predation by herring () and sprat () on in a western baltic sea bay
TRANSCRIPT
ICES Journal of Marine Science, 61: 959e965 (2004)doi:10.1016/j.icesjms.2004.06.016
Short communication
Predation by herring (Clupea harengus) and sprat (Sprattussprattus) on Cercopagis pengoi in a western Baltic Sea bay
Elena Gorokhova, Towe Fagerberg, and Sture Hansson
Gorokhova, E., Fagerberg, T., and Hansson, S. 2004. Predation by herring (Clupea harengus)and sprat (Sprattus sprattus) on Cercopagis pengoi in a western Baltic Sea bay. eICES Journal of Marine Science, 61: 959e965.
Cercopagis pengoi is a pelagic cladoceran that has recently colonized the Baltic Sea and theLaurentian Great Lakes and is recognized as a species with the potential to affect naturalfoodwebs. To study the consumption of C. pengoi by zooplanktivorous fish, stomachcontents of herring (size range 52e252 mm) and sprat (57e116 mm) from a coastal area ofthe northern Baltic proper were examined in parallel with zooplankton samples. The overallproportion of fish preying on C. pengoi was high both for sprat (70%) and for herring(61%), and it accounted for 8G 23% and 20G 33% of prey dry weight in the diets. Thepredation on Cercopagis depends on its abundance and on fish size; herring showeda tendency to become more selective for Cercopagis with increasing size. The majority ofdiapause eggs found in sprat (69%) were immature and appeared digested, while this wasthe case only for 2% of the eggs found in herring. These results suggest that Cercopagis hasbecome a significant component in the diet of zooplanktivorous fish and, therefore, itsabundance may be controlled by fish predation.
� 2004 International Council for the Exploration of the Sea. Published by Elsevier Ltd. All rights reserved.
Keywords: Cercopagis pengoi, diapausing eggs, diet, herring Clupea harengus, invasivespecies, predation, selectivity, sprat Sprattus sprattus.
Received 21 August 2003; accepted 15 June 2004.
E. Gorokhova, T. Fagerberg, and S. Hansson: Department of Systems Ecology, StockholmUniversity, SE-106 91 Stockholm, Sweden. Correspondence to E. Gorokhova: tel.: C46 8164256; fax: C46 8 158417; e-mail: [email protected]
Introduction
Cercopagis pengoi is a pelagic cladoceran that has
recently colonized the Baltic Sea (Ojaveer and Lumberg,
1995; Uitto et al., 1999; Bielecka et al., 2000), including
coastal waters of the Stockholm Archipelago (Gorokhova
et al., 2000). Furthermore, the species has recently ap-
peared in the Great Lakes (MacIsaac et al., 1999), where it
is spreading rapidly (Therriault et al., 2002). The invasion
has raised concerns that it may change the foodweb
structure of native ecosystems (Vanderploeg et al., 2002).
Indeed, when the closely related cercopagid Bythotrephes
longimanus colonized North American lakes, it undoubt-
edly influenced these ecosystems (Yan et al., 2002). When
established, Cercopagis may in a similar way affect resi-
dent zooplankton communities by selective predation, and
this has recently been reported from the Gulf of Riga
(Ojaveer et al., 1999) and Lake Ontario (Benoit et al.,
2002). This may result in decreased grazing pressure on
phytoplankton and enhanced algal blooms. Cercopagis
may also impact fish populations by competing with
1054-3139/$30.00 � 2004 International Coun
0-group fish for small prey, or conversely by becoming
prey itself for older fish (Vanderploeg et al., 2002).
Indeed, zooplanktivorous fish both in the Baltic and in the
Great Lakes have been found to prey on Cercopagis when
available (Ojaveer and Lumberg, 1995; Ojaveer et al.,
1998; Antsulevich and Valipakka, 2000; Bushnoe et al.,
2003). However, the extent to which fish consume
C. pengoi may vary between size classes and species.
Small fish may be unable to use it as prey due to gape
size limitations, while large individuals may consume
a significant proportion of Cercopagis production and
even control its population development.
