General Long-Term Trends of Themisto spp. at LTER HAUSGARTEN in the Eastern Fram Strait
The focus of this study was to verify the trends in dominating pelagic amphipod swimmer abundances recorded by Kraft et al. (2011, 2013) by investigating three more consecutive years (2011–2014) of amphipod sampling at two locations within the LTER observatory HAUSGARTEN. As amphipod species are more abundant in particular water masses (Dalpadado et al., 2001), they may serve as sentinel species to detect changes in the pelagic environment. With this new dataset, we also analyzed the population structure of the three dominating amphipod species.
The trends of respective abundances of the two native Themisto species – the boreal T. abyssorum and the Arctic T. libellula – as well of that of the intruding T. compressa, corroborated the observations from the period 2000–2012 described by Kraft et al. (2013). In the period 2000–2012, total amphipod counts increased by a factor of 14 during and after a warm anomaly (Supplementary Figure S1) observed in the HAUSGARTEN area, and counts remained equally high until 2014 (Figure 2, 3) despite the more stable water temperatures (Supplementary Figure S1).
Bottom-up and top-down processes similarly could have caused this tremendous increase in amphipod counts. Soltwedel et al. (2016) reported a slight increase of chlorophyll a biomass in the Fram Strait’s surface waters as well as in the water column. Hence, they suggested an increase of phytoplankton biomass in the region from 2008 onward (Soltwedel et al., 2016). These findings are complemented by Nöthig et al. (2015) documenting increasing chlorophyll a concentrations in the WSC since the 1990s during summer months. This development provided more food for pelagic herbivorous copepods, which in the Fram Strait mainly comprises the genus Calanus (Hirche et al., 1994; Mumm et al., 1998; Hop et al., 2006; Hildebrandt et al., 2014). It has been shown that egg production in Calanus glacialis was positively correlated to chlorophyll a concentration (Hirche et al., 1994). Supporting this, an increase of ∼50% in copepod abundances in central HAUSGARTEN sediment taps was evident between 2000 and 2014 (Nöthig et al., unpublished data). It is thus possible that raptorial carnivores, such as amphipods of the genus Themisto (Dalpadado et al., 2008b; Kraft et al., 2015), found sufficient prey, which might have caused the continuously increasing amphipod abundances. However, we do not know whether the 50% increase in food availability would have been enough to sustain the very high numbers of amphipods we report. One could speculate that, in addition, amphipods were not so heavily preyed upon by higher trophic levels.
Similarly, a top-down mechanism could have caused increased amphipod abundances. Amphipods are important prey for fish (e.g., capelin, cod), birds (e.g., little auk), and marine mammals (e.g., ringed seal) (Klekowski and Wȩsławski, 1991; Dalpadado et al., 2001, 2008b; Auel and Werner, 2003; Melle et al., 2004). For the Barents Sea, Dalpadado et al. (2001) suggested a strong predator-prey interaction between amphipods and fish such as cod and capelin. They further demonstrated that low abundances in predatory fish were accompanied by increased amphipod stocks, and vice versa (Dalpadado et al., 2001). Hence, it is very likely that in this trophic interaction, the release of predation pressure results in the recovery of amphipods preyed upon. Dalpadado et al. (2001) also suggested that this mechanism mainly controls amphipod populations in the Barents Sea. According to the Norwegian Directorate of Fisheries (2015), catches of capelin, haddock, and herring declined between 2010 and 2014, whereas Atlantic cod (Gadus morhua) catches increased by ∼30% for the same time frame. Given the high commercial value of Atlantic cod and haddock (Norwegian Directorate of Fisheries, 2015), a recently observed northward shift of Atlantic cod (Christiansen et al., 2016) and an enormous increase of fishing vessel sightings near Svalbard (Bergmann and Klages, 2012), the potential fishing pressure may remain high, relieving the amphipods from predation impact.
The three dominant pelagic Themisto species showed significant seasonal, inter-annual and spatial variability (Figure 3), with high abundances in summer and lower numbers in winter. Overall, T. abyssorum dominated the amphipod community by >50% during the period 2011–2014, corresponding to Kraft et al. (2011) results. However, in 2012/13, T. abyssorum and T. libellula were present in nearly equal proportions (∼40%, respectively) at both HAUSGARTEN sites. This sampling period was characterized by extraordinary ocean temperatures starting with a warm winter followed by a pronounced temperature drop, with cold water prevailing the entire summer of 2013 (Walczowski et al., 2017). It has been demonstrated that T. abyssorum was more abundant than T. libellula when warm WSC water predominated (Koszteyn et al., 1995; Dalpadado, 2002; Dalpadado et al., 2008a,b, 2016). Hence, the broad impact of the WSC may explain the observed predominance of T. abyssorum at the HAUSGARTEN, potentially coupled with increased reproductive rates and/or less predation mortality as discussed above. High reproductive rates were observed at both sites over the study period as indicated by the high proportions of juveniles compared with the two other species and their ubiquity over both seasons.
