Introduction

Climatic changes in marine ecosystems are one of the most important issues of current ocean science. The success of long-term forecasts of climate, bioresource potential, and conditions of maritime activity depends on correct interpretation of data on recent hydrophysical and biological processes. Therefore, it is necessary to assess the role of natural and anthropogenic forces in climatic anomalies, which have been particularly pronounced in recent years.

As is well known, the Quaternary period is characterized by intervals of global climate cooling and warming. During the cooling phases, continental glaciations expanded and oceanic sea ice cover advanced to 30–40°N. Consequently, boreal fauna and flora shifted to formerly subtropical zones. In the post-glacial Holocene, specifically the last 10 thousand years, the warmest period was that of the Atlantic optimum, around 6.8 thousand years ago (Matishov and Pavlova 1990; Matishov et al. 1994).

Over the last centuries, there have been cyclic changes in climate and ice cover. For example, numerous climatic extremes during the eighteenth century were well documented in the European territory of Russia (the end of “the Little Ice Age”), with relative stabilization in the nineteenth century (Borisenkov et al. 1986). These changes have been recorded by the Murmansk Marine Biological Institute (MMBI) at the Kola Scientific Center (KSC) of the Russian Academy of Sciences (RAS). The MMBI has been conducting expeditionary research since 1935 in the Arctic, from the Greenland Sea to the Kara Sea (Matishov and Matishov 2012). Information from data sets of the MMBI and other institutions has been assimilated to produce a series of electronic atlases reflecting climatic changes in the marine environment and biota (Matishov et al. 1998b, 2000, 2004).

Organisms inhabiting polar areas are sensitive to cyclic changes in climate and may serve as indicators of such changes (Anon. 1985, 1994; Sakshaug et al. 1994; Wassmann et al. 2006). However, the response of organisms varies depending on their longevity and life cycle stages, active and passive migrations, trophic interactions, and other factors. Hence, a long-term series of oceanographic and hydrobiological observations along the Kola Transect (Fig. 1) is of significant value, allowing trends of species composition and biomass of marine organisms to be determined in relation to climatic dynamics.

Fig. 1
figure 1

Map of Barents Sea, showing main water current patterns (Matishov et al. 2012), depth curves of 50, 100, and 200 m (Matishov 1997), and location of Kola Transect

Full size image

Contemporary research methods are based on the Large Marine Ecosystems (LME) concept (Sherman 1995). Within the Western Arctic, LMEs of the East Greenland Shelf, Norwegian Shelf, Barents Sea (including the White Sea), and Kara Sea have been delineated. Each LME has its own specific features of Pleistocene paleogeography and its own evolution of biogeocenoses in the Holocene, until present day. The Barents Sea covers a specific area of the Western Arctic (Matishov et al. 2003; Sakshaug et al. 2009). During ice ages, glaciers regulated the formation of specific glacial reliefs on the shelf (such as deep troughs, hollows, depressions, moraine ridges, and rises) of Barents Sea coasts. Consequently, these formerly glaciated shelf configurations strongly influence the large-scale interactions of Atlantic and Arctic water masses.

Marine economics in the Barents Sea area is strongly climate dependent. During the twentieth century, commercial fishing in the Barents Sea was important for providing fish products to Russia (USSR), Norway, and other countries. Strategies of sustainable exploitation of marine bioresources, intensification and development of shipping along the Northern Sea Route, offshore development of petroleum hydrocarbons, and other types of maritime activity ultimately depend on understanding the climate dynamics.

Fish stocks and fisheries play a leading role in LME state and dynamics. For instance, the structure of fish within communities may vary with respect to time and space. This process is determined by various factors, the impact of which is also not constant. One such factor is climatic fluctuations of different periodicity, which influence the physical state of hydrobiont habitats. In general, the major abiotic factors that influence diversity in Arctic seas are temperature and salinity, which determine the structure and dynamics of the fish population. Consequently, oscillations of environmental parameters and transformation of water masses in the Barents Sea and adjacent areas result in direct and indirect inputs to the reproductive capacity of fish generations, number of species, and their relative abundance.

A high priority of modern and earlier benthic investigations is understanding the ecological consequences of rapid warming, with emphasis on the autecology of key species and long-term data of benthic community dynamics (Piepenburg 2005; Piepenburg et al. 2010). Marine species respond to changes in their environment, which may be caused by local processes (e.g., pollution) or regional and even global forces (e.g., climate fluctuations). At the ecological levels of organisms, populations, and communities, these responses may serve as indicators of ongoing and future shifts in ecosystem regimes. The macrozoobenthic fauna of the Barents Sea may be categorized according to biogeographic distribution and temperature preference, specifically, cold water (Arctic), temperate water (boreal), and eurythermal (boreal-Arctic and subtropical-boreal-Arctic) species. The proportions of biogeographic types vary in space and time, as well as in relation to water temperature regime, which is driven by the climate (Brotskaya and Zenkevich 1939; Gur’yanova 1951; Nesis 1960). Therefore, monitoring spatiotemporal patterns in dynamics of benthic biogeographic types provides a versatile tool for tracing climate shifts in the Barents Sea. At present, marine biological studies are focused on evaluating the impact of climatic changes on seabed fauna of the Barents Sea, as well as identifying specific adaptations of benthos organisms to changing environmental conditions (Piepenburg 2005).

