Behavioural thermoregulation in a Neotropical frog is detectable only when physiological integrity allows movement

Technical Summary: Behavioural Thermoregulation in Pseudis limellum

Problem and Context
While behavioural thermoregulation is a primary mechanism for ectotherms to maintain optimal body temperatures, its efficacy at the physiological extremes of a species' thermal tolerance remains poorly understood, particularly in Neotropical amphibians. Existing literature heavily relies on basking lizards, leaving a gap in knowledge regarding semi-aquatic frogs that inhabit the water–air interface (neustonic niche). These species face a unique thermal regime combining wide diel temperature fluctuations with the locomotor demands of ballistic jumping. Furthermore, the Pantanal wetland, home to the study species, is projected to warm by 4–7°C by the end of the century, potentially exceeding the thermal safety margins of local fauna. The study addresses whether Pseudis limellum exhibits active behavioural thermoregulation and whether this behaviour is constrained by the physiological capacity for locomotion at thermal extremes.

Methodology
The researchers conducted a controlled experiment with 60 adult Pseudis limellum in a 170 × 55 cm arena featuring a linear thermal gradient from 20°C to 40°C.

• Experimental Design: Frogs were placed at random starting positions along the gradient and allowed a 90-second observation window to move. A mechanical stimulus (touch with a stick) was used to elicit movement.
• Variables: Three response variables were recorded: response time (s), first-jump distance (cm), and total jumping distance (cm). Final resting temperatures were measured immediately after the trial. Morphological covariates (snout-vent length and relative hind-limb length) were also recorded.
• Statistical Approach: The study employed a fully Bayesian framework using the brms package in R.
  • Distance Models: A hurdle log-normal model was used to account for structural zeros (frogs that did not jump) and the continuous distribution of jumping distances. The hurdle component modeled the probability of immobility as a function of initial temperature, while the continuous component modeled total distance as a function of final temperature.
  • Response Time Model: A right-censored Weibull accelerated-failure-time (AFT) model was applied to response times, treating the 90-second limit as a censoring boundary.
  • Null Model: To distinguish active thermoregulation from geometric drift (the tendency of a random walker in a bounded space to end near the center), the authors simulated a two-dimensional Pearson random-walk null model for each individual based on their starting position and observed jump characteristics.
  • Inference: The study utilized weakly informative priors, posterior predictive checks, and power-scaling sensitivity diagnostics to ensure robustness. The "parking temperature" ($T_{park}$) was derived as the final temperature maximizing expected jumping distance.

Key Results

• Locomotor Constraints at Extremes: Non-response was highly concentrated at thermal extremes. 43% of frogs released at cold temperatures (≤20°C) and 25% of those at hot temperatures (32–36°C) failed to jump within the 90-second window. This supports a physiological-limitation hypothesis where locomotor capacity is impaired at both cold and hot extremes.
• Jumping Performance: Total jumping distance followed a concave (inverted-U) relationship with final temperature, peaking at a parking temperature ($T_{park}$) of 27.2°C (95% CrI [19.8, 30.0]).
• Active Thermoregulation: The combined statistical test rejected the geometric random-walk null model ($P < 0.0001$). The signal for active thermoregulation was concentrated in the cold and hot starting bands, where observed final temperatures were significantly closer to $T_{park}$ than predicted by random drift. In intermediate temperature bands, observed distributions matched the null model, indicating that frogs starting near the optimum did not need to actively correct their position.
• Morphology: Relative hind-limb length did not improve predictive performance in any model, suggesting that temperature, rather than morphology, was the primary driver of variation in this dataset.

Significance and Claims
The paper claims that behavioural thermoregulation in P. limellum is conditional on physiological integrity. The species actively selects a thermal niche centered around ~27°C, but this behaviour is only expressed when thermal conditions permit locomotion. At the extremes of the thermal gradient, individuals are simultaneously most motivated to leave (due to thermal stress) and least capable of doing so (due to muscular failure or torpor).

The authors position this finding as a critical update to the cost–benefit models of thermoregulation, highlighting that the "actuator" of thermoregulation (locomotion) is itself a victim of the thermal stress it seeks to escape. Given the projected warming of the Pantanal, the narrow thermal niche of P. limellum and its reliance on locomotor performance at ~27°C suggest high vulnerability. As temperatures rise, the window for effective behavioural thermoregulation may shrink, potentially leading to a decoupling between the need to thermoregulate and the physiological capacity to do so. The study fills a specific taxonomic and ecological gap by providing the first behavioural thermoregulation data for a neustonic hylid, distinguishing its thermal performance curve from fully aquatic models like Xenopus.