Degen16 MEPS546

Vol. 546: 1–16, 2016
doi: 10.3354/meps11662
Mar Ecol Prog Ser
Published March 21
Patterns and drivers of megabenthic secondary
production on the Barents Sea shelf
Renate Degen1,*, Lis Lindal Jørgensen2, Pavel Ljubin3, Ingrid H. Ellingsen4,
Hendrik Pehlke1, Thomas Brey1
Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research, Am Handelshafen 12, 27570 Bremerhaven, Germany
Institute of Marine Research, 9294 Tromsø, Norway
Polar Research Institute of Marine Fisheries and Oceanography, 183038 Murmansk, Russia
SINTEF Fisheries and Aquaculture, Brattørkaia 17C, 7010 Trondheim, Norway
ABSTRACT: Megabenthos plays a major role in the
overall energy flow on Arctic shelves, but information
on megabenthic secondary production on large spatial
scales is scarce. Here, we estimated for the first time
megabenthic secondary production for the entire Barents Sea shelf by applying a species-based empirical
model to an extensive dataset from the joint Norwegian−Russian ecosystem survey. Spatial patterns and
relationships were analyzed within a GIS. The environmental drivers behind the observed production pattern
were identified by applying an ordinary least squares
regression model. Geographically weighted regression
(GWR) was used to examine the varying relationship of
secondary production and the environment on a shelfwide scale. Significantly higher megabenthic secondary production was found in the northeastern, seasonally ice-covered regions of the Barents Sea than in the
permanently ice-free southwest. The environmental
parameters that significantly relate to the observed pattern are bottom temperature and salinity, sea ice cover,
new primary production, trawling pressure, and bottom
current speed. The GWR proved to be a versatile tool
for analyzing the regionally varying relationships of
benthic secondary production and its environmental
drivers (R2 = 0.73). The observed pattern indicates tight
pelagic−benthic coupling in the realm of the productive
marginal ice zone. Ongoing decrease of winter sea ice
extent and the associated poleward movement of the
seasonal ice edge point towards a distinct decline of
benthic secondary production in the northeastern
Barents Sea in the future.
Six environmental parameters significantly relate to the
observed pattern of high megabenthic secondary production in the northeastern Barents Sea.
Image: R Degen
KEY WORDS: Arctic · Benthos · Megafauna · Pelagic−
benthic coupling · Geographically weighted regression ·
GWR · Geostatistics
Benthic secondary production constitutes an important pathway of energy flow on Arctic shelves
(Piepenburg et al. 1995). Accordingly, it is of particular ecological and economical interest to estimate
secondary production in the Barents Sea, which holds
*Corresponding author:
© The authors 2016. Open Access under Creative Commons by
Attribution Licence. Use, distribution and reproduction are unrestricted. Authors and original publication must be credited.
Publisher: Inter-Research ·
Mar Ecol Prog Ser 546: 1–16, 2016
one of the world oceans’ richest fisheries (Wassmann
et al. 2006b). The Barents Sea ecosystem is characterized by the interplay of polar and Atlantic water
masses, and by a seasonal ice cover (Ingvaldsen &
Loeng 2009). However, the ongoing rapid retreat of
sea ice raises questions concerning the current and
future productivity of the area. The joint Norwegian−Russian Ecosystem Survey (Michalsen et al.
2013) produced an extensive dataset on megabenthos
that covers the entire Barents Sea. This dataset is
unique in spatial coverage and resolution and thus
allows us, for the first time, to model megabenthic
secondary production for an entire Arctic shelf.
Macrozoobenthos (i.e. size class of animals <1−2 cm,
usually sampled with grabs or box cores) of Arctic
shelves and the Barents Sea in particular has been
studied extensively in recent decades (Cochrane et
al. 2009 and references therein). It is a significant
player in benthic carbon cycling and serves as food
for a variety of higher trophic level and commercially
important species like cod or halibut (Clough et al.
2005, Renaud et al. 2007). Significantly less information is available on benthic megafauna (animals of a
size visible on photos or caught via bottom trawling)
of Arctic shelves (Bluhm et al. 2009, Roy et al. 2014,
Grebmeier et al. 2015b, Jørgensen et al. 2015), although it represents an important compartment of
benthic energy flow (Piepenburg et al. 1995). Moreover, little is known about Barents Sea benthic secondary production of either mega- or macrobenthos,
i.e. the newly formed biomass per unit of area and
time, despite the general awareness of the benthic
compartment’s role in energy flow and food webs
(Piepenburg et al. 1995, Cochrane et al. 2009). To
date, only 1 study (Ke˛dra et al. 2013) has dealt with
benthic community secondary production, compared
to some available literature on single benthic species
production (e.g. Bluhm et al. 1998) and the numerous
publications on primary production and pelagic secondary production (Sakshaug et al. 2009, Dalpadado
et al. 2014). Kedra et al. (2013) estimated benthic
infauna and epifauna secondary production on the
Spitsbergen Bank to amount to ~2 and ~22 g C m−2
yr−1, respectively. No information is currently available from the Barents Sea region on the megabenthic
community production to biomass (P:B) ratios, which
represent the rate of biomass turnover (Benke 2012).
The first systematic large-scale study on Barents
Sea megafauna results from the joint Norwegian−
Russian Ecosystem Survey (Michalsen et al. 2013)
and was published recently (Anisimova et al. 2010,
Jørgensen et al. 2015). Jørgensen et al. (2015) provided the first explicit, large-scale analysis of Barents
Sea megafauna community composition and distribution patterns and identified a northern and a
southern megafauna assemblage which can be
divided further into 2 subregions each. The subregions are characterized by the environmental
parameters water depth, temperature, salinity, and
the number of ice-days by means of a canonical correspondence analysis. The border between these
assemblages coincides with the encounter of colder
Arctic (< 0°C) and warmer Atlantic (> 3°C) bottom
water, hence, it is termed the ‘Benthic Polar Front’
(see Fig. 1) (Jørgensen et al. 2015). This front runs
slightly differently than the oceanic Polar Front,
which is shaped by surface water masses (Fig. 1)
(Loeng et al. 1997). The northern megafauna assemblage has more taxa, higher abundance, and higher
biomass than the southern assemblage (Jørgensen et
al. 2015). In the north, benthic biomass is dominated
by echinoderms, followed by Crustacea and Porifera.
The southern assemblage can be further separated in
a south-western section where Porifera dominate,
followed by Echinodermata, and a south-eastern section where benthic biomass is more evenly shared
among Echinodermata, Mollusca, and Porifera.
Generally food input is seen as the main driver of
benthic fauna distribution and biomass at large
regional scales, while seabed attributes explain patterns more significantly at local scales (Pearson &
Rosenberg 1978, Piepenburg 2005, Carroll et al.
2008). Arctic shelf communities reflect the primary
production regime of the overlying water column in
terms of biomass, abundance, and production, suggesting a tight pelagic−benthic coupling (Tamelander et al. 2006, Grebmeier et al. 2006, 2015a). This
also holds true for the Barents Sea macrofauna,
which shows the highest biomass on the shallow
Spitsbergen and Central Banks, the Novaya Zemlya
Bank, and the Pechora Sea (Cochrane et al. 2009),
i.e. areas that have high values in models of primary
productivity (Wassmann et al. 2006b). Zenkevitch
(1963) pointed out that in the Barents Sea the highest
benthic biomass along the Polar Front correlates
inversely with water temperature, and presumed that
this relates to the fact that areas with the coolest
bottom water coincide with areas of the most active
mixing and subsequent upwelling. In the Polar Front
area, less saline Arctic water overlies dense and
nutrient-rich Atlantic waters. Turbulent gyres provoke mixing of the different water masses, and iceedge upwelling of nutrients may occur (Loeng et al.
