Responses of Foothill Yellow-legged Frog (Rana boylii) Larvae to an Introduced
Author(s): David J. Paoletti, Deanna H. Olson, and Andrew R. Blaustein
Source: Copeia, 2011(1):161-168. 2011.
Published By: The American Society of Ichthyologists and Herpetologists
BioOne (www.bioone.org) is an electronic aggregator of bioscience research content, and the online home to over
160 journals and books published by not-for-profit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial
inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.
Copeia 2011, No. 1, 161–168
Responses of Foothill Yellow-legged Frog (Rana boylii) Larvae to an
David J. Paoletti1, Deanna H. Olson2, and Andrew R. Blaustein1
The consequences of species introductions into non-native habitats are a major cause for concern in the U.S. Of
particular interest are the effects of predation by introduced fishes on native amphibian communities. We sought to
determine whether Foothill Yellow-legged Frog (Rana boylii) larvae could recognize non-native Smallmouth Bass
(Micropterus dolomieu) as a predatory threat. Through a series of laboratory experiments, we examined the initial and
overall behavioral responses of larvae to native predators (Rough-skinned newts, Taricha granulosa), introduced
predators (M. dolomieu), and native non-predatory fish (Speckled Dace, Rhinichthys osculus). Each experiment examined a
different potential mode of detection including chemical cues; visual cues; or a combination of chemical, visual, and
mechanical cues. Initially, larvae of R. boylii responded with an increase in activity levels when exposed to visual cues of
M. dolomieu. Analyses of overall responses suggested that individual larvae of R. boylii require multiple cues to facilitate
predator detection. When exposed to multiple cues of their native predator, larvae responded with a significant reduction
in activity levels. Those larvae exposed to cues of the non-native predator displayed similar behaviors relative to control
cues. Consequently, larvae of R. boylii appear to be especially vulnerable to predation by non-native M. dolomieu.
HE consequences of species introductions into nonnative habitats are a major cause of concern. An
introduced species may alter native habitats, cause
economic damage, carry pathogens, compete with natives
for resources, or prey on them (Vitousek et al., 1997; Mack et
al., 2000; Kiesecker et al., 2001; Muir and Jenkins, 2002).
These introductions are particularly prevalent in aquatic
systems. From 1950 to present, the number of introduced
aquatic species has more than tripled in the United States
(USGS, 2009). Muir and Jenkins (2002) have documented up
to 140 aquatic non-native species in the Great Lakes alone.
According to the U.S. Geological Survey’s Nonindigenous
Aquatic Species Database, over 42% (n 5 675) of introduced
aquatic species are fishes (USGS, 2009), with many of these
introductions being intentional.
In the U.S. Pacific Northwest, non-native fishes have been
stocked in a variety of habitats from rivers to sub-alpine
lakes to backyard ponds. Schade and Bonar (2005) conducted surveys of 12 western states from 2000–2002 and found
that of the most widely distributed and abundant nonnative fishes (n 5 15), almost all were introduced for sport
fisheries. Species such as Rainbow Trout (Oncorhynchus
mykiss), Brook Trout (Salvelinus fontinalis), Channel Catfish
(Ictalurus punctatus), and Smallmouth Bass (Micropterus
dolomieu) have been, and continue to be, stocked in many
areas of the western U.S. for recreational purposes.
Introduced species are one of the leading causes of
amphibian population declines worldwide (Kats and Ferrer,
2003), in addition to a myriad of other factors such as
habitat loss, disease, environmental changes, and pollutants (Blaustein and Wake, 1990; Alford and Richards,
1999; Stuart et al., 2004; Pounds et al., 2006). In the
western U.S., predation by introduced fishes has been
implicated in affecting distributions of several amphibian
species (for example, Cascades Frogs, Rana cascadae [Fellers
and Drost, 1993]; Mountain Yellow-legged Frogs, Rana
muscosa [Bradford et al., 1993; Knapp and Matthews,
2000]; and Long-toed Salamanders, Ambystoma macrodactylum [Tyler et al., 1998]). Eggs and larvae of many species
are completely dependent on water for periods ranging
from two months to four years, depending on the species
(Jones et al., 2005). Hence, these pre-metamorphic stages
may be particularly susceptible to the effects of novel fish
In Oregon, the distribution of the Foothill Yellow-legged
Frog (Rana boylii) appears to be influenced by introduced M.
dolomieu in some river systems. Rana boylii is a native streamdwelling frog listed as a ‘‘Sensitive Species’’ in the state,
where its range has been reduced to 43% of historical
locations (Fig. 1; Olson and Davis, 2007). In the Umpqua
River system, M. dolomieu was introduced as recently as the
mid-1960s, where their distribution has rapidly expanded
(Simon and Markle, 1999). This spread is particularly
evident in Cow Creek, a major tributary of the South
Umpqua River in Douglas County, Oregon (Fig. 2). As M.
