ArticlePDF Available

Migration dynamics of Pacific herring (Clupea pallasii) and response to spring environmental variability in the southeastern Bering Sea

  • Backwater Research

Abstract and Figures

In the southeastern Bering Sea, Pacific herring (Clupea pallasii) migrate from the Pribilof Islands region where they overwinter, to the Alaska coast where they spawn in spring. The migration sustains a nearshore commercial fishery that targets roe-bearing females just prior to spawning. Herring also are taken as bycatch in groundfish trawl fisheries, where time and area closures in these fisheries are triggered by herring bycatch caps. Using herring bycatch data collected since the 1970s by National Marine Fisheries Service (NMFS) observers aboard groundfish fishing vessels, a retrospective analysis was conducted to describe the seasonal migration pattern of Pacific herring in the southeastern Bering Sea and to study its spatial and temporal variability. Observed changes in herring catch per unit of effort were compared with variability in climate and oceanographic conditions. The seasonal migration is complex, but annual shifts in migration routes and a possible northward shift of the overwintering grounds was identified. Pre-spawning herring aggregated in different areas depending on whether spawning occurred early or late in spring. The thermal structure of the ocean around the ice edge appears to influence herring migration timing and route as well as spawning date. Thus, on the basis of recent changes in sea-ice extent and duration, we suggest that the herring bycatch savings area that was developed from data collected in the 1980s should be revised to reflect prevailing conditions.
Content may be subject to copyright.
This article was published in an Elsevier journal. The attached copy
is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy
Deep-Sea Research II 54 (2007) 2832–2848
Migration dynamics of Pacific herring (Clupea pallasii)
and response to spring environmental variability
in the southeastern Bering Sea
Naoki Tojo
, Gordon H. Kruse
, Fritz C. Funk
Juneau Center, School of Fisheries and Ocean Sciences, University of Alaska Fairbanks, 11120 Glacier Highway, Juneau, AK 99801, USA
Alaska Department of Fish and Game, Division of Commercial Fisheries, P.O. Box 25526, Juneau, AK 99802-5526, USA
Received in revised form 7 March 2007; accepted 31 July 2007
Available online 19 November 2007
In the southeastern Bering Sea, Pacific herring (Clupea pallasii) migrate from the Pribilof Islands region where they
overwinter, to the Alaska coast where they spawn in spring. The migration sustains a nearshore commercial fishery that
targets roe-bearing females just prior to spawning. Herring also are taken as bycatch in groundfish trawl fisheries, where
time and area closures in these fisheries are triggered by herring bycatch caps. Using herring bycatch data collected since
the 1970s by National Marine Fisheries Service (NMFS) observers aboard groundfish fishing vessels, a retrospective
analysis was conducted to describe the seasonal migration pattern of Pacific herring in the southeastern Bering Sea and to
study its spatial and temporal variability. Observed changes in herring catch per unit of effort were compared with
variability in climate and oceanographic conditions. The seasonal migration is complex, but annual shifts in migration
routes and a possible northward shift of the overwintering grounds was identified. Pre-spawning herring aggregated in
different areas depending on whether spawning occurred early or late in spring. The thermal structure of the ocean around
the ice edge appears to influence herring migration timing and route as well as spawning date. Thus, on the basis of recent
changes in sea-ice extent and duration, we suggest that the herring bycatch savings area that was developed from data
collected in the 1980s should be revised to reflect prevailing conditions.
r2007 Elsevier Ltd. All rights reserved.
Keywords: Pacific herring; Bering Sea; Spawning migration; Sea ice; Thermal structure
1. Introduction
Pacific herring (Clupea pallasii) are widely dis-
tributed in the temperate waters of the North Pacific
from the California coast to the Aleutian Islands
and from the Bering Sea across the Pacific Ocean to
the Yellow Sea (Hay, 1985). They are closely
associated with the continental shelf and coastal
environment, although they migrate through a
variety of habitats depending on population and
life stage (Hay and McCarter, 1997). In general,
Pacific herring spawn earliest at southern latitudes
and later at higher latitudes, in association with the
northward progression of the spring bloom. Herring
spawn in shallow subtidal or intertidal areas along
coastlines each spring and then move offshore to
0967-0645/$ - see front matter r2007 Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +1 907 796 2052;
fax: +1 907 796 6447.
E-mail address: (G.H. Kruse).
Author's personal copy
feed (Mecklenburg et al., 2002). In fall, they move to
deeper overwintering grounds where the diel vertical
migration patterns associated with feeding become
less distinct (Hay, 1985). Northern Bristol Bay
herring, the largest herring spawning population in
the eastern Bering Sea (EBS), migrate from an area
northwest of the Pribilof Islands to Bristol Bay to
an area north of Unimak Pass and back, a total
distance of approximately 2100 km (Fig. 1).
The first studies of EBS herring migration were
based upon data collected from foreign fisheries
operating off the Alaskan coast. The demand for
Alaskan herring increased following the collapse of
the Japanese herring fisheries in the 1950s. Intensive
trawl and gillnet fisheries, conducted mainly by
Japanese and Russian fishermen, followed the
seasonal migration of EBS herring. These foreign
fisheries were phased out of the US Exclusive
Economic Zone (EEZ) following passage of the
Magnuson Fisheries Conservation and Manage-
ment Act in 1976.
A clockwise migration pattern of Bristol Bay
herring (Fig. 1) was indicated by many studies of
fishery data collected before 1976 (Dudnik and
Usol’tsev, 1964;Prokhorov, 1970;Rumyantsev
and Darda, 1970;Wespestad, 1978;Barton and
Fig. 1. Locations of spawning stocks of Pacific herring in the eastern Bering Sea and historically postulated migration pattern with
wintering ground (modified from Barton and Wespestad, 1980). The gray lines are the 50-, 100-, and 200-m isobaths. Thick rectangles
show historical spawning locations. The two largest rectangles represent the (largest) northern Bristol Bay spawning stock. The gridded
circle at 581N represents the major wintering ground in the southeastern Bering Sea (Dudnik and Usol’tsev, 1964;Prokhorov, 1970,
Rumyantsev and Darda, 1970;Funk, 1990;Wespestad, 1978;Wespestad and Gunderson, 1991;Wespestad and Barton, 1981). The gray
boxed arrows show pre-spawning movements, and the white boxed arrows show post-spawning movements. The blank circle with question
mark is the speculated minor wintering ground (Barton and Wespestad, 1980).
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2833
Author's personal copy
Wespestad, 1980;Funk, 1990). Herring were found
to overwinter mainly off-bottom to the northwest of
St. Paul Island and to disperse over the shelf
beginning in early spring. From April to May, they
migrated to the coast where they spawned from
May to June, although in recent years spawning has
begun occasionally in late April. After spawning,
herring remained in the spawning area or migrated
southward along the Alaska Peninsula, where most
of the herring accumulated near Unimak Pass to
feed before moving north across broad feeding areas
along the continental shelf and slope in summer
(Rumyantsev and Darda, 1970). In fall, herring
concentrations gradually moved to locations north-
west of the Pribilof Islands to overwinter.
The objective of this study was to re-examine the
seasonal migration of Pacific herring, including its
spatial and temporal variability, using more ex-
tensive information than was available from the
foreign fisheries. Specific objectives were to: (1)
determine herring migration pathways using herring
bycatch data collected from the foreign and
domestic groundfish fisheries, (2) analyze the gen-
eral pre-spawning migration relative to oceano-
graphic conditions in spring, and (3) explain
variability of herring migration patterns relative to
variability in EBS oceanography. This study was
motivated, in part, by the possibility that herring
migration patterns may have changed in association
with climate regime shifts in 1976–1977 and 1989.
The latter shift started a period of rapid warming in
the southeastern Bering Sea, accompanied by
declines in both the extent and concentration of
sea ice cover (Overland et al., 2004). A shift in
herring migration route might be anticipated
because a congener, the Atlantic herring (Clupea
harengus), has changed its wintering and feeding
grounds in the North Sea several times since 1950
(Corten, 2002). Episodic environmental shifts have
been implicated as the cause of changes in northern
Norway herring migrations (Slotte, 1999a, b).
2. Methods
2.1. Data collection and processing
2.1.1. Herring data
Records of herring bycatch in groundfish trawl
fisheries from 1977 to 2003 were obtained from the
Alaska Fisheries Science Center, National Marine
Fisheries Service (NMFS) observer database. Her-
ring bycatch in these fisheries was recorded as
metric tons. Fishing effort associated with this
bycatch was measured as tow duration (minutes),
unique tow number, and number of records. A
‘‘record’’ is the number of fishing trips each day in
same area, and each record can include multiple
tows. Fishing locations were identified by the
location where the gear was retrieved, recorded to
the nearest minute of latitude and longitude.
ArcGIS 9.0 (ESRI) was used to stratify retrieval
locations by 50 50 km
cells. To comply with
confidentiality requirements, cells with fewer than
three tows were excluded from the analyses.
Catch per unit of effort (CPUE) was calculated
by dividing the biomass of herring taken as bycatch
in a haul by the duration of the tow. As hauls
lacking tow duration accounted for 65% of the
records and many of these hauls were made
nearshore, exclusion of these records would have
introduced spatial bias in geographic coverage. For
cases when tow duration was missing, it was
estimated from a linear model that predicted tow
duration from frequency of tows by month in each
50 50 km
cell. There was a statistically significant
relationship between total tow duration and total
numbers of tows for each available fishery type for
both joint-venture and non-joint venture fisheries
before 1990 (Po0.001, r
40.65). If the number of
tows was not available, then duration was estimated
from a linear model based on the number of fishing
records; tow duration and numbers of fishing
records were statistically significantly correlated
(Po0.001, r
40.42). In the case of mothership
fisheries, fishing records may represent deliveries
from multiple tows. Data were excluded if neither
proxy was available. During preliminary analysis, it
was found that tow duration in the 1990s had large
variance, so hauls with missing durations from the
1990s were omitted from the analysis.
