Stomach content analysis indicates multi decadal trophic stability in a temperate coastal fish food web, western Dutch Wadden Sea

Abstract

Information about stomach content composition of fish species of a temperate coastal fish community (western Dutch Wadden Sea) over the period 1930 – 2019 was analysed to reconstruct long-term trends in trophic position of individual species. Stomach data were not evenly distributed but clustered both with respect to years as well as fish species. For 18 fish species, all being omnivorous and belonging to different functional groups (pelagic, benthopelagic, demersal) and guilds [(near)-resident, juvenile marine migrants, marine seasonal visitiors], prey consumption and trophic position over time could be analysed. Prey occurrence in the stomachs of different fish species showed variability over time, most likely due to fluctuations in prey abundance, but without a trend. For all species, individual fish showed variablity in trophic position in the order of 1 unit or even more both within and between years. However, in all 18 species, no significant trend in mean trophic position over time could be found, despite the serious anthropogenic stress (pollution,eutrophication events, climate change) and the decrease in fish abundance in the area during the last 50 years. The present study does not indicate any changes in trophic position of individual species in the western Dutch Wadden Sea over the last 80 years. At the community level, trophic structure varies due to interannual fluctuations in species composition and year-to year fluctuations in the relative abundance of the various fish species. At the ecosystem level the trophic role of the fish community has been degraded due to the decrease in total fish biomass in the area.

Full text

Stomach content analysis indicates multi decadal trophic stability in a
temperate coastal sh food web, western Dutch Wadden Sea
Suzanne S.H. Poiesz
a,b,*
, Johannes IJ. Witte
a
, Henk W. van der Veer
a
a
NIOZ Royal Netherlands Institute for Sea Research, Department of Coastal Systems, P.O. Box 59, 1790, AB, Den Burg, Texel, the Netherlands
b
Faculty of Science and Engineering, Groningen Institute of Evolutionary Life Sciences, University of Groningen, P.O. Box 11103, 9700, CC, Groningen, the Netherlands
ARTICLE INFO
Keywords:
Coastal sh community
Wadden Sea
Food web structure
Stomach content analysis
Long term trends
Stability
ABSTRACT
Information about stomach content composition of sh species of a temperate coastal sh community (western
Dutch Wadden Sea) over the period 1930–2019 was analysed to reconstruct long-term trends in trophic position
of individual species. Stomach data were not evenly distributed but clustered both with respect to years as well as
sh species. For 18 sh species, all being omnivorous and belonging to different functional groups (pelagic,
benthopelagic, demersal) and guilds [(near)-resident, juvenile marine migrants, marine seasonal visitiors], prey
consumption and trophic position over time could be analysed. Prey occurrence in the stomachs of different sh
species showed variability over time, most likely due to uctuations in prey abundance, but without a trend. For
all species, individual sh showed variablity in trophic position in the order of 1 unit or even more both within
and between years. However, in all 18 species, no signicant trend in mean trophic position over time could be
found, despite the serious anthropogenic stress (pollution, eutrophication events, climate change) and the
decrease in sh abundance in the area during the last 50 years. The present study does not indicate any changes
in trophic position of individual species in the western Dutch Wadden Sea over the last 80 years. At the com-
munity level, trophic structure varies due to interannual uctuations in species composition and year-to year
uctuations in the relative abundance of the various sh species. At the ecosystem level the trophic role of the
sh community has been degraded due to the decrease in total sh biomass in the area.
1. Introduction
Coastal systems provide a large variety of ecosystem goods and
services (see Barbier, 2017;Liu et al., 2021) and consequently, their
ecosystem value is high (Liu et al., 2021). Coastal systems are known as
important foraging grounds for a variety of sh, bird and marine
mammal species (e.g. Goodall, 1983;Beck et al., 2001), and in these
areas sh harvesting has been an important marine ecosystem good for
centuries. However, due to human shing and hunting, coastal ecosys-
tems have also been under pervasive human disturbance for centuries
(Jackson et al., 2001;Lotze, 2007). For the future, anthropogenic
pressure in these areas is expected to continue especially due to the
combined pressure of overshing and habitat destruction, pollution and
climate change (Bijma et al., 2013;European Marine Board, 2013).
Predicting the consequences of the still ongoing threats on the future
productivity of coastal areas requires (among other factors) insight into
the food web structure of these systems. The fact that coastal ecosystems
have been under pervasive human disturbance already for centuries
makes it difcult to get insight in their ‘original pristine state’and to
assess the impact of human disturbance over time. First of all, going back
in time, information about ecosystem status becomes more and more
qualitative and anecdotic. Furthermore, our perspective about the past
also suffers from “the shifting baseline phenomenon”: ecosystem
changes are considered relative to the situation the evaluator can
remember and therefore the baseline shifts with each generation (Pauly,
1995;Zeller et al., 2005). This stresses the need for long time series of
reliable information on ecosystem structure, preferably covering mul-
tiple observer generations. In this study, we focus on the sh food web in
the international Wadden Sea, one of the largest estuarine areas in the
world, bordering the Dutch, German and Danish North Sea coast. The
area is an important resting and fuelling area for birds and nursery area
for various (non)commercial sh species (Zijlstra, 1972;Wolff, 1983).
From archaeological, historical, sheries, and ecological records, it is
clear that the Wadden Sea have been under pervasive disturbance for
* Corresponding author. NIOZ Royal Netherlands Institute for Sea Research, Department of Coastal Systems, P.O. Box 59, 1790, AB, Den Burg, Texel, the
Netherlands.
E-mail address: suzannepoiesz@gmail.com (S.S.H. Poiesz).
Contents lists available at ScienceDirect
Estuarine, Coastal and Shelf Science
journal homepage: www.elsevier.com/locate/ecss
https://doi.org/10.1016/j.ecss.2024.108912
Received 21 December 2023; Received in revised form 5 August 2024; Accepted 6 August 2024

centuries already (Lotze, 2005,2007).
Quality status reports about the ecology of the Wadden Sea has been
produced periodically since 1999 (https://qsr.waddensea-worldher
itage.org), with various ecological monitoring series in the western
part of the area on phytoplankton (Philippart et al., 2007;Jacobs et al.,
2020), macrozoobenthos (Beukema and Dekker, 2020) and sh (Tulp
et al., 2008;van der Veer et al., 2015) providing reconstructions over the
last 60 years. From the 1970’s, no changes in sh biodiversity was
found. However, sh abundance of both pelagic and demersal species
showed a 10-fold decrease in catches from 1980s onwards (Tulp et al.,
2008;van der Veer et al., 2015). At present, various stomach content
studies show that most Wadden Sea sh species are omnivorous, feeding
on multiple prey items with a pivotal position of a few key prey species
(Kellnreiter et al., 2012;Whitehouse et al., 2017;Poiesz et al., 2020,
2023). Stable isotope analyses indicates that the sh food web in this
area consists of a spatially stable structure with various trophic levels
(Poiesz et al., 2021a,2023). To what extend the decrease in sh abun-
dance in the 1980s has caused a shift in prey selection and therefore a
temporal change in their trophic positions by the omnivorous predatory
sh species, is unclear.
Ecological information about the sh food web before the 1970’s is
mostly qualitative and anecdotic, except for stomach content informa-
tion of the sh fauna as a by-product of the long-lasting human shing in
the area. Stomach content data is an important source of information
(Kellnreiter et al., 2012;Whitehouse et al., 2017;Poiesz et al., 2020),
despite the fact that it is labour intensive, requires taxonomic expertise
and only offers a small temporal snapshot of recently consumed prey
items and might thus be sensitive to sampling design (Poiesz et al.,
2023). In the absence of stable isotope information, stomach content
information can be used to derive trophic structure of the sh fauna and
its predator−prey interactions (Hynes, 1950;Baker et al., 2014;Poiesz
et al., 2020, 2021). A recent comparison of stomach content information
and stable isotopes of sh in the western Wadden Sea, illustrated that
both resulted in a similar picture of the trophic structure of the sh fauna
(Poiesz et al., 2023).
For long-term stomach content time series, standardised methods of
sampling and analysis are important (see for overview Hyslop, 1980;
Buckland et al., 2017;Amundsen and S´
anchez-Hern´
andez, 2019).
However, time series often suffer from limitations due to differences
over time in sampling strategy, sampling intensity and/or in detail and
methods of the analyses. In case enough data are present General Ad-
ditive Models (GAMs) can be applied to visualise and analyse trends in
stomach content over time (Hastie and Tibshirani, 1995;Kvaarik et al.,
2019;Kordubel et al., 2024). In this study we focus on unpublished
records of sh stomach content data, mainly from the western part of
Wadden Sea form the NIOZ archive, dating back to the 1930’s (de Vooys
et al., 1991,1993). For all species, missing observations and/or gaps in
the time series occurred. Furthermore, not all records contained infor-
mation about number of prey found, prey condition, and prey weight.
Therefore, Buckland et al. (2017) was followed and the simple pre-
sence/absence and frequency of occurrence approach was taken, since it
is not affected by prey condition and hence provides a rapid, unam-
biguous and reliable account of diet composition and prey trophic
position.
This NIOZ archive stomach content information is used to analyse
uctuations in predator-prey relationships and in the trophic position of
individual sh species over the last century with the aim to get insight in
the temporal variability of the Wadden Sea sh food web. The present
trophic position of the various sh species (Poiesz et al., 2020,2021a)
will be used as reference to test whether shifts in trophic position of
individual sh species has occurred over time. The stomach content data
are available for the time span 1930 –present and thus the time series
period covers more than a single scientic career. As such, the results of
this study can also be used to correct for the “shifting baseline
phenomenon”.
2. Material and methods
2.1. Data collection
From 1930 onwards the Royal Netherlands Institute for Sea Research
(NIOZ) registered observations and landings of sh and invertebrate
species from the western Wadden Sea and nearby Dutch coastal waters
(Fig. 1). Most information originated from NIOZ cruises and sh
collected during NIOZ courses. In addition landings of rare sh species
from shermen were recorded. All individual sh were identied and
information about species and stomach content was recorded. From the
beginning, data collection, section and stomach content analysis was
done by specialised NIOZ personnel only. A more detailed description of
the NIOZ archive can be found in de Vooys et al. (1991,1993).
From the 1980’s onwards, stomach content data were collected from
a long-term monitoring programme of the sh fauna with a passive sh
trap near the entrance of the Wadden Sea in spring and autumn (Poiesz
et al., 2020). Until 2010, all sh caught on Fridays were taken to the
laboratory and sorted within an hour, identied up to species level,
counted and length measured. From 2017 onwards, a maximum of three
individuals per species per week were selected and stored at −20 ◦C for
further analysis. Within a month, individuals were defrosted, and
stomach content was taken out and analysed in a Petri dish under a
binocular (20x). Of each individual sh, total stomach content was
determined (wet mass; g) and subsequently, prey items were identied
up to species level or sometimes, up to a higher classication (class,
order, genus). If possible, total length of the prey was measured (mm).
Incomplete specimens, often from species that were eaten in pieces, such
as Alitta virens or Ensis leei, or from species that were in parts, such as the
Crangon crangon, were counted only by the number of ‘heads’.
Fig. 1. Sampling locations from the North Sea coast and Wadden Sea. Size of
black dots indicate the contribution (%) in the amount of individuals caught for
each sampling location.
S.S.H. Poiesz et al.