Herring (Clupea harengus L.) and sprat (Sprattus
sprattus L.) are dominant species both in the commercial
fishery and as zooplanktivores in the Baltic Sea. Before the
invasion, their diets consisted mostly of calanoid copepods
(Acartia spp., Eurytemora affinis, and Temora longicornis)
and cladocerans (Bosmina coregoni maritima and Pleopsis
polyphemoides), varying between the coastal and open sea
areas and between northern and southern parts of the Baltic
proper (Rudstam et al., 1992; Mehner and Heerkloss, 1994;
cil for the Exploration of the Sea. Published by Elsevier Ltd. All rights reserved.
960 E. Gorokhova et al.
Arrhenius, 1996; Antsulevich and Valipakka, 2000). After
its appearance, Cercopagis has become a significant
component in the diet of adult herring but not of the
young-of-the-year (YOY) herring in Estonian and Finnish
coastal waters (Ojaveer and Lumberg, 1995; Ojaveer et al.,
1998; Antsulevich and Valipakka, 2000). These authors
suggested that C. pengoi could improve herring feeding
conditions and growth. However, the extent to which small
fish consume C. pengoi has not been carefully examined
and remains unclear.
The primary objective of this study was to estimate the
contribution of C. pengoi to the diet of herring and sprat in
a coastal area in the northern Baltic proper and to evaluate
the selectivity for Cercopagis of different size classes of
these fish species. In addition, we examined the frequency
and condition of Cercopagis resting eggs in fish stomachs.
The study was carried out during the summer 2002 in
a deep (w30 m), enclosed bay (Himmerfjarden) located in
the southern archipelago of Stockholm. This area was
chosen because (1) the Cercopagis population is estab-
lished in the area since at least 1997 (Gorokhova et al.,
2000), (2) other coastal areas like this are most likely to
become invaded by Cercopagis (Leppakoski and Olenin,
2000), and (3) these habitats are nursery areas for many fish
species, including herring (Axenrot, 2002).
Material and methods
Zooplankton sampling and analyses
Zooplankton were sampled bi-weekly (JulyeOctober 2002,station indicated in Figure 1) from the bottom to the surface
with a 90-mm WP-2 plankton net (B 57 cm). On two
occasions (3 August and 3 September), additional samples
were taken in the upper 5e10 m with a 60-mm plankton net
(B 23 cm). The samples were preserved and analysed
according to the standard protocol of the Baltic Monitoring
Programme (HELCOM, 1988). Dry weights of zooplankton
were calculated according to Rosen (1981) and Hernroth
(1985). When the dates for zooplankton sampling did not
match the dates of fish sampling, zooplankton densities on
the dates of fish sampling were obtained by linear
interpolation between the abundances in the neighboring
days.
Cercopagis individuals were removed from the samples
under a dissecting microscope prior to sub-sampling and
processed separately. Conventional methods of population
analysis of Cercopagidae were employed (Rivier, 1998);
dry weights of Cercopagis were calculated according to
Uitto et al. (1999).
Fish collection and stomach analysis
Fish were collected at three locations (Figure 1) on five
occasions in JulyeSeptember (Table 1). They were caught
during night, in gillnets that were set 4e6 m below the
surface. The nets were 3 m deep and had nine segments
(each 2.5 m) with mesh sizes 5-, 6.25-, 8-, 10-, 12.5-, 15.5-,
19.5-, 24-, and 29-mm bar. Diet analyses focused on, but
were not restricted to, fish from August and September,
because Cercopagis was most abundant during this period.
The collected fish were kept frozen (�18(C) until analysis;sometimes this resulted in deformed caudal fins, and as
a measure of fish size, we used the distance (mm) from the
tip of the nose to the base of the caudal fin. As a measure of
the mouth size, we used the length (mm) of the lower jaw of
the fish. A total of 106 herring and 80 sprat stomachs were
analysed (64 empty; Table 1). When estimating the
presence and condition of C. pengoi resting eggs, all non-
empty stomachs were considered, while for the rest of the
diet analysis, only those with R 8 identifiable objects were
used (95 stomachs).
