Inter- and Intra-Estuarine Comparison of the Feeding Ecology of Keystone Fish Species in the Elbe and Odra Estuaries

Abstract

Food webs in estuarine ecosystems serve as important biological indicators. The feeding ecology of four keystone fish species, pikeperch (Sander lucioperca L.), smelt (Osmerus eperlanus L.), ruffe (Gymnocephalus cernua L.) and flounder (Platichthys flesus L.), in the Elbe and Odra estuaries was analyzed using stomach content analyses. Important prey of pikeperch were fishes and mysids in both estuaries. Amphipods were especially important as prey for smelt in the Elbe estuary, whereas smelt caught in the Odra estuary mainly consumed mysids. Ruffe fed mainly on amphipods in the Elbe estuary, while annelids (lower section) and insect larvae (upper section) were the most important prey in the Odra estuary. Flounder favored copepods as prey in the Elbe estuary, while bivalves were preferred in the Odra estuary. Higher dietary overlaps were found in the Elbe estuary between smelt vs. ruffe, pikeperch vs. ruffe, and pikeperch vs. smelt. In the Elbe estuary, a shift in the diet composition of pikeperch, smelt, and ruffe was observed from 2021 to 2022 compared to food analyses from the 1990s. These shifts included an increased consumption of amphipods, while mysids and copepods had recently decreased in their diets. These changes indicate a restructuring of the food web, potentially linked to environmental changes, which highlights the sensitivity of estuarine ecosystems.

1. Introduction

Fishes are important elements in estuarine food webs, since they link the lower trophic levels with higher trophic levels such as other piscivorous fish, birds [1], marine mammals, or humans, when used commercially. Based on the relationships known to date, key fish species in the estuarine food web are considered suitable indicators for assessing the condition of estuarine ecosystems [2,3,4,5,6], as parameters such as their diet composition, nutritional intensity, growth rate, abundance, and spatial distribution depend on the availability of suitable habitats and food resources. Appropriate food resources availability at each specific stage of their life cycle is crucial for individual development and stock recruitment.

In estuarine ecosystems, multiple mechanisms shape the composition of food webs. First, prey availability and habitat differences play an important role, influenced by salinity gradients, substrate types, primary production, and detritus inputs [7,8,9]. Seasonal variability, including temperature changes and varying riverine freshwater discharge, affects the transport of prey organisms and triggers migration and reproductive cycles of prey species [10,11]. Interspecific competition also plays a significant role in structuring food webs [12]. Lastly, anthropogenic factors, such as habitat destruction, eutrophication, and pollution, can significantly alter prey composition in estuarine ecosystems [13,14].

However, the impacts of climate change can disrupt the timing of the availability of suitable food sources due to phenological mismatches. This occurs when the timing of phyto- and zooplankton blooms are misaligned with the hatching of fish larvae [15]. Other human-induced factors such as the loss of shallow water areas, hypoxic events, river dredging, increased turbidity, and contaminations, are believed to impair estuarine functions [5,16,17,18,19,20,21]. With recent shifts in the environmental factors, the fish abundance in the Elbe estuary, one of Europe’s largest estuaries, was reduced by over 90% [22], which included massive reductions in smelt, ruffe, and flounder. The reasons for this decline are multifactorial, including habitat destruction by river construction leading to increased turbidity, hypoxia events, changes in discharge dynamics, and a reduction in shallow water areas [22,23,24]. The Odra estuary is part of the Baltic Sea, where stocks of commercially important fish species such as cod (Gadus morhua) and herring (Clupea harengus) have been collapsing in recent years, indicating environmental stressors on the ecosystem, thus affecting the food web. The declines have been driven by a combination of climate change, nutrient loads, and overfishing [15,25,26,27]. Temperature is a key environmental factor influencing foraging behavior and prey–predator interactions, particularly in ectotherms. Climate change-induced temperature fluctuations affect metabolic rates, prompting animals to adjust their foraging preferences and efforts to meet their energy demands [28]. Recognizing this dynamic is crucial for understanding shifts in estuarine food webs under changing environmental conditions. Given the environmental challenges facing the Elbe and Odra estuaries, including climate-induced temperature shifts, habitat degradation, and anthropogenic pressures, this study provides insights into the feeding ecology of keystone fish species, ultimately influencing food web dynamics.

Smelt (Osmerus eperlanus L.), ruffe (Gymnocephalus cernua L.), flounder (Platichthys flesus L.), and pikeperch (Sander lucioperca L.) are among the keystone fish species in the Elbe estuary. These species occur throughout the entire longitudinal stretches of the Elbe estuary [22,29,30,31,32] and thus serve as key elements of the food web in the Elbe estuary. In the Oder estuary, Lorenz [33] revealed that smelt, ruffe, and pikeperch belonged to the dominant species in the Odra Lagoon, while according to Thiel et al. [34], flounder was among the dominant species in the coastal areas of the Pomeranian Bay. Therefore, these four species were collected for diet analyses.

To gain deeper knowledge of the feeding ecology of fishes in different temperate estuaries, this study was conducted on four keystone fish species from two temperate estuaries, the Odra as a microtidal estuary in the southern Baltic Sea and the Elbe as a mesotidal estuary in the south-eastern North Sea. This study aims to (1) analyze and compare their diet breadth and overall food composition in the two different estuaries, (2) identify spatial and seasonal trends in their food composition within each of the two estuaries, and (3) estimate the dietary overlap between pairs of the keystone fish species.

2. Materials and Methods

2.1. Study Area

The Elbe estuary is located in north-west Germany and extends from the weir at Geesthacht downstream to its mouth into the southern North Sea close to the city of Cuxhaven (Figure 1, left). It is strongly influenced by tides and is classified as a mesotidal coastal plain estuary with no marked halocline [32,35,36]. In the freshwater region or upper section (Mühlenberger Loch: ML, Twielenfleth: TF), the salinity is below 0.5 PSU, which stretches from the weir at Geesthacht (Ekm 586) to Stade (Ekm 654). In the brackish oligohaline area or middle section (Schwarztonnensand: ST), down to Brokdorf at Ekm 684, the salinity ranges between 0.5 and 5 PSU, and the brackish mesohaline area or lower section (Brunsbüttel: BB, Medemgrund: MG) continues downstream to the city of Cuxhaven at Ekm 725 [37].

