Recipient Organization
UNIV OF MASSACHUSETTS
(N/A)
AMHERST,MA 01003
Performing Department
Environmental Conservation
Non Technical Summary
A pressing problem for forest managers and society as a whole is the need to mitigate rapid environmental changes affecting all of life on earth. Third to only to habitat destruction and direct exploitation, biological invasions are a leading cause of extinction and loss of biodiversity in forest ecosystems and around the world (Dirzo and Raven 2003). One poorly understood aspect of biological invasions is that the introduction of non-native organisms to areas outside their home range leads to novel ecological interactions among species that have not historically co-existed. These new interactions raise key theoretical questions, such as: How do species adapt to new environments as they expand their range? Can novel species interactions alter longstanding interactions between native species? Do native species re-assemble into new food-webs with novel suites of species after invasion? We will address these questions in terms of invasive plants and their effects on native species of New England's forest ecosystems. A focal species of this work is the widespread invasive biennial plant, Alliaria petiolata (garlic mustard) which disrupts native plant-fungal interactions in Northeastern deciduous forests of North America. Unique plant chemicals exuded by the roots of garlic mustard disrupt symbioses between native plants and mycorrhizal fungi that live on their roots (Stinson et al., 2006), alter the diversity and composition of the soil microbiome (Barto et al., 2012; Anthony et al. 2017; Figure 2), and can thereby change the composition of native forest plant communities (Stinson et al., 2006; Haines et al., 2018). The glucosinolate, sinigrin, is one of the chemicals produced by garlic mustard that also disrupts the behavior, reproduction, and population dynamics of native butterflies in the genus Pieris that rely on host plants of moist deciduous forests. The native butterfly Pieris oleracea is a regionally declining native species that relies on a native mustard plant Cardamine dyphilla. Females mistake garlic mustard for their native host, but garlic mustard is toxic to the emerging caterpillars, creating an "ecological trap". Additionally, garlic mustard is known to alter arthropod communities and foodwebs (Smith-Ramesh 2016) while co-occuring invasive shrubs like Lonicera japonica (Japanese honeysuckle), Berberis thunbergii (barberry), and Rosa multiflora (multiflora rose) alter the habitat and food sources of native birds (Schlossberg, S. and King, D. I. 2010, White and Styles 1992, Gallinat et al. 2020; David et al. 2017, Van Der Putten et al. 2007). Moreover, changes in climate, deposition of nitrogen, and precipitation patterns are abiotic factors that could affect the performance of invasive plants and their impacts on native biodiversity (Dukes and Mooney 1999).We propose field observations and experiments to test three over-arching hypotheses:1. Novel species interactions drive forest biodiversity loss during the invasion process via disruption of longstanding native species interactions.2. Invasive plants provide food and habitat resources that change the nature of forest species networks and alter composition of native food-webs, both prior to and following their eradication.3. The impacts of invasive plants on native species interactions are contingent on abiotic conditions and may vary across the landscape and through time. Briefly, our approach will include regional field observations and experiments with the invasive plant species A. petiolata, along with the co-occurring invasive shrubs L. japonica, B. thunbergii, and R. multiflora (multiflora rose). To address H1-H2, we will first monitor the effects of these species on soil biota, birds, and butterflies as applicable, and relate these data to plant community structure, using nearby un-invaded sites as a reference for un-affected native species interactions. We will further test H1-H2 by conducting experimental eradications of garlic mustard to determine whether and how native communities re-assemble after an invasion event has been mitigated. To address H3, we will focus on A. petiolata's responses to experimental warming and nitrogen addition. We will assess potential impacts of altered plant chemistry on plant-butterfly and plant-soil interactions, both of which are predicted, by extension, to affect native forest vegetation structure.
Animal Health Component
(N/A)
Research Effort Categories
Basic
100%
Applied
(N/A)
Developmental
(N/A)
Goals / Objectives
The goals of this study are to provide insight into the effects of biological invasions on native species interactions in ways that reduce biodiversity in Northeastern forests. We plan to use both field and greenhouse experiments to:1. Determine how garlic mustard invasion and eradication affects the taxonomic structure and diversity of arbuscular mycorrhizal fungi and their associated native plant communities, leveraging existing metagenomics and vegetation datasets.2. Determine whether garlic mustard invasion and eradication positively or negatively impact a) b) food availability for forest songbirds; and b) oviposition and larval success of butterflies.3. Test the responses of leaf chemistry in garlic mustard to experimental warming and nitrogen deposition, and the concomitant effects on native butterflies.
