The effects of parasites and pathogens on host behaviors may be particularly important in predator-prey contexts, since few animal behaviors are more crucial for ensuring immediate survival than the avoidance of lethal predators in nature. We examined the effects of an emerging fungal pathogen of amphibians, Batrachochytrium dendrobatidis, on anti-predator behaviors of tadpoles of four frog species. We also investigated whether amphibian predators consumed infected prey, and whether B. dendrobatidis caused differences in predation rates among prey in laboratory feeding trials. We found differences in anti-predator behaviors among larvae of four amphibian species, and show that infected tadpoles of one species (Anaxyrus boreas) were more active and sought refuge more frequently when exposed to predator chemical cues. Salamander predators consumed infected and uninfected tadpoles of three other prey species at similar rates in feeding trials, and predation risk among prey was unaffected by B. dendrobatidis. Collectively, our results show that even sub-lethal exposure to B. dendrobatidis can alter fundamental anti-predator behaviors in some amphibian prey species, and suggest the unexplored possibility that indiscriminate predation between infected and uninfected prey (i.e., non-selective predation) could increase the prevalence of this widely distributed pathogen in amphibian populations. Because one of the most prominent types of predators in many amphibian systems is salamanders, and because salamanders are susceptible to B. dendrobatidis, our work suggests the importance of considering host susceptibility and behavioral changes that could arise from infection in both predators and prey.
Citation: Han BA, Searle CL, Blaustein AR (2011) Effects of an Infectious Fungus, Batrachochytrium dendrobatidis, on Amphibian Predator-Prey Interactions. PLoS ONE 6(2): e16675. https://doi.org/10.1371/journal.pone.0016675
Editor: Howard Browman, Institute of Marine Research, Norway
Received: October 6, 2010; Accepted: December 24, 2010; Published: February 2, 2011
Copyright: © 2011 Han et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funding sources for this work include the National Science Foundation Postdoctoral Research Fellowship in Bioinformatics and the Budweiser Conservation Fellowship to BAH; and the National Science Foundation Integrated Research Challenges in Environmental Biology Program (DEB0213851 and IBN9977063). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
In ecological communities, few responses are more important for immediate prey survival than contending with lethal predators. In this context, infectious agents such as parasites and pathogens can influence host anti-predatory behaviors with important consequences for predator-prey interactions. For example, several studies reveal that infection may lead to behavioral or physiological changes that make prey more conspicuous and therefore more vulnerable to predation , , . In examples where predators can acquire pathogens from prey, evidence exists that predators may decrease their risk of acquiring infection by avoiding infectious prey and/or selectively consuming non-diseased prey , , . Furthermore, theoretical investigations corroborate empirical conclusions that infection can act through both prey and predator behaviors to influence infection prevalence and host population dynamics , , , .
A novel amphibian pathogen provides an ideal system to examine how a generalist infectious agent may impact predator-prey interactions. Batrachochytrium dendrobatidis (Bd) ,  is a globally distributed fungal pathogen that infects a broad diversity of amphibian host species worldwide (lists of host species available at http://www.spatialepidemiology.net/bd/). Though associated with prominent declines and extinctions of some amphibian populations (reviewed in , ), infection is not invariably lethal. For example, while extirpation has occurred for multiple populations in the Sierra Nevada (Rana muscosa, ), and over 90% of stream-dwelling amphibian species disappeared as a result of Bd epizootics in Panama (e.g., ), populations of other host species persist with low infection prevalence and without symptoms of disease in other regions around the world (e.g., , , ). Even with low-level infection, there is mounting evidence that Bd alters a number of important host behaviors such as social aggregation, thermoregulation, and foraging activity in multiple species , , . However, despite evidence from other predator-prey systems suggesting that infection can greatly impact both host population and infectious disease dynamics, the role of Bd in higher level community interactions, such as predator-prey interactions, has not been well-explored (but see , ).
Predator-prey interactions are often governed by a variety of chemical and visual cues , . In aquatic environments, larval amphibians such as frog tadpoles can detect chemical cues that emanate directly from predators in addition to cues emitted from injured conspecifics (alarm cues). In combination with visual stimuli, these chemical signals provide prey with information about the risk of predation, facilitating appropriate behavioral responses , . Cues from predators that depend primarily on vision for prey capture may cause tadpoles to seek refuge, whereas cues from vibration- or motion-detecting predators may cause tadpoles to drastically decrease activity levels . The expression of anti-predator behaviors is threat-sensitive and context-dependent, but combinations of flight and refuge-seeking behaviors are common among many amphibian species , . Accurate detection and adequate behavioral responses to the threat of predation are critical for prey that must balance the costs associated with hiding or inactivity. For amphibian larvae, engaging in anti-predator behaviors decreases the time spent foraging for food or actively thermoregulating to maximize developmental rates which ultimately translate to a smaller size at metamorphosis and decreased fitness .