1997, Slagstad et al. 1999, Denisenko 2002, Mundy et
al. 2009). Especially in shallower areas like Spitsbergen and the Central Bank, wind and tidal forcing are
Degen et al: Arctic sea ice and benthic production
other sources of enhanced vertical mixing (Slagstad
& McClimans 2005, Sundfjord et al. 2008). The
intense ice-edge bloom with associated high biomass
in combination with turbulent mixing can promote
extensive vertical export of high quality particulate
organic matter (POM; Reigstad et al. 2008). However,
Arctic studies that link benthic production patterns to
ecosystem processes on large spatial scales are
scarce (e.g. Highsmith & Coyle 1990), and this is
even more the case for benthic megafauna. Likewise,
little attention has been given to regional patterns of
megafauna production on the Barents Sea shelf, but
exactly this information is needed to develop reliable
energy flux models (e.g. Dommasnes et al. 2001) and
future scenarios for this rapidly changing ecosystem.
Here, we analyzed for the first time megabenthic
community production for an entire Arctic shelf, i.e.
the Barents Sea. We estimated secondary production
by means of a species-based empirical model (Brey
2012) and used a global regression model to identify
significant drivers of the observed production pattern. Owing to the substantial regional variations of
environmental conditions (water depth, temperature,
salinity, sediment structure, and sea-ice concentration) and human impact (commercial trawling), we
applied a geographically weighted regression (GWR)
model to examine the relationship of secondary production with the environment in space (Fotheringham et al. 2002). To our knowledge, this is the first
time such geo-statistical techniques have been used
to map and spatially analyze marine benthic secondary production.
We aimed to (1) estimate total and major group secondary production (P) of megafauna for the entire
Barents Sea shelf, (2) identify regional patterns, (3)
identify the significant environmental drivers behind
the observed patterns, and (4) analyze their regionally varying relationship to P.
Arctic Ocean in the north, the island Novaya Zemlya
in the east, the Norwegian and Russian mainland in
the south, and the Norwegian Sea and Fram Strait
in the west (Ozhigin et al. 2011) (Fig. 1). Three main
water masses characterize the Barents Sea (Fig. 1):
nutrient-rich Atlantic water with temperatures > 3°C
and salinity of > 35 and coastal water with temperatures in a wider range and salinity < 34.7 enter the
Barents Sea in the southwest, and Arctic water with
temperatures < 0°C (core temp. <−1.5) and salinity of
34.4−34.7 enters the shelf between Svalbard and
Franz Josef Land, between Franz Josef Land and
Novaya Zemlya, and via a small inflow from the Kara
Sea south of Novaya Zemlya (Ingvaldsen & Loeng
2009). A Polar Front (grey line in Fig. 1) separates the
warm Atlantic water from the cold Arctic water, and
thus the permanently ice-free areas in the southwest
from the seasonally ice-covered northeastern areas
(Loeng et al. 1997). Regarding bottom temperature,
the front runs slightly differently and is termed the
‘Benthic Polar Front,’ separating a northern from a
southern faunal assemblage (Jørgensen et al. 2015)
(dashed grey line in Fig. 1). Sediment structure on
the shelf is heterogeneous; fine mud dominates
deeper areas, and coarser substrates are found in
shallower areas with stronger currents (Jørgensen et
al. 2015). Current speed on the Barents shelf is moderate, with highest values of > 0.25 m s−1 in the Norwegian Coastal Current, but just ca. 0.1 m s−1 in the
western outflow (Ingvaldsen & Loeng 2009). Pelagic
primary production is highest in the southwestern
regions influenced by nutrient-rich Atlantic water,
with values >100 g C m−2 yr−1, and is thought to be
lowest in the seasonally ice-covered northeast (Wassmann et al. 2006b), although information on annual
rates of sea-ice-associated production is still insufficient. Trawling impacts on the benthos are highest in
the areas harboring rich accessible fish stocks, i.e. the
ice-free southern areas in particular (Lyubin et al. 2011),
but reliable geo-referenced information of trawling
pressure for the entire Barents Sea is currently lacking.
Study area
Faunal dataset
The Barents Sea is the deepest of all circum-Arctic
shelf seas with depths down to 500 m in the western
troughs (Jakobsson et al. 2004). Generally, the
bathymetry is characterized by several shallow shelf
banks that are segregated by a complex pattern of
deeper depressions (> 200 m), and the average depth
is 230 m (Zenkevitch 1963, Piepenburg et al. 1995,
Ingvaldsen & Loeng 2009). The Barents Sea covers
an area of 1.6 million km2 and is surrounded by the
Benthic megafaunal abundance and biomass data
were derived via the joint Norwegian−Russian Ecosystem Survey (Michalsen et al. 2013). The dataset of
398 bottom trawl stations presented in this study was
compiled by experts on 3 Norwegian and 1 Russian
research vessel from August to October 2008 and
August to October 2009. Samples were taken with a
Campelen 1800 bottom trawl, towed for 15 min at 3
Mar Ecol Prog Ser 546: 1–16, 2016
Fig. 1. Barents Sea bathymetry and scheme of the main water masses. The approximate positions of the Polar Front and the
Benthic Polar Front are indicated by a grey and a dashed grey line, respectively. Bathymetry is based on the International
Bathymetric Chart of the Arctic Ocean basemap ( BIC: Bear Island Channel; CB: Central Bank; HID: Hopen
Island Deep; KI: Kolguyev Island; NB: North Bank; NZB: Novaya Zemlya Bank; PS: Pechora Sea; SB: Spitsbergen Bank
knots. The horizontal opening of the trawl was 11.7 m,
the mesh size varied from 80 mm (stretched) at the
front to 16−22 mm at the cod end. The standard distance between stations was 65 km. On board ship,
the benthic megafauna was separated from the fish
and shrimp catch, identified to the species level, and
counted, and wet-weight biomass was measured
with electronic scales. Colonial species were treated
as 1 individual per colony. For more information on
the joint Norwegian−Russian Ecosystem Survey, see
Michalsen et al. (2013) and Jørgensen et al. (2015).
Environmental dataset
Water depth (m) was estimated with the ship’s
depth sounders at each sampling station. Mean bottom water temperature (°C), bottom water salinity
(psu), and current speed (m s−1) for the period
January 2008 to December 2009 were derived from
the numerical ocean model SVIM (Lien et al. 2014).