The effect of fish size on dietary composition was
examined by grouping fish into 50-mm size classes, with
individuals larger than 200 mm forming the group
‘‘O200 mm’’ (Table 1). Herring !100 mm were mainly
YOY fish produced from spawning in the study area (cf.
Arrhenius and Hansson, 1996), while sprat in this size
range were one-year-old fish. The stomach content was
Figure 1. Study sites: zooplankton and fish were collected in
JulyeSeptember 2002 in a coastal area of the northern Baltic
proper. Station H4, Himmerfjarden Bay, w30-m depth; 58(59#N17(44#E e zooplankton collection site; Stations 1, 2 and 3 e fish
collection sites.
961Predation by herring and sprat on Cercopagis pengoi
Table 1. The frequency of occurrence for Cercopagis in the fish stomachs (%), number of prey per stomach (individuals, meanG s.d.), dry
weight of zooplankton prey per stomach (DW, mg, meanG s.d.), and number of fish in each size class from each sampling occasion that
was used for stomach content analysis (total number of fish/individuals with empty stomachs).
Species,
size classes
Frequency of
Cercopagis, %
Number of prey
per stomach;
meanG s.d.
DW of prey
per stomach,
mg; meanG s.d.
Sampling date, 2002
9 July 28 July 13 Aug 1 Sep 16 Sep
Herring (Clupea harengus)
5e10 cm 54 716G 792 3.3G 6.1 d d 11/1 31/10 14/2
10e15 cm 89 429G 564 10.0G 5.3 d d 12/4 20/10 9/1
15e20 cm 50 36G 15 96.5G 105.2 2/0 1/0 d 6/1 1/1
O20 cm 0 16G 1 5.4G 4.1 2/0 d d 2/0 1/1
Total 61 546G 716 12.7G 52.0 4/0 1/0 23/5 59/21 25/5
Sprat (Sprattus sprattus)
5e10 cm 62 589G 564 2.9G 2.9 d d 7/2 1/0 18/11
10e15 cm 71 1327G 1420 8.7G 9.4 d d 10/3 11/7 27/10
Total 70 1044G 1221 6.4G 8.1 d d 17/5 18/7 45/21
analysed using a dissecting microscope and an inverted
microscope and identifying each prey item to the lowest
possible taxonomic level. Cercopagis were recorded in two
ways: (1) by weighing or (2) by counting the identifiable
individuals or their body parts. The first method was used
when the stomach contained a large bolus of individuals
bundled together and the remnants of their bodies were
relatively intact. Each bolus was transferred to a pre-
weighed tin cup, dried at 60(C for 72 h, and weighed
(G2 mg). The number of individuals in the bolus was
calculated by assuming an individual dry weight of 20 mg(Uitto et al., 1999). It is possible that this underestimated
the number of individuals, as some of them were partly
digested, but we are confident that separating individuals
(hundreds to thousands) and counting their remains would
have introduced an even greater error. The second method
was used when the remnants of Cercopagis were not
aggregated and could be counted directly.
The number of Cercopagis resting eggs was determined
from those visible in a bolus and those found free in the
sample. Furthermore, the maturity of each egg was noted
using the characteristics of Bythotrephes diapausing eggs
(Jarnagin et al., 2000), i.e. amount of yolk present,
coloration and thickness of the outer shell. The eggs were
categorized as (1) mature e those with dispersed droplets,
a bright-yellow coloration, thick and distinct shells and (2)
immature e those which were clear or weakly colored on
periphery, with very few or no droplets and thin outer
shells.
Selectivity estimates and statistics
The selection of prey types was estimated using the
Chesson (1983) selectivity index (ai); the index ranges
from 0 to 1, corresponding to complete avoidance and full
selection. In this study the three most abundant prey groups
were considered, i.e. cladocerans (other than C. pengoi),
copepods, and C. pengoi. As data on the in situ mysid
abundance were unavailable, this prey was excluded from
the selectivity estimates.
