 78.3% of the samples contained Ensisexclusively. The contribution of the other prey species wasmarginal(Table 3). In both winters Ensis was the dominant bivalve species,comprising respectively 55% and 95% of bivalve standing stock inthe Dutch coastal area (Craeymeersch, 2007). In 2001 a consider-able proportion of the bivalve biomass wasmade up byDonax (26%)andMacoma (10%), but Ensiswas apparently selected as food by theEiders, as this was relatively more commonly present in the faeces(Table 3).

The sizes of Ensis eaten were reconstructed from shell hingespresent in the faeces (see Leopold et al., 2007 for details). The birds

, scaldfish, dragonet, gobies and lesser weever in spring and autumn 2005 and 2007 insed as the average weight percentage each prey group was represented in all stomachs

050

100150200250300350400450500

AREA 1 AREA 2 AREA 3

den

sit

y (

n/m

2)

Ensis density

0

10

20

30

40

50

60

70

80

90

100

AREA 1 AREA 2 AREA 3

% E

nsis

in

sto

mach

plaiceflounderdabsoledragonet

Fig. 5. Densities of Ensis in the three subareas in the Voordelta area (upper) andpercentage in stomach contents of the most important Ensis consuming demersal fishspecies caught in the three subareas in autumn (lower). See for locations of areas Fig. 1.Bars represent standard errors.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128122

took 0-group animals in December 2001, with lengths ranging from40.9 to 64.4 mm. In that year, mainly 0-group Ensis were availableto the ducks. The reconstructed shell lengths of the 0-group Ensistaken in December 2001 were ca 5 mm larger than the lengths ofthe same cohort sampled at sea twomonths later, in February 2002(t ¼ 5.88, P < 0.01). The shape of the size distribution of Ensispresent in the seabed and of Ensis found in the faeces is very similar,with an offset of 5 mm (Fig. 7). In the next winter, both 0-group and1-group Ensis were found abundantly in the Van Veen grabsamples, but the birds selected the 1-group animals (Fig. 7).Presumed 1-group Ensis (n ¼ 53), with lengths from 65 to 110 mmpredominated. Sizes of 1-group Ensis present in the seabed andtaken by eiders in February 2003 were not statistically different (t-test, NS). In this situation, the ducks clearly selected the one-yearolds, but did not show further, within-year class size selection(Fig. 7).

Several years later a total of 46 eider faeces were collected on 31March and 6 April 2007 at a roost on the “Bollen van de Ooster” inthe Voordelta (Fig. 1). Prey composition was more varied than onthe Razende Bol, but again, E. directus was the dominating preyspecies (Table 3). Only eight Ensis hinges were found sufficientlyintact for estimating shell lengths in these samples, indicatinga size range of 6.0e8.6 cm shell length. Measurements on thethickness of shell fragments provide more, but less precise infor-mation on sizes of Ensis eaten (Leopold et al., 2007). On average,

thickness of 6 fragments was measured in each of the 37 samplescontaining Ensis. As we do not know from which location in theshell these fragments originated, wemay assume that they were allrepresenting average thickness of shell, or, alternatively that theywere the thinnest or thickest fragments. Based on these assump-tions, ingested shell lengths ranged from 6 to 11 cm to 5e9 cm,with6e9 cm on average. Therefore it is likely that here also mainly 1-group Ensis were eaten. For this period and location there is noinformation on available size classes of Ensis.

Cadée (2006) noticed several thousands of eiders feeding onEnsis in the Wadden Sea just off SE Texel in March 2006. He foundEnsis fragments in circa 90% of eider faeces dropped at the sea-dikeand concluded that Ensis was now the major food source of thesebirds, at least at this site. In spring 2007 eider faeces that weredominated by Ensis fragments were collected on the same site on SETexel. These samples yielded 49 Ensis hinges, representing shelllengths from 4.2 to 8.5 cm and three larger specimen, of 10.8, 12.5and 12.9 cm long.

In conclusion Ensis has become an important component of thecurrent diet of Eider in the Dutch coastal zone from 2001 onwards.Eiders are apperently able to handle Ensis of at least up to 12 cm.

3.4. Recent diet of common scoters compared to historic dietdescriptions

The historic diet of the common scoter has been summarized byLeopold et al. (1995b), Fox (2003) and ICES (2005). These sea duckfeed on any locally abundant bivalve, including Ensis (Aulert andSylvand, 1997; Freudendahl and Jensen, 2006; Skov et al., 2008a)and the Ensis-like Pharus legumen (Hughes et al., 1997). Off theDutch coast S. subtruncata was the staple food when scoters wereabundant wintering visitors in the 1990’s (Leopold et al., 1995,Leopold, 1996). Off western Denmark, S. subtruncata has also beenfound as the staple food of common scoters (Durinck et al., 1990).Elsewhere, the diet was more varied. Small tellinids (Macoma,Tellina, Abra) were probably important in Belgium and Wales(Degraer et al., 1999; Kaiser et al., 2006) and juveniles of largershellfish such as Arctica, Astarte, Cerastoderma, Donax, Mactra, Mya,Mytilus and Natica (Kirchhoff, 1979; Meissner and Bräger, 1990;Schricke, 1993; Aulert and Sylvand, 1997) were found as prey inthe Baltic and in France. Scoters appear to take any prey that isabundant, that comes in the right size class and that is accessible,including fish and fish eggs (Frengen and Thingstad, 2002; Evert,2004) and even a lost ship load of beans (Bauer & Glutz vonBlotzheim, 1969). The first evidence that Scoters had startedfeeding on Ensis in the Netherlands is given by Leopold and Wolf(2003) and Wolf and Meininger (2004), from direct observationson feeding birds in the SWof the country. Stomachs of 383 commonscoters were examined after a mass die-off caused by an oil spill inthe Voordelta in January 1988 (Camphuysen, 1989). The diet con-sisted of a variety of prey items, but mostly of S. subtruncata. Ensiswas not found in any of these stomachs (Offringa, 1991).