The standard deviation of mean sea ice concentration
(%) of the period 2001 to 2008 was estimated from
monthly average sea ice concentration maps provided
Degen et al: Arctic sea ice and benthic production
by NORMAP (10 km grid, via
an algorithm in R software. Mean new primary production (NPP; g C m−2 yr−1) for the period 2001 to 2008
was derived from the SINMOD model (see Wassmann
et al. 2006a). New production is a measure of the
maximum harvestable production or export production from the system (Wassmann et al. 2006a). As Arctic invertebrate megafauna are on average long-lived
(Piepenburg 2005), we presumed the integration of
the previous 8 yr to be appropriate for structuring the
community composition in 2008 and 2009. Sediment
types were characterized into 6 classes based on the
classification scheme of Vinogradova & Litvin (1960),
with class 1 being sand, class 2: silty sand, class 3:
sandy silt, class 4: mud, class 5: clay−silt, and class 6:
clay. Insufficient information is available on sediments in the Spitsbergen area, and hence we estimated the sediment class for several stations based
on information from environmentally comparable
reference stations, considering water depth, distance
to coast, and current speed as parameters. As georeferenced information on trawling pressure in the
Barents Sea is lacking for the years prior to sampling,
we categorized the trawling information of the Russian fisheries fleet from 2002 to 2005 (i.e. number of
hauls per region), provided in the illustrations of Lyubin et al. (2011), into 4 classes with 1: no trawling, 2:
low trawling, 3: intermediate trawling, and 4: high
trawling pressure. Table 1 provides minimum, maximum, and mean values of the environmental parameters considered in this study. The total environmental
information for each of the 398 stations can be found
in the PANGAEA open access library (
Estimating secondary production (P) and
productivity (P:B ratio)
The secondary production (P) of Barents Sea
megafauna was estimated with an empirical artificial
neural network model (for detailed information on the
model see Brey 2012; for another application in Arctic
regions see Nilsen et al. 2006 and Degen et al. 2015).
The model is implemented in an excel spreadsheet
and can be freely accessed via Abundance and biomass data given as individuals and biomass (g wet
weight) per 15 min haul, respectively, were recalculated to m−2 by assuming an average trawled area
of 18 000 m2 per station (Anisimova et al. 2010). As
mean body mass (M) in Joules is the main model input
parameter, biomass was divided by abundance for
each species and station and converted to Joules
using the conversion factor database of Brey (2012,
database version 4,
virtualhandbook). Further model input parameters
are bottom temperature (K), water depth (m), 4 taxonomic categories (Mollusca, Annelida, Crustacea,
Echinodermata), 7 lifestyle categories (infauna, sessile, crawler, facultative swimmer, herbivore, omnivore, carnivore), 4 environmental categories (lake,
river, marine, subtidal), and a marker for exploitation.
All categorical variables were binary (0 or 1). The necessary ecological information for each species was extracted from literature and online resources (for a list
of literature and web sources, see the Supplement at
Species that did not belong to any of the 4 taxonomic
categories of the model were grouped by the taxonomic category their body form resembled most. The
output of the model is the population production to
biomass (P:B) ratio (yr−1), including upper and lower
95% confidence limits. Population P was calculated
by multiplying the P:B ratio with population biomass,
previously converted to g C m−2 yr−1. Total community
P was calculated by adding up all population values
per station. Production per phylum was calculated for
Annelida, Arthropoda, Cnidaria, Echinodermata, Mollusca, and Porifera. The phyla Brachiopoda, Bryozoa,
Cephalorhyncha, Chordata, Echiura, Nemertea, Platyhelminthes, and Sipuncula occurred in very low abundances (< 0.1 ind. m−2) and biomasses (< 5 mg C m−2)
and were therefore combined into the group ‘Others.’
Table 1. Environmental parameters considered in this study: longitude (°E),
latitude (°N), water depth (m), temperature (°C), salinity (psu), current speed
(m s–1), standard deviation of mean sea ice concentration from 2001 to 2008
(%), and new primary production (NPP, g C m−2 yr−1)
Mean 35.59
Depth Temp.
68.47 20.00 −1.49
82.05 485.00 5.92
74.51 248.28 1.55
Sea ice
Geostatistical analysis
P, P:B ratio, NPP, trawling pressure,
water depth, temperature, salinity, current speed, standard deviation of mean
sea-ice concentration, and sediment
structure were projected spatially using
a GIS environment (ArcGIS Desktop:
Release 10, Environmental Systems
Mar Ecol Prog Ser 546: 1–16, 2016
Research Institute). The WGS 1984 Stereographic
North Pole projection was used. The data distribution
was visually inspected, and outliers (2 stations) were
eliminated from the dataset. All analytical methods
applied can be found in the spatial statistics toolbox of
ArcGIS (10.1). Grouping analysis based on bottom
temperature was used to separate the dataset into a
southwestern (SW) and a northeastern (NE) group.
Hotspot analysis (Getis-Ord Gi*) was used to identify
regions of significantly higher P and P:B ratio than the
overall mean. This method identifies statistically significant hotspots, i.e. regions where stations with high
(or low) values cluster together. The global regression
model ordinary least squares (OLS) was used to determine the environmental parameters significantly correlated to the observed patterns of P and P:B ratio.
The independent input variables in the OLS model
were water depth, temperature, salinity, current
speed, sea ice concentration, NPP, trawling pressure,
and sediment structure. The significant variables
were consequently used as input in the GWR model.
GWR accounts for the spatial variability of input data
by incorporating spatially varying relationships in the
regression analysis (Fotheringham et al. 2002) and
was used to visualize the regionally varying relationships between P, P:B ratio, and the explanatory variables. This model is appropriate when more than 100
features (here sample stations) are available, no binary outcomes are predicted, and a projected coordinate system is used. A higher R2 value in the GWR
model than in the OLS model and a distinct difference
in the corrected Akaike’s information criterion (AICc)
between the 2 models indicate that the use of GWR
was appropriate for the present dataset. All skewed
input data were transformed to approach a normal
distribution. Salinity data had to be grouped in classes
because transformation did not sufficiently reduce
skewness. The graphical output of the GWR model
are maps of correlation coefficients with hot-to-cold
rendering indicating regional variation in the relationship of the dependent and the independent variables. The models were tested to fulfill all required assumptions necessary to guarantee a reliable model
output. Moran’s I was used to test for potential spatial
autocorrelation of the regression residuals.
Additional statistics
ANOVA was conducted to test for significant
differences between the secondary production and
the productivity of the faunal group in the SW and the
group in the NE Barents Sea (previously identified
with grouping analysis in GIS) using the JMP® software package, Version 10.0 (SAS Institute).
Total community production (P) and productivity
(P:B ratio)
Total community production per station ranged
from 0.015 to 105 mg C m−2 yr−1 (Table 2, Fig. 2a; The grouping analysis based on bottom temperature separated
the dataset into an NE and an SW part (R2 = 0.70).
Production was significantly higher in the NE part
than in the SW part (F = 106.27, p < 0.0001). The
hotspot analysis performed on Box-Cox transformed
P data identified 4 hotspots in the northern region
Table 2. Minimum, maximum, and mean of total community
abundance (ind. m−2), biomass (mg C m−2), secondary production (P; mg C m−2 yr−1), and productivity (production to
biomass ratio, P:B) per major taxonomic group. The group
‘Others’ includes taxa occurring in very low numbers
(< 0.1 ind. m−2) and biomass (< 5 mg C m−2), i.e. Brachiopoda,
Bryozoa, Cephalorhyncha, Echiura, Nemertea, Platyhelminthes, and Sipuncula. Highest mean values for P and P:B ratio
are highlighted in bold
Abundance Biomass Production P:B
Arthropoda Min
< 0.01
< 0.01
< 0.01
< 0.01
Degen et al: Arctic sea ice and benthic production
Fig. 2. Total community (a) production (P) (mg C m−2 yr−1) and (b) productivity (production to biomass ratio, P:B; yr−1)
and 3 coldspots in the south and northwest (Fig. 3a).
