Spatial changes in water depth
Sampling of fish took place during daylight and more than four hours to sample half the number of stations. Thus, sampling at each of the stations generally occurred at different moments of the tidal cycle and under different meteorological conditions. Besides the influence of tides on water depth, transects differed within sampling stations as it was difficult to always maintain one transect due to strong tidal currents and the influence of wind velocity and direction. Therefore, towing directions and transects within the sampling areas differed and covered the edges, slopes, or centers of the tidal channels. Consequently, there were changes in water depth within each sampling station (Fig. 3). For more information on the significant differences in water depth between stations, see Supplementary Information 1.
Species frequency of occurrence and relative abundances
In total, 55 fish species were found during the survey period and showed strong inter- and intra-annual abundance fluctuations. Some species were present all year round, others occurred only occasionally or were season-specific, while others were rare and present only in some years. Thus, only 22 species accounted for more than 95% of the total abundance for the entire survey. Table 1 and Table 2 show the details of the seasonal frequency of occurrence and the fluctuating abundances in both habitats. Herring (Clupea harengus Linnaeus 1758) was the dominant species while the second and subsequent rankings slightly differed between the habitats. Small sand eel (Ammodytes tobianus Linnaeus, 1758) was ranked second in the benthic habitats with 70% occurrence in summer while in the pelagic habitats, it was third-ranked with high occurrences in spring (Tables 1 and 2). Sprat (Sprattus sprattus Linnaeus, 1758) was ranked third in the benthic habitats with high occurrence in spring followed by summer (Table 1). It was ranked second in the pelagic habitats with high occurrence in spring followed by winter (Table 2).
The sequence of species organization in terms of highly ranked to the least of the 22 abundant species differed per season. Clupea harengus was dominant in both habitats in all seasons. In the benthic habitats, A. tobianus was ranked second in spring and summer, sand goby (Pomatoschistus minutus Pallas 1770) in winter, and whiting (Merlangius merlangus Linnaeus, 1758) in autumn (Fig. 4). In the pelagic habitats, S. sprattus was ranked second in winter, summer and autumn, and A. tobianus in spring. Third and subsequent ranks differed per season (Fig. 4 and Fig. 5). The Relative Abundance Curves (RACs) show seasonal changes in species distributions and evenness. The steep gradients between first ranked to the second-ranked species in all habitats in spring and summer indicate uneven communities (Figs. 4 and 5). In the pelagic habitats, the relatively low steepness in winter and spring indicates a moderately even community as species abundances were almost in a similar range (Fig. 5).
Additional analyses of the seasonal changes in community structure showed that all winters compared to all summers had the highest percentage dissimilarity; ANOSIM, R = 0.63, p = 0.0001, SIMPER = 67.55%. The two seasons were 55% similar in terms of species presence and absence (Jaccard’s coefficient = 0.55). The dissimilarities between other seasonal comparisons were lower, SIMPER ranged from 55–58.49%, ANOSIM, R < 0.3, p < 0.001, and Jaccard’s coefficient was relatively higher and ranged from 0.62–0.79. In all seasonal community structure comparisons, six dominant species, C. harengus, A. tobianus, M. merlangus, S. sprattus, P. minutus, and Nilsson's pipefish (Syngnathus rostellatus Nilsson 1855) contributed the highest percentage dissimilarities. See Table S1 for the details of the percentage contributions of different taxa to the dissimilarities between seasons.