Using estimated ai values, intra- and inter-specific
differences in selectivity for Cercopagis were evaluated
with a log-linear analysis (Upton, 1978). The analysis was
based on a three-way frequency table with the dimensions:
species (herring and sprat), length (longer or shorter than
the median length, 97 mm), and position (ai higher or lowerthan the median value for Cercopagis). Position was treated
as a dependent variable, while species and length were
independent variables. The number of fish that fit into
a certain cell (e.g., herring! shorter than median
length! ai lower than median value; Table 2) is the
frequency that is compared with the expected frequencies
derived from the total number of observations and row/
column sums.
Results
Food availability
During the study period, Cercopagis densities were always
low (!48 ind. m�3, Figure 2A), comprising !0.1% of the
total zooplankton abundance and 0.4e1.2% of the biomass.
Zooplankton were dominated by rotifers (Keratella coch-
learis and Synchaeta spp., 19e80% of total zooplankton
abundance) and copepods (mostly juvenile stages of
Acartia bifilosa and Eurytemora affinis, 14e70%). Clado-
cerans (Bosmina coregoni maritima, Pleopsis polyphe-
moides, and Podon intermedius) contributed at most 10% to
the total abundance, with usually less than 1e2% (Figure
2B). Cladocerans and copepods were abundant in Ju-
neeJuly, decreasing during August, while the rotifers
showed a very drastic decline during the mid-August
(Figure 2B). Cercopagis was common from midsummer
962 E. Gorokhova et al.
Table 2. The three-way frequency table for the log-linear analysis. To test for differences in prey selectivity between the species,
aCercopagis-values for each fish (Chesson selectivity index) were classified as below or above the median-aCercopagis-value (0.335).
Fish length a Below median-a a Above median-a
Herring Below the median 24 13
Above the median 8 16
Sprat Below the median 5 8
Above the median 11 10
The frequencies of observations above and below the median were then analysed by log-linear statistics. The first step (A) was to generate
a saturated model, in which all possible interactions are included. This model has 0 degrees of freedom (d.f.) and explains all variations in
counts in the frequency table (c2-valueZ 0). In the following steps (BeD), interactions between categories are removed, producing
models that are less efficient in explaining the observed frequencies. As the category position was a response variable, it was only species
and length that could be excluded from interactions. The goal was to find interactions, the exclusion of which has significant negative
effects on the remaining model. The only significant result was obtained when species and size effects on selectivity are combined
(B; shown in bold face).
Considered interactions
between variables d.f. c2 Change in d.f. Change in c2 p-value
A e complete model
included: species! length! position 0 0
B e three-way
interaction excluded
included: species! positionC length! position 1 4.37 1 4.37 0.037
C e three-way
and species! position excluded
included: length! position 2 4.41 1 0.04 O0.05
D e three-way
and length! position excluded
included: species! position 2 6.52 1 2.15 O0.05
and showed a distinct peak on 10 September. The
proportion of gametogenic females was also highest in this
sample (16%, Figure 2A). Most of the gametogenic females
(88%) had one-egg broods and hence we used a 1:1 ratio for
egg:female, when using egg number to estimate the number
of Cercopagis in a fish stomach.
Stomach contents
Copepods (Acartia and Eurytemora) and cladocerans
(mostly Bosmina) were dominant prey for sprat and the
smallest of herring (Figure 3). Besides zooplankton, herring
O114 mm had also eaten mysids (Neomysis integer) and
small fish (data not shown). Irrespective of their size
(herring of 52e252-mm length, sprat of 57e116-mm
length), both fish species fed on Cercopagis and it occurred
in 61% of the analysed herring and in 70% of the sprat.
