For colonization success, should hosts and microbes travel alone, together, or swap partners along the way?

Microbiomes that enhance the performance of host plants are likely to be co-introduced with their host during colonization because of their intimate association. Yet, it is unclear how co-introduced microbes will impact host colonization, as both the microbiome and its effects could vary upon introduction into a new habitat. Using the duckweed Lemna japonica - a cosmopolitan, freshwater angiosperm - and its microbiome, we tracked the colonization of both plants and microbes during an experimental co-introduction in the wild. We tested how plant performance varied during colonization when plants were co-introduced with microbes from their home habitat or with microbes local to the introduced habitat. We found that plant performance was substantially reduced when plants were co-introduced with microbes from their home habitat (i.e. with microbes that are non-local to the introduced habitat), relative to hosts with a local microbiome. Moreover, negative impacts from the initial, non-local microbiome persisted for multiple host generations despite a rapid turnover in microbiome composition. Our results suggest that the initial microbiome plants are co-introduced and can leave lasting impacts on plant performance during colonization. Considering the identity of the co-introduced microbiome will therefore be critical to predicting plant colonization dynamics in an era of global change.