S. subtruncata was the main prey species found (in 23 of 31scoters found dead) on Terschelling in 1993; the remaining 8stomachs were empty. Next to Spisula, remains of Cerastodermaedule were found in one bird and remains of Donax vittatus and ofMacoma balthica in two birds each (in all cases next to Spisula, thatwas the main prey). In 1995 the majority of the scoters still fed on S.subtruncata. Spisula was found in 13 (of the 23) common scoters(the remaining 10 stomachs were empty). However, a few smallfragments of Ensis were found in three of the common scoterstomachs; these mark the first incidence of Ensis in the diet ofscoters in the Netherlands.

In the stomachs of five of the birds collected in the period2001e2007, 21 measurable Ensis hinges could be collected

Fig. 6. Percentage Ensis in diet in relation to fish size in five demersal fish species in spring and autumn. The lines are logistic regressions on the individual fishes (% in diet ¼ exp(a þ b*size)/((1 þ exp(a þ b*size)) with a and b: 2005 spring: plaice: a ¼ �6.065, b ¼ 0.2086; sole: a ¼ �5.320, b ¼ 0.1856; dab: a ¼ �2.857, b ¼ 0.1963; flounder: a ¼ 7.5500,b ¼ �0.2181; 2005 autumn: plaice: a ¼ �2.1940, b ¼ 0.1405; dab: a ¼ �3.316, b ¼ 0.1716; dragonet: a ¼ �3.278, b ¼ 0.1350; 2007 autumn: plaice: a ¼ �4.730, b ¼ 0.2553). Onlysignificant relations are indicated.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128 123

(frequency of occurrence ¼ 26%). The 21 Ensis hinges found in thebirds at the Wadden Sea isles corresponded to a bimodal shelllength distribution ranging from 3 to 9 cm, with two maxima (at 4and 8 cm). One bird with Ensis that was found on Texel at 17September 2004 also contained a few shell fragments of bluemussel. A combination ofDonax and Spisula fragments was found inone other bird wit Ensis, found on Texel, March 2001.

In conclusion the fragmentary information that is availableindicates that the first indications that Ensis is taken by commonscoters originate from 1995.

4. Discussion

4.1. Complications of methods used

Results obtained from diet studies in fish based on stomachanalyses may be biased by several factors. Firstly different preyitems may digest at different speeds, therefore the results will bebiased towards prey species that remain recognizable in thestomachs the longest. Shellfish have the advantage that hard preyitems are often well recognizable. Also stomach analysis onlyprovides information on the diet of the past few hours. Fishstomachs were only collected in the area with the highest densitiesof Ensis (the Voordelta), and may therefore not be representativefor the complete coastal zone. The studies on recent sea duck dietswere also conducted in situations where Ensis was the mostabundant prey species, therefore the results may be biased towardsEnsis. Other studies have shown that common scoters and eiderswill take any abundant bivalve that is of the right size (large enough

to be profitable and small enough for swallowing) (Leopold et al.,1995, 2001).

The contribution of different prey to the fish diet was expressedaspercentageof the totalweightof the stomachcontent. Because thediet studies inbirdsweremainly basedonanalyses of faeces, thedietcontent was presented as frequency of occurrence. But for thepurpose of this study: to describe the relative importance of Ensis inthe diet of fish and birds this is not considered a major problem.

The fact that the size distributions of Ensis present in the seabedand of Ensis found in the faeces showed an offset of 5 mm (Fig. 7)suggests that the difference should not be attributed to selectivesampling (either through size selection by the birds or by selec-tively missing relatively large Ensis in the grab samples). Morelikely, a systematic error was involved in estimating the size ofthese small 0-group Ensis from the hinges still present in the faecesand we conclude that the eiders took the available 0-group Ensiswithout size selection.

In our analyses we considered all Ensis found to be E. directus.This assumption is corroborated by data from the national moni-toring program of macrobenthos (using box-corers) that indicatethat E. directus is by far the most common and numerically domi-nant razor clam species in the Dutch coastal waters. Other nativespecies (E. ensis, Ensis arcuatus) are only found sporadically(Holtmann et al., 1996; Tempelman et al., 2009).

4.2. How suitable and important is Ensis as a prey source?

Since 2001 the Ensis stock increased rapidly in the Dutch coastalzone. At the same time S. subtruncata gradually disappeared

Table 1Historic diet composition in Northeastern Atlantic coastal areas, by percentage ofweight, for the most important prey groups.

Predator Polychaetes Crustacea Fish Molluscs Ensisensisf

Lesser weevera e 46 53 e e

Dragonetb 19e28 3e8 0e6 0e5 e

Gobiesc

(Pomatoschistusminutus)

41 47 3 1 e

Scaldfish e e e e e

Plaiced, e, f 15e66 4e76 0e3 0e28 0e15Flounderg 15e50 1e10 e e e

Dabe, h 0e7 0e58 0e18 0e6 e

Soled, e 52e100 0e28 0e14 0e1 e

Solenettei 39 44 e 21 e

a Vasconcelos et al., 2004.b Van der Veer et al., 1990.c Leitão et al., 2006.d Rijnsdorp and Vingerhoed, 2001.e Beyst et al., 1999.f Basimi and Grove, 1985.g Hampel et al., 2005.h Hoines and Bergstad, 2002.i Amara et al., 2004

- means no data on percentage of weight is available.

Table 3Occurrence of different prey species in Eider faecal samples collected at the RazendeBol in the winters of 2001 (n ¼ 47) and 2003 (n ¼ 45) and 46 Eider faecal samplescollected at the Bollen van de Ooster, March/April 2007. The numbers of samplescontaining each of the prey species and the frequency of occurrence (FO) are given.