To describe monthly spatial distributions of
herring, an index of their relative abundance was
developed using CPUE of herring bycatch in each
fishing operation. Because herring CPUE varies as a
function of several factors, in addition to their
abundance in a grid cell, the CPUE data were
stratified and standardized to accommodate the
influence of three factors: (1) type of fishery (e.g.,
mothership, small trawler, or large trawler opera-
tions for joint venture fisheries and foreign fisheries
during 1977–1990 and non-pelagic, pelagic, mixed,
pair, and shrimp trawls for domestic fisheries during
1990–2003), (2) general level of herring biomass in
the EBS (three annual levels created using 33rd and
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482834
Author's personal copy
67th percentiles of annual spawning biomass
estimates from the ADF&G Togiak herring man-
agement area over 1977–2003 so that years of high
herring abundance did not dominate the interpreta-
tion of herring distributional patterns), and (3)
month (to adjust for seasonal differences). To do
this, hauls with no herring bycatch were omitted
and herring CPUE values in the remaining hauls
were natural-log transformed (lnCPUE). Next, for
each grid cell in a given year and month, each
lnCPUE observation was standardized by subtract-
ing the mean within each fishery type, biomass level,
and month, dividing by the corresponding standard
deviation, and adding 3.0 (m73sencompasses all
measurements) to convert all standard normal
deviates to positive values
Zg;b;f;y;m;i¼ln CPUEg;b;f;y;m;i¯
þ3, (1)
where Z
is the standardized log-transformed
deviate for each CPUE iobserved in grid cell gin
month mof year yfor herring biomass level bin
fishery f. A CPUE observation is typically an
individual haul or trawl tow except for foreign
motherships where fishing records (deliveries) some-
times represent multiple tows. Eq. (1) simply scales
the CPUE deviates in each grid cell to have a
common mean and variance to allow comparison
across different herring population levels and fish-
ery types. For a given fishery type, a simple average
of these standardized CPUE deviates for each grid
cell, biomass level, year and month is ¯
combine these fishery-specific values into an overall
mean value (representing all fisheries combined) for
each cell in each month and each year, we calculated
weighted averages as
N, (2)
where n
is the number of observations in
each grid cell for each biomass level and fishery type
by year and month, and Nis the overall total
number of observations, including hauls with no
bycatch. Weighted averages at this level Eq. (2)
allowed analysis of herring relative abundance for
each month and year, whereas a simple average of
these weighted averages ( ¯
Zg;m) across all years
(1977–2003) allowed analysis of average monthly
herring distributions independent of year.
To summarize herring distributions from the
CPUE data, kernel density contours were estimated
in ArcGIS Spatial Analyst. The kernel (f
estimated by
nh X
, (3)
where his the smoothing parameter or bandwidth,
is the ith observation, xis the exact location over
the study area, nis sample size, and Kis the kernel
function giving the weights to the values for
estimation (Silverman, 1986). Kdetermines the
basic shape of the smoothed lines and hdetermines
the degree of horizontal generalization and is set as
the ‘‘searching radius’’ (Silverman, 1986). We set the
kernel function to the default value used by ArcGIS
9.0 (ESRI), which approximates the Gaussian
kernel. The searching radius was defined as
142,000 m (the diagonal of two 50-km cell widths),
which allowed us to generalize the CPUE values
using at least two neighborhood points even in the
isolated corners of the grid.
To visualize both the general movement and the
variability of herring distribution and migration,
two types of plots were generated from the kernel
density surface: one from the abundance index in
each calendar month combined over years (biomass
concentration plot), and another from the abun-
dance index in individual months from individual
years (peak plots). In the biomass concentration
plots, monthly density surfaces revealed the general
seasonal movement regardless of the interannual
variability of herring population parameters. The
peak plots showed the detailed monthly distribu-
tions of CPUE in each year.
For the biomass concentration plots, contours of
constant density were fitted to the kernel density
surfaces. Using these contours, the spatial arrange-
ments of the monthly upper quartiles were assumed
to represent the general movement of herring and its
seasonal variability. For the peak plots, any
individual month with o30 grid cells with CPUE
data was discarded. Then, a kernel density surface
was fit to the remaining months, by year. From the
fitted density contours, we calculated the upper
quartiles to display areas of peak herring abundance
in each month in each year. The locations of the
centroids of these peaks was determined for every
year and month. The percentiles of the kernel
density surfaces were classified with ArcMap
based upon the estimated density index (f
). A
Visual Basic in Application (VBA) function in
ArcGIS was used to calculate the centroids. An
analysis of the dispersion of these centroids in each
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2835
Author's personal copy
month and year was used to reveal interannual
variability of the geographic location of herring and
identified the months of greatest variability in
herring location.
2.1.2. Environmental data
Two environmental variables were hypothesized
to affect herring pre-spawning (April–May) migra-
tion dynamics: sea-surface temperature (SST) and
sea-ice total concentration (SITC). The SST data
were collected by vessels of opportunity and
reported by latitude and longitude in the Compre-
hensive Ocean-Atmosphere Dataset (COADS) from
the National Climate Data Center (NCDC). The
SITC data were taken from the Environmental
Working Group Joint US-Russian Sea Ice Atlas
(Arctic Climatology Project) from the National
Snow and Ice Data Center (NSIDC) and archived
sea ice charts by National Ice Center (NIC).
The spatial overlap between the COADS SST
data and the herring distribution data was not great,
so the SST data were smoothed by kriging with a
geostatistical analyst (ESRI), one of the ArcGIS
extensions. Distance and directional autocorrela-
tions were used in deriving the estimated SST field,
and extrapolations were cross-validated to obtain
the most representative estimates (Tojo, 2006). For
this analysis, grid cells were defined as 1 km
The SITC dataset with 12.5 km
grid cells was
produced in raster format until 1994. After 1994 the
data were recorded in vector format with smooth
lines to partition the polygons. Vector format data
were converted to raster format for consistency.
Missing or corrupted GIS ice charts in the NIC data
archive were repaired or replaced by digitizing the
original NIC ice charts. Because the conversion
algorithm produces misleading wiggly lines con-
necting the centers of the individual cells, an inverse
distance weighting (IDW) of the extrapolated sur-
face was used. The IDW method simply weights the
values as an inverse linear function of the distances.
2.2. Analyses
2.2.1. General migration pattern
To visualize the general migration pattern of EBS
herring, the centers of statistical quartiles of herring
distributions from CPUE density surfaces were
plotted by month for years with sufficient data for
analysis. For this purpose, the same equal-area
projection was applied as was used for the data
processing. By plotting the upper quartile of the
density surfaces, we aimed to clarify the areas where
herring were concentrated.
2.2.2. Variability of migration patterns
The dispersion of annual peaks of herring
distribution was calculated by the average nearest-
neighbor distance among the centroids of those
peaks. The larger the average nearest-neighbor
distance indicates more dispersion among these
centroids. The month with the greatest dispersion
was identified as the month with the most inter-
annual variability of migration.
The pre-spawning migration season (April and
May) was known from previous studies (Barton and
Wespestad, 1980;Hay, 1985) and from our ob-
servations of spawning sites in 2003 and 2004.
Therefore, herring distributions were compared
with environmental conditions by overlaying the
GIS layers for herring (centroids of CPUE peaks)
with the corresponding SST and SITC interpolated
surfaces. Monthly SITC and SST surfaces were
compared with the timing of herring arrival at the
northern shore of Bristol Bay, determined from
historical aerial surveys conducted by the Alaska
Department Fish and Game (ADF&G). We binned
years as early (1982, 1985, 1986, 1988, 1990, 1992,
1994) or late (1981, 1983, 1987, 1993, 1998, 2000,
2001, 2003) based on the upper and lower quartiles
of arrival date on the spawning grounds (Table 1).
3. Results
3.1. General migration pattern during 1977– 2003
The spatial distribution of fishing effort was
sufficiently broad (Fig. 2) to allow the use of fishery
bycatch CPUE data to examine the seasonal
migrations of herring. Overall, from 1977 to 2003,
fishing effort was distributed throughout the EBS,
extending into the Aleutian Basin. Fishing effort
was most heavily concentrated in the middle
domain (50–100 m), especially in spring and sum-
mer, and in the outer domain (depth 4100 m) of the
southeastern Bering Sea. Fishing in deep areas such
as the Pribilof Canyon, southwest of the Pribilof
Islands, was associated with the walleye pollock
(Theragra chalcogramma) fishery.
Monthly distributions of herring biomass often
had peaks in two regions, one located west of
St. Matthew Island and the other between
St. George Island (the southernmost Pribilof Island)
and Unimak Pass throughout most of the year
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482836
Author's personal copy
(Fig. 3). This bimodal distribution was least
apparent in June, when herring were concentrated
north of Unimak Pass. The spatial extent of
area in the upper quartiles was minimum in
December (125,042.5 km
) and maximum in Sep-
tember (266,366.8 km
). The average size of the area
encompassing the upper quartile was 190,190.7 km
(95% c.i.159,208.8 to 221,172.6 km
). From No-
vember to March, the areas covered by the upper
quartiles were smaller than the lower confidence
interval of the monthly mean, indicating tighter
concentration of herring from winter to early
Among the pre-spawning months in spring, the
greatest variability in the location of annual peaks
was in April when the centroid peaks began to
expand shoreward over the southeastern shelf
(Fig. 3). Among months with adequate observations
(n430), the high ratio of the observed average
monthly nearest-neighbor distances to the expected
average nearest-neighbor distances based on a
random distribution indicates greater dispersion in
April compared to other months (Fig. 4). The
z-score of the observed average nearest-neighbor
distance in April was 1.71, which is not signifi-
cantly different from the expected average nearest-
neighbor distance in random distribution (z*71.96,
a¼0.05). In general, the largest concentrations of
herring biomass occurred approximately 250 km
east of the Pribilof Islands. In years of maximum
eastern extent, the distribution in April reached the
vicinity of major coastal spawning areas. In May,
herring were concentrated along the shoreline in
areas such as northern Bristol Bay (58.51N,
160.51W), and between the Pribilof Islands and
western Yukon-Kuskokwim area (591N, 1631W). In
June, the major concentrations of herring extended
along the coast of the Alaska Peninsula to Unimak
Pass (Fig. 3). The concentration was broadly spread
over Bristol Bay. The centroids of annual peaks
developed as small clusters in various locations. The
western ‘‘tail’’ of the herring concentrations shifted
eastward by about 125 km compared to May. In
July, herring concentrations showed a broad
O-shaped pattern on the southeastern Bering Sea
shelf (Fig. 3), and monthly peaks became more
prominent at Unimak Pass. A minor concentration
appeared south and offshore of St. Matthew Island
along the slope. This bimodal feature intensified
through fall and early winter. The northern mode
was located west of St. Matthew Island (591N,
1771W) and the southern mode extended northwest
of Unimak Pass.