Project Methods
Experiment 1: We currently maintain 45 3x3m long-term research plots in moist deciduous forests, where garlic mustard is invading: Harvard Forest, Petersham MA; Pittsfield State Forest, Pittsfield, MA; Trustees of Reservations properties Questing and McLennan Reserves, in Tyringham, MA and a private site in Deerfield, MA. Each site consists of 3 invaded plots, 3 non-invaded reference plots, and 3 plots where garlic mustard has been eradicated (5 sites x 3 treatments x 3 replicates). Vegetation is surveyed each year by identifying and tallying all species at every plot, and soil samples were collected for chemical and microbial analyses in 2013, 2015, and 2017. We will extract DNA from the soil samples we collected between 2013-2017 and utilize our newly-developed bioinformatics pipeline (Trautwig et al. in prep) to evaluate AMF taxonomic composition in response to garlic mustard invasion and eradication over time. Samples were collected as described in Anthony et al. 2017; We will assess relationships between key fungal, plant taxa, and edaphic soil features through PERMANOVA and regression modeling (generalized linear models) in order to evaluate abundance of key OTUs and plant species to derive key mechanisms of garlic mustard invasion. We will use NMDS and phylogenetic bar plots to visualize whether and how specific taxa associate according to the different treatments and relate vegetation to fungal community datasets.Experiment 2a:We will conduct surveys of food availability and behavioral observations of bird foraging on the dominant plants in the plots. Surveys for food availability will include a) larval Lepidoptera (caterpillars) and b) fruits from woody plants. We will conduct three timed-search sampling on a set of 8 woody plants and on 4 subplots of herbaceous plants per plot. Larvae will be measured to the nearest 0.5 mm. A similar approach will be used on the herbaceous subplots; we will conduct timed searches of all the herbaceous vegetation within 4 subplots per plot. Subplots will be selected at random within the plot, but at least one patch of garlic mustard will be included when that species is present in order to ensure we have sufficient samples to compare larval counts between garlic mustard and patches of native herb cover. Fruit surveys will follow the approach of Gallinat and colleagues (2020). Target bird species will include three forest breeding songbirds that forage on shrubs and herbaceous cover, the Black-capped Chickadee (Poecile atricapillus), Wood Thrush (Hylocichla mustelina), and Gray Catbird (Dumetella carolinensis). We will visit each plot for a minimum of 1 hour, aiming to get at least one foraging observation from each of the three target species. The frequency of foraging on native species versus invasive species will be compared for each species.Experiment 2b: We will survey the abundance of toothwort, native host plant of butterfly P. oleracea, butterfly oviposition rates, and larval abundances to determine how the presence and eradication of garlic mustard affects native plant-butterfly interactions. Toothwort abundances will be assessed as part of the ongoing botanical surveys of each plot in which all plant species are identified and tallied as described in Experiment 1; Oviposition rates by P. oleracea will be assessed weekly beginning in April, 2021, by surveying all Brassicaceae (native and invasive mustard plants) in each plot for the presence and abundance of eggs during each search. Caterpillar surveys will be conducted as part of the Lepidopteran counts and measurements described above in Experiment 2a. We will use GLM to compare abundance of toothwort in eradicated, invaded, and reference plots and to determine whether there are differences among the treatments in oviposition rates and larval abundance/larval size.Experiment 3a: We will grow garlic mustard from seeds collected at the study sites. Groups of 30 seedlings will be potted into 4" pots with ProMix soil and grown in one of four treatments in a randomized factorial design. Treatments will be: warming only, nitrogen only, combined warming + nitrogen, and a control (3 populations x 30 plants x 4 treatments = 360 plants). The plants will be grown in the CNS Research and Education Greenhouse at the University of Massachusetts Amherst, with temperatures set to +5ºC ambient conditions using hydroponic seedling warming mats (iPower) and those without seedling mats set atop a piece of black vinyl that matches the warming mats. An ammonium nitrate solution (NH4NO3) applied in an amount equivalent to ~40-50 kg 2 to plants in the nitrogen and combined warming + nitrogen treatments. Sinigrin concentrations at three points during the experiment will be assessed using High Performance Liquid Chromatography (HPLC) protocols (Popova and Morra 2014). Light/CO2 response curves will be generated using an LI-6400 ( Li-COR Biosciences, Lincoln, Nebraska, USA); C:N leaf tissue ratios will be quantified with an elemental CN analyzer; and leaf, root, and reproductive biomass will be measured.To test how changes in sinigrin affect phytotoxicity to tree species during the seedling stage, we will grow two-year old seedlings of Acer rubrum (red maple) and Quercus rubra (red oak) obtained from local seedstock in the greenhouse with extracts of garlic mustard from the global change treatments described above, with a water treatment as a control. We will quantify mychorrizal colonization of the roots using a published staining and microscopy protocol established in our lab (Wheeler et al., 2016).Experiment 3b: We will test the effects of our treatments on butterfly oviposition and larval survival by subjecting female P. oleracea to a subset of garlic mustard plants from the experiment: We will use naïve females raised in the lab on radish leaf (Raphanus raphaistrum sativus) to avoid the possibility that prior host exposure to garlic mustard or native toothwort affects oviposition choice (Steward and Boggs 2020). F1 females will be derived from butterflies caught at each of 3 field sites described above. We will test the preference for A. petiolata from the different treatments by 20 non-sib females from each butterfly source population, using a simultaneous choice assay with plants from the four treatments offered to the butterflies in 1x1m mesh cages. Eggs on each plant will be counted and removed once a day, until 50 eggs have been laid. Relative attractiveness of A. petiolata from each treatment will be calculated as the proportion of eggs laid on that plant. All plants used in the assays will be at the budding stageWe will test larval survival and growth rates of newly hatched first instar larvae from the females used in the oviposition trials above. We will use larvae from seven females from each origin population. Ten sibling larvae from each female will be placed individually on cuttings of A. petiolata from each treatment. Growth rates will be calculated for the duration of the first instar, or until death, whichever comes first. Growth rate will be measured as ((length of the newly hatched larva)-(length of the newly molted second instar larva or at death))/(# days to molt to the second instar or death). Mortality will be recorded twice daily. Leaf sinigrin concentration, combining cuttings from the same plant, will be analyzed as above for oviposition tests. Oviposition will be analyzed using generalized linear mixed models, with treatments as the fixed effect and collection site as a random effect (Steward and Boggs, 2020). Growth rate data will be analyzed using ANCOVA, with sinigrin concentration and recorded temperature in the environmental chamber as a covariate, and mortality data will be analyzed with a Cox proportional hazards test. We will thus help determine how future abiotic conditions might affect phytotoxicity of garlic mustard to native forest-reliant butterflies.