Prey species Scientific nameMarch/April 2007

Razende bol Bollen vande Ooster

Dec-01 Feb-03 March/April 2007

n FO (%) n FO (%) n FO (%)

Annelids Nereis longissima 0 0.0 0 0.0 2 4.0Brown shrimp Crangon crangon 0 0.0 0 0.0 1 2.0European green crab Carcinus maenas 7 14.9 0 0.0 0 0.0Undet. crab Carcinus/Liocarcinus 0 0.0 0 0.0 3 6.0Periwinkle Littorina littorea 2 4.3 0 0.0 0 0.0

Natica alderi 0 0.0 1 2.2 0 0.0Edible cockle Cerastoderma edule 1 2.1 0 0.0 2 4.0Venus clam Venus sp. 0 0.0 0 0.0 2 4.0Rayed trough-shell Mactra corallina 0 0.0 0 0.0 1 2.0Through shell Spisula subtruncata 4 8.5 1 2.2 1 2.0American razor clam Ensis directus 45 95.7 42 93.3 37 79.0Blue mussel Mytilus edulis 0 0.0 5 11.1 3 6.0Banded wedge-shell Donax vittatus 0 0.0 1 2.2 2 4.0Baltic tellin Macoma balthica 1 2.1 0 0.0 0 0.0Banded Brittle

StarfishOphiolepis superba 0 0.0 0 0.0 15 32.0

Heart urchin Echinocardium cordatum 0 0.0 0 0.0 6 13.0

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128124

(Goudswaard et al., 2008). The distribution of the two speciesshows similarities: they occur in highest concentrations in theVoordelta and off the Wadden Sea islands. Although Ensis stocksremained high in the last 6 years, there is a strong year to yearfluctuation. Good spat fall occurred irregularly (Daan and Mulder,2006). The actual driving factors behind recruitment variabilityare unknown but food availability and predator density are candi-dates, while winter temperatures have been shown to affectrecruitment success in other species (Beukema et al., 2009). Thesurvival of juveniles during their first winter is thought to be low,but can reach levels higher than 50% (Armonies and Reise, 1998).Survival during winter can be influenced by temperature, storms orby oxygen deficiency after algae blooms (Armonies and Reise,1998). Survival of older individuals can also be influenced bya poor physiological condition after reproduction (Cadée, 2001). Onthe Dutch North Sea coasts high numbers of dead and dying Ensiscan be found during certain periods (Cadée et al., 1994; Cadée,2001). The cause of these periods of mass mortality is not clear(Cadée et al., 1994).

Plaice, dab and sole have been found to feed on siphons of Ensisspp. or on E. ensis in the Irish Sea (Braber and Groot, 1973).However the importance of Ensis spp. or molluscs in the diet byweight never exceeded 28%. Lesser weever, dragonet and gobies

Table 2Food items found (presence/absence) in 255 eiders with non-empty stomachs,collected during a mass die-off in the Wadden Sea and adjacent North Sea in2000e2002. The numbers of samples containing each of the prey species and thefrequency of occurrence (FO) are given.

Prey species Scientific name n birds FO (%)

Undet. crab Carcinus/Liocarcinus 3 1.2Edible crab Cancer pagurus 1 0.4Flying crab Liocarcinus holsatus 1 0.4European green crab Carcinus maenas 15 5.9Hermit crab Eupagurus bernhardus 1 0.4Blue mussel Mytilus edulis 174 68.2Periwinkle Littorina littoralis 44 17.3Edible cockle Cerastoderma edule 34 13.3American razor clam Ensis directus 6 2.4Through shell Spisula subtruncata 4 1.6Fish Pisces 1 0.4

have not been reported to feed on Ensis or molluscs prior thisstudy. Most fish species are considered selective and opportunisticpredators (Mittelbach, 2002). This is also reflected in the predationon E. directus. In 1998 Amara et al. (2001) showed that E. directuswas available in low densities (5.1 g m�2 AFDW) in the coastal areaof the southern North Sea. However, stomachs of flatfish caught inthe same area did not contain any remains of Ensis. This is inaccordance with our finding that the occurrence in the stomachsdepended on the local densities (Fig. 5). Despite higher densities in2007, the role of Ensis in the fish diet was lower. This is probablypartly due to the fact that in 2007 we sampled less of the largersized fish (Fig. 6). In spring 2005 the fish most likely fed on dyingEnsis, based on the observations that the fish nets were full of Ensisshells still containing meat. In autumn they fed on the 0-group. Inspring and autumn 2007 large Ensis shells were found in thecatches, but these were all empty. So it is likely that the availabilityrather than the densities as such determine the importance ofEnsis as a prey species for fish. This brings us to the question howfish manage to get hold of such a deep burying species (seeSection 4.3).

Scoters and eiders have been found to feed on different speciesof razor clams, in Norway (Thingstad et al., 2000), Denmark (Fox,2003; Skov et al., 2008a), The Netherlands (Leopold and Wolf,2003; Wolf and Meininger, 2004; Ens et al., 2006; Leopold et al.,2007), France (Aulert and Sylvand, 1997) and Wales (Hugheset al., 1997). The importance of Ensis for the diet clearly dependson local abundances, both of Ensis and alternative prey species.Ensis, particularly larger specimen must be an awkward prey tocatch, handle, swallow and crush. Sea duck stomachs are strong“shell crushing machines” but their shape is basically round andthey seem ill-adapted to dealing with very long and hard-shelledprey, even though Ensis up to about 11 or 12 cm shell length maybe swallowed whole (Swennen et al., 1985; Thingstad et al., 2000).Both common scoters and eiders have been found dead afterattempts to swallow razor clams that were apparently too large(Swennen and Duiven, 1989). On the other hand, small Ensis maybe suitable prey, certainly for the comparatively large eiders. Eidersmay recently even have escaped mass starvations during a period