There are numerous historical and novel publications with original data on large-scale benthos surveys, in addition to reviews on species diversity (Deryugin 1925; Gur’yanova 1928; Filatova 1938; Brotskaya and Zenkevich 1939; Nesis 1960; Antipova 1975; Bryazgin et al. 1981; Anon. 1986; Brattegard et al. 1997; Kiyko and Pogrebov 1997; Galkin 1998; Sirenko 2001; Palerud et al. 2004; Denisenko 2005, 2006, 2008; Cochrane et al. 2009; Dahle et al. 2009; Gulliksen et al. 2009; Anisimova et al. 2010). Analysis of changes to fauna composition caused by temporal climatic fluctuations remains unresolved, because of different approaches to the principles of biogeography nomenclature (Semenov 1990 ). The ecological approach of the biogeography discipline takes ecological habitats into account, specifically conditions under which a species has the highest biomass per surface unit (Zenkevich 1977).

During warm periods, boreal species disperse into the Barents Sea from the west, with their northern distribution boundary shifting northeast, while Arctic species retreat to colder, northern regions. During cold periods, the shift in distribution ranges is reversed. Such periodic climate-driven alternations were repeatedly registered during the first half of the twentieth century (Deryugin 1925; Gur’yanova 1928; Shorygin 1928; Deryugin 1930; Gur’yanova 1951; Nesis 1960) for zoobenthos, in addition to the migration routes of Atlantic cod, stock sizes of the Kamchatka crab, and other commercially used organisms (Drinkwater 2009, 2011). Therefore, climate dynamics may be reflected by populations of marine organisms, for which stock distribution and fluctuations have been comparatively well monitored during the course of their commercial exploitation.

Based on our own data and the literature (published mostly in Russian), we analyzed a series of indices for the state of marine ecosystems, which allowed a connection (or absence of a connection) to be established with actual changes to the Barents Sea climate over more than a century. Most Norwegian and other non-Russian literature have been covered in extenso by Sakshaug et al. (2009).

Materials and methods

Studies of the climate dynamics of the Barents Sea ecosystem were carried out using MMBI oceanographic and hydrobiological databases, which have been assimilated via expeditions by the institute for more than 75 years and through international data exchange (Table 1). The thermohaline database of secular transects in the Barents Sea contains more than 220,000 hydrologic stations for the period 1900–2011 (Matishov et al. 1998b, 2000, 2004). An electronic database of Barents Sea ice conditions contains average values for each month from 1960 to 2011. From these MMBI data sets, long-term mean values (norms) and anomalies of water temperature and ice conditions were computed.

Table 1 Origin of oceanographic and hydrobiological data analyzed
Full size table

Using these data sets, we investigated the thermohaline structure of Barents Sea water. Mean temperature of Atlantic water in layers 0–50 m, 0–200 m, and 0 m–bottom is a general indicator of climate variations, which reflects processes in the active upper layers and in the entire water column of northern European seas (Matishov et al. 2007, 2009).

As an assessment criterion of the thermal state of Barents Sea water, a correlation value (“climate index”) between water temperature anomalies (ΔT) in the 0–200 m layer along the Kola Transect every year and the mean square deviation of annual water temperature (σ) is used to define the following categories:

$$ \begin{array}{*{20}c} {{\text{Abnormally}}\,{\text{cold}}\,{\text{year}}} \hfill & { - \Updelta T^\circ {\text{C}} > 1.5\sigma } \hfill \\ {{\text{Cold}}\,{\text{year}}} \hfill & {0.5\sigma < - \Updelta T^\circ {\text{C}} \le 1.5\sigma } \hfill \\ {{\text{Normal}}\, ( {\text{regular)}}\,{\text{year}}} \hfill & { \pm \Updelta T^\circ {\text{C}} \le 0.5\sigma } \hfill \\ {{\text{Warm}}\,{\text{year}}} \hfill & {0.5\sigma < \Updelta T^\circ {\text{C}} \le 1.5\sigma } \hfill \\ {{\text{Abnormally}}\,{\text{warm}}\,{\text{year}}} \hfill & {\Updelta T^\circ {\text{C}} > 1.5\sigma } \hfill \\ \end{array} $$

The most obvious response of fish communities to local climatic fluctuation is a change in number of species; however, fish diversity may be represented by species lists or by quantitative indices of complexity in the composition of communities, specifically the Shannon Index of organization. This index summarizes data on species composition, number of organisms, and degree of their dominance. The higher the index, the more organized the system. Quantitative indices of fish diversity (Shannon Index H, uniformity E, or evenness of distribution J) were obtained from ICES data based on the catch of pelagic, demersal, and benthic fish species (Eurostat/ICES database on catch statistics; ICES 2011) for three areas: I (Barents Sea), IIa (Northern Norwegian Sea), and IIb (Northeastern Greenland Sea) (ICES 2012). The calculations consider 21 species, of which total biomass in the boreal and Arctic parts of specified areas was 99.74 and 98.82 %, respectively (Karamushko 2005).