Taxonomic identication was based on an internal reference collection
and Hayward and Ryland (2017) for polychaetes, bivalves and crabs and
Wheeler (1978) for sh species. For more details see van der Veer et al.
(2015) and Poiesz et al. (2020).
2.2. Data processing
All sh records were checked for species name and, if necessary,
updated according to WoRMS (http://www.marinespecies.org). Next,
sh species were assigned in line with previous work (van der Veer et al.,
2015;Poiesz et al., 2020) into: pelagic (occurring mainly in the water
column between 0 and 200 m, not feeding on benthic organisms);
benthopelagic (living and/or feeding on or near the bottom, as well as in
midwater, between 0 and 200 m) and benthic (living and/or feeding on
the bottom) according to FishBase (Froese and Pauly, 2021). Species
were also classied according to their use of the area into near-resident
and resident species, marine juvenile migrants and seasonal visitors
based on Zijlstra (1983).Dicentrarchus labrax (bass) was considered to
have become a resident species in the Wadden Sea in recent time, due to
the presence of small juveniles and adults almost year-round (Cardoso
et al., 2015).
All prey items found in the stomachs of the sh were checked and
scientic name, family, order and class were updated according to
WoRMS (http://www.marinespecies.org)
Level of taxonomic identication of prey items was variable over the
years, often Class level from 1930 to 1980 versus species level from 1980
onwards. For all prey Classes, Families and species found, trophic po-
sition was taken from FishBase (Froese and Pauly, 2021).
Per year, for each sh species, the mean percentage of occurrence (=
number of stomachs containing a prey species divided by total number
of stomachs with content examined) of each class of prey items was
determined as a measure of diet composition following Baker et al.
(2014). Furthermore, the trophic position of each individual sh j(TP
j
)
was calculated from the stomach content as the mean trophic position of
the different prey species kfound in a stomach, according to:
TPj=1+∑TPk
k[1]
where:
TPj: being the calculated trophic position of the individual sh j;
TPk: the trophic position of prey species kin the stomach of sh j.
k: the number of different prey species in the stomach of sh j.
The bias introduced by not correcting for differences in mass of the
various prey items in the stomachs is small (Poiesz et al., 2021a). Next,
for each sh species, the mean trophic position per year was calculated.
2.3. Data analysis
The impact of level of detail of prey identication on estimated
trophic level of stomach content was analysed for the 2010–2019 data
(Poiesz et al., 2020). Estimated trophic levels of the stomach contents
based on trophic values of identied prey species were compared with
estimates after a rerun with Class values instead of species values.
In all species, missing observations and gaps in the time series
occurred. For sh species, with minimum 15 years of observation with at
least 5 stomach contents analysed were present to apply General Addi-
tive Models (GAMs) to visualise and analyse trends over time (Hastie
and Tibshirani, 1995). For these sh species i, trends over time in the
most common prey items (PO
i
) and in mean trophic position (TP
i
) were
analysed by tting GAMs using locally weighted least squares regression
(LOESS), an identity link function and the Gaussian error distribution
according to:
Fig. 2. NIOZ archive.
A: Number of stomachs contents analysed over the years 1932–2019.
B: Number of Bivalve species identied in the Wadden Sea sh stomachs.
C: Number of Malacostraca species identied in the Wadden Sea sh stomachs.
D: Number of Pisces species identied in the Wadden Sea sh stomachs.
E: Number of Polychaete species identied in the Wadden Sea sh stomachs.
S.S.H. Poiesz et al.

Table 1
Overview of prey items found in the stomachs of the various sh species of the NIOZ archive between 1931 and 2019, together with trophic position according to
FishBase (www.shbase.com).
Class Order Family Scientic name Common name Trophic position (−)
Eggs (Crab, shrimp, sh) 1,00
Asteroidea Forcipulatida Asteriidae Asteriidae Sea stars 2,00
Asteroidea Spatangoida Loveniidae Echinocardium Sea urchins 2,00
Bivalvia Adapedonta Bivalves 2,10
Bivalvia Adapedonta Myridae Mya spec Solf shell clams
Bivalvia Adapedonta Pharidae Ensis Razor clams 2,10
Bivalvia Adapedonta Pharidae Ensis leei Atlantic jackknife clam 2,10
Bivalvia Mytilida Mytilidae Mytilus edulis Blue mussel 2,10
Bivalvia Cardiida Tellinidae Limecola balthica Baltic macoma 2,10
Caenogastropoda Littorinimorpha Hydrobiidae Peringia ulvae Laver spire shell 2,40
Chlorophyta Algae Algae Algae Algae 1,00
Chlorophyta Cladophorales Cladophoraceae Chaetomorpha melagonium Chaetomorpha melagonium 1,00
Chlorophyta Ulvales Ulvaceae Ulva lactuca Sea lettuce 1,00
Coleoidea Cephalopoda Loliginidae Loligo vulgaris European squid 3,50
Coleoidea Cephalopoda Loliginidae Sepia ofcinalis Common cuttlesh 3,50
Coleoidea Teuthida Teuthida Teuthida Squid 3,50
Cydippida Ctenophora Ctenophora Ctenophora Ctenophora 3,00
Cydippida Cydippida Pleurobrachiidae Pleurobrachia pileus Sea-gooseberry 3,00
Discomedusae Rhizostomeae Rhizostomatidae Rhizostoma pulmo Giant jellysh 3,00
Gastropoda Littorinimorpha Littorinimorpha Littorinimorpha Littorinimorpha 2,40
Heterobranchia Nudibranchia Nudibranchia Nudibranchia Nudibranchs 2,40
Hydrozoa 2,30
Hydrozoa Anthoathecata Tubularia Tubularia Tubularia 2,30
Hydrozoa Anthoathecata Corynidae Sarsia tubulosa Clapper medusa 2,50
Insecta Insecta Insecta Insecta Insects 1,00
Malacostraca Amphipoda Isopoda Hyperia galba Big-eye amphipod 2,30
Malacostraca Amphipoda Hyperiidae Talitrus saltator Sand hopper 2,30
Malacostraca Amphipoda Gammaridae Gammarus spec Gammarus 2,30
Malacostraca Balanomorpha Thoracica Semibalanus balanoides Barnacle 2,10
Malacostraca Balanomorpha Thoracica Thoracica Barnacles 2,30
Malacostraca Copepoda Copepoda Copepoda Copepods 2,00
Malacostraca Decapoda Anomura Paguroidea Hermit crabs 3,20
Malacostraca Decapoda Brachyura Corystes Helmet crabs 2,50
Malacostraca Decapoda Brachyura Macropipus Macropipus 2,50
Malacostraca Decapoda Brachyura Macropodia rostrata Long-legged spider crab 2,50
Malacostraca Decapoda Brachyura Portunidae Swimming crabs 2,50
Malacostraca Decapoda Carcinidae Cancer pagurus Edible crab 2,50
Malacostraca Decapoda Carcinidae Carcinus maenas Shore crab 2,50
Malacostraca Decapoda Corophiidae Corophium sp Corophium sp 2,60
Malacostraca Decapoda Corophiidae Corophium volutator Mud shrimp 2,60
Malacostraca Decapoda Crangonidae Caprella linearis Skeleton shrimp 2,60
Malacostraca Decapoda Crangonidae Crangon allmanni Crangon allmanni 2,60
Malacostraca Decapoda Crangonidae Crangon crangon Brown shrimp 2,60
Malacostraca Decapoda Crangonidae Gastrosaccus spinifer Gastrosaccus spinifer 2,20
Malacostraca Decapoda Crangonidae Mysidae Mysidae 2,20
Malacostraca Decapoda Crangonidae Palaemon serratus Aesop prawn 2,60
Malacostraca Decapoda Crangonidae Pontophilus bispinosus Philocheras bispinosus bispinosus 2,60
Malacostraca Decapoda Crangonidae Pontophilus trispinosus Philocheras trispinosus 2,60
Malacostraca Decapoda Crangonidae Praunus exuosus Chameleon shrimp 2,20
Malacostraca Decapoda Crangonidae Processa Processa 2,60
Malacostraca Decapoda Crangonidae Processa canaliculata Processa canaliculata 2,60
Malacostraca Decapoda Cumacea Cumacea Hooded shrimp 2,60
Malacostraca Decapoda Nephropidae Homarus gammarus European lobster 3,20
Malacostraca Decapoda Palaemonidae Palaemon elegans Grass prawn 2,60
Malacostraca Decapoda Polybiidae Macropipus holsatus Swimming crab 2,50
Malacostraca Isopoda Isopoda Idotea sp Idotea sp 2,30
Mollusca Mollusca Mollusca Mollusca Mollusca 2,60
Nematoda Nematoda Nematoda Nematoda Nematodes 2,10
Ophiuroidea Ophiurida Ophiuroidea Ophiura ophiura Serpent star 2,00
Ophiuroidea Spatangoida Loveniidae Echinocardium cordatum Sea-potato 2,00
Pisces Pisces Pisces Pisces 3,60
Pisces Atheriniformes Aterinidae Atherina presbyter Sand-smelt 3,70
Pisces Clupeiformes Clupeidae Alosa fallax Twaite shad 2,92
Pisces Clupeiformes Clupeidae Clupea harengus Herring 3,40
Pisces Clupeiformes Clupeidae Sprattus sprattus Sprat 3,09
Pisces Cyprinodontiformes Belonidae Belone belone Garsh 3,68
Pisces Gadiformes Gadidae Ciliata mustela Five-bearded rockling 3,53
Pisces Gadiformes Gadidae Merlangius merlangus Whiting 3,83
Pisces Gasterosteiformes Gasterosteidae Gasterosteus aculeatus Stickleback 3,30
Pisces Mugiliformes Mugilidae Liza aurata Golden grey mullet 2,05
Pisces Perciformes Moronidae Dicentrarchus labrax Bass 3,60
Pisces Perciformes Ammodytidae Ammodytes tobianus Sandeel 4,15
Pisces Perciformes Ammodytidae Hyperoplus lanceolatus Greater sandeel 4,00
(continued on next page)
S.S.H. Poiesz et al.

