Content uploaded by Christer Brönmark
Author content
All content in this area was uploaded by Christer Brönmark on Sep 18, 2014
Content may be subject to copyright.
1023
Ecology,
82(4), 2001, pp. 1023–1039
q
2001 by the Ecological Society of America
THE INFLUENCE OF MULTIPLE INTRODUCED PREDATORS
ON A LITTORAL POND COMMUNITY
P
ER
N
YSTRO
¨
M
,O
LA
S
VENSSON
,B
JO
¨
RN
L
ARDNER
,C
HRISTER
B
RO
¨
NMARK
,
AND
W
ILHELM
G
RANE
´
LI
Department of Ecology, Ecology Building, Lund University, S-223 62 Lund, Sweden
Abstract.
In a replicated field experiment we studied the effects of natural densities
of two exotic consumers, the predatory and herbivorous signal crayfish (
Pacifastacus len-
iusculus
) and the predatory rainbow trout (
Oncorhynchus mykiss
), on multiple trophic levels
of a pond community. The goals were to: (1) determine the individual and combined effects
of predators on macroinvertebrates, macrophytes, and periphytic algae; (2) evaluate the
strength of direct and indirect interactions in a food web influenced by omnivores; and (3)
evaluate the relative importance of direct and indirect predator effects on mortality and
growth of a native frog species,
Rana temporaria.
The experiment showed that both signal crayfish and rainbow trout had strong effects
on multitrophic levels of a littoral pond community, through direct consumption and indirect
effects on lower trophic levels. Crayfish had weak but significant negative effects on the
biomass of predatory invertebrates and greatly reduced the biomass of snails, the most
abundant invertebrate grazers. Although the number of active herbivorous tadpoles tended
to be higher in crayfish cages relative to control cages, the proportion of surviving froglets
was lower in crayfish cages than in control cages, possibly due to crayfish predation on
injured tadpoles. The size of surviving froglets did not differ from controls, but tadpoles
in crayfish cages often suffered tail injuries. Macrophyte coverage decreased as a result of
crayfish consumption and nonconsumptive fragmentation. However, the biomass of pe-
riphyton increased in crayfish cages relative to controls, probably due to reduced grazing
from snails. In contrast, trout had strong negative effects on the biomass of both predatory
invertebrates and insect grazers, whereas trout had less effect on snail biomass than did
crayfish. Also, in contrast to crayfish cages, the number of active tadpoles in trout cages
was lower than in controls, probably due to a combination of trout predation and trout-
induced reduced tadpole activity. Trout had a strong negative impact on froglet survival,
and froglets in trout cages metamorphosed at a smaller size and had reduced growth rates
compared to froglets in crayfish and control cages. As with crayfish, the biomass of pe-
riphyton increased in trout cages relative to controls, which may be due to a combination
of both density and trait-mediated trout effects on tadpole grazing.
In treatments with multiple predators the effects of crayfish and trout on caged com-
munities were independent, and there were few interactions. Mostly effects of combined
predators reflected those in single predator cages. Our results demonstrate that noninter-
acting, introduced multiple predators can have strong direct and indirect effects on multiple
trophic levels in pond communities. Trophic cascades may develop in aquatic food webs
even with omnivores such as crayfish, and in complex habitats with trout. These strong
indirect effects are mediated through both predation on important grazers (i.e., the crayfish–
snail–periphyton link) and a combination of density and behavioral responses of grazers
to predators (i.e., the trout–tadpole–periphyton link). When two noninteracting predators
have strong but different effects on prey survival or activity, their combined effects on
intermediate trophic levels reflect responses to the more dangerous predator.
Key words: exotic species; freshwater pond; grazers; growth rate; littoral pond; omnivory;
Oncorhynchus mykiss; Pacifastacus leniusculus;
predators, multiple;
Rana temporaria;
Sweden; tro-
phic cascade.
I
NTRODUCTION
Predation is one of the major factors influencing prey
population dynamics and community structure (e.g.,
Sih 1987, Lima and Dill 1990). In the joint evolution
of prey and predators, prey evolve morphological struc-
tures, chemical repellents, crypsis, and antipredator be-
Manuscript received 23 July 1999; revised 6 January 2000;
accepted 16 February 2000; final version received 21 March 2000.
haviors that reduce predation risk (e.g., Lima and Dill
1990, Endler 1991). These adaptations affect the ability
of predators to regulate prey abundance (Power 1992
a,
Polis and Strong 1996, Scheffer 1997). However, in-
troduced predators may have particularly strong effects
on native prey that do not recognize new predators, do
not show appropriate avoidance behaviors (Shave et al.
1994, Kiesecker and Blaustein 1997), or if the intro-
duced predators have foraging strategies that differ
from those of native predators (McIntosh and Town-
1024
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
send 1996). Subsequently, introduced predators can po-
tentially affect the distribution and abundance of or-
ganisms at multiple trophic levels in a community (e.g.,
Schoener and Spiller 1999).
In community ecology, both theoretical (e.g., Hair-
ston et al. 1960, Oksanen et al. 1981, Fretwell 1987)
and experimental work has traditionally focused on the
direct and indirect effects of predators in tightly linked
food chains with distinct trophic levels. For example,
several studies have revealed the importance of pred-
ator-induced trophic cascades (e.g., Spiller and Schoe-
ner 1990, Bro¨nmark 1994, Letourneau and Dyer 1998),
and studies of introduced predators have further in-
creased our knowledge of how trophic cascades may
develop (e.g., McIntosh and Townsend 1996, Schoener
and Spiller 1999, Nystro¨m and A
˚
bjo¨rnsson 2000).
However, the applicability of the traditional linear food
chain theory to species interactions in natural food
webs has recently been questioned, because many com-
munities have multiple interacting predators and high
connectance due to the prevalence of omnivory (e.g.,
Polis and Strong 1996, Rosenheim 1998). Further, the
mechanism explaining trophic cascades in food chains
has traditionally been viewed as a result of prey re-
moval, but recent studies have shown that cascading
effects can also be behaviorally transmitted (Lima
1998). Examples from both terrestrial (Schmitz et al.
1997, Gastreich 1999) and freshwater systems (Mc-
Intosh and Townsend 1996, Turner 1997, McCollum et
al. 1998, Peckarsky and McIntosh 1998) suggest that
if herbivores shift their foraging behavior in response
to predation risk, basal resources may increase.
Studies of introduced and native predators typically
examine the effects of one predator at a time. Many
communities, however, have been invaded by multiple
predators (Ross 1991, Elvira et al. 1996, Gamradt and
Kats 1996, Kiesecker and Blaustein 1998). Results
from experimental studies of multiple native predators
suggest the combined effects can be very strong, al-
though quite complex due to a combination of direct
and indirect trophic effects (Peckarsky and McIntosh
1998), the prevalence of omnivory (the consumption
of resources from more than one trophic level; Diehl
1993, 1995, Pringle and Hamazaki 1998), and inter-
actions among predators (Sih et al. 1998). Therefore,
to understand the effects of multiple predatorinvasions,
we need to consider all predator types and how they
interact directly and indirectly with other trophic levels
in the community.
To predict the effects of multiple predators on the
structure of natural communities, the traditional ap-
proach that examines interactions between individual
prey and predator species may not be appropriate (see
review in Sih et al. 1998). That approach assumes pred-
ators do not interact in their effects on prey, i.e., that
interaction modifications are absent (Billick and Case
1994, Wootton 1994). Indeed, work in both terrestrial
(e.g, Spiller and Schoener 1994) and freshwater sys-
tems (Pringle and Hamazaki 1998, Van Buskirk 1988)
have revealed that multiple predators may have non-
interactive effects on prey populations. However, other
studies have shown that multiple predators often have
complex effects that can not be predicted simply by
observing interactions between pairs of species in iso-
lation (Sih et al. 1998). Most often, the combined ef-
fects of predators on prey are lower than expected from
their individual effects, due to interactions among pred-
ators (Soluk and Collins 1988, Soluk 1993, Rosenheim
1998). Less commonly, the combined predation rate of
predators is higher than expected from their individual
effects (Sih et al. 1998). Facilitation among multiple
predators can occur when antipredator responses to one
predator increase risk of predation to another predator
(Soluk and Collins 1988, Losey and Denno 1998, Kie-
secker and Blaustein 1998).
Because many communities have more than one in-
troduced predator species, the resulting effects on these
invaded communities may be complex and hard to pre-
dict, but potentially very strong (e.g., Kiesecker and
Blaustein 1998). The mechanisms behind the strong
effects of exotic predators are rarely determined, but
may provide results that can be successfully integrated
into models of community ecology (Vitousek 1990,
Lodge et al. 1998).
In this study, we examined the impact of two exotic
predators, the signal crayfish (
Pacifastacus leniuscu-
lus
) and the rainbow trout (
Oncorhynchus mykiss
)on
trophic interactions in a natural pond community. The
signal crayfish is both a herbivore and a predator where-
as rainbow trout is a predator on different consumer
trophic levels. The goals of this study were to: (1)
determine the individual and combined effects of these
two commonly introduced exotic predators on multiple
trophic levels in a pond; (2) evaluate the strength of
direct and indirect interactions in a food web influenced
by omnivores; and (3) evaluate the relative importance
of direct and indirect predator effects on mortality and
growth of a native frog species,
Rana temporaria.
The introduced predators
The signal crayfish, native to northwestern United
States, is resistant to the fungus that causes crayfish
plague, and was therefore imported to Sweden in 1960
as a replacement for the susceptible native noble cray-
fish (
Astacus astacus
) (Lowery and Holdich 1988). The
signal crayfish is an omnivore that feeds on inverte-
brates, macrophytes, periphyton, and detritus(Nystro¨m
et al. 1996, Lira˚s et al. 1998).
Rainbow trout, native to the Pacific coast of North
America, was introduced to Sweden and Europe at the
end of the 19th century. Although rainbow trout are
often the top predators in lakes and ponds where they
are introduced, their feeding behavior in lakes and
ponds in Europe is not well studied. Data from British
lakes suggest that they eat primarily large active prey
such as amphipods and water bugs, whereas insect
April 2001 1025
MULTIPLE INTRODUCED PREDATORS
F
IG
. 1. Hypothesized food-web links, in the littoral pond
(A) without introduced predators, (B) with signal crayfish,
(C) with rainbow trout, and (D) with both crayfish and trout
assuming these predators have independent effects. Arrow
thickness indicates expected interaction strengths. Solid lines
indicate consumption, and dashed lines indicate competition.
grazers and snails are consumed to a lesser extent (Hunt
and O’Hara 1973). Similarly, data from
;
200 fishes
caught by anglers in three Swedish lakes, show that
rainbow trout feed on the immature and adult stages
of predatory invertebrates (e.g., Odonata, Heteroptera,
and Coleoptera), and on aquatic insect grazers (e.g.,
Ephemeroptera and Trichoptera) (Lindqvist 1981).
