Environmental and Physiological Factors Influence Antipredator Behavior in Scinax hiemalis (Anura: Hylidae)

Individual variation in antipredator behavior may be influenced by ecological factors, such as ambient temperature and predation pressure, and by intrinsic factors, such as physiological condition. We tested the hypothesis that both types of factors interact to induce a specific antipredator response in the treefrog, Scinax hiemalis. When disturbed by the threat of predation, individuals of this species exhibit either a passive response by feigning death or immobility, or an active escape response by jumping away. We examined the responses of 24 adult male S. hiemalis to simulated predation in the laboratory at 10, 15, and 20 C. In addition, responses from a single stimulus were compared with those from a series of stimuli. We determined physiological condition from measures of body length and mass, jumping performance, aerobic metabolism, and estimated energy reserves. Temperature had the most influence on antipredator behavior, with more frogs exhibiting passive responses at 10 C than at higher temperatures. If stimulated more than once, the proportion of active responses increased at all three temperatures. Larger individuals were more likely to exhibit an active response, but no mass-independent physiological variables were related to response type. These results suggest that frogs respond to both extrinsic and intrinsic factors that may affect their behavioral performance.

Fig. 1.

Percent of active antipredator responses in frogs at three different temperatures. White bars represent first stimulus (ejection), and black bars represent second stimulus (tapping)

Fig. 2.

Number of active responses of 10 trials in 16 individuals tested in random order under controlled conditions at 15 C. Behavioral responses to tube ejection and tapping were remarkably different among individuals

Fig. 3.

Relationship between the percent of active antipredator responses produced by frogs and four indicators on physical status. White bars represent first stimulus (ejection) and black bars represent second stimulus (tapping). The letters A, B, and C represent three ordinal categories for the physical trait analyzed, such that A<B<C. Values in each panel represent the P-value of chi-square tests comparing the proportion of active responses and each indicator

Fig. 4.

Maximum jumping performance for the species (black circles) and longest jump during voluntary fleeing in antipredator behavior experiments (open circles) for frogs at three temperatures. Vertical bars show ± SE

Fig. 5.

Distance jumped by frogs at 15 C under three conditions: Max = mean maximum distance jumped in laboratory tests designed to elicit maximum possible responses; APB = mean maximum distance jumped under our antipredator behavior protocol; Field = mean maximum distance jumped by frogs exposed to approach or touch by a predator snake. Vertical bars show ± SE. Asterisk indicates P < 0.001, ANCOVA (F_2,66 = 13.819)

Fig. 6.

Proportion of behavioral responses of frogs in the field to the approach of and touch by a predator snake. Numbers above each bar represent actual number of responses. Death-feigning was not observed in the field. The response behavior referred to as alert (frog raises anterior portion of body, repositions rear legs, and appears ready to jump) was observed only in the field and was probably obscured by our lab protocol

Received: 31 Dec 2001

Accepted: 09 May 2002