Habitat loss and simplification lower arthropod richness but not diversity

Habitat loss is rarely truly random and often occurs selectively with respect to the plant species comprising the habitat. Such selective habitat removal that decreases plant species diversity, that is, habitat simplification or homogenization, may have two negative effects on other species. First, the reduction in plant community size (number of individuals) represents habitat loss for species at higher trophic levels who use plants as habitat. Second, when plants are removed selectively, the resulting habitat simplification decreases the diversity of resources available to species at higher trophic levels. It follows that habitat loss combined with simplification will reduce biodiversity more than habitat loss without simplification. To test this, we experimentally implemented two types of habitat loss at the plant community level and compared biodiversity of resident arthropods between habitat loss types. In the first type of habitat loss, we reduced habitats by 50% nonselectively, maintaining original relative abundance and diversity of plant species and therefore habitat and resource diversity for arthropods. In the second type of habitat loss, we reduced habitats by 50% selectively, removing all but one common plant species, dramatically simplifying habitat and resources for arthropods. We replicated this experiment across three common plant species: Asclepias tuberosa, Solidago altissima, and Baptisia alba. While habitat loss with simplification reduced arthropod species richness compared with habitat loss without simplification, neither type of habitat loss affected diversity, measured as effective number of species (ENS), or species evenness as compared with controls. Instead, differences in ENS and evenness were explained by the identity of the common plant species. Our results indicate that the quality of remaining habitat, in our case plant species identity, may be more important for multi-trophic diversity than habitat diversity per se.