Many ponds in Sweden are devoid of both native
crayfish and predatory fish. In addition, rainbow trout
are often introduced, even though they are predators
on native and introduced crayfish species (Foster and
Slater 1995; P. Nystro¨m,
unpublished data
).
The natural system
Permanent fishless habitats often have communities
composed of large predatory invertebrates and large
moderately active prey (Wellborn et al. 1996). In per-
manent ponds in southern Sweden without predatory
fish and crayfish, predatory invertebrates such as bugs,
beetles, and leeches are top predators (Nystro¨m et al.
1996, Wagner 1997, Fig. 1A), and the grazer com-
munity is dominated by large thin-shelled pulmonate
snail species (Bro¨nmark 1994, Bro¨nmark and Weisner
1996). Predatory invertebrates typically have weak ef-
fects on these grazers (Bro¨nmark 1992, 1994), and
large pulmonate snail species efficiently regulate pe-
riphyton biomass (see review in Bro¨nmark 1989). How-
ever, insect grazers mostly have a negligible impact on
periphyton biomass (Bro¨nmark 1994, Bro¨nmark and
Vermaat 1998, Fig. 1A).
During spring and summer, up to seven amphibian
species breed in these ponds, including both widely
distributed species such as the common frog
Rana tem-
poraria
(Loman 1988) and more rare species (e.g., the
European tree frog,
Hyla arborea,
Bro¨nmark and Ed-
enhamn 1994). Experimental studies have shown that
common frog larvae have effects on periphyton com-
parable to those of snails (Bro¨nmark et al. 1991, Fig.
1A). In many ponds predatory invertebrates cause sig-
nificant mortality of amphibian larvae (e.g., Smith
1983, Wellborn et al. 1996, Fig. 1A).
Predicted impact of the introduced predators
Crayfish typically reduce the biomass of slow-mov-
ing benthic invertebrates, particularly snails, but have
less impact on mobile nonmolluscan invertebrates
(Lodge et al. 1994, Nystro¨m et al. 1996, Perry et al.
1997, Nystro¨m et al. 1999, Fig. 1B). Also, crayfish can
have strong direct impacts on primary producers by
grazing macroalgae (Hart 1992, Creed 1994), and by
grazing and nonconsumptive destruction of macro-
phytes (e.g., Lodge et al. 1994, Nystro¨m and Strand
1996, Fig. 1B). Effects of crayfish on amphibian larvae
are not well known, but experimental studies suggest
that while some crayfish species, including signal cray-
fish, may consume or injure tadpoles (e.g., Axelsson
et al. 1997, Nystro¨m and A
˚
bjo¨rnsson 2000), crayfish
are generally weak predators on tadpoles (Holomuzki
1989, Fauth 1990, Lefcort 1996). Periphytic algae are
also consumed by signal crayfish (Flint and Goldman
1975), but crayfish are less efficient grazers on mi-
croalgae than are snails (Luttenton et al. 1998, Nystro¨m
et al. 1999, Fig. 1B). In fact macrophytes may indi-
rectly benefit from snail removal of shading periphyton
(Bro¨nmark 1989, 1994, Bro¨nmark and Vermaat 1998,
Fig. 1B).
1026
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
In its native habitats, predation by the visually hunt-
ing rainbow trout has been shown to affect large active
prey such as predatory invertebrates, herbivorous tad-
poles, stoneflies, and amphipods (Ware 1973, Cooper
1988, Feltmate and Williams 1989). Moreover, intro-
duced trout species usually have strong negative im-
pacts on amphibian larvae (e.g., Macan 1966, Bradford
et al. 1998) and weak effects on snails (Macan 1966,
Fig. 1C).
In a field enclosure experiment, we tested the single
and combined effects of the introduced signal crayfish
and rainbow trout on the interactions in a littoral pond
community. Our design enabled us to test whether cray-
fish and trout had interactive or noninteractive effects
on pond communities (Fig. 1D). Interactive effects
were expected if; (1) trout reduce crayfish effects by
predation or by altering crayfish foraging behavior
(e.g., Stein and Magnuson 1976, Hill and Lodge 1995);
(2) a change in behavior of tadpoles to one predator
affects predation risk to the other predator (e.g., Morin
1995, Peacor and Werner 1997, Kiesecker and Blau-
stein 1998); or (3) crayfish reduce the biomass of mac-
rophytes, facilitating fish effects on predatory inver-
tebrates, tadpoles, and insect grazers (Crowder and
Cooper 1982, Diehl 1992). Finally, reduction of large
grazers (i.e., snails or tadpoles) was expected to have
strong positive effects on the biomass of periphyton
(i.e., a trophic cascade).
M
ETHODS
Field experiment
A field experiment was carried out in a shallow eu-
trophic pond in southern Sweden (13
8
50
9
E, 55
8
50
9
N),
with no fish or crayfish (area, 2500 m
2
; maximum
depth, 2 m) from 9 May to 14 July 1997. During sum-
mer, water is alkaline (pH
;
8, alkalinity
5
2 mmol/L,
total phosphorus of 34
m
g/L and total nitrogen of 0.9
mg/l). Pond sediment was covered by submerged mac-
rophytes, especially
Elodea canadensis.
Emergent veg-
etation was dominated by
Carex rostrata
and
Glyceria
fluitans.
We set up 20 cages (6 m
2
,2
3
3 m) in the pond in
early April, each including natural emergent vegetation
and a terrestrial zone of the shoreline. Cobbles and
stones were abundant in the shallow part of enclosures,
providing a suitable habitat for crayfish (Lodge and
Hill 1994). Maximum cage depth was
;
1.5 m. Each
cage was constructed with nylon-reinforced plastic
walls attached to wooden frames. At the sediment sur-
face, 20 cm of the plastic was folded into the cage, and
covered with gravel and sediments to create a tight seal
against the sediment. Two triangular mesh windows
(0.5
3
1.3 m, mesh size: 1.8
3
3.2 mm) in each of the
longer walls allowed water circulation and immigra-
tion/emigration of smaller invertebrates (but not tad-
poles). To prevent disturbance from birds but not in-
sects, cages were covered with a net (mesh-size 20 mm)
(Nystro¨m et al. 1999). Four treatments were randomly
assigned to the 20 cages, each replicated five times:
control (no crayfish and no trout, C), crayfish (CR),
trout (T), and crayfish
1
trout (CR
1
T). When the
experiment was terminated, cages were removed and
predators released into the pond.
Crayfish were collected from nearby ponds and add-
ed to cages on 16 May 1997, except for young-of-the-
year crayfish that were stocked on 6 June, when hatch-
ing normally occurs in south Swedish ponds. Natural
densities of signal crayfish are similar to those of the
native noble crayfish (Anonymous 1993), and densities
(12 crayfish/m
2
) and size distributions of crayfish used
in this experiment mimicked those found for noble
crayfish in the littoral zone of Swedish lakes (Appel-
berg and Odelstro¨m 1986). In each crayfish cage, we
stocked ten 3-yr-old crayfish (carapace length, 48.8
6
0.4 mm [mean
6
1
SE
];
n
5
100), eight 2-yr-oldcrayfish
(40.4
6
0.6 mm,
n
5
80), 17 1-yr-old crayfish (18.8
6
0.3 mm,
n
5
170) and 37 hatchlings. Five dead adult
crayfish were found during the experiment and were
replaced by crayfish of similar sizes. On 24 July, we
placed one cylindrical trap (mesh 15 mm) baited with
roach (
Rutilus rutilus
) in each cage overnight to esti-
mate the relative abundance of adult crayfish at the end
of the experiment. On average, 5.0
6
0.9 adult crayfish
were caught in each cage, and the number of crayfish
caught did not differ between crayfish and crayfish
1
trout cages (ANOVA,
F
1,8
5
0.57,
P
5
0.47).
Rainbow trout with an average wet mass of 90
6
7
g (mean
6
1
SE
,
;
18 cm, estimated from five randomly
chosen fish) were also introduced into cages on 16 May.
According to Swedish legislation, trout had to originate
from an approved hatchery in order not to spread dis-
eases. Because no data exist on densities of rainbow
trout in Swedish ponds and lakes, we placed two fish
in each cage, a density corresponding to that of pred-
atory perch (
Perca fluviatilis
), in nearby lakes (Persson
1986). This density is also well within the ranges of
rainbow trout in southern Ontario streams (Feltmate
and Williams 1989), and brown trout in southern Swed-
ish streams (Eklo¨v 1996). Three trout were replaced
during the experiment, since they were caught in the
nets covering the enclosures. When the experiment was
terminated, all fish were alive.
In Europe, the common frog (
Rana temporaria
)
breeds in a wide range of habitats from temporary to
permanent ponds and lakes with predatory fish (Beebee
1985, Marnell 1998). Although the common frog does
not avoid ovipositing in habitats with fish, fish intro-
duction may cause local population declines (Meyer et
al. 1998). In Sweden the common frog breeds in ponds
with dense populations of signal crayfish as well as in
ponds with both rainbow trout and signal crayfish (Nils-
son 1999; P. Nystro¨m,
personal observations
). Fifteen
Rana
egg masses were collected from a nearby pond
and placed in a plastic wading pool (4.5 m
2
) near the
experimental pond on 10 April. Hatching occurred on
April 2001 1027
MULTIPLE INTRODUCED PREDATORS
29 April and on 12 May; 600 tadpoles (100 tadpoles/
m
2
, stages: 26–27 according to Gosner 1960) were ran-
domly chosen from the pool and stocked in each cage.
This initial density is within the range of natural tad-
pole densities of common frogs found in most ponds,
including those with signal crayfish and rainbow trout
(B. Lardner,
unpublished data
).
Sampling
Macroinvertebrates
.—We destructively sampled
macroinvertebrates inside cages on three occasions:
once before predator stocking (9–14 May) and twice
after predator stocking (16–18 June and 1–7 July). To
ensure efficient sampling of macroinvertebrates with
different behaviors and activity periods we combined
three sampling methods. Macroinvertebrates on cage
walls were sampled by scraping a 0.6-m
2
area on each
of the longer walls with a small hand net (Lodge et al.