The Odra estuary is located in north-east Germany and drains to the Baltic Sea (Figure 1, right) with very low influences from the tide with tidal oscillation below 20 cm [38]; it can therefore be classified as a microtidal estuary. We categorized the Odra estuary into three sections: the upper section (Odra Lagoon: OL), with salinities ranging between 1 and 3 PSU, the middle section (Greifswald Bodden: GB), with a salinity range of 6–7 PSU, and the lower section (Pomeranian Bay: PB), with a salinity of 5–8 PSU [39,40].

2.2. Collection of Sample Material

We collected samples throughout the year, at least once per season: spring 2021 (May/June 2021), summer (August 2021), autumn (November 2021), winter (February/March 2021), and spring 2022 (May/June 2022). Fish samples were collected from commercial fishermen using the following catching techniques: stow net, fish traps, gill nets, trawl net. Fish were frozen at −20 °C immediately after catching or, at the latest, after landing, for preparation in the laboratory.

2.3. Analysis of Fish Diet

Diet analysis was carried out following Debus and Winkler [41] and Thiel [42]. The stomachs of the fishes were opened and the abundance, length, and weight of the prey species were determined. Through digestive processes, the biomass of the consumed prey organisms is reduced; hence, the actual consumed biomass at the moment of feeding cannot be determined by weighing. Therefore, the biomass was reconstructed by regression formulas following Debus and Winkler [41] as found in the literature (Table S1). Fragments were measured with a stereomicroscope Leica MZ 9.5 (±0.01 mm) combined with Leica Flexacam C1 (Leica LAS X software version 3.7.4.23463). From the fragment length, the total length and weight of the organism were calculated. Fragments or prey items that exceeded the maximum length of the stereomicroscope (>20 mm) were measured using millimeter paper. If no regression formula was available for an organism, standard weight was used. Whenever no regression formula and no standard weight was available, the available standard weight in the literature (Table S1) of the taxonomically and morphologically closest relative was used [41].

For species identification of consumed fish, characteristic bony structures (vertebral columns, singular vertebra, mandibular, pharyngeal teeth, otoliths) were used. Fish severely disintegrated by digestion were x-rayed (Faxitron MultiFocus) to portray identification features such as number of vertebrae or number of fin rays. The taxonomic identification of consumed prey objects was determined by characteristic features found in the literature: refs. [43,44,45,46,47,48,49,50,51,52] for fish and [53,54,55,56,57,58,59,60,61] for invertebrates. The identified prey organisms were assigned to taxonomic groups for later analyses (Gastropoda, Bivalvia, Annelida, Copepoda, Cladocera, Isopoda, Amphipoda, Thecostraca, Mysida, Decapoda, Insecta, Fish).

2.4. Calculation of Parameters in Feeding Ecology

For the characterization of the food composition, the following three parameters were calculated [62]: (I) Abundance (N_i): the abundance of prey types (i) to the number of all prey individuals, (II) frequency of occurrence (F_i): number of fish that ingested one prey type i in relation to the total number of fished which were analyzed, and (III) biomass (B_i): compares one prey individual’s weight per type i to the total weight of all prey individuals in percent.

$$N_i(\%)={\displaystyle \frac{n_i}{n}}\times 100$$

(1)

$${ F}_i\left(\%\right)={\displaystyle \frac{N_i}{N}}\times 100$$

(2)

$$B_i\left(\%\right)={\displaystyle \frac{{RB}_i}{RB}}\times 100$$

(3)

n = Amount of prey items

i = Number of prey types i

N = Number of fishes with full stomachs

N_i = Abundance of prey type i

RB = Reconstructed prey biomass

RB_i = Proportion of prey type i in reconstructed prey biomass

Single index approaches for describing the diet can be biased [63], since they fail to capture the significance of each food component within the diet. The relative importance of prey taxa varies when only using single indices such B_i, N_i, or F_i alone. Small prey such as copepods occur in large numbers and are therefore represented with larger abundances (N_i) in the diet, while larger prey such as fish represent a larger biomass (B_i) although they occur in lower quantities. Using only the abundance of prey would overestimate the importance of the small prey ingested. To address this limitation, the Index of Relative Importance (IRI) proposed by George and Hadley [64] was calculated. The IRI combines multiple dietary metrics, providing a more comprehensive and accurate assessment of the importance of food components in the diet.

$$IRI \left(\%\right)={\displaystyle \frac{(B_i+N_i+F_i)}{\sum (B_i+N_i+F_i)}}\times 100$$

The diet breadth (DB) indicates whether a species is a feeding generalist or specialist and was calculated following Levins [65]. The value of the DB can range from 1 to n. A high DB corresponds with higher diversity in the diet, and when DB is equal to 1 only one type of food was consumed. The Levins’ DB not only provides information on the number of food components consumed but also about the distribution or evenness of the diet.

$$DB={\displaystyle \frac{1}{\sum _{i=1}^n{pi}^2}}$$

$pi$ = Proportion of prey item $i$

$n$ = Total number of prey types

Dietary overlap (C_xy) was calculated with Schoener’s [66] equation.

$$C_{xy}=\sum _{i=1}^n\mathrm{m}\mathrm{i}\mathrm{n}(p_{xi};p_{xy})$$

(4)

p_xi; p_xy = Proportion of usage (IRI) for the resource i by the fish species x and y

n = Number of prey species

2.5. Statistical Analysis

Non-metric multidimensional scaling (NMDS) was conducted on untransformed diet composition data (IRI%) from each station using Bray–Curtis dissimilarities for the four keystone fish species. From the individual specimens, subsamples were created for each station: for smelt, 10 specimens per subsample, for pikeperch, 2 specimens per subsample, for ruffe, 5 specimens per subsample, and for flounder, 3 specimens per subsample. Analysis of similarities (ANOSIM) and later post hoc pairwise ANOSIM were conducted (999 permutations) to identify differences between groups (stations). For the pairwise comparisons, p-values were corrected via the Hochberg adjustment. Similarity percent (SIMPER) was conducted in order to determine the contribution of each prey taxon to the variation found in the NMDS ordination. The NMDS, ANOSIM, and SIMPER analyses were performed using the R-package ‘vegan’ (R version 4.1.2) [67].