Changes in the distribution of commercial fish species, as indicated by the location of vessels in the sea, may serve as an indirect index of shifts in the marine environment. The major commercial species in the Barents Sea is Atlantic cod. Analysis of the fisheries geography of this species was based on consolidated reports and data on remote monitoring of fishing vessel locations over the period 1977–2006. Analysis of changes in the area of commercial fishing sites and the commercial stock of cod was evaluated in relation to the chronological schedule of water temperature anomalies at the Kola meridian transect (Fig. 1). This research is focused on marine areas north of 73°N, because the greatest changes in migration directions and fish distribution were in the northern part of the sea when cycles of climatic fluctuation changed.

Adaptation conditions of the Kamchatka crab, which was introduced to the Barents Sea in the 1960s, were analyzed separately. The abundance of alien species was analyzed with the authors’ own data and published material (Kuzmin and Gudimova 2002; Berenboim 2003; Anon. 2008; Sundet and Berenboim 2008; Anisimova et al. 2010; Lyubina et al. 2012).

Quantitative assessment of bottom fauna response to variability in water temperature on the Kola Transect was performed using the example of polychaetes as the most abundant, evenly distributed, and diverse taxonomic group of benthic animals. The analysis was performed following the method of Nesis (1960), whereby changes in ratio of the number of species inhabiting the boundary of their habitat (only boreal and arctic) were evaluated, owing to their dependence on water temperatures. It is assumed that variability of species diversity, abundance, and biomass in the settlement conditions of polychaetes reflects the overall state and dynamics of bottom communities. This analysis was based on data gathered along the Kola Transect in 1995, 1997, 1999, 2000, 2001, and 2007. Samples in 1995 were taken by a grab “Ocean” with 0.25 m2 sampled area in two replicates at each station. In 1997, 1999, 2000, 2001, and 2007, samples were taken by a van Veen grab with 0.1 m2 sampled area in five replicates at each station. The relative percentage of boreal and Arctic polychaete species was calculated at each station. The total ratio of boreal and Arctic species was determined for all stations along the Kola Transect for each year of the investigation. The total percentage of boreal and Arctic species was compared with data of water temperature anomalies in the 0–200 m layer along the transect (Fig. 1), through simple regression analysis based on student’s t distribution.

To identify correlations in biomass dynamics of zoobenthos with changes in climate conditions, simple regression analysis based on student’s t distribution was again performed. The number of samples at one station (five) did not provide statistical confidence of the arithmetic mean of benthic biomass. Hence, to minimize error of the mean, the number of samples was increased by combining samples from six stations along the Kola Transect. These stations were in the middle part of the transect, with similar environmental conditions and benthic species composition, and the dominant polychaete species was Spiochaetopterus typicus. Analysis data were collected from 1995 to 2001. The biomass of benthos was compared against water temperature anomaly data in the 0–200 m layer along the transect (Frolova et al. 2007).

Results

Hydrographic parameters and climate index

The leading determinant of the climate of Eurasian polar seas is advection of the North Atlantic current. Warm salt waters of Atlantic origin enter the Barents Sea from the west. According to different assessments, inflow strengths of these waters range from 1.6 (O’Dwyer et al. 2001) to 2 Sv (106 m3/s) (Blindheim 1989; Ingvaldsen et al. 2002). In the early twenty-first century, advection of Atlantic waters increased sharply (Matishov et al. 2009). This inflow is determined by both the North Atlantic Oscillation (NAO) and wind field across the transect, from northern Norway to Bear Island (Ingvaldsen and Loeng 2009). The regional climate is further influenced by feedback mechanisms between the extent of ice cover and thermal regime and circulation of water masses.

Because of the distribution of Atlantic water in the Barents Sea, there is a stable frontal zone (polar front) up to 1,500 km long. This front separates the non-freezing southwestern Barents Sea from the remainder of the sea, where water is transformed from Atlantic and Arctic waters to local water masses with observed seasonal ice cover (Fig. 1). The Atlantic waters, which traverse underwater glacial troughs, exhibit spatiotemporal thermohaline mosaics. Hence, to obtain an objective picture of climatic trends, it is necessary to use average values of hydrologic characteristics (Matishov et al. 1998a, 2009).