Table 1
The total abundance, the percentage of each species to the total abundance, seasonal mean abundances, and the percentage seasonal frequency of occurrence of the 22 species in the benthic habitats for the entire survey. Spp_code show the species names in other analyses
Species | Total abundance | % of the total abundance | Seasonal mean abundance | % Seasonal frequency of occurrence |
Common names | Scientific names | Spp_code | Spring | Summer | Autumn | Winter | Spring | Summer | Autumn | Winter |
Herring | Clupea harengus | C_har | 818841 | 67.6 | 1091.2 | 1989.7 | 368.1 | 137.6 | 30 | 56 | 10 | 4 |
Small sand eel | Ammodytes tobianus | A_tob | 176934 | 14.6 | 180.8 | 540.8 | 47.6 | 2.7 | 23 | 70 | 6 | 0 |
Sprat | Sprattus sprattus | S_spr | 69791 | 5.8 | 160.4 | 75.1 | 50.7 | 16.2 | 53 | 25 | 17 | 5 |
Whiting | Merlangius merlangus | M_mer | 48971 | 4.0 | 10.6 | 137.0 | 71.0 | 0.7 | 5 | 63 | 32 | 0 |
Sand goby | Pomatoschistus minutus | P_min | 35714 | 3.0 | 58.7 | 15.0 | 51.3 | 39.3 | 36 | 9 | 31 | 24 |
Nilsson's pipefish | Syngnathus rostellatus | S_ros | 11717 | 1.0 | 15.4 | 17.5 | 16.6 | 2.4 | 30 | 34 | 32 | 5 |
Plaice | Pleuronectes platessa | P_pla | 7346 | 0.6 | 5.4 | 13.4 | 7.3 | 8.0 | 16 | 39 | 21 | 23 |
Great sand eel | Hyperoplus lanceolatus | H_lan | 5304 | 0.4 | 14.6 | 7.4 | 0.4 | 0.0 | 65 | 33 | 2 | 0 |
Bull-rout | Myoxocephalus scorpius | M_sco | 4764 | 0.4 | 2.7 | 1.4 | 2.5 | 18.3 | 11 | 6 | 10 | 74 |
Common goby | Pomatoschistus microps | P_mic | 4602 | 0.4 | 4.1 | 0.2 | 4.6 | 14.4 | 17 | 1 | 20 | 62 |
Three-spined stickleback | Gasterosteus aculeatus | G_acu | 3934 | 0.3 | 4.6 | 4.5 | 1.0 | 8.9 | 24 | 24 | 5 | 47 |
Hooknose | Agonus cataphractus | A_cat | 3715 | 0.3 | 3.5 | 2.0 | 5.9 | 6.4 | 20 | 11 | 33 | 36 |
Dab | Limanda limanda | L_lim | 3331 | 0.3 | 4.5 | 0.8 | 6.4 | 3.8 | 29 | 5 | 41 | 25 |
Cod | Gadus morhua | G_mor | 2923 | 0.2 | 0.4 | 7.0 | 4.5 | 1.6 | 3 | 52 | 33 | 12 |
Smelt | Osmerus eperlanus | O_epe | 2667 | 0.2 | 3.0 | 1.2 | 1.7 | 7.3 | 23 | 9 | 13 | 56 |
Eelpout | Zoarces viviparus | Z_viv | 2642 | 0.2 | 2.8 | 3.2 | 0.8 | 6.1 | 22 | 25 | 6 | 48 |
European anchovy | Engraulis encrasicolus | E_enc | 1768 | 0.1 | 0.2 | 0.5 | 7.4 | 0.0 | 3 | 7 | 91 | 0 |
Fluonder | Platichthys flesus | P_fle | 1278 | 0.1 | 2.1 | 0.4 | 0.6 | 3.0 | 35 | 7 | 11 | 45 |
Horse mackerel | Trachurus trachurus | T_tra | 885 | 0.1 | 0.0 | 3.8 | 0.1 | 0.0 | 0 | 96 | 4 | 0 |
Gunnel | Pholis gunnellus | P_gun | 814 | 0.1 | 1.6 | 0.7 | 0.5 | 1.0 | 43 | 19 | 12 | 26 |
Striped seasnail | Liparis liparis | L_lip | 639 | 0.1 | 1.6 | 0.8 | 0.2 | 0.2 | 59 | 30 | 6 | 6 |
Scaldfish | Arnoglossus laterna | A_lat | 261 | 0.0 | 0.1 | 0.0 | 1.1 | 0.0 | 7 | 3 | 89 | 1 |