The largest hotspot area is located west of Novaya
Zemlya, and the hotspot with the highest local benthic production is located SW of Franz Josef Land
(North Bank). Another hotspot is located on the
southern slope of Novaya Zemlya, and the smallest
resides in the central Barents Sea (Hopen Island
Deep and western slopes of the Central Bank). The
geographically largest coldspot is located in the SW
Barents Sea. The 2 other and geographically smaller
coldspots are located northwest of Spitsbergen and in
the southeast, west of Kolguyev Island. Total community productivity (P:B ratio) ranged from 0.038 to 0.841
(Table 2, Fig. 2b). The P:B ratio was not significantly
different between the SW and the NE region (F =
3.00; p = 0.084), but the highest P:B ratios were found
in the SW region (Fig. 2b). The hotspot analysis identified 1 large hotspot in the central Barents Sea
Fig. 3. Interpolated (inverse distance weighting method) standard deviation of G* scores of the hotspot analysis of total community (a) production (P) and (b) the production to biomass ratio (P:B). Red indicates significantly higher values than the
mean; blue indicates significantly lower values than the mean
Mar Ecol Prog Ser 546: 1–16, 2016
Table 3. Results of the ordinary least squares (OLS) and geographically
weighted regression (GWR) models for total community production and
the production to biomass ratio (P:B). Significant parameters (p < 0.05) are
highlighted in bold. NPP: new primary production, AICc: corrected
Akaike’s information criterion
Depth (m)
Temperature (°C)
Salinity (class)
Current speed (m s−1)
Sea ice concentration (SD)
NPP (mg C m−2 yr−1)
Trawling pressure
Sediment (class)
< 0.0001
< 0.0001
< 0.0001
Spitsbergen and north of Novaya
Zemlya, Arthropods have a hotspot in
the southwestern Barents Sea, north of
Novaya Zemlya, and in the Pechora Sea,
and Mollusca have a P:B ratio hotspot in
the SW and southern Barents Sea and in
the Pechora Sea. The figures of P and P:B
ratio per major group can be found in
Figs. S1 & S2 in the Supplement, and detailed information about regional megafauna community composition based on a
dataset from the year 2011 can be found
in Jørgensen et al. (2015).
Global model (OLS)
The OLS model for secondary production
fitted the data with R2 = 0.41 and
an AICc of 1800.12. The AICc is a
measure of the relative quality of a statistical model for a given dataset; accor dingly, it can be used to compare OLS and GWR
(Hopen Island Deep and western slopes of the Cenmodels based on the same input parameters. The
tral Bank) and 2 smaller hotspots northwest of SpitsOLS model identified 6 parameters that explained
bergen and in the southwest deep Bear Island Channel (Fig. 3b).
Major group production (P) and
productivity (P:B ratio)
Echinodermata clearly dominate the
megabenthic production in the Barents Sea
by contributing 50% to the total production,
followed by Arthropoda (18%), Annelida
(12%), and Mollusca (7%). Cnidaria, Porifera, and all other taxa contribute below
5%. The overall pattern of P is mainly shaped by Echinodermata and Arthropoda,
with both showing clear hotspots in the
western area of Franz Josef Land and in the
southeastern Barents Sea (see Fig. S1 in the
Supplement at
suppl/m546p001_supp.pdf). Arthropods have
a third hotspot on the Novaya Zemlya Bank.
Regarding biomass (mg C m−2), the pattern
is different, with Echinodermata contributing 61%, Arthropoda 14%, and Porifera 8%
to the overall biomass. Highest productivity
was found in the phyla Annelida (mean
0.61), Arthropoda (0.20), and Mollusca
(0.17) (Table 2). Annelida have productivity
hotspots in the SW Barents Sea, north of
Fig. 4. Mapped R2 values from the geographically weighted regression
(GWR) model of megabenthic secondary production (P). Dark red points
indicate areas with highest model fit (R2 values 0.62−0.82); blue points
indicate areas of low model fit
Degen et al: Arctic sea ice and benthic production
the observed production patterns significantly
(Table 3). The OLS model for the P:B ratio had a
model fit of R2 = 0.15 and AICc = −1114.12 and
identified 5 parameters that significantly explained
the variance in the P:B ratios (Table 3). Here, a
significant Jarque-Bera statistic (p = 0.0036) indicated severe model bias.
Local model (GWR)
The GWR model for production (P) based on the 6
significant variables identified with OLS (Table 3)
displayed an overall model fit of R2 = 0.73 and AICc =
88.40. The GWR model for productivity (P:B ratio)
based on the 4 significant variables identified with
OLS had an overall model fit of R2 = 0.53 and AICc =
−59.51. The spatial distribution of stations with the
highest model fit is shown in Fig. 4. The 6 significant
correlation coefficients identified with GWR (shown
in Fig. 5a−f) highlight the regionally varying relationships of production and the environment. As the
GWR model of P:B ratio is based on potentially
biased assumptions from the OLS model, it should be
interpreted cautiously. Consequently, we restrain
from interpreting spatial P:B ratio patterns and model
output in this study.
Remarks on the methodology
Data from trawl samples are generally considered
semi-quantitative and gross estimates (Eleftheriou &
Moore 2005). However, when trawling is carried out
consistently over a large number of stations, as was
the case in the joint Norwegian−Russian Ecosystem
Survey (Michalsen et al. 2013), relative spatial patterns can be identified (Mueter & Litzow 2008, Fossheim et al. 2015, Jørgensen et al. 2015). In this study,
we accordingly do not present estimated secondary
production per station or in detail, but rather focus on
the regional differences and relationships.
Patterns of megabenthic secondary production
Secondary production of Barents Sea megafauna is
significantly higher in the NE seasonally ice-covered
areas than in the permanently ice-free SW areas. We
detected 4 regional hotspots of megabenthic secondary production: the area west of and on the southern
slope of Novaya Zemlya, the region southwest of
Franz Josef Land, and a smaller hotspot in the central
Barents Sea. All of these regions of high megafaunal
production correspond to the regions of high benthic
biomass reported in previous studies (Zenkevitch
1963, Denisenko 2001, 2002, Wassmann et al. 2006b).
The region southwest of Franz Josef Land was only
mentioned by Zenkevitch (1963), but not in other
studies, probably because the regions > 78° N were
not included in later expeditions due to dense ice
cover that impeded sampling (see results assembled
by Wassmann et al. 2006b). The region of Spitsbergen Bank (Fig. 1), previously reported as a region of
high benthic biomass (Zenkevitch 1963, Denisenko
2001, 2002, Wassmann et al. 2006b, Cochrane et al.
2012, Kedra et al. 2013), was not identified as a
hotspot, neither in biomass nor in production, in our
megafauna study. Jørgensen et al. (2015) found intermediate biomasses in this area in their analysis of
trawl data from the year 2011. One possible explanation for this discrepancy was proposed by Denisenko
(2001): the author found a significant negative correlation of benthic biomass with the intensity of bottom
trawling and concluded that trawling is one of the
main causal factors of long-term fluctuations of bottom communities in the Barents Sea.
The relative contribution of major taxonomic
groups to overall benthic secondary production was
clearly dominated by Echinodermata, leading with
50%, followed by Arthropoda with 18% and Annelida with 12% (see also Table 2). Regarding biomass
(mg C m−2), the dominance of Echinodermata was
even clearer with 61% of the overall biomass, followed by Arthropoda with 14%, a pattern also
reported to be typical for epibenthic communities in
the Pacific Arctic (Bluhm et al. 2009, Grebmeier et al.
