taxonID	type	description	language	source
0393B316414FFFF6E460E762FB2A14C8.taxon	description	The number of identifiable prey items per individual was usually fewer than 100 in all expeditions, with a few exceptions (up to> 1200 individuals) during PS 92 and PS 106 / 2. The stomach contents of B. saida caught during three expeditions were numerically dominated by copepods. Calanus spp. numerically dominated the diet during the two spring expeditions (PS 92 and P 106 / 2), whereas the harpacticoid Tisbe sp. dominated during late summer / autumn (PS 80; Kohlbach et al., 2017) followed by Calanus spp. (Figure 2). In terms of MD, Calanus spp. was also the dominant food item during expeditions PS 92 and PS 106. During PS 80, MD of the stomach contents were dominated by the ice-associated amphipod A. glacialis (Table 4). High numbers of amphipods were found in the stomachs of fish collected only during this expedition. Appendicularians were not common in the stomachs of fish from expeditions other than PS 122. They accounted for 1.3 % of the total recognizable food items in fish from PS 92, although with a relatively high FO. Chaetognaths were also a relatively common food item during expedition PS 92, occurring in 66.7 % of the stomachs, albeit in low numbers. Krill (Euphausiidae) were only found in the stomach contents of fish collected during PS 106 / 2 (Figure 2; Table 4). During PS 80, 27.5 % of the stomachs were infested with the trematode parasite Hemiurus levinseni (Kohlbach et al., 2017), which is usually hosted by calanoid copepods (KØie, 2009). No parasites were found in the stomachs of fish from PS 92. From the fish of PS 106 / 2, one stomach contained two trematode parasites, and one stomach contained a single nematode parasite. Identifiable prey items in B. saida stomachs from PS 122 were limited to four taxonomic groups, namely copepods (62.5 % FO), appendicularians (56.3 % FO), chaetognaths (42.8 % FO), and amphipods (25 % FO). Two of the 16 individuals had empty stomachs, in one of which a trematode parasite was found. On average, appendicularians numerically dominated the diet of B. saida during the summer / autumn months (PS 122 / 4 – 5) of PS 122 (53.8 %) followed by copepods (17.9 %; Figure 2), whereas copepods, appendicularians, and chaetognaths dominated the diet in terms of MD (30.5 %, 30.5 %, and 20.9 %, respectively; Table 4). Identifiable copepods consisted mainly of the genus Calanus (Table 4). Of the amphipods, A. glacialis and T. abyssorum could be identified to species level. The stomach of the single winter individual, collected during PS 122, contained little food (Figure 2; Table 4), and the few food items that were recognizable comprised 80 % copepods and 20 % chaetognaths numerically. The IRI (Figure 3) indicated that amphipods (IRI: 70.5 %) and copepods (IRI: 29.3 %) dominated by the species A. glacialis and Tisbe sp. were the most important food items during PS 80 (Kohlbach et al., 2017). Copepods (IRI: 92 %), dominated by Calanus spp., were the most important food item during PS 92. During PS 106 / 2, copepods (IRI: 49.9 %), again dominated by Calanus spp., were most important, followed by euphausiids, including Thysanoessa sp. (IRI: 45.8 %). Appendicularians were the most important food item for B. saida during the summer of PS 122 (IRI: 59.0 %) followed by copepods (IRI: 25.7 %) and chaetognaths (IRI: 11.4 %). Detailed IRI on the lowest identified taxonomic levels can be found in Table 4.	en	Schaafsma, Fokje L., Flores, Hauke, David, Carmen L., Castellani, Giulia, Sakinan, Serdar, Meijboom, André, Niehoff, Barbara, Cornils, Astrid, Hildebrandt, Nicole, Schmidt, Katrin, Snoeijs-Leijonmalm, Pauline, Ehrlich, Julia, Ashjian, Carin J. (2024): Insights into the diet and feeding behavior of immature polar cod (Boreogadus saida) from the under-ice habitat of the central Arctic Ocean. Journal of Fish Biology 105 (3): 907-930, DOI: 10.1111/jfb.15836, URL: https://doi.org/10.1111/jfb.15836
