Patterns of association in a co-introduced insect herbivore and parasitoid wasp
Date
2017
Authors
Bowker, Cheryl Lindsay, author
Ode, Paul J., advisor
Black, William C., IV, committee member
Bjostad, Louis, committee member
Bowers, Deane, committee member
von Fischer, Joseph C., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Species within trophic networks that experience a range expansion can lead to populations of coevolved species re-associating in some areas, not associating in others, or creating novel associations with native species at one or more trophic level(s). Although most studies have focused predominantly on coevolved interactions between pairs of species, a growing number of studies investigate interactions within a broader community context across geographically widespread areas. By tracing the invasion routes of co-introduced species that share a close evolutionary history in their native range, I address the relative importance of long-term historical associations versus ecological fitting in community assembly. In the context of invasion ecology, ecological fitting refers to the process whereby exotic species form novel associations with native species that lack a shared coevolutionary relationship. In Chapter 1 of my dissertation, I review the literature that has taken a population ecology approach to understanding community assembly of introduced coevolved species. Collectively, these studies suggest that multiple factors ranging from local adaptation and ecological fitting are important in shaping multi-species associations in the introduced range and can occur simultaneously in one system. For example, coevolution can be an important factor in the re-association of many exotic plant-herbivore species, but ecological fitting also readily occurs as evidenced by the prevalence of native species host switching to exotics, and vice-versa. For the remainder of my dissertation, I used molecular methods to determine the invasion routes of the herbivorous parsnip webworm (Depressaria pastinacella; Lepidoptera: Depressariidae), and its primary parasitoid, Copidosoma sosares (Hymenoptera: Encyrtidae) to address the importance of shared population history and ecological fitting in shaping species interactions in introduced ranges. The coevolved reciprocal interaction between wild parsnip, parsnip webworm, and C. sosares has been well documented. Throughout its native range in Europe and much of its introduced range in North America, the parsnip webworm attacks wild parsnip (Pastinaca sativa; Apiaceae). Wild parsnips produce furanocoumarins, allelochemicals that are broadly biocidal because they intercalate DNA and interfere with transcription. Selection for chemically-based resistance occurs in plant populations by increasing concentrations of three furanocoumarins: xanthotoxin, bergapten, and sphondin. Genetic variation also exists in webworm populations in the rate at which these furanocoumarins are metabolized indicating that plant chemistry can act as a selective force on insect physiology. The wild parsnip was introduced to the United States (U.S.) as a food crop in the 17th century, where it escaped cultivation and spread throughout the U.S. By the 1860s, parsnip webworm was accidentally introduced to the U.S. and has re-associated with its coevolved wild parsnip plant, and formed novel associations with cow parsnip (Heracleum maximum; Apiaceae), a plant native to the U.S. Copidosoma sosares is also found in the U.S., but its arrival date is unclear. Copidosoma sosares attacks webworms on both wild parsnip and cow parsnip, but its distribution is patchy compared to parsnip webworms; with rare exception, C. sosares populations are restricted to the western U.S. In Chapter 2, I used a mitochondrial molecular marker to determine: 1) the source population(s) of U.S. and New Zealand parsnip webworms, and 2) whether parsnip webworm populations in the U.S. or Europe are locally adapted to the host plant species they attack. I found that parsnip webworms experienced a genetic bottleneck during introduction to the U.S. and New Zealand. U.S. parsnip webworm populations were founded by a single (or a few) population(s) of parsnip webworms from the British Isles. British Isles populations of parsnip webworm are themselves genetically isolated from continental European parsnip webworm populations. The introduction pattern of webworms stands in contrast to that of its host plant, wild parsnip, in the introduced ranges of the U.S. and New Zealand. The acquisition of the non-coevolved cow parsnip by webworms in the U.S has led to genetic divergence from webworms feeding on their coevolved plant species. In contrast, European webworm populations are not restricted to the host plant species on which they feed. In Chapter 3, I describe the development of novel microsatellite markers for tracing the routes of C. sosares within the introduced range of the U.S. Thirty-four candidate loci were identified, 12 of which were ultimately chosen for detailed screening. Seven of the 12 loci were polymorphic, but only 5 of the 7 were successfully amplified across samples and ranges. Of these 5 polymorphic loci, there were 3 - 9 alleles per locus. Inbreeding coefficients and the null allele frequency ranged from -0.04 to 0.74 and 0 to 0.73, respectively. After Bonferroni correction, only one locus (Csos 4) significantly deviated from Hardy-Weinberg Equilibrium (HWE, P < 0.05) across all populations sampled. When the data were partitioned by location (European and U.S. populations), loci Csos 2, Csos 3, and Csos 4 in European populations conformed to HWE (P > 0.05), whereas loci Csos 1, Csos 2, and Csos 3 in the U.S. conformed to HWE (P>0.05). No pairs of loci demonstrated linkage disequilibrium, neither across all populations, nor when the data was partitioned into European and U.S populations (P > 0.05). These results indicate the developed microsatellite markers for C. sosares are well suited for use as genetic markers for elucidating the population structure of C. sosares. In Chapter 4, I used the microsatellite markers developed in Chapter 3 and an additional mitochondrial marker to test three hypotheses: 1) C. sosares populations in the U.S. came from the same European location as their webworm hosts in Europe (Host-Pursuit Hypothesis), 2) once in the U.S., C. sosares followed a similar host plant switch onto cow parsnip as parsnip webworms (Continued Host-Pursuit Hypothesis), and 3) C. sosares populations will have limited geneflow between sites that are geographically farther apart, and differ in elevation (isolation by distance and elevation). The molecular data indicate the Netherlands and several other European sites served as sources for U.S. C. sosares populations, following the predictions from the Host-Shift Hypothesis. European C. sosares populations from hogweed have directly established on U.S. cow parsnip, and U.S. C. sosares populations on wild parsnip have host plant switched to cow parsnip. However, C. sosares geneflow is not restricted to either host plant species in the U.S., contrary to our findings with U.S. webworms. Overall, the factors that can influence the establishment of an herbivore (i.e., host plant species, concentrations of plant toxins) in an introduced area, may not be important for its primary parasitoid.