To examine whether Bd infection affects amphibian predator-prey interactions, we measured two important anti-predator behaviors, activity rate and refuge use, in a series of controlled laboratory experiments testing the responses of larvae of four host species (frogs: Anaxyrus boreas, Rana aurora, Rana cascadae, Pseudacris regilla) to Bd infection. In many temperate aquatic habitats, the dominant predators of tadpoles are other amphibian species, which are also susceptible to Bd infection. In particular, salamanders are among the most important tadpole predators in many amphibian communities  and have been found infected with Bd in the wild , . Since attacking and consuming infected prey could increase the risk of Bd transmission to these predators, salamanders might be expected to display different predation patterns based on the infection status of their tadpole prey. To examine this possibility we also conducted feeding trials to determine whether two salamander species (Ambystoma macrodactylum and Taricha granulosa) indeed consume potentially infectious prey; and to compare the rates of predation on infectious vs. uninfected tadpoles of R. aurora, R. cascadae, and P. regilla.
Bd does not infect amphibian embryos, presumably because they lack keratin, a protein which is required by Bd for growth, reproduction, and survival . To ensure infection-free animals, we collected partial clutches of several egg masses (>10) of R. aurora from permanent ponds in Monmouth, OR (Polk County, elevation 61 m; latitude/longitude: 44.84/-123.30). Several whole egg masses (>25) of P. regilla were collected from temporary ponds and wetlands in Corvallis, OR (Benton County, elevation 87 m; latitude/longitude: 44.56/-123.26). Partial clutches of several R. cascadae (>5) and A. boreas (>25) egg masses were collected from permanent water bodies in Deschutes County, OR (elevation 1951 m; latitude/longitude: 44.29/-121.55). Eggs were collected at early developmental stages 11–15  and reared in 37 L aquaria containing filtered water with aeration. Upon hatching, conspecifics from separate clutches were mixed between multiple 37 L aquaria at densities of approximately 100 tadpoles per tank. Complete water changes were conducted approximately every 7 days. Tadpoles were fed a 2∶1 ratio of ground alfalfa pellets and fish flakes ad libitum. All animals were kept at 14–16°C on a 14∶10 hr light:dark photoperiod for the duration of the experiment. At developmental stages 28–29 , tadpoles were moved into several 11 L aquaria at densities of 20 tadpoles each for inoculation with either Bd or a control treatment.
We used inoculation methods known to produce infection in larvae of these species . Culture plates (1% Tryptone and agar) containing Bd (JEL strain 215) were flooded with 15 mL of filtered water for 20 minutes to allow the discharge of infectious zoospores from sporangia. The liquid contents from 2 flooded culture plates were added to each of six 11 L aquaria containing 20 tadpoles each. The culture plates were submerged for 1–2 seconds in water to ensure that zoospores were transferred from the culture plate to the aquarium. The procedure was repeated with control culture plates (1% Tryptone and agar without Bd) for six additional 11 L aquaria containing 20 animals each. Using a hemacytometer we counted an average concentration of 6.18×106 zoospores/mL for three Bd culture plates. Tadpoles were exposed to treatments for 10 days and taken haphazardly from treatment aquaria for experiments. For brevity, we refer to tadpoles exposed to Bd as “Bd+” and tadpoles exposed to the control agar wash as “Bd−”. A subsample of Bd+ and Bd− tadpoles used in activity rate and refuge use trials were analyzed using real-time quantitative PCR techniques based on . DNA was extracted from excised mouthpart tissue from the following species (sample sizes analyzed from each treatment, Bd+ and Bd−): R. aurora (N = 5); A. boreas (N = 8); P. regilla (N = 6); R. cascadae (N = 8). We used a DNAeasy 96 well kit (Qiagen, Valencia, California) for DNA extraction, and quantified DNA concentration using a spectrophotometer (Nanodrop Technologies, Wilmington, Delaware). We used the ABI 7300 Real-time PCR system (Applied Biosystems, Foster City, California) for PCR reactions. Twenty five µL reactions contained: 5 µL of 20 ng/L template DNA and 20 µL of master mix (containing 900 nM forward and reverse primers, 125 nM MGB probe, and Taqman Master Mix). We obtained Batrachochytrium genome equivalent standards from D. Boyle  and included triplicates of each standard serially diluted on each plate (10−1, 100, 101, 102) and a duplicate of the high standard (103). Unknown samples were run in triplicate and values that differed by a coefficient of variation greater than 0.2 were rerun for greater accuracy. Values obtained from the real-time PCR reaction are mean BD zoospore genome equivalents (ge) per nanogram of excised mouthpart tissue. This measure accounts for differences in BD infection severity between species based on size alone (i.e., more mouthpart tissue containing more zoospores).
