Tritrophic interactions in forests: Direct and indirect interactions between birds, insect herbivores, and oaks

I used bird exclusion experiments to determine how the direct and indirect effects of bird predation vary spatially within forests. In Chapter 1 I used a randomized-block design to show that both bird and insect herbivore abundances varied through space in similar habitat. Yet the direct effects of birds on herbivores and the indirect effects on leaf damage of oaks did not vary between blocks, suggesting that the biological control services of birds are robust to variations in population abundance of both the birds and their prey.;In Chapter 2 I used a light manipulation to produce oak saplings with different leaf quality and factorially excluded bird predators in a common light environment to measure the effects of birds and leaf quality on herbivore abundance and herbivory. Sun leaves appeared to be lower quality food, yet herbivores were significantly more abundant and caused greater leaf damage on sun-exposed trees. Bird exclusion did not change herbivore abundance, but did increase leaf damage. The effects of birds did not vary with light manipulation. The higher abundance and damage on sun trees may have been due to ovipositing females preferring hosts with greater leaf and shoot growth. Birds may have reduced leaf damage through non-consumptive effects on herbivore feeding behavior.;In Chapter 3 I combined a two-year bird exclusion experiment with measurement of natural variation in oak leaf quality. Although herbivore community composition varied over time, birds had little effect. Leaf quality influenced the total abundance and richness of herbivores as well as the abundance of different feeding guilds. These effects of leaf quality were strongest at the end of the growing season, when leaf quality is presumably lowest overall.;In Chapter 4 I studied the role of the physical light environment and foliage characteristics in determining abundance of the oak lacebug ( Corythuca arcuata Hemiptera: Tingidae). Using an information-theoretic approach, I evaluated a priori hypotheses of the relationship between light, plant traits, and C. arcuata abundance. Abundance was best predicted by light environment and leaf carbon content. Adult C. arcuata prefer trees growing under an open canopy and trees with low carbon content; abundance also positively correlated with leaf water content. Although carbon and water did not vary with light in this study, low carbon and high water content are often associated with shadier conditions, suggesting that C. arcuata faces a trade-off between preferences for physical habitat conditions and host plant characteristics.;Insect prey abundance can also affect the distribution of avian predators. In Chapter 5 I compared the annual distribution of native cuckoos to outbreaks of invasive gypsy moths (Lymantria dispar Lepidoptera: Lymantriidae). Populations of cuckoos, one of the few bird species that feeds on gypsy moth caterpillars, spike within outbreaks and are significantly below average abundance for tens to hundreds of kilometers in all directions. This pattern and timing of abundance support the hypothesis that cuckoos locate concentrated food resources during a post-migratory nomadic phase and represents one of the few cases of native predator distribution being influenced by exotic prey.;These studies indicate that complex interactions exist beyond a simple, unidirectional consumption model of plants, herbivores, and avian predators. The indirect positive effect of birds on plants appears robust to variation in the abundance and traits of the three trophic levels, but the mechanism for this effect may vary through time and space. The impact of birds, however, did not vary with plant characteristics. These characteristics, which can depend on environmental context, likely play a larger role in determining the abundance, structure, and impacts of herbivores than do insectivorous bird predators. (Abstract shortened by UMI.)...