Fig. 7. Length-frequency distributions of Ensis as reconstructed from hinges in eider faeces collected at the Razende Bol in December 2001 (upper panel) and in February 2003(lower panel), compared to sizes of Ensis found in bottom samples near the roost (black bars) in the same winters. No growth is thought to occur in shellfish in midwinter, i.e.between sampling faeces and live Ensis. Shell lengths in mm, in 5 mm bins.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128 125

when their preferred prey, mussels and cockles were scarce in theWadden Sea, by switching to Ensis (Ens et al., 2006). Scoters maybe less well equipped to eat Ensis due to their smaller size,although scoters may thrive on smaller species or individuals ofrazor clams (Hughes et al., 1997). Numbers of scoters wintering inDutch coastal waters have been decimated after the decline ofSpisula (Baptist and Leopold, 2009) and have not been rebuiltconcurrently with the growth of the Ensis stock. In Denmarkscoters seem to prefer Spisula. Here Ensis and Spisula occur indisaggregated habitats. Early in winter they feed in the areasdominated by Spisula and only later in the season when Spisulapresumably become depleted they move to the Ensis areas(Petersen and Skov, 2007; Skov et al., 2008b). This does not seemto have occurred in The Netherlands, probably because the speciesnever co-occurred in high densities.

4.3. How do fish and birds get hold of Ensis?

This study shows that Ensis can be an important prey for certainbird and fish species. It is not knownhow Ensis is caught by differentpredators and how Ensis react to this predation pressure. Ensis canflee as a result of chemical or physical disruption by burying them-selves deep in the sediment or by fleeing from the sediment andswimming away (Armonies and Reise, 1998; Muir, 2003). The latterhas also been shown as an avoidance response to predation bynaticid snails (Schneider, 1982). In North America (New Jersey toNorth Carolina) the nemertean worm Cerebratulus lacteus feeds onEnsis by entering the burrow from below and engulfing the anteriorend of the clam. This forces the clam to project much of its bodyabove the surface (sometimes leaving the burrow), thus becomingsubject to surface predators (McDermott, 1976).

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128126

There are no direct observations of predation on Ensis by fish. Itis clear though that only larger fish seem to get hold of Ensis (Fig. 6).The fish species that take Ensis include the larger flatfish and toa lesser degree dragonet. For the smaller species gobies, lesserweever and scaldfish Ensis only makes up a small proportion of thediet (Fig. 4). The stomach contents provide no information on thesize of ingested prey. Hinges are generally not found and only fleshis ingested. In autumn 2005 however the stomachs contained shellfragments of spat. Based on the information of 2005 and 2007 itseems that Ensis is only catchable if they are at the surface becauseof a massive die-off of adult clams or if spat is available. Presumablythe young Ensis bury less deep and can be caught more easily.

The manner in which birds exploit Ensis may occur in threedirect ways and possibly also indirectly:

(1) Direct feeding on live Ensis in the subtidal. Scoters and Eidershave been found to feed on Ensis in subtidal waters of around10 m depth (Cadée, 2006). The feeding method used is to divedown to the sea floor, extract the Ensis from the sediment,swallow themwhole one by one and crush and digest the foodin the muscular stomach.

(2) Direct feeding on live Ensis in the intertidal. Waders, particu-larly oystercatchers have learnt to extract live Ensis fromemerged sand flats (Swennen et al., 1985). Ensis is found by eyeand the bill is quickly inserted in the whole in the sand flat thatreveals its presence. Oystercatchers eat only the flesh that theyextract from the shells before swallowing.

(3) Direct feeding on dead or dying Ensis washed up at the hightide mark. Mass strandings of Ensis occur regularly and gulls(up to >10.000 per event) may feast on these shellfish e.g.(Cadée, 2001). Smaller birds such as waders or corvids also usethis new food source regularly, but such cases have only beendocumented on pictures on the internet.

(4) Indirectly, if dense Ensis stands attract an associated fauna ofsmall fish, such as gobies. Coastal, diving seabirds that prey onsuch small fish, like divers Gavia, grebes Podiceps or fish-eatingducks like mergansers Mergus merganser or long-tailed ducksClangula hyemalis may benefit from such a mechanism buttheir feeding ecology has not yet been studied in relation toshellfish occurrence.

4.4. The role of Ensis in the ecosystem

Because of the lack of historic data on distribution and stock sizeof the native Ensis species it is impossible to evaluate whether ornot there is a competition between the native and introduced Ensisspecies and what the result of this may be regarding Ensis preda-tion. Prior to the arrival of E. directus two Ensis species werecommon in the Dutch coastal area, namely E. arcuatus and Ensisminor (Backeljau, 1986). The former seems to have withdrawn toareas further offshore, while the latter seems to be extirpated.Although this hypothesis is based on erratic observations, it is clearthat since the introduction of E. directus far less encounters withboth indigenous Ensis species are reported.

In comparison with the pre-Ensis era diets of both fish andseabirds, the introduction of Ensis clearly had an impact on thecurrent diet. For eiders and possibly also for common scoters thedominance of Spisula in the diet in the coastal zone has been takenover by Ensis. Especially for the flatfish species the large share ofpolychaetes, crustaceans and fish has partly been replaced by Ensis.It is also clear that Ensis both replaced other prey species thatdeclined such as Spisula (for sea birds) and added to the existingprey field (for fish).We have shown that it can be an important preyfor fish and birds. Even though it is not preferred prey it can provide

an alternative to native species such as trough shells, blue musselsand cockles. However because of its behaviour the availability asa food source is not predictable. So far the role of Ensis in the coastalecosystem has been descriptive in nature. To evaluate the actualrole of Ensis in the ecosystem there is a need to quantify thecontribution of Ensis to the total benthic biomass production. Forthat aim we need more precise biomass estimates and productionrates. Recently a sampling device capable of distracting completeEnsis (instead of only the top ends) has been developed that wouldenable a more precise biomass estimate (Craeymeersch et al.,2007). Recent data in the Wadden Sea point to an underestima-tion of the absolute biomass estimate of at least a factor 7 (Jansen,unpublished data IMARES).