In feeding trial experiments, tadpole host tissues could not be retrospectively sampled to quantify infection severity, thus Bd+ and Bd− designations refer only to exposure status in feeding trials.
We chose predators that naturally co-occur in habitats from which eggs of the four host/prey species described above were collected. From low elevation ephemeral pools, we collected adult rough-skinned newts (Taricha granulosa; hereafter referred to as “Taricha”; mean snout-vent length of adults in this population: 69.4±1.4 mm (Biga et al, unpublished); Benton County, OR; elevation 87 m; latitude/longitude: 44.56/-123.26) as predators of R. aurora and P. regilla . From permanent water bodies at higher elevations, we collected larval long-toed salamanders (Ambystoma macrodactylum; hereafter referred to as “Ambystoma”), a common predator of R. cascadae  and A. boreas tadpoles (pers. obs.). The mean snout-vent length of larvae used in this experiment was 37.2±0.41 mm. Larvae were collected from Deschutes County, OR (elevation 1951 m; latitude/longitude: 44.29/-121.55). Taricha were held in 37 L aquaria, two individuals per tank. Six Ambystoma larvae were housed in each of several 37 L aquaria separated by mesh compartments (12×20×25 cm) within aquaria to prevent cannibalism . Since evidence is accumulating that many prey species can assess predation risk based on cues from predator diet , predators were fed a single species diet of tadpoles ad libitum for approximately one month prior to activity and refuge use experiments to accumulate species-specific chemical cues in the water. For example, one group of Ambystoma predators was fed a diet of only R. cascadae tadpoles for one month, while a separate group of Ambystoma was fed a strict diet of B. boreas for one month prior to trials on each of these prey species. Similarly, one group of Taricha predators was fed a strict diet of P. regilla, while another group was fed a strict diet of R. aurora for one month prior to trials on each of these prey species. Water containing predators contained a combination of predator cues and species-specific alarm cues from consumption of tadpoles of a single species, and we used water from these predator-holding tanks for our behavioral trials. We refer to this mixture of predator and tadpole alarm cues as “predator cue” hereafter. Neutral (control) cues were harvested from two 37 L aquaria containing approximately 28 grams of Tubifex worms in each tank (Tubifex tubifex) for all species except R. aurora. We chose Tubifex worms because they are not consumed by tadpoles, and are non-predatory detritivores that are common in many freshwater ecosystems . The neutral cue for R. aurora trials was filtered water (as in ) because Tubifex worms were unavailable during trials for R. aurora. Predators were starved for 5 days prior to feeding trials to standardize hunger levels among individual predators.
Activity rates and refuge use of tadpoles
To examine activity and refuge use, we filled opaque plastic chambers (31.5×11×20 cm) with 1 L filtered water. The chambers were placed atop grids visible through the bottom of the chambers (grid squares = 25 mm2). Refuges were constructed using black polyvinyl chloride pipe cut into segments and fitted inside 20 mL glass beakers to form a dark tunnel. A single refuge was placed haphazardly in one corner of each chamber. Chambers containing 1 refuge and 1 tadpole from the appropriate treatment (Bd+ or Bd−, and predator cue (+/−) were left overnight to acclimate before commencing the experiment. Our experiment employed a 2×2 factorial design with four treatment combinations for each host species tested separately: Bd+ with predator cue, Bd+ with neutral cue, Bd− with predator cue, Bd− with neutral cue. Treatments were randomly assigned to chambers and replicated 30 times for each species except R. aurora (20 replicates for predator cue treatments, 25 replicates for neutral cue treatments). 500 mL of predator or neutral cue was added to chambers and left for 60 minutes prior to data collection. Prior to trials, we observed tadpoles of each species near, underneath, or inside of refuges (pers. obs.) ensuring that tadpoles were not repelled by the presence of constructed refuges and able to use refuges if they sensed a threat.
We recorded the number of gridlines crossed and the number of times each tadpole used refuge in 30-second intervals over 4 hours for a total of 8 observations per tadpole. To ensure independence of activity data from refuge use data, we counted the number of times the tadpole used refuge during the 30 second period and recorded the number of gridlines crossed by the tadpole when it was not using refuge. Black curtains surrounded test chambers to prevent observers from disrupting the animals. All trials were conducted between 1000 and 1400 hours. We used 2-way analyses of variance for statistical analyses of the mean activity rate and refuge use of tadpoles from Bd+ and Bd− treatments for each species.