Herring stomachs contained up to 1299 and those of sprat eup to 116 C. pengoi. The proportion of Cercopagis in the
diet was highest in 100e150-mm herring (Table 1, Figure
3), among which some fish had nothing but Cercopagis in
their stomach. The higher proportion of Cercopagis in
herring stomachs (Figure 3A) may at least in part be
explained by their relatively larger mouth compared to that
of sprat (Figure 4). Thus, although rare in the water column,
Cercopagis contributed substantially to the stomach content
of both fish species (Figure 3). The proportion of
Cercopagis in diets increased with its abundance, with
the strongest correlation derived for 100e150-mm herring
(Pearson’s rZ 0.68, p! 0.003).
Selectivity
There were statistically significant inter-specific differences
in selectivity for Cercopagis between herring and sprat
(Chesson selectivity index; Table 2). The difference was
that the selectivity for Cercopagis increased with fish size
for herring, but not in sprat.
Resting eggs
Resting eggs of Cercopagis were found in the stomachs of
all size classes of the both species (Figure 5), sometimes in
large quantities, e.g. 1299 eggs in a 113-mm herring. They
were more frequent in herring than in sprat, with the highest
frequency of eggs occurring (89%) in 100e150-mm herring
963Predation by herring and sprat on Cercopagis pengoi
(Figure 5A). Most of the eggs in sprat (69%) were immature
(Figure 5B), while this was the case only for 2% of the eggs
found in herring stomachs (6% if the stomach with 1299
eggs was excluded from the calculations). Most of the
0
15
30
45
Parthenogenic females
Gametogenic females
Males
Cerco
pag
is
a
bu
nd
an
ce
, in
d. m
-3
18 Ju
n
23 Ju
n
30 Ju
l
3 A
ug
13 A
ug
28 A
ug
3 S
ep
10 S
ep
23 S
ep
9 O
ct
0
100
200
300
400
500
Rotatoria
Cladocera
nauplii
C I-III
C IV-adults
Sampling date
Zo
op
lan
kto
n ab
un
dan
ce, in
d. 10
3 m
-3
Copepoda
A
B
* * * * *
Figure 2. Abundance and population structure of Cercopagis
pengoi (A) and the native zooplankton community (B) during the
study period. Note the different scales on the Y-axes. Asterisks
indicate fish sampling occasions.
immature eggs found in sprat appeared as at least partially
digested, with only thin empty outer shells containing
essentially no yolk.
Discussion
It has been suggested that Cercopagis will compete for
herbivorous zooplankton with young stages of planktivo-
rous fish, if these are unable to prey on Cercopagis due to
its long caudal spine (Vanderploeg et al., 2002). In our
study, however, fish down to a size of 52 mm fed on
Cercopagis, including YOY herring. Despite its rarity in
plankton (!1% of total zooplankton abundance and
biomass), Cercopagis was ingested with high frequency
by all fish !200 mm (Table 1). The predation on
Cercopagis appears to depend on its abundance and to
some extent on fish size; in particular, herring showed
a tendency to become more selective for Cercopagis with
increasing size. Thus, the colonization by Cercopagis has,
at least in some areas, led to a shift in the feeding ecology
0 50 100 150 200 250 300
0
10
20
30
40
Herring: a=0.134; b=0.58; r2=0.97
Sprat: a=0.08; b=3.10; r2=0.60
F1,179
= 26.4455; P < 0.0001
Body length, mm
Jaw
len
gth
, m
m
Figure 4. Relationships between mouth size measured as lower jaw
length (mm) and body length (BL, mm) in herring and sprat;
regression line: jaw lengthZ aBLC b. The difference between the
slopes is extremely significant.
Size classes, cm
Herring
Copepods
Cladocerans
Mysids
Cercopagis
Sprat
5-10 10-15
B
5-10 10-15 15-20 >20
0
25
50
75
100A
% o
f fo
od
b
io
mass
Figure 3. Diets of different size classes of herring (A) and sprat (B), expressed as percentages (based on dry weight) of different prey types
in their stomachs. Contribution of Cercopagis is shown separately from the rest of the cladocerans.