Measurements of temperature and salinity along the Kola Transect (Fig. 1) have been conducted almost continuously since 1900, with interruptions from 1907 to 1920 and 1942–1944 (Knipowitsch 1905; Matishov et al. 2009). In general, the southern part of the transect (south of 74°N) has been studied with regularity (Karsakov 2009). However, study of the Central Bank and area north of 74°N is of greater interest (Fig. 1).

In the active upper sea layer along the Kola Transect, two warm cycles have been the most pronounced in the secular dynamics of climate (Fig. 2). A climatic phenomenon in the first half of the twentieth century, known as the Arctic warming, was characterized by a temperature increase at the coast and islands, reducing seasonal ice cover in the Arctic seas (Zubov 1945; Matishov et al. 2009). Two waves of warming along the transect, with temperature anomalies up to +1.0 °C, were registered in the early 1920s and during most of the 1930s. Warming during that period might be explained by the coincidence of phases with respect to 11- and 90-year cycles in solar activity and a 250-year cycle in the change of earth’s rotation velocity (Maksimov 1954). The warming period through the 1930s was followed by four decades of cooler conditions, interspersed with some short warm events.

Fig. 2
figure 2

Average weighted anomalies of a water temperature (°C) and b salinity (‰) of Kola Transect in the 0–50 m layer from 1900 to 2011, based on MMBI data (“The Great Salinity Anomaly” period of Belkin et al. 1998)

Full size image

A new trend in warming of the Arctic seas began to be observed at the end of the 1980s. At the beginning of the twenty-first century, advection of the North Atlantic current continued to increase. This was also observed in the core of the North Cape Current from 1980 onwards (Ingvaldsen and Loeng 2009). In the Barents Sea, the warm anomaly in the 0–200 m water layer reached its maximum between 2001 and 2006, followed by a decrease in recent years.

During the 1990s, relatively warm weather was typical off the coasts of Novaya Zemlya, Franz Josef Land, and Spitzbergen. However, the temperature increase was not uninterrupted. Short-term cooling period of the active layer was registered from 1997 to 1999 (Matishov et al. 1999, 2009), when Kola Bay became frozen for the fourth time in the twentieth century. During 2007–2010, hydrologic indices again showed cooling of Barents Sea water (Fig. 2). Since then, the Barents Sea ecosystem functional regime has gradually returned to conditions of the mean long-term climatic state. In the future, a shift to a cooling period is possible. According to Ingvaldsen and Loeng (2009), a 70-year period of natural variability, as indicated on the Kola Transect, would lead to stable temperatures or negative trend over the next 20–30 years. The authors assume that the expected cooling over the next two to three decades would mask an eventual anthropogenic warming of the Barents Sea.

Based on data over the period 1951–2010, the mean square deviation of annual water temperature σ in the 0–200 m layer along the Kola Transect was 0.51 °C. This criterion was used to develop an anomalies calendar of Barents Sea water thermal state (Fig. 3). The calendar allows definition of qualitative and quantitative dependencies between the hydrographical (thermal) regime and specific distribution features of the main groups of organisms in that sea. For example, dependencies may be determined for populations of commercial species, based on long-term fisheries statistics, as shown in Fig. 3 for cod.

Fig. 3
figure 3

a Inner-secular cyclic characteristic of Barents Sea climate (thermal regime of Murmansk current exemplified). b Long-term changes in cod stock and fishing area north of 73°N (in % of total area of this part of Barents Sea)

Full size image

Ice conditions

Ice conditions, in terms of seasonal ice cover and location of the ice edge in the Barents Sea, are characterized by large seasonal and interannual variability (Wassmann et al. 2006). In general, the last 50 years may be divided into two major periods: (1) a cold period from 1960 to the early 1990s (considerable positive anomalies of ice covered area) and (2) a warm period from the early 1990s to the present (prevalence of high negative anomalies) (Fig. 4).

Fig. 4
figure 4

Anomalies of mean annual ice cover (%) in Barents Sea for the period 1960–2011

Full size image

The charts, based on the MMBI electronic database of Barents Sea ice conditions, demonstrate interannual variability in location of the ice edge. A gradual decrease in total area of ice cover has been typical of the Barents Sea for the last half-century, with minimum values recorded in 2006 (Fig. 4). However, an increase in total area of ice cover has been observed during the last 4 years (Fig. 5).

Fig. 5
figure 5

Location of ice edge in Barents Sea during periods of minimum ice cover and regular (normal) conditions, 1 February 2006, and 2 February 2011, respectively

Full size image

Ichthyofauna diversity

It has been established for the research period that the most diverse and balanced fish component of the studied communities was in ICES area IIa. The maximum value of the Shannon Index for this period was 2.94 in the Northern Norwegian Sea, 2.33 in the Barents Sea, and 2.2 in the Northeastern Greenland Sea. Minimum values for these three regions were 1.77, 0.88, and 0.46, respectively (Fig. 6). Since the quantity of analyzed species remained constant, the Shannon Index value was only affected by their correlation in the catch. For this reason, the dynamics of the evenness of shares, or species abundance, almost completely coincides with Shannon Index fluctuations.