dolomieu have spread upstream through Cow Creek, a
coincident reduction of R. boylii has been observed (Borisenko and Hayes, unpubl; Rombough, unpubl.). In the
federal Conservation Assessment for R. boylii in Oregon,
Olson and Davis (2007) proposed a variety of causes for the
decline of this species, including habitat loss, pollutants,
change in hydrologic regimes, and introduced species
including M. dolomieu. However, these potential factors in
the frog’s decline in Oregon remain unstudied.
We investigated possible mechanisms behind declines of
R. boylii in Cow Creek, Oregon, by experimentally assessing
their responses when exposed to M. dolomieu. We chose a
laboratory setting for our experiments in order to maintain
consistent environmental conditions between experimental
treatments, including the chemical composition of the
water, while eliminating a majority of the biotic and abiotic
variation that might confound interpretations of larval
behavior in the field. Amphibian larvae often ‘‘recognize’’
predators via waterborne chemical cues (Kats and Dill,
1998). Upon detection of a potential threat, larvae often
respond with antipredator behaviors such as refuge use or a
reduction in activity levels (Petranka et al., 1987; Kiesecker
et al., 1996). We have observed both of these behaviors in R.
boylii during field observations; upon our approach they
often took cover by retreating under cobble, and have been
Environmental Science Graduate Program, 104 Wilkinson Hall, Oregon State University, Corvallis, Oregon 97331; E-mail: (DJP)
[email protected]; and (ARB) [email protected] Send reprint requests to ARB.
USDA Forest Service, Pacific Northwest Research Station, 3200 SW Jefferson Way, Corvallis, Oregon 97331; E-mail: [email protected]
Submitted: 11 September 2009. Accepted: 27 October 2010. Associate Editor: G. Haenel.
F 2011 by the American Society of Ichthyologists and Herpetologists
Copeia 2011, No. 1
Fig. 1. Past and present distributions of Rana boylii in Oregon, USA. Modified from Olson and Davis (2007). MCP 5 Minimum Convex Polygon of
species range from known site locations.
observed motionless when in close proximity to a native
predator, the Rough-skinned Newt (Taricha granulosa; Jones
et al., 2005). The newt appeared to be a stealthy predator,
slowly stalking its prey, although no studies of their
predation behavior under natural conditions are known to
us. It is possible that the more rapid predatory behavior of
introduced bass is novel to R. boylii.
We conducted Experiment I to determine if chemical cues
alone are sufficient for larvae of R. boylii to detect a potential
threat. Kats et al. (1994) showed that amphibian larvae may
change antipredator behaviors based on their developmental stage. To determine if these ontogenetic differences were
evident in R. boylii, Experiment Ia was conducted using
early-stage larvae, whereas Experiment Ib focused on the
responses of late-stage larvae. Because chemical cues may be
insufficient to detect a potential threat in the stream
environment in which larvae of R. boylii live, we designed
Experiment II to determine the role visual cues play in
response to predator stimuli. In a lotic environment,
amphibian larvae may need to use several modes of
detection in tandem to identify a potential threat. Experiment III was designed to expose larvae of R. boylii to
chemical, visual, as well as mechanical (i.e., water movement) cues of the stimulus animal concurrently.
We hypothesized that larvae of R. boylii would respond to
cues of a native predator by reducing activity levels and/or
increasing refuge use, whereas those exposed to cues of the
non-native, unfamiliar predator would display activity levels
similar to larvae exposed to control cues. Studies examining
the effects of introduced species on native amphibians are
numerous, but to our knowledge only one (Kupferberg,
1997) has been conducted investigating the effects of nonnative species (American Bullfrog, Lithobates catesbeianus,
formerly Rana catesbeiana) on Rana boylii. Our study is the
first attempt to determine empirically the effects of M.
dolomieu on larval behavior of R. boylii. Absence of an
appropriate antipredator response by larvae may directly
impact R. boylii where their populations overlap with M.