1994; size, 20
3
10 cm; mesh, 300
m
m). Macroinver-
tebrates associated with macrophytes or appearing in
the water column were sampled by placing a metal
frame (1
3
1.4 m) in the center of the cage. Using the
same hand net we collected invertebrates by seven stan-
dard sweeps (during 1 min) through the whole water
column within the frame area. In order to collect in-
vertebrates living in or on the sediment surface as well
as invertebrates active during hours of darkness, we
used active-fauna traps. The trap is a modified plastic
beaker with a cone-shaped net funnel (1.0-mm mesh
size) welded onto the opened bottom end. At the top
end, a removable lid is furnished with an identical plas-
tic net (Wagner 1997). These traps are particularly use-
ful for collecting invertebrates such as bugs and beetles
that otherwise may be missed using conventional sam-
pling methods (e.g., nets and corers). Traps were placed
at the sediment surface at five locations within each
cage. Trapped invertebrates were collected after 48 h.
All collected invertebrates were frozen for later iden-
tification. After thawing, invertebrates were counted
and their total length measured to the nearest 0.1 mm.
Dry mass (shell-free dry mass for snails) was deter-
mined for all invertebrates using length–mass regres-
sions (P. Nystro¨m,
unpublished data
). Macroinverte-
brate biomass from the three sampling methods was
pooled, giving invertebrate biomass per cage (Lodge
et al. 1994).
Tadpoles
.—Relative abundance and tadpole activity
were estimated nondestructively using the same active-
fauna traps as for invertebrates. Five traps were placed
at the sediment surface of each cage on three occasions:
once before predator stocking (14 May, 2 d after tad-
poles were stocked) and twice after predators had been
stocked (27 May and 16 June). After 48 h we counted
the number of tadpoles in each trap, recorded tail in-
juries and then released the tadpoles in the shallow part
of the cages.
Froglets
.—Starting in mid-June we checked cages
daily for metamorphs when tadpoles had reached larval
stage 40 (according to Gosner 1960) after which fore-
limbs appear (froglets). We counted and collected fro-
glets daily (above the water line, or swimming without
using their tail), recorded date of collection, and de-
termined their size at metamorphosis (snout–vent
length) to the nearest 0.1 mm. Froglets were then re-
leased in a nearby pond. Response variables for each
enclosure included percentage of stocked individuals
that metamorphosed, mean size at metamorphosis,
mean duration of the larval period (number of days
between stocking and metamorphosis), and mean
growth rate (froglet size divided by larval period). Af-
ter 14 July we observed no froglets in any of the cages,
and thus assumed metamorphosis to be completed.
Macrophytes
.—Macrophyte coverage was estimated
within the metal frame where invertebrates were sweep
netted. Coverage was expressed as percentage total area
(large stones excluded). The nonconsumptive destruc-
tion of macrophytes by crayfish was assessed weekly
by netting live macrophyte fragments floating on the
water surface within each cage (Lodge et al. 1994).
Collected plants were dried (65
8
C, until constant mass)
and weighed.
Periphytic algae
.—We used two different substrates
to evaluate the effects of treatments on algae. In each
cage we put nine glazed ceramic tiles (10
3
10 cm)
on the sediment surface (0.5 m depth), and nine poly-
ethylene strips (1.6
3
50 cm), which were anchored to
the sediment at one end and kept floating by a piece
of cork nailed to the other end. Tiles may be grazed
by all grazers, whereas strips were inaccessible to sig-
nal crayfish (Nystro¨m et al. 1999). To allow algal col-
onization of substrates before the experiment, sub-
strates were placed in the pond on 13 March and then
transferred to the cages in early April. On the same
dates as macrophytes and invertebrates were sampled,
three tiles and three strips were randomly collected
from each cage and frozen. Plastic strips were pre-
served in vials, whereas periphyton on tiles was
scraped off with a razor blade and then transferred to
vials. Chlorophyll
a
per unit surface area was then
measured spectrophotometrically after extraction with
ethanol (Jespersen and Christoffersen 1987).
Statistical analyses
Statistical analyses of response variables were per-
formed on the means from each cage. In most cases a
two-way factorial design with repeated measures was
used to examine effects of crayfish, trout, and the in-
teractions between these two groups (e.g., Fauth 1990,
Pringle and Hamazaki 1998). Prior to analysis, all data
used in two-way MANOVAs or ANOVAs were natural
log- or log(
x
1
1)-transformed to normalize distribu-
tions and stabilize variances. The use of log-trans-
formed data also changes the underlying model of mul-
tiple predator effects from an additive model to a mul-
tiplicative risk model. The multiplicative model is more
appropriate to apply to experimental data on prey bio-
1028
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
T
ABLE
1. Two-way, repeated-measures MANOVA for effects of crayfish, trout, and their
interactive effects on communities in cages (predatory invertebrates, snails, insect grazers,
tadpoles, periphyton on tiles and strips, and macrophyte coverage) after predator introduction
in June and July (Time).
Source df Wilks’ lambda
FP
Between cages
Crayfish
Trout
Crayfish
3
Trout
7, 10
7, 10
7, 10
0.1566
0.0456
0.4792
7.6907
29.8749
1.5521
0.0023
,
0.0001
0.2548
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
7, 10
7, 10
7, 10
7, 10
0.0057
0.1143
0.0226
0.4829
247.5448
11.0656
61.8320
1.5299
,
0.0001
0.0005
,
0.0001
0.2614
Note:
We assessed tadpoles by using the number caught in traps after predator introduction
on 27 May and 16 June 1997.
mass and survivorship measured over time, particularly
when predators have strong effects on prey survival
(Wootton 1994, Sih et al. 1998). Percentage data used
in one-way ANOVAs were arcsine-transformed prior
to analysis.
The analysis of predator effects on community struc-
ture was complicated by the occurrence of several re-
sponse variables, which most likely were correlated,
and which were sampled repeatedly over time. Multiple
univariate analyses for a large number of species made
in the same community increase the risk of type I error
(e.g., Morin 1987). Therefore, we first performed a
two-way MANOVA with time as a repeated measure
to test for overall effects of crayfish, trout, and both
predators on multiple response variables in enclosures
(the dependent variables in the MANOVA were the
biomass of predatory invertebrates, snails, insect graz-
ers, number of trapped tadpoles, macrophyte coverage,
and periphyton biomass on strips and tiles). Finally, to
clarify the interpretation of the MANOVA, two-way,
repeated-measures ANOVAs for each dependent var-
iable tested the effects of crayfish, trout, and their com-
bined effect over time.
Because very few froglets survived in cages with
crayfish
1
trout, this treatment was omitted from cal-
culations of tadpole size at metamorphosis, growth rate,
and length of the larval period. Instead, the effect of
each predator on the performance of froglets was tested
by a one-way MANOVA. Individual ANOVAs of each
variable then clarified the interpretation of the MAN-
OVA, and Tukey’s post hoc test revealed differences
between control, crayfish, and trout cages. Because
growth rate among tadpoles may be density dependent,
we compared growth rate in control, crayfish, and trout
cages including froglet survival as a covariate (AN-
COVA).
R
ESULTS
Because enclosures were installed in a natural pond
community, we tested if initial conditions in May were
significantly different between cages assigned to dif-
ferent treatments, using a one-way MANOVA.Because
neither MANOVA (
F
21,29
5
0.56,
P
5
0.91) nor any
individual ANOVAs showed significant differences be-
tween cages assigned to different treatments (
P
.
0.09
in all cases), differences between response variables on
the subsequent sampling dates (June and July for in-
vertebrates and primary producers and May and June
for tadpoles) were assumed to reflect effects of the
introduced predators. Therefore, only the samplings in
June and July (May and June for tadpoles) were in-
cluded in the analysis of treatment effects over time.
MANOVA indicated that both crayfish and trout had
significant effects on enclosed communities over time
(Table 1). Furthermore, crayfish and trout effects on
caged communities were independent indicated by non-
significant crayfish-by-trout interactions (Table 1). Be-
low we interpret these results using two-way, repeated-
measures ANOVAs by examining the effects of cray-
fish, trout, and their combined effect on each of the
dependent variables when samples taken after predator
stocking in June and July are included.
Macroinvertebrates
Before predators were added, macroinvertebrate bio-
mass in the pond (results from all 20 cages pooled)
was dominated by predatory heteropterans such as No-
tonectidae (29%), Naucoridae (21%), and Corixidae
(2%). Dystiscidae constituted 28% of the total biomass.
Macroinvertebrate grazers made up 16% of the total
biomass, and were dominated by thin-shelled pulmo-
nate snails (
Lymnaea
spp., 60%) and the mayfly
Cloeon
sp. (28%).
Trout dramatically reduced the biomass of predatory
macroinvertebrates, whereas crayfish had a less strong
yet significant effect (Table 2, Fig. 2A). Predatory ma-
croinvertebrates were almost eliminated in cages with
trout present, and in two of the crayfish
1
trout cages
no predatory macroinvertebrates were collected on the
last sampling date. However, the effects of crayfish and
trout on the biomass of predatory invertebrates were
independent of each other (no significant interaction,
April 2001 1029
MULTIPLE INTRODUCED PREDATORS
T
ABLE
2. Two-way, repeated-measures ANOVAs for effects of crayfish, trout, and their in-
teractive effects on the biomass of (A) predatory invertebrates (B) snails, and (C) insect
grazers after predator introduction in June and July 1997 (Time).