In November and December, the herring dis-
tribution was mostly limited to an area of the
northern EBS centered near 59.51N, 1791W
although annual peaks occurred in the southeast
in some years. Herring appeared to be sparse along
the 65 m isobath but monthly fishing effort data
were insufficient to provide any certainty for these
detailed features. The overwintering concentration
west of St. Mathew Island in November and
December (59.51N1791W) was located more to
the northwest after 1976 than in earlier years
(Wespestad, 1978; Appendices in Tojo, 2006).
3.2. Interannual variability in herring migration:
comparisons to variability in ocean conditions
As the greatest interannual variability in the
location of herring distribution peaks occurred in
April, it may be a key month for determining
Table 1
Arrival timings of herring to northern Bristol Bay spawning area,
sorted by statistical percentile
Julian day Year Percentile Bin
111 2003 0 Early
112 1981 4 Early
114 1987 8 Early
114 1993 8 Early
115 2000 16 Early
116 1983 20 Early
116 1998 20 Early
116 2001 20 Early yearly 25th percentile
117 1997 32 Normal
119 1980 36 Normal
119 1996 36 Normal
120 1979 44 Normal
121 1978 48 Normal
122 1995 52 Normal
122 2002 52 Normal
123 1984 60 Normal
124 1990 64 Normal
125 1989 68 Normal
126 1991 72 Normal ylate 25th percentile
127 1986 76 Late
128 1994 80 Late
129 1988 84 Late
132 1982 88 Late
135 1999 92 Late
137 1992 96 Late
139 1985 100 Late
Upper and lower 25th percentiles (thick underlines) were used for
the borders to define both early and late bin years in analyses.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2837
Author's personal copy
whether herring migrate early or late in the season,
perhaps owing to environmental conditions that
affect gonadal maturation. The ice edge in April was
150 km farther north in early versus late arrival
years, especially in the coastal areas of northern
Bristol Bay (Fig. 5A). For both arrival years,
Fig. 2. Historical and seasonal distributions of fishing effort from 1977 to 2003. Total duration of all tows is used as an index of fishing
effort for all years combined. Gray lines show the 50-, 100-, and 200-m isobaths.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482838
Author's personal copy
centroids of peaks occurred in areas of less ice from
sea-ice edge to the open sea; 17 of 18 centroids in
early years and 11 of 13 in late years in April were
present in the area where SITC was less than 25%
(Fig. 5A).
In April the ratio of observed to expected average
nearest-neighbor distances among peak centroids
was 0.83 in early arrival years and 0.65 in late
arrival years. The average nearest-neighbor distance
among peak centroids in late arrival years
(z-score ¼2.14, a¼0.05) indicated statistically
significant clustering of centroids, indicating a
clumped distribution of centroids and less detect-
able variability in migration patterns in late arrival
years. On the other hand, the z-score (1.45) in
early arrival years fell within the 95% confidence
interval, indicating that the among-centroid dis-
tances in early arrival years could not be distin-
guished from random dispersions in these years.
Coastal waters, extending northward from the
northeast corner of Bristol Bay, were enclosed by
the 455% isoclines of SITC in late arrival years.
The SST observations generally revealed a larger
expanse of cold water on the southeastern con-
tinental shelf in the late arrival years than in the
early arrival years. On the other hand, there was no
obvious difference in ice distribution/concentration
or SST along the continental shelf edge, Aleutian
Basin area, or northern portion of the EBS among
early and late arrival years.
Where they occurred, differences in sea-ice
distribution among early and late arrival years
persisted into May (Fig. 5B). In early arrival years,
sea ice receded more to the north and the coastal
area was mostly ice-free in the spawning areas
except in Norton Sound. In May in late arrival
years, the coastal zone remained covered with
sparse sea ice, and a distinctive feature was the
Fig. 3. General migration pattern of herring in the eastern Bering Sea from 1977 to 2003. Gray shaded areas represent the upper 25th-
percentile monthly contours of catch-per-unit-effort (biomass concentration plot: gray shaded) averaged across all years and black dots
represent the annual peak locations within yearly centroids of herring concentrations by month. Cross marks indicate fishing efforts
yielding no reported herring catches.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2839
Author's personal copy
southward tongue-like extension of 410% sea ice
isoclines (and cold SST) between Cape Newenham
and the Pribilof Islands. In early arrival years,
herring tended to occur in open water, whereas
herring seemed to be more associated with the ice
edge in late arrival years (Fig. 5B).
4. Discussion
4.1. General migration pattern of EBS herring
We confirmed the general inshore–offshore sea-
sonal migration pattern of Bristol Bay herring
found in earlier studies, but there were some
important exceptions. Previous studies (e.g., Pro-
khorov, 1970;Rumyantsev and Darda, 1970;
Barton and Wespestad, 1980;Funk, 1990) con-
cluded that herring migrated in clockwise pattern in
the EBS (Fig. 1), overwintering mainly off-bottom
to the northwest of St. Paul Island. In contrast, we
identified two concentrations in fall/winter. One was
located farther to the northwest of the Pribilof
Islands than had been reported previously, and the
other was a new location north of Unimak Pass,
extending to the northwest in the direction of the
Pribilofs (Fig. 6). There was some speculation about
the existence of a southern overwintering ground
(Barton and Wespestad, 1980), but it was not clearly
indicated in the pre-1976 analyses (Wespestad,
1978). Our concept of seasonal migration places
herring in both southern and northern overwinter-
ing areas, spreading eastward over the middle shelf
in spring and heading toward the coast (Fig. 6).
Spawning generally occurred from May to June, as
indicated in previous studies, but it has begun to
occur in late April in recent years. As previously
reported, herring were sparse around the spawning
area after spawning as at that time they were
migrating southward along the Alaska Peninsula.
The majority of the herring spawners, mainly from
northern Bristol Bay, accumulated near Unimak
Pass to feed. Some of the other northern stocks
Fig. 3. (Continued)
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482840
Author's personal copy
migrated across the shelf, as well. Whereas previous
studies have indicated that herring move northward
along the continental shelf and slope in summer and
fall (Rumyantsev and Darda, 1970), we found that
some herring followed this pattern, whereas others
remained behind to overwinter north of Unimak
Pass. Thus, the migration of Bristol Bay herring
cannot be simply characterized as ‘‘clockwise’’
pattern as was previously proposed. Instead, we
suggest a modified conceptual model of EBS herring
migration with both northern and southern over-
wintering and feeding grounds (Fig. 6).
In Atlantic herring, large-scale migrations have
been related to schooling behavior (Huse et al.,
2002) and spawning timing is related to body length
of pre-spawning individuals (Slotte and Fiksen,
2000;Slotte, 1999a) as well as responses to
oceanographic factors (Messieh, 1987). The persis-
tence of general migration patterns, perhaps learned
by younger age classes from older age classes,
was proposed for Atlantic herring (Corten, 2002).
In British Columbia, spawning waves occur in
Fig. 3. (Continued)
Distance (m)
Observed / Expected
Fig. 4. Monthly change of average nearest neighborhood
distances (average NND: black dots) and observed average
NND to expected average NND ratio (solid rectangles). The
average NDD is computed between centroids for consecutive
months. The expected average NND were calucurated with the
ArcGIS argorithm by assuming random distribution within same
area extent of observed datasets. The numbers next to the dots or
rectangular are the numbers of observations each month. NND
in the months without adequate observations (np30: January,
November, and December) were noted with open symbols. The
spike in NND for April indicates movement between March and
April. Low NND values in March and fall indicate that herring
were most clustered in those months.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2841
Author's personal copy
association with gonad maturation level and migra-
tion speed depends upon the differences of body size
or age (Ware and Tanasichuk, 1989). Our data do
not allow us to conclude whether migration patterns
of young herring are learned from old herring, but
from 1978 to 2004, larger herring do tend to spawn
first in northern Bristol Bay (Tojo and Kruse,
unpublished data).
A conceptual model of fish migration was
developed by Harden Jones (1968, 1980),and
Fig. 5. Comparisons of monthly herring distributions for early and late spawning years in (A) April and (B) May. Herring distributions
are shown for early (left panels: 1981, 1983, 1987, 1993, 1998, 2000, 2001, 2003) and late spawning years (right panels: 1982, 1985, 1986,
1988, 1992, 1994, 1999). For each comparison, the top panel is a comparison between the monthly peaks of herring (dots) and average
total concentration of sea ice (%) and the bottom panel is a comparison of catch-per-unit-effort with average SST (1C) for the same years.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482842
Author's personal copy
improved by Cushing (1981) and others (e.g., Secor,
2002) to incorporate an ontogenetic component of
migration. In overview, this conceptual model
involves a triangle of fish movement throughout
their life history. It includes advection of larvae
from a spawning area (one point of the triangle) to a
nursery area (a second point of the triangle). Once
herring attain sexual maturity, they join the adult
population, with repeated seasonal migrations
between feeding (the third point of the triangle)
and the spawning areas. Based on our analysis, it
seems that this migration model does not fit the EBS
herring very well. Although after spawning some
herring probably return to the major overwintering
grounds northwest of the Pribilof Islands for
feeding, another major herring concentration can
Fig. 5. (Continued)
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2843
Author's personal copy
be found north of Unimak Pass in summer.