Acknowledgements

We would like to thank the crew of the Goeree 58 and theLuctor. Several colleagues from IMARES provided assistance ineither logistics, data collection or analysis of the fish stomachsampling: Dirk den Uijl, Ronald Bol, André Dijkman Dulkes, SaschaFassler, Peter Groot, Kees Groeneveld, Sarah Kraak, Remment terHofstede, Betty van Os-Koomen, Simon Rijs, Gerrit Rink, Marcel deVries, Jan van Willigen and Jakob Asjes. Peter Spannenburg, MarcelKoen and Sandra van Loon helped sorting through the eider faecessamples. Erik Binnedijk analysed fish stomachs. The data on whichthis paper is based originate from different studies that werefinanced by the Centre for Water Management (formerly NationalInstitute for Coastal and Marine Management), the NationalClimate Change and Spatial Planning program BSIK “Klimaat voorRuimte”. The study was supported by the strategic research pro-gramme "Sustainable spatial development of ecosystems, land-scapes, seas and regions" funded by the Dutch Ministry ofAgriculture, Nature Conservation and Food Quality. This studycontributes to the project EnSIS, financed by the Belgian SciencePolicy (contract SD/NS/09A http://www.belspo.be/)

References

Alexander, R., 1979. The Invertebrates. Cambridge University Press, Cambridge.Amara, R., Laffargue, P., Dewarumez, J.M., Maryniak, C., Lagardere, F., Luczac, C.,

2001. Feeding ecology and growth of 0-group flatfish (sole, dab and plaice) ona nursery ground (Southern Bight of the North Sea). Journal of Fish Biology 58,788e803.

Amara, R., Mahe, K., LePape, O., Desroy, N., 2004. Growth, feeding and distributionof the solenette Buglossidium luteum with particular reference to its habitatpreference. Journal of Sea Research 51, 211e217.

Armonies, W., Reise, K., 1998. On the population development of the introducedrazor clam Ensis americanus near the island of Sylt (North Sea). HelgolanderMeeresuntersuchungen 52, 291e300.

Arts, F.A., Berrevoets, C.M., 2007. Midwintertellingen van zee-eenden in de Wad-denzee en Nederlandse kustwateren, januari 2007. Report No. 2007.010, RIKZ,DPM. Culemborg en RIKZ, Middelburg.

Aulert, C., Sylvand, B., 1997. Les Macreusess noires (Melanitta nigra) et brunes(Melanitta fusca) hivernant au large de Côtes dus Calvados: relation entre leregime alimentaire et les peuplements macrozoobenthiques marins littoraux.Ecologie 28, 107e117.

Backeljau, T., 1986. Lijst van de recente mariene mollusken van België Report No. 29.Studiedocumenten van het KBIN.

Baptist, M.J., Leopold, M.F., 2009. The effects of shoreface nourishments on Spisulaand scoters in The Netherlands. Marine Environmental Research 68, 1e11.

Basimi, R.A., Grove, D.J., 1985. Estimates of daily food intake by inshore populationof Pleuronectes platessa L. off eastern Anglesey, North Wales. Journal of FishBiology 27, 505e520.

Bauer, K.M., Glutz von Blotzheim, U.N., 1969. Handbuch der Vögel Mitteleuropas 3.Akademischer Verlagsgesellschaft, Frankfurt am Main.

Beukema, J.J., 1974. The efficiency of the Van Veen grab compared with the Reineckbox sampler. J. Cons. perm. int. Explor. Mer 35, 319e327.

Beukema, J.J., Dekker, R., Jansen, J.M., 2009. Some like it cold: populations of thetellinid bivalve Macoma balthica (L.) suffer in various ways from a warmingclimate. Marine Ecology-Progress Series 384, 135e145.

Beyst, B., Cattrijsse, A., Mees, J., 1999. Feeding ecology of juvenile flatfishes of thesurf zone of a sandy beach. Journal of Fish Biology 55, 1171e1186.

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128 127

Blegvad, H., 1916. On the food of fish in the Danish waters within the Skaw.Beretning fra den Danske Biologiske Station 24, 19e72.

Braber, L., Groot, S.J.D., 1973. The food of five flatfish species (Pleuronectiformes) inthe Southern North Sea. Netherlands Journal of Sea Research 6, 163e172.

Cadée, G.C., 2001. Zilvermeeuwen profiteren van sterfte van Ensis directus. Zeepaard61, 133e140.

Cadée, G.C., 2006. Eidereenden op Ensis directus dieet. Spirula 351, 86e87.Cadée, G.C., Cadée-Coenen, J., 1994. Hoe zilvermeeuwen Amerikaanse Zwaard-

scheden (Ensis directus) vangen. Correspondentieblad Nederlandse Mala-cologische Vereniging 278, 64e67.

Cadée, G.C., Cadée-Coenen, J., Witte, J.Y., 1994. Massale sterfte van Ensis directus opde Schanserwaard en elders blijft raadselachtig. Correspondentieblad Neder-landse Malacologische Vereniging 279, 86e93.

Camphuysen, C.J., 1989. Beached Bird Surveys in the Netherlands 1915e1988;Seabird mortality in the Southern North Sea since the early days of oil pollution.Vogelbescherming, werkgroep Noordzee, Amsterdam.

Camphuysen, C.J., Berrevoets, C.M., Cremers, H., Dekinga, A., Dekker, R., Ens, B.J., vander Have, T.M., Kats, R.K.H., Kuiken, T., Leopold, M.F., van der Meer, J.,Piersma, T., 2002. Mass mortality of common eiders (Somateria mollissima) inthe Dutch Wadden Sea, winter 1999/2000: starvation in a commerciallyexploited wetland of international importance. Biological Conservation 106,303e317.