Table 2
The total abundance, the percentage of each species to the total abundance, overall seasonal mean abundances, and the percentage seasonal frequency of occurrence of the 19 abundant species in the pelagic habitats for the entire survey period
Species | Total abundance | % of the total abundance | Seasonal mean abundances | % seasonal frequency of occurrence |
Common names | Scientific names | Spring | Summer | Autumn | Winter | Spring | Summer | Autumn | Winter |
Herring | Clupea harengus | 300829 | 87.9 | 10087 | 12621 | 1423 | 401 | 29 | 22 | 25 | 23 |
Sprat | Sprattus sprattus | 18472 | 5.4 | 331 | 518 | 685 | 111 | 36 | 14 | 22 | 28 |
Small sand eel | Ammodytes tobianus | 10867 | 3.2 | 634 | 45 | 206 | 2 | 57 | 23 | 12 | 7 |
Great sand eel | Hyperoplus lanceolatus | 3282 | 1.0 | 245 | 5 | 5 | 0 | 54 | 23 | 22 | 1 |
Nilsson's pipefish | Syngnathus rostellatus | 2436 | 0.7 | 33 | 50 | 131 | 9 | 30 | 28 | 33 | 9 |
Three-spined stickleback | Gasterosteus aculeatus | 1764 | 0.5 | 45 | 21 | 22 | 71 | 35 | 9 | 21 | 35 |
Horse mackerel | Trachurus trachurus | 1752 | 0.5 | 0 | 143 | 3 | 0 | 0 | 82 | 18 | 0 |
Sand goby | Pomatoschistus minutus | 716 | 0.2 | 32 | 8 | 6 | 14 | 25 | 7 | 26 | 43 |
Smelt | Osmerus eperlanus | 496 | 0.1 | 17 | 2 | 6 | 19 | 20 | 11 | 16 | 53 |
Whiting | Merlangius merlangus | 451 | 0.1 | 0 | 23 | 18 | 0 | 4 | 65 | 31 | 0 |
Plaice | Pleuronectes platessa | 197 | 0.1 | 9 | 4 | 1 | 2 | 27 | 30 | 16 | 27 |
Common goby | Pomatoschistus microps | 193 | 0.1 | 1 | 0 | 2 | 16 | 11 | 0 | 26 | 63 |
Fluonder | Platichthys flesus | 159 | 0.0 | 9 | 1 | 0 | 3 | 49 | 8 | 8 | 36 |
Bull-rout | Myoxocephalus scorpius | 92 | 0.0 | 3 | 0 | 1 | 3 | 30 | 11 | 19 | 41 |
Hooknose | Agonus cataphractus | 82 | 0.0 | 3 | 1 | 2 | 2 | 22 | 22 | 19 | 38 |
Garfish | Belone belone | 65 | 0.0 | 4 | 1 | 0 | 0 | 41 | 47 | 12 | 0 |
Lumpfish | Cyclopterus lumpus | 63 | 0.0 | 1 | 0 | 4 | 2 | 16 | 3 | 63 | 19 |
Dab | Limanda limanda | 62 | 0.0 | 1 | 0 | 5 | 0 | 40 | 13 | 40 | 7 |
Eelpout | Zoarces viviparus | 49 | 0.0 | 2 | 0 | 1 | 1 | 46 | 8 | 15 | 31 |
Temporal changes in fish diversity
The diversity indices, species richness (S), evenness (J), dominance (D), and Shannon-Wiener Index (H) varied both spatially and temporally over the study period. In general, there were higher taxa numbers in the benthic habitats compared to pelagic habitats (Fig. S1). Species richness (S) in benthic habitats was the highest in spring and summer. In contrast, S in pelagic habitats was low in summer (Fig. S1). Inter-annual differences in S were observed with high variabilities in both benthic and pelagic habitats (numerous outliers in Fig. S1).
