Seed Feeding, Seed Caching, and Burrowing Behaviors ofHarpalus rufipesDe Geer Larvae (Coleoptera: Carabidae) in the Maine Potato Agroecosystem

Seed feeding, seed caching, and burrowing behaviors ofHarpalus rufipeslarvae were investigated in Presque Isle, Maine during 1994–1996. Larval feeding preference for seeds was tested in the laboratory with 10 plant species in both choice and no-choice feeding trials. The feeding rate of seeds varied with seed species. The biomass and number of seeds attacked were related to seed size. All larval instars attacked more small seeds as compared to large seeds and mid-sized seeds in both choice and no-choice feeding trials. Seedling emergence of redroot pigweed (Amaranthus retroflexusL.) under field conditions was affected by a significant interaction between the presence of first instar larvae and the depth of seed burial. Fewer seedlings emerged from 0 to 3 cm depth in the soil when first instar larvae were present than when they were not present. Seedling emergence from seeds sown below a 3-cm depth was not affected by the presence of larvae. A parallel study investigating the effects of first instar feeding on a less preferred species, wild mustard (Brassica kaber[DC] L. C. Wheeler.), showed that the presence or absence of larvae had no impact on the emergence of wild mustard across all burial depths. Laboratory studies of larval seed caching and burrowing behavior using the preferred common lambsquarters (Chenopodium albumL.) and the less preferred birdsrape mustard (Brassica rapaL.) showed no difference between burrowing depths nor the number of seeds cached by the first instar larvae. However, the number of seeds cached by the second instar larvae was dependent on species identity (1.5 common lambsquarters per day vs 0.5 birdsrape mustard per day).