Source df
MS
FP
A) Predatory invertebrates
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
23.9085
63.4731
6.1686
2.1793
10.9705
29.1249
2.8305
0.0044
,
0.0001
0.1119
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
1
1
1
1
16
5.4870
0.7813
14.2803
0.0026
1.0325
5.3143
0.7567
13.8307
0.0025
0.0349
0.3972
0.0019
0.9604
B) Snails
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
155.5614
11.5948
0.0750
68.1280
5.0779
0.0329
,
0.0001
0.0386
0.8584
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
1
1
1
1
16
0.0039
0.1043
0.0001
0.4329
0.2546
0.0152
0.4096
0.0004
1.7002
0.9034
0.5312
0.9837
0.2107
C) Insect grazers
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
0.1090
25.0506
0.3951
0.4630
0.2354
54.0991
0.8532
0.6341
,
0.0001
0.3694
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
1
1
1
1
16
3.2877
0.1866
1.9334
0.0155
0.4944
6.6504
0.3774
3.9109
0.0314
0.0202
0.5476
0.0655
0.8615
Table 2). The biomass of snails was greatly reduced by
crayfish (Table 2, Fig. 2B), whereas crayfish did not
significantly reduce insect grazers (mainly the summer
generation of
Cloeon
sp., Fig. 2C). In contrast to cray-
fish, trout had a much weaker, though significant effect
on the biomass of snails, but had strong effects on
insect grazers (Table 2, Fig. 2B, C). Similarly, there
were no significant crayfish-by-trout interactions (Ta-
ble 2) indicating that effects of each predator were
independent. Generally, the effect of combined pred-
ators (crayfish
1
trout) reflected the most dangerous
predator in each case (trout for predatory invertebrates
and insect grazers, crayfish for snails).
Tadpoles and froglets
The number of tadpoles caught in active-fauna traps
was significantly reduced by trout (Table 3, Fig. 3). In
crayfish cages, more tadpoles tended to be caught com-
pared to controls, but the lowest number of tadpoles
was caught in crayfish
1
trout cages. Thus there was
a significant crayfish-by-trout interaction. Nonetheless,
the effect of combined predators was similar to the trout
effects (Fig. 3). In trout, and crayfish
1
trout cages,
so few tadpoles were collected that we could not in-
clude them in the analysis of tadpole tail injury. On
the last sampling date, however, the proportion of tad-
poles with tail injuries was significantly higher in the
crayfish cages (18
6
3.3%, mean
6
1
SE
) compared to
control cages (0.9
6
0.1%, one-way ANOVA,
F
1,8
5
25.80,
P
5
0.0010).
In total 725 froglets survived through metamorpho-
sis that occurred from 22 June to 14 July. A two-way
ANOVA indicated that the proportion of froglets meta-
morphosing in control cages (15.5
6
2.9%) was sig-
nificantly higher than in crayfish (5.4
6
1.6%,
F
1,16
5
14.47,
P
5
0.0016) and trout cages (3.1
6
1.2%,
F
1,16
5
27.39,
P
,
0.0001). The lowest number of surviving
froglets was found in crayfish
1
trout cages (0.2
6
0.1%). In two of the crayfish
1
trout cages no indi-
viduals metamorphosed. Even though the effects of
crayfish and trout individually on froglet survival were
strong, a marginally nonsignificant crayfish-by-trout
1030
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
F
IG
. 2. Total biomass of (A) predatory invertebrates, (B)
snails, and (C) insect grazers (means
6
1
SE
) over time sam-
pled in control cages (C), crayfish cages (CR), trout cages
(T), and in crayfish
1
trout cages (CR
1
T). Note thedifferent
scales on the
y
-axes (statistics in Table 2).
interaction indicated that these predator effects were
independent (two-way ANOVA, crayfish-by-trout in-
teraction;
F
1,16
5
4.11,
P
5
0.0596).
Due to low survival in crayfish
1
trout cages (six
froglets in total), this treatment was omitted from anal-
yses of size at metamorphosis, growth rate, and length
of the larval period. A one-way MANOVA showed that
overall froglet performance differed among control,
crayfish, and trout cages (Table 4). Individual ANOVAs
showed that mean length of the larval period did not
contribute to the significance of the MANOVA (54
6
0.5 d in control cages, 52
6
1.1 d in crayfish cages,
and 53
6
1.0 d in trout cages [all mean
6
1
SE
). How-
ever, both size at metamorphosis and growth rate dif-
fered between treatments (Table 4). Mean size of fro-
glets in control cages was similar to that in crayfish
cages, but froglets in trout cages were significantly
smaller than in both control and crayfish cages (Table
4, Fig. 4). To test if differences in growth rate of tad-
poles persisted after controlling for variation in final
froglet density, we analyzed growth rates using AN-
COVA, with number of surviving froglets as covariate
in the analysis. The ANCOVAshowed that growthrates
had statistically indistinguishable slopes (treatment-by-
survivor interaction:
F
2,9
5
1.36,
P
5
0.30); therefore,
we tested if population growth rates differed between
treatments by omitting the interaction term in the next
analysis. With increasing number of survivors, growth
rate declined linearly (ANCOVA,
F
1,9
5
8.58,
P
5
0.0168, Fig. 5), but the intercepts differed between
treatments (ANCOVA,
F
2,11
5
23.04,
P
5
0.0001, Fig.
5). Growth rate was significantly lower in trout cages
in comparison with control (Tukey’s post hoc test,
P
5
0.0017) and crayfish cages (Tukey’s post hoc test,
P
5
0.0003), but there was no difference between cray-
fish and control cages (Tukey’s post hoc test,
P
5
0.80).
Macrophytes
Crayfish significantly reduced macrophyte coverage,
whereas trout had no impact on macrophytes (Table 5,
Fig. 6A). Due to crayfish feeding activity, floating mac-
rophyte parts (mainly live stems of
Glyceria fluitans
)
accumulated in cages with crayfish and in crayfish
1
trout cages (Fig. 6B). The variation in macrophyte cov-
erage and the biomass of floating macrophytes reflected
the effect of crayfish, as indicated by the nonsignificant
crayfish-by-trout interactions (Table 5, Fig. 6).
Periphyton
The biomass of periphyton on strips was strongly
dependent on treatment. Chlorophyll
a
increased in
both crayfish and trout cages. Again, there was no sig-
nificant crayfish-by-trout interaction, but the combined
effect of crayfish
1
trout was higher than each predator
separately (Table 6, Fig. 7A). On the June sampling
date the periphyton biomass on tiles followed the same
pattern as biomass on strips. However, the chlorophyll
a
levels in all predator treatments conveyed by the end
of the experiment, resulting in a significant crayfish-
by-trout interaction (Table 6, Fig. 7B).
D
ISCUSSION
This study shows that both rainbow trout and signal
crayfish have strong multitrophic level effects on ben-
thic communities, through direct consumption and in-
direct effects on lower trophic levels. In most cases,
predators had independent effects on prey (e.g., Fig.
1D). Signal crayfish had a strong impact on snails and
April 2001 1031
MULTIPLE INTRODUCED PREDATORS
T
ABLE
3. Two-way repeated-measures ANOVAs for effects of crayfish, trout, and their in-
teractive effects on the abundance of tadpoles caught in active-fauna traps after predator
introduction in May and June 1997 (Time).
Source df
MS
FP
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
0.4738
7.1030
1.2896
0.1486
3.1885
47.8030
8.6787
0.0931
,
0.0001
0.0095
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
1
1
1
1
16
0.0288
0.2017
0.0141
0.1675
0.0501
0.5751
4.0271
0.2810
3.3452
0.4593
0.0620
0.6033
0.0861
T
ABLE
4. One-way MANOVA, univariate one-way ANO-
VAs and Tukey’s post hoc tests for effects of treatment
(control, crayfish, and trout) on length of the larval period
of
Rana temporaria
, size at metamorphosis, and growth
rate.
Response variable df
MS
FP
Multivariate analysis
Larval period
6, 20
2, 12
0.1234†
0.0025
6.1540
1.6577
0.0009
0.2314
Size at metamorphosis
Control vs. Crayfish
Control vs. Trout
Crayfish vs. Trout
2, 12 0.0455 27.8297
,
0.0001
0.7683
0.0003
0.0002
Growth rate
Control vs. Crayfish
Control vs. Trout
Crayfish vs. Trout
2, 12 0.0026 14.4142 0.0006
0.4859
0.0052
0.0008
†Wilks’ lambda.
F
IG
. 3. The number of tadpoles (means
6
1
SE
) caught
per active-fauna trap over time (CPUE
5
catch per unit effort)
in control cages (C), crayfish cages (CR), trout cages (T), and
in crayfish
1
trout cages (CR
1
T) (statistics in Table 3).
macrophytes as predicted (Fig. 1B), but had less of an
impact on predatory invertebrates, tadpoles, and insect
grazers. Signal crayfish have weak direct effects on
actively moving prey such as those predatory inver-
tebrates and insect grazers that dominated the inver-
tebrate biomass in this pond (Nystro¨m et al. 1999).
Several crayfish species have been found to have a
negative impact on snails and aquatic macrophytes, in-
troduced species in particular (e.g., Nystro¨m 1999).
Correlative studies also suggest that signal crayfish
have a strong negative impact on aquatic macrophytes
and snails in Swedish ponds (Nystro¨m et al. 1996).
In contrast, trout had a strong negative impact on the
biomass of large active predatory invertebrates. More-
over, trout also reduced biomass of invertebrate graz-
ers, such as insect grazers, but had less of an impact
on snail biomass. The number of tadpoles caught in
traps and the number of surviving froglets were also
low in cages with trout; froglets had a reduced growth
rate compared to control and crayfish cages. The effect
of trout in our experiment is consistent with the results
from previous studies showing that native (e.g., Ware
1973, Cooper 1988, Wiseman et al. 1993) and intro-
duced rainbow trout (Bradford et al. 1998) have strong
effects on large active prey. Similarly, snails were rel-
atively unaffected by trout in other studies (cf. Macan
1966, Hemphill and Cooper 1984).
Overall, our results suggest that when two nonin-
teracting predators have strong but different effects on
prey survival or behavior, their combined effects on
intermediate trophic levels reflect responses to the more
dangerous predator (e.g., Kurzava and Morin 1998).
Our data suggest several different mechanisms that ex-
plain the direct and indirect predator effects on these
pond communities.
Trophic cascades and effects of predators on algae
In food chains, trophic cascades occur if primary
consumers that regulate primary producers are signif-
icantly reduced by predators, or if these primary con-
sumers respond to predators by reduced feeding activ-
ity (McIntosh and Townsend 1996, Turner 1997, Peck-
arsky and McIntosh 1998). However, trophic cascades
may not be readily observed if top predators are om-
nivorous, having direct effects on primary producers
(Nystro¨m et al. 1996, Pringle and Hamazaki 1998).
1032
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
F
IG
. 4. Size at metamorphosis of surviving froglets (means
1
1
SE
) in (A) different treatments, and (B) the number of
froglets surviving from different size classes (all froglets pooled from five replicates) (statistics in Table 4).