The General Additive Model (GAM) comprising smoothing terms of water temperature, sampling station and the factors year and season explained 15% and 24% of the variations in S in the benthic and pelagic habitats, respectively. Water temperature was significant and explained 4% and 11% variations in S in the benthic and pelagic habitats, respectively (Table 3 and Table 4). Smoothing curves show higher S in temperatures between 4°C to 7°C and 13°C to 16°C in both habitats and remain stable when temperatures are above 19°C (Fig. 6). However, S ranged from 2 to 15 species (mean of 7.6±3.1 (SD), n = 100) between 19°C and 22.6°C in the benthic habitat and from 1 to 7 species (mean of 2.6±1.5 (SD), n = 42) in the pelagic habitats (Fig. S2). Thus, the wide 95% confidence bands indicate high uncertainty in S prediction when water temperature is > 19°C. Numerical output of different GAMs for both habitats are summarized in Table 3 and Table 4, best models are shown in bold. The tables show various information on the performance of various models in explaining the variations in diversity indices. For instance, deviance explained which is equivalent to R2 in linear regressions and the Akaike Information Criterion (AIC). The GAMs show that the factor year explained higher variability in S at 7% (benthic) and 10% (pelagic) habitats compared to the seasonal effect that explained only 2% and 3%.
Table 3
Estimated parameters, standard errors, t-values, p-values, deviance explained, scale estimates, and Akaike Information Criterion (AIC) of the parametric components of the general additive models of the explanatory variables to the diversity indices in benthic habitats. Best models are shown in bold, s() represents the smoothing terms of water temperature
Diversity index | Explanatory variables | Estimate/ Intercept | SE | t-value | p-value | Deviance explained | Scale est. | AIC |
Species richness (S) | s(Temperature) + Year + Season + Station | 7.47 | 0.51 | 14.57 | *** | 15% | 10.8 | 4385.1 |
s(Temperature) + Year + Season | 8.11 | 0.43 | 18.95 | *** | 13% | 10.9 | 4390.8 |
| s(Temperature) + Year | 9.16 | 0.36 | 25.46 | *** | 10% | 11.2 | 4405.6 |
| s(Temperature) + Season | 6.88 | 0.26 | 26.70 | *** | 6% | 11.6 | 4429.3 |
| s(Temperature) | 7.84 | 0.12 | 65.84 | *** | 4% | 11.8 | 4441.4 |
| Year + Season | 5.99 | 0.41 | 21.01 | *** | 8% | 11.4 | 4419.1 |
| Year | 9.40 | 0.36 | 26.12 | *** | 7% | 11.5 | 4426.8 |
| Season | 7.13 | 0.24 | 29.69 | *** | 2% | 12.1 | 4458.2 |
| Station | 7.28 | 0.32 | 22.60 | *** | 2% | 12.1 | 4465.6 |
Evenness (J) | s(Temperature) + Year + Season + Station | 0.56 | 0.03 | 16.69 | *** | 17% | 0.0 | -167.2 |
| s(Temperature) + Year + Season | 0.53 | 0.03 | 18.87 | *** | 15% | 0.0 | -169.2 |
| s(Temperature) + Year | 0.47 | 0.02 | 19.98 | *** | 13% | 0.0 | -156.3 |
| s(Temperature) + Season | 0.49 | 0.02 | 29.97 | *** | 12% | 0.0 | -169.2 |
| s(Temperature) | 0.44 | 0.01 | 58.08 | *** | 11% | 0.0 | -157.5 |
