1
Electronic Supplementary Material
Individual and combined effects of multiple pathogens on Pacific treefrogs
John M. Romansic1, Pieter T. J. Johnson2, Catherine L. Searle1, James E. Johnson3,
Tate Tunstall4, Barbara A. Han5, Jason R. Rohr6 and Andrew R. Blaustein1
1Department of Zoology, Oregon State University, 3029 Cordley Hall, Corvallis, OR 97331, USA
2Department of Ecology and Evolutionary Biology, University of Colorado, Boulder, CO 80309, USA
3Department of Biological Sciences, Central Washington University, Ellensburg, WA 98926, USA
4Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, Santa Barbara, CA 93106, USA
5Odum School of Ecology, University of Georgia, Athens, GA 30602, USA
6Department of Integrative Biology, University of South Florida, Tampa, FL 33620, USA
Corresponding author:
John M. Romansic
Email:
Collection and maintenance of Pseudacris regilla
For experiment 1, approximately 220 P. regilla larvae (developmental stage 25, Gosner 1960) were collected on 25July 2005 from a pond in the Potholes subalpine meadow (44.034˚N, 121.668˚W). For experiment 2, 30 masses of P. regilla embryos (stages12–16, Gosner 1960) were collected on 26 June 2006 at Little Three Creek Lake (44.102˚N, 121.642˚W). Larvae were fed a mixture (3:1 by volume) of rabbit chow and Tetramin fish flakes and the photoperiod was natural. Prior to experimentation,P. regilla were kept in 38-l glass aquaria filled with approximately 35 l of aerated water (experiment 1: 2.3–3.4 larvae/l of water; experiment 2: 15 embryo clutches/aquarium, 12 larvae/l of water after hatching) at 13.5–17.0˚ C and transferred to new tanks with new water every 7–9 days. Water, unless otherwise noted, was tapwater conditioned with NovAqua® and Amquel® (0.14 ml of each conditioner/l of water). Larvae were moved to an 18.5–22.5˚ C laboratory one and seven d before the start of experiments 1 and 2, respectively. The acclimation period was shorter for experiment 1 to ensure that larvae were within developmental stages susceptible to limb deformities from attacking Ribeiroiasp. cercariae (stages 26–28, P.T.J.J.unpublished data). After transfer to the warmer temperature, larvae were transferred to new containers with new water every 5–7 days except where noted in Materials and methods. Metamorphs in experiment 1 were not fed until tail resorption was complete, after which they each received 15 1–2 week old crickets per day, with Reptocal mineral and vitamin supplement dusted onto the crickets once per week.
Pathogen sources
Snails (Planorbella sp.) were collected on 20August 2005 (experiment 1) and 26July 2006 (experiment 2) from a pond in the E. E. Wilson Wildlife Refuge (44.700˚ N, 123.213˚ W) with a history of high prevalence of Ribeiroiasp. (hereafter, Ribeiroia) infection and severe limb deformities in P. regilla (P. Johnson et al. 2002; J.M.R. andP.T.J.J., unpublished data). Cercariae from three (experiment 1) or five (experiment 2) snails shedding Ribeiroia cercariae were gathered and mixed following Johnson et al. (1999). BD isolates JEL 215 (experiment 1) and JEL 274 (experiment 2) were grown on Petri dishes containing 1% tryptone agar media and incubated for 9 and 13 days, respectively, at 22–23˚ C. For experiment 1, BD dishes were maintained at 4–5˚ C for 12 days before experimentation to prevent overgrowth. Water moldwas isolated from water taken on 2 July 2006 at Scott Lake (44.217˚ N, 121.891˚ W) following Romansic et al. (2006) and identified as Achlyaflagellata using available keys and standard methods (Johnson 1956; T. Johnson et al. 2002). Identification was confirmed using a DNA barcoding procedure and comparison to available DNA sequences in Genbank (J.E.J,unpublished data). Achlya-inoculated hemp seeds were added to sterile, standard-sized (diameter9 cm) Petri dishes (35 seeds/dish) filled with 46 ml of sterile ultrapure water and incubated at 22–22.5˚ C for 3 days.
