Research

Currently, my research focuses on interactions between a threatened plant, an invasive plant, and several introduced weevil species. Rather than go into an introduction here, I would refer you to the “Current Research” section of the About tab, and will go into detail about my current projects here.

I have four sections:
Volatile Compounds and Host Plant Selection
Weevil Feeding Preferences on Heads and Leaves
Weevil Feeding Preferences on Heads and Leaves
Weevil Overwintering and Population Size Estimate

Volatile Compounds and Host Plant Selection
Reason for Investigating
We noticed that not all Pitcher’s thistle plants were attacked evenly in the field, and thought that this may be due to differences in volatile compound emissions by the plants in question.

scent_vials

Vials containing plant volatile samples awaiting analysis.

Methods
To examine the composition of volatile compounds in C. pitcheri, we used dynamic headspace methods as outlined in Theis and Raguso (2005). We used three different classes of heads to determine differences between these plants. The classes were: without weevils, with weevils, and with weevils introduced (a plant that initially had no weevils, to which we introduced weevils). In addition to these three classes, we also sampled fully-open heads of two phenological stages. These two stages are mostly male (>50% of florets exhibiting pollen without an excised stigma) and mostly female (>50% of florets with excised stigma; lacking pollen). In each category, we used 10 plants, which is a typical sample size to capture most of the variation that would exist within a given class.

For each plant, we bagged one head and sampled for six hours (starting at 8am), using personal air pumps that had been calibrated to the same flow rate. These moved the air through filters containing Porapak absorbent, which captures the scent compounds. After the six hours, we eluted the samples into hexane for storage and analysis. Samples are kept on ice until they are analyzed. We collaborate with Nina Theis (Elms College) on this project, and she runs the samples on a GC-MS.

Anticipated Results
We expect to see differences in volatile emission between the plants that are prone to weevils and those that are not. In addition, we expect to see differences as the plants progress through their phenological stages. Additionally, as Nina Theis has previously quantified volatiles of Cirsium arvense and Cirsium repandum, we expect to find that Cirsium pitcheri is closer to C. repandum in volatile composition than to C. arvense. This is because, in the phylogeny of the genus, New World Cirsium species typically cluster together apart from Old World species. In keeping with that pattern, we do predict that C. pitcheri will contain, as a large fraction of its blend, the compound phenylacetaldehyde, which has been shown to attract L. planus and also to be important in both pollinator and florivore attraction (Theis 2006).

Why This Matters
If we know more about what attracts the weevil L. planus to its host plants, we can better craft management strategies to limit its damage to C. pitcheri populations. It may be possible to identify compounds that are repellant to the weevils, and which may be applied to the plants, thus preventing weevil damage, while allowing pollinators access. Additionally, the change of compound emission as the heads develop likely has an impact on which pollinator species are present. With this data, in conjunction with pollinator observations, it may be possible to identify which stages are more important for successful pollination, and which stages would be more conducive to weevil management, again to limit pollinator impact.

As for the relationship between the compounds of C. pitcheri and the two other Cirsium spp. that have been examined, this knowledge would help clarify the relationship between these species and help with determining how generalizable the results are to other US Cirsium spp.

Top of Page

Weevil Feeding Preferences on Heads and Leaves
Reason for Investigating
The reasons for investigating feeding preferences are twofold because of the species involved. First, L. planus is noted as utilizing Centaurea spp. in its native range. It has not been recorded as using these species in the US. Due to the prevalence of the invasive Centaurea stoebe in and around many C. pitcheri populations, there were concerns that the weevil could be using Centaurea as a reservoir, which would complicate potential weevil management strategies.

The second reason is because of the related weevil species Larinus minutus and Larinus obtusus. These weevils are approved biocontrol agents for Centaurea diffusa and Centaurea stoebe in the US, and are generally specific to knapweeds. However, as with L. planus, there are reports of them using other genera, including Cirsium, in their native range. Because they are actively distributed in Wisconsin and Michigan, states which make up the vast majority of the population of Pitcher’s thistle, there were concerns that they might jump to Cirsium pitcheri. As the thistle faces a large number of threats already, this is clearly not a desirable outcome.

Methods
To get at these feeding preferences, we used standard biocontrol choice and non-choice tests. In non-choice tests, weevils were presented with only one plant species of interest. In choice tests, a Cirsium species (C. arvense for L. planus and C. pitcheri for the others) and C. stoebe were present. We substituted C. arvense for C. pitcheri due to insufficient C. pitcheri to use in all tests. These tests took place in a growth chamber located at the Chicago Botanic Garden. The choice of a growth chamber rather than a field test was necessary to prevent the accidental introduction of any of the weevil species to the northeast Illinois area. The growth chambers were tuned to match the environmental conditions, particularly light and temperature, typically present in the field at that time of year.

obtusus_on_cipi

A Larinus obtusus individual on a Cirsium pitcheri head.