| Year + Season | 0.49 | 0.03 | 18.03 | *** | 11% | 0.0 | -139.5 |
| Year | 0.46 | 0.02 | 19.18 | *** | 3% | 0.1 | -79.7 |
| Season | 0.46 | 0.02 | 30.22 | *** | 8% | 0.0 | -134.2 |
| Station | 0.47 | 0.02 | 22.20 | *** | 1% | 0.1 | 69.9 |
Dominance (D) | s(Temperature) + Year + Season + Station | 0.43 | 0.04 | 11.39 | *** | 20% | 0.1 | 25.2 |
| s(Temperature) + Year + Season | 0.42 | 0.03 | 13.47 | *** | 18% | 0.1 | 29.5 |
| s(Temperature) + Year | 0.45 | 0.03 | 17.43 | *** | 17% | 0.1 | 27.3 |
| s(Temperature) + Season | 0.51 | 0.02 | 26.83 | *** | 11% | 0.1 | 76.2 |
| s(Temperature) | 0.54 | 0.01 | 62.30 | *** | 10% | 0.1 | 74.2 |
| Year + Season | 0.43 | 0.03 | 14.22 | *** | 15% | 0.1 | 55.1 |
| Year | 0.45 | 0.03 | 16.46 | *** | 9% | 0.1 | 103.0 |
| Season | 0.53 | 0.02 | 29.93 | *** | 7% | 0.1 | 103.1 |
| Station | 0.55 | 0.02 | 22.42 | *** | 2% | 0.1 | 156.4 |
Shannon (H) | s(Temperature) + Year + Season + Station | 1.22 | 0.08 | 14.54 | *** | 22% | 0.3 | 1361.3 |
| s(Temperature) + Year + Season | 1.25 | 0.07 | 17.75 | *** | 19% | 0.3 | 1374.3 |
| s(Temperature) + Year | 1.22 | 0.06 | 20.90 | *** | 19% | 0.3 | 1369.8 |
| s(Temperature) + Season | 1.01 | 0.04 | 23.54 | *** | 12% | 0.3 | 1425.5 |
| s(Temperature) | 0.98 | 0.02 | 49.87 | *** | 11% | 0.3 | 1424.0 |
| Year + Season | 1.23 | 0.07 | 18.06 | *** | 15% | 0.3 | 1406.4 |
| Year | 1.24 | 0.06 | 20.45 | *** | 9% | 0.3 | 1454.5 |
| Season | 0.97 | 0.04 | 24.27 | *** | 7% | 0.3 | 1457.7 |
| Station | 0.96 | 0.05 | 17.43 | *** | 2% | 0.4 | 1508.9 |
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 |
Table 4
Estimated parameters, standard errors, t-values, p-values, deviance explained, scale estimates, and Akaike Information Criterion (AIC) of the parametric components of the general additive models of the explanatory variables to the diversity indices in the pelagic habitats. Best models are shown in bold, s() represents the smoothing terms of water temperature
Diversity index | Explanatory variables | Estimate/ Intercept | SE | t-value | p-value | Deviance explained | Scale est. | AIC |
Species richness (S) | s(Temperature) + Year + Season + Station | 4.65 | 0.34 | 13.65 | *** | 24% | 3.74 | 1831.91 |
s(Temperature) + Year + Season | 5.08 | 0.29 | 17.32 | *** | 19% | 3.90 | 1845.48 |
| s(Temperature) + Year | 5.08 | 0.29 | 17.38 | *** | 19% | 3.90 | 1842.11 |
| s(Temperature) + Season | 3.86 | 0.10 | 38.91 | *** | 11% | 4.18 | 1863.83 |
| s(Temperature) | 3.87 | 0.10 | 39.45 | *** | 11% | 4.18 | 1861.08 |
| Year + Season | 5.31 | 0.35 | 15.11 | *** | 14% | 4.12 | 1863.96 |
| Year | 5.30 | 0.30 | 17.58 | *** | 10% | 4.27 | 1876.53 |
| Season | 3.89 | 0.21 | 18.17 | *** | 3% | 4.49 | 1889.15 |
| Station | 3.55 | 0.22 | 15.96 | *** | 4% | 4.50 | 1892.58 |
Evenness | s(Temperature) + Year + Season + Station | 0.71 | 0.04 | 16.21 | *** | 17% | 0.06 | 48.27 |