Pathogen counting methods
Using a hemacytometer, we measured concentrations of BD and Achlya in randomly selected, unused Petri dishes from the same pathogen stocks used in experimentation.Extrapolation using the total volume of water in the beakers containing larvae provided an estimate of initial concentrations of BD and Achlya in BD+ and Achlya+ treatments, respectively.
Swabbing methods
Swabs were passed over the entire body surface. Larval mouthparts and the underside of hindlimbs were swabbed after the rest of the body. Each swab were passed 10 times on the underside of each femoral hindlimb segment (hip to knee), 10 times on theunderside of each tibiofibular hindlimb segment (knee to ankle), and 15 times on the underside of each hind foot.
Additional controls in PCR assays for BD
Similar to the method of testing for PCR inhibition described in Hyatt et al. (2007), we ran all 67 tissue samples from experiment 1 and 73 out of 121 (60%) of tissue samples from experiment 2 in assays that used an exogenous internal positive control (Exo IPC, Applied Biosystems) in every reaction well. A reaction well containing a sample was scored as inhibited if its Exo IPC Ct score was > 6 cycles higher than the mean Exo IPC Ct of the three procedural controls (containing only master mix and water) on its plate.
In addition, randomly chosen samples of seven BD-exposed P. regilla (5 larvae and 2 metamorphs) and 10 unexposed conspecifics (5 larvae and 5 metamorphs) from a previous experiment (Blaustein et al. 2005) provided positive and negative controls, respectively, to demonstrate that our methods were capable of detecting BD infection. Four of the seven positive control individuals and half of the negative control individuals were assayed with Exo IPC to test for inhibition as described above. BD infection prevalence in randomly selected samples of 15 exposed and 15 unexposed P. regilla larvae from Blaustein et al. (2005), as measured by histology, was 87 and 0%, respectively (Blaustein et al. 2005). All positive control individuals showed infection, with totalsample BDloadsranging from 740 to 4484 genome equivalents (mean± 1SE = 1676 ± 583; range of BD loads in reaction wells: 9.25–56.05 genome equivalents), while no negative control individuals showed infection. Furthermore, none of thesamplesrun with Exo IPC displayed PCR inhibition.
Cox Proportional Hazards models of developmental rate
Cox Proportional Hazards (Cox PH) models 1 and 2 included all individuals in experiments 1 and 2, respectively, and tested for effects of the experimental treatments. Then, for experiment 2, we dropped the non-significant treatment factors and used a multiple comparisons procedure to test for pairwise differences among the remaining treatment combinations (models 3–8), using a Bonferroni adjustment tomaintain α = 0.05 while making multiple pairwise comparisons (Quinn and Keough 2002).Each Cox PH model in the multiple comparisons procedure (models 3–8) used only the experimental units in the two treatment combinations being compared.Model 9 excluded individuals in the Ribeiroia+ treatment that displayed mortality or deformities to test whether Ribeiroia altered rate of development in lesser-affected individuals. Model 10 used only individuals in the Ribeiroia+ treatment and tested for effects of limb deformity status, Achlya, and BD. In models 9–10, individuals not reaching stage 37 were excluded to ensure that development was sufficiently advanced to detect most deformities.
Fig. S1 Percentage of Pseudacris regillalarvae reaching metamorphosis over time in experiment 2. Percentages are out of the total number of individuals at the start of the experiment. All individuals that did not metamorphose died. Dashed and solid lines represent, control (–) and exposure (+) treatments of Ribeiroia, respectively, while open and closed circles represent control (–) and exposure (+) treatments of Batrachochytrium dendrobatidis (BD). Line labels use the abbreviation R for Ribeiroia. Achlya– and Achlya+ treatments are combined as in Fig 1. Overlapping data points are offset, by a maximum of 0.6%. Rates of development were analyzed using a Cox PH model that incorporated whether or not metamorphosis occured and, if so, the time to metamorphosis
Fig. S2 Length at tail resorption, final length, and post-metamorphic growth of Pseudacrisregilla
in experiment 1(n= 23, 22, and 21 individuals, respectively). Mean ±1SE is given for each variable
Fig. S3 Mass at metamorphosis (mean ±1SE) of Pseudacris regilla
in experiment 2.Achlya– and Achlya+ treatments are combined as in Fig. 1