Trade-offs in host choice of an herbivorous insect based on parasitism and larval performance

Abstract

Herbivore diet breadth is predicted to evolve in response to both bottom-up and top-down selective pressures, including host plant abundance, quality and natural enemy pressure. As the relative importance and strength of interactions change over an herbivore's geographic range, local patterns of host plant use should change in response, altering local diet breadths. Fall webworm (Hyphantria cunea) is a widespread, polyphagous moth species that feeds on hundreds of plant species worldwide. Populations of fall webworm in Colorado remain polyphagous, but their diet breadth is restricted compared to other populations and thus present an ideal opportunity to test the ecological drivers of host use by a polyphagous herbivore. We investigated how host abundance, larval performance, and parasitism affect host use for fall webworm to test how these selective pressures may act individually or in concert, as well as the role of any trade-offs among fitness components, to explain diet breadth and host use. We found that host abundance was a significant predictor of host use, which suggests a selective pressure to reduce search time for oviposition sites by adult females. We also detected an important trade-off between bottom-up and top-down selective pressures: higher quality host plants also had a greater proportion of larval mortality due to parasitism. Local patterns of host plant abundance appear to narrow the set of hosts used by fall webworms in Colorado, while the trade-off between host quality and risk of parasitism helps explain the maintenance of a generalized feeding strategy within this restricted set of hosts.

Geography is more important than host plant use for the population genetic structure of a generalist insect herbivore

“…Larvae of the two types of FW are relatively easy to distinguish from one another, but it remains unclear whether adults are morphologically differentiated. Previous authors who have examined this issue focused on wing length (e.g., Yang et al, ); however, wing length can vary with environmental and dietary factors; for example, individuals are usually smaller when feeding on low‐quality plants (e.g., Murphy & Loewy, ; Vidal & Murphy, ). A more reliable way to measure such differences is to use morphological “landmarks” and to then calculate the ratio of distances between these landmarks as a measure of overall body shape (Zelditch, Swiderski, Sheets, & Fink, ).…”

“…Larvae from Colorado were collected across the state in the following counties: Arapahoe, Baca, Boulder, Chaffee, El Paso, Garfield, Jefferson, Larimer, Las Animas, Mesa (letters a–j in Figure ). Sample and host plant data were collected as explained in Murphy and Loewy () and Vidal and Murphy (), and we recorded host plant information for each sample (Tables and A1).…”

“…First, FW is one of the most generalized plant‐feeding insect species known, feeding on more than 600 woody plant species worldwide (Warren & Tadić, ). Yet, its diet breadth varies considerably over its geographic range, with some populations clearly being more specialized than others (e.g., western Colorado, Murphy & Loewy, vs. northeastern coast of the United States, Mason, Wilkes, Lill, & Singer, ; but even the “specialized” populations remain generalists). Second, FW has a broad native range across North America.…”

“…and Malus sp. are used in the west and east of the United States; Murphy & Loewy, ; Mason et al, ). Third, FW has previously been documented to consist of two morphologically and genetically distinct forms (Yang, Kawabata, Tufail, Brown, & Takeda, ) that differ in their life history traits, behaviour, host plant use and geographic distribution (e.g., Ito & Warren, ; Oliver, ).…”

Quantitative measure of fitness in tri‐trophic interactions and its influence on diet breadth of insect herbivores

“…We calculated relative abundance and relative host use following Mason et al. () and Murphy and Loewy (), and we analyzed the relationship between host plant use and abundance using linear regression for populations that used three or more host plant species.…”