2015a). Studies based on quantitative sampling
methods have reported a different biomass pattern,
with Mollusca (predominantly bivalves) dominating
with 35% before Echinodermata with 19% and
Arthropoda with 15% (Wassmann et al. 2006b). In
accordance with Jørgensen et al. (2011), who compared the sample efficiency of Van Veen grabs
(quantitative) and beam trawls (qualitative), we recommend the use of complementary techniques in
future studies to ensure the sampling of the entire
benthic community at each station. The studies by
Grebmeier et al. (2015b) from the Chukchi Sea or by
Kedra et al. (2013) from the Spitsbergen Bank are
good examples of how the use of multiple sampling
devices leads to a better understanding of the entire
benthic community. Nevertheless, several studies
performed on smaller spatial scales confirm the dom-
Mar Ecol Prog Ser 546: 1–16, 2016
Fig. 5. Correlation coefficients derived from the geographically weighted regression (GWR) megafauna production model for
the 6 significant parameters: (a) bottom temperature, (b) salinity, (c) sea-ice concentration (standard deviation), (d) new
primary production (NPP), (e) trawling pressure, and (f) current speed. Red circles in left panels indicate significant positive
correlations; blue circles indicate significant negative correlations of secondary production (P) and the respective parameter.
The maps on the right show the interpolated (inverse distance weighting method) environmental parameters, with red, blue,
and yellow areas indicating high, low, and intermediate values, respectively
Degen et al: Arctic sea ice and benthic production
Fig. 5 (continued)
Mar Ecol Prog Ser 546: 1–16, 2016
inance of echinoderms in the overall benthic biomass
and energy flow on Arctic shelves (Piepenburg 2005
and references therein, Renaud et al. 2007).
Drivers of megabenthic secondary production
At large regional scales, food input is reported to
be the main driver of distribution and biomass of
benthos, and of benthic production accordingly
(Grebmeier et al. 1988, Piepenburg 2005). Arctic
benthic biomass hotspots were previously reported
to coincide with areas of high primary production
and with ice-edge areas (Denisenko 2002, Wassmann et al. 2006b, Carroll et al. 2008, Grebmeier et
al. 2015a). We used NPP (Wassmann et al. 2006a) as
a proxy of food input to the benthos in our regression
model and expected a positive correlation with
megabenthic production, i.e. high P values in areas
of high NPP. However, we found a reverse pattern: P
was negatively related with NPP and was significantly higher in regions that are seasonally sea ice
covered and reported to be of distinctly lower pelagic
primary production (Wassmann et al. 2006b).
Besides the negative correlation with NPP, we
found P to be negatively correlated with bottom
water temperature, positively with salinity, and positively with the standard deviation of sea ice concentration. All of these factors relate to some extent to
the marginal ice zone (MIZ). Temperature relates to
the MIZ because in the Barents Sea, the maximum
extent of colder Arctic water masses and the Polar
Front often coincides with the sea ice extent in winter
or early spring (Wassmann et al. 2006b). Salinity
effects on P were significant in areas where strong
mixing between Atlantic and Arctic water masses
occurs, i.e. along the Polar Front (Wassmann et al.
2006b). The standard deviation of sea ice concentration is an obvious proxy for the MIZ. The bulk of the
total annual primary production of the northern and
NE Barents Sea takes place in the MIZ (Denisenko
2002, Wassmann et al. 2006a, Sakshaug et al. 2009).
Spring ice melt gives rise to a nutrient-rich euphotic
zone that supports a distinct phytoplankton bloom in
the MIZ which moves constantly poleward while
receding from its winter position at the Polar Front
(Piepenburg et al. 1995, Wassmann et al. 2006a, Leu
et al. 2015). This ice-edge bloom induces vertical carbon flux, which might be especially high in spring,
before pelagic production and consumption are balanced, or in years where a mismatch between primary and secondary producers occurs (Tamelander
et al. 2006, Eiane & Tande 2009, Leu et al. 2015). The
recent study by Leu et al. (2015) suggested highest
fluxes of ice-related POM and thus strongest sympagic−pelagic−benthic coupling in the post-bloom
phase in late spring, when large aggregates of ice
algae are released from the ice. Additionally, deep
vertical mixing in the area of the Polar Front
enhances the export of high-quantity and high-quality particulate organic carbon (POC), thus favoring
tight pelagic−benthic coupling (Olli et al. 2002, Reigstad et al. 2008). Wiedmann et al. (2014) emphasized
the role of mixing and turbulence for POC export to
deeper areas: they showed that export of POC can be
high in post-bloom situations, given that vertical mixing occurs. Tamelander et al. (2006) analyzed the
pelagic−benthic coupling in the Barents Sea MIZ
during summer and detected tight coupling between
surface production and the benthic community over
relatively small scales. Additionally, they detected a
high degree of heterogeneity, determined by water
mass properties like local upwelling and primary
production regimes. Our benthic secondary production estimates reflect this patchiness (Fig. 2), as there
are 4 significant hotspots and high variability between stations that are on average just 65 km apart
(Fig. 3,
We further investigated the strength of the
pelagic−benthic coupling in the northern seasonally
ice-covered region compared to the ice-free southern
regions by means of an ‘inverse’ approximation. The
ratio of mean secondary production (P) to mean
NPP (0.003:76 g C m−2 yr−1) in the southern region
is 0.00004; in the northern region it is 0.00016
(0.008:50 g C m−2 yr−1). This distinctly higher ratio in
the north indicates that here, either a larger part of
the NPP is channeled to the benthos (i.e. tighter
pelagic−benthic coupling), as also suggested by
Reigstad et al. (2008), or there is additional NPP that
is as yet not accounted for. Kedra et al. (2013) estimated a benthic carbon demand of up to 70 g C m−2
yr−1 to sustain the mean epibenthic production of
~22 g m−2 yr−1 at the Spitsbergen Bank. As this region
was a coldspot in our study, we assume a significantly higher benthic carbon demand in the regions
south and west of Novaya Zemlya or SW of Franz
Josef Land, which we identified as hotspots of megabenthic production (Fig. 3a). The mean estimated
NPP for these regions is 50 g C m−2 yr−1. If we add a
suggested contribution of ice algae primary production of a maximum of 25% (Wassmann et al. 2006b),
we would reach around 60 g C m−2 yr−1, not enough
to fulfill a carbon demand of potentially much more
than 70 g C m−2 yr−1(Kedra et al. 2013). Consequently, we assume that the production of sea-ice
Degen et al: Arctic sea ice and benthic production
algae and accordingly also the overall NPP in these
regions might be considerably higher than previously anticipated. Although reliable estimates of ice
algal production on large spatial scales are lacking,
our assumption is supported by several studies from
other Arctic regions (see Leu et al. 2015 and references therein). In the Canadian Arctic, Matrai &
Apollonio (2013) found ice and sub-ice microalgae to
contribute up to 50% to total net community production. Gradinger (2009) reported that on the shelves
and slope regions of the Chukchi and Beaufort Seas,
sea-ice primary production exceeds water-column
primary production significantly in early spring. The
few studies that have analyzed the benthic utilization
of sea-ice algae could indicate that benthic communities use sea-ice algae as rapidly as phytoplankton
(McMahon et al. 2006, Sun et al. 2007).