3. Results

3.1. Food Composition and Diet Breadth

In the Elbe estuary, smelt (n = 643) with a size range from 75 to 226 mm, pikeperch (n = 84) with a size range from 76 to 190 mm, ruffe (n = 199) with a size range from 50 to 220 mm, and flounder (n = 287) with a size range from 24 to 383 mm were used for dietary analyses. In the Odra estuary, smelt (n = 102) from 76 to 210 mm, pikeperch (n = 121) from 105 to 219 mm, ruffe (n = 194) from 72 to 198 mm, and flounder (n = 118) from 84 to 339 mm were used. The highest diet breadth (DB) was estimated for flounder (6.8), followed by smelt (5.6), pikeperch (4.9), and ruffe (3.2). In the Odra, estuary flounder (6.5) was again the keystone species with the highest DB, followed by pikeperch (4.9), ruffe (4.2), and smelt (3.0). Pikeperch and flounder had similar DB in both estuaries, while for smelt the DB in the Elbe estuary was larger than in the Odra estuary. In contrast, the DB of ruffe was larger in the Odra estuary than in the Elbe estuary.

For pikeperch from the Elbe estuary (

Table 1. Food composition (% IRI) of keystone fish species in the Elbe estuary. Single prey components were categorized by their affiliation to higher groups (mesozooplankton, macrozoobenthos, zoobenthos, nekton). Higher groups in capitalized letters display the sums of the individual related prey taxa.

	Pikeperch	Smelt	Ruffe	Flounder
Diet breadth (B)	4.9	5.6	3.2	6.8
MESOZOOPLANKTON		11.77		34.36
COPEPODA		9.82		34.36
Eurytemora affinis		4.77		16.59
Calanoida indet.				17.77
Cyclopoida		0.20
Copepoda—eggs		2.16
Copepoda—spermatophores		0.75
Copepoda indet.		1.94
CLADOCERA		1.95
Daphnia galatea		0.78
Daphnia longispina		0.56
Daphnia spp.		0.50
Bosmina longirostris		0.11
Cladocera indet.		0.08
MAKROZOOPLANKTON and	51.56	70.60	91.62	34.58
mainly HYPOZOOBENTHOS MYSIDA	23.65	19.40	23.11	17.01
Neomysis integer	23.06	15.82	21.19	17.01
Mesopodopsis slabberi	0.59	3.58	1.92
AMPHIPODA	13.03	38.84	49.57	10.93
Gammarus spp.	13.03	35.61	48.43	10.93
Corophium volutator		3.23	1.14
INVERTEBRATE EGGS		0.76		2.89
DECAPODA	14.88	11.60	18.94	3.75
Crangon crangon	5.13	6.21	13.15	3.75
Palaemon longirostris	9.75	4.67	4.49
Eriocheir sinensis		0.72	1.30
ENDOZOOBENTHOS and mainly EPIZOOBENTHOS		1.30	1.85	24.67
BIVALVIA			0.20	0.59
Mya sp.				0.59
Pisidium spp.			0.20
ISOPODA		0.46	0.17	8.00
Idotea balthica				6.56
Idotea emarginata		0.07	0.17
Idotea granulosa		0.09
Sphaeroma sp.				1.44
Isopoda indet.		0.30
ANNELIDA		0.72		2.19
Hediste diversicolor		0.48
Annelida indet.		0.24
Tubificidae				2.19
INSECTA		0.12	1.48	13.89
Chaoborus sp.		0.07
Chironomidae		0.05	0.63	13.89
Insecta indet.			0.85
NEKTON (Fish)	48.42	16.27	6.50	6.34
Osmerus eperlanus	34.13	11.79	6.50
Clupea harengus	6.27	3.27
Sprattus sprattus	1.49
Clupeidae	0.56
Gymnocephalus cernua	2.99	0.52
Sander lucioperca		0.05
Pomatoschistus microps		0.06
Pomatoschistus minutus	1.49	0.58		1.44
Merlangius merlangus				0.65
Platichthys flesus	1.49
Fish Eggs				4.25

), fish as prey displayed 48.42% in total (index of relative importance = IRI, all following values referring the food composition are displayed as IRI). Most important prey species were smelt (34.31%) and herring (6.27%), while sprat (1.49%), ruffe (2.99%), and sand goby (1.44%) were of lesser importance. The second most important prey taxon were mysids (23.65%), mainly displayed by Neomysis integer (23.06%). Amphipods displayed a moderately important prey taxon with 13.06% (Gammarus spp.), together with decapods (14.88%) (Crangon crangon: 5.13% and Palaemon longirostris: 9.75%).

Amphipods (38.84%) were the most important prey organisms for smelt, which was mainly displayed by Gammarus spp. (35.61%). The second most important food taxon for smelt were other fishes (16.26%), which consisted mainly of other smelt (11.79%, cannibalism) and herring Clupea harengus (3.27%). The third most important prey taxon were mysids, displayed mainly by Neomysis integer (15.82%). Copepods were also an important prey of smelt (9.82%), which mainly consisted of Eurytemora affinis. In smaller amounts, other benthic invertebrates such as the annelid Hediste diversicolor and isopods (Idotea emarginata, Idotea granulosa) with 0.46% were consumed by smelt.

For ruffe, amphipods (49.57%) were the most important prey organisms (Gammarus spp.: 48.43%). These were followed by mysids (23.11%), consisting mainly of Neomysis integer (21.19%), decapods with 18.94% (Crangon crangon: 13.15%, Palaemon longirostris: 4.49%, Eriocheir sinensis: 1.30%), and fish with 6.50% (smelt). Insect larvae (Chironomidae: 0.63%, insect larvae indet. 0.85%) were of minor importance for ruffe from the Elbe estuary.

For flounder, copepods (34.35%) were the most important prey taxon, which were followed by mysids with 17.01% (Neomysis integer), insect larvae with 13.89% (Chironomidae), amphipods with 10.93% (Gammarus spp.), isopods 8.00% (Idotea balthica: 6.56%, Sphaeroma sp.: 1.44%), fish 4.90% (fish-eggs: 4.25%, Merlangius merlangus: 0.65%). Bivalves were of minor importance 0.59% (Mya sp.) for flounder from the Elbe estuary and gastropods were not consumed at all.

In the Odra estuary (

Table 2. Food composition (% IRI) of keystone fish species in the Odra estuary. Single prey components were categorized by their affiliation to higher groups (mesozooplankton, macrozoobenthos, zoobenthos, nekton). Higher groups in capitalized letters display the sums of the individual related prey taxa.