Fig. 6
figure 6

Dynamics in a diversity of fish component of community and b evenness of species abundance distribution in several ICES northern regions—Barents Sea (I), Northern Norwegian Sea (IIa), and Northeastern Greenland Sea (IIb) (from ICES Catch Statistics data)

Full size image

Geography of commercial fishing

In total, 182 species and subspecies of fish have been classified in the Barents Sea (Karamushko 2008). Only a few of these species are of major commercial interest. These are cod (Gadus morhua), saithe (Pollachius virens), Polar cod (Boreogadus saida), haddock (Melanogrammus aeglefinus), capelin (Mallotus villosus), and herring (Clupea harengus). The fisheries for each of these species have fluctuated considerably for various reasons, including changes in stock size, which are due to fishing and recruitment success or failure. Interactions between species have been described, such as higher predation rates of capelin larvae by juvenile cod (Gjoesaeter 2009). Changes in regional climate regime, and hence marine environment, have a clear impact on migrations of commercial fish species and, as a result, the geography of commercial fishing. The same applies to commercially important crustaceans, such as prawns and Kamchatka crabs (Paralithodes camtchatica). The following section focuses on the distribution and migration of cod and Kamchatka crab.

Arcto-Norwegian cod (Gadus morhua)

A generalized integration of climatic and fishing information for 30 years based on MMBI data (Zhichkin 2009) permitted patterns of cod migration and distributions of fisheries to be determined with respect to water temperature. For example, during an extremely cold period of the late 1970s to early 1980s, boundaries of cod feeding areas in the east were restricted to shallow water areas of the Murmansk Bank. By the end of the feeding period in the northwest part of the sea, cod distributions only just reached the Zuidkapp Trough and eastern slope of the Spitsbergen Bank (Fig. 7a).

Fig. 7
figure 7

Migration and distribution of commercial cod in Barents Sea during May–October of a abnormally cold and b abnormally warm years

Full size image

During the 1980s–1990s, with initiation of a warm phase of climatic fluctuation (Drinkwater 2009; Zhichkin 2009), principal cod migration routes began to redistribute. During the abnormally warm period of 2004–2006, migrations were directed both east and north (Fig. 7b). In the east, cod distributions during September–October reached the coastal waters of Novaya Zemlya, between 70 and 74°N. However, the main migration flow was still in a northerly direction (in the area of the Island of Nadezhdy or “Hope Island,” and the Perseus Bank).

Analysis of changes to commercial cod stocks and of total fishing area showed that negative anomalies of water temperature observed in the late 1970s to early 1980s, combined with a significant increase in total catch volume, decreased the cod stock to nearly one-third its original size. The area of fishing sites north of 73°N during that period did not exceed 1–2 % of the total area of that region (Fig. 3). During the warm phase in the early 1990s, the fishing area expanded eastward and, especially, northward. The commercial cod stock was thereby restored to maximum values, to which there was also a contribution by decreased fishing pressure (Borisov et al. 2001). A short cooling period in 1997–1998 reduced the fishing area. Finally, during the warming period of the early 2000s, the commercial stock stabilized and cod habitat enlarged.

Therefore, the migration of cod and changes to its fishing areas reflect several factors, with the primary ones comprising changes in thermal state of the sea, together with commercial extraction and strict fishing quotas under current conditions. In general, warm anomalies contribute to generating ecologically sustainable commercial fishing.

Kamchatka (Red King) crab (Paralithodes camtschatica)

Of special interest among Barents Sea commercial resources is an alien species, the Kamchatka crab. The species was introduced to the Barents Sea between 1961 and 1969. During this period, MMBI biologists introduced about 1.5 million larvae, 10 thousand juveniles, and 5 thousand adults from Kamchatka to waters of the Western Murman coast of the Kola Peninsula (Kuzmin and Gudimova 2002; Berenboim 2003). Additional specimens were not introduced, and population abundance remained insignificant for a long time. Until the late 1980s, population abundance did not exceed 100 thousand specimens. Then, the Kamchatka crab population increased rapidly, particularly in the late 1990s, reaching a maximum near 21 million specimens in 2003. This population growth occurred 9–12 years after the warm period of 1989–1990 and coincided with the warming period of the first years of the twenty-first century (Fig. 8) (Kuzmin and Gudimova 2002; Berenboim 2003). It is assumed that pelagic larvae of the crab survived under the favorable thermal conditions of 1989–1990, and these later developed into juveniles without commercial extraction. By the early 2000s, the rich year-class reached sexual maturity and commercial size at 8–10 years (McCaughran and Powell 1977).