Carter, C.G., Grove, D.J., Carter, D.M., 1991. Trophic resource partitioning betweentwo coexisting flatfish species off the North coast of Angelsey, North Wales.Netherlands Journal of Sea Research 27, 325e335.

Consortium, V.A.S., 2008. Amerikaanse zwaardschede - Ensis directus. Niet-inheemse soorten van het Belgisch deel van de Noordzee en aanpalendeestuaria. VLIZ Information Sheets 7.

von Cosel, R., 1982. Die amerikanische Schwertmuschel Ensis directus (Conrad) inder Deutschen Bucht. Senckenbergiana Maritima 14, 147e173.

Craeymeersch, J., 2007. Long term series in the Netherlands. In: Report of theBenthos Ecology Working Group (BEWG), 23e27 May 2007. BEWG, Wilhelm-shaven, Germany, pp. 39e43. ICES CM 2007/MHC:10.

Craeymeersch, J.A., van Stralen, M.R., Wijsman, J.W., Kesteloo, J., Perdon, J.,de Mesel, I., 2007. Ontwikkeling van een monstertuig voor bestandsopnamesvan mesheften Report No. C084/07. IMARES.

Craeymeersch, J.A., Wijsman, J.W.N., 2006. Ruimtelijke verschillen en temporelefluctuaties in het voorkomen van een aantal schelpdieren in de VoordeltaReport No. C013/06.

Creutzberg, F., Witte, J.I., 1989. An attempt to estimate the predatory pressureexerted by the lesser weever, Trachinus vipera Cuvier, in the southern north-Sea.Journal of Fish Biology 34, 429e449.

Daan, R., Mulder, M., 2006. The macrobenthic fauna in the Dutch sector of the NorthSea in 2005 and a comparison with previous data. Report No. 2006e3. RoyalNetherlands Institute for Sea Research.

Daisie, 2009. Handbook of Alien Species in Europe. Springer, Dordrecht.Degraer, S., Vincx, M., Meire, P., Offringa, H., 1999. The macrozoobenthos of an

important wintering area of the Common Scoter Melanitta nigra. Journal of theMarine Biological Association of the United Kingdom 79, 243e251.

Durinck, J., Christensen, K.D., Skov, H., Danielsen, F., 1990. Diet of common scoterMelanitta nigra and velvet scoter M. fusca wintering in the north sea. OrnisFennica 70, 215e218.

Durinck, J., Christensen, K.D., Skov, H., Danielsen, F., 1993. Diet of the common scoterMelanitta nigra and velvet scoter Melanitta fusca wintering in the North Sea.Ornis Fenn 70, 215e218.

Ens, B.J., Borgsteede, F.H.M., Camphuysen, C.J., Dorrestein, G.M., Kats, R.K.H.,Leopold, M.F., 2002. Eidereendensterfte in de winter 2001/2002 Report No. 521.Alterra.

Ens, B.J., Kats, R.K.H., Camphuysen, C.J., 2006. Waarom zijn Eiders niet massaalgestorven in de winter van 2005/2006? Limosa 79, 95e106.

Essink, K., Detmann, C., Farke,H., Laursen, K., Lüerssen,G.,Marencic,H.,Wiersinga,W.,2005. Wadden Sea Status Report 2004. Wadden Sea Ecosysteem 19, 359.

Evert, U., 2004. Nahrungsökologie von Meeresenten in der Pommerschen Bucht.Christian-Albrechts-Universität.

Fox, A.D., 2003. Diet and habitat use of scotersMelanitta in the Western Palearctic ea brief overview. Wildfowl 54, 163e182.

Frengen, O., Thingstad, P.G., 2002. Mass occurrence of sandeels (Ammodytes spp.)causing aggregations of diving ducks. Fauna Norvegica 22, 32e36.

Freudendahl, A.S., Jensen, K.T., 2006. Population Dynamics of the American RazorClam (Ensis americanus) at Horns Rev. Report by University of Aarhus, Depart-ment of Marine Ecology for Energi2.

Gollasch, S., Minchin, D., Rosenthal, H., Voigt, M., 1999. Exotics across the ocean.case histories on introduced species: their general biology, distribution, rangeexpansion and impact. Logos Verlag, Berlin.

Goudswaard, P.C., Kesterloo, J.J., Perdon, K.J., Jansen, J.M., 2008. Mesheften (Ensisdirectus), halfgeknotte strandschelpen (Spisula subtruncata), kokkels (Cera-stoderma edule) en otterschelpen (Lutraria lutraria) in de Nederlandse kust-wateren in 2008 Report No. C069/08. IMARES.

Groot, S.J.D., 1964. Diurnal activity and feeding habits of plaice. Rapp P-v Reun ConsPerm Int Explor Mer 155, 48e51.

Gürs, I., Gürs, K., Paustian, J., Wiese, V., 1993. Ensis directus (Conrad 1843) zumersten Mal in der südwestlichen Ostsee (Bivalvia: Solenidae). Schriften aus demMalakozoologischen Haus der Natur 79.

Gwyther, D., Grove, D.J., 1981. Gastric emptying in Limanda limanda (L.) and thereturn of appetite. Journal of Fish Biology 18, 245e259.

Hagmeier, A., 1930. Eine fluktuation von Mactra (Spisula) subtruncata da Costa ander ostfriesischen kuste. Ber Dt Wiss Komm Meeresfors 5, 126e155.

Hampel, H., Cattrijsse, A., Elliott, M., 2005. Feeding habits of young predatory fishesin marsh creeks situated along the salinity gradient of the Schelde estuary,Belgium and The Netherlands. Helgoland Marine Research 59, 151e162.

Himmelman, J.H., Dutila, C., Gaymerb, C.F., 2005. Foraging behavior and activitybudgets of sea stars on a subtidal sediment bottom community. Journal ofExperimental Marine Biology and Ecology 322, 153e165.

Hoines, A.S., Bergstad, O.A., 2002. Food partitioning by flatfishes on a herringspawning ground. Sarsia 87, 19e34.