	Pikeperch	Smelt	Ruffe	Flounder
Diet breadth (B)	4.9	3.0	4.2	6.5
MESOZOOPLANKTON			2.15
CLADOCERA			2.15
Daphnia longispina			2.15
MAKROZOOPLANKTON and mainly HYPOZOOBENTHOS	41.70	69.50	24.80	15.31
MYSIDA	15.27	53.03	2.64	2.46
Neomysis integer	15.27	53.03	2.64	2.46
AMPHIPODA	7.68	2.07	19.5	4.54
Gammarus spp.	7.68	2.07	11.95	4.54
Corophium volutator			7.55
INVERTEBRATE EGGS		0.49
DECAPODA	18.75	13.91	2.66	8.31
Crangon crangon	18.09	13.91	1.30	4.02
Palaemon longirostris	0.66
Eriocheir sinensis			1.36
Carcinus maenas				4.29
ENDOZOOBENTHOS and mainly EPIZOOBENTHOS	3.40	7.44	72.82	84.72
GASTROPODA	0.43	0.36	2.32	11.71
Potamopyrgus jenkinsi	0.43
Hydrobiidae		0.36	2.32	11.71
BIVALVIA	0.74	0.40	2.70	48.03
Mya arenaria	0.74
Mya sp.				30.91
Mytilus edulis				7.95
Cerastoderma edule				8.85
Mysella bidentata		0.40
Limecola balthica				0.32
Bivalvia indet.			2.70
THECOSTRACA				0.53
ISOPODA		1.49	1.56	3.77
Idotea balthica			0.40	3.77
Idotea granulosa			1.16
Idotea chelipes		1.49
ANNELIDA	1.75	5.19	25.50	13.58
Hediste diversicolor	1.75	5.19	25.50	13.58
INSECTA	0.48		40.74	7.10
Chironomidae	0.48		39.85	7.10
Insecta indet.			0.89
NEKTON (Fish)	54.89	23.07	0.28
Osmerus eperlanus			0.28
Clupea harengus	1.75
Gymnocephalus cernua	0.58
Pomatoschistus minutus		14.30
Pomatoschistus spp.	17.15	8.77
Neogobius melanostomus	0.97
Alburnus alburnus	0.76
Rutilus rutilus	33.68

), pikeperch fed mostly on other fishes, with a proportion of 54.89% consisting mainly of Rutilus rutilus (33.68%) and Pomatoschistus spp. (17.15%). The second most important prey taxon was displayed by decapods (18.75%), which consisted almost exclusively of Crangon crangon (18.09%). Mysids with 15.27% were exclusively displayed by Neomysis integer. Amphipods, displayed by Gammarus spp. (7.68%), were of minor importance in the diet of pikeperch from Odra estuary.

The most important prey for smelt in the Odra estuary were mysids, with an IRI of 53.03% (Neomysis integer), followed by fishes, which were composed of Pomatoschistus minutus (14.30%) and Pomatoschistus spp. (8.77%). Decapods occurred at 13.91% (Crangon crangon), while annelids (Hediste diversicolor: 5.19%) and amphipods (Gammarus spp. 2.07%) were less important prey taxa for smelt.

In the Odra estuary, insect larvae were the most important food component for ruffe. The insects were mainly displayed by larvae from the family Chironomidae (38.85%). The second and third most important food components of ruffe were annelids, which were represented by Hediste diversicolor (25.5%) and amphipods with 19.5% (Gammarus spp.: 11.95%, Corophium volutator 7.55%).

For flounder in the Odra estuary, bivalves (48.03%) were the most important food component (Mya sp.: 30.91%, Mytilus edulis: 7.95%, Cerastoderma edule: 8.85%). The second most important prey organisms were annelids (Hediste diversicolor: 13.58%), followed by gastropods (Hydrobiidae 11.71%) and decapods with 8.31% (Crangon crangon: 4.02%, Carcinus maenas: 4.29%).

3.2. Spatio-Temporal Variability of Food Composition Within the Estuaries

3.2.1. Pikeperch

Polygons representing the stations in the NMDS ordination showed overlap for the samples from all stations in the diet composition of pikeperch (Figure 2). The pairwise ANOSIM indicated that significant differences existed solely between stations ST (Elbe) and GB (Odra) (R = 0.721, p = 0.028). In the Elbe estuary, the proportion of piscivorous diet of pikeperch was largest in the lower and upper sections (MG: 75.9%, ML: 59.9%), while in the middle section (ST: 12.5%) the amount of fish was smallest (Figure 3). In the middle section, the importance of mysids (47.6%) and amphipods (30.8%) was substantially greater compared to the other regions in the Elbe estuary. In the Odra estuary, the proportion of fish in the diet of pikeperch was largest in the middle section (GB: 81.8%), compared to the upper section (OL: 59.7%) and lower section (PB: 35.7%). In contrast, the proportion of decapods in the diet of pikeperch was higher in the lower section (35.2%) of the Odra estuary, while in the upper (4.1%) and middle section (2.1%), the proportions of decapods were substantially smaller. In both estuaries, the proportion of fish in the diet of pikeperch was higher during the summer compared to the autumn samples (Figure 3). In autumn, the proportion of decapods and mysids increased in both estuaries, whereby in the Odra estuary the general diversity of ingested prey taxa increased.

3.2.2. Smelt

The NMDS ordination of the diet composition of smelt revealed that most stations in the ordination space overlap, except for the PB station which is orientated separately from the other stations (Figure 4). Significant dissimilarities in food composition were found at the stations in between the two estuaries: BB vs. PB (R = 0.659, p = 0.012), ML vs. PB (R = 0.733, p = 0.012), ST vs. PB (R = 0.703, p = 0.012), and TF vs. PB (R = 0.684, p = 0.012). In the Elbe estuary, copepods were absent in the diet of smelt caught at the lower section (MG). The proportion of copepods in the diet of smelt gradually increased in the upstream direction: BB (1.5%), ST (7.3%), TF (15.3%), ML (29.2%). Amphipods reached the highest proportion of the IRI in the middle section at ST (47.5%). The importance of mysids in the diet of smelt was largest at the lower sections (MG: 25.3%, BB 26.7%). Isopods (MG: 0.6%, BB: 1.6%) and annelids (MG: 1.0%, BB: 1.3%) were exclusively found in the diet of smelt from the lower sections. Cladocerans were exclusively found in samples from the middle and upper sections of the Elbe estuary. In the Odra estuary smelt could only obtained from the lower section (PB), so the overall description of the food composition corresponds to “3.1 Food composition and diet breadth”. In the Elbe estuary, seasonal differences in the food composition of smelt were found. In particular, during spring (45.5%) and autumn (41.5%), the percentages of amphipods were higher in the diet of smelt. In spring and autumn samples (Figure 5), amphipods were the dominant prey, while mysids dominated in summer samples with 40.0%, and copepods were the most important prey in winter (43.5%). In the Odra estuary, samples from summer and autumn were available. Notably, the amount of fish in the diet of smelt was larger in summer (49.7%) compared to autumn (14.7%). On the contrary, mysids were less important as a food resource during summer and had increased later in the autumn season.