POior TPi=
α
+f(Year) +
ε
i
ε
i∼N(0,
σ
2)[2]
The model was cross-validated with different degrees of smoothing
(SPAN) to determine the optimal SPAN based on the minimum residual
sum of the root mean square error (RMSE). The evaluation of the GAM
results was done following Swartzman et al. (1992) and MacKenzie and
Schiedek (2007): The trend of the GAM model was drawn with 95%
condence limits. If a horizontal line could be drawn between the 95%
condence area of the tted trend, the results of the GAM model was
judged as no changes over time (P >0.05).
In addition, the whole sh data set (including all species) was ana-
lysed, whereby the present range of trophic position (TP) of the various
species (2010–2019) as described by Poiesz et al. (2020) was taken as
reference. For all years and all species, the estimates of TP were
compared with the reference period and scored as (1) above, (2) within
or (3) below the 2010–2019 range. Next, trends in these scores over time
were analysed per 5 year period.
All computations and analyses were done in R (R Core Team, 2021).
The graphics were made using the ggplot package (Wickham, 2009).
3. Results
3.1. Fish data
The NIOZ archive contained information about 7031 stomachs of 43
sh species over the years 1932–1979. Data for the years 1980–2019
included information about another 5217 stomachs of 60 sh species, in
total information about 12248 stomachs of 64 sh species. Records were
not evenly distributed but clustered both with respect to years as well as
sh species. Also, records of some species were only present in the
1940–1960’s (skates and shark species), records of bass Dicentrarchus
labrax (bass) only appeared in the samples in recent times and for some
species only few records were available (see Supplementary materials
Table S1). The archive data cluster around a few intervals: period
1947–1951; period 1962–1969; period 1975–1981; period 2005–2009
and the reference period 2010–2019 (Fig. 2A).
In total 117 different prey items were described over the years
(Table 1). For detailed information see Supplementary materials
Table S2. Number of species identied did not show a trend for the
various Classes except for slightly higher number Pisces and Polychaetes
in recent years (Fig. 2B,C,D,E).
For the analysis, sh species from the various functional groups and
guilds were selected (Table 2). Trends with GAM in stomach content
could be analysed for 15 species and trends in trophic position could be
determined in 16 species.
3.2. Stomach content
3.2.1. All data
Within the reference period (2010–2019), Malacostraca were the
most important Class of prey in the stomachs of the analysed Wadden
Sea sh species based on the mean relative occurrence, followed by
Pisces, Polychaetes and Bivalves (Fig. 3). The various periods each
showed a larger interannual variability of prey mean relative occurrence
than the reference period. During the period 2005–2009, the relative
mean occurrence of the various prey classes was within the range of the
reference period. In the period 1975–1981, more Polychaetes and more
Bivalves were found as prey. During the period 1962–1969, also more
Polychaetes were found as prey but less Pisces. The period 1947–1951
displayed a large variability: some years had more Pisces while other
years had hardly any Pisces but more Malacostraca as prey (Fig. 3).
The Malacostraca prey (3706 records) mainly consisted of the family
Crangonidae (2156 records, brown shrimps and other shrimps) and
furthermore Copepods (415 records). Pisces (2601 records) were partly
unidentied species (772 records) and furthermore Clupidae (394 re-
cords, mainly herring) and Gobiidae (339 records, mainly sand goby). In
addition there were 533 records of Callionymidae prey, however this
Table 1 (continued )
Class Order Family Scientic name Common name Trophic position (−)
Pisces Perciformes Trachinidae Echiichthys vipera Lesser weever 4,40
Pisces Perciformes Callionymidae Callionymus lyra Dragonet 4,41
Pisces Perciformes Gobiidae Gobius niger Black goby 3,30
Pisces Perciformes Gobiidae Pomatoschistus minutus Sand goby 3,20
Pisces Petromyzontiformes Petromyzontidae Petromyzon marinus Lamprey 3,11
Pisces Pleuronectiformes Pleuronectidae Limanda limanda Dab 3,40
Pisces Pleuronectiformes Pleuronectidae Platichthys esus Flounder 3,26
Pisces Pleuronectiformes Pleuronectidae Pleuronectes platessa Plaice 3,29
Pisces Pleuronectiformes Pleuronectidae Reinhardtius hippoglossoides Greenland halibut 4,60
Pisces Pleuronectiformes Solidae Buglossidium luteum Solenette 3,25
Pisces Pleuronectiformes Solidae Solea solea Sole 3,20
Pisces Salmoniformes Osmeridae Osmerus eperlanus Smelt 3,31
Pisces Scorpaeniformes Cottidae Myoxocephalus scorpius Bull-rout 3,90
Pisces Scorpaeniformes Cottidae Myoxocephalus quadricornis Four-horn sculpin 3,60
Pisces Scorpaeniformes Liparidae Liparis liparis Sea-snail 3,89
Pisces Scorpaeniformes Gobiidae Pomatoschistus lozanoi Lozano’s goby 3,30
Pisces Scorpaeniformes Gobiidae Pomatoschistus microps Common goby 4,45
Pisces Scorpaeniformes Gobiidae Pomatoschistus sp Pomatoschistus sp 4,45
Pisces Zeiformes Zeidae Zeus faber Dory 4,50
Polychaeta Annelida Annelida Annelida Annelida 2,10
Polychaeta Arenicolidae Arenicolidae Arenicolidae Arenicolidae 2,10
Polychaeta Phyllodocida Aphrodita Aphrodita Sea mouse 2,10
Polychaeta Phyllodocida Nereididae Alitta virens Sandworm 2,10
Polychaeta Phyllodocida Nereididae Nereididae Nereididae 2,10
Polychaeta Phyllodocida Nereididae Nereis Nereis 2,10
Polychaeta Phyllodocida Opheliidae Ophelia limacina Ophelia limacina 2,10
Polychaeta Phyllodocida Phyllodocidae Arenicola marina Lugworm 2,10
Polychaeta Phyllodocida Phyllodocidae Lanice conchilega Sand mason worm 2,10
Polychaeta Phyllodocida Phyllodocidae Marenzelleria viridis Marenzelleria viridis 2,10
Polychaeta Phyllodocida Phyllodocidae Nephtys hombergii Catworm 2,10
Polychaeta Phyllodocida Phyllodocidae Phyllodoce maculata Phyllodoce maculata 2,10
Polychaeta Phyllodocida Phyllodocidae Scoloplos armiger Scoloplos armiger 2,10
Polychaeta Polychaeta Polychaeta Polychaeta Bristle worms 2,10
Polychaeta Terebellida Pectinariidae Lagis koreni Trumpet worm 2,10
S.S.H. Poiesz et al.