MATERIALS AND METHODS
Eggs of R. boylii were collected on 13 June 2008 from an
area populated by M. dolomieu (Cow Creek, Douglas Co., OR;
42u52930N, 123u349320W). On 24 June 2008, a second set of
eggs was collected from a bass-free area (Carberry Creek,
Jackson Co., OR; 42u4960N, 123u10970W; Fig. 2). Larvae from
these populations will be referred to as BASS and BASSLESS
Paoletti et al.—Antipredator behavior of Rana boylii
larvae may exhibit reduced activity levels, making it difficult
to discern between inactivity due to satiation or the
presence of a potential threat. In each experiment, food
was withheld from all animals 24–48 h prior to experimental
trials to avoid the effects of satiation. Blinds around test
tanks prevented disturbance of animals during experimental
Fig. 2. Location of foothill yellow-legged frog (Rana boylii) study site
and egg-collection sites at Cow Creek, Douglas County, and Carberry
Creek, Jackson County, Oregon, USA.
hereafter. Eggs were transported to Oregon State University
and maintained in aerated, 38-L aquaria filled with dechlorinated water. Animals were exposed to a natural 14L:10D
photoperiod at a constant temperature of 16uC. Upon
hatching, larvae were fed a mixture of fish flakes (TetraFin)
and ground rabbit chow ad libitum. Source animals for the
stimulus cues were collected from Cow Creek and tributaries
of the South Umpqua River. These included: T. granulosa,
Speckled Dace (Rhinichthys osculus), and M. dolomieu. We
compared larval responses between treatments, exposing
them to native and non-native predators, and to R. osculus, a
non-predatory native minnow which served as our ‘‘positive
fish control.’’ All animals were maintained in the conditions
listed above at Oregon State University. Rhinichthys osculus
were fed fish flakes and tubifex worms; T. granulosa and M.
dolomieu were fed earthworms and crickets ad libitum. Sated
Experiment Ia: Detection via chemical cues by early-stage
larvae.—Early-stage larvae (Stage 20–24; Gosner, 1960) were
exposed to one of four chemical cue treatments: Control (no
cue); Positive Control (non-lethal, native R. osculus); Native
Predator (T. granulosa); and Non-native Predator (M. dolomieu). Individual larvae (n 5 104) were randomly assigned to
one of the four treatments and were not re-used in trials to
address this objective. In addition, larvae from BASS and
BASSLESS populations were alternately tested. There were a
total of 13 larvae for each population (13 replicates 3 4
treatments 5 52 animals tested per population 5 104 total).
Chemical cues were supplied through a series of gravitational ‘‘flow-through’’ tanks (Fig. 3; similar to those in
Petranka et al., 1987) and were designed to simulate the
stream environment in which these animals are found.
Three clear plastic tanks were arranged in a linear fashion, in
sequentially descending heights. Filtered, dechlorinated
water was supplied via a 144-L source tank. Water was
gravity-fed through 2.5 cm polyvinyl chloride (PVC) tubing
into a ‘‘treatment tank’’ (23 3 37 3 16 cm). This tank held
the animal responsible for supplying the chemical cue
treatment. From here, water was gravity-fed to a third tank,
the ‘‘experimental tank’’ (23 3 37 3 16 cm), which held a
single frog larva. A 5-cm (2.5 cm diameter) piece of black
plastic pipe was affixed to the bottom of the tank to provide
a standardized potential refuge for larvae. A drain pipe was
inserted to maintain a constant water height of 11 cm (mean
flow rate 5 2.0 L/min). Four of these systems ran
concurrently, with each of the four treatment tanks
supplying one of the four cues. The cue assigned to each
treatment tank was randomly assigned. Experimental tanks
were rinsed thoroughly between each replication.
Prior to running each trial, animals supplying the cue
were placed in the treatment tank, and a randomly selected
larva was placed in the experimental tank and allowed to
acclimate for ten minutes. After this the valve from the
source tank was slowly opened, followed by the treatment
Fig. 3. Design of flow-through tanks addressing the potential response of larvae of Rana boylii from Cow Creek, Oregon, USA, to chemical cues of
native and non-native predators, and native non-predators. Animals supplying chemical cues were placed in the treatment tank, and larvae were
placed in the experimental tank.
Copeia 2011, No. 1
tank valve. Observations began immediately and locations of
larvae were recorded during ‘‘spot-checks’’ conducted every
two minutes for ten minutes. A grid divided into six quadrants
(10 3 10 cm each) was placed under each experimental tank
to track movement (Garcia et al., 2009). Grids were designed
to be large relative to larval body size to ensure that any
movement captured was deliberate. For example, an earlystage larva would have to travel the equivalent of ten bodylengths in order to cross one gridline, therefore providing a
highly conservative representation of activity levels. Furthermore, this grid size approximates larval movements observed
in their natural environment. At each observation we
recorded the location of each larva, whether there was
movement, number of gridlines crossed following their
previous position, and whether or not they were utilizing
the refuge. The experiment was conducted over three days.
populations were alternately tested by exposing them to one
of four different treatments (Control, Dace, Newt, or Bass).