(J) | s(Temperature) + Year + Season | 0.65 | 0.04 | 17.46 | *** | 13% | 0.06 | 58.18 |
| s(Temperature) + Year | 0.65 | 0.04 | 17.11 | *** | 9% | 0.07 | 70.51 |
| s(Temperature) + Season | 0.76 | 0.03 | 26.78 | *** | 9% | 0.06 | 51.46 |
| s(Temperature) | 0.66 | 0.01 | 53.84 | ** | 6% | 0.07 | 60.62 |
| Year + Season | 0.69 | 0.05 | 15.23 | *** | 5% | 0.07 | 81.18 |
| Year | 0.64 | 0.04 | 16.65 | *** | 3% | 0.07 | 86.05 |
| Season | 0.70 | 0.03 | 26.67 | *** | 2% | 0.07 | 73.54 |
| Station | 0.70 | 0.03 | 25.82 | *** | 3% | 0.07 | 70.78 |
Dominance | s(Temperature) + Year + Season + Station | 0.54 | 0.05 | 11.15 | *** | 15% | 0.08 | 147.67 |
(D) | s(Temperature) + Year + Season | 0.51 | 0.05 | 10.08 | *** | 12% | 0.08 | 150.36 |
| s(Temperature) + Year | 0.56 | 0.04 | 13.61 | *** | 11% | 0.08 | 150.84 |
| s(Temperature) + Season | 0.61 | 0.01 | 44.49 | *** | 8% | 0.08 | 148.81 |
| s(Temperature) | 0.61 | 0.01 | 44.84 | *** | 6% | 0.08 | 149.15 |
| Year + Season | 0.52 | 0.05 | 10.47 | *** | 8% | 0.08 | 163.78 |
| Year | 0.56 | 0.04 | 13.24 | *** | 6% | 0.08 | 166.45 |
| Season | 0.59 | 0.03 | 19.91 | *** | 2% | 0.08 | 166.53 |
| Station | 0.59 | 0.03 | 19.33 | *** | 2% | 0.09 | 170.97 |
Shannon | s(Temperature) + Year + Season + Station | 0.86 | 0.08 | 10.21 | *** | 20% | 0.23 | 627.33 |
(H) | s(Temperature) + Year + Season | 0.94 | 0.09 | 10.63 | *** | 18% | 0.24 | 627.85 |
| s(Temperature) + Year | 0.87 | 0.07 | 12.08 | *** | 17% | 0.24 | 626.28 |
| s(Temperature) + Season | 0.71 | 0.02 | 29.29 | *** | 12% | 0.25 | 633.68 |
| s(Temperature) | 0.70 | 0.02 | 29.50 | *** | 11% | 0.25 | 633.15 |
| Year + Season | 0.95 | 0.09 | 10.95 | *** | 12% | 0.25 | 649.46 |
| Year | 0.89 | 0.07 | 11.96 | *** | 8% | 0.26 | 661.48 |
| Season | 0.75 | 0.05 | 14.33 | *** | 4% | 0.27 | 662.26 |
| Station | 0.72 | 0.06 | 13.02 | *** | 1% | 0.28 | 680.01 |
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 |
Evenness (J) varied both annually and seasonally in both habitats. Evenness was higher in winter and spring than in summer in most of the years. Evenness was higher in the pelagic habitats than in the benthic habitats (Fig. S3). Water temperature explained 11% and 6% variations in J in the benthic and pelagic habitats, respectively (Table 3 and Table 4). In the benthic habitats, the smoothing terms of water temperature show that J decreases with increasing temperature up to 15°C followed by a minor increase. This is followed by a decline at temperatures above 18°C (Fig. 7, a). In the pelagic habitats, J declines with increasing temperatures up to 8°C then remains constant up to 11°C. This is followed by a further decrease to 15°C then an increase up to 19°C after which it remains constant (Fig. 7, b). The factor season explained a higher variability in J (8%) than the factor year (4%) in the benthic habitats (Table 3). In the pelagic habitat, both factors year and season explained low variability in J at 3% and 2%, respectively (Table 4).