Holtmann, S.E., Groenewold, A., Schrader, K.H.M., Asjes, J., Craeymeersch, J.A.,Duineveld, G.C.A., Bostelen, A.J.v., Meer, J.v.d., 1996. Atlas of the zoobenthos ofthe Dutch Continental Shelf. Ministry of Transport, Public Works and WaterManagement, North sea directorate, Rijswijk.

Howlett, D.J., 1990. The arrival in Britain of Ensis americanus. Conchologist’sNewsletter 114, 301e302.

Hughes, B., Stewart, B., Brown, M.J., Hearn, R.D., 1997. Studies of Common ScoterMelanitta Nigra Nigra killed during the Sea Empress Oil Spill. Countryside forWales, Bangor, Wales.

ICES, 2005. Report of the Working Group on Seabird Ecology (WGSE), 29 March e

1 April 2005. Texel, The Netherlands. Report No. 2005/G:07, ICESJones, N.S., 1952. The bottom fauna and the food of flatfish off the Cumberland

coast. Journal of Animal Ecology 21, 182e205.Kaiser, M.J., Galanidi, M., Showler, D.A., Elliott, A.J., Caldow, R.W.G., Rees, E.I.S.,

Stillman, R.A., Sutherland, W.J., 2006. Distribution and behaviour of CommonScoterMelanitta nigra relative to prey resources and environmental parameters.Ibis 148, 110e128.

Kats, R., 2007. Common Eiders Somateria mollisima in the Netherlands: The rise andfall of breeding and wintering populations in relation to the stocks of shellfish.PhD, University of Groningen.

Kirchhoff, K., 1979. Nahrungsölologische Untersuchungen an benthosfressendenEnten in der Hohwachter Bucht. MSc, Universität Kiel.

Klimpel, S., Seehagen, A., Palm, H.W., 2003. Metazoan parasites and feedingbehaviour of four small-sized fish species from the central North Sea. Parasi-tology Research 91, 290e297.

Leitão, R., Martinho, F., Neto, J.M., Cabral, H.N., Marques, J.C., Pardal, M.A., 2006.Feeding ecology, population structure and distribution of Pomatoschistusmicrops (Krøyer, 1838) and Pomatoschistus minutus (Pallas, 1770) in a temperateestuary, Portugal. Estuarine. Coastal and Shelf Science 66, 231e239.

Leopold, M.F., 1993. Spisula’s, zeeëenden en kokkelvissers: een nieuw milieupro-bleem op de Noordzee. Sula 7, 24e28.

Leopold, M.F., 1996. Spisula subtruncata als voedselbron voor zee-eenden inNederland Report No. 96e102.

Leopold, M.F., 2002. Eiders Somateria mollissima scavenging behind a lugworm boat.Journal of Sea Research 47, 75e82.

Leopold, M.F., Baptist, H.J.M., Wolf, P.A., Offringa, H., 1995. De Zwarte ZeeëendMelanitta nigra in Nederland. Limosa 68, 49e64.

Leopold, M.F., Kats, R.K.H., Ens, B.J., 2001. Diet (preferences) of common eidersSomateria mollissima. Wadden Sea Newsletter 1, 25e31. 2000.

Leopold, M.F., Spannenburg, P.C., Verdaat, H.J.P., Kats, R.K.H., 2007. Identificationand size estimation of Spisula subtruncata and Ensis americanus from shellfragments in stomachs and faeces of Common Eiders Somateria mollissima andCommon Scoters Melanitta nigra. In: Kats, R.K.H. (Ed.), Common EidersSomateria Mollissima in the Netherlands the Rise and Fall of Breedingand Wintering Populations in Relation to the Stocks of Shellfish PhD UnivGroningen, pp. 63e85.

Leopold, M.F., Wolf, P., 2003. Zee-eenden eten ook Ensis. Nieuwsbrief NederlandseZeevogelgroep 4, 5.

Luczak, C., Dewarumez, J.M., Essink, K., 1993. 1st record of the American jack-knifeclam Ensis directus on the French coast of the North Sea. Journal of the MarineBiological Association of the UK 73, 233e235.

Marshall, S., Elliott, M., 1997. A comparison of univariate and multivariate numericaland graphical techniques for determining inter- and intraspecific feedingrelationships in estuarine fish. Journal of Fish Biology 51, 526e545.

McDermott, J.J., 1976. Predation of the razor clam Ensis directus by the nemerteanworm Cerebratulus lacteus Chesapeake Science 17, 299e301.

Meissner, J., Bräger, S., 1990. The feeding ecology of wintering eiders Somateriamollissima and common scoters Melanitta nigra on the Baltic Sea coast ofSchleswig-Holstein. FRG. Wader Study Group Bulletin 58, 10e12.

Mittelbach, G.G., 2002. Fish foraging and habitat choice: a theoretical perspective.In: Hart, P.J.B., Reynolds, J.D. (Eds.), Handbook of Fish and Fisheries. Blackwellscience, Oxford, pp. 251e265.

Muir, S.D., 2003. The Biology of Razor Clams (Ensis spp.) and Their Emergent Fisheryon the West Coast of Scotland. University of London.

Muus, B.J., Nielsen, J.G., Dahlstrom, P., Nystrom, B.O., 1999. Zeevissen van Noord-en West-Europa, Schuyt & Co Uitgevers en Importeurs. Haarlem.

Nehls, G., Ketzenberg, C., 2002. Do Common Eiders Somateria mollissima exhausttheir food resources? A study on natural mussel Mytilus edulis beds. Dan RevGame Biol 16, 47e61.

Noordhuis, R., Spaans, A.L., 1992. Interspecific competition for food between herringLarus argentatus and lesser black-backed gulls L. fuscus in the Dutch WaddenSea area. Ardea 80, 115e132.