F
IG
. 5. Growth rate of froglets (mm/d) from
stocking until metamorphosis in relation to the
final number of surviving froglets in control
cages (C), crayfish cages (CR), and trout cages
(T) (ANCOVA, see
Results: Tadpoles and frog-
lets
).
Since both
Lymnaea
snails and
R. temporaria
tadpoles
reduce the biomass of periphytic algae (Bro¨nmark et
al. 1991), we expected strong indirect effects of pred-
ators on periphyton if those predators significantly re-
duced grazer biomass or activity. In crayfish cages,
snail densities decreased dramatically and periphyton
increased on both tiles and strips. Interestingly, signal
crayfish is an omnivore that also consumes periphyton
in Swedish ponds (Lira˚s et al. 1998), which may have
been expected to conceal any cascading effects. How-
April 2001 1033
MULTIPLE INTRODUCED PREDATORS
T
ABLE
5. Two-way repeated-measures ANOVAs for effects of crayfish, trout, and their in-
teractive effects on (A) macrophyte coverage after predator introduction in June and July
1997 (Time) and (B) the cumulative biomass of floating live macrophyte parts in cages.
Source df
MS
FP
A) Macrophyte coverage
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
1.0582
0.0068
0.0014
0.0366
28.9446
0.1852
0.0378
,
0.0001
0.6727
0.8484
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
1
1
1
1
16
0.0020
0.0225
0.0027
0.0043
0.0018
1.1078
12.7970
1.5230
2.4677
0.3082
0.0025
0.2350
0.1358
B) Floating macrophytes
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
70.8141
1.6065
1.0025
88.5051
2.0078
1.2530
,
0.0001
0.1757
0.2795
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
5
5
5
5
80
0.6413
0.1266
0.1466
0.0508
0.0699
9.1735
1.8116
2.0964
0.7269
0.0036
0.1933
0.1589
0.4409
Note:
ANOVA for effects on the cumulative biomass of floating live macrophyte parts (B)
includes data from six weekly samplings after predator introduction (Time).
ever, in this and other studies, crayfish grazing on pe-
riphyton did not outweigh the positive indirect effect
of a reduction in snail densities (Weber and Lodge
1990, Lodge et al. 1994, Nystro¨m et al. 1999). Due to
a mismatch of their feeding morphology, crayfish are
not as efficient grazers on microalgae as are snails (Lut-
tenton et al. 1998); and therefore indirect effects of
omnivorous crayfish on algae may be strong. In con-
trast, in systems with algal communities having fila-
mentous forms, crayfish grazing may be significant
(Hart 1992, Creed 1994) and may outweigh reduced
grazing from snails (Nystro¨m et al. 1996).
Furthermore, presence of refuges in heterogenous
habitats (e.g., Power 1992
b
) and effective prey de-
fenses (e.g., snails with thick shells) may also prevent
cascading effects of top predators (Bro¨nmark 1994,
Bro¨nmark and Weisner 1996). Here, crayfish reduced
snail abundance and habitat complexity simultaneously
by reducing macrophytes, but even in complex habitats
signal crayfish have strong effects on
Lymnaea
snails
(Nystro¨m and Pe´rez 1998). As opposed to mollusci-
vorous fish, crayfish are not gape-limited and may re-
duce the abundance of even large thick-shelled snail
species (Lodge et al. 1998). Therefore, cascading ef-
fects of crayfish were observed even though crayfish
are omnivorous, habitats were heterogenous, and prey
were apparently defended.
The trophic cascade observed in cages with trout was
due to a different mechanism. The high periphyton bio-
mass in cages with trout in this study may be a com-
bination of both density and trait-mediated trout effects
on tadpole grazing. In cages with trout, both tadpole
survival to metamorphosis and the number of trapped
tadpoles (activity) were reduced, even though habitats
were complex. In streams, trout may have positive in-
direct effects on periphyton by consuming large im-
portant invertebrate grazers (Bechara et al. 1992), or
by forcing grazers to reduce their feeding activity (e.g.,
McIntosh and Townsend 1996). In another system,
predatory newts in pond communities indirectly affect
phytoplankton (Morin 1995) and periphyton (Leibold
and Wilbur 1992, Wilbur 1997) by consuming tadpoles.
Peacor and Werner (1997) found that the nonlethal
presence of a caged dragonfly led to increased periph-
yton levels by changing tadpole activity. In addition,
Rana
tadpoles from the same population we used in
our experiment increased their time spent in refuges in
the presence of chemical cues released by rainbow
trout. This change in behavior was sufficient to allow
periphyton to increase (Nystro¨m and A
˚
bjo¨rnsson
2000).
Our results suggest that trophic cascades may de-
velop in aquatic food webs even with omnivores such
as crayfish, and in complex habitats with trout. These
strong indirect effects are mediated through both pre-
dation on important grazers (i.e., the crayfish–snail–
periphyton link) and a combination of density and be-
havioral responses of grazers to predators (i.e., the
trout–tadpole–periphyton link).
1034
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
F
IG
. 6. (A) Percentage macrophyte coverage and (B) ac-
cumulated macrophyte dry biomass (means
6
1
SE
) floating
in control cages (C), crayfish cages (CR), trout cages (T), and
in crayfish
1
trout cages (CR
1
T) (statistics in Table 5).
F
IG
. 7. Areal periphyton on (A) strips and on (B) tiles
measured as chlorophyll
a
(means
6
1
SE
) over time in control
cages (C), crayfish cages (CR), trout cages (T), and in crayfish
1
trout cages (CR
1
T). Note the different scales on the
y
-
axes (statistics in Table 6).
Effects of predators on tadpoles and froglets
Predator avoidance behaviors are very common in
nature (Lima and Dill 1990) and may reduce direct
effects on prey mortality (Scheffer 1997). However,
predator avoidance often has associated fitness costs
(Lima 1998, McPeek and Peckarsky 1998). In our
study, tadpoles associated with crayfish suffered more
tail injuries than in controls, in agreement with exper-
imental studies (Figiel and Semlitsch 1991, Axelsson
et al. 1997). Tail length affects tadpole swimming speed
(Wassersug and Sperry 1977) and, thus, the risk of
being caught by crayfish (Figiel and Semlitsch 1991).
Moreover, the swimming ability of tadpoles is greatly
reduced during metamorphosis and froglets are more
sensitive to predation than pre-metamorphic stages
(Wassersug and Sperry 1977, Huey 1980, Brown and
Taylor 1995). The sublethal effect of crayfish may ex-
plain why fewer tadpoles metamorphosed in crayfish
cages than control cages.
In trout cages, however, survivorship, growth rate, and
size at metamorphosis were lower than in control and
crayfish cages. Size at metamorphosis is positively cor-
related to juvenile survivorship and fitness in the adult
state (Berven and Gill 1983, Smith 1987). In the com-
mon frog, large females produce more eggs than small
females, and egg fertility is also positively correlated to
the length of the male (Gibbons and McCarthy 1986).
Thus, although the direct mortality effects of trout and
crayfish on froglets were significant in this study, the
negative effect of trout on
Rana
populations is probably
stronger than for crayfish since surviving froglets were
smaller in trout cages. The size distribution of surviving
froglets in trout cages indicates that selective predation
was unlikely since the mean size of the entire population
of froglets had shifted toward smaller sizes, rather than
being truncated due to selective removal of larger tad-
poles (Fig. 4B, see also Fauth 1990). Furthermore, pred-
ator-induced reduction in the density of tadpoles could
have reduced intraspecific competition, resulting in
higher growth rates of surviving tadpoles in cages with
low survivorship (Fig. 5, see also Morin 1983). How-
ever, if tadpoles alter their microhabitat to avoid pred-
ators, reduced feeding rate may decrease tadpole growth
rate and size at metamorphosis (Skelly 1992).
Rana
tad-
poles reduce their activity when exposed to waterborne
April 2001 1035
MULTIPLE INTRODUCED PREDATORS
T
ABLE
6. Two-way repeated-measures ANOVAs for effects of crayfish, trout and their in-
teractive effects on periphyton measured as chlorophyll
a
(A) strips and (B) tiles afterpredator
introduction in June and July 1997 (Time).
Source df
MS
FP
A) Periphyton (strips)
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
10.2707
8.3586
0.3773
0.2995
34.2906
27.9067
1.2596
,
0.0001
,
0.0001
0.2783
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
1
1
1
1
16
4.3926
0.0238
0.0084
0.0024
0.0936
46.9326
0.2542
0.0901
0.0253
,
0.0001
0.6210
0.7679
0.8757
B) Periphyton (tiles)
Between cages
Crayfish
Trout
Crayfish
3
Trout
Error
1
1
1
16
1.5632
3.4244
0.7848
0.1246
12.3666
27.0907
6.2086
0.0029
,
0.0001
0.0241
Within cages
Time
Time
3
Crayfish
Time
3
Trout
Time
3
Crayfish
3
Trout
Error
5
5
5
5
80
3.6895
0.0052
0.2089
0.3044
0.0153
241.3404
0.3386
13.6622
19.9094
,
0.0001
0.5688
0.0020
0.0004
chemical cues from rainbow trout (Nystro¨m and A
˚
b-
jo¨rnsson 2000).
Effects of multiple predators
Several studies have revealed that multiple predators
often have interactive effects on prey due to risk re-
duction or risk enhancement (Sih et al. 1998). Here, the
overall impact of crayfish and trout on components of
the pond community reflected different responses by in-
termediate consumers to the most dangerous predator.
Predators primarily had independent effects on the com-
munity, because predator–predator interactions were
weak and neither risk reduction nor enhancement oc-
curred. Similar multiple predator effects have been ob-
served in systems where the two predators differ in their
effects on prey, so that one predator has weak effects
on a certain prey species, and the other has strong effects
(Sih et al. 1998). Based on their studies of stream-living
mayflies, McIntosh and Peckarsky (1999) predict that
predator avoidance should take place in the form of a
hierarchy when one predator presents a substantially
higher predation risk than the other. In this study, the
effects on snails and macrophytes were due to crayfish,
whereas the effects on insect grazers and predatory in-
vertebrates were mostly due to trout. Rainbow trout con-
sume signal crayfish (Frantz and Cordone 1970; P. Nys-
tro¨m,
unpublished data
), and juvenile crayfish often re-
duce their activity levels and increase the use of refuges
in the presence of fish (Stein and Magnuson 1976, Butler
and Stein 1985, Blake and Hart 1993). Predatory fish
could, therefore, reduce crayfish impact on lower trophic
levels (Stein and Magnuson 1976, Hill and Lodge 1995).