3.2.3. Ruffe

The NMDS ordination shows that samples from the Elbe estuary were grouped in clusters, while being separated from samples of the Odra estuary. Samples from the Odra estuary were less overlapping compared to samples from the Elbe estuary (Figure 6). Pairwise ANOSIM shows significant dissimilarities between all stations between Elbe and Odra estuaries: BB vs. GB (R = 0.781, p = 0.036), BB vs. PB (R = 0.913, p = 0.016), BB vs. OL (R = 0.755, p = 0.028), ST vs. GB (R = 0.848, p = 0.016), TF vs. GB (R = 0.702, p = 0.049), ML vs. GB (R = 0.624, p = 0.016) ML vs. PB (R = 0.926, p = 0.028), ML vs. OL (R = 0.659, p = 0.049), ST vs. PB (R = 0.947, p = 0.016), TF vs. PB (0.856, p = 0.049), ST vs. OL (R = 0.834, p = 0.016), TF vs. OL (R = 0.605, p = 0.049). Within the Elbe estuary, ML vs. ST (R = 0.305, p = 0.049), was the only comparison that showed significant dissimilarities, while within the Odra estuary, PB vs. OL (R = 0.784, p = 0.016), and GB vs. PB (R = 0.663, p = 0.036) were significantly different. For ruffe in the Elbe estuary, amphipods were the most important prey organisms at all stations, while in the upstream regions, the IRI of amphipods was slightly elevated (ML: 63.1%, TF: 55.8%). Fish in the diet of ruffe increased alongside the salinity gradient, with highest IRI in the upstream regions (ML: 15.1%, TF: 10.5%). Insect larvae were only found in small amounts in the middle (2.4%) and upper (2.0%) region of the Elbe estuary in the diet of ruffe.

Decapods were important prey organisms from the middle to the lower sections. In the Odra estuary, insect larvae were the most important food resource of ruffe. Here, the IRI was highest in the upper region (OL: 66.9%) and declined alongside the salinity gradient, with 39.9% in the middle region (GB) and 4.5% in the lower region (PB). In contrast, the most important prey in the lower region were annelids, which decreased along the salinity gradient when moving upstream: middle region (21.4%), upper region (no annelids found in the diet). Amphipods were the most important prey organisms during all four seasons (Figure 7), although in slightly lower amounts during summer (46.8%) and autumn (45.0%) compared to spring (54.5%) and winter (53.9%). Seasonal differences in the diet of ruffe were most pronounced in the winter season, while in all other season, mysids were an important part of the diet; during winter, no mysids were found. Ruffe consumed other fish throughout the year, except during winter. In the Odra estuary, the amount of insect larvae was particularly lower in autumn (32.4%), while in the same season, the proportion of annelids (37.4%) was highest. Amphipods were consumed throughout all seasons, and the proportion was largest in winter (34.9%) and smallest during summer (6.7%).

3.2.4. Flounder

Food composition reflected by the NMDS ordination showed that most samples of flounder were partly overlapping while some were not. In particular, samples from the Odra estuary (PB and GB) were separated in the ordination space (Figure 8). Considerable differences were found between stations in the Elbe and Odra estuaries: BB vs. PB (R = 0.958, p = 0.018), MG vs. GB (R = 0.448, p = 0.011), ML vs. GB (R = 0.633, p = 0.016), MG vs. PB (R = 0.876, p = 0.011), and ML vs. PB (R = 0.954, p = 0.032). Within the Elbe estuary, MG vs. ML (R = 0.598, p = 0.018) and MG vs. TF (R = 0.641, p = 0.032) revealed significant dissimilarities, while in the Odra estuary dissimilarities of GB vs. PB (R = 0.405, p = 0.018) were significant. For flounder in the Elbe, the proportion of copepods in the diet was smallest in the lower regions of the estuary (MG: 4.8%, BB: 3.0%). In the upstream direction, the proportion of copepods successively increased: ST (37.5%), TF (50.0%), and ML (67.3%). Insects in the diet were only found in the upstream region (TF: 38.3%, ML 25.0%), while mysids were found in the diet of flounder caught at the middle and lower regions of the estuary. In the lower region at BB, the mysids displayed a proportion of 48.5%. Isopods were found in small amounts in all regions, except for the most upstream station, ML. In the Odra estuary, in the lower region, bivalves were the dominant food resource (PB: 64.3%). In the middle section, the food composition was not as dominated by one food resource compared to the lower region (Figure 6). In the upper regions of the Odra estuary (OL), no flounder were captured, hence no feeding data were available. At GB, bivalves (25.5%) were the most important prey, followed by annelids (23.7%) and insects (17.0%). The proportion of mysids in both regions was comparatively small, with 3.9% in the lower region and 0.5% in the middle region. Isopods were of larger importance for flounder from the middle section (7.6%) compared to the lower section (1.0%). The diet composition of flounder from the Elbe estuary displayed high seasonal variability (Figure 9). During spring (47.1%) and winter (48.0%), copepods were the most important prey organisms. The proportion of mysids was substantially larger in spring (22.8%), summer (42.6%), and autumn (21.6%) than in winter (3.6%). Insect larvae were predominantly found in flounders from autumn, with an IRI of 53.7%, compared to summer (12.5%), and with no insects during spring and winter. Amphipods, in contrast, were only found in Elbe flounder caught in spring (15.7%) and winter (17.4%). In the Odra estuary, high seasonal variability was found in the diet of flounder. Bivalves were predominantly found in flounders from summer (63.0%) and autumn (54.8%). Isopods were an important food resource during winter (26.8%) compared to the other seasons (spring: 3.7%, summer: 3.8%, autumn: 1.1%). Annelids were especially important prey organisms during spring (45.6%) and winter (31.7%), while decapods were particularly important during autumn (18.2%).