Presumably, these post-spawning herring are ac-
tively feeding so the major feeding and overwinter-
ing grounds do not always coincide. Likewise,
although there is a clear northward movement of
herring in the fall, some herring remain in the
southern habitat around the southeastern EBS shelf
area in the winter (Fig. 6), counter to the simple
Harden Jones migration triangle.
A northwestward shift of the northern over-
wintering concentration has occurred in recent years
that could be related to regional warming that
began in 1989 with a shift in the Arctic Oscillation
(Overland et al., 2004). However, such ideas require
cautious interpretation because of the potential
influence of other factors. The distribution of
herring that spawn in the Gulf of Anadyr can
overlap the distribution of EBS herring in their
overwintering area northwest of the Pribilof Islands
(Wespestad and Gunderson, 1991). Therefore, all
herring bycatch in groundfish fisheries in the US
EEZ is not necessarily of US origin.
Because of their long-distance migration and their
abundance, Pacific herring play important roles in
energy transfer throughout the Bering Sea ecosys-
tem. Firstly, their seasonal migration facilitates
cross-shelf transport of energy among offshore
and inshore environments. Secondly, as forage
fishes they are important conveyors of energy from
lower trophic levels (planktonic organisms) to
various upper trophic levels, including various bird
species (Bishop and Green, 2001;Suryan et al.,
Fig. 6. Revised eastern Bering Sea (EBS) herring migration pattern and spatial variability in pre-spawning migration. The gray lines are
the 50-, 100-, and 200-m isobaths. Thick rectangles show historical spawning locations. The two gridded circles represent the major
wintering grounds. The gray boxed arrows show pre-spawning movements, and the open boxed arrows show post-spawning movements.
The thick black arrow indicates postulated spatial variability of major pre-spawning migration passage of herring (dotted gray box arrow)
due to sea ice edge thermal dynamics.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482844
Author's personal copy
2002;Rodway et al., 2003) and large megafauna
(Sharpe and Dill, 1997;Sharpe, 2002;Gende et al.,
2001;Thomas and Thorne, 2001).
Environmental conditions associated with the
northern and southern feeding/wintering grounds
likely affect predator and prey interactions and
bioenergetics. The northward shift of overwintering
aggregations may have both costs and benefits to
herring. Costs may include some energetic expendi-
tures necessary to swim greater distances between
feeding and spawning areas. Potential benefits
include the avoidance of predation by groundfish
species that are abundant in the south (Adlerstein
and Trumble, 1998), better local feeding conditions,
and/or lower basal metabolic rates associated with
colder temperatures to the north. Norwegian
Atlantic herring show size- and age-specific differ-
ences in migration distances due to costs and
benefits during the feeding migration (Kvamme
et al., 2003). Perhaps EBS herring schools make a
choice of migration route based on similar trade-
Shifts in location of herring overwintering areas
have practical management implications. In 1991,
the North Pacific Fishery Management Council
created a herring savings area (58–601N,
172–1751W) to protect overwintering herring from
incidental bycatch in other fisheries, particularly the
winter walleye pollock fishery. The herring savings
area is closed to fishing if observed herring bycatch
reaches the Prohibited Species Catch (PSC) limit,
which is set annually at 1% of the estimated EBS
herring biomass. However, closure of the herring
savings area has rarely been triggered. With
apparent shifts in the overwintering grounds of
herring, it may be necessary to realign the winter
herring savings area to protect herring in the future.
Also, relationships of other overwintering herring
populations in this region may need additional
considerations, particularly for determinations of
the ‘‘unit stock’’ and accounting of total fishing
mortality experienced during the herring’s life
In conducting this analysis of patterns, herring
bycatch data were used as an index of relative
abundance. By doing so, it was assumed that trawl
fishing gears sampled herring in proportion to their
local densities and that representative samples of the
catches were taken by onboard observers. Data
were not collected in a way that would allow us to
examine potential day/night differences in the
vertical distributions of herring and associated
differences in catch rates nor to examine differences
among juvenile and adult herring. Although there is
no way to test these assumptions with the available
data, the general similarity between monthly herring
distributions inferred from herring bycatch
estimates (Fig. 3) and those inferred from the
historical foreign fisheries that targeted herring
(e.g., Wespestad, 1978, also see Appendix 1.A in
Tojo, 2006), suggest that inferences about adult
herring biomass distributions from these observa-
tions are reasonable. Because multiple observations
were smoothed over time and space, the results are
not very sensitive to deviations from these assump-
tions for individual tows.
4.2. Relationship between interannual herring
migration dynamics and EBS ice edge variability
We posit that interannual variability of pre-
spawning EBS herring migration timing and migra-
tion route in spring, particularly in April, is an
adaptive response by herring to ice edge thermal
dynamics (Fig. 5). In the western Atlantic Ocean,
the size of the sub-zero water mass in spring
determines whether shoreward movements of her-
ring are inhibited during their spawning migration
in the southern Gulf of St. Lawrence (Messieh,
1987). In the southeastern Bering Sea, cold melt-
water extending from the ice edge (Alexander and
Niebauer, 1981) and a stable thermocline (Muench
and Schumacher, 1985) may guide herring schools
along optimal isotherms as they migrate to the
spawning grounds.
Responses of herring to sea ice are probably
complex due to schooling behaviors and variable
availability of optimal habitats. Temperature gra-
dients appear to affect the distribution of Atlantic
herring off Norway (Misund et al., 1998;Nøttestad
et al., 1999), and temperature may influence Pacific
herring distribution in the EBS (Barton and
Wespestad, 1980). Energy expenditures during
herring pre-spawning migrations are higher than
during overwintering in northern Norwegian her-
ring (Slotte, 1999b). It appears that Pacific herring
follow specific isotherms along their migration route
from offshore overwintering areas to coastal spawn-
ing grounds, but it is unclear whether there is a
bioenergetic basis for this phenomenon.
There are records of pre-spawning herring occur-
ring in sub-zero temperatures in the EBS
(Rumyantsev and Darda, 1970), including observa-
tions of herring spawning under the ice in cold years
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2845
Author's personal copy
in northern Bristol Bay (Jim Browning, 2003,
personal communication). Similar instances were
observed in other coastal regions of the south-
eastern Bering Sea (Christie Hendrich, 2003, perso-
nal communication). Likewise, information
provided by local residents identified some small
wintering aggregations under the ice in Norton
Sound (Barton and Wespestad, 1980). These contra-
dictory observations might be explained by herring
schooling behavior (Ferno
¨et al., 1998;Huse et al.,
2002). For instance, a simulation study of Atlantic
herring revealed changes in direction of an entire
school when more than 7% of herring moved in a
certain direction (Huse et al., 2002). The coarse
sampling scheme of our data does not allow us to
examine such school dynamics for EBS herring.
Physiological status and ecological experiences of
segments of herring schools need to be taken into
account to realistically predict migration dynamics
or distribution. Such sampling can only be attained
through a directed at-sea sampling program. Based
on the aggregate samples available for this analysis,
we propose that pre-spawning herring tend to avoid
cold temperatures associated with high ice concen-
trations, unless a proportion of a school reaches the
physiological requirements for spawning. When a
proportion of herring in a school reaches a
spawning-ready threshold, the whole school may
move into cold water that they otherwise would
have avoided.
An ongoing decline of sea ice in the EBS
(Overland and Stabeno, 2004) presents an impor-
tant research opportunity to investigate changes in
the distribution and role of herring in the marine
ecosystem. How will herring respond to a future
lack of sea ice? Shifts in overwintering grounds to
the northwest of the Pribilofs, and the apparent
bifurcation of herring distribution in winter includ-
ing a concentration north of Unimak Pass, are
precursors, perhaps, of further responses. Should
the system proceed toward ice-free conditions,
herring will experience a new thermal regime
beyond the range of historical observations. Addi-
tional future research should include detailed field
investigations, including herring genetic identifica-
tion studies to examine the habitat overlap of Gulf
of Anadyr and EBS spawning populations in
winter, and effects of climate-driven oceanographic
changes on herring distributions and abundance, as
well as the impacts of such changes in this
ecologically important species on upper trophic
5. Conclusions
EBS herring have a basic migration pattern that is
characterized by southern and northern overwinter-
ing areas, an inshore movement in spring, and
migration along the Alaska Peninsula after spawn-
ing in summer and early fall. Significant interannual
variability exists in the pre-spawning herring migra-
tion pathways, largely in response to sea ice
variability. The exact response of herring to sea
ice is probably related to water conditions in the
migration route between the offshore northern and
southern wintering areas and coastal spawning sites.
The recent and continuing decline of sea ice in the
EBS will undoubtedly affect herring in the future.
The potential impacts of a warmer EBS with no sea
ice on herring migration are not predictable,
because they are outside the range of historical
observations. However, continuing environmental
changes provide opportunities to conduct new
research on the interactions between sea ice and
herring and their environment. Given the ecological
importance of herring to upper trophic levels, as
well as the economic and social importance of
herring fisheries to coastal residents of this region,
such studies will be important, both to understand
the effects of global warming on the Bering Sea
ecosystem, as well as to fishery management.
Possible changes in the northern overwintering
grounds and details of the southern overwintering
concentrations pose scientific and management
challenges and research opportunities. The under-
standing of herring migration dynamics that we
have tried to advance with our research should be
validated and expanded by in-situ field studies, as
well as laboratory investigations of reproductive
physiology and behavior.