There were eight replicates for each population (8 replicates
3 4 tanks 5 32 animals tested per population 5 64 total).
Orientation of tanks was randomly assigned, as was the
treatment. Treatment animals were randomly selected and
placed in their pre-designated half of the tank. Following a
ten-minute acclimation period, larvae were gently placed in
the center of their half and observations began immediately
every two minutes for ten minutes. We recorded the
location of each larva, whether there was movement,
number of gridlines crossed following their previous
position, and whether or not they were utilizing the refuge.
Ten individuals used in this study were randomly selected
from larvae used in previous experiments. The experiment
was run over two days.
Experiment Ib: Detection via chemical cues by late-stage
larvae.—Methods were identical to those listed above, using
late-stage larvae (Stage 30–37; Gosner, 1960). Over 70% of
these individuals were stages 34 to 36. There were nine
replicates for each population (9 replicates 3 4 tanks 5 36
animals tested per population 5 72 total).
Statistical analysis.—We examined initial and overall responses of larvae to treatments. Initial responses were
activity data from the first two observations (the first four
min of a trial); overall responses were data from all five
observations over ten minutes. The sum of the number of
gridlines crossed across observations for a single larva was
used as an indicator of activity level in analyses. For all
experiments, we tested for differences in activity levels
between the four treatment groups using a full generalized
linear model (GLM) examining the effects of population,
treatment, and their interaction. A GLM was selected to
account for non-normality of the observations and potential heteroscedasticity. A Poisson distribution was applied
when data appeared non-normal; otherwise a Gaussian
distribution was used. Drop in deviance Chi-square tests
were used to test the statistical significance of population,
treatment, and their interaction. One-way ANOVAs and
Welch modified two-sample t-tests were applied to population and treatment means to assess which populations or
treatments were responsible for significant main effects
indicated by the GLM. In Experiment II, the proportion of
the overall 10 min time spent in each section was analyzed
using a Contingency Table Analysis, with the Control
treatment used as our baseline for comparison. Tank size
and grid size varied among experiments, and the chemical
cue concentrations in Experiment I may have differed from
those in Experiment III. Hence, each experiment should be
considered independently. Analyses were performed using
S-Plus 8.0 statistical software (Insightful Corp.; http://www.
Experiment II: Detection via visual cues.—We used four pairs of
9.5-L glass tanks, with an opaque divider placed between
each pair. Glass tanks were used instead of plastic tanks from
the previous experiment to allow for maximum visibility.
Beneath one of each pair was placed a six-square grid (4.5 3
10.5 cm each) which allowed the observer to track tadpole
movement. In addition to these gridlines, the treatment
tank was divided into three sections (4.5 3 21 cm each)
situated relative to the treatment tank—Near Section,
Middle Section, and Far Section. Late-stage larvae (stages
33–40; Gosner, 1960) from the BASSLESS population were
exposed to one of four different visual cues: Control (no
animal); Positive Control (native R. osculus); Native Predator
(T. granulosa); and Non-native Predator (M. dolomieu). A total
of 40 larvae were tested (10 replicates 3 4 tank-pairs 5 40
animals), none of which were re-used within the study.
All animals were randomly assigned to a treatment, as well
as their position within each of the four pairs (i.e., left- or
right-hand tank). Randomly selected treatment animals and
larvae were placed into tanks and allowed 15 minutes to
acclimate. Opaque dividers were then slowly removed from
in between each pair of tanks to allow animals to see one
another, and observations began immediately. Observations
were taken every two minutes for ten minutes. During every
observation, we recorded within which section each larva
was located, whether there was movement, and number of
gridlines crossed from their previous position. The experiment was conducted over one day.
Experiment III: Detection via chemical/visual/mechanical cues.—
This experiment was designed to reflect the same exposure
to cues larvae might receive in the side pools of streams in
which they are often found. Four plastic tanks (50 3 32 3
14 cm) were divided in half width-wise by fiberglass mesh to
separate larvae from the treatment animal. A 5-cm (2.5 cm
diameter) piece of black plastic pipe was affixed to the
bottom of the larva half to provide potential refuge. A ninesquare grid (8.3 3 10.7 cm) was positioned under the larva
half to monitor movement (or lack thereof). Late-stage
larvae (Stage 31–37; Gosner, 1960) from BASS and BASSLESS
In Experiment Ia, examining the responses of early-stage
larvae to chemical cues, analyses of initial responses
revealed no differences among treatments (df 5 3, P 5
0.38), but in analyses of overall responses, we found an
interaction between population and treatment (Poisson
distribution, df 5 3, P 5 0.06). Follow-up one-way ANOVAs
indicated a difference at the treatment level (df 5 3, P 5
0.05; Fig. 4A) but not between the two populations (df 5 1,
P 5 0.38). However, no significant differences between pairs
of treatments were detected (post hoc Welch t-tests, 6
comparisons conducted, P . 0.17).
