Fig 1.
Representation of eight treatment groups in the experimental design.
The smaller tadpole represents a single Pacific chorus frog (Pseudacris regilla) larva, while the larger tadpole represents a single American bullfrog (Lithobates catesbeianus) larva. Dividing lines within the tanks represent the cage treatment (open or closed). Tadpole images are from artist Parkjisun, sourced from The Noun Project website, under a Creative Commons-BY license.
Fig 2.
Representative daily measured maximum temperature in a historic treatment tank and a future treatment tank.
Daily maximum measured temperatures did not always reach the daily set temperature due to power issues with heaters (see [42]), but a consistent difference between the treatment groups was maintained. Temperature scenarios were designed to provide variation in daily maximum temperature, while maintaining an overall positive slope to represent the increase in average daily temperature over the summer months.
Table 1.
Analysis of deviance table for all generalized linear regression models of treatment effects on survival, mass, and development.
Fig 3.
Bullfrog larvae reduced the absolute relative growth of Pacific chorus frog larvae under simulated future conditions.
Relative growth represents the proportion of the initial body length (snout-vent length [SVL]) obtained during the experimental period. Under historic (cooler, less evaporation) conditions, no effect of bullfrogs was observed on growth of chorus frogs. Under future conditions, bullfrog presence reduced growth of chorus frog larvae (t-ratio = 2.254; df = 2,35; p = 0.0304; marked with an asterisk in plot). Boxplots are standard boxplots, with median and standard quartiles ± 1.5*IQR whiskers.
Table 2.
Analysis of covariance for absolute relative growth (proportion of initial length) of Pacific chorus frogs.
Fig 4.
Pacific chorus frog weight at metamorphosis (gosner stage 42).
A) Pacific chorus frog larvae were heavier at stage 42 [49] under simulated future climate conditions (p = 0.04). B) Pacific chorus frog larvae metamorphosed at a heavier weight in the absence of bullfrog larva (p = 0.03). C) Pacific chorus frog larvae were heavier in open cage treatments at stage 42 (p = 0.008). Boxplots are standard boxplots, with median and standard quartiles ± 1.5*IQR whiskers. Data were analyzed using a Gaussian GLM with a log link.
Fig 5.
Pacific chorus frog development under simulated future conditions.
Pacific chorus frog larvae reached metamorphosis (Gosner stage 42 [49]) an average of 5 days sooner under future climate conditions (quasi-Poisson GLM with log link; p<0.0001). Boxplots are standard boxplots, with median and standard quartiles ± 1.5*IQR whiskers.