We wish to express our thanks to Professors
Terry Quinn and Dave Musgrave for very helpful
comments on a draft of this manuscript. We are
particularly grateful to Professor Quinn for his
advice on methods for CPUE standardization. We
appreciate insightful discussions with Vidar We-
spestad and Doug Hay regarding EBS herring
biology. This publication is the result of research
sponsored in part by the North Pacific Research
Board and by Alaska Sea Grant with funds from the
National Oceanic and Atmospheric Administration
Office of Sea Grant, Department of Commerce,
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482846
Author's personal copy
under Grant no. NA 16RG2321 (Project no. R/02-
02), and from the University of Alaska with funds
appropriated by the state. It was initially presented
at the GLOBEC ESSAS Symposium on the effects
of climate change on sub-arctic marine ecosystems
in June 2005. This paper was first presented in the
GLOBEC-ESSAS Symposium on Effects of climate
variability on sub-arctic marine ecosystems, hosted
by PICES in Victoria, BC, May 2005.
Adlerstein, S.A., Trumble, R.J., 1998. Pacific halibut bycatch in
Pacific cod fisheries in the Bering Sea: an analysis to evaluate
area–time management. Journal of Sea Research 39, 153–166.
Alexander, V., Niebauer, H.J., 1981. Oceanography of the
eastern Bering Sea ice-edge zone in spring. Limnology and
Oceanography 26, 1111–1125.
Barton, L.H., Wespestad, V.G., 1980. Distribution, biology, and
stock assessment of western Alaska’s herring stocks. In:
Proceedings of the Alaska Herring Symposium. Alaska Sea
Grant College Program Report 80-4. Fairbanks: University of
Alaska Fairbanks, pp. 27–53.
Bishop, M.A., Green, S.P., 2001. Predation on Pacific herring
(Clupea pallasi) spawn by birds in Prince William Sound,
Alaska. Fisheries Oceanography 10 (suppl. 1), 149–158.
Corten, A., 2002. The role of ‘‘conservatism’’ in herring
migrations. Reviews in Fish Biology and Fisheries 11,
Cushing, D.H., 1981. Fisheries Biology: A Study in Population
Dynamics, second ed. University of Wisconsin Press, Madi-
son, 295pp.
Dudnik, Y.I., Usol’tsev, E.A., 1964. The herring of the eastern
part of the Bering Sea. In: Moiseev, P.A. (Ed.), Soviet
Fisheries Investigations in the Northeastern Pacific, Part II.
Israel Program for Scientific Translations, Jerusalem, pp.
¨, A., Pitcher, T.J., Mele, W., Nøttestad, L., Mackinson, S.,
Hollingworth, C., Misund, O.A., 1998. The challenge of the
herring in the Norwegian Sea: making optimal collective
spatial decisions. Sarsia 83, 149–167.
Funk, F.C., 1990. Migration of eastern Bering Sea herring, as
inferred from 1983 to 1988 joint venture and foreign trawl
bycatch rates. Alaska Department of Fish & Game, Division
of Commercial Fisheries, Regional Information Report 5J90-
04, Juneau.
Gende, S.M., Womble, J.N., Wilson, M.F., Marston, B.H., 2001.
Cooperative foraging by Steller sea lions, Eumetopias jubatus.
Canadian Field-Naturalist 115, 355–356.
Harden Jones, F.R., 1968. Fish Migration. Edward Arnold,
London, 325p.
Harden Jones, F.R., 1980. The nekton: production and migration
patterns. In: Barnes, R.K., Mann, K.H., (Ed.). Blackwell,
Oxford, pp. 119–142.
Hay, D.E., 1985. Reproductive biology of Pacific herring (Clupea
harengus pallasi). Canadian Journal of Fisheries and Aquatic
Sciences 42 (Suppl. 1), 111–126.
Hay, D.E., McCarter, P.B., 1997. Continental shelf area and
distribution, abundance, and habitat of herring in the North
Pacific. In: The Role of Forage Fishes in Marine Ecosystems.
University of Alaska Sea Grant College Program Report 97-
01, University of Alaska Fairbanks, Fairbanks, pp. 559–572.
Huse, C., Railback, S., Ferno
¨, A., 2002. Modeling changes in
migration pattern of herring: collective behavior and numer-
ical domination. Journal of Fish Biology 60, 571–582.
Kvamme, C., Nøttestad, L., Ferno
¨, A., Misund, O.A., Dom-
masnes, A., Axelsen, B.E., Dalpadado, P., Misund, O.A.,
2003. Migration patterns in Norwegian spring-spawning
herring: why young fish swim away from the wintering area
in late summer. Marine Ecology Progress Series 247, 197–210.
Mecklenburg, C.W., Mecklenburg, T.A., Thorsteinson, L.K.,
2002. Pacific herring. In: Fishes of Alaska. American Fisheries
Society, Bethesda, p. 134.
Messieh, S.N., 1987. Some characteristics of Atlantic herring
(Clupea harengus) spawning in the Southern Gulf of St.
Lawrence. Northwest Atlantic Fisheries Organization Scien-
tific Council Studies 11, 53–61.
Misund, O.A., Vilhjalmsson, H., Jakupsstovu, S.H., Rottingen, I.,
Belikov, S., Asthorsson, O., Blindheim, J., Jo
´nsson, J., Krysou,
A., Malmberg, S.A., Sveinbjørnsson, S., 1998. Distribution,
migration and abundance of Norwegian spring spawning
herring in relation to the temperature and zooplankton
biomass in the Norwegian Sea as recorded by coordinated
surveys in spring and summer 1996. Sarsia 83 (2), 117–127.
Muench, R.D., Schumacher, J.D., 1985. On the Bering Sea ice
edge front. Journal of Geophysical Research, Section C:
Oceans 90 (C2), 3185–3197.
Nøttestad, L., Misund, O.A., Orvik, K.A., Hoddevik, B., 1999.
Influence of sea temperature on herring distribution and
migration in the Norwegian Sea in April. International
Council for the Exploration of the Sea CM 1999/M:03.
Overland, J.E., Spillane, M.C., Soreide, N.N., 2004. Integrated
analysis of physical and biological pan-arctic change. Climatic
Change 63, 291–322.
Overland, J.E., Stabeno, P.J., 2004. Is the climate of the Bering
Sea warming and affecting the ecosystem? Eos. Transactions
of the American Geophysical Union 85 (33), 309–316.
Prokhorov, V.G., 1970. Winter period of life of herring in the
Bering Sea. Proceedings of the Pacific Scientific Research
Institute of Fisheries & Oceanography 64, 329–338 [The
Translation Bureau (MJK) Foreign Language Division,
Department of the Secretary of State of Canada, Ottawa].
Rodway, M.S., Regehr, H.M., Ashley, J., Clarkson, P.V.,
Goudie, R.I., Hay, D.E., Smith, C.M., Wright, K.G., 2003.
Aggregative response of harlequin ducks to herring spawning
in the Strait of Georgia, British Columbia. Canadian Journal
of Zoology 81, 504–514.
Rumyantsev, A.I., Darda, M.A., 1970. Summer herring in the
eastern Bering Sea. In: Moiseev, P.A. (Ed.), Soviet Fisheries
Investigations in the Northeastern Pacific. Part 5. Pishchevaya
Promyshlennost. Israel Program for Scientific Translations,
Jerusalem, pp. 409–441.
Secor, D.H., 2002. Historical roots of the migration triangle.
ICES Marine Science Symposia 215, 329–335.
Sharpe, F.A., 2002. Social foraging of the southeast Alaskan
humpback whale, Megaptera novaengliae. Dissertation Ab-
stracts International Part B: Science and Engineering 62,
Sharpe, F.A., Dill, L.M., 1997. Behavior of Pacific herring
schools in response to artificial humpback whale bubbles.
Canadian Journal of Zoology 75, 725–730.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–2848 2847
Author's personal copy
Silverman, B.W., 1986. Density Estimation for Statistics and
Data Analysis, Monographs of Statistics and Applied
Probability, vol. 26. Chapman & Hall/CRC, New York,
Slotte, A., 1999a. Effects of fish length and condition on
spawning migration in Norwegian spring spawning herring
(Clupea harengus L.). Sarsia 84, 111–127.
Slotte, A., 1999b. Differential utilization of energy during
wintering and spawning migration in Norwegian spring-
spawning herring. Journal of Fish Biology 54, 338–355.
Slotte, A., Fiksen, O., 2000. State-dependent spawning migration
in Norwegian spring-spawning herring. Journal of Fish
Biology 56, 138–162.
Suryan, R.M., Irons, D.B., Kaufman, M., Benson, J.,
Jodice, P.G.R., Roby, D.D., Brown, E.D., 2002. Short-term
fluctuations in forage fish availability and the effect on prey
selection and brood-rearing in the black-legged kittiwake
Rissa tridactyla. Marine Ecology Progress Series 236,
Thomas, G.L., Thorne, R.E., 2001. Night-time predation by
Steller sea lions. Nature 411 (6841), 1013.
Tojo, N., 2006. Environmental cues for Pacific herring (Clupea
pallasii) spawning in northern Bristol Bay. M.S., University of
Alaska Fairbanks, Fairbanks, 140pp.
Ware, D.M., Tanasichuk, R.W., 1989. Biological basis of
maturation and spawning waves in Pacific herring (Clupea
harengus pallasi). Canadian Journal of Fisheries and Aquatic
Sciences 46, 1776–1784.
Wespestad, V.G., 1978. Exploitation, distribution and life history
features of Pacific herring in the Bering Sea. National Marine
Fisheries Service, Northwest and Alaska Fisheries Center,
Seattle, Washington, USA.
Wespestad, V.G., Barton, L.H., 1981. Distribution, migration,
and status of Pacific herring. In: Hood, D.W., Calder, J.A.
(Eds.), The Eastern Being Sea Shelf: Oceanography and
Resources, vol. 1. National Oceanic and Atmospheric
Administration, pp. 509–525.
Wespestad, V.G., Gunderson, D.R., 1991. Climatic induced
variation in Eastern Bering Sea herring recruitment. In:
Proceeding of the International Herring Symposium. Alaska
Sea Grant College Program Report 91-01. University of
Alaska Fairbanks, Fairbanks, pp. 127–140.