We tested whether two predators (Ambystoma and Taricha) consume infected prey, and whether predation rates differed between Bd+ and Bd− prey for three species (P. regilla, R. cascadae, R. aurora). Feeding trials were not conducted on A. boreas because tadpoles and Ambystoma predators were not available in sufficient numbers during the timeline of the experiment. Rectangular plastic tubs (capacity: 39.7 L, size: 0.16×0.86×0.42 m) were filled with approximately 7.6 L filtered water and acclimated to cold-room temperature (14–16°C) for a minimum of 4 hours. 10 tadpoles from either Bd+ or Bd− treatments were then added to each tub and allowed to acclimate to novel conditions overnight. At 0900 the next morning a single predator was added to each tub (Ambystoma for R. cascadae trials, and Taricha for R. aurora and P. regilla trials), and predation was allowed to occur without replacement of consumed prey. We recorded the number of tadpoles remaining in each tub every 60 minutes for a total of 240 minutes (4 observations). There were 7 replicates (predators) per treatment, and we used each predator only once.
We measured the number of tadpoles consumed by predators over time, and applied generalized linear model (GLM) analyses for each predator-prey combination, including both “time” and “predator ID” as sources of variation in our models. Models for each species identified a Poisson error structure and a log link function. The general model structure for all three predator-prey combinations was: Number tadpoles consumed ∼Bd treatment + Time + ID + (Bd treatment*Time).
Within feeding trials, the survival of each individual tadpole was not monitored over time. We therefore measured the time until at least half of the tadpoles (5 or more) had been consumed by the predator and fit Cox proportional hazard models to estimate how Bd affects the risk of at least half of the tadpoles being consumed by the predator. We also tested for differences between Bd+ and Bd− survival curves for each prey species. All analyses were implemented in the R statistical computing environment, version 2.11.1 , and survival analyses used the survival package .
Quantitative PCR reactions indicate that inoculation methods produced infection in tadpoles exposed to Bd in our experiments, and ensured that controls were not infected with Bd. There was variation in infection severity between species, with A. boreas showing the most severe infections (0.53 ge), followed by R. aurora (0.35 ge), P. regilla (0.11 ge), and R. cascadae (0.01 ge). None of the Bd− (control) tadpoles were infected with Bd.
Tadpole host behaviors in response to Bd and predator chemical cues varied across the four prey species tested. Rana aurora tadpoles exposed to predator cues were less active than those exposed to neutral cues. There was no effect of Bd on activity. There was also no effect of either Bd exposure or cue treatment on refuge use in R. aurora (Figure 1a,b). In A. boreas tadpoles there was a significant interaction between the Bd and cue treatments on activity rates, and significant main effects of Bd and cue treatments on refuge use (Table 1; Figure 1c,d). Tukey HSD post-hoc tests revealed that Bd+ A. boreas tadpoles exposed to predator cue were more active and used refuge more frequently compared to tadpoles in neutral cue treatments and compared to uninfected tadpoles. Neither activity rate nor refuge use differed between Bd or predator cue treatments for the remaining two species, P. regilla or R. cascadae (Table 1).
Rana aurora tadpoles decreased activity in the presence of chemical cues from Taricha predators. Bd+ Anaxyrus boreas tadpoles were more active and used refuge more frequently when exposed to chemical cues of Ambystoma predators compared to neutral chemical cues.
In feeding trials, predators consumed tadpoles from both Bd+ and Bd− treatments in all three prey species tested (P. regilla, R. cascadae, and R. aurora; Figure 2). More Bd+ R. aurora tadpoles were consumed by predators compared to controls (z = 2.94, p = 0.003), with variation between individual predators contributing to this effect (z = 2.53, p = 0.012). For R. cascadae tadpoles, there was a decrease in the total number of tadpoles consumed over time (z = −2.51, p = 0.012), but no effect of the Bd treatment. Predation of P. regilla was not explained by Bd treatment, Time, or variation between individual predators. Also, Bd*Time interactions were not significant for any of the three predator-prey combinations tested, indicating that predation rates were not influenced by Bd status of the prey. A two-sample power analysis shows the minimum detectable effect size (i.e., the estimated magnitude of difference between Bd+ vs. Bd− prey consumed) is 0.44 given our sample sizes for each predator-prey combination (power = 0.8, significance level = 0.05). This retrospective power analysis shows that our design was sufficient to detect ‘medium’ effects of predators on prey populations (sensu ), but suggests that larger sample sizes may increase the ability to detect smaller effect sizes.
The symbol (+) at the end of survival curves for P. regilla and R. aurora denotes censored data, where some prey survived beyond the end of the feeding trial. All R. cascadae tadpoles were consumed by Ambystoma predators during feeding trials.
