Performance of parasite larvae
To investigate if the survival of L3 was affected by exposure to skin secretions of cane toads, or of a native frog (Litoria caerulea ), we placed <7 day old L3 in a Petri dish containing the treatment solution for 24 h. To obtain skin secretions of cane toads (n = 10 range-core toads and 12 invasion-front toads) and frogs (n = 4), we swabbed individuals as described above. The head of the cotton bud was then diluted in 1 ml water to obtain our treatment solution. We used multiple individuals (but not mixing secretions from different individuals) to avoid pseudoreplication, and used the treatment solution of individual toads and frogs for L3 originating both from the range-core and invasion-front to avoid biases induced by individual variation of the secretions. Aged tap water was used as control. After keeping L3 in a treatment solution for 24 h, we individually placed them in 96 well microplates (Greiner Bio-One, Kremsmünster, Austria) with 200 µl water. Twice a week we refilled the water level in the microplates and checked whether each L3 was alive by touching it with a human eyelash that was sterilized in 70% ethanol and then washed in sterilized water to minimize the risk of introducing pathogens. If the L3 did not move after 5 attempts, we considered it dead (verified within 3 days as the body began to decompose), and calculated longevity as the number of days from hatching to death. We used a crossed design, testing L3 from different geographic regions with toad skin secretion from different geographic regions as well as frog skin secretion (n = 170 L3; Table S3). L3 for this experiment were sourced from multiple toads (n = 15 range-core toads and 8 invasion-front toads) to avoid pseudoreplication.
To investigate if L3 use cane toad skin secretions as a cue, we obtained skin secretion solutions as described above. The experimental setup consisted of a 15 mm wide and 50 mm long strip of filter paper, in a Petri dish. One end of the strip was used as the ‘control’ zone and the other end as the ‘treatment’ zone. We pipetted 20 µl of skin secretion onto the treatment zone and 20 µl water onto the control zone. For procedural controls (i.e., to check that L3 randomly move into both zones), we applied water to both zones. The location of control and treatment zones was assigned randomly. We then pipetted 10 L3 (also contained in 20 µl water) onto the middle of the strip of filter paper. After 2 h, we cut the filter paper in half, washed out the L3 from the paper (same method as in experiment 1), and counted the number of L3 in the control and the treatment zones. We also washed out the Petri dish to count the remaining L3 to estimate the ‘recovery rate’ (i.e., how many L3 we were able to find). For each scent trial (procedural control, range-core toad skin secretions, invasion-front toad skin secretions) and L3 origin (range-core, invasion-front), we ran 30 replicates, totaling 180 trials using 1,800 L3 (Table S4).