N. Tojo et al. / Deep-Sea Research II 54 (2007) 2832–28482848
... Shark A's return migration and residence in the Bering Sea is also likely influenced by foraging opportunities. Interestingly, Shark A returned to the Bering Sea in August, well after most Pacific salmon (Oncorhynchus spp.) from that region return to their spawning streams [27][28][29][30][31]. Given this, Shark A's return migration to the Bering Sea may coincide with fall aggregations of Pacific herring (Clupea pallasii) in the western Bering Sea [15,31] or a return north for mating. ...
... Interestingly, Shark A returned to the Bering Sea in August, well after most Pacific salmon (Oncorhynchus spp.) from that region return to their spawning streams [27][28][29][30][31]. Given this, Shark A's return migration to the Bering Sea may coincide with fall aggregations of Pacific herring (Clupea pallasii) in the western Bering Sea [15,31] or a return north for mating. Shark A's PSAT provided a single year of movement information, so it is unknown whether a similar pattern would be conducted annually. ...
... In contrast to year one, after visiting the Emperor Seamount Chain region during fall of its second year, Shark B made a brief northerly foray into the southeastern Bering Sea shelf during the months of February to March. While the reasons for this observed movement pattern is speculative, the timing and location of this movement closely overlaps spatially and temporally with immature Chinook salmon from western Alaska [35], overwintering Pacific herring [31], and walleye pollock (Theragra chalcogramma) [36], all of which are known prey species for salmon sharks [4,7,8,16]. In both years Shark B initiated a movement northward to the Bering Sea in June, which may be related to large-scale migration patterns of Pacific salmon species returning north to their spawning rivers in the Bering Sea region [27][28][29][30][31], or possibly a return to north to mate. ...
Full-text available
Background The salmon shark ( Lamna ditropis ) is a widely distributed apex predator in the North Pacific Ocean. Many salmon sharks from the eastern North Pacific, specifically Prince William Sound, Alaska, have been satellite tagged and tracked, but due to the sexual segregation present in salmon sharks, most of these tagged sharks were female. Consequently, little information exists regarding the migration patterns of male salmon sharks. To better understand the migration and distribution of this species, information on the male component of the population as well as from sharks outside of Prince William Sound, Alaska, is needed. In this study, we deployed satellite transmitters on two mature male salmon sharks caught in the Bering Sea. Results The two mature male salmon sharks tagged in the Bering Sea exhibited distinct migration patterns. The first male, tagged in August 2017, traveled to southern California where it remained from January to April after which it traveled north along the United States’ coast and returned to the Bering Sea in August 2018. The second male, tagged in September 2019, remained in the North Pacific between 38° N and 50° N before returning to the Bering Sea in July of year one and as of its last known location in year two. The straight-line distance traveled by the 2017 and 2019 sharks during their 12 and 22 months at liberty was 18,775 km and 27,100 km, respectively. Conclusions Before this study, our understanding of salmon shark migration was limited to female salmon sharks satellite tagged in the eastern North Pacific. The 2017 male salmon shark undertook a similar, but longer, north–south migration as tagged female sharks whereas the 2019 shark showed little overlap with previously tagged females. The different migration patterns between the two male sharks suggest distinct areas exist for foraging across the North Pacific. The return of both sharks to the Bering Sea suggests some fidelity to the region. Continued tagging efforts are necessary to understand the population structure of salmon sharks in the North Pacific. This tagging study highlights the importance of opportunistic efforts for obtaining information on species and sex with limited distribution data.
... Seasonal migrations are also common for fish species in the highly seasonal North Pacific ecosystem. Pacific herring (Clupea pallasii) and yellowfin sole (Limanda aspera), for example, exploit abundant food resources as they migrate between summer feeding grounds and offshore overwintering grounds (Nichol 1998;Tojo et al. 2007). Both the ontogenetic and seasonal migration patterns of polar cod in the Pacific Arctic are not well established and could be improved with additional sampling beyond the August and September open water sampling season. ...
... The migration of a small-bodied, high-latitude fish species is not unprecedented, but there remains much uncertainty surrounding the migration patterns of polar cod in the Chukchi Sea. Other marine fish species, such as Pacific herring and walleye pollock in the Bering Sea, exhibit seasonal migrations between feeding grounds and spawning grounds (Kotwicki et al. 2005;Tojo et al. 2007). Additionally, telemetry studies in the Atlantic Arctic found that polar cod is physically capable of traveling over 100 km in response to rapidly evolving ice conditions (Kessel et al. 2015). ...
Full-text available
Polar cod (Boreogadus saida) is a key forage fish in the Arctic marine ecosystem and provides an energetic link between lower and upper trophic levels. Despite its ecological importance, spatially explicit studies synthesizing polar cod distributions across research efforts have not previously been conducted in its Pacific range. We used spatial generalized additive models to map the distribution of polar cod by size class and relative to environmental variables. We compiled demersal trawl data from 21 cruises conducted during 2004–2017 in the Chukchi and Beaufort seas, and investigated size-specific patterns in distribution to infer movement ecology of polar cod as it develops from juvenile to adult life stages. High abundances of juvenile polar cod (≤ 70 mm) in the northeastern Chukchi Sea and western Beaufort Sea were separated from another region of high abundance in the eastern Beaufort Sea, near the US and Canadian border, suggesting possible population structure in the Pacific Arctic. Relating environmental correlates to polar cod abundance demonstrated that temperature and salinity were related to juvenile distribution patterns, while depth was the primary correlate of adult distribution. A comparison of seasonal 2017 abundances of polar cod in the southern Chukchi Sea found low demersal abundance in the spring when compared to the summer. Seasonal differences in polar cod abundance suggest that polar cod migration may follow a classical ‘migration triangle’ route between nursery grounds as juveniles, feeding grounds as subadults, and spawning grounds as adults, in relation to ice cover and seasonal production in the Chukchi Sea.
... Such consequences may impede the functions, durability, and resilience of ecosystems and increase the risk of population extinction [3,48]. Pacific herring, for example, received the highest phenology (P) score in this study and showed changes in migration timings and life cycle pathways due to rising seawater temperature and marine environmental changes, which have affected the Pacific herring resources in the eastern Bering Sea [57]. ...
Full-text available
Climate change is expected to cause changes in marine biota and ecosystems, thereby directly affecting fishery production. To establish policies to respond to climate change, the importance of climate change vulnerability assessments is growing. In South Korea, annual fishery production has decreased since 1986, and climate change has caused changes in compositions of species and the ecological structure. Therefore, we assessed the vulnerability to climate change for 36 species with sensitivity and exposure. Based on this result, the vulnerability of 24 fisheries to climate change was evaluated. In this study, as exposure factors, we considered relationships between future seawater temperatures and spawning/habitat temperature of each species. Species with high scores both in sensitivity attributes and climate exposure factors are evaluated as highly vulnerable species and fisheries with high catch ratios of such species are assessed to be relatively more vulnerable. Hence, it is required to prioritize fisheries with high catch ratios of relatively vulnerable species when establishing policies to manage offshore and coastal fisheries in Korea.
... Thus, increased freshwater input in the western sound could result in cooler temperatures and delayed spawning. For example, in the Bering Sea, migration of herring to spawning grounds is highly correlated with ice melt, as herring likely use thermoclines to orientate themselves (Tojo et al. 2007). Ward et al. (2017) showed that herring productivity is negatively correlated with freshwater discharge in the Prince William Sound region. ...
Full-text available
Shifts in spawning phenology may impact the early life stages of small pelagic fishes, affecting their first-year survival and recruitment. In Prince William Sound, Pacific herring is a key forage species that once supported commercial and subsistence fisheries for many decades, but collapsed in 1993 and has yet to recover. Starting in 1980, spawn timing shifted earlier by approximately 2−4 wk over a 27 yr period, then abruptly shifted later by approximately 3 wk over the next 7 yr. We quantified the influence of 15 environmental and population-level covariates on these spawn timing shifts using generalized linear models. Earlier spawn timing was associated with higher biomass in the eastern sound and older mean age in the western sound. Across the entire sound, earlier spawning was associated with weaker downwelling, weaker meridional winds, and the positive phase of the Pacific-North American teleconnection pattern, which is characterized by warmer North Pacific waters. These results are a critical first step towards assessing how changes in spawning phenology impact first-year survival of herring offspring and potentially contribute to persistent poor recruitment that has inhibited the recovery of the Prince William Sound population.
... This difference in herring size can result both from changes in human fishing activities and Pacific herring's behavioral change due to the cooling climate event, as was the case with cod. Previous studies suggest that reproductive behaviors and migration patterns of Pacific herring can change due to changes in the marine environment, including water temperature (Hay et al., 2008;Tojo et al., 2007), though how these would affect stable isotope values is unknown. In fact, Pacific herring around Hokkaido experienced a large change in population and migration patterns in the 20th century due to changes in the marine ecosystem likely caused at least in part by intensive modern commercial fishing (Kobayashi, 2002;Nagasawa, 2001). ...
Stable isotope analysis is one of the most effective methods of reconstructing human fishing practices and changes in past marine ecosystems. The effectiveness of this method can be further improved when considering diachronic changes in stable isotope ratios of archaeological remains of several different fish species that exhibit different behavioral or ecological traits. In this study, diachronic changes in human fishing practices and marine ecosystems were investigated for Epi-Jomon (299–258 BC) and Okhotsk (489–1200 AD) periods in prehistoric Hokkaido, northern Japan, by utilizing the stable isotope analysis of archaeological fish bone collagen. Carbon and nitrogen stable isotope ratios of 242 fish bone samples, representing 12 taxa, excavated from the site of Hamanaka 2 on Rebun Island revealed significantly lower (p < 0.05) nitrogen isotope ratios in cod from the Okhotsk period than the Epi-Jomon period. This difference could be related to the development of fishing gear and/or to changes in fishing strategies in the Okhotsk period, as well as to changes in the behavior of cod because of the rapid cooling climate event separating the two periods. Our results demonstrate that some aspects of past human fishing practices and marine ecosystem change can be reconstructed by considering diachronic changes in the stable isotope ratios of several fish species together.
