EXPERIMENTAL TESTS OF EFFECTS OF PLANT PRODUCTIVITY AND DIVERSITY ON GRASSLAND ARTHROPOD DIVERSITY

Abstract

Because the quantity, quality, and heterogeneity of resources should affect the diversity of consumers, plant productivity, plant composition, and plant diversity may influence the diversity of trophic levels higher up the food chain (''bottom-up'' control of diversity). Increasing plant productivity may increase herbivore diversity by: increasing the abundance of rare resources (''resource rarity hypothesis''), increasing herbivore abun- dance and local persistence (''consumer rarity hypothesis'') or increasing intraspecific den- sity dependence (''density dependence hypothesis''). Increasing plant diversity may increase the diversity of herbivores because herbivores specialized on these new plant species can persist locally. Increasing the diversity of herbivores could likewise increase the diversity of predators and parasites, although parasite and predator diversity may also respond directly to changes in vegetation. Here I use data from a well-replicated grassland experiment at Cedar Creek, Minnesota, to analyze the responses of arthropod diversity to independent manipulations of (1) plant productivity and (2) plant diversity and composition. Long-term nitrogen addition (''historical treatment'') decreased plant diversity and favored C 3 grasses. Short-term nitrogen addition (''modern treatment'') increased plant productivity without appreciably changing plant diversity or plant composition. Arthropods were sampled using both sweep nets and vacuum samplers. Total arthropod species richness and abundance, as well as species richness and abun- dance of every trophic group (detritivores, herbivores, parasites, and predators), were sig- nificantly greater in plots with higher levels of modern fertilization and greater plant pro- ductivity. Path analysis supported the consumer rarity hypothesis, because modern fertil- ization increased herbivore species richness only indirectly by increasing herbivore abun- dance. Surprisingly, higher levels of historical fertilization that lowered plant species richness (but did not change plant productivity) significantly increased total arthropod species richness, did not affect detritivore or herbivore species richness, and significantly increased parasite and predator species richness. Direct and indirect effects of historical fertilization on the species richness of trophic groups were separated statistically using block regression chain modeling. Herbivore species richness was decreased through the direct pathway, but this was offset through indirect effects so that there was no overall response to historical fertilization. In plots with low plant diversity, similar numbers of herbivore species supported a greater diversity of parasites and predators. These results suggest that the diversity and composition of plants control the diversity of consumers not only directly by determining available resources, but also indirectly by influencing the interactions between herbivores and their parasites and predators.