Ecological interactions between Euphydryas editha larvae and their host plants
Haan, Nathan L.
MetadataShow full item record
I examined ecological interactions between larvae of Euphydryas editha (Lepidoptera: Nymphalidae) and their host plants. These caterpillars, and the plants they eat, provide an intriguing system for studying several aspects of basic and applied ecology. In various chapters I focus on plant-mediated indirect effects, multi-trophic chemical interactions, ontogenetic niche shifts, the ecology and conservation of early-instar caterpillars, and the management and recovery of rare species. Euphydryas editha larvae are oligophagous herbivores, specializing on a few related host plant species. Two hosts I focus on are in the genus Castilleja, and the third host is Plantago lanceolata, an exotic species which E. editha recently incorporated into its diet. The plants E. editha specializes on produce iridoid glycosides, secondary compounds which are deterrent to many organisms, but which Euphydryas and some other specialists co-opt, sometimes accumulating them at high concentrations to defend against predators. Members of the genus Castilleja are hemiparasites; they form connections to other plants’ roots and extract resources from them. Therefore, Castilleja traits could depend on interactions with host plants, creating an indirect interaction pathway in which the plants Castilleja parasitizes affect herbivores (E. editha) by modifying the quantity or quality of food available to them. I grew C. levisecta with six different hosts, as well as without a host, while E. editha larvae fed on it. Castilleja size and leaf N concentrations depended on the host it parasitized, and larger, more N-rich plants resulted in larger E. editha larvae with higher survival rates. The ratio of two iridoid glycosides the larvae sequestered also depended on the identity of the host used by Castilleja. This work shows that hemiparasitic plant traits can mediate strong indirect interactions. In a field study, I compared outcomes for E. editha ssp. taylori larvae as they fed on C. levisecta, C. hispida, and P. lanceolata. This subspecies of E. editha is endangered, and inhabits grasslands in the Pacific Northwest. Managers involved in recovery efforts need information about the suitability of its host plants. Therefore, I placed clusters of E. e. taylori eggs on each species, and tracked larval survival from instar to instar. I also measured larval phenology, mass, and sequestration of iridoid glycosides. I tracked the senescence rates, pigmentation, and leaf nutrition (C:N ratios) for plants in each host species, and measured several environmental variables that could influence them. I found that survival depended on the host species that was eaten; it was highest on P. lanceolata, intermediate on C. hispida, and considerably lower on C. levisecta. Importantly, the factors influencing survival depended strongly both on the plant species larvae ate and their larval instar, with different predictors of survival for different instars. The overall differences in survival were mostly because of a disparity in survival during second instar. Larvae feeding on C. levisecta were less likely to survive from hatching to second instar, and from second to third instar, when plants were senescing, but this did not occur when they fed on the other two species. Group size was important to larvae feeding on P. lanceolata (but not on either Castilleja species); they were more likely to survive from second to third instar, and developed to fourth instar faster, when they were members of larger sibling groups. Survival from third to fourth instar was higher than for previous stages, and was not related to any of the variables that were measured. These findings related to larval survival show the importance of assessing survival instar by instar, as well as the importance of measuring outcomes for early-instar caterpillars. Larval mass was unaffected by any of the variables that were measured. Contrary to expectations, environmental variables like slope, aspect, and vegetation structure had no discernable effects on mass or development rate of the larvae. However, larvae that reached fourth instar earlier spent much more time feeding before entering diapause, suggesting butterflies that fly earlier (whose larvae consequently develop earlier) could have higher reproductive success. Environmental variables in this study had no measurable direct effects on larvae, but they could still influence them by changing the quality of their host plants: senescence of C. levisecta was faster in dry microsites than mesic ones, indicating plants growing in mesic microsites could be more phenologically compatible with E. e. taylori. There were also strong differences in the amounts of iridoid glycosides larvae were able to sequester from their hosts. They sequestered the compounds aucubin and catalpol from P. lanceolata, and when they fed on either Castilleja species, they sequestered these two compounds plus two others, macfadienoside and (putatively) methyl shanzhiside. The overall amounts sequestered from C. levisecta were lower than for the other two species, and may be low enough to leave them undefended against predators. In summary, I found that several outcomes for E. editha larvae are attributable to differences that occur within and among their various host plants. These differences can be attributable to innate species characteristics, but also to intraspecific differences caused by parasitic interactions and environmental factors. In this system, differences in host plants strongly influenced mass, growth rate, survival, and secondary chemical sequestration by the herbivore E. editha.
- Forestry