Dominance (D) varied both annually and seasonally in both habitats. The highest values were found in most summers. However, in 2015 spring and autumn had higher values than summer in the benthic habitat while in the pelagic habitat, springs of 2009 and 2017 showed relatively higher values (Fig. S3). In the benthic habitats, smoothing terms showed that D is unstable at temperatures below 5°C then gradually increases up to 14°C after which it remains stable to 18°C. This is followed by an increase at temperatures above 19°C (Fig. 8, a). A similar pattern occurred in the pelagic habitat. However, the wide 95% confidence interval indicates the uncertainties in D at temperatures above 16°C (Fig. 8, b). Water temperature explained 10% and 6% variations in D in the benthic and pelagic habitats, respectively. The factor year explained 9% and 6% of the variability in D in benthic and pelagic habitats, respectively. The factor season on the other hand explained 7% variability in benthic and only 2% in the pelagic habitats (Tables 3 and 4).
Shannon-Wiener Index (H) varied annually and across seasons in both habitats. Generally, H was high in winter and low in summer. However, in 2017 spring showed the highest H values in the benthic habitat, and springs of 2012, 2016, and 2017 had the highest values in the pelagic habitats (Fig. S4). Smoothing terms of water temperature showed a gradual increase in H from 1°C obtaining a maximum value at 5°C in both benthic and pelagic habitats (Fig. 9), which represents an equal proportion of all species at this temperature. As temperatures increase beyond 5°C, H gradually declines in all the habitats. However, in the benthic habitat, H was stable between 14°C and 18°C after which further decline occurs at higher temperatures. Water temperature explained 11% variability in H in both habitats. The factor year explained 9% and 8% variations in H while season explained 7% and 4% of the same in the benthic and pelagic habitats, respectively (Table 3 and Table 4).
Spatial distribution of fish with changes in water depth
The tidal effects and changes in water depth within sampling stations on the distribution of fish were analyzed using generalized linear mixed-effects models (GLMMs). The different slopes and intercepts showed the distribution patterns in different sampling stations with changes in water depth. Generally, S decreased with increasing water depth in all sampling stations in the benthic habitats even though sampling was always slightly above the sediments irrespective of the water depth. The intercepts and slopes show minimal variations in S (variance (σ2) = 0.59; sampling station, and σ2 = 0.004; depth) between sampling stations in the benthic habitats regardless of whether the station was located in or adjacent to the deep tidal channels (Fig. 10, a). In the pelagic habitats, variations in S with water depth between sampling stations (σ2 = 1.14 (sampling station) and σ2 = 0.01 (depth) were observed. For instance, S decreased with water depth in Sylt_6 while Sylt_1 and Sylt_9 showed positive relationships between S and water depth (Fig. 10, b).
Evenness (J) increased with water depth in all sampling stations in the benthic habitats with no variations between stations (same intercept and slope, Fig. 11, a). In the pelagic habitats, J varied between sampling stations. For instance, Sylt_2 and Sylt_6 showed positive relationships between J and water depth while no depth effects were observed in Sylt_8. In the stations adjacent to deep tidal channels (i.e. Sylt_1, Sylt_4, and Sylt_9), J decreased with an increase in water depth (Fig. 11, b). Dominance on the other hand showed a slight decrease with an increase in water depth in all sampling stations in all habitats (Fig. S5). Individuals of various species utilized the submerged areas at high water depths resulting in equal abundance proportions in the sampled areas. There were minor variations in D with water depth between sampling stations. For instance, in the benthic habitats, Sylt_2 and Sylt_4 showed relatively steeper slopes than the other stations. Only Sylt_2 in the pelagic habitat showed such steepness (Fig. S5). There were no significant effects of water depth on H.