Disturbance and biotic resistance are important factors driving plant invasions, but how these factors interact for plants with different modes of colonization (i.e., sexual and asexual) is unclear. We evaluated factors influencing the invasion of nonnative Phragmites australis, which has been rapidly expanding in brackish tidal wetlands in Chesapeake Bay. We conducted a survey of naturally occurring small-scale disturbances (removal of vegetation and/or sediment deposition) across four plant communities; determined the effects of small-scale disturbance and biotic resistance on P. australis seedling and rhizome emergence; and tested the effects of size and frequency of small-scale disturbances on seedling emergence and survival of transplanted seedlings. The results of our study demonstrate that the invasion window for seeds is in disturbed areas in high-marsh plant communities that flood less frequently; seedling emergence in undisturbed areas was negligible. Establishment of shoots from rhizome segments was low in all plant communities. Disturbance size and frequency had no significant impact on seed germination and seedling survival. Our findings provide evidence that small-scale within-wetland disturbances are important for the invasion of the nonnative lineage of P. australis by seeds in brackish tidal wetlands in Chesapeake Bay. Efforts to reduce disturbances, large and small, in wetlands can be used to limit P. australis invasion by seed, but invasion by rhizome is still likely to occur across many plant communities irrespective of the presence of disturbance.

Of 311 papers on wetland restoration, only 15 concerned large-scale experimentation in restoration sites. Most papers described what happened, reported on small field experiments, or discussed restoration targets. While these are important topics, our opinion is that we lose significant opportunities to learn how to recover populations, community structure, and ecosystem processes, and we limit our ability to document variability and whole-system responses, when we do not experiment at large scales. We suggest that, wherever possible, large projects facilitate field tests of alternative restoration approaches. Furthermore, we encourage researchers to take advantage of major restoration efforts by conducting large field experiments, assessing multiple responses, and offering restoration guidance in an adaptive framework.