Fig. 8
figure 8

Outbreak in abundance of Kamchatka crab in Barents Sea, caused by warming in early twenty-first century (based on data from Murmansk Marine Biological Institute and Knipovich Polar Research Institute of Marine Fisheries and Oceanography)

Full size image

At present, distribution of the Paralithodes camtschatica population has shifted eastward and westward along the Murman and Norwegian coasts of the Barents Sea (Sundet and Berenboim 2008). Current declines in crab abundance may be interpreted as a result of the actions of fisheries.

The high abundance of the large, predatory Kamchatka crab decreased the biomass of indigenous benthic organisms, including echinoderms, mollusks, and bristle worms, which formed the main food source for this species (Falk-Petersen et al. 2011). The problem of crab impact on benthos was discussed in recent articles (Sundet and Berenboim 2008). A more visible impact of P. camtschatica on benthos was noted in the Varangerfjord, which had been inhabited by this species for a long period (Anisimova et al. 2010). Warming of the sea was one reason for considerable growth in crab abundance and habitat expansion (Berenboim 2003). Therefore, it became evident that there was a connection between crab stock and climate conditions, which also influenced biological interactions and fisheries (Anon. 2008).

Macrozoobenthos

Overall, 190 polychaete species were identified along the Kola Transect during the research period. Of these, 38 (20 %) occurred on the boundary of the research area, with their distribution potentially usable as an indicator of warm or cold temperature conditions. These species are comprised of the following:

  • 19 boreal species: Asychis biceps, Clymenura borealis, Eteone suecica, Eunice dubitata, E. norvegica, E. pennata, Exogone verrugera, Filograna implexa, Goniada norvegica, G. maculata, Hydroides norvegica, Neoleanira tetragona, Orbinia norvegica, Paramphinome jeffreisii, Paraonella nordica, Scolelepis korsuni, Sosane wireni, Streblosoma bairdi, and S. intestinalis

  • 19 Arctic species: Aglaophamus malmgreni, Axiothella catenata, Clymenura polaris, Diplocirrus hirsutus, D. longosetosus, Euchone analis, Glyphanostomum pallescens, Lanassa nordenskioldi, Leaena abranchiata, Lysippe labiata, Marenzelleria wireni, Melinna elisabethae, Micronephtys minuta, Nephtys paradoxa, Ophelina cylindricaudata, Praxillura longissima, Pseudoscalibregma parvum, Sabellides borealis, and Typosyllis fasciata

Among the many polychaetes, there are neither dominant nor typical forms of bottom communities.

After several warm years beginning in 1989, the highest percentage of boreal species was measured in 1995 on the seabed in the coastal and central branches of the Murmansk current. An increase in the number of Arctic species was seen north of the main branch of the Murmansk current and in the northern branch of the Nordkapp current. During the short cool period of 1997, the number of boreal polychaete species was similar to that recorded in 1995. However, the number of Arctic species increased considerably at all stations of the Kola Transect. In the warm year of 2000, the proportion of boreal species decreased relative to 1997, most notably in the coastal branch of the Murmansk current. The percentage of Arctic polychaetes decreased in the warm current streams at this time. Despite being a relatively warm year, 2001 was colder than 2000; consequently, the proportion of boreal species decreased notably in the coastal and main branches of the Murmansk current. The boreal species Asychis biceps, which was previously registered in all study years, disappeared from the coastal branch region during this period (Figs. 2, 3, and 9).

Fig. 9
figure 9

Species percentages of a boreal and b Arctic polychaetes in relation to total number of polychaete species at 11 stations of Kola Transect in different years (1995, 1997, 1999, 2000, 2001, and 2007)

Full size image

After a period of abnormally warm years (2001–2007), the relative number of boreal species exceeded the percentage of Arctic species in the main branch of the Murmansk current (Fig. 9). The total percentage of boreal species from all stations along the section noticeably increased until 2007, reaching 8.2 % (Fig. 10a). The total amount of Arctic species (for all stations) decreased to 9.1 % during the same period (Fig. 10b). The boreal species Clymenura borealis and Filograna implexa were first recorded in the Kola Transect area during 2007. In this same year, the Arctic species Aglaophamus malmgreni was not found in the coastal and main branch areas of the Murmansk current. The cold water species Nephtys paradoxa shifted its distribution further north, to 73°N.

Fig. 10
figure 10

Relative abundance (%) of a boreal and b Arctic species of polychaetes in southern part (540 km) of Kola Transect and water temperature anomalies of Murmansk current

Full size image

It has been noted that the ratio of boreal and Arctic polychaete species on the shelf along the Kola Transect changes unevenly from year to year. The zoogeographical structure of zoobenthos reflects differences in heat content of different branches of the Murmansk and Nordkapp currents at various times (Figs. 2, 3, 9, and 10). At the start of the research period, prevalent warming was observed in the coastal branch of the Murmansk current. In the first years of the twenty-first century (through 2007), the capacity of the coastal branch decreased, with the main branch of this current becoming the warmest.