0393B316414FFFF6E460E762FB2A14C8.taxon	description	n 53 12 Prey item Average MD stomach 1 (%) ± SD Copepod UNID 4.2 ± 14.6 7.4 ± 15.6 Harpacticoid UNID 0.3 ± 2.0 0 Calanus sp. 4.9 ± 17.6 71.1 ± 27.8 Pseudocalanus sp. 0 0.0 ± 0.01 Paraeuchaeta sp. 1.9 ± 13.7 0.4 ± 1.3 Metridia sp. 0 0.5 ± 1.7 Oncaea sp. 0.0 ± 0.02 0 Tisbe sp. 8.3 ± 25.7 0 Amphipod UNID 10.8 ± 25.8 0.6 ± 2.2 Apherusa glacialis 52.7 ± 43.2 1.7 ± 6.0 Themisto sp. 4.1 ± 13.3 0.4 ± 1.5 Euphausiid UNID 0 0 Thysanoessa sp. 0 0 Decapods 1.9 ± 13.7 0 Chaetognaths 0.6 ± 3.7 17.4 ± 18.2 Appendicularians 0.8 ± 5.6 0.5 ± 0.7 FO (%) Copepod UNID 18.9 50.0 Harpacticoid UNID 20.8 0 Calanus sp. 20.8 100.0 Pseudocalanus sp. 0 8.3 Paraeuchaeta sp. 1.9 8.3 Metridia sp. 0 8.3 Oncaea sp. 1.9 0 Tisbe sp. 49.1 0 Amphipod UNID 24.5 8.3 Apherusa glacialis 64.2 8.3 Themisto sp. 17.0 8.3 Euphausiid UNID 0 0 Thysanoessa sp. 0 0 Decapods 1.9 0 Chaetognaths 3.8 66.7 Appendicularians 3.8 41.7 IRI (%) Copepod UNID 1.6 4.7 Harpacticoid UNID 1.2 0 Calanus sp. 3.2 87.3 Pseudocalanus sp. 0 0.0 Paraeuchaeta sp. 0.1 0.0 Metridia sp. 0 0.0 Oncaea sp. 0.0 0 Tisbe sp. 23.2 0 Amphipod UNID 5.4 0.0 Apherusa glacialis 63.4 0.1 PS 106 / 2 PS 122 / 4 – 5 (s / a) PS 122 / 1 (w) 5 15 1 1.7 ± 3.8 17.23 ± 25.55 44.2 0 0 0 40.1 ± 39.2 7.89 ± 16.8 0 0 0 0 0 1.3 ± 5.2 0 0 0 0 0 0 0 0 0 0 0.3 ± 1.0 8.4 ± 19.2 0 2.3 ± 19.7 6.7 ± 25.8 0 0 0.5 ± 2.1 0 33.0 ± 29.1 0 0 21.4 ± 21.9 0 0 0 0 0 0 18.1 ± 27.0 55.8 0 26.5 ± 33.7 0 20.0 53.3 0 0 80.0 33.3 0 0 0 6.7 0 0 0 0 0 0 20.0 20.0 40.0 6.7 0 6.7 60.0 0 80.0 0 0 0 0 40.0 0 60.0 0.5 21.5 0 0 49.4 4.1 0 0 0 0.1 0 0 0 0 0 0 0.4 2.8 3.9 1.1 T A B L E 4 (Continued) PS 80 PS 92 Themisto sp. 1.7 0.0 Euphausiid UNID 0 0 Thysanoessa sp. 0 0 Decapods 0.1 0 Chaetognaths 0.1 7.4 Appendicularians 0.1 0.4 PS 106 / 2 PS 122 / 4 – 5 (s / a) PS 122 / 1 (w) 0 0.1 22.9 0 22.9 0 0 0 0 11.4 0 59.0 Note: Presented are averages of relative contribution to dry mass (MD) and SDs, frequency of occurrence (FO), and index of relative importance (IRI). Note that many unidentified copepods likely also belong to the genus Calanus, which would influence the FO and IRI. For PS 122, winter (w) and summer / autumn (s / a) fish were separated. For the winter fish, FO and IRI were not calculated as only one fish was caught during this season. n = the number of fish analysed from each expedition. not further determined, this suggests the presence of multiple species, including C. hyperboreus. The size frequency of Calanus spp. from the stomach contents can be found in Figure S 3 b. The cluster analysis separates the fish into four groups distinguished by the food items that were numerically dominant in the stomach contents: calanoid copepods (mainly Calanus spp.), krill, appendicularians, and a combination of A. glacialis and / or Tisbe sp. (Figure 4 a). Stations located north of Svalbard and the Fram Strait (between 20 west and 30 east), as well as station 248 (PS 80) located in the Nansen basin, were dominated by calanoid copepods, whereas the other stations, located more north and east, were dominated by A. glacialis and / or Tisbe sp. (Figure 4 b). There were some exceptions from this general pattern. In the former region (north of Svalbard / western Nansen basin), stomach contents of fish collected at station 67 _ 5 during PS 106 / 2 were dominated by krill or A. glacialis. During PS 80, stomach contents of all fish at station 216 and most fish at station 223 were also dominated by A. glacialis. In the latter region (north / east), the PS 122 stations (located east of 100 east) contained few recognizable food items, with the exception of one stomach that was dominated by appendicularians and the stomach of the fish collected in winter, which was dominated by calanoid copepods.	