... Pacific herring also exhibit variability in migration distance. The prime example of long migration in these waters is Togiak herring in the Alaskan Bering Sea, which migrate into inshore areas (Bristol Bay) during the spawning season but stay offshore (Unimak Pass and between Pribilof Islands and St. Matthew Island) during the overwintering and feeding period, migrating * 2100 km over this time period (Tojo et al. 2007). At the other end of the scale in these waters are BCH, a herring of smaller body size, which spawn in the Strait of Georgia and feed in shelf regions off the west coast of Vancouver Island, covering a total distance of 400 km (Hay et al. 2001). ...
Full-text available
Life-history traits of Pacific (Clupea pallasii) and Atlantic (Clupea harengus) herring, comprising both local and oceanic stocks subdivided into summer-autumn and spring spawners, were extensively reviewed. The main parameters investigated were body growth, condition, and reproductive investment. Body size of Pacific herring increased with increasing latitude. This pattern was inconsistent for Atlantic herring. Pacific and local Norwegian herring showed comparable body conditions, whereas oceanic Atlantic herring generally appeared stouter. Among Atlantic herring, summer and autumn spawners produced many small eggs compared to spring spawners, which had fewer but larger eggs—findings agreeing with statements given several decades ago. The 26 herring stocks we analysed, when combined across distant waters, showed clear evidence of a trade-off between fecundity and egg size. The size-specific individual variation, often ignored, was substantial. Additional information on biometrics clarified that oceanic stocks were generally larger and had longer life spans than local herring stocks, probably related to their longer feeding migrations. Body condition was only weakly, positively related to assumingly in situ annual temperatures (0–30 m depth). Contrarily, body growth (cm × y−1), taken as an integrator of ambient environmental conditions, closely reflected the extent of investment in reproduction. Overall, Pacific and local Norwegian herring tended to cluster based on morphometric and reproductive features, whereas oceanic Atlantic herring clustered separately. Our work underlines that herring stocks are uniquely adapted to their habitats in terms of trade-offs between fecundity and egg size whereas reproductive investment mimics the productivity of the water in question.
... Despite their importance in the marine ecosystem, little is known about the movement of forage fish. Because of their small size, and in some species the sensitivity to handling [3], most movement studies of forage fish have relied on fishery-dependent methods including traditional mark-recapture [4,5] or catch-per-unit effort (cpue) [6] analyses. However, with characteristically low recapture rates (e.g. ...
Full-text available
Background Over the past two decades, various species of forage fish have been successfully implanted with miniaturized acoustic transmitters and subsequently monitored using stationary acoustic receivers. When acoustic receivers are configured in an array, information related to fish direction can potentially be determined, depending upon the number and relative orientation of the acoustic receivers. However, it can be difficult to incorporate directional information into frequentist mark-recapture methods. Here we show how an empirical Bayesian approach can be used to develop a model that incorporates directional movement information into the Arnason-Schwarz modeling framework to describe survival and migration patterns of a Pacific herring ( Clupea pallasii ) population in coastal Alaska, USA. Methods We acoustic-tagged 326 adult Pacific herring during April 2017 and 2018 while on their spawning grounds in Prince William Sound Alaska, USA. To monitor their movements, stationary acoustic receivers were deployed at strategic locations throughout the Sound. Receivers located at the major entrances to the Gulf of Alaska were arranged in parallel arrays to determine the directional movements of the fish. Informative priors were used to incorporate the directional information recorded at the entrance arrays into the model. Results A seasonal migratory pattern was found at one of Prince William Sound’s major entrances to the Gulf of Alaska. At this entrance, fish tended to enter the Gulf of Alaska during spring and summer after spawning and return to Prince William Sound during the fall and winter. Fish mortality was higher during spring and summer than fall and winter in both Prince William Sound and the Gulf of Alaska. Conclusions An empirical Bayesian modeling approach can be used to extend the Arnason-Schwarz modeling framework to incorporate directional information from acoustic arrays to estimate survival and characterize the timing and direction of migratory movements of forage fish.
... During years of extensive southerly ice movements, sea ice constrains the operating area of the groundfish vessels, as well as likely influencing the timing of the northward eider migrations. Sea ice information was obtained from the National Ice Center, summarized by amount of ice cover in ten percent increments and entered as a GIS layer as described in Tojo et al. (2007). ...
Technical Report
Full-text available
Spectacled eiders breeding in Alaska were listed as threatened in 1993, followed by the listing of the Alaska-breeding population of Steller’s eider in 1997. Primary reasons for concern for both eider species were the near disappearance of the species from the Yukon-Kuskokwim Delta and the low numbers of breeding birds on the North Slope. Causes of the declines in the Alaska-breeding birds are unknown. The larger Russian component of both eider species is more abundant and mixes with the Alaska breeding population during the late-summer molt and on the wintering grounds. Spectacled eiders molt in both northwestern Alaska and eastern Siberia, and have been only recently discovered wintering in leads in the pack ice south of St. Lawrence Island. Steller’s eiders move to locations in southwestern Alaska during the summer molt, and the mixture of Asian and Alaskan breeding populations winters from the eastern Aleutian Islands to Lower Cook Inlet and Kodiak Island, with Izembek Lagoon being the center of molting and wintering abundance. The potential impacts of fisheries on these eider species has been unknown. The current project was undertaken to describe which commercial fisheries occur in space and time near habitat used by the threatened Alaska components of the two eider populations. Steller’s and spectacled eiders could be envisioned to interact with fisheries in four possible ways: 1) entanglement, 2) competition for prey, 3) collisions or other interference with fishing vessels or fishing-related structures, and 4) habitat exclusion or avoidance because of fishing-related activity. Entanglement of Steller’s and spectacled eiders has not been reported to be a problem to date, but effort distributions were investigated as one way to get a handle on the potential risk from this source. Prey competition is unlikely because commercial fisheries do not directly take the shallow-water mollusk and crustacean species in the Steller's eider diet. Vessel strikes have been reported by observers and are investigated in this report. Habitat exclusion is difficult to quantify, but knowledge of nearby fishing effort distributions is essential for a first step. Seventeen records of collisions between fishing vessels and Steller's eiders appeared in fisheries observer reports which also had reliable time-of-day information, along with nineteen king eider strikes. Plotting the date and time of day of these observations revealed that all Steller's eider collisions except one had occurred at night during the winter and spring, along with all nineteen of the King eider observations. The collisions occurred in light levels well below astronomical twilight. As a result, the vessel collision problem is framed by the annual light regime for further analyses. Combining the presence of substantial numbers of eiders along the Alaska peninsula with the annual light regime, the time period September 1 through April 30 likely represents the greatest risk of collisions between eiders and fishing vessels. Between May 1 and August 30 there is little or no deep darkness in these areas and bird strikes are highly unlikely. Commercial catch records were examined from ADFG fish tickets and NMFS observers to determine the location and time of catch for salmon, herring, shellfish and groundfish fisheries. Very little fishing effort occurs near Spectacled eider critical habitat so detailed analysis of fishing distributions was organized around Steller's eider distributions. In addition to active fishing operations, there is potential for vessel collisions and habitat exclusion when vessels are offloading, anchored up, or transiting to or from fishing grounds. No direct records are kept of these activities. However, port of landing can serve as proxy for fishing-related activities such as offloading catch, fueling and staging areas. For the September through April period, Kodiak and Dutch Harbor/Unalaska were by far the busiest fishing ports, with landings dominated by shellfish and groundfish vessels. Along the Alaska Peninsula, King Cove was ranked 5th, and Port Moller 7th for the number of September through April landings. However, Port Moller landings consisted of late-season salmon, almost all in the early part of September. There were no other landings reported from the north Alaska Peninsula during the September through April period. The locations, timing, and gear used in Pacific herring fisheries are described but nearly all of the fishing effort occurs during the well-lit period of the year when vessel collisions are unlikely. Pacific herring fisheries occur at herring spawning locations throughout the coastal migration route of Steller's eiders. Very little Dungeness crab are harvested from areas near Steller's eider critical habitat; in addition, what Dungeness fishing effort occurs is mostly during the well-lighted summer months. Shrimp landings have been minimal in recent years. Scallop harvests occur offshore in the Bering Sea, but most effort occurs during the summer months. Bristol Bay and the Red King Crab savings area are closed to scallop dredging. The Bristol Bay red king crab fishery, with its brightly lit vessels, could well have the most potential for interactions with late-migrating eiders or overwintering eiders while anchored up or transiting nearshore areas. This fishery has occurred from late October through early December in recent years. Salmon fisheries occur during the summer period with minimal periods of darkness, so that there is very low potential for eider-vessel collisions. No eider collisions with salmon fishing vessels have been reported. Although gillnet entanglements with Steller's eiders have never been reported, other diving seabirds do become entangled in gillnets, so it is conceivable that there is some entanglement risk. Drift gillnet fishing occurs along the north Alaska Peninsula, primarily from Port Moller to Port Heiden. The largest amount of set gillnet effort near the general Steller's eider molting/summering areas occurs at Nelson Lagoon. During the spring migration, no groundfish effort occurred in Bristol Bay when it was open to trawling from April 1 to June 15. Bristol Bay is closed to groundfish trawling during the fall Steller's eider migration. The only groundfish fishing found near areas used by Steller's eiders was the yellowfin sole fishery along the northern shore of Kuskokwim Bay. However, this fishery occurs in May, past the time when astronomical twilight has disappeared from the fishing grounds so that the risk of vessel collisions is minimal. Eider-vessel collisions appear to involve bright lights during periods of darkness, so that it may be possible to detect the distribution of fishing vessels at sea using low-light satellite-borne sensors. NOAA archives of DMP/OLS satellites, which have been used to detect squid fleets at sea, were screened for cloud free periods when Bering Sea fishing fleets might be detected. Among other possibilities, the opening of the 1997 Bristol Bay red king crab fishery on November 1 coincided with a satellite pass over the fishing grounds when cloud cover was less than 10%, providing an ideal opportunity to determine whether crab vessel lights can be detected. However, time constraints precluded examining the imagery from that date. Further advancements in our knowledge of eider migration routes, from telemetry and other sources, could be used to further narrow the scope of analyses of potential conflict between fishing vessels and eiders. Observer records of vessel-eider collisions provided by far the most valuable information in this analysis. However, the seabird collision data is obtained anecdotally in field notes of fishery observers whose primary mission is to estimate groundfish and shellfish catches. A more dedicated, directed, effort to document seabird encounters would greatly enhance our knowledge of how fisheries and seabirds interact.
