The ongoing warming were associated with increased water and air temperature, larger area covered by Atlantic and Mixed warm water masses and decreased ice coverage. The warming was also associated with
increased macro zooplankton such as krill and jellyfish biomass, increased fish recruitment (age 0) which trigger positive development of fish stocks (cod, haddock, deep water redfish, capelin and herring). Increased production and adequate fishing pressure in relation to stock size led to cod and haddock stock size increasing to record high levels and the capelin stock withstanding the high predation level. Several sea birds populations from coastal colonies stay or migrate in to the Barents Sea to feed and moult, possibly increasing the total number.
Figure 2.2.1 clearly shows the warming of the Barents Sea during the period 1970–2017. Temperature related variables, such as temperature itself, the area occupied by Atlantic Water (T>3°C) and Mixed Water (0<T°C>3), and the ice area at minimum and maximum sea-ice extent all show a clear trend towards warmer conditions. The variables connected to the dynamics, i.e. the flow of Atlantic Water (e.g. BSO – the inflow of Atlantic Water to the Barents Sea) and the atmospheric forcing represented by the NAO does not show any clear trend. These results corroborate the findings that, while both the temperature and volume transport of the Atlantic Water into the Barents Sea contribute to the oceanic heat transport into the Barents Sea, they (temperature and volume transport) vary on different timescales. While the volume transport fluctuates from year-to-year and even shorter timescales, the temperature shows a clear signal of multidecadal variability, i.e. climate variability of addition to an underlying trend induced by climate change.
The eastern Barents Sea has experienced the largest warming trend over the last 47 years, with considerably less warming along the boundaries (ICES 2018). There is, however, a slightly different geographical pattern at different depth intervals, with the warming trend being skewed towards the south with depth. Here, one should note that the upper-layer is more in contact with the atmosphere and, thus, governed by the air-sea fluxes, whereas the deepest depth interval, i.e. 100–200 m, is more strongly dictated by advection of Atlantic Water. Furthermore, there is considerable interannual variability superimposed onto the underlying, multidecadal trend.
Overall, the temperature and salinity averaged over the whole Barents Sea and at different depth levels were lower in 2017 than in 2016, albeit above long-term average values (Figure 2.2.2), which is in line with the future expectations stated in last year’s report.
Zooplankton and 0-group fish
Figure 2.2.3 shows a clear increase of krill, jellyfish and capelin, cod and haddock recruitment during the period 1988–2017, and the Mann-Kendall trend test showed that these positive trends were statistically significant. In contrast, polar cod, which is an arctic fish, showed a significant and negative trend, as these were at a high level during the 1990s and decreased in recent decades (Mann-Kendall trend test). However, neither zooplankton (all size fractions), nor herring or saithe recruitment show any clear trend.
Pelagic and demersal fish
Figure 2.2.4 shows increasing trend of cod stock biomass (age 3+), haddock spawning stock and recruites (age 3), deepwater redfish (S. mentelle) stock and Greenland halibut stock during the period 1990–2017, and the Mann-Kendall trend test showed that these positive trends were statistically significant. Recruitment to the fisheries exploited stocks (age 3) for cod does not show any clear trend. Polar cod stock was at high level in the beginning of 2000s and decreased after that. Herring biomass decreased since 2007 due to strong 2004-year class moved to the Norwegian Sea and since that, no strong year classes occurred.
Figure 2.2.4. Time-series of fish biomass (both recruited at age 3, spawning and stock) and landings in the Barents Sea sorted by trend. Blue means value below the 1980–2017 average and red means above the 1980–2017 average.
Figure 2.2.5 show increasing trend of snowcrab stock biomass during the period 2006-2017, and the Mann-Kendall trend test showed that these positive trends were statistically significant. The other parameters did not show any significant trends, most likely do to too short time series, but a change from mostly negative (before 2011) to more positive (from 2012) anomalies for the large animal-groups Porifera (sponges), Mollusca (bivalves, snails etc), and Echinodermata (sea stars, brittle stars, etc).
During the period 2004-2018 we found a positive northward trend in the spatial distribution (centre of gravity) of black-legged kittiwake and a negative trend in the abundance of this species (Mann-Kendal trend test). We found no significant trends in the northward displacement nor in the abundance of Brünnich’s guillemot and Atlantic puffin (Figures 2.2.6).
Figure 2.2.6. Time-series of sea birds abundance and spatial distribution (centre of gravity) in the Barents Sea sorted by trend. Blue means value below the 2004–2018 average and red means above the 2004–2018 average.
Human pressure decreased since 1970s in the Barents Sea. Figure 2.2.7 shows statistically significant decreasing trend in cod (ages 5-10), haddock (ages 4-7), capelin and shrimp fishing mortalities and capelin, Greenland halibut, polar cod and S. mentelle landings (Mann-Kendall trend test) during the period 1970–2016. Haddock landings showed increasing trend due to strong increase in haddock stock since 2005.
Observed data mostly indicate nonstationary. For the time series analysis, a structural time series model, observation = trend component + cyclic component + noise, is applied (Harvey 1989, West and Harrison 1997). The trend component is modelled a structural trend model using stochastic differential equation (Kitagawa and Gersch 1996). The model is represented by a state space form and the state corresponding to trend is predicted and filtered by Kalman filter algorithm. Figure 2.2.8 show the original data, estimated prediction (red) and filtering values (blue).
For the estimated trends, the common trends among the data are estimated by a classification and similarity analyses (Solvang et al. 2008). The classified groups indicating rising, flat fluctuation and decline are summarized in Figure 2.2.9. The configurations are recognized as dominant common trend for the data. Change of the trends for biotic and human impact data during observed year could be investigated by the corresponded common trends for abiotic data. The cyclic components are obtained by extracting the trend component from the data. The obtained cyclic component are considered directional relationships among data by the feedback system analysis (Solvang and Subbey 2019).