en	Schaafsma, Fokje L., Flores, Hauke, David, Carmen L., Castellani, Giulia, Sakinan, Serdar, Meijboom, André, Niehoff, Barbara, Cornils, Astrid, Hildebrandt, Nicole, Schmidt, Katrin, Snoeijs-Leijonmalm, Pauline, Ehrlich, Julia, Ashjian, Carin J. (2024): Insights into the diet and feeding behavior of immature polar cod (Boreogadus saida) from the under-ice habitat of the central Arctic Ocean. Journal of Fish Biology 105 (3): 907-930, DOI: 10.1111/jfb.15836, URL: https://doi.org/10.1111/jfb.15836
0393B3164159FFEFE728E45AFDDA15DA.taxon	description	The bioenergetic model provides insights in the sensitivity of B. saida growth to variation in prey fields. It shows that growth rate may be more sensitive to the amount of food ingested, defined in the model through stomach fullness and, to a lesser degree, the energetic content of prey. This is primarily explained by the data distribution of the two parameters in the model having different coefficients of variation, that is, ratio between SD and the mean. Nonetheless, the high importance of stomach fullness in the model could indicate that the benefit of high-energy prey may be counterbalanced by other factors, such as energy expenditure for prey capture or minimizing predation risk. Such findings agree with our microscopic analysis indicating preferred feeding on prey that is abundant and / or easily available and / or prey that can be collected with relatively low effort. Although highest growth rates were found when B. saida fed predominantly on energy-rich prey, such as Calanus spp. or krill (Figure 8), our comparison of prey composition with the abundances of various key prey taxa indicates that the energetic trade-off for such high-energy prey depends strongly on its abundance and the abundance of alternative prey (Figure 5; Table 5). The effect of subzero temperatures on metabolic rates at sampling could explain the low modeled growth rates. Near-zero temperatures have been experimentally shown to induce low stomach evacuation rates and high assimilation rates in B. saida (Hop & Tonn, 1998), resulting in enhanced feed conversion efficiency at low feed intake (Kunz et al., 2016). This could explain the good body condition of B. saida caught at subzero temperature, while the model indicated low growth rates, suggesting model constrains in its temperature-limitation functions. Unfortunately, we lack crucial data from subzero temperatures to correctly model B. saida adaptation to ice habitat. Results from model simulations should be interpreted while considering the simplifications and specific assumptions that have been made (David et al., 2022), including a fixed ratio between active and basal metabolic rates and a constant energetic content of fish and prey types. An averaged energetic content per prey species was used for all sampled fish, although several prey species are known to have varying energy contents over the course of the year or with size (e. g., Kraft et al., 2015; Nowicki et al., 2023; Percy & Fife, 1981). This could reduce the variability in modeled growth rates. However, seasonal variation in the energy content of a species is likely much less than the variation between species. Although it is not surprising that both stomach fullness and prey energy content influence growth rate, the results indicate that changes in the abundance and catchability of the prey, and thus the amount of prey ingested, may have a larger impact than changes in prey energy content. With the warming of the Arctic Ocean and a shift toward smaller copepods and higher abundances of gelatinous zooplankton species, a number of prey field characteristics will change for B. saida such as the density, the size spectrum, the energy content, and the catchability of prey. The latter is due to potential changes in the density and behavior of the prey, as well as the loss of the under-ice habitat as a major feeding ground. Changes in such characteristics should thus be taken into account when anticipating consequences of environmental changes for B. saida.	