In Experiment 1b, where we examined late-stage larvae
responses to chemical cues, we found no differences among
treatments initially (df 5 3, P 5 0.80) or overall (df 5 3, P 5
0.47). However, in post hoc analyses, when all treatment data
Paoletti et al.—Antipredator behavior of Rana boylii
Fig. 4. Average number of gridlines crossed per treatment for each experiment testing the overall response of larvae of Rana boylii from Oregon,
USA, to potential native predators (Rough-skinned Newts, Taricha granulosa), non-native predators (Smallmouth Bass, Micropterus dolomieu), and
non-predatory native fish (Speckled Dace, Rhinichthys osculus). Larvae from locations with bass (BASS) and without bass (BASSLESS) were tested.
(A) Experiment Ia: Detection via chemical cues by early-stage larvae. (B) Experiment Ib: Detection via chemical cues by late-stage larvae. (C)
Experiment II: Detection via visual cues. (D) Experiment III: Detection via chemical/visual/mechanical cues. When exposed to all sensory cues, larvae
reduced activity levels when exposed to T. granulosa.
were combined, late-stage larvae from the BASSLESS population moved more often (Welch t-test, t 5 22.52, df 5 70, P
5 0.01), and traveled greater distances (Welch t-test, t 5
22.53, df 5 70, P 5 0.01), than individuals from the BASS
population (Fig. 4B).
In Experiment II, examining visual cues, we found initial
responses differed among treatments (df 5 3, P 5 0.03; Fig. 5).
Post hoc t-tests revealed differences in number of gridlines
crossed between larvae exposed to Bass (mean 5 2.7) versus
the Control (mean 5 1.1; t 5 2.73, P 5 0.01), as well as Bass
(mean 5 2.7) versus Dace (mean 5 1.0; t 5 2.61, P 5 0.02). In
analyses of overall responses, we found no differences in
activity levels among treatments (df 5 3, P 5 0.14; Fig. 4C).
Compared to the Control treatment, larvae did not exhibit a
preference in tank Section when visually exposed to the other
three treatments (x20.05,2, P . 0.25 for all tests).
In Experiment III, where we examined a combination of
stimuli, we did not detect an initial response (df 5 3, P 5
0.55), but overall, we found that when larvae of R. boylii were
exposed to T. granulosa, the native predator, larvae from both
BASS and BASSLESS populations responded strongly by
decreasing both the frequency of movements (F3,60 5 6.02,
P , 0.01) and the average number of gridlines crossed (F3,60 5
7.77, P , 0.01; Fig. 4D). Larvae exposed to Control, Dace, and
Bass cues responded similarly to one another, exhibiting
higher activity levels relative to larvae exposed to T. granulosa.
Throughout all trials, only two larvae were observed utilizing
refuge; therefore data on refuge use were disregarded.
We found no behavioral differences among treatments
when larvae of R. boylii were exposed to chemical cues only
(Experiments Ia and Ib). The overall activity level differences
in Experiment Ia were influenced by larvae exposed to Bass
Fig. 5. Experiment II: Initial response in average number of gridlines
crossed by late-stage (Stage 33–40; Gosner 1960) larvae of Rana boylii
from Oregon, USA, to visual cues of potential native predators (Roughskinned Newts, Taricha granulosa), non-native predators (Smallmouth
Bass, Micropterus dolomieu), and non-predatory native fish (Speckled
Dace, Rhinicththys osculus).
cues. However, our follow-up analyses revealed no significant relationship. In some systems, chemical cues alone are
often sufficient for predator detection in amphibians
(Petranka et al., 1987; Kiesecker et al., 1996). However,
many of these studies focused on amphibian larvae from
lentic environments (but see Sih and Kats, 1994; Jowers
et al., 2006). The stream environment in which larvae of
R. boylii live could moderate their reliance on chemical cues
alone to detect potential threats.
Although tadpole activity levels did not differ at the
treatment level in Experiment Ib, late-stage larvae from the
BASSLESS population exhibited higher activity levels overall
relative to late-stage larvae from the BASS population. Eggs
from the BASSLESS population were collected from a higher
elevation (707 m) compared with those from the BASS
population (273 m), which likely results in a shorter larval
period. Since larvae of R. boylii do not overwinter, increased
activity levels may be necessary at higher elevations to
acquire the resources necessary to speed development and
metamorphose prior to the onset of cooler weather. These
results may reveal population-level differences in life history
and behavior that were previously unknown for the species.