Prey survival curves did not differ significantly between Bd treatments for any prey species (R. aurora, X2 = 0.5, df = 1, p = 0.498; R. cascadae, X2 = 0.8, df = 1, p = 0.365; P. regilla, X2 = 0.2, df = 1, p = 0.644). Cox proportional hazards tests showed that Bd increased predation risk in R. aurora (by 1.8 times ±2.4), and decreased predation risk in P. regilla (by 0.75 times ±2.1) and R. cascadae (by 0.6 times ±1.8) compared to the control treatments, but these differences were not statistically significant (Table 2).
It is clear that the interactions between parasites, predators, and prey in an ecological community are complex, and may be an important driver of infection and population dynamics in many systems , . As a first investigation of the potential impacts of a widespread infectious fungus on predator-prey behavioral interactions in amphibian communities, we examined how two common anti-predator behaviors in four species of prey were affected by predator chemical cues and exposure to Bd. We also examined whether predators respond to infection risk by consuming uninfected prey more frequently compared to infected prey. Our results showed clear interspecific differences in the anti-predator behaviors of amphibian larvae in response to chemical cues of predators. We found that Bd exposure changed anti-predator behaviors in the tadpoles of one species, A. boreas, but not in the other three species we examined (P.regilla, R. aurora, and R. cascadae). Feeding trials revealed that salamander predators (Taricha and Ambystoma) consumed both Bd+ and Bd− tadpole prey of these latter three species at similar rates, and that predation risk among prey was not influenced by infection status.
Anaxyrus boreas tadpoles in Bd+ treatments reacted strongly to predator cues by increasing their activity rates and refuge use. A number of studies have demonstrated that A. boreas is particularly susceptible to Bd infection at both larval and post-metamorphic life stages (e.g., , ), and corroborate our observations of hyperactivity among infected tadpoles . Although we were unable to formally quantify this behavior, we note that hyperactivity may more accurately be described as erratic and unpredictable swimming behavior characterized by darting quickly around the test chamber and visiting refugia seemingly randomly during these bouts of erratic activity. We speculate that hyperactivity/erratic activity of Bd+ A. boreas tadpoles exposed to predator cues may arise through a synergism between incidental physiological effects of Bd infection and a behavioral response to predation risk. Tadpoles with increased activity rates that are also frequenting available refugia can decrease predation risk by making capture mechanically more difficult for predators , . Moreover, large aggregations of A. boreas that typically occur in nature ([e.g., 49]) can create a confusion effect that helps to evade predation (reviewed in , ), though we note that group behavior may differ from that of individuals observed in a laboratory setting. In contrast, Bd− tadpoles exposed to predator chemical cues may reflect behaviors similar to the other two species we examined (P. regilla, R. cascadae), requiring more information before responding with potentially costly anti-predator behaviors (e.g., ).
Based on several previous studies (reviewed in ), we expected all four species to respond to predator chemical cues by decreasing activity rates and hiding more frequently. However, only two of the species we examined (R. aurora and A. boreas) fulfilled these expectations. The other two species (P. regilla, R. cascadae) showed no detectable differences in activity or refuge use in response to predator chemical cues. The anti-predator behaviors that we measured, reducing activity and utilizing refugia, present significant costs to larval amphibians by decreasing the amount of time tadpoles spend foraging and actively thermoregulating to maximize growth rates. In ephemeral habitats, long periods of inactivity or hiding can depress growth and developmental rates, ultimately decreasing fitness , . To reduce the costliness of these anti-predator behaviors, prey assess predation risk through multiple sources of information, using both visual and chemical cues in the environment . Thus, one possibility is that the latter two species required more information signaling predation threat to elicit any detectable changes in anti-predator behaviors (e.g., more concentrated and/or longer exposure to predator chemical cues, or the addition of visual stimuli , ).