3.3. Dietary Overlap in the Elbe and Odra Estuaries

Generally, the dietary overlap between the keystone species was substantially larger in the Elbe estuary than in the Odra estuary (Figure 10). This was true for all species combinations except for ruffe/flounder (Elbe Cxy = 0.325, Odra Cxy = 0.319) and pikeperch/smelt (Elbe Cxy = 0.553, Odra Cxy = 0.472), in which the dietary overlap was only slightly larger in the Elbe compared to the Odra. A high similarity in diet composition was found between ruffe and smelt (Cxy = 0.703), which was the highest calculated dietary overlap for the keystone fish species compared in pairs in the Elbe estuary. In the Oder estuary, the dietary overlap between the keystone fish species, especially ruffe and smelt, was substantially smaller (Cxy = 0.119).

4. Discussion

4.1. Spatial Feeding Variability

Notably, pikeperch consumed less fish and more invertebrates (amphipods, mysids, decapods) at station ST, compared to the other stations. The reduced consumption of fish at ST may in turn be a consequence of the lower food availability of fish (smelt), especially in the middle region of the Elbe estuary. The preference of pikeperch for smelt as a prey species of pikeperch in the Elbe estuary is consistent with findings of Kafemann and Thiel [68]. In a previous study, we found that smelt densities were much lower in the middle region (ST: 6905 individuals per 1 Mio m³) compared to the upper region (ML: 32,307 individuals per 1 Mio m³) and the lower region (MG: 30,307 individuals per 1 Mio m³) [22]. Koll et al. [69] showed that conditions and food supply were unfavorable for juvenile pikeperch in the middle estuary. At the station ST, the condition factor (Fultons’ body condition factor K) was lower compared to the other stations, and pikeperch expressed starvation on a molecular level (transcriptomes). This also reflects the findings of van Densen [70], who described a large growth difference between piscivorous and non-piscivorous juvenile pikeperch. Juvenile pikeperch appear to be flexible predators that specialize in different prey taxa, preferably fish, depending on the food resources of the ecosystem they inhabit. Although a piscivore diet is important for juvenile pikeperch regarding their survival and growth, the lack of preferred prey organisms can lead to juvenile pikeperch feeding predominantly on less nutritious taxa such as mysids (Neomysis integer) and amphipods (Gammarus spp.).

The food composition of smelt in the Elbe and Odra estuaries differed primarily in terms of the proportions of fish, mysids, and gammarids. The smelt in the Elbe estuary consumed fewer fish (Elbe estuary: IRI = 16.26%, Odra estuary: IRI = 23.07%), fewer mysids (Elbe estuary: IRI = 19.4%, Odra estuary: IRI = 53.03%), and more gammarids (Elbe estuary: IRI = 35.61%, Odra estuary: 2.07%) than in the Odra estuary. In smelt, a gradual change in diet along the longitudinal gradient of the Elbe estuary was observed, with copepods becoming more important upstream, while the importance of mysids declined. Eurytemora affinis displays the most abundant copepod species in the Elbe estuary and was found in the highest densities in the freshwater zone in southern marginal areas close to ML [71]. Temperature has a larger affect on the reproductive parameters of Eurytemora affinis compared to salinity, as population densities were described to decrease drastically when water temperature exceeds 20° C [72,73]. The primary production in the maximum turbidity zone (MTZ) of the Elbe estuary was low, with a year-round low quality of particulate organic matter (POM) [74,75]. As a result, there is limited food available for Eurytemora affinis in this region. The high turbidity results in rather unfavorable living conditions for phototrophic organisms, where phagotrophy provides crucial advantages for mixotrophic flagellates [76], leading to trophic lengthening [75]. Enriched δ¹⁵N in juvenile smelt was found in the MTZ but not in adults, which means that the adults avoid areas with inappropriate environmental conditions and limited food availability (Hauten 2024, [75]). In the diet of smelt in the Eru Bay (Gulf of Finland), amphipods and mysids were the most common prey, while fish in the diet were important prey organisms, particularly for larger smelt from August to October [77]. In the Gulf of Riga, juvenile smelt strongly selected the cladoceran Bosmina longispina as prey. It was concluded that the fish may switch to consume other prey organisms when the preferred food is limited [78].

Fish in the diet of ruffe increased with lower salinities, with the highest IRI observed in the upstream regions. Insect larvae were only found in small amounts in the diet of ruffe in the middle and upper regions of the Elbe estuary. Decapods were important prey organisms from the middle to the lower sections but not in the upper estuary. In the Odra estuary, insect larvae were the most important food resource for ruffe. The IRI was highest in the upper region and declined with higher salinities in the middle region (GB). In the lower region (PB) of the Odra estuary, insect larvae were of minor importance in the diet of ruffe. The most important prey in the lower region were annelids, which decreased in the upstream regions. This is due to the fact that insect larvae and the annelid Hediste diversicolor have a certain salinity requirement [79]. In non-tidal estuaries such as the Couronian Lagoon, adult ruffe prey on juvenile smelt and perch, alongside meso- and macrozooplankton. Similarly, in the Darß-Zingst Bodden, ruffe also consume nine-spined sticklebacks (Pungitius pungitius) [80]. In other freshwater habitats (e.g., Lake Võrtsjärv, Lake Aydat), chironomid larvae and pupae were the dominant prey of ruffe [81,82].

In the Elbe estuary, copepods became more prominent in the diet of flounder in the middle and upper regions, likely due to the preference of Eurytemora affinis for the freshwater zone in the southern marginal areas [71]. As flounders are migratory species [83], they move through these regions where copepods are more abundant. In the Odra estuary, bivalves (Mya arenaria) were the dominant prey in the lower region (PB), where the Odra Bank is located. This shoal, characterized by sandy sediments, supports a benthic community dominated by Mya arenaria [84]. A proportion of benthic food components (approx. 50%) has also been determined previously for the Elbe estuary [85] and has also been found in other estuaries, such as the Schlei [86], the Westerschelde [87], the Humber estuary [88], the Severn estuary [89], and the Tagus estuary [90]. This shows that the feeding preference of flounder from the Elbe estuary for copepods and zoobenthic taxa aligns with findings from other European estuaries. In the Odra estuary, no copepods were consumed, while zoobenthic taxa accounted for a higher proportion of the IRI (84.72%), indicating a stronger preference for benthic prey in this system. Flounder in the northern Baltic Sea (Åland archipelago) also showed strong seasonal variations in the diet of flounder [91]. In the south western Baltic Sea, the main prey of flounder was Limecola balthica [92].