However, adult crayfish have a size refuge from most
fish predators and do not respond to predatory fish (Stein
and Magnuson 1976, Butler and Stein 1985). In this
study, catches of adult crayfish did not differ between
crayfish and crayfish
1
trout cages, suggesting that the
abundance of adult signal crayfish was unaffected by
trout. Two years after the experiment, 12 of the rainbow
trout used in the experiment were recaptured in the pond
and their stomach contents analyzed. None of the trout
had consumed crayfish (P. Nystro¨m,
unpublished data
)
indicating weak direct effects of trout on crayfish.
Behavioral differences between prey species may af-
fect their vulnerability to predators with different for-
aging strategies (Rahel and Stein 1988, Blois-Heulin
et al. 1990, McPeek 1990, Kurzava and Morin 1998).
Hence, presence of multiple predators may alter an an-
imal’s mortality risk, if avoiding one predator affects
the predation risk from another predator (Soluk and
Collins 1988, Kiesecker and Blaustein 1998, Losey and
Denno 1998). Here, decreased tadpole activity in the
presence of rainbow trout did not increase predation
risk from signal crayfish, similar to results in Nystro¨m
and A
˚
bjo¨rnsson (2000). Tadpoles in this study respond-
ed to trout by reducing their activity, but tended to be
more active in crayfish cages. The number of trapped
tadpoles in the combined treatment was similar to the
numbers active in trout cages, suggesting that tadpoles
responded to the more dangerous predator.
Depending on substrate type, the indirect effects on
periphyton by crayfish and trout differed on the last
sampling date. Periphyton on strips in cages with both
predators was higher than when crayfish or trout were
1036
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
alone, due to reduced grazing from snails in crayfish
cages and reduced grazing from tadpoles in trout cages.
Since tadpoles responded to the more dangerous trout
by reduced feeding, trait-mediated indirect effects on
periphyton persisted in the combined predator treatment.
There are presently few demonstrations of indirect ef-
fects on primary producers in food webs influenced by
multiple predators. Morin (1995) showed that two spe-
cies of salamanders had noninteractive effects on phy-
toplankton biomass by predation on tadpoles. Nystro¨m
and A
˚
bjo¨rnsson (2000) found that direct lethal effects
of signal crayfish on common frog tadpoles had positive
indirect effects on periphyton, and nonlethal presence
of trout had similar effects on periphyton by trait-me-
diated effects on tadpole behavior. These predatoreffects
were independent and in the combined predator treat-
ment periphyton biomass was the product of single pred-
ator effects. Trait-mediated indirect effects on periph-
yton have also been shown in other systems with mul-
tiple predators (Peacor and Werner 1997, Peckarsky and
McIntosh 1998). Periphyton on tiles conveyed at the last
sampling date resulting in a significant crayfish-by-trout
interaction. It is possible that periphyton on tiles (sam-
pled at a water depth of 0.5 m) became more light limited
than periphyton growing on strips (sampled just below
the water surface) in cages with predators. At the end
of the experiment we observed an increase in turbidity
in predator cages that might have prevented thick algal
mats from building up on tiles.
Effects of multiple introduced predators
Predictions about the impact of exotic species are hard
to make, since the characteristics of the target com-
munity are as important as the characteristics of the
introduced species, each of which is rarely known in
detail (e.g., Lodge 1993, but see Holway 1999 for an
exception). Theoretical and empirical studies suggest
that the impact of an introduced predator will most likely
be severe in species-poor systems where native predators
and competitors are absent (Diamond and Case 1986,
Pimm 1989, Lodge 1993). Native prey may also be vul-
nerable to new predators if they have no prior experience
with functionally similar predators. Prey that have not
been exposed to predators or to native predators only,
may not have evolved efficient defense systems against
nonnative predators. Thus, to predict the effects of mul-
tiple introduced predators on community structure we
first need to determine interaction strength between spe-
cies in the native food web; we also need to determine
if the exotic predators interact with each other and with
lower trophic levels. Finally it is important to know if
native prey have any experience with the introduced
predator type and show appropriate avoidance behaviors
or other defense adaptations.
In this and other systems, introduced crayfish species
may have stronger effects on snails and macrophytes
than native crayfish due to higher consumption rates
(Olsen et al. 1991, Nystro¨m and Strand 1996, Lodge
et al. 1998, Nystro¨m et al. 1999). The strong impact
of crayfish on snails observed in this study suggests
that snails lack or have inefficient antipredator defenses
against crayfish, possibly even less efficient to an in-
troduced species. In this study, the coevolutionary his-
tory of crayfish and snails in Sweden may not have
been sufficiently long to allow the development of ef-
ficient antipredator behaviors in snails. Similarly, the
predatory invertebrates in these ponds may have little
experience with the ‘‘trout type predator’’ because na-
tive brown trout in southern Sweden are mostly found
in streams. For example, introduced trout in fishless
California alpine lakes has eliminated or greatly re-
duced the numbers of large, mobile, and epibenthic-
limnetic taxa including invertebrates and amphibians
(Bradford et al. 1998).
Predictions and future investigations
Based on the results from this study, we predict that
effects of multiple introduced predators on community
structure can be strong when the introduced predators:
(1) experience little predation or competition from native
predators, (2) do not affect each other’s biomass or for-
aging efficiency, and (3) have prey that lack efficient
defense adaptations to the introduced predators. Our data
further suggest that lethal predator effects should be
most important for naive prey (e.g., snails in this study),
but trait-mediated effects on prey should be more im-
portant if prey show significant antipredator responses
to the predators (e.g., the common frog tadpoles in this
study). Trophic cascades may develop if vulnerable prey
are strong interactors in the native food web, and the
introduced predators, though omnivorous, are less effi-
cient consumers on prey resources. These cascades may
be strong if the multiple introduced predators affect dif-
ferent grazers in the native food web, each of which has
the potential to control plant resource levels.
To understand the mechanisms that structure more
complex food webs we have to consider both direct and
indirect effects of multiple predators, and the relative
importance of density and trait-mediated effects of mul-
tiple predators on prey populations. Also as a result of
antipredator behavior, prey mortality rate may not in-
crease linearly with increasing predator density (Abrams
1993). Future studies, should, therefore address inter-
active effects of predators by including variations in
predator and prey density (e.g., Diehl 1995). Further, in
an era of faunal homogenization due to anthropogenic
effects on dispersal rates, the role of exotic species, and
especially exotic predators, will become more and more
important to understanding the structure of communities.
A
CKNOWLEDGMENTS
We are very grateful to Go¨ran Bo¨rjesson for letting us use
the experimental pond. Jan A
˚
bjo¨rnsson helped us construct the
enclosures and Elsa Ma˚nsson assisted during the installation.
Jonas Dahl, Olof Berglund, Barbara Peckarsky, and two anon-
ymous reviewers made valuable comments on an earlierversion
of the manuscript. The investigation was supported by grants
April 2001 1037
MULTIPLE INTRODUCED PREDATORS
from the Swedish Environmental Protection Agency (to Wil-
helm Grane´li), Ebba and Sven Schwartz foundation (to Wil-
helm Grane´li),and by the Royal Swedish Academy of Sciences.
L
ITERATURE
C
ITED
Abrams, P. A. 1993. Why predation rate should not be pro-
portional to predator density. Ecology 74:726–733.
Anonymous. 1993. Institute of Freshwater Research of the
Swedish National Board of Fisheries. Mo¨jligheter att o¨ka
flodkra¨ftbesta˚nd i svenska vatten. Information fra˚n So¨t-
vattenslaboratoriet, Drottningholm, nr 2 (
in Swedish
).
Appelberg, M., and T. Odelstro¨m. 1986. Habitat distribution,
growth and abundance of the crayfish
Astacus astacus
L.
in the littoral zone of four neutralized lakes. Pages 1–30
in
M. Appelberg, editor. The crayfish
Astacus astacus
L.
in acid and neuralized environments. PhD thesis, Uppsala
University, Sweden.
Axelsson, E., P. Nystro¨m, J. Sidenmark, and C. Bro¨nmark.
1997. Crayfish predation on amphibian eggs and larvae.
Amphibia-Reptilia 18:217–228.
Bechara, J. A., G. Moreau, and D. Planas. 1992. Top-down
effects of brook trout (
Salvelinus fontinalis
) in a boreal
forest stream. Canadian Journal of Fisheries and Aquatic
Sciences 49:2093–2103.
Beebee, T. J. C. 1985. Discriminant analysis of amphibian
habitat determinants in south-east England. Amphibia-Rep-
tilia 6:35–43.
Berven, K. A., and D. E. Gill. 1983. Interpreting geographic
variation in life-history traits. American Zoologist 23:85–
97.
Billick, I., and T. J. Case. 1994. Higher order interactions in
ecological communities: What are they and how can they
be detected? Ecology 75:1529–1543.
Blake, M. A., and P. J. B. Hart. 1993. The behavioural re-
sponses of juvenile signal crayfish
Pacifastacus leniusculus
to stimuli from perch and eels. Freshwater Biology 29:89–
97.
Blois-Heulin, C., P. H. Crowley, M. Arrington, and D. M.
Johnson. 1990. Direct and indirect effects of predators on
the dominant invertebrates of two freshwater littoral com-
munities. Oecologia 84:295–306.
Bradford, D. F., S. D. Cooper, T. M. Jenkins Jr., K. Kratz, O.
Sarnelle, and A. D. Brown. 1998. Influences of natural
acidity and introduced fish on faunal assemblages in Cal-
ifornia alpine lakes. Canadian Journal of Fisheries and
Aquatic Sciences 55:2478–2491.
Bro¨nmark, C. 1989. Interactions between epiphytes, mac-
rophytes and freshwater snails: a review. Journal of Mol-
luscan Studies 55:299–311.
Bro¨nmark, C. 1992. Leech predation on juvenile freshwater
snails: effects of size, species and substrate. Oecologia 91:
526–529.
Bro¨nmark, C. 1994. Effectsof tench and perch on interactions
in a freshwater, benthic food chain. Ecology 75:1818–1828.
Bro¨nmark, C., and P. Edenhamn. 1994. Does the presence of
fish affect the distribution of tree frogs (
Hyla arborea
)?
Conservation Biology 8:841–845.
Bro¨nmark, C., S. D. Rundle, and A. Erlandsson. 1991. In-
teractions between freshwater snails and tadpoles: com-
petition and facilitation. Oecologia 87:8–18.