964 E. Gorokhova et al.
Herring
A
5-10 10-15 15-20 >200
15
30
45
60
Freq
uen
cy, %
Sprat
B
5-10 10-15
Mature
Immature
Size classes, cm
Figure 5. Proportions of fish stomachs that contained Cercopagis resting eggs and relative distribution of mature and immature eggs in
herring (A) and sprat (B) of different size.
of herring and sprat, the major zooplanktivores in the
Baltic. Moreover, this shift occurs not only in adult herring
(as suggested by Ojaveer et al., 1998), but also in sprat and
in fish as young as metamorphosed YOY herring.
Cercopagis pengoi is present in the northern Baltic proper
from July to October, being most abundant in AugusteSep-tember, i.e. at the same time of year as other zooplankton
decline. This decline also occurred prior to the invasion of
Cercopagis and is probably caused by factors such as
resource availability or fish predation (Johansson et al., 1993;
Adrian et al., 1999). Indeed, the consumption peak by fish
occurs in AugusteSeptember and the major zooplanktivores
are YOY clupeoids (Rudstam et al., 1992; Arrhenius and
Hansson, 1993). Therefore, as fish consumption increases
and densities of native zooplankton decline, Cercopagis
reaches its abundance peak and may become an important
food source for zooplanktivorous fish during this period.
Like most cladocerans, C. pengoi is a cyclic parthe-
nogen, and new population can be established from a single
egg. The primary mode of reproduction is clonal, in-
terspersed with periods ( primarily in autumn) of sexual
reproduction that involves formation of resting eggs
(Rivier, 1998). It has also been suggested that resting eggs
of Cercopagis can survive a passage through the fish
digestive system (Antsulevich and Valipakka, 2000), and
therefore fish could act as a dispersal vector. However, eggs
carried by Instar IeIII gametogenic females are in various
stages of development when consumed by fish, and this
may potentially affect the viability of those eggs following
passage through the gut. We found a tendency in sprat, but
not in herring, to ingest mostly females with immature eggs
(Figure 5) and many of these eggs appeared partially
digested. Because sprat has a smaller mouth compared to
that of herring (Figure 4), it is possible that sprat feeds
preferentially on smaller Instar IeII gametogenic females,
which are more likely to have immature eggs, while herring
selects larger Instar III females with dark, highly visible
mature eggs as shown for many planktivorous fish and
egg-carrying zooplankters. Jarnagin et al. (2000) showed
experimentally that fully mature resting eggs of Bytho-
trephes survived the passage through the alimentary canal
of yellow perch (Perca flavescens), while the hatching
success of immature eggs was decreased. The impact of fish
predation on the production of viable over-wintering eggs
may therefore be influenced by the species composition of
the zooplanktivorous fish and this may influence the
recruitment of Cercopagis the following year.
Once invasive species are established, one of the few
countermeasures that can be taken is to try to suppress the
invader by managing its predators. This approach was taken
in Lake Michigan, to control alewife and rainbow smelt
(Rand and Stewart, 1998). The two dominant zooplanktiv-
orous fish in the Baltic are herring and sprat, the
populations of which are strongly influenced by the fishery
and for which annual catch quotas are set by the
International Baltic Sea Fisheries Commission. Our study
shows that both these species are predators on Cercopagis
and this may need to be accounted for in the management
of the fishery, if we would like to reduce the abundance of
Cercopagis. This would then be an excellent example of an
application of the ecosystem-based approach to fisheries
management, which is a cornerstone in the recently adopted
common fisheries policy of the European Union (Anon.,
2002). It should be acknowledged, however, that it is not
fully understood to which extent herring, sprat and other
zooplanktivores, like for example smelt (Osmerus eperla-
nus), jellyfish, and mysid shrimps, actually control
Cercopagis through predation. Neither do we know if
Cercopagis actually competes with the fish for prey or
constitutes a new and important food web link to higher
trophic levels.
