Boreogadus saida (Lepechin, 1774)

3.2 | Diet of B. Saida View in CoL from the central Arctic Ocean

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 PS92 and PS106/2. 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 (PS92 and P106/2), whereas the harpacticoid Tisbe sp. dominated during late summer/autumn (PS80; Kohlbach et al., 2017) followed by Calanus spp. (Figure 2). In terms of MD, Calanus spp. was also the dominant food item during expeditions PS92 and PS106. During PS80, 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 PS122. They accounted for 1.3% of the total recognizable food items in fish from PS92, although with a relatively high FO. Chaetognaths were also a relatively common food item during expedition PS92, occurring in 66.7% of the stomachs, albeit in low numbers. Krill ( Euphausiidae ) were only found in the stomach contents of fish collected during PS106/2 (Figure 2; Table 4). During PS80, 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 PS92. From the fish of PS106/2, one stomach contained two trematode parasites, and one stomach contained a single nematode parasite.

Identifiable prey items in B. saida stomachs from PS122 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 (PS122/4 – 5) of PS122 (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 PS122, 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 PS80 ( Kohlbach et al., 2017). Copepods ( IRI : 92%), dominated by Calanus spp. , were the most important food item during PS92. During PS106/2, 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 PS122 ( 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.

Calanus spp. in B. saida View in CoL stomachs from PS80 and PS92 had LP s of 1.5 – 4 mm. Some were staged as adult females (PS80), and as CIV/CV copepodites and adult females (PS92). This suggests that they belonged to either the species Calanus finmarchicus View in CoL or Calanus glacialis ( Madsen et al., 2001) View in CoL . In stomachs from fish collected at station 47_23 (PS92), adult female Calanus sp. with LP > 6 mm were also found, which is equivalent to the size of female C. hyperboreus ( Madsen et al.,2001) View in CoL . The individuals from PS106/2 stomachs were almost all adult females, and the LP of these females suggested that the species was C. glacialis ( Madsen et al., 2001) View in CoL . A wide range of Calanus spp. LP s was found during PS122 ( 1.5 – 6.4 mm) and, although

PS80 PS92

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

PS106/2 PS122/4–5 (s/a) PS122/1 (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)

PS80 PS92

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

PS106/2 PS122/4–5 (s/a) PS122/1 (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 PS122, 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 S3b.

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 4a). Stations located north of Svalbard and the Fram Strait (between 20 west and 30 east), as well as station 248 (PS80) 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 4b). 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 PS106/2 were dominated by krill or A. glacialis . During PS80, 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 PS122 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.