Additionally to the POC fluxes from the sea ice,
there might also be a substantial contribution of advected material originating from pelagic production
or from shallower macroalgae areas (Rosenberg
1995, Renaud et al. 2007, 2015), like along the shelf of
Novaya Zemlya and Franz Josef Land. Renaud et al.
(2007) pointed out that advected matter originating
tens of kilometers away could in fact explain the
enhanced benthic community oxygen demand they
observed in their study from the Beaufort Sea, but
also underpin the coincidence of the onset of algal
growth with the increase in benthic respiration they
observed in the same region. Renaud et al. (2015)
showed that most benthic taxa in an Arctic fjord feed
on a broad mixture of POM and macroalgal detritus,
even at depths > 400 m, thus strengthening the
importance of macroalgae as food source at least for
near-shore benthic communities.
The exclusion of smaller benthic size classes, the
meio- and macrofauna, from this study could be
another explanation for the non-correlation of food
input — here NPP— and benthic secondary production. A recent study from the Pacific Arctic indicates
that macrofaunal biomass is linked more closely to
food supply than the megabenthic epifauna (Grebmeier et al. 2015b). The authors presumed that
mobility may explain this pattern: while the more stationary macrofauna reflects the local food input
(either from sedimentation or from advection) more
closely, the larger mobile species are capable of
exploiting wider areas and a greater variety of food
sources. This opportunism may also be an important
strategy during the polar night, when photosynthetic
production is absent and alternative food sources or
stored reserves must be exploited (Berge et al. 2015).
Berge et al. (2015) found evidence of this oppor-
tunism in the rapid response of benthic scavengers to
food-fall items, as well as in the continuous annual
growth of a suspension feeder.
A similar pattern of clearly higher biomass in the
NE Barents Sea was reported for the mesozooplankton sampled during the joint Norwegian−Russian
Ecosystem survey in 2009 (Orlova et al. 2011). The
Arctic copepods Calanus glacialis and C. hyperboreus, both grazers on phytoplankton and ice algae,
clearly dominated the mesozooplankton biomass, a
pattern also reported from the zooplankton community in the Beaufort Sea (Darnis et al. 2008). The
calanoid copepods have an up to 2 yr life cycle and,
at least at certain developmental stages, spend the
overwintering period near the bottom (Melle & Skjoldal 1998, Berge et al. 2015). We assume that especially during the polar night the accumulated biomass of more than 1 generation of copepods might
constitute a substantial food source to benthic predators and scavengers (Jørgensen et al. 2015).
Apart from the environmental parameters that are
related to the MIZ and to food input, we further
detected a significant negative correlation between P
and current speed and a positive correlation between
P and trawling pressure. The former can be related
to the fact that regions with high current speed are
often dominated by filter feeders, such as Porifera,
Bryozoa, Hydrozoa, ascidians, and epifaunal bivalves
that contributed little to overall P in this study. The
positive correlation to trawling pressure was unexpected, as a previous study along the Kola transect
showed a significant negative correlation between
benthic biomass and bottom trawling (Denisenko
2001). Our findings probably relate to the fact that
long-term bottom trawling changes the age and size
spectrum of benthic communities from long-lived,
large sized individuals to short-lived, smaller sized
individuals and species (Callaway et al. 2007), an
effect also observed after ice scour disturbances (e.g.
Conlan et al. 1998). As the P:B ratio is inversely
related to body size, this shift increases the community P:B ratio and most likely community P as well.
Regionally varying relationships
Overall, our geo-statistical approach shows that the
tight pelagic−benthic coupling along the productive,
seasonally moving ice edge is of crucial importance
for northern Barents Sea megabenthic production.
Apart from the differences on large spatial scales, the
GWR approach enabled us further to investigate the
regional variation in the correlation between P and
Mar Ecol Prog Ser 546: 1–16, 2016
environmental parameters (Fig. 5a−f). Here, we focus
on salinity (Fig. 5b) and on sea-ice concentration
(Fig. 5c). Salinity is correlated positively with P in a
regional band resembling the approximate position
of the Benthic Polar Front (Figs. 1 & 5b), thus indicating the relevance of mixed water masses along the
MIZ for benthic production. The standard deviation
of sea ice concentration shows the highest positive
correlation with P southwest of Franz Josef Land
(Fig. 5c), the region where the highest P values were
observed in this study (Fig. 3a). This may point towards the importance of sea-ice algae for benthic
secondary production in this region, as the ice-algal
contribution to overall primary production was reported to be proportionally more important in areas
where sea ice retreats later in the year (Gosselin et al.
1997, Carroll & Carroll 2003). Denisenko (2002)
showed for the Barents Sea, but also for the Kara,
Laptev, East-Siberian, and Chukchi Seas, that areas
of high benthic biomass matched well with areas of
the longest duration of average multiyear ice cover.
Either way, our findings highlight the effectiveness of
GWR in identifying areas where particular relationships between environmental and ecological features
exist that should be studied in more detail.
enced by ice-edge processes and dynamics. Simultaneously, primary and secondary production in the
southern, ice-free regions might decrease as well,
due to increasing thermal stratification (Slagstad et
al. 2015). Accordingly, we expect cascading effects of
these changes on all levels of the Barents Sea ecosystem. These prospects stress the importance of
continuous integrated monitoring programs, such as
the joint Norwegian−Russian Ecosystem Survey, that
can provide sound scientific advice on ecosystem
Acknowledgements. We thank the graduate program POLMAR for funding R.D., and Vidar S. Lien (IMR) for provision
of modeled environmental data. Thanks to the Institute of
Marine Research and PINRO for including benthic experts
in the annual Ecosystem Survey. The crews and the Russian
and Norwegian benthic taxonomists onboard the Russian
and Norwegian research vessels are thanked for the collection, identification, and quantification of the benthic fauna.
We also thank the Norwegian Ministry of Foreign Affairs for
economic support of exchange of Russian and Norwegian
scientists between the research vessels.
We found significantly higher megabenthic secondary production in the northern, seasonal icecovered Barents Sea than in the southern, ice-free
region. We conclude that tighter pelagic−benthic
coupling in the northern area, and regionally varying
environmental conditions, like local upwelling and
tidal mixing, explain the observed pattern. The negative correlation of P to NPP, used here as a proxy for
food input to the benthos, might be explained to a
certain extent by an underestimation of the modeled
NPP. Additionally, the (unknown) influence of advection processes must be taken into account. A further
explanation for the non-correlation between primary
and secondary production might be that this relationship is less distinct in the highly mobile epibenthic
megafauna (Grebmeier et al. 2015b). Hence, we presume that a more holistic approach including all benthic size classes would yield the expected positive
correlation of benthic secondary production with primary production on large spatial scales.
Regarding the anticipated further poleward movement of the MIZ, our findings indicate that benthic
production in the Barents Sea might decrease significantly, owing to the diminishing shelf area influ-
Anisimova NA, Jørgensen LL, Lyubin PA, Manushin IE
(2010) Mapping and monitoring of benthos in the
Barents Sea and Svalbard waters: results from the joint
Russian-Norwegian benthic programme 2006–2008. IMRPINRO Joint Report Series 1-2010
Benke A (2012) Secondary production. Nat Educ Knowl 3:23
Berge J, Daase M, Renaud PE, Ambrose WG and others
(2015) Unexpected levels of biological activity during the
polar night offer new perspectives on a warming Arctic.
