Abstract
Stream periphyton is an ideal study system for explaining how dispersal shapes community patterns. Few studies have tried to investigate periphyton metacommunities at the reach scale, and studies comparing local versus upstream periphyton propagule sources are lacking. We aimed to address these knowledge gaps by disentangling environmental constraints and dispersal sources, including dispersal hypotheses related to periphyton functional guilds. We covered 25-m sections of streambed with plastic silage cover sheets in three streams in Southern New Zealand, allowing river water to flow over the sheets. Samples on top of these sheets allowed periphyton colonisation only by drifting upstream propagules, while 'control' samples placed directly upstream of the plastic sheets were colonised by local and upstream propagules. We collected samples after 7, 14, and 25 days of colonisation. Response variables included periphyton biomass, community structure, and relative abundances of functional guilds. Control samples showed 1.5-6 times higher cell densities than plastic-cover samples, suggesting that local colonisation is very important for biomass accrual. Periphyton communities on both tile types became more similar to each other with time, indicating that environmental filters overcame effects of colonisation sources. While motile and flagellated taxa showed the ability to reach their preferred microhabitats in all streams, the responses of the remaining functional guilds did not follow the expected patterns. We conclude that periphyton community assembly strongly depends on reach-scale connectivity, which results in higher biomass accrual and community structure. These findings suggest that the mass effect paradigm is likely to be the principal metacommunity process shaping stream periphyton communities at the reach scale.