Acknowledgements
This research was supported by research grants from The
Swedish Research Council for Environment, Agricultural
Sciences and Spatial Planning (Formas) and Swedish
965Predation by herring and sprat on Cercopagis pengoi
Environmental Protection Agency. We thank B. Larsson for
his invaluable help in collecting fish for analyses. We also
thank M. Petersson (Ar Research Station, Gotland Univer-
sity, Sweden), L. Lundgren and B. Abrahamsson (Systems
Ecology, Stockholm University, Sweden), and Karolina
Eriksson-Gonzales (Linkoping University, Sweden) for
technical assistance and logistical support.
References
Adrian, R., Hansson, S., Sandin, B., DeStasio, B., and Larsson, U.1999. Effects of food availability and predation on a marinezooplankton community e a study on copepods in the BalticSea. International Review of Hydrobiology, 84: 609e626.
Anon. 2002. Council Regulation (EC) No 2371/2002 of 20December 2002 on the Conservation and Sustainable Exploita-tion of Fisheries Resources under the Common Fisheries Policy.Official Journal of the European Communities, 2371: 1e22.
Antsulevich, A., and Valipakka, P. 2000. Cercopagis pengoi e newimportant food object of the Baltic herring in the Gulf ofFinland. International Review of Hydrobiology, 85: 609e619.
Arrhenius, F. 1996. Diet composition and food selectivity of0-group herring (Clupea harengus L.) and sprat (Sprattussprattus L.) in the northern Baltic Sea. ICES Journal of MarineScience, 53: 701e712.
Arrhenius, F., and Hansson, S. 1993. Food consumption of larval,young and adult herring and sprat in the Baltic Sea. MarineEcology Progress Series, 96: 125e137.
Arrhenius, F., and Hansson, S. 1996. Growth and seasonal changesin energy content of young Baltic Sea herring (Clupea harengusL). ICES Journal of Marine Science, 53: 792e801.
Axenrot, T. 2002. Seasonal dynamics and recruitment predictionfor herring in the Baltic proper. Licentiate in Philosophy thesis,Department of Systems Ecology, Stockholm University, Sweden(available from Stockholm University Library). 24 pp.
Benoit, H. P., Johannsson, O. E., Warner, D. M., Sprules, W. G.,and Rudstam, L. G. 2002. Assessing the impact of a recentpredatory invader: the population dynamics, vertical distribution,and potential prey of Cercopagis pengoi in Lake Ontario.Limnology and Oceanography, 47: 626e635.
Bielecka, L., _Zmijewska, M. I., and Szymborska, A. 2000. Anew predatory cladoceran Cercopagis (Cercopagis) pengoi(Ostroumov 1891) in the Gulf of Gdansk. Oceanologia, 42:371e374.
Bushnoe, T. M., Warner, D. M., Rudstam, L. G., and Mills, E. L.2003. Cercopagis pengoi as a new prey item for alewife (Alosapseudoharengus) and rainbow smelt (Osmerus mordax) in LakeOntario. Journal of Great Lakes Research, 29: 205e212.
Chesson, J. 1983. The estimation and analysis of preference and itsrelationship to foraging models. Ecology, 64: 1297e1304.
Gorokhova, E., Aladin, N., and Dumont, H. 2000. Further rangeextensions of Cercopagis genus (Crustacea, Branchiopoda,Onychopoda) with notes on taxonomic composition and ecology.Hydrobiologia, 429: 207e218.
HELCOM. 1988. Guidelines for the Baltic Monitoring Programmefor the third stage. Baltic Sea Environment Proceedings, 27D, 161 pp.
Hernroth, L. 1985. Recommendations on methods for marinebiological studies in the Baltic Sea. Mesozooplankton biomassassessment. The Baltic Marine Biologists, Publ. 10, 32 pp.
Jarnagin, S. T., Swan, B. K., and Kerfoot, W. C. 2000. Fish asvectors in the dispersal of Bythotrephes cederstroemi: diapaus-ing eggs survive passage through the gut. Freshwater Biology,43: 579e589.