Bro¨nmark, C., and J. E. Vermaat. 1998. Complex fish–snail–
epiphyton interactions and their effects on submerged
freshwater macrophytes. Pages 47–68
in
E. Jeppesen, Ma.
Søndergaard, Mo. Søndergard, and K. Christoffersen, ed-
itors. The structuring role of submerged macrophytes in
lakes. Ecological Studies 131. Springer, New York, New
York, USA.
Bro¨nmark, C., and S. E. B. Weisner. 1996. Decoupling of
cascading trophic interactions in a freshwater, benthic food
chain. Oecologia 108:534–541.
Brown, R. M., and D. H. Taylor. 1995. Compensatory escape
mode trade-offs between swimming performance and ma-
neuvering behavior through larval ontogeny of the wood
frog,
Rana sylvatica.
Copeia 1995:1–7.
Butler, M. J., and R. A. Stein. 1985. An analysis of the
mechanisms governing species replacements in crayfish.
Oecologia 66:168–177.
Cooper, S. D. 1988. The responses of aquatic insects and
tadpoles to trout. Internationale Vereinigung fu¨r Theore-
tische und Angewandte Limnologie, Verhandlungen 23:
1698–1703.
Creed, R. P., Jr. 1994. Direct and indirect effects of crayfish
grazing in a stream community. Ecology 75:2091–2103.
Crowder, L. B., and W. E. Cooper. 1982. Habitat structural
complexity and the interaction between bluegills and their
prey. Ecology 63:1802–1813.
Diamond, J., and T. J. Case. 1986. Community ecology.Harp-
er and Row, New York, New York, USA.
Diehl, S. 1992. Fish predation and benthic community struc-
ture: the role of omnivory and habitat complexity. Ecology
73:1646–1661.
Diehl, S. 1993. Relative consumer sizes and the strengths of
direct and indirect interactions in omnivorous feeding re-
lationships. Oikos 68:151–157.
Diehl, S. 1995. Direct and indirect effects of omnivory in a
littoral lake community. Ecology 76:1727–1740.
Eklo¨v, A. G. 1996. Effects of habitat size and species richness
on anadromous brown trout,
Salmo trutta
L., populations.
Fisheries Management and Ecology 3:97–101.
Elvira, B., G. G. Nicola, and A. Almodovar. 1996. Pike and
red swamp crayfish: a new case on predator–prey relation-
ship between aliens in central Spain. Journal of Fish Bi-
ology 48:437–446.
Endler, J. A. 1991. Interactions between predators and prey.
Pages 169–196
in
J. R. Krebs and N. B. Davis, editors.
Behavioural ecology: an evolutionary approach. Blackwell
Scientific, Oxford, UK.
Fauth, J. E. 1990. Interactive effects of predators and early
larval dynamics of the treefrog
Hyla chrysoscelis.
Ecology
71:1609–1616.
Feltmate, B. W., and D. D. Williams. 1989. Influence of rain-
bow trout (
Oncorhynchus mykiss
) on density and feeding
behavior of a perlid stonefly. Canadian Journal of Fisheries
and Aquatic Sciences 46:1575–1580.
Figiel, C. R., Jr., and R. D. Semlitsch. 1991. Effects of non-
lethal injury and habitat complexity on predation in tadpole
populations. Canadian Journal of Zoology 69:830–834.
Flint, R. W., and C. R. Goldman. 1975. The effects of a
benthic grazer on the primary productivity of the littoral
zone of Lake Tahoe. Limnology and Oceanography 20:
935–944.
Foster, J., and F. M. Slater. 1995. A global review of crayfish
predation with observations on the possible loss of
Aus-
tropotamobius pallipes
in the Welsh wye due to crayfish
plague. Freshwater Crayfish 8:589–613.
Frantz, T. C., and A. L. Cordone. 1970. Food of lake trout
in Lake Tahoe. California Fish and Game 56:21–35.
Fretwell, S. D. 1987. Food chain dynamics: the central theory
of ecology? Oikos 50:291–301.
Gamradt, S. C., and L. B. Kats. 1996. Effect of introduced
crayfish and mosquitofish on California newts (
Taricha to-
rosa
). Conservation Biology 10:1155–1162.
Gastreich, K. R. 1999. Trait-mediated indirect effects of a
theridiid spider on an ant-plant mutualism. Ecology 80:
1066–1070.
Gibbons, M. M., and T. K. McCarthy. 1986. The reproductive
output of frogs
Rana temporaria
(L.) with particular ref-
erence to body size and age. Journal of Zoology (London)
209:579–593.
Gosner, K. L. 1960. A simplified table for staging anuran
1038
PER NYSTRO
¨
M ET AL.
Ecology, Vol. 82, No. 4
embryos and larvae with notes on identification. Herpe-
tologica 16:183–190.
Hairston, N. G., F. E. Smith, and L. B. Slobodkin. 1960.
Community structure, population control and competition.
American Naturalist 94:421–425.
Hart, D. D. 1992. Community organization in streams: the
importance of species interactions, physical factors, and
chance. Oecologia 9:220–228.
Hemphill, N., and S. D. Cooper. 1984. Differences in the
community structure of stream pools containing or lacking
trout. Internationale Vereinigung fu¨r Theoretische und An-
gewandte Limnologie, Verhandlungen 22:1858–1861.
Hill, A. M., and D. M. Lodge. 1995. Multi-trophic-level im-
pact of sublethal interactions between bass and omnivorous
crayfish. Journal of the North American Benthological So-
ciety 14:306–314.
Holomuzki, J. R. 1989. Predation risk and macroalga use by
the stream-dwelling salamander
Ambystoma texanum.
Cop-
eia 1989:22–28.
Holway, D. A. 1999. Competitive mechanisms underlying
the displacement of native ants by the invasive Argentine
ant. Ecology 80:238–251.
Huey, R. B. 1980. Sprint velocity of tadpoles (
Bufo boreas
)
through metamorphosis. Copeia 1980:537–540.
Hunt, P. C., and K. O’Hara. 1973. Overwinter feeding in
rainbow trout. Journal of Fish Biology 5:277–280.
Jespersen, A. M., and K. Christoffersen.1987. Measurements
of chlorophyll-
a
from phytoplankton using ethanol as ex-
traction solvent. Archiv fu¨r Hydrobiologie 109:445–454.
Kiesecker, J. M., and A. R. Blaustein. 1997. Population dif-
ferences in responses of red-legged frogs (
Rana aurora
)to
introduced bullfrogs. Ecology 78:1752–1760.
Kiesecker, J. M., and A. R. Blaustein. 1998. Effects of in-
troduced bullfrogs and smallmouth bass on microhabitat
use, growth and survival of native red-legged frogs (
Rana
aurora
). Conservation Biology 12:776–787.
Kurzava, L. M., and P. J. Morin. 1998. Test of functional
equivalence: complementary roles of salamanders and fish
in community organization. Ecology 79:447–489.
Lefcort, H. 1996. Adaptive, chemically mediated fright re-
sponse in tadpoles of the southern leopard frog,
Rana utri-
cularia.
Copeia 1996:455–459.
Leibold, M. A., and H. M. Wilbur. 1992. Interactionsbetween
food-web structure and nutrients on pond organisms. Na-
ture 360:341–343.
Letourneau, D. K., and L. A. Dyer. 1998. Experimental test
in lowland tropical forest shows top-down effects through
four trophic levels. Ecology 79:1678–1687.
Lima, S. L. 1998. Nonlethal effects in the ecology of pred-
ator–prey interactions: What are the ecological effects of
anti-predator decision making? BioScience 48:25–34.
Lima, S. L., and L. M. Dill. 1990. Behavioral decisions made
under the risk of predation: a review and prospectus. Ca-
nadian Journal of Zoology 68:619–640.
Lindqvist, S. A
˚
. 1981. Food selection by rainbow trout (
Sal-
mo gairdneri
) in some western lakes. Information fra˚nSo¨t-
vattenslaboratoriet, Drottningholm, nr 2 (in Swedish with
English summary).
Lira˚s, V., M. Lindberg, P. Nystro¨m, H. Annadotter, L. A.
Lawton, and B. Graf. 1998. Can ingested cyanobacteriabe
harmful to the signal crayfish (
Pacifastacus leniusculus
)?
Freshwater Biology 39:233–242.
Lodge, D. M. 1993. Biological invasions: lessons for ecol-
ogy. Trends in Ecology and Evolution 8:133–137.
Lodge, D. M., and A. M. Hill. 1994. Factors governing spe-
cies composition, population size, and productivity of cool-
water crayfishes. Nordic Journal of Freshwater Research
69:111–136.
Lodge, D. M., M. W. Kershner, J. E. Aloi, and A. P. Covich.
1994. Effects of an omnivorous crayfish (
Orconectes rus-
ticus
) on a freshwater littoral food web. Ecology 75:1265–
1281.
Lodge, D. M., R. A. Stein, K. M. Brown, A. P. Covich, C.
Bro¨nmark, J. E. Garvey, and S. P. Klosiewski. 1998. Pre-
dicting impact of freshwater exotic species on native bio-
diversity: challenges in spatial scaling. Australian Journal
of Ecology 23:53–67.
Loman, J. 1988. Breeding by
Rana temporaria
; the impor-
tance of pond size and isolation. Memoranda Societes Fau-
na Flora Fennica 64:113–115.
Losey, J. E., and R. F. Denno. 1998. Positive predator–prey
interactions: enhanced predation rates and synergistic sup-
pression of aphid populations. Ecology 79:2143–2152.
Lowery, R. S., and D. M. Holdich. 1988.
Pacifastacus len-
iusculus
in North America and Europe, with details of the
distribution of the introduced and native crayfish species
in Europe. Pages 283–308
in
D. M. Holdich and R. S.
Lowery, editors. Freshwater crayfish, biology,management
and exploitation. Croom Helm, London and Sydney, UK.
Luttenton, M. R., M. J. Horgan, and D. M. Lodge. 1998.
Effects of three
Orconectes
crayfishes on epilithic microal-
gae: a laboratory experiment. Crustaceana 71:845–861.
Macan, T. T. 1966. The influence of predation on the fauna
of a moorland fishpond. Archiv fu¨r Hydrobiologie 61:432–
452.
Marnell, F. 1998. Discriminant analysis of the terrestrial and
aquatic habitat determinants of the smooth newt (
Triturus
vulgaris
) and the common frog (
Rana temporaria
) in Ire-
land. Journal of Zoology (London) 244:1–6.