Table 2
Overview of selected sh species from the NIOZ Archive, together with functional group and guild. (Near)-resident: Near-resident or resident species; JMM: juveniel marine migrants; MSV: Marine seasonal visitor. For
each species, total number of stomachs and number of years with observations, split up into unpublished data (1932–2009) and reference data (2010–2019 after Poiesz et al., 2020), are listed. Type of analysis is indicated
by X.
Scientic name Common name Functional
group
Guild Number of
stomachs
Number of years with
observations
Stomach
content
Trophic
position
Total 1932–2009 2010–2019 Total 1932–2009 2010–2019 Composition GAM Estimate GAM
Belone belone Garsh Pelagic (Near)-
resident
32 10 22 16 8 8 x x
Clupea harengus Herring Pelagic JMM 243 49 194 20 10 10 x x x x
Sprattus sprattus Sprat Pelagic JMM 51 20 31 15 7 8 x x x x
Trachurus trachurus Scad Pelagic MSV 109 29 80 20 10 10 x x x x
Osmerus eperlanus Smelt Pelagic MSV 120 14 106 15 7 8 x x x x
Merlangius
merlangus
Whiting Benthopelagic MSV 220 90 130 23 13 10 x x x x
Trisopterus luscus Bib Benthopelagic MSV 147 93 54 22 13 9 x x x x
Gadus morhua Cod Benthopelagic MSV 119 77 42 19 13 6 x x x x
Anguilla anguilla Eel Benthopelagic MSV 13 10 3 7 6 1 x x x
Ciliata mustela Five-bearded
rockling
Benthic (Near)-
resident
239 88 151 19 9 10 x x x x
Platichthys esus Flounder Benthic (Near)-
resident
456 182 274 28 18 10 x x x x
Myoxocephalus
scorpius
Bull-rout Benthic (Near)-
resident
156 103 53 23 14 9 x x x x
Zoarces viviparus Viviparous
blenny
Benthic (Near)-
resident
144 134 10 16 13 3 x x x x
Pomatoschistus
minutus
Sand goby Benthic (Near)-
resident
133 97 36 14 7 7 x x x x
Pleuronectes
platessa
Plaice Benthic JMM 1048 942 106 33 24 9 x x x x
Solea solea Sole Benthic JMM 59 39 20 17 12 5 x x x x
Limanda limanda Dab Benthic MSV 1260 1224 36 29 21 8 x x x x
S.S.H. Poiesz et al.

record is doubtful since it was based on a single observation of 512 prey
items in one year (1949). The Polychaetes (2917 records) mainly
referred to Annilida (2334 records) and furthermore Phyllodocidae (345
records, mainly Lanice spec. and Nereis spec.). The Bivalvia prey (1831
records) were mainly unidentied (949 records) and other Ensis spec.
(829 records). For detailed information see Supplementary Material
Table S2.
3.2.2. Individual species
The group of pelagic species (Fig. 4A) contained one (near)resident
species (garsh Belone belone), two juvenile marine migrants (herring
Clupea harengus and sprat Sprattus sprattus) and two marine seasonal
visitors (scad Trachurus trachurus and smelt Osmerus eperlanus). Garsh
mainly consumed Pisces (herring and to a lesser extent sandeel) but also
regularly Malacostraca (mainly brown shrimp). Prey items for herring
were mainly Malacostraca (mainly Copepods and to a lesser extent
Gammarus, Corophium and Mysidae) and some Pisces (herring and
sandeel), Polychaete, (mixture of species) and Bivalves (razor clams).
Sprat mainly consumed Malacostraca (mainly consisting of Copepods
and to a lesser extent shore crab and brown shrimp). For scad, main prey
items were Malacostraca (mainly brown shrimp and shore and swim-
ming crabs) and Pisces (mainly herring and sandeel). Malacostraca
(mainly shrimps and swimming crabs and some Copepods) and Pisces
(herring and various goby species) were also the main prey items of
smelt.
The group of benthopelagic species (Fig. 4B) only contained four
marine seasonal visitor species (whiting Merlangius merlangus, bib Tri-
sopterus luscus, cod Gadus morhua and eel Anguilla anguilla). Whiting
focused on Pisces (mainly herring, some sandeel and various goby spe-
cies), Malacostraca (mainly shrimps, to a lesser extend crabs and some
Mysis) and from 2000 on Autobranchia (Ensis spec.). Bib mainly
consumed Malacostraca (shrimps, crabs and some Mydidae) and some
Pisces (herring and to a lesser extend sand goby and sandeel). Main prey
items for cod were Malacostraca (mainly brown shrimp and some shore
crabs), some Pisces (mainly herring and some sandeel and goby species)
and Polychaeta. Eel mainly preyed upon Polychaeta and Malacostraca
(brown shrimps and shore crabs).
The group of demersal species (Fig. 4C) contained ve (near)-resi-
dent species (ve-bearded rockling Ciliata mustela, ounder Platichthys
esus, bull-rout Myoxocephalus scorpius, viviparous blenny Zoarces
viviparus and sand goby Pomatoschistus minutus), two juvenile marine
migrant species (plaice Pleuronectes platessa and sole Solea solea) and a
marine seasonal visitor species (dab Limanda limanda). Five-bearded
rockling focussed on Malacostraca (mainly brown shrimp but also
crabs) and in recent years sometimes on Pisces (mainly herring and also
goby species). Flounder consumed a variety of prey items but especially
Polychaeta, Pisces (mainly herring and some goby species), Malacos-
traca (mainly brown shrimps and to a lesser extend Corophium and shore
and swimming crabs) and some Bivalves. Bull-rout preyed mainly on
Malacostraca (brown shrimp and shore and swimming crabs) and to a
lesser extent on Pisces (mainly herring). Main prey items of viviparous
blenny were Malacostraca (Amphipods, brown shrimp and some crabs)
and some Polychaeta, Pisces (herring) and Bivalves. Sand goby preyed
especially upon Malacostraca (Copepods, Amphipods, small shrimp and
shore crabs) and also on some Polychaeta and Pisces (herring). Plaice
consumed a variety of prey species, especially Polychaeta, Malacostraca
(mainly shrimps and shore and swimming crabs and some Amphipods
and Mysis), Bivalves and Caenogastropoda (Hydrobia). Sole focused on
Polychaeta and Malacostraca (mainly shrimps and crabs and some
Mysis). Dab consumed a variety of prey items with a focus on Poly-
chaeta, Pisces (mainly herring and furthermore some sandeel), and
Malacostraca (mainly brown shrimps and furthermore shore and
swimming crabs).
For the group of pelagic species, for all prey items (except for one
year for the occurrence of Malacostraca in the diet of herring) for which
a GAM with 95% condence limits could be calculated, a horizontal line
could be drawn between the 95% condence limits of the tted trend,
implying that the frequency of occurrence had not changed over time
(Supplementary Material Fig. S1A and Table 3). For three benthopelagic
species (whiting, bib and cod) a GAM with 95% condence limits could
be calculated for the Malacostraca and Pisces and in all cases a hori-
zontal line could be drawn between the 95% condence limits of the
tted trend, implying that the frequency of occurrence of Malacostraca
and Pisces had not changed over time for whiting, bib and cod
(Supplementary Material Fig. S1B and Table 3). For the group of
demersal species, for all prey items (except for the occurrence of Bi-
valves in the diet of plaice) for which a GAM with 95% condence limits
could be calculated, a horizontal line could be drawn between the 95%
condence limits of the tted trend, implying that the frequency of
occurrence had not changed over time (Supplementary Material Fig. S1C
and Table 3).
All GAM parameters of the various trends in prey occurrence of the
various sh species [smoother span, number of observations, number of
parameters, standard error, smoother matrix, effective degrees of
freedom (edf) and the p-value] are presented in Supplementary Material
Table S3.
3.3. Trophic position
For all individual years of the reference period 2010–2019, esti-
mated trophic levels of the stomach contents based on a rerun with
trophic values of identied prey Class were signicantly related with
original estimates based on trophic values of identied prey Species
(Fig. 5).
Fig. 3. Relative mean occurrence (%) of the most abundant prey classes in the
stomachs of Wadden Sea sh species within the NIOZ archive (1932–2019).
Only years with al least 50 observations are listed.
S.S.H. Poiesz et al.

Fig. 4A. Mean occurrence of various prey items in the stomachs of the selected pelagic species; the (near)-resident species garsh Belone belone; the juvenile marine
migrants herring Clupea harengus and sprat Sprattus sprattus and the marine seasonal visitors scad Trachurus trachurus and smelt Osmerus eperlanus.
S.S.H. Poiesz et al.

3.3.1. All data
Variability in trophic position of the different prey species was low
for most Classes, except for Malacostraca (23%) and Pisces (38%)
(Table 1). For the reference period (2010–2019) in almost all sh spe-
cies, the estimated trophic position showed variation over a range of ~2
units (Table 4).
The estimated mean trophic position (TP) of the various sh species
over the years can be found in Supplementary Material Table S4. In the
period 2005–2009, estimated mean trophic position were within those
of the period 2010–2019 (Fig. 6). Between 1970 and 1990, the per-
centage of species with estimates of trophic position within the reference
range was lower, around 50–60 % and a higher percentage of species
had estimates below the reference range compared with above the range
(Fig. 6). During the period 1945–1950 the percentage of species with
estimates respectively below and above the reference range were almost
similar (Fig. 6).
3.3.2. Individual species
In 16 species, enough data were present to apply General Additive
Models (GAMs) to visualise and analyse trends over time (Fig. 7ABC).
In all pelagic (twaite shad, herring, sprat, scad and smelt), bentho-
pelagic (whiting, bib, cod and eel) and demersal species (ve-bearded
rockling, ounder, bull-rout, viviparous blenny, sand goby, plaice, sole
and dab), a horizontal line could be drawn between the 95% condence
limits of the tted trend, implying no change over time (Fig. 7ABC). All
GAM parameters of the various trends in trophic position of the various
sh species [smoother span, number of observations, number of pa-
rameters, standard error, smoother matrix, effective degrees of freedom
(edf) and the p-value] are presented in Supplementary Material Table
S5.
4. Discussion
4.1. Quality and limitations of the NIOZ archive data
Long-term series are unique and in principle valuable data sets,
however a precondition is that the quality and the limitations of the data
can be judged and that potential pitfalls can be identied. Wiltshire and
Dürselen (2004) carried out a quality control of the Helgoland Reede
long-term phytoplankton data archive (1962 –present) and listed a
number of typical general problems they came across. The most
important issues that can be expected for all long-term series can be
summarized as:
- lack of meta-information, especially from the past;
- the mismatch between the original records on paper and the elec-
tronical archive;
- outdated taxonomic nomenclature and synonyms;
- different procedures over time;
- different investigators over time with different taxonomic
knowledge.
Also the NIOZ archive data suffers from some of these problems. The
NIOZ archive also lacks meta-information with respect to information
about potential digestion between time of catch and of stomach analysis.
However, most of the records originate from NIOZ courses where sh
were dissected immediately after being caught. Stomach content of rare
sh species from shermen might have suffered from digestion: often
these stomachs were empty or could not be identied. The NIOZ archive
did not suffer from a mismatch between the original records on paper
and the electronical archive, because the data were never electronically
archived in the past. The problem of outdated taxonomic nomenclature
and synonyms occurred but was solved by using WoRMS (http://www.
Fig. 4B. Mean occurrence of various prey items in the stomachs of the selected benthopelagic species; the marine seasonal visitor species (whiting Merlangius
merlangus, bib Trisopterus luscus, cod Gadus morhua and eel Anguilla Anguilla.
S.S.H. Poiesz et al.