Johansson, S., Hansson, S., and Araya-Nunez, O. 1993. Temporaland spatial variation of coastal zooplankton in the Baltic Sea.Ecography, 16: 167e173.
Leppakoski, E., and Olenin, S. 2000. Non-native species and ratesof spread: lessons from the brackish Baltic Sea. BiologicalInvasions, 2: 151e163.
MacIsaac, H. J., Grigorovich, I. A., Hoyle, J. A., Yan, N. D., andPanov, V. E. 1999. Invasion of Lake Ontario by the Ponto-Caspian predatory cladoceran Cercopagis pengoi. CanadianJournal of Fisheries and Aquatic Sciences, 56: 1e5.
Mehner, T., and Heerkloss, R. 1994. Direct estimation of foodconsumption of juvenile fish in a shallow inlet of thesouthern Baltic. International Review of Hydrobiology, 79:295e304.
Ojaveer, H., and Lumberg, A. 1995. On the role of Cercopagispengoi (Ostroumov) in Parnu Bay and the NE part of the Gulf ofRiga ecosystem. Proceedings of Estonian Academy of Sciences.Ecology, 5: 20e25.
Ojaveer, E., Lumberg, A., and Ojaveer, H. 1998. Highlights ofzooplankton dynamics in Estonian waters (Baltic Sea). ICESJournal of Marine Science, 55: 748e755.
Ojaveer, H., Lankov, A., Eero, M., Kotta, J., Kotta, I., andLumberg, A. 1999. Changes in the ecosystem of the Gulf of Rigafrom the 1970s to the 1990s. ICES Journal of Marine Science,56: 33e40.
Rand, P. S., and Steward, D. J. 1998. Prey fish exploitation,salmonine production, and pelagic food web efficiency in LakeOntario. Canadian Journal of Fisheries and Aquatic Sciences, 55:318e327.
Rivier, I. K. 1998. The predatory Cladocera (Onychopoda:Podonidae, Polyphemidae, Cercopagidae) and Leptodorida ofthe world. In Ed. by H. Dumont. Guides to the Identification ofthe Micro-invertebrates of the Continental Waters of the World,vol. 13: Backhuys Publishers, Leiden, 213 pp.
Rosen, R. A. 1981. Lengthedry weight relationships of somefreshwater zooplankton. Journal of Freshwater Ecology, 1:225e229.
Rudstam, L. G., Hansson, S., Johansson, S., and Larsson, U. 1992.Dynamics of planktivory in a coastal area of the northern BalticSea. Marine Ecology Progress Series, 80: 159e173.
Therriault, T. W., Grigorovich, I. A., Kane, D. D., Haas, E. M.,Culver, D. A., and MacIsaac, H. J. 2002. Range expansion of theexotic zooplankter Cercopagis pengoi (Ostroumov) into westernLake Erie and Muskegon Lake. Journal of Great Lakes Research,28: 698e701.
Uitto, A., Gorokhova, E., and Valipakka, P. 1999. Distribution ofnon-indigenous Cercopagis pengoi in the coastal waters of theEastern Gulf of Finland. ICES Journal of Marine Science,56(Supplement): 49e57.
Upton, G. J. G. 1978. The Analysis of Cross-Tabulated Data. JohnWiley & Sons Ltd, Bristol, 148 pp.
Vanderploeg, H. A., Nalepa, T. F., Jude, D. J., Mills, E. L., Holeck,K. T., Liebig, J. R., Grigorovich, I. A., and Ojaveer, H. 2002.Dispersal and emerging ecological impacts of Ponto-Caspianspecies in the Laurentian Great Lakes. Canadian Journal ofFisheries and Aquatic Sciences, 59: 1209e1228.
Yan, N. D., Girard, R., and Boudreau, S. 2002. An introducedinvertebrate predator (Bythotrephes) reduces zooplankton spe-cies richness. Ecology Letters, 5: 481e485.