Timing and duration of spawning migration of Pacific herring is known to vary by fish age and length. Older fish mature earlier, but there are few verified examples of differences in the pre-spawning migration of individuals varying in age or size. Here we use sonic tagging to examine the fluctuations in age, fish length and maturity throughout the migration of herring in Akkeshi, Japan. We tracked the horizontal and vertical movement of individual fish using acoustic telemetry and biologging methodology. In Akkeshi Lake, older, larger individuals migrated early, reached peak maturity and remained in the lake until early April, after spawning. The timing and duration of the stay in Akkeshi Lake coincided with the freezing and melting of the surface layer of the lake. Fish tracking indicated no differences in herring in different length classes migrated to and from Akkeshi Lake and offshore and stayed in Akkeshi Bay. For the vertical movement, larger, older herring occupied a wide depth range during diurnal vertical movements as they moved to Akkeshi Lake and to offshore. In contrast, the smaller group occurred frequently in depths shallower than 5 m throughout the day and night. Greater depth variation could result in beneficial effects in predator avoidance and increased prey encounter opportunities though vertical migration. In contrast, smaller herring that were more confined to surface waters could experience more frequent sensing of the environmental cues of spawning grounds such as low-salinity water and olfactory stimulation near the surface. Therefore, there are differences in the benefits and risks during migration for each fish length, which may result in superior spawning grounds and individual mortality in this area.
Full-text available
To understand the population structure of the Japanese jack mackerel Trachurus japonicus in coastal areas adjacent to the Kuroshio Current (referred to as the “CAK”), we analyzed size composition and commercial landing data of juvenile fish in these areas for the period 2005–2015. Trachurus japonicus does not undergo population‐scale spawning migration, and thus, the connectivity between the spawning and juvenile/adult habitat areas is important. Therefore, our primary aim was to assess the origin of juveniles landed in a number of subareas, including those spawned in local spawning grounds in January–May in the western part of the CAK (w‐CAK), those spawned in May–July in the eastern part (e‐CAK), and those spawned in February–March in the remote spawning ground in the southern East China Sea (s‐ECS). Fishing periods starting in spring (spring onset) were commonly observed in the CAK, which involved relatively small size classes (50–100 mm fork length [FL]). Back estimates based on the growth rate of T. japonicus suggested that the contributions from the s‐ECS probably dominated most of the spring onsets in April–June because the smallest size class (50–70 mm FL) occurred almost exclusively in April–May. In autumn, onset signals were associated with the landing of juveniles from the local spawning ground in an eastern subarea of the e‐CAK. Despite the asymmetric transport and migration flows between the habitat areas of T. japonicus, its population levels may be sustained because the local and remote spawning grounds are used in different seasons.
Full-text available
Steller Sea Lions were observed cooperatively foraging for Eulachon (Thaleichthyes pacificus) and possibly Herring (Clupea pallasi) in Berners Bay, southeast Alaska in spring, 1996-1999.
Full-text available
The distribution and migration of Norwegian spring spawning herring (Clupea harengus) in the Norwegian Sea in spring and summer 1996 were mapped during 13 coordinated surveys carried out by Faroese, Icelandic, Norwegian and Russian research vessels.After spawning at the banks of the Norwegian Coast in February-March, most ofthe spent herring migrated out in the Norwegian Sea through a corridor between 67°N and 68°N. In May, 4 and 5 year old herring, which form the younger part of the spawning stock, were distributed in small schools or scattered layers at 25-100 m depth over large areas of the central Norwegian Sea. Older and larger herring formed large schools, generally at 250-400 m depth near the cold front along the eastern part ofthe Icelandic Exclusive Economic Zone (EEZ). The total abundance of herring in the Norwegian Sea was estimated to be about 47 billion individuals or about 8 million tonnes. In June, the older and larger herring had migrated northwards into the Jan Mayen zone, while the younger herring remained in the southern and central Norwegian Sea. In July, the younger herring had migrated back to the area offVesterMen, northern Norway. In July/August, the larger herring were found in small schools near the surface in the northern Norwegian Sea.Relationships between the temperature distribution, zooplankton abundance and herring distribution and migration are considered. In May, the lowest zooplankton biomass was observed in the central and southern Norwegian Sea. At that time, there were high zooplankton concentrations in the westernmost part of the Norwegian Sea, within the domain of the East Icelandic Current. The herring did not enter this body of cold water with temperatures of 1–2°C in the uppermost 300 m, but migrated to the north and north-east in search of food.
Full-text available
Norwegian spring-spawning herring, Clupea harengus harengus L., are long-lived multiple spawners subject to strong variation in recruitment success. They tend to adopt low-risk, preferred-conservative strategies, yet they display considerable plasticity in migratory behaviour and associated spatial dynamics. Although their migration patterns have long been investigated, few studies have analysed the factors and mechanisms governing spatial dynamics. In this study an ecological and evolutionary perspective is adopted, emphasizing proximate mechanisms that restrict the extent to which herring can localize an optimal habitat.The starting point is the assertion that the herring's migratory behaviour can be explained by an interplay of a few key factors. Despite spatial and temporal variations, the Norwegian Sea has consistent and predictable features, such as the distribution of water masses and timing of seasonal plankton production. Herring may locate favourable habitats by using a combination of predictive orientation mechanisms, based upon genetic factors and leaming, and of reactive mechanisms, such as memorybased state-location comparisons and orientation to gradients in the sea.Changes in herring distribution and density occur on micro-, meso- and macroscale. After reviewing the available information on school density, school size, school size adjustments, synchronized behaviour patterns and swimming speed of both individual schools and school clusters, the authors attempt to form a link across spatio-temporal scales to explain patterns in distribution. Existing theory seems inadequate to explain the dynamic behaviour exhibited by large herring schools, though it must in some way reflect optimal decisions by individual fish. It is suggested that the appropriate resolution for the analysis of herring spatial dynamics in meso- and macroscale could be the school unit, and that we have to analyse how individual fish behaviours bring about these dynamics.
Full-text available
Norwegian spring-spawning herring Clupea harengus make extensive feeding migrations in the Norwegian Sea during summer. At the end of the feeding season herring may meet conflicting demands between moving to areas with uncertain food abundance and predation risk and migrating to the wintering area at the coast. Previous studies on the distribution of herring indicate a general eastward migration in late summer, but do not reveal actual migration patterns and orientation mechanisms. About 700 schools of herring were tracked using multi-beam sonar during 2 surveys in the Norwegian Sea in July 1995 and 1996. In coastal areas off Northern Norway, migrating schools of young herring (2 to 6 yr) swam contrary to expectations westward with a mean migration speed of 0.65 m s(-1). MOCNESS samples of zooplankton biomass, herring stomach samples and condition factors indicated better feeding conditions westward off the continental slope than in coastal areas, strongly suggesting that young herring migrated westward to improve feeding conditions. This migration could be released by low food abundance and be based on a large-scale genetic predictive orientation mechanism. West of the continental shelf and slope, in waters with elevated abundance of larger prey items, migration directions were variable. Schools of young herring reaching higher food supply seemed to stop their westward migration and stay in these waters to feed. A small number of older herring (>Age 6) were located even farther west near the Arctic Front where food concentrations were greatest. Predictive orientation mechanisms toward richer feeding areas may improve with age as a result of experience and learning.
In general the following facts apply for the Norwegian spring spawning herring in the 1990s. It does not feed from the onset of wintering (September-October) in Vestfjorden, northern Norway (68°N), until spawning is completed (March-April). In mid-January it migrates towards spawning grounds within a range of approximately 1500 km along the coast (58°-70°N). The relative weekly energy loss is 3-4 times higher during the spawning migration than during the wintering, and it increases with decreasing fish length. Survival of progeny may increase southwards due to beneficial environmental conditions. It is widely accepted that herring return to spawn at the same spawning grounds year after year (homing). However, this study demonstrated that Norwegian spring spawning herring may deviate from this general homing tendency, due to a trade-off between survival of progeny and physiological migration constraints related to fish length and condition (energy storage). During the spawning seasons in 1995 and 1996 the mature herring were mainly distributed between Sogn and Lofoten (61°-70°N), with the shelf area off Møre (62°-64°N) being most important. Both the fish length, condition and stage of maturity increased southwards, but the condition appeared to be the most important variable influencing the distance migrated and spawning time.
New insight into the feeding habits of these mammals will help conservation attempts.
The ice-edge region of the southeast Bering Sea was studied in terms of the hydrographic regime, phytoplankton biomass, and primary productivity during the springs of 1975 through 1977. The results showed that a phytoplankton bloom occurs at the ice edge just as the spring ice-decay period begins, and that this accounts for a significant proportion of the annual carbon input over the shallow shelf. The bloom is intensified in time and space by the influence of the ice edge on the physical structure of the water column. Specifically, melting ice seems to increase the stability of the water column, near and under the ice, by lowering the salinity. Frontal structure in salinity and temperature are apparent at the ice edge and are attributed to the melting ice but also, at times, to wind-driven Ekman-type upwelling. These data are also related to recent short term (ca. months-year) climatic fluctuations that seem to control the seasonal position of the ice-edge zone relative to the shelf break. In "cold" years, the ice edge comes southward to the shelf break and overlies the more nutrientrich Alaska Stream/Bering Sea source water. In "warm" years, the ice-edge zone does not reach this nutrient-rich water. This may be important to the biology of the ice-edge ecosystem. © 1981, by the American Society of Limnology and Oceanography, Inc.