Genetic structure of invasive baby’s breath (Gypsophila paniculata) populations in a freshwater Michigan dune system

Abstract

29Coastal sand dunes are dynamic ecosystems with elevated levels of disturbance, and as such they 30 are highly susceptible to plant invasions. One such invasion that is of major concern to the Great 31Lakes dune systems is that of perennial baby's breath (Gypsophila paniculata). The invasion of 32 baby's breath negatively impacts native species such as the federal threatened Pitcher's thistle 33 (Cirsium pitcheri) that occupy the open sand habitat of the Michigan dune system. Our research 34 goals were to (1) quantify the genetic diversity of invasive baby's breath populations in the 35Michigan dune system, and (2) estimate the genetic structure of these invasive populations. We 36 analyzed 12 populations at 14 nuclear and 2 chloroplast microsatellite loci. We found strong 37 genetic structure among populations of baby's breath sampled along Michigan's dunes (global 38 F ST = 0.228), and also among two geographic regions that are separated by the Leelanau 39 peninsula. Pairwise comparisons using the nSSR data among all 12 populations yielded 40 significant F ST values. Results from a Bayesian clustering analysis suggest two main population 41 clusters. Isolation by distance was found over all 12 populations (R = 0.755, P < 0.001) and 42 when only cluster 2 populations were included (R = 0.523, P = 0.030); populations within cluster 43 1 revealed no significant relationship (R = 0.205, P = 0.494). Private nSSR alleles and cpSSR 44 haplotypes within each cluster suggest the possibility of at least two separate introduction events 45 to Michigan. 46 47

Old Meets New: Combining Herbarium Databases with Genetic Methods to Evaluate the Invasion Status of Baby’s Breath (Gypsophila paniculata) in North America

“…Leimbach-Maus et al, 2018a) show that the SBD-MI and AD-MI populations have distinct308 DNA haplotypes compared to the PS-MI population and other more northern Michigan 309 populations not included in this study. The combination of these data suggest that SBD-MI and 310 AD-MI are likely not the result of serial founding events from the source population of PS-MI.311 A more likely explanation for the distinct patterns observed among our populations could 312 be a signature of G. paniculata's horticultural past.…”

“…Table 1). Samples from two additional locations in Sleeping Bear Dunes 134 National Lakeshore, MI and Arcadia Dunes, MI were collected in the summer of 2016 (Table 1) 135 (Leimbach-Maus, Parks, & Partridge, 2018a). Leaf tissue was collected from 15-30 individuals 136 per location (5-10 leaves per plant).…”

Effective dispersal patterns in prairie plant species across human‐modified landscapes

“…Interpretation of effective dispersal requires consideration of spatial and temporal scales affecting populations (Robledo-Arnuncio et al, 2014;Twyford et al, 2020); for most plants, individual genetic relationships within fine scales (e.g., scale of less than 1 km radius) will be dominated by annual dispersal events (Grasty et al, 2020), while population structure at large scales (e.g., scale of 100 km radius) will be formed by cumulative multi-generational gene flow (Elleouet & Aitken, 2019). Depending on the dispersal ecology of a species, studies conducted at the mesoscale (e.g., scale of 10 km radius) may capture the interface between drivers of dispersal, such as dispersal vector behaviour or landscape features, and evolutionary consequences, such as prolonged gene flow, drift and colonization dynamics (Arredondo et al, 2018;Leimbach-Maus et al, 2018;Schweiger et al, 2004).…”

“…Isolation by distance models are most reliable in fine-scale homogeneous landscapes that experience consistent conditions for dispersal or large scales at which coalescence relationships dominate genetic structure estimates. At the mesoscale, evolutionary relationships are often faint, and IBD models do not capture spatial variation in landscape features, habitat quality and fragmentation, or dispersal vector behaviour (Arredondo et al, 2018;Leimbach-Maus et al, 2018;Mateo-Sánchez et al, 2015). In this context, dispersal may be influenced by topographical or land use features, such as land elevation, tree canopy coverage, rivers or streams, urbanization, agricultural conversion and meteorological events Sork & Waits, 2010).…”

“…Isolation by resistance is based on circuit theory and postulates some landscape features are more favourable to dispersal, while other landscape features constrain dispersal (termed 'conduits' and 'resistors', respectively) (McRae et al, 2008). By comparing landscape features to population genetic differentiation, we can infer how the rate of gene flow fluctuates across heterogeneous landscapes (Chiappero et al, 2023;Leimbach-Maus et al, 2018;Segelbacher et al, 2010;Sork & Waits, 2010).…”