Predator avoidance snakes

In their predator avoidance, salamanders use complex odors that combine chemicals from both predator and prey. In the laboratory, red-backed salamanders, P. cinereus, avoid filter papers soaked with water extracts from garter snakes that had been preying on salamanders, while earthworm-fed snakes lacked this effect. Exudations from unfed snakes and extracts from homogenized salamanders had no such alarming effect (Madison etal, 2002). 

Burger, J. (1989). Following of conspecific and avoidance of predator chemical cues by pine snakes (Pittwphis melanoleucus). Journal of Chemical Ecology 15,799-806. 

Flowers, M. A. and Graves, B. M. (1997). Juvenile toads avoid chemical cues from snake predators. Animal Behaviour 53,641-646. 

Red-backed salamanders Plethodon cinereus) respond to distilled water rinses of garter snakes (Thamnophis sirtalis) that have recently attacked other red-backed salamanders (Madison et al., 2002). These anti-predator behaviors consist of avoidance (Madison et al., 1999a Mcdarby et al., 1999) and altered activity (Madison et al., 1999a), and have been documented in the laboratory and in the field (Sullivan et al., 2002). The responses persist for up to 36 hours in laboratory trials, but no longer than 3 hours in the field (Sullivan et al., 2002). Whether the attenuated response in the field is due to a diminishing of the cue or to field-based shifts in salamander behavior remains unclear. 

Amphibians are vulnerable to predation and show diverse predator defenses. We hypothesized that the common red-backed salamander should try to evade two common terrestrial predators, the garter snake and the spotted salamander, by avoiding locations containing their scent. When given a choice of clean substrates or those soiled by these predators, the red-backed salamander avoided the soiled substrates. To test whether they avoided the waste products and not the predator scent, we gave test subjects a choice between substrates soiled by predators and those soiled by red-backed salamanders. Salamanders pref-ered conspecific substrates to predator substrates, and in combination with other findings, our data show that red-backed salamanders probably reduce the likelihood of predation by avoiding locations with the chemical traces of predators. 

The alarm substance hypothesis cannot be excluded in our trials, but it can be rejected based on other studies. We have shown that red-backed salamanders do not avoid substrate traces from red-backed salamanders that have been induced to autotomize their own tail (a stress-induced, anti-predator response Arnold, 1988 Lancaster Wise, 1996), and other red-backed salamanders even attempt to eat the autotomized tails (Madison, unpublished). In addition, the broken skin of sacrificed red-backed salamanders fails to negate the chemosensory avoidance of substrates from garter snakes that had been fed red-backed salamanders (McDarby, 1997). 

The intraspecific competitor avoidance response can be excluded in the trials involving both predators. The spotted salamanders did not have diets that included red-backed salamanders, so pheromones of red-backed salamanders could not have elicited the avoidance observed. The garter snake trials included one comparison that specifically control-... 

In summary, we believe our experiments indicate that red-backed salamanders identify and avoid chemical substances deposited by spotted salamanders and garter snakes, and since the latter two species are either known or probable predators of red-backed salamanders, we believe the avoidance response is an anti-predator mechanism that decreases predation risk. 

The lower activity levels at night during the goldfish-fed snake trials were probably normal, low-level activity rather than freeze responses. The latter term implies immobility and certainly a reduction in activity below normal levels, which in the absence of a significant avoidance response was unlikely to have occurred. Some of the salamanders in our study may have momentarily shifted into an anti-predator defense mode upon initial exposure to snake odors, but quickly shifted back into an avoidance mode once contact with the predator appeared less imminent. This ambivalence in some salamanders may also have contributed to the 18 minute delay to peak activity. 

One explanation for the partial dissociation observed and why avoidance scores were not more extreme during the day is that confinement in the test dishes could have forced the more active salamanders to cross between samples at such a rate that side avoidance was abolished. If a single rushed response from each salamander was the basis for the preference score, a dilution or cancellation effect could have occurred for some salamanders. However, since salamanders in our study had many opportunities to correct for crossings onto snake substrates, and since only 11 out of 20 positions on the water side were needed to count as an avoidance of snake substrate, any tendency to avoid substances on this substrate should have been expressed. In addition, the movements that occurred were not rushed, as in an escape ( anti-predator ) response, but slower deliberate actions with pauses and head movements typical of chemosensory investigation (reviewed by Jaeger, 1986). Thus, we conclude that space restrictions in the petri dishes did not abolish avoidance among the more active salamanders. 

---