In both test types, plants were represented by cut stems with flowering heads taken from individuals grown in a common garden on site, and placed in vases containing nutrient solution. Weevils were released at 9 AM and the weevils were watched continuously for two hours to note initial behaviors. During the remainder of the daylight hours, observations of the weevils took place for 30 minute increments every hour. After this first day, weevils are observed for 10 minute periods three times daily. During these observational periods, we keep track of the number and duration of an assortment of activities, including feeding, mating, tasting, oviposition, etc. The trials run for a total of four days, a time period chosen because past research had shown this to be the period of peak ovipositions on plants used in this style of test (Louda et al. 2005). Each weevil species will go through two rounds of both choice tests and non-choice tests.

 

 

Anticipated Results
We expect that these three weevil species will utilize the species in non-choice tests to some extent, while in choice tests, they will tend to use their preferred host.

Why This Matters
The knowledge of how these weevils utilize these plants is critical to their use and management in the field. Because invading insects tend to move through a landscape following their preferred host plants, control (in the case of L. planus) and release plans (in the case of the others) depend on knowing which plants can be successfully used for feeding and/or oviposition. Because of the interplay between these non-native weevils, the invasive C. stoebe, and the threatened C. pitcheri, this information may lead to improved control strategies for the invasive plant, management of the rare plant against the weevils, and better screening potential for biocontrol insects.

Top of Page

Weevil Feeding Preferences on Leaf Tissue
Reason for Investigating
Due to the inherent limitations of sample size with the methods described in the previous section, we decided to run the same tests using leaf material in petri dishes. This method has been established in previous studies (Coyle et al. 2012).

obtusus_petri

Larinus obtusus in a feeding trial in petri dishes.

Methods
For these tests, we used only Larinus minutus and Larinus obtusus because by this time of the summer, the L. planus individuals were at the end of their lifecycle and were no longer behaving in ways that were typical during peak breeding season. We took leaf samples of known area and placed them on moistened filter paper. Each dish received one weevil. The weevils were randomly assigned to a given treatment (with the treatments being the versions of choice and non-choice mentioned in the above experiment). Petri dishes were stored in an incubator set to match external environmental conditions. Weevils were allowed to feed for a period of 48 hours. Deionized water was used to re-moisten the filter paper, as needed. After this period of 48 hours, weevils were removed and the leaf fragments were scanned so that the area consumed could be calculated.

Expected Results
We expect similar results with these tests as with the clipped plant feeding trials from the section above. Because this type of test uses only leaf tissue, it is not able to inform about oviposition or other tendencies, but can demonstrate breadth of feeding.

Why This Matters
These tests are confirmatory to the tests mentioned in the previous section. As with those tests, they are important to elucidate the tendencies of these weevils to feed on certain plants in this system. Because feeding on the vegetative stage is known to influence the overall fecundity of an individual, knowledge of feeding strategies is especially important in evaluating any potential risk to C. pitcheri from L. minutus and L. obtusus.

Top of Page

Weevil Overwintering and Population Size Estimate
Reason for Investigating
During preliminary observations in October of 2013, we were unable to locate any Larinus planus individuals in areas where they were known to occur in the previous summer. In the literature, L. planus is reported to overwinter in the leaf litter at the base of its host plants. In the dunes, this litter layer is reduced/non-existent. With these observations, we were curious where/how the weevils were overwintering. The population size estimate comes from mark-recapture formulae and came into play based on marked individuals and multiple surveys. Originally, we were taking data that would almost be used for mark-recapture if we changed our methods slightly, so we changed our methods and now will take data for both.

marking_planus

Preparing to mark three Larinus planus individuals in the field.

Methods
In fall of 2014, we marked weevils, both in the dune C. pitcheri population and in known C. arvense populations with L. planus present. Weevils were marked with nail polish on their ventral side, with different colors corresponding to different locations. Weevils were released where they were marked. In spring of 2015, we will return to determine potential movements of weevils between different habitat patches. Additionally, we plan to mark more weevils and do a mark-recapture study to determine the population size of weevils at our primary field site.

Expected Results
We expect that there is some movement in late fall away from the inhospitable dunes and that this may reverse in the spring. Additionally, based on some preliminary estimates, we estimate that there will be at least 500 weevils present in the main population at our field site.

Why This Matters
Knowledge of wintering patterns and population size is important for effective management of this threat to C. pitcheri. Depending on their movements, strategies could be employed to prevent the majority of weevils from re-entering the dune system (traps, etc). Additionally, a population estimate would allow better management as it would provide a better resolution as to the scale of the problem in populations that are threatened by the weevil.

 

Top of Page

Works Cited
Coyle, D. R., W. J. Mattson, M. S. Jordan, and K. F. Raffa. 2012. Variable host phenology does not pose a barrier to invasive weevils in a northern hardwood forest. Agricultural and Forest Entomology 14:276–285.

Louda, S. M., T. A. Rand, A. E. Arnett, A. S. Mcclay, K. Shea, and A. K. McEachern. 2005. Evaluation of Ecological Risk to Populations of a Threatened Plant from an Invasive Biocontrol Insect. Ecological Applications 15:234–249.

Theis, N. 2006. Fragrance of Canada thistle (Cirsium arvense) attracts both floral herbivores and pollinators. Journal of Chemical Ecology 32:917–927.

Theis, N., and R. Raguso. 2005. The effect of pollination on floral fragrance in thistles. Journal of Chemical Ecology 31:2581–2600.

Top of Page