Coevolution Between Grasshopper mice (Onychomys spp.) and Bark and Striped Scorpions (Centruroides spp.)

Abstract

Asymmetrical selection has been proposed as the strongest argument for rejecting hypotheses of coevolutionary arms races between predators and prey. In many cases there is evidence of increased investment by the prey in response to the predator, but no evidence of increased investment by the predator in response to the prey, thus producing asymmetry in selection. However, selection against a predator may be increased when the interaction is with a "dangerous" prey. Predators are most likely to respond evolutionarily to potentially lethal prey. This study employs grasshopper mice (Onychomys spp.) and bark and striped scorpions (Centruroides spp.) as a model to test the hypothesis that interspecific interaction between a predator and a potentially lethal prey will result in behavioral and physiological adaptations that reciprocally mediate their interaction (i.e., coevolution). Bark scorpions (Centruroides exilicauda) and striped scorpions (Centruroides vittatus) produce a potent venom containing neurotoxins that selectively bind to the ion-channels of vertebrates. Vertebrate-specific neurotoxins may produce lethal effects in mammals, especially small mammals. Southern grasshopper mice (Onychomys torridus) and Mearns' grasshopper mice (O. arenicola) are known to be voracious predators on scorpions. Southern grasshopper mice are broadly sympatric with bark scorpions in the Sonoran Desert, and Mearns' grasshopper mice are broadly sympatric with striped scorpions in the Chihuahuan Desert. The third species in this genus, the northern grasshopper mouse (O. leucogaster) is broadly allopatric with Centruroides spp. In a preliminary study, both southern and Mearns' grasshopper mice demonstrated resistance to bark and striped scorpion neurotoxins. The evolution of toxic venom and resistance to that venom strongly suggests a coevolutionary relationship between Onychomys spp. and Centruroides spp. To test this hypothesis, I evaluated the predator-prey relationship between grasshopper mice and their toxic scorpion prey during staged feeding trials. Additionally, I compared the geographic patterns of venom resistance in all three species of grasshopper mice with geographic patterns of venom toxicity in bark and striped scorpions. Results from the feeding study demonstrated that grasshopper mice do not distinguish between toxic and non-toxic species of scorpions; mice attacked, incapacitated, and consumed bark and striped scorpions without hesitation and as effectively as they attacked crickets and non-toxic scorpions in the genus Vaejovis. The feeding experiments indicate that grasshopper mice have the ability to prey on bark and striped scorpions in habitats where they co-occur. Venom resistance analyses demonstrated that all three species of grasshopper mice have evolved some resistance to the vertebrate-specific neurotoxins produced by Centruroides spp. The assays show that patterns of venom toxicity in Centruroides and venom resistance in Onychomys co-vary geographically, both within and among species; i.e., populations of Onychomys interacting with the most toxic populations of Centruroides were extremely resistant; populations of Onychomys interacting with only moderately toxic populations of Centruroides were only moderately resistant; and populations of Onychomys not sympatric with Centruroides were only weakly resistant. Such systematic covariation between venom toxicity in the scorpions and venom resistance in the mice is consistent with a coevolutionary, arms race hypothesis.