en	Schaafsma, Fokje L., Flores, Hauke, David, Carmen L., Castellani, Giulia, Sakinan, Serdar, Meijboom, André, Niehoff, Barbara, Cornils, Astrid, Hildebrandt, Nicole, Schmidt, Katrin, Snoeijs-Leijonmalm, Pauline, Ehrlich, Julia, Ashjian, Carin J. (2024): Insights into the diet and feeding behavior of immature polar cod (Boreogadus saida) from the under-ice habitat of the central Arctic Ocean. Journal of Fish Biology 105 (3): 907-930, DOI: 10.1111/jfb.15836, URL: https://doi.org/10.1111/jfb.15836
0393B316415AFFEFE460E508FB32161D.taxon	description	Energy content measurements on most species from our study generally corresponded well with findings from the literature (Båmstedt, 1981; Norrbin & Båmstedt, 1984; Nowicki et al., 2023; Percy & Fife, 1981). Variations in energy content with season and body size have been found for various species. The energy content of the krill species T. inermis did not vary between winter and summer, but larger specimens had a higher energy content than smaller ones (Nowicki et al., 2023). For another krill species, Meganyctiphanes norvegica, variation between summer and winter was found, with values of 22.71 and 24.60 kJ g 1 MD, respectively (Nowicki et al., 2023). Energy contents decreased in winter (18.77 and 17.83 kJ g 1 MD) compared to summer (23.07 and 21.13 kJ g 1 MD) for both T. libellula and T. abyssorum, respectively (Nowicki et al., 2023). Individuals of T. libellula larger than 30 mm were found to have a higher energy content compared to smaller individuals, but for individuals <20 mm, there was no difference between size classes of T. libellula nor was there a difference with similar-sized T. abyssorum (Nowicki et al., 2023). This suggests that the energy content measurements we performed on T. libellula and T. abyssorum can be regarded as representative for the individuals ingested by the B. saida in our study. In the growth rate model, the energy content of Calanus prey was based on measurements performed on C. hyperboreus. Size measurements performed on Calanus spp. from the stomachs suggest, however, that the stomach contents likely contained a majority of C. finmarchicus and / or C. glacialis. The energy contents of the three Calanus species may vary, largely because lipid contents depend on developmental stage, season, and region due to the different life cycles and distribution patterns of the three Calanus species (e. g., Graeve et al., 2005; Kattner et al., 1989; Kirkesaeter, 1978 in Båmstedt, 1981; Scott et al., 2000; Swalethorp et al., 2011). Although some studies indicate that the energy density or lipid content per weight unit can be similar among the three species (Davies et al., 2012; Kohlbach et al., 2016), the smaller sizes indicate that energy content per individual would be less for C. glacialis and C. finmarchicus compared to C. hyperboreus.	en	Schaafsma, Fokje L., Flores, Hauke, David, Carmen L., Castellani, Giulia, Sakinan, Serdar, Meijboom, André, Niehoff, Barbara, Cornils, Astrid, Hildebrandt, Nicole, Schmidt, Katrin, Snoeijs-Leijonmalm, Pauline, Ehrlich, Julia, Ashjian, Carin J. (2024): Insights into the diet and feeding behavior of immature polar cod (Boreogadus saida) from the under-ice habitat of the central Arctic Ocean. Journal of Fish Biology 105 (3): 907-930, DOI: 10.1111/jfb.15836, URL: https://doi.org/10.1111/jfb.15836