Offringa, H., 1991. Verspreiding en Voedseloecologie van de Zwarte Zeeëend (Mel-anitta nigra) in Nederland Report No. 1991e13. Netherlands Intsitute for SeaResearch, NIOZ, (Texel).

I. Tulp et al. / Estuarine, Coastal and Shelf Science 90 (2010) 116e128128

Ortega-Salas, A.A., 1980. Seasonal changes in the common dab, Limanda limanda (L.)in Isle of Man waters. Journal of Fish Biology 16, 75e82.

Palmer, D., 2003. The introduced razor fish Ensis directus in the Wash and NorthNorfolk. Shellfish New 16, 13e15.

Pentreath, R.J., Lovett, M.B., 1978. Transuranic nuclides in plaice (Pleuronectesplatessa) from the North-eastern Irish Sea. Marine Biology 48, 19e26.

Perdon, K.J., Goudswaard, P.C., 2007. Mesheften (Ensis directus), halfgeknottestrandschelpen (Spisula subtruncata) en kokkels (Cerastoderma edule) inde Nederlandse kustwateren in 2007 Report No. C087/07. IMARES.

Petersen, I.K., Skov, H., 2007. Changes in bird habitat utilisation around the HornsRev I Offshore Wind Farm, with particular emphasis on Common Scoter. Dan-marks Miljøundersøgelser. Report request commissioned by Vattenfall A/S.

Pihl, L., Rosenberg, R., 1984. Food selection and consumption of the shrimp Crangoncrangon in some shallow marine areas in western Sweden. Marine EcologyProgress Series 15, 159e168.

Rijnsdorp, A.D., Vingerhoed, B., 2001. Feeding of plaice Pleuronectes platessa L. andsole Solea solea L. in relation to the effects of bottom trawling. Journal of SeaResearch 45, 219e229.

Schneider, D., 1982. Escape response of an infaunal clam Ensis directus Conrad 1843,to a predatory snail, Policines duplicatus Say 1822. Veliger 24, 371e372.

Schricke, V., 1993. La baie du Mont-Saint-Michel, premiere zone de mue en Francepour la macreuse noire (Melanitta nigra). Alauda 61, 35e38.

Severijns, N., 2001. De verspreiding van Ensis directus (Conrad, 1843) in Europa, 23jaar na de introductie: de opmerkelijke opmars van een immigrant.De Strandvlo 21, 123e146.

Skov, H., Durinck, J., Erichsen, A., Kloster, R.M., Møhlenberg, F., Leonard, S.B., 2008a.Horns Rev II offshore wind farm food basis for common scoter. Baseline studies2007e2008. Orbicon/DHI/Marine Observers Report commissioned by DONGenergy.

Skov, H., Piper, W., Leonhard, S.B., 2008b. Horns Rev II wind farm. Monitoring ofwaterbirds. Baseline studies 2007e08. Orbicon A/S & DHI Report request.Commissionde by DONG Energy.

Smaal, A., Craeymeeersch, J., Kamermans, P., van Stralen, M., 2001. Is food shortagethe cause of Eider Duck mortality? Shellfish and crab abundance in the DutchWadden Sea 1994e1999. Wadden Sea Newsletter 1, 35e38.

Smaal, A., van Stralen, M., Craeymeersch, J., 2005. Does the introduction of thePacific Oyster Crassostrea gigas lead to species shifts in the Wadden Sea?In: Dame, R.F., Olenin, S. (Eds.), The comparative roles of suspension-feeders in ecosystems. Kluwer Academic Publishers, Dordrecht, theNetherlands, pp. 277e289.

Steenbergen, J., Escaravage, V., 2006. Baseline study MEP-MV2, Lot 2: bod-emdieren. Eindrapportage Campagnes 2004e2005 Report No. C053/06,IMARES/NIOO.

Swennen, C., 1976. Populatiestructuur en voedsel van de Eidereend Somateriamollissima in de Nederlandse Waddenzee. Ardea 64, 311e371.

Swennen C., 1991. Ecology and population dynamics of the common eider in theDutch Wadden Sea. PhD, Rijksuniversiteit Groningen.

Swennen, C., Duiven, P., 1989. Eidereend Somateria mollisima gestikt in amerikaansezwaardschede Ensis directus. Limosa 62, 153e154.

Swennen, C., Leopold, M.F., Stock, M., 1985. Notes on growth and behaviour of theAmerican razor clam Ensis directus in the Wadden Sea and the predation on itby birds. Helgolander Meeresuntersuchungen 39, 255e261.

Tempelman, D., van der Wal, J.T., van Moorsel, G., de Kluijver, M., Lewis, W.,Verduin, E., Vanagt, T., 2009. The macrobenthic fauna in the dutch sector of theNorth Sea in 2008 and a comparison with previous data Report No. 202462e3.Grontmij / AquaSense.

Thingstad, P.G., Hokstad, S., Frengen, O., 2000. Nye opplysninger om aerfuglensSomateria mollissima naeringsbiologi (Some new notes considering thecommon eider’s Somateria mollissima foraging biology). Fauna-Oslo 53,66e71.

Todd, R.A., 1915. Report on the Food of Plaice. In: Fisheries Investigation Series 2 2 3.Van der Veer, H.W., Creutzberg, F., Dapper, R., Duineveld, G.C.A., Fonds, M.,

Kuipers, B.R., Vannoort, G.J., Witte, J.I.J., 1990. On the ecology of the dragonetCallionymus lyra L. in the southern North-Sea. Netherlands Journal of SeaResearch 26, 139e150.

Vasconcelos, R., Prista, N., Cabral, H., Costa, M.J., 2004. Feeding ecology of the lesserweever, Echiichthys vipera (Cuvier, 1829), on the western coast of Portugal.Journal of Applied Ichthyology 20, 211e216.

Wolf, P.A., Meininger, P.L., 2004. Zeeën van zee-eenden bij de Brouwersdam.Nieuwsbrief Nederlandse Zeevogelgroep 5, 1e2.