High correlation was identified through comparison of the time series of total percentages of boreal and Arctic polychaete species against fluctuations in water temperature (Table 2). There was a 1–5 year lag in the response of indicator species to changes in water temperature regime. Arctic and boreal fauna respond with different periods of delay. It is estimated that the delay for boreal species is 4 years, while that of Arctic species is 1 year (Table 2; Fig. 10).

Table 2 Correlation coefficients of ratio of boreal and Arctic polychaete species with water temperature anomalies of Murmansk current (150–200 m layer)
Full size table

Results of the dynamics of biogeographic structure of polychaetes along the Kola Transect allow conclusions to be drawn about variations in response to velocity of boreal and Arctic species to changes in water temperature regime. This variation may contribute to explaining different reproductive strategies of boreal and Arctic species (Nesis 1960; Galkin 1998).

Quantitative time series data (1995–2001) on benthos from the Kola Transect allow biomass dynamics to be studied relative to variations in seawater temperature in the southern Barents Sea (Frolova et al. 2007). Total benthic biomass was negatively correlated with water temperature in the 0–200 m layer. Biomass of boreal-Arctic and subtropical–boreal–Arctic species showed a significant positive relationship. The biomass of Arctic species was negatively correlated with temperature (Table 3) (Frolova et al. 2007). Correlation analysis permitted precise intervals of biomass response to be determined. The time lag in response to total benthic biomass and biomass of some abundant and evenly distributed species differs with temperature variations (Table 3). The time lag for mollusks was determined in correlation with values calculated for the Pechora Sea (Denisenko 2007). The time lag in response to echinoderm biomass is also believed to correspond to the life cycle of these invertebrates. The time lag of 8 years, which was determined for the polychaetes Spiochaetopterus typicus and Maldane sarsi, requires further clarification with longer time series. Therefore, cyclic changes in the vector of mean annual temperatures correspond to benthic biomass with time lag about 3 years for the entire community and 3–8 years for certain species (Fig. 11) (Frolova et al. 2007).

Table 3 Correlation between biomass of dominant benthic species and total benthic biomass with temperature anomalies
Full size table
Fig. 11
figure 11

Delay in response of biomass and migration for indicator species with change in vector of thermal cycles along Kola Transect (33°30′E)

Full size image

Discussion

The climatic situation in the Barents Sea reflects large-scale processes that affect the entire Arctic, such as the warming periods of 1920–1930 and present (Frolov et al. 2009; Matishov et al. 2012). However, the causes and trends of recent warming in the sea area remain under dispute. The centennial series of water temperature and ice observations is not long enough to prove that cyclic changes observed in the twentieth century are stable phenomena. The extent of anthropogenic impact on recent climate also requires further evaluation.

Nevertheless, the pronounced warming of 1990–2000 has already been accompanied by changes in the Barents Sea ecosystem, some of which may be prolonged. The results of our studies contribute to ecosystem indices of high and low sensitivity.

The analysis of fish diversity showed that for different ICES areas (I, Barents Sea; IIa, Northern Norwegian Sea; and IIb, Northeastern Greenland Sea), the amplitude of temporal fluctuations of community parameters is high. For example, the greatest fish biodiversity within the community of studied area during 1973–2006 was in the northern Norwegian Sea, where the Shannon Index exceeded 1.77. In this region, interannual fluctuations in diversity were also the smallest. The diversity of species decreased sharply in this region during 1993–1997, when a considerable increase in Atlantic herring biomass was observed, causing this species to dominate the community in the study area. In subsequent years, a diversity decrease was determined by short-term fluctuations in the abundance of species, such as blue whiting (1980), capelin (1983), cod (1990), and herring (2005).

In the other two areas (I and IIb), the processes of species diversity formation followed the same pattern in different years. However, quantitative indices in the Barents Sea were generally higher, with amplitude of diversity and abundance evenness lower than in the northeast Greenland Sea. The main species causing significant decrease in diversity and evenness in abundance distribution were capelin (1977, 1981, 1984, 1992, 2002) and, to a lesser degree, cod (1994, 2004).

Consequently, regardless of geographic latitude of the evaluated areas, the most notable fluctuations in biodiversity of the fish component of the community were observed during periods of considerable change in biomass of pelagic species (Karamushko 2010, 2012).

Northern ecosystems are characterized by a high degree of dominance of a limited number of fish species and relatively low indices of natural diversity. Hence, a decrease in the number of dominant species for any reason, including climatic, increases diversity because of a shared alignment of other, less numerous representatives of the community. Therefore, in ecosystems such as the Barents Sea, an increase in diversity, particularly a sharp one, may serve as an indicator of negative change (including decreasing fish harvests) in dominant fish populations. This in turn would lead to structural and functional alterations of the entire fish community and ecosystem.

