Effectiveness of cardenolides as feeding deterrents toPeromyscus mice.

I compared the feeding responses of five species ofPeromyscus mice (aztecus, polionotus, melanotis, leucopus, andmaniculatus) to three bitter-tasting cardenolides (ouabain, digoxin, and digitoxin) that differ greatly in lipophilic character.Peromyscus, like other muroid rodents, are unusual in that they can ingest relatively large amounts of cardenolides without adverse physiologic effects. In experiment 1, I determined avoidance thresholds for the three cardenolides with 48 hr, two-choice tests. Mice exhibited large interspecific differences in avoidance threshold, and the interspecific ranking of the thresholds (maniculatus=leucopus >melanotis >polionotus >aztecus) was the same for each of the cardenolides. In experiment 2, I reevaluated the avoidance thresholds, but this time monitored the pattern of intake (i.e., bout lengths) during initial feeding encounters with cardenolidelaced diets. For each cardenolide, mice were subjected to three tests. In test 1, they received a control diet; in test 2, a diet containing the cardenolide at a concentration 1 log, unit below the avoidance threshold (as determined in experiment 1); and in test 3, a diet containing the cardenolide at the avoidance threshold concentration. Results were similar across all species and cardenolide types: Bout lengths in tests 1 and 2 were statistically equal, whereas those in test 3 were significantly shorter than those in test 1. The rapid rejection of cardenolide-laced diets in test 3 is consistent with a preingestive (i.e., gustatory) mechanism underlying the avoidance thresholds. I conclude (1) thatPeromyscus species differ substantially in taste sensitivity to cardenolides and that these differences may influence each species' respective ability to eat cardenolide-laced insects; and (2) that a species' relative taste sensitivity to one cardenolide predicts its sensitivity to other cardenolides.