Whereas the two common species in the study area were present and dominating throughout the year, T. compressa was absent in the trap samples over long periods in winter (November–February). The reappearance of this species in spring may thus indicate an “allochthonous origin” as speculated by Kraft et al. (2013). Abundances of this North Atlantic species remained elevated compared to the mid-2000s, with noteworthy counts in late summer 2011, which may be attributed to inflow of warmer Atlantic water causing higher ocean temperatures in the eastern Fram Strait between 2011 and 2012 (Beszczynska-Möller et al., 2012; Walczowski, 2013; Gluchowska et al., 2017). This warm event has also been shown to be related to a substantial increase in abundances of the Atlantic-associated copepod Calanus finmarchicus in the WSC (Gluchowska et al., 2017). Irrespective of its winter absences, T. compressa appears to have become a common species in the eastern Fram Strait. To date, however, the species has not yet been recorded in the central Arctic (Kosobokova et al., 2011; Kosobokova, personal communication). L. clausii was the most abundant amphipod species after the three Themisto spp. with 19 specimens collected between 2011 and 2014. Previously, it was absent, but, – similarly to T. compressa in July 2004 – it became more abundant, although on a lower scale.
In this study we observed erratic peaks in amphipod abundances, especially in T. abyssorum (see Figure 3). Swarms or high density aggregations of T. abyssorum, T. compressa and T. libellula have previously been recorded both on the seafloor and in the water column (Lampitt et al., 1993; Vinogradov, 1999; Angel and Pugh, 2000). We assume that swarms of Themisto spp. were present at the sediment traps no later than the end of July 2013 (Figure 3), accumulating in the instruments’ funnels, filling the sample cups to the top, and thus exceeding the poison’s capacity to preserve the samples and resulting in degraded samples (Lee et al., 1992). A funnel full of swimmers would also explain the prolonged event lasting until early September 2013, with swimmer material filling up the next sample cup entirely when it was exposed for sampling. Strikingly, degraded samples occurred concurrently at both the central and the northern locations between July and early September 2013, for which similar mechanisms causing the degradation can be assumed. Furthermore, we suggest that due to degradation, significant numbers of amphipods are not included in the data, possibly leading to an underestimation of the maximum abundances recorded. In this context, the occasional T. abyssorum abundance peaks as in June 2012 and September 2013 (75.3 and 53.9 Ind. m−2d−1, respectively; Figure 3) were unexpected and can similarly be considered records of swarming events.
Population Structure of Themisto spp. in the Eastern Fram Strait During 2011–2014
Noteworthy proportions of juveniles of T. abyssorum (up to 8% of the total individuals) were recorded, which is consistent with Kraft et al. (2012). These were higher at the northern HAUSGARTEN site compared to the central location (Figure 4). This difference between sites coincides with a difference of ca. + 0.1°C in mean water temperature between August 2011 and June 2014 at the northern station compared to the central station and differences in sea ice cover. Juveniles were present in both seasons at both sites (contradicting Kraft, 2010), indicating more than one spawning period per year, as discussed by Koszteyn et al. (1995). This outcome contrasts with Kraft (2010), who obtained a seasonal pattern for juvenile T. abyssorum abundances, hence suggesting a seasonal migration of juveniles or lower reproductive rates. Not a single juvenile specimen of T. compressa was recorded in this study during winter, and only very few occurred in summer (max. 1% of the total specimen count), indicating no or only limited reproduction in the area; this is supported by Kraft et al. (2012). However, the record of Kraft et al. (2013) of brooding females in the traps may indicate that reproduction of T. compressa in the area is possible, but still rare. Very low numbers of juvenile T. libellula were recorded, irrespective of location and season (max. 1% of the total specimen count), whereas other investigations reported elevated numbers of juveniles between May and June (Percy, 1993; Koszteyn et al., 1995; Dale et al., 2006; Kraft et al., 2012). This may imply unfavorable reproductive conditions for the true Arctic T. libellula in a warming environment, as suggested by Dalpadado et al. (2016).