Zoobenthos communities respond to interannual climate variability, as documented by a series of repeated surveys in the Barents Sea along the Kola Transect since 1921 (Deryugin 1925; Nesis 1960; Galkin 1998; Denisenko 2008). The search for vectors of changes to total benthic biomass in relation to temperature regime is a focal theme in hydrobiology and is addressed within the problem of climate-driven fluctuations in stocks of marine natural resources. Early studies indicated a general decrease in benthic biomass during cold periods (Bochkov and Kudlo 1973). However, another explanation for reduced benthic biomass after cooling was found in the central and northeast Barents Sea (Denisenko 2006). These results showed that the response of benthos to climate fluctuation depends on specific regional features of the environment.

In summary, long-term benthic studies documented that benthic communities track oceanographic variability in terms of abundance and ratios of biogeographic affiliation with a lag of several years. We did not, however, observe shifts to alternate stages or dramatic changes in the composition biomass of bottom communities. Any long-term ecosystem management plan for the Barents Sea and adjacent waters (Matishov et al. 2003) must incorporate these natural cyclic changes.

The results of the dynamics of zoogeographical composition and benthos biomass along the Kola Transect showed these two characteristics to have similar changes, confirming previously recorded trends (Nesis 1960; Denisenko 2006, 2008).

The close relationship between indicator species and fluctuations of ambient temperatures in natural habitats of the Barents Sea was shown previously, using the example of mollusks (Galkin 1998). During periods of positive temperature anomalies, the distribution of warm water bivalves, such as Clinocardium pinnulatum (syn. Parvicardium elegantulum), Mytilus edulis, and Modiolus modiolus, as well as gastropods such as Gibbula tumida, extended northeast by 40–150 nautical miles. Furthermore, ranges of subtropical-boreal species (such as Chlamys sulcata, Anomia ephippium, Helcion pellucidus, and others) extended into the Barents Sea, with samples being taken along the Kola Transect. Arctic fauna (such as Buccinum hydrophanum, Limatula hyperborea, and others) were confined to areas that remained under the influence of cold currents. The reverse process was observed during cooling phases (Fig. 11) (Galkin 1998).

Summary and conclusions

Polar areas of the earth have been subject to natural cyclic changes in climate and ice cover during the geologic past and present epochs. This cyclic characteristic might be somewhat disturbed by anthropogenic factors, such as the greenhouse effect. Thus, it is worth studying both natural and anthropogenic components of climate variations.

Based on results of MMBI RAS expeditionary research and analysis of information from oceanographic and hydrobiological databases, patterns of recent climatic changes in the Western Arctic seas were determined. Warming of the Barents Sea during the first decade of the twenty-first century may be compared in terms of its intensity and duration to the warming of the 1920s–1930s. During 2007–2010, however, this process was reversed, and oceanographic indices approached the long-term average. It is possible that future changes in the climate system will continue to be cyclic. Therefore, the recent reduction in ice cover observed in the Russian Arctic may cease in the next two to three decades (Frolov et al. 2009). Within a framework of long-term forecasting and retrospective analysis of climate alterations, hydrobiological indicators may be incorporated. These include species composition, density, biomass, and migration routes of abundant marine organisms, because they are significantly affected by climate fluctuations.

Rising water temperatures lead to migration of Atlantic cod from the Norwegian Sea to the central and southeast Barents Sea, hence the northward extension of fishing areas. In contrast, during periods of cooling, the distribution of commercially exploited cod localizes in the southern and southwest Barents Sea. Thus, changes in direction of cod migrations and geography of commercial fishing may serve as indicators of climatic fluctuations in the Barents Sea.

The rapid increase in abundance of introduced Kamchatka crab during the early 2000s may be interpreted as a response to general warming and to strengthening advection of Atlantic waters.

Bottom fauna of the Barents Sea mainly responded to strong and long-term climatic anomalies. Warm water fauna composition changes according to climatic cycles, with a reduction in organism abundance, species number, and total habitat area. Total benthic biomass and biomass of dominant species respond after time lags of 4–8 years. During periods of maximum warm and cold states of the near-bottom layer of the sea, distributions of benthic species differ substantially. For instance, the natural habitat of boreal species expands during warming, whereas that of Arctic species expands during cooling. With increasing mean annual temperatures, biomass of boreal-Arctic species increases, whereas that of Arctic species decreases.

Overall, marine ecosystems are well adapted to long-term variability of environmental conditions. This is especially confirmed by the 75 years of MMBI research. During this period, the reproductive state and food chain abundance of marine mammals, colonial seabirds, commercial fish, and invertebrates have depended on both regional climate and scale of exploitation. In terms of biodiversity, artificial introduction of species from the Far East (humpback salmon and Kamchatka crab) and the invasion of other alien species also contribute to structuring the marine ecosystem.