In Experiment II, larvae exhibited an initial response to
visual cues within the first few minutes of the experiment,
but not an overall response. The initial response was
reflected as increased activity of the Newt- and Bass-exposed
larvae. This is not surprising given that both M. dolomieu and
T. granulosa were occasionally observed behaving aggressively towards larvae (for example, T. granulosa and M.
dolomieu followed larvae and lunged at them). Over the
longer course of the observation period, this response
dampened, suggesting that visual cues alone were not a
sustained stimulus for antipredator behaviors. Our results
agree with those of similar studies examining the role of
visual cues for predator detection by amphibian larvae
(Stauffer and Semlitsch, 1993; Kiesecker et al., 1996; Hickman et al., 2004).
Results of Experiment III, examining combined sensory
cues, indicated that multiple cues may be necessary to elicit
a sustained antipredator response in larvae of R. boylii.
Larvae from both BASS and BASSLESS populations significantly reduced their overall activity levels when placed in
the same tank as their native predator, T. granulosa.
Micropterus dolomieu readily detect motion to capture prey
(Sweka and Hartman, 2003), and we observed M. dolomieu
lunging at larvae during several observations. However,
those larvae placed in the same tank as M. dolomieu did not
show any difference in activity levels relative to controls
which supports our original prediction that larvae exposed
to the cues of a non-native, unfamiliar predator would
display activity levels similar to larvae exposed to control
cues. Therefore, the failure to reduce activity levels to avoid
detection in the presence of M. dolomieu could increase
predation events on larvae of R. boylii.
It is unclear why larvae did not utilize refuge during trials.
In the field we regularly witnessed larvae of R. boylii
retreating between stones and under cobble upon our
approach. We can only surmise that larvae avoided artificial
refugia used in laboratory trials due to some quality of the
material used or the tube-design of the refugia. In future
studies, the use of natural cover objects could be tested.
The ability to adapt behaviorally to an unfamiliar predator
is contingent upon a range of factors such as time spent in
sympatry, experience, amount of predation pressure, and
Copeia 2011, No. 1
the behavioral plasticity of the species. Our findings
contribute to the growing body of work examining the
responses of native amphibians to introduced predators.
Analogous studies examining other native U.S. Pacific
Northwest amphibians suggest that species differ greatly in
their responses to predators. For instance, 10 of the 21
studies we reviewed (Paoletti, 2009) reported no response to
the chemical cues of an introduced predator. Our results
suggest that after approximately 40 years of coexistence
(approx. 20 generations; Hayes et al., 2005), R. boylii lack the
ability to respond to M. dolomieu as a threat. On the other
hand, Kiesecker and Blaustein (1997), found that in
approximately 30 generations (McAllister and Leonard,
2005), larvae of the Red-legged Frog (Rana aurora) developed
the ability to detect and avoid non-native American
Bullfrogs (L. catesbeianus), a potential predator. The behavioral responses witnessed in this study emphasize the
difficulty in predicting how a naı̈ve, native species might
respond to an unfamiliar predator. In the brief 40 years the
two species have coexisted, M. dolomieu may have not
exerted selection pressure strong enough to modify behavioral patterns of R. boylii. Given the chance, M. dolomieu will
readily consume larvae and adults of R. boylii (pers. obs.).
However, we do not know whether R. boylii is the preferred
prey or simply an opportunistic food source.
Several studies have shown that amphibian larvae respond
to the chemical cues of injured conspecifics, or those
chemicals emitted from predators that have consumed
conspecifics (Laurila et al., 1997; Chivers and Mirza, 2001).
Like many ranids, larvae of R. boylii are found in loose
aggregations until metamorphosis (pers. obs.). It is possible
that an individual cannot maintain the vigilance necessary
to detect predators, and so rely on the alarm cues produced
by one or a few group members to detect a predatory threat.
Our study focused on the response of an individual tadpole
to the cues of the treatment animal, without the aid of
group effects or supplemental dietary cues produced by
predators. In addition, our study monitored larval behavior
in an artificially homogeneous habitat. The transmission
and detection of chemical, visual, and mechanical cues may
be altered by the physically heterogeneous habitat in which
larvae are normally found. Therefore, the strength of the
responses elicited by stimulus animals in the experiments
reported herein should be considered highly conservative.
Our results indicate that M. dolomieu are capable of directly
affecting populations of R. boylii in areas where they overlap.