Maturity studies based on net hauls (e.g., Williams and Robins, 1981; Koszteyn et al., 1995; Dalpadado, 2002; Dale et al., 2006; Dalpadado et al., 2016) do not provide year-round data sets as do sediment trap catches (e.g., Kraft et al., 2012). Whereas net catches conducted in the Barents Sea by Dalpadado (2002) between August-September 1993 yielded juvenile proportions of up to 88% for T. abyssorum and 80% for T. libellula, a maximum percentage of 8% was recorded for T. abyssorum herein. Large relative numbers of juveniles have also been found in similar studies in the Greenland and Barents seas (e.g., Koszteyn et al., 1995; Dale et al., 2006). On the other hand, size distributions found herein agree with other sediment trap-approaches, such as Kraft et al. (2012), who generally observed very few juveniles. The different approaches target different depths, and it is known that Themisto spp. are often segregated by depth in the water column according to sex and life stage (Williams and Robins, 1981; Wȩsławski et al., 2006). Varying criteria for classifying life stages (juveniles, immature adults, mature adults) (Williams and Robins, 1981; Percy, 1993; Wȩsławski et al., 2006) and seasonal migratory behavior (Percy, 1993; Kraft, 2010) may also account for the large discrepancies in the outcomes.
Community Changes of Themisto spp. at LTER HAUSGARTEN in the Eastern Fram Strait
By combining data on pelagic amphipods species from this study (2011–2014) with Kraft (2010) Themisto spp. and other pelagic amphipod data sets from her Ph.D. thesis (2000–2011), we obtained a broader view of the amphipod community development between 2000 and 2014. Within this time frame, a separation of older and more recent samples is apparent (Figure 5), indicating a development of the system.
According to the similarity analyses of species composition between the different years and sites, the trend of increased T. compressa proportions most probably caused the dissimilarity over years. In general, the system has changed to a state of higher T. compressa abundances reflected by an increasing contribution of this species to the observed cluster patterns. Thus, the lack of T. compressa in the 2000/01 sample resulted in its strong dissimilarity with other samples. Furthermore, increased T. libellula abundances mainly contributed to the vertical separation of the samples with increasing numbers in the upper domain. Similarly, high T. abyssorum abundances contributed to the vertical clustering with highest values in the lower domain of the chart. However, these trends do not seem to indicate a continuous temporal development as discussed for T. compressa. These trends are reflected in the 2012/13 samples because their dissimilarity was mainly due to high cluster contributions of increased abundances of T. compressa and T. libellula as well as to low abundances of T. abyssorum. Interestingly, the 2012/13 winter was characterized by notably high ocean temperatures followed by a subsequent cold summer (Supplementary Figure S1, as mentioned in Gluchowska et al., 2017). We speculate that the abundances of T. libellula did not appear to be affected by increased abundances of the intruding T. compressa, at least under cold water conditions; however, abundances of T. abyssorum appeared to be adversely influenced (see also Stempniewicz et al., 2007). We further speculate that even though T. abyssorum is assumed to tolerate a high temperature gradient and show high abundances in Atlantic water masses (Dalpadado, 2002), competition between T. compressa and T. abyssorum may play a role, given their similar sizes and ecological roles (e.g., carnivorous feeding type – see Kraft et al., 2015).
Variations and possible shifts in amphipod proportions of the three dominant pelagic hyperiids are evident based on our long-term data series. The occurrence of the North Atlantic species T. compressa (Kraft et al., 2013) continued until 2014, which may be attributed to higher water temperatures (Supplementary Figure S1; Walczowski et al., 2017). The latter was confirmed by the observation that the abundances observed at the central HAUSGARTEN site were considerably higher than at the northern location. The system is evidently shifting toward a warm, more North Atlantic-influenced state (Gluchowska et al., 2017), potentially causing Arctic species to decline (Dalpadado et al., 2016). This is corroborated by the low numbers of juveniles of the true Arctic T. libellula in the sampling period 2013/14, even though previously, these were commonly detected. The replacement of the larger, lipid-rich T. libellula by the sub-Arctic and temperate species T. abyssorum and T. compressa in the Arctic could change the food chain pattern with possible consequences for fish, whale and bird populations that depend on this species as major prey. For example, little Auks (Alle alle) feed predominantly on the largest size class of T. libellula (Lønne and Gabrielsen, 1992), and hence, the other Themisto species cannot eventually act as substitutes because of their smaller size.
Warming water temperatures are a likely cause of the increasing amphipod abundances between 2000 and 2014 that are potentially affecting trophic interactions and increasing competition between Themisto spp. Thus, more temperate species evidently extended their range into the Arctic, as we demonstrated by the seasonal establishment of the North Atlantic species T. compressa. Other, previously unsampled species are newly appearing in the Fram Strait sediment traps, with the most abundant being the hyperiid L. clausii. These outcomes suggest ongoing environmental shifts taking place in the seasonally ice-covered eastern Fram Strait. For a better understanding of species interactions and for firm predictions regarding future pelagic communities, more regionally and temporally extensive studies on the topic are urgently needed.