Differences in food composition between the two estuaries can be attributed to the prey availability and generalist feeding behavior of the keystone species, since they exploit a wide range of prey resources. Consequently, fluctuations in prey availability across different locations result in distinct food compositions. The Elbe and Odra estuaries are ecologically and hydrologically quite different: The Elbe is a mesotidal estuary with strong tidal influences and high flow velocities, which shape its ecological conditions. In contrast, the Odra estuary is microtidal, with minimal tidal influence. Its lagoon-like structure, rather than a prominent channel, results in larger water residence time in the OL [93]. This creates conditions that support different prey species, whereas in the Elbe estuary, tidal currents and daily mixing result in fluctuations in salinity. In the Odra estuary, for example, bivalves are prevalent on sandbanks, while insects reproduce in large numbers under the oligohaline conditions of the OL.

4.2. Temporal Feeding Dynamics

Seasonal variations in the food composition for pikeperch were more pronounced in the Odra estuary. In the Elbe estuary, food compositions between summer and autumn were quite similar. In the Odra estuary, pikeperch consumed less fish and a larger variety of prey taxa, including benthic invertebrates such as annelids, gastropods, and isopods during autumn. Additionally, the proportion of decapods was substantially larger in autumn. This suggests that the preferred fish species were less available in autumn than in summer. Studies conducted in the Kiel Canal (NOK) by Kafemann and Thiel [68] found that pikeperch experienced reduced growth because the NOK lacked a sufficiently high supply of fish as their preferred food, including herring [94] (Winkler 1980) and smelt in particular [70,95]. In the 2021–2022 period, pikeperch consumed gammarids at nearly all stations (except MG at the river mouth), while gammarids were not found at all in the food of pikeperch in the Elbe estuary during the 1992–1993 period [70].

In the Elbe estuary, seasonal differences in the food composition of smelt were most pronounced in winter, with copepods being the dominant prey. This suggests a significantly higher availability of copepods during late winter. In contrast, in the Odra estuary, fish were the most important prey during summer but were replaced by mysids in autumn. Smelt fed predominantly on gobies from the genus Pomatoschistus, which occur most abundantly during late summer months in the Baltic. Pikeperch [96] showed that the biomass of Pomatoschistus minutus in the Baltic peaked in August/September and decreased substantially later in October. As the availability of Pomatoschistus sp. as a food resource decreased in autumn, smelt shifted their main diet to mysids. Compared to the findings from Thiel [42], a lower proportion of preyed upon smelt were found in the food of smelt (cannibalism) in the Elbe estuary in the 2021–2022 period than in the 1991–1993 period (IRI = 11.79% vs. IRI = 23.74%). Overall, fish made up a substantially lower proportion of the smelt’s diet composition in 2021–2022 than in 1991–1993 (IRI = 16.27% vs. IRI = 26.03%). Similarly to the pikeperch, the proportion of the consumed gammarids in 2021–2022 (IRI = 35.61%) was notably higher than in 1991–1993 (IRI = 3.08%), while the proportion of the Neomysis integer was similarly high during both periods (IRI 2021–2022: 15.82%, IRI 1992/1993: 14.31%). The proportion of mesozooplankton, especially that of calanoid copepods of the species Eurytemora affinis, was also highly variable. The number of copepods in the diet of smelt in the Elbe estuary from 1991 to 1993 was IRI = 11.78% [42], while the percentage during 2021–2022 was significantly smaller (IRI = 4.77%). Ladiges [97] described smelt from the Elbe estuary as planktonic feeders, which are highly dependent on Eurytemora affinis but periodically consume benthic invertebrates. It was further stated that between November and March, benthic organisms were exclusively consumed, mainly displayed by the genus Gammarus, and it was concluded that smelt undergo a hunger period in winter, during which they shift to less desirable prey organisms such as gammarids. In the samples from 2021 to 2022, gammarids were a major food component all year round, indicating a limitation in the preferred food resources for smelt in the Elbe estuary. The primary production in the maximum turbidity zone (MTZ) of the Elbe estuary is low, with year-round low quality of particulate organic matter (POM) [74,75], providing limited food resources for grazing copepods. Hauten [75] found enriched δ¹⁵N in juvenile smelt in the MTZ but not in adults and concluded that the adults avoid areas with inappropriate environmental conditions and limited food availability. The consumption of indigestible objects such as small pieces of wood, stones, plant matter, and sand was observed most frequently during winter in our samples, which was also described by Ladiges [97] during starvation periods.

The seasonal diet variation in ruffe was not as pronounced as in the other keystone species in both estuaries. Across all four seasons, a similar proportion of amphipods was consumed, while in winter, ruffe did not feed on mysids. In the Odra estuary, especially during autumn, ruffe fed on annelids. In summer, ruffe from the Odra estuary fed on cladocerans. Long-term studies in the Baltic Sea have shown that temperature and salinity influence the abundance of cladocerans, and that densities were highest during summer [98]. By comparing the recently analyzed diet of ruffe with previous studies from the Elbe estuary, further indications of a change in food availability were found: Thiel [42] determined very different food compositions for ruffe in this area for the 1991–1993 period. In that period, ruffe from the Elbe estuary consumed a relatively high proportion of copepods of the species Eurytemora affinis (IRI = 26.21%), while in our study, no copepods were found in the diet of ruffe from 2021 to 2022. In the 1990s, gammarids were consumed by ruffe, with an IRI = 7.55%. In the samples from 2021 to 2022, samples were estimated to contain a substantially larger proportion of gammarids (IRI: 48.43%). Amphipods have a sturdy cuticula, which is formed by chitin and calcium (ash, i.e., amount of inorganic non-combustible material), which makes them of lower nutritional value compared to other prey organisms with no cuticula (e.g., fish) or those with less sturdy cuticula (e.g., mysids). The study by Thiel [42] did not report fish in the diet of ruffe, which changed in the samples from 2021 to 2022, where fish (smelt) were ingested by ruffe (IRI = 6.5%).