Curr Biol 25:2555−2561
Bluhm BA, Piepenburg D, von Juterzenka K (1998) Distribution, standing stock, growth, mortality and production of
Strongylocentrotus pallidus (Echinodermata: Echinoidea) in the northern Barents Sea. Polar Biol 20:325−334
Bluhm BA, Iken K, Mincks Hardy S, Sirenko BI, Holladay
BA (2009) Community structure of epibenthic megafauna in the Chukchi Sea. Aquat Biol 7:269−293
Brey T (2012) A multi-parameter artificial neural network
model to estimate macrobenthic invertebrate productivity and production. Limnol Oceanogr Methods 10:
Callaway R, Engelhard GH, Dann J, Cotter J, Rumohr H
(2007) A century of North Sea epibenthos and trawling:
comparison between 1902−1912, 1982−1985 and 2000.
Mar Ecol Prog Ser 346:27−43
Carroll ML, Carroll J (2003) The Arctic Seas. In: Black K,
Shimmield G (eds) Biogeochemistry of marine systems.
Blackwell Publishing, Oxford, p 127−156
Carroll ML, Denisenko SG, Renaud PE, Ambrose WG Jr
(2008) Benthic infauna of the seasonally ice-covered
western Barents Sea: patterns and relationships to environmental forcing. Deep-Sea Res II 55:2340−2351
Clough LM, Renaud PE, Ambrose WG (2005) Impacts of
water depth, sediment pigment concentration, and benthic macrofaunal biomass on sediment oxygen demand
Degen et al: Arctic sea ice and benthic production
in the western Arctic Ocean. Can J Fish Aquat Sci 62:
Cochrane SKJ, Denisenko SG, Renaud PE, Emblow CS,
Ambrose WG Jr, Ellingsen IH, Skarðhamar J (2009) Benthic macrofauna and productivity regimes in the Barents
Sea — ecological implications in a changing Arctic. J Sea
Res 61:222−233
Cochrane SKJ, Pearson TH, Greenacre M, Costelloe J, Ellingsen IH, Dahle S, Gulliksen B (2012) Benthic fauna
and functional traits along a Polar Front transect in the
Barents Sea — Advancing tools for ecosystem-scale
assessments. J Mar Syst 94:204–217
Conlan KE, Lenihan HS, Kvitek RG, Oliver JS (1998) Ice
scour disturbance to benthic communities in the Canadian High Arctic. Mar Ecol Prog Ser 166:1−16
Dalpadado P, Arrigo KR, Hjøllo SS, Rey F and others (2014)
Productivity in the Barents Sea — response to recent
climate variability. PLoS One 9:e95273
Darnis G, Barber DG, Fortier L (2008) Sea ice and the
onshore-offshore gradient in pre-winter zooplankton
assemblages in southeastern Beaufort Sea. J Mar Syst 74:
Degen R, Vedenin A, Gusky M, Boetius A, Brey T (2015) Patterns and trends of macrobenthic abundance, biomass
and production in the deep Arctic Ocean. Polar Res
Denisenko SG (2001) Long-term changes of zoobenthos biomass in the Barents Sea. Proc Zool Inst Russ Acad Sci
Denisenko S (2002) Zoobenthos and ice distribution in the
Arctic seas. Proc Zool Inst Russ Acad Sci 296:39−46
Dommasnes A, Christensen V, Ellertsen B, Kvamme C and
others (2001) An Ecopath model for the Norwegian Sea
and Barents Sea. 2001: fisheries impacts on North
Atlantic ecosystems. Fish Cen Res Rep 9:213−239
Eiane K, Tande KS (2009) Meso and macrozooplankton. In:
Sakshaug E, Johnsen G, Kovacs K (eds) Ecosystem Barents Sea. Tapir Academic Press, Trondheim, p 209−234
Eleftheriou A, Moore DC (2005) Macrofauna techniques. In:
Eleftheriou A (ed) Methods for the study of marine benthos. Blackwell Science, Oxford, p 160−228
Fossheim M, Primicerio R, Johannesen E, Ingvaldsen Randi
B, Aschan MM, Dolgov AV (2015) Recent warming leads
to a rapid borealization of fish communities in the Arctic.
Nat Clim Change 5:673−678
Fotheringham AS, Brunsdon C, Charlton ME (2002) Geographically weighted regression: the analysis of spatially
varying relationships. Wiley, Chichester
Gosselin M, Levasseur M, Wheeler PA, Horner RA, Booth
BC (1997) New measurements of phytoplankton and ice
algal production in the Arctic Ocean. Deep-Sea Res II
Gradinger R (2009) Sea ice algae: major contributors to primary production and algal biomass in the Chukchi and
Beaufort Seas during May/June 2002. Deep-Sea Res II
Grebmeier JM, McRoy CP, Feder HM (1988) Pelagic-benthic coupling on the shelf of the northern Bering and
Chukchi Seas. I. Food supply source and benthic biomass. Mar Ecol Prog Ser 48:57−67
Grebmeier JM, Cooper LW, Feder HM, Sirenko BI (2006)
Ecosystem dynamics of the Pacific-influenced Northern
Bering and Chukchi Seas in the Amerasian Arctic. Prog
Oceanogr 71:331−361
Grebmeier JM, Bluhm BA, Cooper LW, Danielson S and
others (2015a) Ecosystem characteristics and processes
facilitating persistent macrobenthic biomass hotspots
and associated benthivory in the Pacific Arctic. Prog
Oceanogr 136:92−114
Grebmeier JM, Bluhm BA, Cooper LW, Denisenko SG, Iken
K, Ke
˛ dra M, Serratos C (2015b) Time-series benthic
community composition and biomass and associated
environmental characteristics in the Chukchi Sea during
the RUSALCA 2004−2012 Program. Oceanography 28:
Highsmith RC, Coyle KO (1990) High productivity of
northern Bering Sea benthic amphipods. Nature 344:
Ingvaldsen R, Loeng H (2009) Physical oceanography. In:
Sakshaug E, Johnsen G, Kovacs K (eds) Ecosystem
Barents Sea. Tapir Academic Press, Trondheim, p 33−64
Jakobsson M, Grantz A, Kristoffersen Y, Macnab R (2004)
Bathymetry and physiogeography of the Arctic Ocean
and its constituent seas. In: Stein R, Macdonald RW (eds)
The Arctic Ocean organic carbon cycle: present and past.