Amphibian predators (including Ambystoma and Taricha species) are susceptible to Bd infection , , . Thus, Bd transmission to salamander predators through consuming infected prey is a realistic possibility that may contribute to infection dynamics in amphibian communities. To our knowledge, this possibility remains unexplored. The results from feeding trials suggest that neither Taricha nor Ambystoma salamander species we tested avoid consuming infected tadpole prey. On the contrary, predators consumed tadpoles from both Bd+ and Bd− treatments at statistically indistinguishable rates, and showed no differences in the total number of tadpoles consumed among the three prey species tested (P. regilla, R. aurora, R. cascadae). In addition, neither the rate nor the risk of predation differed across Bd treatments among the three predator-prey combinations we examined. These results suggest that selective predation (detectable through differences in predation rates on tadpoles among Bd treatments) may not be acting on this system, although it is important to note that the low infection levels found in R. cascadae may partially explain why Ambystoma predators are non-selective on this prey species. In contrast to a growing number of studies investigating the impacts of selective predation on infectious disease dynamics [e.g., 7,59], the ecological consequences of non-selective predation have not been examined widely. However, recent theoretical works suggest that even when predators do not select between hosts on the basis of infection status, equilibrium disease prevalence can increase in a population in a number of ecologically plausible scenarios, such as when infection is non-regulatory (i.e., when it does not influence population dynamics through host demography) . Furthermore, selective predation may not allow for long-term persistence of prey populations that are also affected by infectious disease , implying the possibility that non-selective predation may be one mechanism contributing to the persistence of multiple host species and a shared infectious agent. We suggest that more detailed examinations of the patterns of selective vs. non-selective predation in natural communities may shed light on why some infected amphibian populations persist while others go extinct. Extending the patterns observed in our study to any persistent effects on prey populations and predator-prey interactions in natural systems will require closer examination of the degree of predator selectivity per se (e.g., through tests where predators make a distinct choice between consuming infected vs. uninfected prey); and the identification of additional infection-related drivers of prey population dynamics, such as how Bd infection affects predator as well as prey behaviors in semi-natural settings (e.g., capture efficiency of predators, and escape efficiency of prey).
Population declines of frogs and toads associated with Bd infection are precipitating important long-term changes to natural communities (e.g., ecosystem dynamics and community structure , ). Though Bd has often been considered a relatively superficial and non-lethal infection for larvae of most amphibian host species (but see ), it is clear that sub-lethal infection is sufficient to alter even basic behaviors such as schooling, thermoregulation , and foraging efficiency  in multiple species. Our study provides a timely first examination of how non-lethal Bd exposure can induce important behavioral changes contributing to predator-prey interactions in some host species (e.g., Anaxyrus boreas); and highlights an unexplored possibility that this pathogen may affect infection prevalence and community dynamics through insidious effects on amphibian predator-prey interactions.
Authors thank J Ng, H Lee, L Linn, J Hubbard, R Hill, T Young, J Romansic, B Bancroft, D Paoletti, J Tyburczy, and L Biga for assistance with this study, and L Pintor, D Kramer, and S Carver for helpful discussion. This study was carried out in accordance with guidelines set by the Institutional Animal Care and Use Committee (IACUC) at Oregon State University under approved proposal numbers 3034 and 3181.
Conceived and designed the experiments: BAH. Performed the experiments: BAH CLS. Analyzed the data: BAH. Contributed reagents/materials/analysis tools: BAH ARB. Wrote the paper: BAH.
- 1. Barber I, Huntingford FA (1996) Parasite infection alters schooling behavior: deviant positioning of helminth-infected minnows in conspecific groups. Proceedings of the Royal Society of London, Series B: Biological Sciences 263: 1095–1102.
- 2. Bakker TCM, Mazzi D, Zala S (1997) Parasite-induced changes in behavior and color make Gammarus pulex more prone to fish predation. Ecology 78: 1098–1104.
- 3. Hudson PJ, Dobson AP, Newborn D (1992) Do parasites make prey vulnerable to predation? Red grouse and parasites. Journal of Animal Ecology 61: 681–692.
- 4. Roy S, Chattopadhyay J (2005) Disease-selective predation may lead to prey extinction. Mathematical Methods in the Applied Sciences 28: 1257–1267.
- 5. Jones GA, Sieving KE, Avery ML, Meagher RL (2005) Parasitized and non-parasitized prey selectivity by an insectivorous bird. Crop Protection 24: 185–189.
- 6. Al-Zyoud F, Sengonca C (2004) Prey consumption preferences of Serangium parcesetosum Sicard (Col., Coccinelidae) for different prey stages, species and parasitized prey. Journal of Pest Science 77: 197–204.
- 7. Packer C, Holt RD, Hudson PJ, Lafferty KD, Dobson AP (2003) Keeping the herds healthy and alert: implications of predator control for infectious disease. Ecology Letters 6: 797–802.
- 8. Hudson PJ, Dobson AP, Newborn D (1998) Prevention of population cycles by parasite removal. Science 282: 2256–2258.
- 9. Roy M, Holt RD (2008) Effects of predation on host-pathogen dynamics in SIR models. Theoretical Population Biology 73: 319–331.
- 10. Holt RD, Roy M (2007) Predation can increase the prevalence of infectious disease. The Americal Naturalist 169: 690–699.
- 11. Longcore JE, Pessier AP, Nichols DK (1999) Batrachochytrium dendrobatidis gen. et sp. nov., a chytrid pathogenic to amphibians. Mycologia 91: 219–227.
- 12. Berger L, Speare R, Daszak P, Green DE, Cunningham AA, et al. (1998) Chytridiomycosis causes amphibian mortality associated with population declines in the rain forests of Australia and Central America. Proceedings of the National Academy of Sciences 95: 9031–9036.