Removal of non-native fish has led to recovery of local
amphibian populations in some areas (Hoffman et al., 2004;
Vredenburg, 2004). Despite being recognized as an introduced species by Pacific Northwestern state agencies, M.
dolomieu remains an actively managed sport fishery. Because
of this and the difficulty involved with removal, eradication
is not a feasible option. If larvae of R. boylii are unable to avoid
predation by an unfamiliar predator, as our findings show,
then local populations may disappear before they are able to
adapt. It is therefore critical to understand the behavioral
responses of native wildlife when faced with a novel predator.
Developing the ability to predict these responses could aid
fisheries managers when developing stocking plans to
mitigate amphibian declines at the regional level.
We would like to thank D. Simon, D. Markle, C. Rombough,
G. Weaver, C. Baldwin, M. Kluber, and T. Garcia for their
Paoletti et al.—Antipredator behavior of Rana boylii
assistance. Financial support for this project was provided by
the USDA Forest Service, Pacific Northwest Research Station
(Corvallis, Oregon), and Oregon State University. These
experiments comply with the current laws of the United
States and with Oregon State University animal care
regulations (ACUP #3732). Animals were collected according to Oregon Department of Fish and Wildlife regulations
(permit #068-08 and permit #OR2008-4516).
Alford, R. A., and S. J. Richards. 1999. Global amphibian
declines: a problem in applied ecology. Annual Review of
Ecology, Evolution, and Systematics 30:133–165.
Blaustein, A. R., and D. B. Wake. 1990. Declining amphibian populations: a global phenomenon. Trends in
Ecology and Evolution 5:203–204.
Bradford, D. F., F. Tabatabai, and D. M. Graber. 1993.
Isolation of remaining populations of the native frog,
Rana muscosa, by introduced fishes in Sequoia and Kings
Canyon National parks, California. Conservation Biology
Chivers, D. P., and R. S. Mirza. 2001. Importance of
predator diet cues in responses of larval wood frogs to fish
and invertebrate predators. Journal of Chemical Ecology
Fellers, G. M., and C. A. Drost. 1993. Disappearance of the
cascades frog (Rana cascadae) at the southern end of its
range, California, USA. Biological Conservation 65:177–
Garcia, T. S., D. J. Paoletti, and A. R. Blaustein. 2009.
Correlated trait response: comparing amphibian defense
strategies across a stress gradient. Canadian Journal of
Gosner, K. L. 1960. A simplified table for staging anuran
embryos and larvae with notes on identification. Herpetologica 16:183–190.
Hayes, M. P., S. J. Kupferberg, and A. J. Lind. 2005. Foothill
yellow-legged frog: Rana boylii Baird, p. 182–185. In:
Amphibians of the Pacific Northwest. L. L. C. Jones,
W. P. Leonard, and D. H. Olson (eds.). Seattle Audubon
Society, Seattle, Washington.
Hickman, C. R., M. D. Stone, and A. Mathis. 2004. Priority
use of chemical over visual cues for detection of predators
by graybelly salamanders, Eurycea multiplicata griseogaster.
Hoffman, R. L., G. L. Larson, and B. Samora. 2004.
Responses of Ambystoma gracile to the removal of introduced nonnative fish from a mountain lake. Journal of
Jones, L. L. C., W. P. Leonard, and D. H. Olson (eds.).
2005. Amphibians of the Pacific Northwest. Seattle
Audubon Society, Seattle, Washington.
Jowers, M. J., R. Campell-Palmer, P. T. Walsh, and J. R.
Downie. 2006. Intraspecific variation in the avoidance
response of stream frog (Mannophryne trinitatis) tadpoles to
fish and prawn predators. Herpetological Journal 16:337–346.
Kats, L. B., J. A. Breeding, K. M. Hanson, and P. Smith.
1994. Ontogenetic changes in California newts (Taricha
torosa) in response to chemical cues from conspecific
predators. Journal of the North American Benthological
Kats, L. B., and L. M. Dill. 1998. The scent of death:
chemosensory assessment of predation risk by prey
animals. Ecoscience 5:361–394.
Kats, L. B., and R. P. Ferrer. 2003. Alien predators and
amphibian declines: review of two decades of science and
the transition to conservation. Diversity and Distributions
Kiesecker, J. M., and A. R. Blaustein. 1997. Population
differences in responses of red-legged frogs (Rana aurora)
to introduced bullfrogs. Ecology 78:1752–1760.
Kiesecker, J. M., A. R. Blaustein, and C. L. Miller. 2001.
Transfer of a pathogen from fish to amphibians. Conservation Biology 15:1064–1070.
Kiesecker, J. M., D. P. Chivers, and A. R. Blaustein. 1996.
The use of chemical cues in predator recognition by
western toad tadpoles. Animal Behaviour 52:1237–1245.
