Genetic reconstruction of a bullfrog invasion to elucidate vectors of introduction and secondary spread

Abstract

Reconstructing historical colonization pathways of an invasive species is critical for uncovering factors that determine invasion success and for designing management strategies. The American bullfrog (Lithobates catesbeianus) is endemic to eastern North America, but now has a global distribution and is considered to be one of the worst invaders in the world. In Montana, several introduced populations have been reported, but little is known of their sources and vectors of introduction and secondary spread. We evaluated the genetic composition of introduced populations at local (Yellowstone River floodplain) and regional (Montana and Wyoming) scales in contrast to native range populations. Our objectives were to (1) estimate the number of introductions, (2) identify probable native sources, (3) evaluate genetic variation relative to sources, and (4) characterize properties of local‐ and regional‐scale spread. We sequenced 937 bp of the mitochondrial cytochrome b locus in 395 tadpoles collected along 100 km of the Yellowstone River, from three additional sites in MT and a proximate site in WY. Pairwise ΦST revealed high divergence among nonnative populations, suggesting at least four independent introductions into MT from diverse sources. Three cyt b haplotypes were identical to native haplotypes distributed across the Midwest and Great Lakes regions, and AMOVA confirmed the western native region as a likely source. While haplotype (H d = 0.69) and nucleotide diversity (π = 0.005) were low in introduced bullfrogs, the levels of diversity did not differ significantly from source populations. In the Yellowstone, two identified haplotypes implied few introduction vectors and a significant relationship between genetic and river distance was found. Evidence for multiple invasions and lack of subsequent regional spread emphasizes the importance of enforcing legislation prohibiting bullfrog importation and the need for continuing public education to prevent transport of bullfrogs in MT. More broadly, this study demonstrates how genetic approaches can reveal key properties of a biological invasion to inform management strategies.

Novel application of explicit dynamics occupancy models to ongoing aquatic invasions

“…Leung, Drake, & Lodge, 2004). For Yellowstone River bullfrogs, human-mediated propagule pressure is likely minimal since only two haplotypes occurred within this study frame, there was no molecular evidence for jump-dispersal events within the Yellowstone River, and these haplotypes are distinct from others in the region (Kamath et al, 2016).…”

“…Leung, Drake, & Lodge, ). For Yellowstone River bullfrogs, human‐mediated propagule pressure is likely minimal since only two haplotypes occurred within this study frame, there was no molecular evidence for jump‐dispersal events within the Yellowstone River, and these haplotypes are distinct from others in the region (Kamath et al., ). Consequently, I focused on natural spread and used covariates that relate the distance of a site from potential sources of initial and contemporary bullfrog propagules.…”

“…Similarly, human secondary translocations associated with the aquaculture trade spread bullfrogs non‐continuously across much of China, where the same two haplotypes occurred at multiple sites (Bai, Ke, Consuegra, Liu, & Li, ). In contrast, molecular data do not support evidence for human secondary translocations of bullfrogs in Montana and Wyoming, where bullfrogs are largely limited to their drainages of initial introduction and each drainage has distinct haplotypes (Kamath et al., ). The contrast in spread extent between bullfrog populations with and without evidence of human secondary translocations suggests that bullfrog management programs should not only focus prevention and control efforts on habitats with or proximate to permanent waters but also allocate resources to education and regulatory efforts to not transport bullfrogs.…”

“…Indeed, most of Montana has high suitability for bullfrog establishment since the environmental features of its waters closely match those of bullfrogs in their native range (Ficetola, Thuiller, et al, 2007). Yet, bullfrogs have been documented in relatively few waters even though they have occurred in Montana since as early as the 1920s (Kamath et al, 2016).…”

“…Panels A and B also show artificial waterbodies adjacent to the floodplain. The Yellowstone River flows southwest to northeast [Colour figure can be viewed at wileyonlinelibrary.com] following initial introduction events, is now likely driving this rapid invasion, as there is no molecular evidence for human-mediated jumpdispersal events (Kamath, Sepulveda, & Layhee, 2016). In this study, I used 5 years of annual bullfrog monitoring data at 92 sites to test if the probabilities of colonization and extinction for breeding bullfrog life stages were related to covariates that describe local habitat quality, propagule pressure and environmental stochasticity.…”