- 13. Fisher MC, Garner TWJ, Walker SF (2009) Global emergence of Batrachochytrium dendrobatidis and amphibian chytridiomycosis in space, time, and host. Annual Review of Microbiology 63: 291–310.
- 14. Kilpatrick AM, Briggs CJ, Daszak P (2010) The ecology and impact of chytridiomycosis: an emerging disease of amphibians. Trends in Ecology & Evolution 25: 109–118.
- 15. Vredenburg VT, Knapp RA, Tunstall TS, Briggs CJ (2010) Dynamics of an emerging disease drive large-scale amphibian population extinctions. Proceedings of the National Acadamy of Sciences 107: 9689–9694.
- 16. Lips KR, Brem F, Brenes R, Reeve JD, Alford RA, et al. (2006) Emerging infectious disease and the loss of biodiversity in a Neotropical amphibian community. Proceedings of the National Academy of Sciences 103: 3165–3170.
- 17. Retallick RWR, McCallum H, Speare R (2004) Endemic infection of the amphibian chytrid fungus in a frog community post-decline. PLoS Biology 2: e351.
- 18. Kielgast J, Rodder D, Veith M, Lotters S (2010) Widespread occurrence of the amphibian chytrid fungus in Kenya. Animal Conservation 13: 36–43.
- 19. Rothermel BB, Walls SC, Mitchell JC, Dodd CK Jr, Irwin LK, et al. (2008) Widespread occurrence of the amphibian chytrid fungus Batrachochytrium dendrobatidis in the southeastern USA. Diseases of Aquatic Organisms 82: 3–18.
- 20. Venesky MD, Parris MJ, Storfer A (2010) Impacts of Batrachochytrium dendrobatidis infection on tadpole foraging performance. EcoHealth 6: 565–575.
- 21. Richards-Zawacki CL (2009) Thermoregulatory behaviour affects prevalence of chytrid fungal infection in a wild population of Panamanian golden frogs. Proceedings of the Royal Society of London, Series B: Biological Sciences 277: 519–528.
- 22. Han BA, Bradley PW, Blaustein AR (2008) Ancient behaviors of larval amphibians in response to an emerging fungal pathogen, Batrachochytrium dendrobatidis. Behavioral Ecology and Sociobiology 63: 241–250.
- 23. Parris MJ, Reese E, Storfer A (2006) Antipredator behavior of chytridiomycosis-infected northern leopard frog (Rana pipiens) tadpoles. Canadian Journal of Zoology 84: 58–65.
- 24. Parris MJ, Beaudoin JG (2004) Chytridiomycosis impacts predator-prey interactions in larval amphibian communities. Oecologia 140: 626–632.
- 25. Griffin AS (2004) Social learning about predators: a review and prospectus. Learning & Behavior 32: 131–140.
- 26. Lima SL, Dill LM (1990) Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology 68: 619–640.
- 27. Kats LB, Dill LM (1998) The scent of death: chemosensory assessment of predation risk by prey animals. Ecoscience 5: 361–394.
- 28. Chivers DP, Smith RJF (1998) Chemical alarm signalling in aquatic predator/prey systems: a review and prospectus. Ecoscience 5: 335–352.
- 29. Skelly DK, Werner EE (1990) Behavioral and life-historical responses of larval American toads to an odonate predator. Ecology 71: 2313–2322.
- 30. Duellman WE, Trueb L (1986) Enemies and defense. Biology of amphibians. Baltimore, , MD: Johns Hopkins University Press. pp. 241–260.
- 31. Padgett-Flohr GE, Longcore JE (2005) Ambystoma californiense: fungal infection Herpetological Review 36: 50–51.
- 32. Padgett-Flohr GE, Longcore JE (2007) Taricha torosa (California newt), fungal infection. Herpetological Review 78: 176–177.
- 33. Gosner KL (1960) A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica 16: 183–190.
- 34. Boyle DG, Boyle DB, Olsen V, Morgan JAT, Hyatt AD (2004) Rapid quantitative detection of chytridiomycosis (Batrachochytrium dendrobatidis) in amphibian samples using real-time Taqman PCR assay. Diseases of aquatic organisms 60: 141–148.
- 35. Calef GW (1973) Natural mortality of tadpoles in a population of Rana aurora. Ecology 54: 741–758.
- 36. Lanoo M (2005) Amphibian Declines: The Conservation Status of United States Species. University of California Press.
- 37. Wildy EL, Chivers DP, Kiesecker JM, Blaustein AR (1998) Cannibalism enhances growth in larval long-toed salamanders (Ambystoma macrodactylum). Journal of Herpetology 32: 286–289.