Fig. 4C. Mean occurrence of various prey items in the stomachs of the selected benthic species; the (near)-resident species ve-bearded rockling Ciliata mustela,
ounder Platichthys esus, bull-rout Myoxocephalus scorpius, viviparous blenny Zoarces viviparus and sand goby Pomatoschistus minutus; the juvenile marine migrant
species plaice Pleuronectes platessa and sole Solea solea and the marine seasonal visitor species dab Limanda limanda, thick-lipped grey mullet Chelon labrosus and
lesser weever Echiichthys vipera.
S.S.H. Poiesz et al.

marinespecies.org) for checking species, family, genus and class name of
all sh and prey records. As far as we can check in the records, all an-
alyses have always been supervised and/or carried out by qualied
NIOZ staff with taxonomic knowledge.
Most striking are the differences in amount of data and in level of
taxonomic identication of stomach content over time. From 1980 on-
wards, sh were collected regularly in spring and summer and until
1980 only randomly as part of NIOZ courses and landings from sher-
men. This means that despite the more than 12.000 records of stomach
content analysis for the sh community of the Wadden Sea, the dataset
shows a large patchiness and variability both with respect to the years in
which data were collected and in the sh species analysed. Surprisingly,
level of taxonomic identication of prey over time hardly affected the
estimate of the trophic position. A sensitivity analysis for the reference
period 2010–2019 showed that estimated trophic levels based on prey
Class were signicantly related with original estimates based on prey
Species.
By interpreting the results of the analysis of the NIOZ archive data,
these restrictions should be kept in mind.
4.2. Wadden Sea baseline
What would be a realistic baseline for the Wadden Sea system and in
particular for its sh fauna is open for debate. The Wadden Sea has been
under the inuence of anthropogenic stress for centuries (see for
instance Lotze, 2005,2007). Stress factors caused by human-induces
activities (overshing and pollution events), could theoretically be
reduced or stopped. This, however, might be not pragmatic. However,
other factors such as habitat loss are even more difcult to reverse. The
last extensive habitat loss in the western Dutch Wadden Sea took place
in the Marsdiep tidal basin in the western part in 1932 with the exclo-
sure of the Zuiderzee estuary by the Afsluitdijk and in the eastern Dutch
Wadden Sea in 1964 with the exclosure of the Lauwers (Wolff, 1983).
This means that for the Marsdiep tidal basin a baseline before 1932 is
unrealistic with respect to any analyses with more recent data, including
the present situation. The low shing pressure and low level of pollution
(nutrients, chemicals) during the second world war would plea for a
realistic baseline around 1945 for the Wadden Sea system.
Quantitative information about the Dutch Wadden Sea system for the
period around 1945 is scarce except for water temperature and salinity
data (van Aken, 2008a,2008b) and remains fragmentary until the
beginning of the 1970s, despite the start of nutrient measurements
(phosphorus) from 1949 onwards (Postma, 1954) and primary produc-
tion estimates (Postma and Rommets, 1970) and demersal sh surveys
in 1963–1965 (Creutzberg and Fonds, 1971). Only for the last half
century from the 1970’s onwards more systematic information is
available with presently time series about various abiotic and biotic
ecosystem components such as water temperature and salinity, primary
production, the benthic community, sh fauna, wading birds and marine
mammals for various parts of the Wadden Sea. For an overview see the
various quality status reports of the Wadden Sea (https://qsr.waddensea
-worldheritage.org/).
The present study contains information on trophic structure based on
stomach content dating back to the early 1930’s. The Wadden Sea
ecosystem in the 1930’s will have been a system with lower nutrient
concentrations (van der Veer et al., 1989;van Raaphorst and van der
Veer, 1990;van Raaphorst and de Jonge, 2004) but nevertheless a
system with a higher sh abundance compared to the present ecosystem.
A higher sh abundance in the past is supported by the fact that, before
and after the second world war, there was a protable commercial fyke
net shing in the area. However catches and protability decreased
rapidly until the last shing company was terminated in 1966 and taken
over by NIOZ to start the long-term monitoring series (van der Veer
et al., 2015).
Fig. 4C. (continued).
S.S.H. Poiesz et al.

The much and varying variability in the stomach content for any
given species within the period 1930–2019 raises the question whether
stomach content data of sh species is absent in particular years and
decades because the sh species were absent or rare in the ecosystem, or
because they were simply not targeted during that time. For most spe-
cies, missing data indicate that they were not targeted: from the 1980’s
onwards, stomach content data were collected from a long-term
monitoring programme of the sh fauna with a passive sh trap near the
entrance of the Wadden Sea in spring and autumn and during that period
no species went extinct and common species were caught almost every
year (Poiesz et al., 2020). Only most of the skate and shark species
disappeared from the Wadden Sea from the 1960’s onwards, similar as
in other areas (Walker and Heessen, 1996;Dulvy and Reynolds, 2002;
Reynolds et al., 2005;Heessen et al., 2015;Bom et al., 2020;Poiesz
et al., 2021b).
4.3. Prey consumption
A variety of sources are available for the reconstruction of the sh
food web structure, ranging from anecdotal and semi-quantitative in-
formation about species composition (see for instance Roberts, 2007) to
quantitative analysis of archaeological remains such as of bones and
otoliths. The latter can include stable isotope analysis (Fry, 2006;Mid-
delburg, 2014;Phillips et al., 2014;Tsutaya et al., 2021), genetics, age
and growth analyses (see for example Bolle et al., 2004,Cuveliers et al.,
2007) and stomach content analysis [such as deriving trophic structure
and predator−prey interactions (Hynes, 1950;Baker et al., 2014)].
Stomach content analysis provides information about recently ingested
prey items only, while especially regurgitation and digestion are factors
that may cause prey items to be missed or overlooked. The extended
period of sampling may have partly overcome these limitations, how-
ever, for rare species an insufcient number of stomachs may have been
sampled to cover all possible prey species (Karachle and Stergiou, 2017;
Mulas et al., 2015).
Recent studies in two different parts of the Wadden Sea reveal that,
although most of the Wadden Sea sh species are rather omnivorous,
their food requirements are fuelled by a few key prey species (Kellnreiter
et al., 2012;Poiesz et al., 2020). This omnivorous feeding behaviour can
also be recognized in the stomach content compositions of the Wadden
Sea sh fauna over the last half century. Interannual variations in
stomach composition do occur due to variations in the level of detail of
the stomach content analysis over the years as well as variations in prey
abundance. Nevertheless, a few groups, Bivalvia/Autobranchia, Poly-
chaeta, Malacostraca (mainly Decapoda: shrimps and crabs) and Pisces,
were the main prey items from the 1930’s onwards to recent decades. A
few key species as main pathways of energy ow to higher trophic levels
might be a general characteristic for estuarine systems; it has been
described for other areas also, such as Amphipods and Copepods in the
French Chanche estuary (Selleslagh et al., 2012).
Trends in prey occurrence in the stomachs could be determined for
some prey items in some individual sh species. However, the analysis
Table 3
Changes over time in main prey groups (grey) of selected sh species from the NIOZ Archive 1931–2019. (Near)-resident: Near-resident or resident species; JMM:
juveniel marine migrants; MSV: Marine seasonal visitor. Signicance of the GAM trend is indicated (n.s.: not signicantly deviating from horizontal line or P <0.05).
Only years with more than 5 observations are included.
Scientic name Common name Functional group Guild Algae Bivalves Malacostraca Pisces Polychaeta
Clupea harengus Herring Pelagic JMM n.s. P <0.05 n.s. n.s.
Sprattus sprattus Sprat Pelagic JMM n.s.
Trachurus trachurus Scad Pelagic MSV n.s. n.s.
Osmerus eperlanus Smelt Pelagic MSV n.s. n.s.
Merlangius merlangus Whiting Benthopelagic MSV n.s. n.s.
Trisopterus luscus Bib Benthopelagic MSV n.s. n.s.
Gadus morhua Cod Benthopelagic MSV n.s. n.s.
Ciliata mustela Five-bearded rockling Benthic (Near)-resident n.s. n.s.
Platichthys esus Flounder Benthic (Near)-resident n.s. n.s.
Myoxocephalus scorpius Bull-rout Benthic (Near)-resident n.s. n.s.
Zoarces viviparus Viviparous blenny Benthic (Near)-resident n.s.
Pomatoschistus minutus Sand goby Benthic (Near)-resident n.s. n.s.
Pleuronectes platessa Plaice Benthic JMM P <0.05 n.s. n.s.
Solea solea Sole Benthic JMM n.s. n.s.
Limanda limanda Dab Benthic MSV n.s. n.s. n.s.
Fig. 5. Comparsion of estimated TP values of individual sh species per 5 year
period with the range of TP of the reference period 2010–2019.
S.S.H. Poiesz et al.