The diet composition of flounder was the most variable among the keystone fish species, likely due to its generalist feeding behavior [99]. In the Elbe estuary, copepods were a major dietary component during spring and winter. In the Odra estuary, annelids made up a larger proportion of the diet in spring and winter, while bivalves were more dominant in summer and autumn. This seasonal difference may be due to a preference for annelids, as they lack a hard shell, making them easier to digest and a more favorable prey option. Although the prey composition of flounder changed between these two periods, the differences were not as pronounced as for the other species. In both periods, 1991–1993 [42] and 2021–2022, flounder consumed copepods in high proportions in the Elbe estuary (IRI 1991–1993: 41.29%, IRI 2021–2022: 34.36%). The zoobenthos species changed from an IRI of 14.4% (1991–1993) to an IRI of 24.7% (2021–2022). Both groups (copepods, zoobenthic taxa) together accounted for more than 50% of the IRI of flounder during both periods in the Elbe estuary.

Seasonal shifts in salinity, temperature, and freshwater inflow across seasons can induce the reproductive cycles of both predators and prey. The temperature affects the metabolic rate of fish, influencing their feeding requirements [100]. In contrast to the spatial salinity gradient in estuaries, seasonal changes in river discharge affect the salinity gradient and therefore the distribution of biota. The changes in river discharge also affect plankton dynamics [101], as higher discharge increases the downstream transport of planktonic organisms. The salinity gradient in estuaries plays a crucial role in shaping the composition and distribution of biota in estuaries [21,30,42,102]. Generalist feeders opportunistically exploit a wide range of prey resources. As a result, fluctuations in prey availability lead to temporal dietary shifts, with opportunistic predators adapting to seasonally abundant or accessible food sources [103]. The combination of seasonal shifts in temperature, salinity, and river discharge can trigger spawning and recruitment dynamics, which shape the fish composition [21,30,32,104]. In the Elbe estuary, the spawning migration of smelt in late winter is induced by the increased river discharge [105]. By comparing the food composition of keystone fish species from the Elbe estuary with historical data, a shift from a predominantly plankton-based food web to a more benthos-based food web is observed. This transition could be an indication of possible responses to shifting environmental conditions, since the Elbe estuary has a long history of anthropogenic interventions.

Hypoxia has been shown to affect feeding behavior in several fish species [106,107,108], primarily reducing foraging intensities and metabolic rates, which leads to growth slowing. The oxygen situation in the Elbe estuary has been documented by several authors within the last 40 years [21,29,32,109,110]. Especially in the Hamburg Port region (upper estuary), severe hypoxic events have occurred. Although the oxygen situation improved compared to the 1980s [24], the situation has declined in recent years leading to recurring hypoxic events. Especially during years with lower river runoff, areas with low oxygen concentrations have extended upstream to Bunthäuser Spitze at Ekm 609. These hypoxic events during summer months could have impaired the feeding activities of the keystone fish species in the Elbe estuary. Accordingly, changes in feeding strategies can occur under different turbidity conditions, with visual predators being more affected by turbidity than fish that feed on benthic macroinvertebrates [111]. For visual predators, the turbidity of the water columns plays a crucial role in their predation success. Increased turbidity could also be an explanation for the dietary shift towards extended feeding on amphipods. Generally, pikeperch and smelt feed on more pelagic organisms, particularly smelt, which primarily consume calanoid copepods, and pikeperch, which primarily feed on other fish. However, as turbidity increased [24], visual predation might have been impaired, causing a shift towards more benthic foraging. We suspect these shifts in food composition were induced by the availability of prey, with accessibility changing foraging behavior as a result of changed environmental conditions.

4.3. Dietary Overlap

Sharing an abundant food source is not necessarily direct evidence of competition, but it can be caused by high abundances of certain prey organisms, leading to food selection [112]. Schoener’s values that are higher than 0.6 or lower than 0.4 are considered to be of ecological relevance [113,114,115,116,117]. When comparing the dietary overlaps of the key species in the Elbe estuary from 2021 to 2022 to historical data from the 1990s [42], the general overlaps in the 1990s were smaller overall, especially due to lower abundances of amphipods in the diet. The high dietary overlap (0.703) of smelt and ruffe in the Elbe estuary is an indication for relevant competition. The findings from the Elbe estuary indicate either a competition for amphipods due to a lack of other suitable prey taxa or a general high availability of this prey taxon.

5. Conclusions

The differences in the food composition of keystone fish species from Elbe and Odra estuaries were probably primarily influenced by the availability of prey organisms. The preference for a piscivorous diet was clear in pikeperch from both estuaries, although the availability of fish as prey fluctuated within the estuaries. In the Elbe estuary, smelt fed extensively on amphipods, while in the Odra estuary, amphipods played a subordinate role and mysids were the preferred food objects. Ruffe in the Elbe estuary preferred mainly amphipods as prey, while in the Odra estuary, insect larvae and annelids were the preferred food, with preferences changing along the salinity gradient. Flounder had the most diverse diet, especially in the Elbe estuary, with preferences for copepods, mysids, and insect larvae, while in the Odra estuary, bivalves, annelids, and gastropods were preferred. Seasonal differences in food composition were driven by fluctuations in prey availability, leading to temporal dietary shifts as opportunistic predators adapted to seasonally abundant or accessible food sources. This effect was most pronounced in flounder, the most generalist feeders, as reflected by the highest diet breadth (DB) among the keystone species. Compared to investigations from the 1990s, the food composition in the Elbe estuary has shifted, with a notable increase in amphipods as a dietary component in pikeperch, smelt, and ruffe. Therefore, the food composition of keystone fish species in the Elbe estuary showed indications of shifting from a plankton-based (copepods, mysids) food web to a more benthos (amphipods)-orientated system. Together with trophic elongation of the food chain, these shifts could have been induced by changed environmental factors, which may further have ecological impacts on phytoplankton and zooplankton that ultimately affect higher trophic levels, promoting altered foraging behavior. The increased dietary overlap between species in the Elbe estuary indicate an increased competition for existing food resources, which could be the result of limited food availability in the Elbe estuary.