Faye Manning, P. Jefferson Curtis, Ian J. Walker , Jason Pither

1. Waterfowl are potential long-distance dispersal vectors for aquatic microbes such as diatoms, but supporting empirical data are scarce, especially concerning external transport on feathers.
2. We conducted an experiment designed to partially emulate diatom dispersal via adherence to waterfowl, and to evaluate the effects of relative humidity (RH) and exposure time on viability. We dipped individual breast feathers from mallards (Anas platyrhynchos) in a pure culture of the freshwater diatom Nitzschia pusilla Grunow, then, at room temperature (22.7 °C), subjected them to one of four contrasting levels of RH (ca. 8, 35, 70, 88%) crossed with one of four exposure times (10, 60, 120, 240 minutes) within a chamber through which air was passed continuously, mimicking light wind that might be experienced by diatoms adhered to subsurface feathers. We then gently placed the feather on sterile growth medium. After two weeks we used spectrofluorometry to detect diatom growth and thus diatom viability.
3. We found that exposure time and RH interacted significantly to affect diatom viability: the negative effect of exposure time was strongest under low RH conditions, but under high RH (88%) the probability of being viable was 0.84 for a ten minute exposure (95% confidence interval: 0.64 to 0.94), and 0.45 for four hours of exposure (95% confidence interval: 0.18 to 0.75).
4. We combined our experimental findings with geospatial data to predict the probability of potential dispersal via adherence to mallards throughout Nebraska, South Dakota, and North Dakota, which are situated within the central waterfowl migration flyway in North America, and host important mallard breeding grounds. Using published data about (i) mallard flight speeds, (ii) the geographic distribution of surface waters and of N. pusilla, and (iii) vapour pressure deficit (VPD; calculated using RH and air temperature) during the months of April and May, our geospatial model predicted high probabilities of potential dispersal, over tens to hundreds of kilometres, among water bodies of the central migration flyway.
5. Taken together, the results of our experiment and geospatial models provide novel insights into ectozoochory of freshwater diatoms, specifically that long-distance dispersal of diatoms via adherence to waterfowl feathers is highly plausible, particularly during the near-dawn hours when waterfowl flight activity peaks and VPD is low. Considered alongside previous evidence suggesting successful internal transport by waterfowl, we conclude that for freshwater diatoms ectozoochory is likely commonplace among waterbodies frequented by waterfowl.