McCollum, E. W., L. B. Crowder, and S. A. McCollum.1998.
Complex interactions of fish, snails, and littoral zone pe-
riphyton. Ecology 79:1980–1994.
McIntosh, A. R., and B. L. Peckarsky. 1999. Criteria deter-
mining behavioural responses to multiple predators by a
stream mayfly. Oikos 85:554–564.
McIntosh, A. R., and C. R. Townsend. 1996. Interactions
between fish, grazing invertebrates and algae in a New
Zealand stream: a trophic cascade mediated by fish-induced
changes to grazer behaviour? Oecologia 108:174–181.
McPeek, M. A. 1990. Behavioral differences between
En-
allagama
species (Odonata) influencing differential vul-
nerability to predators. Ecology 71:1714–1726.
McPeek, M. A., and B. L. Peckarsky. 1998. Life histories
and the strengths of species interactions: combining mor-
tality, growth, and fecundity effects. Ecology 79:867–879.
Meyer, A. H., B. R. Schmidt, and K. Grossenbacher. 1998.
Analysis of three amphibian populations with quarter-cen-
tury long time-series. Proceedings of the Royal Society of
London Britain 265:523–528.
Morin, P. J. 1983. Predation, competition, and the compo-
sition of larval anuran guilds. Ecological Monographs 53:
119–138.
Morin, P. J. 1987. Salamander predation, prey facilitation and
seasonal succession in microcrustacean communities. Pag-
es 174–187
in
C. W. Kerfoot and A. Sih, editors. Predation:
direct and indirect impacts on aquatic communities. Uni-
versity Press of New England, Hanover and London, New
Hampshire, USA.
Morin, P. J. 1995. Functional redundancy, non-additive in-
teractions, and supply-side dynamics in experimental pond
communities. Ecology 76:133–149.
Nilsson, F. 1999. Skaraborgsinventeringen av sma˚vatten
1997. Sjo¨ar & vattendrag, a˚rsskrift fra˚n miljo¨o¨vervaknin-
gen 1997 Swedish Environmental Protection Agency,
Stockholm, Sweden (in Swedish).
Nystro¨m, P. 1999. Ecological impact of introduced and native
crayfish on freshwater communities: European perspec-
tives. Pages 63–85
in
F. Gherardi and D. M. Holdich, ed-
itors. Crayfish in Europe as alien species. CrustaceanIssues
11, Balkema, The Netherlands.
April 2001 1039
MULTIPLE INTRODUCED PREDATORS
Nystro¨m, P., and K. A
˚
bjo¨rnsson. 2000. Effects of fish chem-
ical cues on the interactions between tadpoles and crayfish.
Oikos 88:181–190.
Nystro¨m, P., C. Bro¨nmark, and W. Grane´li. 1996. Patterns in
benthic food webs: a role for omnivorous crayfish? Fresh-
water Biology 36:631–646.
Nystro¨m, P., C. Bro¨nmark, and W. Grane´li. 1999. Influence
of an exotic and a native crayfish species on a littoral ben-
thic community. Oikos 85:545–553.
Nystro¨m, P., and J. R. Pe´rez. 1998. Crayfish predation on the
common pond snail (
Lymnaea stagnalis
): the effect of hab-
itat complexity and snail size on foraging efficiency. Hy-
drobiologia 368:201–208.
Nystro¨m, P., and J. A. Strand. 1996. Grazing by a native and
an exotic crayfish on aquatic macrophytes. Freshwater Bi-
ology 36:673–682.
Oksanen, L., S. D. Fretwell, J. Arruda, and P. Niemela¨. 1981.
Exploitation ecosystems in gradients of primary produc-
tivity. American Naturalist 118:240–261.
Olsen, T. M., D. M. Lodge, G. M. Capelli, and R. Houlihan.
1991. Mechanisms of impact of an introduced crayfish (
Or-
conectes rusticus
) on littoral congeners snails and macro-
phytes. Canadian Journal of Fisheries and Aquatic Sciences
48:1853–1861.
Peacor, S. D., and E. E. Werner. 1997. Trait-mediatedindirect
interactions in a simple aquatic food web. Ecology 78:
1146–1156.
Peckarsky, B. L., and A. R. McIntosh. 1998. Fitness and
community consequences of avoiding multiple predators.
Oecologia 113:565–576.
Perry, W. L., D. M. Lodge, and G. A. Lamberti. 1997. Impact
of crayfish predation on exotic zebra mussels and native
invertebrates in a lake-outlet stream. Canadian Journal of
Fisheries and Aquatic Sciences 54:120–125.
Persson, L. 1986. Effectsof reduced interspecificcompetition
on resource utilization in perch (
Perca fluviatilis
). Ecology
67:355–364.
Pimm, S. L. 1989. Theories of predicting success and impact
of introduced species. Pages 351–367
in
J. A. Drake, H.
A. Mooney, F. di Castri, R. H. Groves, F. J. Kruger, M.
Rejma´nek, and M. Williamson, editors. Biological inva-
sions: a global perspective. John Wiley and Sons, New
York, New York, USA.
Polis, G. A., and D. R. Strong. 1996. Food web complexity
and community dynamics. American Naturalist 147:813–
846.
Power, M. E. 1992
a
. Top-down and bottom-up forces in food
webs: do plants have primacy? Ecology 73:733–746.
Power, M. E. 1992
b
. Habitat heterogeneity and the functional
significance of fish in river food webs. Ecology 73:1675–
1688.
Pringle, C. M., and T. Hamazaki. 1998. The role of omnivory
in a neotropical stream: separating diurnal and nocturnal
effects. Ecology 79:269–280.
Rahel, F. J., and R. A. Stein. 1988. Complex predator–prey
interactions and predator intimidation among crayfish, pi-
scivorous fish, and small benthic fish. Oecologia 75:94–
98.
Rosenheim, J. A. 1998. Higher-order predators and the reg-
ulation of insect herbivore populations. Annual Review in
Entomology 43:421–447.
Ross, S. T. 1991. Mechanisms structuring stream fish assem-
blages: Are there lessons from introduced species? Envi-
ronmental Biology of Fishes 30:359–368.
Scheffer, M. 1997. On the implication of predator avoidance.
Aquatic Ecology 31:99–107.
Schmitz, O. J., A. P. Beckerman, and K. M. O’Brien. 1997.
Behaviorally mediated trophic cascades: effects of preda-
tion risk on food web interactions. Ecology 78:1388–1399.
Schoener, T. W., and D. Spiller. 1999. Indirect effects in an
experimentally staged invasion by a major predator. Amer-
ican Naturalist 153:347–358.
Shave, C. R., C. R. Townsend, and T. A. Crowl. 1994. Anti-
predator behaviours of a freshwater crayfish(
Paranephrops
zealandicus
) to a native and an introduced predator. New
Zealand Journal of Ecology 18:1–10.
Sih, A. 1987. Predators and prey lifestyles: an evolutionary
and ecological overview. Pages 203–224
in
C. W. Kerfoot
and A. Sih, editors. Predation: direct and indirect impacts
on aquatic communities. University Press of New England,
Hanover and London, New Hampshire, USA.
Sih, A., G. Englund, and D. Wooster. 1998. Emergentimpacts
of multiple predators on prey. Trends in Ecology and Evo-
lution 13:350–355.
Skelly, D. K. 1992. Field evidence for a cost of behavioral
antipredator response in a larval amphibian. Ecology 73:
704–708.
Smith, D. C. 1983. Factors controlling tadpole populations
of the chorus frog (
Pseudacris triseriata
) on Isle Royale,
Michigan. Ecology 64:501–510.
Smith, D. C. 1987. Adult recruitment in chorus frogs: effects
of size and date at metamorphosis. Ecology 68:344–350.
Soluk, D. A. 1993. Multiple predator effects: predicting com-
bined functional response of a stream fish and invertebrate
predators. Ecology 74:219–225.
Soluk, D. A., and N. C. Collins. 1988. Synergistic interac-
tions between fish and stoneflies: facilitation and interfer-
ence among stream predators. Oikos 52:94–100.
Spiller, D. A., and T. W. Schoener. 1990. A terrestrial field
experiment showing the impact of eliminating top predators
on foliage damage. Nature 347:469–472.
Spiller, D. A., and T. W. Schoener. 1994. Effects of top and
intermediate predators in a terrestrial food web. Ecology
75:182–196.
Stein, R. A., and J. J. Magnuson. 1976. Behavioral response
of crayfish to a fish predator. Ecology 57:751–761.
Turner, A. M. 1997. Contrasting short-term and long-term
effects of predation risk on consumer habitat use and re-
sources. Behavioral Ecology 8:120–125.
Van Buskirk, J. 1988. Interactive effects of dragonfly pre-
dation in experimental pond communities. Ecology 69:
857–867.
Vitousek, P. M. 1990. Biological invasions and ecosystem
processes: towards an integration of population biology and
ecosystem studies. Oikos 57:7–13.
Wagner, B. M. A. 1997. Influence of fish on the breeding of
the red-necked grebe
Podiceps grisegena
(Boddaert, 1783).
Hydrobiologia 344:57–63.
Ware, D. M. 1973. Risk of epibenthic prey to predation by
rainbow trout (
Salmo gairdneri
). Journal of Fisheries Re-
search Board of Canada 30:787–797.
Wassersug, R. J., and D. G. Sperry. 1977. The relationship
of locomotion to differential predation on
Pseudacris tris-
eriata
(Anura: Hylidae). Ecology 58:830–839.
Weber, L. M., and D. M. Lodge. 1990. Periphytic food and
predatory crayfish: relative roles in determining snail dis-
tribution. Oecologia 82:33–39.
Wellborn, G. A., D. K. Skelly, and E. E. Werner. 1996. Mech-
anisms creating community structure across a freshwater
habitat gradient. Annual Reviews in Ecology and System-
atics 27:337–363.
Wilbur, H. M. 1997. Experimental ecology of food webs:
complex systems in temporary ponds. Ecology 78:2279–
2302.
Wiseman, S. W., S. D. Cooper, and T. L. Dudley. 1993. The
effects of trout on epibenthic odonate naids in streampools.
Freshwater Biology 30:133–145.
Wootton, J. T. 1994. Putting the pieces together: testing the
independence of interactions among organisms. Ecology
75:1544–1551.
