Rare, long-distance dispersal events are a key process in generating and maintaining patterns in biological diversity and species distributions across space and time. The 9.0 magnitude earthquake that struck the eastern coast of Japan in 2011, and the subsequent 38 m high tsunami washed large amounts of shoreline debris into the Pacific Ocean that led to a large-scale biological rafting event carrying nearly 300 marine species to the western shores of North America. Whether oceanic, trans-Pacific dispersal via rafting generates long distance dispersal events for small, flightless, terrestrial species is unknown. By sampling beach debris associated with known hot-spots of tsunami debris along the north and east shores of Graham Island, Haida Gwaii, Canada, I document significantly dissimilar invertebrate communities associated with tide-line beach debris and the occurrence of several putative Japanese species of soil-dwelling mites (Acari: Oribatida). Previous explanations of Haida Gwaii’s unique flora and fauna have been attributed to a proximity to the Beringian land bridge and the accumulated evidence of near-offshore glacial refugia during the last glacial period. However, my research also suggests that stochastic, trans-Pacific rafting events contribute to the biodiversity and biogeography of soil communities on the west coast of North America.

Species abundance data were generated at locations with previously documented Japan Tsunami Marine Debris (JTMD) along the north and east shorelines (North Beach and East Beach) of Graham Island, Haida Gwaii, British Columbia, Canada. Samples were collected from beach, tidal edge, and forest locations in Summer 2017 over a one-week period following the morning high tide.  Beach samples consisted of approx. 20 cm2 of debris from multiple items collected along the shoreline representing the daily tide line.  Shoreline vegetation and leaf litter at the tidal edge zone were sampled to represent the first possible, but likely non-optimal habitat that would allow for colonization of new terrestrial migrants.  Forest floor samples were also collected for comparison with long-term resident species.  Where possible beach, edge and forest samples were collected at the same location.  All samples were extracted on portable Berlese faunal extractors over 72 hrs into 75% ethanol, and specimens observed were subsequently identified to species-level using keys from the primary literature, and compared against >3000 curated specimens in the author’s taxonomic collection.