Table 4
Size frequency distribution of estimated trophic position (TP) of individual sh for the period 2010–2019.
Scientic name Common name Trophic position (TP)
1.00–1.49 1.50–1.99 2.00–2.49 2.50–2.99 3.00–3.49 3.50–3.99 4.00–4.49 4.50–4.99 5.00–5.49 Total
Agonus cataphractus Hooknose 1 28 1 2 32
Alosa fallax Twaite shad 7 138 23 33 201
Ammodytes tobianus Sandeel 3 6 9
Anguilla anguilla Eel 1 8 9
Aphia minuta Transparent goby
Arnoglossus laterna Scaldsh 24 1 25
Aspitrigla cuculus Red gurnard
Atherina presbyter Sand-smelt 1 1 6 30 2 2 42
Belone belone Garsh 2 2 1 26 31
Callionymus lyra Dragonet 1 1 2
Callionymus
reticulatus
Reticulated
dragonet
1 1
Chelon auratus Golden grey mullet 1 88 1 4 5 99
Chelon labrosus Thick-lipped grey
mullet
55 2 2 16 2 5 82
Chelon ramada Thin-lipped grey
mullet
3 3
Ciliata mustela Five-bearded
rockling
3 3 390 17 25 438
Clupea harengus Herring 8 1 26 220 12 33 300
Cyclopterus lumpus Lumpsucker 1 26 2 29
Dicentrarchus labrax Bass 3 1 52 446 33 100 635
Dipturus batis Skate
Echiichthys vipera Lesser weever 2 3 1 6
Engraulis encrasicolus Anchovy 2 2
Eutrigla gurnardus Grey gurnard 1 1
Gadus morhua Cod 1 67 10 10 88
Gasterosteus aculeatus Stickleback 2 1 11 58 2 1 75
Gobius niger Black goby
Hyperoplus
lanceolatus
Greater sandeel 1 9 10
Limanda limanda Dab 11 27 1 3 42
Liparis liparis Sea-snail 86 2 0 88
Lipophrys pholis Shanny 1 1
Melanogrammus
aeglenus
Haddock
Merlangius merlangus Whiting 1 19 170 24 25 239
Microstomus kitt Lemon sole 1 1
Mustelus mustelus Smooth hound
Myoxocephalus
scorpius
Bull-rout 1 2 100 12 3 118
Neogobius
melanostomus
Round goby 7 1 8
Osmerus eperlanus Smelt 1 11 110 13 33 168
Pholis gunnellus Buttersh 2 5 7
Phrynorhombus
norvegicus
Norwegian topknot
Platichthys esus Flounder 2 2 114 276 14 49 457
Pleuronectes platessa Plaice 1 2 66 39 3 10 121
Pollachius pollachius Pollack 46 13 7 66
Pollachius virens Saithe 1 2 6 7 8 24
Pomatoschistus
lozanoi
Lozano’s goby 1 10 11
Pomatoschistus
microps
Common goby 1 4 5
Pomatoschistus
minutus
Sand goby 14 22 4 40
Salmo salar Salmon
Salmo trutta Sea trout 3 47 8 236 294
Sardina pilchardus Pilchard 4 14 3 21
Scomber scombrus Mackerel 5 6 8 19
Scophthalmus
maximus
Turbot 1 2 58 11 19 91
Scophthalmus
rhombus
Brill 13 1 10 24
Solea solea Sole 1 16 6 5 28
Sparus aurata Gilt-head sea
bream
2 2
Sprattus sprattus Sprat 1 1 37 3 42
Squalus acanthias Spurdog
Syngnathus acus Greater pipesh 18 2 2 22
Syngnathus rostellatus Nilsson’s pipesh 33 33
Taurulus bubalis Sea scorpion 1 21 22
(continued on next page)
S.S.H. Poiesz et al.

was hampered by large patchiness and variability in the data and in the
variability in the level of detail of the stomach content analysis. In all
sh species that could be analysed, prey occurrence showed uctuations
over time. The most important prey species of the Malacostraca, the
brown shrimp and the shore crab, both showed large interannual uc-
tuations in the Dutch Wadden Sea, with a general increase in both
species over a 40 yr period (Tulp et al., 2012). Similar uctuations were
observed in secondary production of intertidal bivalves and polychaetes,
however without a clear trend over time (Beukema and Dekker, 2022).
Also the Wadden Sea sh community showed strong interannual uc-
tuations in abundance, added to a clear decline from the 1980s until the
early 2000s (Tulp et al., 2008;van der Veer et al., 2015). Herring, the
most important sh prey species, showed strong variation among years
and uctuated in abundance within one order of magnitude (van der
Veer et al., 2015). Therefore, the uctuations in stomach content
composition partly reects interannual variability in absolute and
relative abundance of the most important prey groups.
The large patchiness and variability in the data resulted in large
condence intervals of the GAM smoother over time. Despite the large
uctuations in prey occurrence in the stomachs of the various sh spe-
cies, hardly any signicant differences between years were found
(except for the occurrence of Pisces in the diet of smelt and sand-smelt
and the occurrence of Malacostraca in the diet of sole and dab).
Furthermore, no trends in prey occurrence over time were found in the
various species analysed. This means that amphipod crustaceans, brown
shrimps and crabs, juvenile herring and gobies and to a lesser extend
bivalves and polychaetes are not only the key prey species presently
(Poiesz et al., 2020) but already had a pivotal position in the sh food
web in the past, at least from the 1930’s onwards (this study).
4.4. Trophic position
The large patchiness in the data for all Wadden Sea sh species with
respect to years of sampling, results in a mozaik of snapshots of trophic
positions of individual species over time and in a number of species with
enough data to apply General additive models (GAMs) to visualise and
analyse trends over time. The analysis of the complete data set and the
analyses of the individual species both indicated that trophic positions
during the period 1930–2010 were variable but did not signicantly
differ from those in the present reference period (as described in Poiesz
et al., 2020). The variability in individual stomach contents, and hence
in the estimates of trophic position, illustrates the omnivorous character
of most of the sh species in the Wadden Sea: current day estimations of
trophic position varies by 2 units for most sh species (Poiesz et al.,
2020,Table 3). It cannot be excluded that the present dataset with high
sampling variability might be not robust enough to identify trends over
time for these sh species with an inherent large individual variability in
trophic position.
On the other hand, network analyses indicate that estuaries are
rather stable systems, where a few species such as for instance clupeids,
atsh and gobies are able to cope with the inherent cyclical and sea-
sonal perturbations: those species are robust and are responsible for a
stable system (Lobry et al., 2008). In the western Dutch Wadden Sea
there are also no trend indications in the number of species caught over
the period 1960–2010 (van der Veer et al., 2015). The fact that in this
study no trend in trophic position was found in species belonging to
different modes of life (pelagic, benthopelagic and demersal) and guild
(near-resident, juvenile marine migrant or seasonal visitor) might imply
that this could also hold true for the other species not analysed in this
study.
Although estuaries might be rather stable systems, serious impacts of
anthropogenic stress have nevertheless been documented for many of
these systems (see for instance Kennish, 1991,2002;Chapman and
Wang, 2001), including the Wadden Sea (Lotze, 2005,2007). With
respect to the sh fauna, this has led to the disappearance of most skate
and shark species in the area, causing a loss of biodiversity in the
Wadden Sea from the 1960’s onwards, similar to those reported in other
areas (Walker and Heessen, 1996;Dulvy and Reynolds, 2002;Reynolds
et al., 2005;Heessen et al., 2015;Bom et al., 2020;Poiesz et al., 2021b).
Before the 1960s, the Wadden Sea sh community did include skate and
sharks, top predators with a relatively high trophic position.
The present study does not indicate any changes in trophic position
of individual species in the western Dutch Wadden Sea over the last 80
years. This may be different at the community level. Although sh
species composition in the western Wadden Sea has shown to be rather
robust, species composition does show some interannual variation (van
der Veer et al., 2015). Some species have also disappeared in the past,
Table 4 (continued )
Scientic name Common name Trophic position (TP)
1.00–1.49 1.50–1.99 2.00–2.49 2.50–2.99 3.00–3.49 3.50–3.99 4.00–4.49 4.50–4.99 5.00–5.49 Total
Trachinus draco Greater weever
Trachurus trachurus Scad 4 35 12 79 130
Trigla lucerna Tub gurnard 1 19 2 22
Trisopterus luscus Bib 3 106 2 6 117
Trisopterus minutus Poor cod 1 3 4
Zeus faber Dory
Zoarces viviparus Viviparous blenny 1 3 4 1 2 11
Fig. 6. Comparison of estimated mean TP values of all sh species per 5-year
period to the reference period TP range (2010–2019). For more information
see text.
S.S.H. Poiesz et al.

