How does abiotic heterogeneity shape plant-caterpillar-parasitoid communities?
Variation in abiotic resources can have profound effects on living communities. One way in this can occur is by altering anti-herbivore resistance traits of the plants that form the base of the food web, which can in turn change herbivore preference and performance. These traits can also affect the relationship between herbivores and their enemies; by prolonging development or lowering abundance, plant quality can increase or decrease herbivore susceptibility to predation or parasitism.
In this project, I used a natural mosaic of soil fertility in Northern California to explore 1) if soil resource availability alters size, composition, and structure of multitrophic interaction networks, and 2) understand which mechanisms could be driving these differences. This unique soil mosaic is characterized by having large patches of low resource serpentine soils that are low in Nitrogen and micronutrient content, have poor water holding capacity, and prevalent heavy metals. These patches transition within meters to areas of non-serpentine soil, which are higher in organic matter, more balanced in micronutrient stoichiometry, and generally more supportive of plant growth. Growing throughout this mosaic of distinct soil types are conspecific (Adenostoma fasciculatum) and congeneric (Arctostaphylos sp., Ceanothus sp., Quercus sp.) woody shrubs, which in turn host a diverse assemblage of > 80 species of Lepidopteran larvae, and their many associated species of parasitoid wasp and fly.
By collecting caterpillars from plants across soil types, we found that herbivore communities were both smaller and less diverse on serpentine soils. We found evidence for a bottom-up, plant trait based explanation for this pattern, and a lack of evidence for predation or parasitism as a top-down force shaping these differences. Finally, we found that soil type ramifies to alter network-level properties; in serpentine soils, generalist herbivore species interact with fewer plant genera, creating a more specialized or "modular" network than in neighboring non-serpentine soils.
Soils shape herbivore communities from the bottom-up
Insect herbivore communities are often shaped by two non-exclusive mechanisms: plant resources and predation. From the 'bottom-up', host plant traits can act directly on herbivores by changing preference of ovipositing females and/or larval mortality; simultaneously, natural enemies exert predation pressure from the 'top-down'. In addition to their independent effects, these two forces can work together to alter insect assemblages. When host plants are of low quality, herbivores may develop more slowly and be more available to their enemies in time. Alternatively, low quality plants often harbor lower abundances of herbivores, and may be passed over by predators and parasitoids in favor of higher quality plants with more prey.
To explore whether these forces were behind differences in herbivore community size across the serpentine-non serpentine habitat mosaic, I first quantified host plant quality by measuring larval growth of native herbivores on all focal plants. I then reared over 2000 larvae in the lab to quantify parasitism rates, and measured bird predation by placing bird exclusion cages around 96 shrubs in McLaughlin Reserve. Together, these results showed that the top-down effects of predation and parasitism are of equal magnitude across soil types, and that plant quality alone is the more likely driver of herbivore community size and diversity in this soil mosaic system.
Effects on structure of species interactions & ecological networks
An exciting aspect of this project has been using the natural history of native caterpillars to better understand their role within their community. For example, species that can consume only a single plant genus (specialists) create compartments of specialized interactions within the plant-herbivore network, linking in only to their specific host. In contrast, species that consume many genera (generalists) have the potential to act as important connectors across the interaction network, as they can form links between many different host plants. Environmental factors that alter the way generalists interact with potential hosts - by making them have a narrower host range in some areas or habitats than others, for example - can subsequently alter how the entire network is connected. In this project, I have found that generalist species in harsh serpentine soils exhibit a narrower "realized" host range, reducing the prevalence of generalized interactions in the community and making the entire plant-herbivore network more 'modular', or specialized, in these low-resource patches.