Supplementary Methods
Collection and Animal Husbandry
Bullfrog eggs (10 clutches) were collected from Fish hatchery pond (41.64° N 80.43° W) and Geneva pond (41.59° N 80.25° W) near Pymatuning Laboratory of Ecology in Pittsburgh, Pennsylvania in August 2011. After hatching, size selected bullfrog tadpoles (Gosner stage 27) were added to mesocosms. Bullfrog tadpoles overwintered in 1200-L plastic cattle watering tanks filled with ~900 L of water containing leaf litter (primarily Quercus), a suite of micro-invertebrates (zooplankton, amphipods, isopods, snails), phytoplankton and periphyton. On 30 May 2012, we drained the cattle tanks and randomly distributed bullfrog tadpoles into 100-L pools that were set up on 16 May 2012. These pools were filled with 90 L of well water and we added pond water from four nearby ponds and visually screened for invertebrate predators. After removing predators, we combined the pond water and added equal aliquots to each tank to provide a natural source of algae and zooplankton to all pools. We added 5 g of rabbit chow and 100 g of dry oak leaves (primarily Quercus spp.) to provide a source of nutrients and a substrate for the periphyton. To ensure that the bullfrog tadpoles had enough periphyton resources, we set up new pools every two weeks using the same procedures described above. Once the periphyton algae had established in these new pools, we randomly transferred bullfrog tadpoles to these new resource rich pools. All tanks and pools were covered using 60% shade cloth to prevent organisms from entering or leaving. We collected bullfrog metamorphs from 8 June until 4 September 2012.Animals were randomized into holding containers before overnight shipment to Oregon State University (OSU). All animals were between 2 and 4 weeks old before the start of experiments at OSU.
Upon arrival to OSU, bullfrogs were placed in glass terraria held at 20°C with a 13h: 11h light to dark photoperiod. Animals acclimated for at least 24h before initiation of the experiment. We randomly assigned individuals to either Bd-exposed or unexposed (control treatments) for each separate strain experiment. All animals were weighed to the nearest 0.001g and measured to the nearest 0.1 mm before the start of the experiment. Animals were placed in Petri dishes (140 X 30 mm) with holes in the lid and a thin film of water covering the bottom for the duration of the experiment. Animals were changed (fresh water) and re-inoculated once a week for 4 weeks and were fed 2-week crickets (5 crickets per feeding) twice per week. We monitored animals for signs of mortality, daily. We ran the first group of 33 bullfrogs in the first experimental trial with Bd strain JEL 274 and ran the next group of 31 bullfrogs at the same developmental time-point (several weeks after metamorphosis) for the second experimental trail with Bd strain JEL 630. Thus, we ensured that animals in both trials were the same age after metamorphosis.
Inoculation
We used Bd strain JEL 274 (originally isolated from a Western toad in Colorado, USA) and JEL 630 (originally isolated from an American bullfrog in Oregon, USA). The collection history of each strain differs (JEL 274 was originally isolated in 1999 and JEL 630 was isolated in 2009, Joyce Longcore, personal communication). We obtained both strains from J. Longcore in May 2011. Strains that had been cryogenically preserved at -80°C were thawed, streaked onto tryptone agar plates, and express mailed to Oregon (JEL 630 to Portland State University and JEL 274 to Oregon State University). After approximately 1 week of growth, we sterilely removed Bd colonies from original plates into sterile 1% tryptone broth. Both JEL 274 and JEL 630 went through approximately 3-5 passages (into new liquid culture) before they were used in our bullfrog strain experiments. Bd strains were maintained for a period of early growth after transfer to new liquid culture at room temperature (approximately 2 weeks), and then stored at 4°C until the next serial passage (approximately once per month). We grew Bd from liquid cultures on 1% tryptone-agar plates for use in our experiments and for both strains our methods were identical. Plates were inoculated 7 days prior to inoculation of animals and held at approximately 20 °C. To harvest Bd from agar, we flooded plates with 15 mL dechlorinated water for 5 minutes. To standardize inoculation dose among exposed animals, we pooled water from at least 10 Bd-inoculated plates for each inoculation and quantified the number of zoospores in the inoculums with a hemocytometer. After quantifying zoospores, we diluted inoculum to a concentration of 1.7 × 104 zoospores/mL inoculum. Then, we added 15 mL inoculums to each Petri dish. Thus, Bd-exposed animals were exposed to 15 ml of 1.7 × 104 zoospores/ml inoculate broth. Control animals were exposed to 15 mL inoculum without Bd culture created using sterile agar plates. We exposed all species to the same Bd dose at each inoculation, and we conducted each inoculation during the experiment in the same manner. Water in Petri dishes was changed every 7 days, with re-inoculation occurring simultaneously. We placed an additional 10 mL of water into each Petri dish at each water change, which combined with the inoculate was sufficient to completely cover the bottom of the dish with a thin film of water. Animals therefore remained in direct contact with water on their ventral side, but could not submerge themselves. We monitored mortality twice daily and removed dead animals from their dishes immediately and preserved them in 95% ethanol.
We did not observe any qualitative differences in zoospore activity or density on 7d old Bd plates prepared from JEL 274 or JEL 630, although we did not quantify zoospore size or other phenotypic differences in strains. For both strains, we observed primarily active zoospores (versus full or empty sporangia) on 7 d old plates. Because of our standardization of zoospore concentration, we controlled for the amount of zoospores that animals were initially exposed to. However, innate differences of each strain (e.g. differences in temporal patterns of zoospore growth and reproduction in each Bd strain while in contact with animals and in the water within the Petri dishes) are likely to have acted after initial inoculation. Because all other aspects were controlled, differences in host responses to Bd were a function of both host and strain-specific characteristics.
At the end of the 30 day trial, all surviving animals were euthanized by immersion in MS-222 and preserved in 95% ethanol. We then determined Bd-infection levels with quantitative polymerase chain reaction (qPCR) using methods of Boyle et al. 2004 for all Bd-exposed animals and 5 randomly selected control individuals from each experimental trial. We analyzed each sample in triplicate and calculated the average number of genome equivalents per individual.