Knapp, R. A., and K. R. Matthews. 2000. Non-native fish
introductions and the decline of the mountain yellowlegged frog from within protected areas. Conservation
Kupferberg, S. J. 1997. Bullfrog (Rana catesbeiana) invasion
of a California river: the role of larval competition.
Laurila, A., J. Kujasalo, and E. Ranta. 1997. Different antipredator behaviour in two anuran tadpoles: effects of predator
diet. Behavioral Ecology and Sociobiology 40:329–336.
Mack, R. N., D. Simberloff, W. M. Lonsdale, H. Evans, M.
Clout, and F. A. Bazzaz. 2000. Biotic invasions: causes,
epidemiology, global consequences, and control. Ecological Applications 10:689–710.
McAllister, K. R., and W. P. Leonard. 2005. Northern redlegged frog: Rana aurora Baird and Girard, p. 182–185. In:
Amphibians of the Pacific Northwest. L. L. C. Jones, W. P.
Leonard, and D. H. Olson (eds.). Seattle Audubon Society,
Muir, T. A., and P. T. Jenkins. 2002. Economic impacts of
exotics in the Great Lakes—a review. International
Association for Great Lakes Research Conference Program
and Abstracts 45:90–91.
Olson, D. H., and R. Davis. 2007. Conservation assessment for the Foothill Yellow-legged Frog (Rana boylii)
in Oregon, Version 1.0. USDA Forest Service, Region 6,
and Oregon BLM, Special Status Species Program. Available at: http://www.fs.fed.us/r6/sfpnw/issssp/species-index/
fauna-amphibians.shtml (accessed 19 June 2009).
Paoletti, D. P. 2009. Responses of Foothill Yellow-legged Frog
(Rana boylii) larvae to an introduced predator. Unpubl.
Master’s thesis, Oregon State University, Corvallis, Oregon.
Petranka, J. W., L. B. Kats, and A. Sih. 1987. Predator–prey
interactions among fish and larval amphibians: use of
chemical cues to detect predatory fish. Animal Behaviour
Pounds, J. A., M. R. Bustamente, L. A. Coloma, J. A.
Consuegra, M. P. L. Fogden, P. N. Foster, E. La Marca,
K. L. Masters, A. Merino-Viteri, R. Puschendorf, S. R.
Ron, G. A. Sánchez-Azofeifa, C. J. Still, and B. E. Young.
2006. Widespread amphibian extinctions from epidemic
disease driven by global warming. Nature 439:161–167.
Schade, C. B., and S. A. Bonar. 2005. Distribution and
abundance of nonnative fishes in streams of the Western
United States. North American Journal of Fisheries
Sih, A., and L. B. Kats. 1994. Age, experience, and the
response of streamside salamander hatchlings to chemical
cues from predatory sunfish. Ethology 96:253–259.
Simon, D. C., and D. F. Markle. 1999. Evidence of a
relationship between smallmouth bass (Micropterus dolo-
mieu) and decline of Umpqua chub (Oregonichthys kalawatseti) in the Umpqua Basin, Oregon. Northwestern
Stauffer, H. P., and R. D. Semlitsch. 1993. Effects of visual,
chemical and tactile cues of fish on the behavioural
responses of tadpoles. Animal Behaviour 46:355–364.
Stuart, S. N., J. S. Chanson, N. A. Cox, B. E. Young, A. S. L.
Rodrigues, D. L. Fischman, and R. W. Waller. 2004.
Status and trends of amphibian declines and extinctions
worldwide. Science 306:1783–1786.
Sweka, J. A., and K. J. Hartman. 2003. Reduction of reactive
distance and foraging success in smallmouth bass, Micropterus dolomieu, exposed to elevated turbidity levels.
Environmental Biology of Fishes 67:341–347.
Tyler, T., W. J. Liss, L. M. Ganio, G. L. Larson, R. Hoffman, E. Deimling, and G. Lomnicky. 1998. Interaction
Copeia 2011, No. 1
between introduced trout and larval salamanders (Ambystoma macrodactylum) in high-elevation lakes. Conservation Biology 12:94–105.
USGS. 2009. Nonindigenous Aquatic Species Database.
United States Geological Survey. Available at: http://nas.
er.usgs.gov/graphs/Group.asp?SpeciesGroup5Fishes (accessed 7 May 2009).
Vitousek, P. M., C. M. D’Antonio, L. L. Loope, M.
Rejmánek, and R. Westbrooks. 1997. Introduced species:
a significant component of human-caused global change.
New Zealand Journal of Ecology 21:1–16.
Vredenburg, V. T. 2004. Reversing introduced species
effects: experimental removal of introduced fish leads to
rapid recovery of a declining frog. Proceedings of the
National Academy of Sciences of the United States of