- 38. Chivers DP, Mirza RS (2000) Predator diet cues and the assessment of predation risk by aquatic vertebrates: a review and prospectus. In: Marchlewska-Koj A, Lepri JJ, Muller-Schwarze D, editors. Krakow, Poland: Springer. pp. 277–284.
- 39. Gilbert MA, Granath WO Jr (2003) Whirling disease of salmonid fish: life cycle, biology, and disease. Journal of Parasitology 89: 658–667.
- 40. Hews DK, Blaustein AR (1985) An investigation of the alarm response in Bufo boreas and Rana cascadae tadpoles. Behavioral and Neural Biology 43: 47–57.
- 41. Team RDC (2010) R: a language and environment for statistical computing Vienna, Austria: R Foundation for Statistical Computing.
- 42. Therneau T, Lumley T (2009) survival: survival analysis, including penalised likelihood. R package version 2.35-8. http://CRAN.R-project.org/package=survival.
- 43. Cohen J (1988) Statistical power analysis for the behavioral sciences. Hillsdale,, NJ: Lawrence Earlbaum Associates, Inc.
- 44. Anderson RM, May RM (1980) Infectious disease and population cycles of forest insects. Science 210: 658–661.
- 45. Blaustein AR, Romansic JM, Scheessele EA, Han BA, Pessier AP, et al. (2005) Interspecific variation in susceptibility of frog tadpoles to the pathogenic fungus Batrachochytrium dendrobatidis. Conservation Biology 19: 1460–1468.
- 46. Carey C, Bruzgul JE, Livo LJ, Walling ML, Kuehl KA, et al. (2006) Experimental exposures of Boreal toads (Bufo boreas) to a pathogenic chytrid fungus (Batrachochytrium dendrobatidis). EcoHealth 3: 5–21.
- 47. Hews DK (1988) Alarm response in larval western toads, Bufo boreas: release of larval chemicals by a natural predator and its effect on predator capture efficiency. Animal Behaviour 36: 125–133.
- 48. Niell SR, Cullen JM (1974) Experiments on whether schooling by their prety affects the hunting behaviour of cephalopods and fish predators. Journal of Zoology 172: 549–569.
- 49. O'Hara RK, Blaustein AR (1982) Kin preference behavior in Bufo boreas tadpoles. Behavioral Ecology and Sociobiology 11: 43–49.
- 50. Watt PJ, Nottingham SF, Young S (1997) Toad tadpole aggregation behaviour: evidence for a predator avoidance function. Animal Behaviour 54: 865–872.
- 51. Hoff Kv, Blaustein AR, McDiarmid RW, Altig R (1999) Behavior: interactions and their consequences. In: McDiarmid RW, Altig R, editors. Tadpoles: the biology of anuran larvae. Chicago: University of Chicago Press. pp. 215–239.
- 52. Parris MJ, Cornelius TO (2004) Fungal pathogen causes competitive and developmental stress in larval amphibian communities. Ecology 85: 3385–3395.
- 53. Martin J, Lopez P (1999) When to come out from a refuge: risk-sensitive and state-dependent decisions in an alpine lizard. Behavioral Ecology 10: 487–492.
- 54. Dupré RK, Petranka JW (1985) Ontogeny of temperature selection in larval amphibians. Copeia 1985: 462–467.
- 55. Mathis A, Vincent F (2000) Different use of visual and chemical cues in predator recognition and threat-sensitive predator-avoidance responses by larval newts (Notophthalmus viridenscens). Canadian Journal of Zoology 78: 1646–1652.
- 56. Bernard MF (2006) Survival trade-offs between two predator-induced phenotypes in Pacific treefrogs (Pseudacris regilla). Ecology 87: 340–346.
- 57. Ferrari MCO, Chivers DP (2009) Temporal variability, threat sensitivity and conflicting information about the nature of risk: understanding the dynamics of tadpole antipredator behaviour. Animal Behaviour 78: 11–16.
- 58. Davidson EW, Parris M, Collins JP, Longcore JE, Pessier AP, et al. (2003) Pathogenicity and transmission of chytridiomycosis in tiger salamanders (Ambystoma tigrinum). Copeia 2003: 601–607.
- 59. Duffy M (2007) Selective predation, parasitism, and trophic cascades in a bluegill-Daphnia-parasite system. Oecologia 153: 453–460.
- 60. Whiles MR, Lips KR, Pringle CM, Kilhara SS, Bixby RJ, et al. (2006) The effects of amphibian population declines on the structure and function of Neotropical stream ecosystems. Frontiers in Ecology and the Environment 4: 27–34.
- 61. Smith KG, Lips KR, Chase JM (2009) Selecting for extinction: nonrandom disease-associated extinction homogenizes amphibian biotas. Ecology Letters 12: 1069–1078.