Fig. 7A. Mean trophic position (−) of the selected pelagic species; the (near)resident species twaite shad Alosa fallax; the juvenile marine migrants herring Clupea
harengus and sprat Sprattus sprattus and the marine seasonal visitors scad Trachurus trachurus and smelt Osmerus eperlanus.
S.S.H. Poiesz et al.

such as most of the skate and shark species. Furthermore, year-to year
uctuations in the relative abundance of the various sh species (Tulp
et al., 2008;van der Veer et al., 2015) will be reected in interannual
variations in the trophic structure of the sh community. In the western
Wadden Sea, the trophic structure of this community showed indeed
some uctuations from 1980 to 2011. For both the demersal and ben-
thopelagic sh fauna the trophic position remained the same, while for
pelagic sh the mean fell from about 3.9 to 3.1., mainly due to the
decrease in abundance of predatory pelagic sh such as cod and garsh
(van der Veer et al., 2015).
The 10-fold decrease in total biomass of the catches of both pelagic
and demersal species from 1980 to 2011 (van der Veer et al., 2015) il-
lustrates the degradation of the trophic role of the sh community at the
ecosystem level in the western Wadden Sea. To what extent this has
affected ecosystem functioning is unclear. In the North Sea, the deple-
tion of demersal sh species in the period 1973–2000 appears to have
released the benthos from “top-down”biomass control, leading to an
increase in benthic production and invertebrates (Heath, 2005). To what
extent the trophic structure of the sh community in the western
Wadden Sea are a reection of a more general pattern also in the other
tidal basins of the Wadden Sea is unclear. The fact that most species are
omnivorous and species composition appears to be largely the same at a
large scale (Kühl and Kuipers, 1983;Kellnreiter et al., 2012;Meyer et al.,
2016;Poiesz et al., 2020) might suggest a general pattern in trophic
position of the sh species in the Wadden Sea. However, the fact that
local and interannual differences were found in the abundance of
demersal sh in the western, central and eastern part of the Wadden Sea
and in its coastal regions (Tulp et al., 2017) implies that at the
community and ecosystem level the trophic structure of the sh com-
munity may differ to some extent.
4.5. Conclusive remarks
In this study, trends in prey species consumed and in trophic position
were analysed and by means of stomach content information compared
to the present situation (2010–2019) for 18 omnivorous sh species in
the western Dutch Wadden Sea. Prey consumption of different sh
species showed variability over time, but without a change over time.
Also, in all 18 species, no signicant change in mean trophic position
over time could be found. Despite the general decrease in sh abundance
in the area (van der Veer et al., 2015). The present study does not
indicate any changes in trophic position of individual species in the
western Dutch Wadden Sea over the last 80 years despite the serious
level of anthropogenic stress (pollution, eutrophication events, climate
change) and the decrease in sh abundance in the area.
CRediT authorship contribution statement
Suzanne S.H. Poiesz: Writing –review &editing, Writing –original
draft, Visualization, Validation, Supervision, Project administration,
Formal analysis, Data curation, Conceptualization. Johannes IJ. Witte:
Methodology. Henk W. van der Veer: Writing –review &editing,
Writing –original draft, Validation, Supervision.
Fig. 7B. Mean TP (−) of the selected benthopelagic species; the marine seasonal visitor species whiting Merlangius merlangus, bib Trisopterus luscus, cod Gadus morhua
and eel Anguilla Anguilla.
S.S.H. Poiesz et al.

Fig. 7C. Mean TP (−) of the selected benthic species; the (near)-resident species ve-bearded rockling Ciliata mustela, ounder Platichthys esus, bull-rout Myox-
ocephalus scorpius, viviparous blenny Zoarces viviparus and sand goby Pomatoschistus minutus; the juvenile marine migrant species plaice Pleuronectes platessa and sole
Solea solea and the marine seasonal visitor species dab Limanda limanda.
S.S.H. Poiesz et al.

Declaration of competing interest
The authors declare that they have no competing interests.
Data availability
Data will be made available on request.
Acknowledgements
Thanks are due to all our colleagues who assisted in the collection
and analyses of the samples, especially Rob Dapper, Ewout Adriaans,
Willem Jongejan, Sieme Gieles and Marco Kortenhoeven. All recent sh
sampling and handling was done under CCD project number:
AVD8020020174165.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ecss.2024.108912.
References
Amundsen, P.-A., S´
anchez-Hern´
andez, J., 2019. Feeding studies take guts –critical
review and recommendations of methods for stomach contents analysis in sh.
J. Fish. Biol. 95, 1364–1373.
Baker, R., Buckland, A., Sheaves, M., 2014. Fish gut content analysis: robust measures of
diet composition. Fish Fish. 15, 170–177.
Barbier, E.B., 2017. Marine ecocystem services. Curr. Biol. 27, R431–R510.
Beck, M.W., Heck, J.R.K.L., Able, K.W., Childers, D.L., Eggleston, D.B., Gillanders, B.M.,
Halpern, B., Hays, C.G., Hoshino, K., Minello, T.J., Orth, R.J., Sheridan, P.F.,
Weinstein, M.P., 2001. The identication, conservation, and management of
estuarine and marine nurseries for sh and invertebrates. Bioscience 51, 633–641.
Beukema, J.J., Dekker, R., 2022. Bottom-up as well as top-down processes govern
zoobenthic secondary production in a tidal-at ecosystem. Limnol. Oceanogr. 67,
2547–2556.
Beukema, J.J., Dekker, R., 2020. Half a century of monitoring macrobenthic animals on
tidal ats in the Dutch Wadden Sea. Mar. Ecol. Prog. Ser. 656, 1–18.
Bolle, L.J., Rijnsdorp, A.D., van Neer, W., Millner, R.S., van Leeuwen, P.I., Ervynck, A.,
Ayers, R., Ongenae, E., 2004. Growth changes in plaice, cod, haddock and saithe in
the North Sea: a comparison of (post)medieval and present-day growth rates on
otolith measurements. J. Sea Res. 51, 313–328.
Bom, R.A., van de Water, M., Camphuysen, C.J., van der Veer, H.W., van Leeuwen, A.,
2020. Historical ecology and demise of the iconic angelshark Squatina squatina in the
southern North Sea. Mar. Biol. 167, 1–10.
Buckland, A., Baker, R., Loneragan, N., Sheaves, M., 2017. Standardising sh stomach
content analysis: the importance of prey condition. Fish. Res. 196, 126–140.
Bijma, J., P¨
ortner, H.-O., Yesson, C., Rogers, A.D., 2013. Climate change and the oceans -
what does the future hold? Mar. Poll. Bull. 74, 495–505.
Cardoso, J.F.M.F., Freitas, V., Quilez, I., Jouta, J., Witte, J.IJ., van der Veer, H.W., 2015.
The European Sea bass Dicentrarchus labrax in the Dutch Wadden Sea: from visitor to
resident species. J. Mar. Biol. Ass. UK 95, 839–850.
Chapman, P.M., Wang, F., 2001. Assessing sediment contamination in estuaries. Environ.
Toxicol. Chem. 20, 3–22.
Creutzberg, F., Fonds, M., 1971. The seasonal variation in the distribution of some
demersal sh species in the Dutch Wadden Sea. Thalassia Jugosl. 7, 13–23.
Cuveliers, E.L., Bolle, L.J., Volkaert, F.A.M., Maes, G.E., 2007. Inuence of DAN isolation
from historical otoliths on nuclear-mitochondrial amplication and age
determination in an overexploited sh, the common sole (Solea solea L.). Mol. Ecol.
Resourc. 9, 725–732.
de Vooys, C.G.N., Witte, J.IJ., Dapper, R., van der Meer, J., van der Veer, H.W., 1991.
Lange termijn veranderingen in zeldzame vissoorten op het Nederlands continentaal
plat van de Noordzee. Nioz-rapport 6, 81.
de Vooys, C.G.N., Witte, J.IJ., Dapper, R., van der Meer, J., van der Veer, H.W., 1993.
Lange termijn veranderingen op het Nederlands continentaal plat van de Noordzee:
trends in invertebraten van 1931–1990. Nioz-rapport 17, 68.
Dulvy, N.K., Reynolds, J.D., 2002. Predicting extinction vulnerability in skates. Conserv.
Biol. 16, 440–450.
European Marine Board, 2013. Navigating the future IV. Position paper 20 of the
European Marine Board. Ostend, Belgium, ISBN 9789082093100.
Froese, R., Pauly, D., 2021. FishBase. www.shbase.org.
Fry, B., 2006. Stable Isotope Ecology. Springer-Verlag, New York, USA.
Goodall, D.W., 1983. Estuaries and Enclosed Seas. Elsevier, Amsterdam.
Hastie, T.J., Tibshirani, R.J., 1995. Generalized Additive Models. Chapman and Hall,
London.
Hayward, P.J., Ryland, J.S., 2017. Handbook of the Marine Fauna of North-West Europe.
Oxford University Press, Oxford, UK.
Heath, M.R., 2005. Changes in the structure and function of the North Sea sh foodweb,
1973-2000, and the impacts of shing and climate. ICES J. Mar. Sci. 62, 847–868.
Heessen, H.J.L., Daan, N., Ellis, J.R., 2015. Fish Atlas of the Celtic Sea, North Sea and
Baltic Sea. Wageningen Academic Publishers, Wageningen.
Hynes, H.B.N., 1950. The food of the freshwater sticklebacks (Gastrosteus aculeatus and
Pygosteus pungitius) with a review of methods used in studies of the food of shes.
J. Anim. Ecol. 19, 36–58.
Hyslop, E.J., 1980. Stomach contents analysis-a review of methods and their application.
J. Fish. Biol. 17, 411–429.
Jackson, J.B., Kirby, M.X., Berger, W.H., Bjorndal, K.A., Botsford, L.W., Bourque, B.J.,
Bradbury, R.H., Cooke, R., Erlandson, J., Estes, J.A., Hughes, T.P., Kidwell, S.,
Lange, C.B., Lenihan, H.S., Pandol, J.M., Peterson, C.H., Steneck, R.S., Tegner, M.J.,
Warner, R.R., 2001. Historical overshing and the recent collapse of coastal
ecosystems. Science 37, 293–629.
Jacobs, P., Kromkamp, J.C., van Leeuwen, S.M., Philippart, C.J.M., 2020. Planktonic
primary production in the western Dutch Wadden Sea. Mar. Ecol. Prog. Ser. 639,
53–71.
Kvaarik, C., ´
Oskarsson, G.J., Danielsd´
ottir, A.K., Marteinsd´
ottir, G., 2019. Diet and
feeding strategy of Northeast Atlantic mackerel (Scombrus scomber) in Icelandic
waters. PLoS One 14 (12), e0225552.
Karachle, P.K., Stergiou, K.I., 2017. An update on the feeding habits of sh in the
Mediterranean Sea (2002-2015). Mediterr. Mar. Sci. 18, 43–52.
Kennish, M.J., 1991. Ecology of Estuaries: Anthropogenic Effects. CRC Press, Boca Raton
USA.
Kennish, M.J., 2002. Environmental threats and environmental future of estuaries.
Environ. Conserv. 1, 78–107.
Kellnreiter, F., Pockberger, M., Asmus, H., 2012. Seasonal variation of assemblage and
feeding guild structure of sh species in a boreal tidal basin. Estuar. Coast Shelf Sci.
108, 97–108.
Kordubel, K., Martínez-Rinc´
on, R.O., Baschek, B., Boersma, M., Hieronymi, M., Johns, D.
G., Kirstein, I.V., Voynova, Y.G., M˝
oller, K.O., 2024. Long-term changes in
spatiotemporal distribution of Noctiluca scintillans in the southern North Sea.
Harmful Algae. https://doi.org/10.1016/j.hal.2024.102699.
Kühl, H., Kuipers, B.R., 1983. Qualitative food relationships of Wadden Sea shes. In:
Dankers, N., Wolff, W.J., Zijlstra, J.J. (Eds.), Fishes and Fisheries of the Wadden Sea.
Balkema, Rotterdam, pp. 122–123.
Liu, C., Liu, G., Yang, Q., Luo, T., He, P., Franzese, P.P., Lombardi, G.V., 2021. Emergy-
based evaluation of world coastal ecosystem services. Water Res. 204, 117656.
Lobry, J., David, V., Pasquaud, S., Lepage, M., Sautour, B., Rochard, E., 2008. Diversity
and stability of an estuarine trophic network. Mar. Ecol. Prog. Ser. 358, 13–25.
Lotze, H.K., 2005. Radical changes in the Wadden Sea fauna and ora over the last 2000
years. Helgol. Mar. Res. 59, 71–83.
Lotze, H.K., 2007. Rise and fall of shing and marine resource use in the Wadden Sea,
southern North Sea. Fish. Res. 87, 208–218.
MacKenzie, B.R., Schiedek, D., 2007. Daily ocean monitoring since the 1860s shows
record warming of northern European seas. Global Change Biol. 13, 1335–1347.
Meyer, J., Kr¨
oncke, I., Bartholom¨
a, A., Dippner, J.W., Schückel, U., 2016. Long-term
changes in species composition of demersal sh and epibenthic species in the Jade
area (German Wadden Sea/Southern North Sea) since 1972. Estuar. Coast Shelf Sci.
181, 284–293.
Middelburg, J.J., 2014. Stable isotopes dissect aquatic food webs from the top to the
bottom. Biogeosciences 11, 2357–2371.
Mulas, A., Bellodi, A., Cannas, R., Cau, A., Cuccu, D., Marongiu, M.F., Porcu, C.,
Follesa, M.C., 2015. Diet and feeding behaviour of longnosed skate Dipturus
oxyrinchus. J. Fish. Biol. 86, 121–138.
Pauly, D., 1995. Anecdotes and the shifting baseline syndrome in sheries. Trends Ecol.
Evol. 10, 430.
Philippart, C.J.M., Beukema, J.J., Cad´
ee, G.C., Dekker, R., Goedhart, P.W., van Iperen, J.
M., Leopold, M.F., Herman, P.M.J., 2007. Impacts of nutrient reduction on coastal
communities. Ecosystems 10, 96–119.
Phillips, D.L., Inger, R., Bearhop, S., Jackson, A.L., Moore, J.W., Parnell, A.C., Ward, E.J.,
2014. Best practices for use of stable isotope mixing models in food-web studies.
Can. J. Zool. 92, 823–835.
Poiesz, S.S.H., Witte, I.IJ., van der Meer, M.T.J., van der Veer, H.W., Soetaert, K.E.R.,
2021a. Trophic structure and resource utilization of the coastal sh community in
the western Wadden Sea; evidence from stable isotope data analysis. Mar. Ecol. Prog.
Ser. 677, 115–128.
Poiesz, S.S.H., Witte, J.IJ., van der Meer, M.T.J., Jager, Z., Soetaert, K.E.R., van der
Heide, T., van der Veer, H.W., 2023. Stomach content and stable isotopes illustrate
large spatial similarity in the Wadden Sea sh food-web structure. Mar. Ecol. Prog.
Ser. 707, 57–76.
Poiesz, S.S.H., Witte, J.IJ., van der Veer, H.W., 2020. Only a few key prey species fuel a
temperate coastal sh food web. Mar. Ecol. Prog. Ser. 653, 153–166.
Poiesz, S.S.H., Witte, J.IJ., van der Veer, H.W., 2021b. Historical trophic ecology of some
divergent shark and skate species in the Dutch coastal North Sea zone. Mar. Biol.
168, 165.
Postma, H., 1954. Hydrography of the Dutch Wadden Sea. Arch. Neerl. Zool. 10,
405–511.
Postma, H., Rommets, J.W., 1970. Primary production in the Wadden Sea. Neth. J. Sea
Res. 4, 470–493.
R Core Team, 2021. R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. URL. http://www.R-project.
org/.
Reynolds, J.D., Dulvy, N.K., Goodwin, N.B., Hutchings, J.A., 2005. Biology of extinction
risk in marine shes. Proc. R. Soc. A B. 272, 2337–2344.
Roberts, C., 2007. The unnatural history of the sea. The Past and Future of Humanity and
Fishing. Gaia. Octopus Publishing Group, London.
S.S.H. Poiesz et al.