Springer, Berlin, p 1−6
Jørgensen LL, Renaud PE, Cochrane SKJ (2011) Improving
benthic monitoring by combining trawl and grab surveys. Mar Pollut Bull 62:1183−1190
Jørgensen LL, Ljubin P, Skjoldal HR, Ingvaldsen RB, Anisimova N, Manushin I (2015) Distribution of benthic megafauna in the Barents Sea: baseline for an ecosystem
approach to management. ICES J Mar Sci 72:595−613
˛ dra M, Renaud PE, Andrade H, Goszczko I, Ambrose WG
Jr (2013) Benthic community structure, diversity, and
productivity in the shallow Barents Sea bank (Svalbard
Bank). Mar Biol 160:805−819
Leu E, Mundy CJ, Assmy P, Campbell K and others (2015)
Arctic spring awakening − steering principles behind the
phenology of vernal ice algal blooms. Prog Oceanogr
Lien VS, Gusdal Y, Vikebø FB (2014) Along-shelf hydrographic anomalies in the Nordic Seas (1960−2011): locally
generated or advective signals? Ocean Dyn 64:1047−1059
Loeng H, Ozhigin V, Aadlandsvik B (1997) Water fluxes
through the Barents Sea. ICES J Mar Sci 54:310−317
Lyubin P, Anisimova NA, Manushin IE (2011) Long term
effects on benthos of the use of bottom fishing gears. In:
Jakobsen T, Ozhigin VK (eds) The Barents Sea. Ecosystem, resources, management. Half a century of RussianNorwegian cooperation. Tapir Academic Press, Trondheim, p 768−775
Matrai P, Apollonio S (2013) New estimates of microalgae
production based upon nitrate reductions under sea ice
in Canadian shelf seas and the Canada Basin of the
Arctic Ocean. Mar Biol 160:1297−1309
McMahon KW, Ambrose WG Jr, Johnson BJ, Sun MY,
Lopez GR, Clough LM, Carroll ML (2006) Benthic community response to ice algae and phytoplankton in Ny
Ålesund, Svalbard. Mar Ecol Prog Ser 310:1−14
Melle W, Skjoldal HR (1998) Reproduction and development
of Calanus finmarchicus, C. glacialis and C. hyperboreus
in the Barents Sea. Mar Ecol Prog Ser 169:211−228
Michalsen K, Dalpadado P, Eriksen E, Gjøsæter H and
others (2013) Marine living resources of the Barents
Sea — ecosystem understanding and monitoring in a climate change perspective. Mar Biol Res 9:932−947
Mueter FJ, Litzow MA (2008) Sea ice retreat alters the biogeography of the Bering Sea continental shelf each winter. Ecol Appl 18:309−320
Mar Ecol Prog Ser 546: 1–16, 2016
➤ Mundy CJ, Gosselin M, Ehn J, Gratton Y and others (2009)
Contribution of under-ice primary production to an iceedge upwelling phytoplankton bloom in the Canadian
Beaufort Sea. Geophys Res Lett 36, L17601, doi:10.1029/
Nilsen M, Pedersen T, Nilssen EM (2006) Macrobenthic
biomass, productivity (P/B) and production in a highlatitude ecosystem, North Norway. Mar Ecol Prog Ser
Olli K, Wexels Riser C, Wassmann P, Ratkova T, Arashkevich E, Pasternak A (2002) Seasonal variation in vertical
flux of biogenic matter in the marginal ice zone and the
central Barents Sea. J Mar Syst 38:189−204
Orlova EL, Dalpadado P, Knutsen T, Nesterova VN, Prokopchuk IP (2011) Zooplankton. In: Jakobsen T, Ozhigin VK
(eds) The Barents Sea. Ecosystem, resources, management. Half a century of Russian-Norwegian cooperation.
Tapir Academic Press, Trondheim, p 91−119
Ozhigin VK, Ingvaldsen RB, Loeng H, Boitsov V, Karsakov A
(2011) Introduction to the Barents Sea. In: Jakobsen T,
Ozhigin VK (eds) The Barents Sea. Ecosystem, resources,
management. Half a century of Russian-Norwegian cooperation. Tapir Academic Press, Trondheim, p 768−775
Pearson TH, Rosenberg R (1978) Macrobenthic succession
in relation to organic enrichment and pollution of the
marine environment. Oceanogr Mar Biol Annu Rev 16:
Piepenburg D (2005) Recent research on Arctic benthos:
common notions to be revised. Polar Biol 28:733−755
Piepenburg D, Blackburn TH, von Dorrien CF, Gutt J and
others (1995) Partitioning of benthic community respiration in the Arctic (northwestern Barents Sea). Mar Ecol
Prog Ser 118:199−213
Reigstad M, Wexels Riser C, Wassmann P, Ratkova T (2008)
Vertical export of particulate organic carbon: attenuation, composition and loss rates in the northern Barents
Sea. Deep-Sea Res II 55:2308−2319
Renaud PE, Riedel A, Michel C, Morata N, Gosselin M,
Chiuchiolo A (2007) Seasonal variation in benthic
community oxygen demand: a response to an ice algal
bloom in the Beaufort Sea, Canadian Arctic? J Mar Syst
Renaud PE, Løkken TS, Jørgensen LL, Berge J, Johnson BJ
(2015) Macroalgal detritus and food-web subsidies along
an Arctic fjord depth-gradient. Front Mar Sci 2:31
Rosenberg R (1995) Benthic marine fauna structured by
hydrodynamic processes and food availability. Neth J
Sea Res 34:303−317
Roy V, Iken K, Archambault P (2014) Environmental drivers
of the Canadian Arctic megabenthic communities. PLoS
One 9:e100900
Editorial responsibility: James McClintock,
Birmingham, Alabama, USA
Sakshaug E, Johnsen G, Kristiansen S, von Quillfeldt C, Rey
F, Slagstad D, Thingstad F (2009) Phytoplankton and primary production. In: Sakshaug E, Johnsen G, Kovacs K
(eds) Ecosystem Barents Sea. Tapir Academic Press,
Trondheim, p 167−208
Slagstad D, McClimans TA (2005) Modeling the ecosystem
dynamics of the Barents sea including the marginal ice
zone: I. Physical and chemical oceanography. J Mar Syst
Slagstad D, Tande KS, Wassmann P (1999) Modelled carbon
fluxes as validated by field data on the north Norwegian
shelf during the productive period in 1994. Sarsia 84:
Slagstad D, Wassmann PFJ, Ellingsen I (2015) Physical constraints and productivity in the future Arctic Ocean. Front
Mar Sci 2:85
Sun M, Carroll ML, Ambrose WG Jr, Clough LM, Zou L,
Lopez GR (2007) Rapid consumption of phytoplankton
and ice algae by Arctic soft-sediment benthic communities: evidence using natural and 13C-labeled food materials. J Mar Res 65:561−588
Sundfjord A, Ellingsen I, Slagstad D, Svendsen H (2008)
Vertical mixing in the marginal ice zone of the northern
Barents Sea—results from numerical model experiments.
Deep-Sea Res II 55:2154−2168
Tamelander T, Renaud PE, Hop H, Carroll ML, Ambrose
WG Jr, Hobson KA (2006) Trophic relationships and
pelagic−benthic coupling during summer in the Barents
Sea Marginal Ice Zone, revealed by stable carbon and
nitrogen isotope measurements. Mar Ecol Prog Ser
Vinogradova PS, Litvin VM (1960) Study of the topography
and sediments in the Barents and Norwegian seas. In:
Marty Y (ed) Soviet fisheries investigations in north
European Seas. PINRO-VNIRO Press, Moscow, p 101−110
(in Russian)
Wassmann P, Slagstad D, Wexels Riser C, Reigstad M
(2006a) Modelling the ecosystem dynamics of the
Barents Sea including the marginal ice zone II.
Carbon flux and interannual variability. J Mar Syst
Wassmann P, Reigstad M, Haug T, Rudels B and others
(2006b) Food webs and carbon flux in the Barents Sea.
Prog Oceanogr 71:232−287
Wiedmann I, Reigstad M, Sundfjord A, Basedow S (2014)
Potential drivers of sinking particle’s size spectra and
vertical flux of particulate organic carbon (POC): turbulence, phytoplankton, and zooplankton. J Geophys Res
Zenkevitch L (1963) Biology of the seas of the USSR. George
Allen and Unwin, London
Submitted: August 28, 2015; Accepted: February 10, 2016
Proofs received from author(s): March 9, 2016
Download PDF