Interannual variability in abundance of North Sea jellyfish and links to the North
Atlantic Oscillation
Jellyfish abundance is increasing in numerous marine ecosystems
worldwide, perhaps as a consequence of ‘‘regime
shifts’’ associated with climatic change, increasing fishing
pressure, or both (Goy et al. 1989; Brodeur et al. 1999;
Brierley et al. 2001; Mills 2001). Scyphomedusae can consume
large numbers of zooplankton (including copepods,
Behrends and Schneider 1995; Purcell 2003) that could otherwise
be food for fish, and can also consume fish eggs and
larvae (Purcell and Arai 2001). Jellyfish are therefore potentially
detrimental to fisheries, both as competitors with, and
predators of, fish. It has been suggested that environmental
variation affects the abundance and distribution of jellyfish
medusae over a range of temporal and spatial scales (Goy
et al. 1989; Graham et al. 2001). It has also been shown that
food availability, light, salinity, and temperature are all important
to the strobilation of scyphistomae (i.e., the development
of ephyrae by the sessile polyp) and the survival of
medusae (Russell 1970; Purcell et al. 1999; Lucas 2001).
There is a growing awareness that large-scale climatic variation
can lead to changes in zooplankton community composition and abundance, and that this, in turn, can have major
consequences for fisheries (Edwards et al. 2002; Reid et
al. 2003). With ever-declining world fish stocks (Hutchings
2000; Pauly et al. 2002) and possible concomitant increases
in jellyfish (Mills 2001), it is important to understand the
role that climatic forcing might play in influencing jellyfish
abundance and the mechanisms by which it could contribute
to ecosystem variability (Attrill and Power 2002; Brodeur et
al. 2002; Parsons and Lalli 2002).
The North Atlantic Oscillation (NAO) is a climatic phenomenon
that contributes greatly to variability in the weather
system over the North Atlantic, North Sea, and Europe and
has been shown to affect both marine and terrestrial ecosystems
(Marshall et al. 2001; Ottersen et al. 2001; Beaugrand
et al. 2002). The NAO Index (NAOI) is a measure of the
difference in pressure between the Icelandic Low and the
Azores High (Hurrell 1995). When the NAO is in its ‘‘high’’
phase and the NAOI is positive, the pressure dipole is pronounced
and strong westerlies transport warm air to Northern
Europe, resulting in warmer-than-average winters and
relatively warm water in the North Sea. The ‘‘low’’ phase
(negative NAOI) is characterized by a much weaker Azores
High and approximately the reverse of the system described
above. The NAO is significantly related to sea level pressure,
surface winds, wave heights, sea surface temperature, and
current influx to the North Sea (Beaugrand 2003) and has
the potential to influence profoundly the pelagic environment.
Jellyfish are able to respond rapidly to favorable environmental conditions, and blooms in their abundance could
provide a biological indication of interannual variation in
hydroclimatic conditions. Thus, it is expected that, if influential,
NAO-related effects on the ecosystem could rapidly
become evident in jellyfish populations.
Using data on the abundance of medusae of Aurelia aurita,
Cyanea lamarckii, and Cyanea capillata from the North
Sea, we explore here the possibility that jellyfish populations
respond to climatic variation as quantified by the NAOI.
Methods
Jellyfish sampling—Medusa abundance data were collected
over 15 yr (1971–1986, excluding 1984) during the routine
summer International Council for the Exploration of the
Sea (ICES) International 0-group Gadoid Surveys of the
North Sea (Fig. 1) (Hay et al. 1990). Surveys, of which
jellyfish were bycatch, were conducted with the International
Young Gadoid Pelagic Trawl (IYGPT). Every year, trawls
were conducted during June and July, and for the years 1971
and 1972, hauls were also made in August. The maximum
number of trawls made in any one year was 215 (in 1979)
and the minimum was 43 (in both 1985 and 1986). From
2,030 IYGPT trawls throughout the North Sea in this period,
more than 430,000 medusae were caught, identified, and
measured. Three species of jellyfish were particularly conspicuous:
A. aurita, C. lamarckii, and C. capillata.
The IYGPT had mesh sizes of 100 mm in wings, bosom,
and belly, tapering through intermediate mesh size to 10-
mm knotless meshing in the extension piece and codend.
The mouth opening of the net was ;14 m2
. When fished at
a maximum speed of 2.5 knots for 1 h, about 65,000 m3 of
water was filtered, assuming 100% filtration efficiency (Hay
et al. 1990). During the gadoid surveys, the trawl was fished
for 1 h in a standard depth profile. For the first third (0–20
min) of the trawl duration, the net was fished close to the
seabed. The net was then hauled to midwater, or to thermocline
depth where this was known, and fished for a further
20 min. During the third period of the trawl, the net was
fished close to the surface (5–10 m). In depths .150 m, 125
m was taken as the bottom depth; in depths ,30 m, only
the bottom and surface were fished (for 30 min each, Hay
et al. 1990).
The catches of medusae probably provided conservative
estimates of actual abundance because of the variable mesh
sizes and the variable sizes of medusae. The stepped vertical
haul profile could also have led to bias in estimations (Hay
et al. 1990). Nevertheless, the same method was used each
year, and a broad range of medusa diameters (1–47 cm) were
caught in the samples. The sampling efficiency was internally
consistent, and catch data are therefore directly comparable
between years.
Hay et al. (1990) chose to report median medusa catch
values to provide an index of jellyfish abundance (the distributions
of catches were nonnormal, rendering the mean
unacceptable as a maximum likelihood estimator). The
heavily skewed distributions were attributable to the patchiness
and sudden blooms of jellyfish populations, resulting
in a high proportion of nil catches and a few extremely high
catches.
Hay et al. (1990) identified four areas of the North Sea
(Fig. 1) that were representative of the major areas of jellyfish
abundance. However, the regions were not sampled
uniformly, but as follows: east of Shetland (ESh) 1971–1986
excluding 1984 (14 yr), north of Scotland (NoS) 1974 and
1976–1986 excluding 1984 (11 yr), east of Scotland (EoS)
1971–1983 (13 yr), and west of northern Denmark (WND)
1972 and 1975–1983 (10 yr).
Analysis of the interannual variation in abundance of jellyfish
with the NAOI—To test for links between medusa
abundance and the NAO, we investigated data for each species
in each area separately, except ESh, where too few A.
aurita and C. lamarckii medusae were caught to enable robust
statistical analysis. A normalized index of the NAO was
obtained from the National Center for Atmospheric Research
(Climate and Global Dynamics Division [U.S.A.], http://
www.cgd.ucar.edu/;jhurrell/). The winter (December–
March) NAOI of the normalized sea level pressure (SLP)
difference between Lisbon, Portugal, and Stykkisholmur/
Reykjavik, Iceland coincides with the biologically important
period of ephyra production and development in the North
Sea (Russell 1970). The SLP anomalies at each station were
normalized by division of each mean pressure for the period
December–March by the long-term (1865–1984) mean standard
deviation. Normalization was used to avoid domination
of the series by the greater variability of the Stykkisholmur/
Reykjavik station (Hurrell et al. 2003). To compare temporal
changes in abundance of medusae between the regions, the
species data were scaled (by subtracting the mean abundance
from the yearly median abundance and then dividing by the
maximum value of the resulting time series) and plotted
against the inverted NAOI (scaled and inverted by dividing
by the minimum value of the time series). For A. aurita and
C. lamarckii from WND and A. aurita from EoS, an association
was apparent (Fig. 2).
All medusa abundance data were natural logarithm transformed
to normalize the distributions prior to further statis-