Selleslagh, J., Lobry, J., Amara, R., Brylinski, J.M., Bo¨
et, P., 2012. Trophic functioning of
coastal ecosystems along an anthropogenic pressure gradient: a French case study
with emphasis on a small and low impacted estuary. Estuar. Coast Shelf Sci. 112,
73–85.
Swartzman, G., Huang, C., Kaluzny, S., 1992. Spatial analysis of Bering Sea groundsh
survey data using generalized additive models. Can. J. Fish. Aquat. Sci. 49,
1366–1378.
Tsutaya, T., Takahash, T., Omori, T., Yamazaki, K., Sato, T., Yoneda, M., Schulting, R.J.,
Kato, H., Weber, A.W., 2021. Reconstruction of diachronic changes in human shing
activity and marine ecosystems from carbon and nitrogen stable isotope ratios of
archaeological sh remains. Quat. Int. https://doi.org/10.1016/j.
quaint.2021.12.005.
Tulp, I., Bolle, L.J., Rijnsdorp, A.D., 2008. Signals from the shallows: in search of
common patterns in long-term trends in Dutch estuarine and coastal sh. J. Sea Res.
60, 54–73.
Tulp, I., Bolle, L.J., Meesters, E., de Vries, P., 2012. Brown shrimp abundance in
northwest European coastal waters from 1970 to 2010 and potential causes for
contrasting trends. Mar. Ecol. Prog. Ser. 458, 141–154.
Tulp, I., van der Veer, H.W., Walker, P.A., van Walraven, L., Bolle, L.J., 2017. Can guild-
or site-specic contrasts in trends or phenology explain the changed role of the
Dutch Wadden Sea for sh? J. Sea Res. 127, 150–163.
van Aken, H.M., 2008a. Variability of the salinity in the western Wadden Sea on tidal to
centennial time scales. J. Sea Res. 59, 121–132.
van Aken, H.M., 2008b. Variability of the water temperature in the western Wadden Sea
on tidal to centennial time scales. J. Sea Res. 60, 227–234.
van der Veer, H.W., Dapper, R., Henderson, P.A., Jung, A.S., Philippart, C.J.M., Witte, J.
IJ., Zuur, A.F., 2015. Changes over 50 years in sh fauna of a temperate coastal sea:
degradation of trophic structure and nursery function. Estuar. Coast Shelf Sci. 155,
156–166.
van der Veer, H.W., van Raaphorst, W., Bergman, M.J.N., 1989. Eutrophication of the
Dutch Wadden Sea: external nutrient loadings of the Marsdiep and vliestroom basin.
Helgol. Meeresunters. 43, 501–515.
van Raaphorst, W., de Jonge, V.N., 2004. Reconstruction of the total N and P inputs from
the IJsselmeer into the western Wadden Sea between 1935–1998. J. Sea Res. 51,
109–131.
van Raaphorst, W., van der Veer, H.W., 1990. The phosphorus budget of the Marsdiep
tidal inlet (Dutch Wadden Sea) in the period 1950 - 1985: importance of the
exchange with the North Sea. Hydrobiologia 195, 21–38.
Walker, P.A., Heessen, H.J.L., 1996. Long-term changes in rays populations in the North
Sea. ICES J. Mar. Sci. 53, 1085–1093.
Wickham, H., 2009. Ggplot2. Springer, New York, New York, NY.
Wheeler, A., 1978. The Fishes of the British Isles and North-West Europe. Frederick
Warne, London.
Whitehouse, G.A., Buckley, T.W., Danielson, S.L., 2017. Diet compositions and trophic
guild structure of the eastern Chukchi Sea demersal sh community. Deep Sea Res.
Part II 135, 95–110.
Wiltshire, K.H., Dürselen, C.-D., 2004. Revision and quality analyses of the Helgoland
Reede long-term phytoplankton data archive. Helgol. Mar. Res. 58, 252–268.
Wolff, W.J., 1983. Ecology of the Wadden Sea. Balkema Press, Rotterdam, The
Netherlands.
Zeller, D., Froese, R., Pauly, D., 2005. On losing and recovering sheries and marine
science data. Mar. Pol. 29, 69–73.
Zijlstra, J.J., 1972. On the importance of the Waddensea as a nursery area in relation to
the conservation of the southern North Sea shery resources. Symp. Zool. Soc. Lond.
29, 233–258.
Zijlstra, J.J., 1983. The function of the Wadden Sea for the members of its sh-fauna. In:
Dankers, N., Wolff, W.J., Zijlstra, J.J. (Eds.), Fishes and Fisheries of the Wadden Sea.
Balkema, Rotterdam, pp. 20–25.
S.S.H. Poiesz et al.