Metacommunity studies in running waters have progressed rapidly in recent decades, and periphyton metacommunities have received special attention. Stream periphyton are an ideal study system due to their ecological relevance and the relative ease of testing theoretical questions on these communities of small species with short generation times. While most studies conclude that environmental controls are the principal driver in periphyton communities, there is also evidence supporting the importance of spatial processes. Several authors have pointed out that a key challenge in stream metacommunity research is to disentangle the influences of environmental filters and dispersal constraints on metacommunity organization. In this context, two important existing knowledge gaps are related to the role of three-dimensional biofilm mats and dispersal processes in periphyton metacommunities. Biofilms are matrix-enclosed microhabitats that periphyton species inhabit. Inside these biofilms, competition for resources and metabolic cooperation occurs between hundreds of species at the same time, and thick biofilms increase carrying capacity, functional diversity, and species interactions, hindering our ability to make predictions about periphyton community dynamics. On the other hand, periphyton species traits and running-water ecosystems favor long dispersal processes, and periphyton species show high capability to colonize adjacent substrates, thus enabling short dispersal processes as well. However, a lack of studies on dispersal processes in periphyton makes it difficult to understand how species traits and propagule sources interact with metacommunity processes. To help address these knowledge gaps, I carried out three novel field experiments to investigate how periphyton species interact at different spatial scales and with environmental-dispersal filters under semi-controlled but realistic scenarios, complementing existing survey-based studies and laboratory experiments.

 The aim of my first data chapter (The long) was to estimate dispersal distances of stream periphyton taxa in the field. As done for other organism groups, periphyton dispersal patterns could potentially be assessed via propagule traps below a “source point”. However, identifying such a source point for algal species within a river continuum would be extremely challenging. We therefore proposed using lake phytoplankton algae as a proxy for the dispersal of periphyton algae in the stream, assuming the lake outlet as a source point. We estimated kernel distributions as dispersal models, using phytoplankton density in periphyton and settling velocity from laboratory experiments as predictors. Median travel distances were estimated to be around 900 m, and dispersal probability dropped abruptly after 1.5 km downstream of the outlet. Settling velocity dispersal models suggested a shorter median travel distance, around 500 m, but with higher probabilities of long-distance dispersal events. This chapter provided novel insights into understanding the connectivity of periphyton metacommunities.

To determine the roles of different dispersal sources during habitat colonization, we focused on two key sources in Chapter 3 (The short): local propagules, namely colonization from the nearby streambed, versus upstream propagules, namely periphyton colonization by drifting taxa. We covered 25-m reaches of streambed with plastic silage cover sheets in three streams, then placed ceramic tiles over these sheets for drifting algae colonization. We also exposed ‘control’ tiles on the surface of periphyton-covered streambeds directly upstream of the plastic sheets. Tiles were sampled after 7, 14 and 25 days of colonization, and taxa were classified using trait-based guilds related to dispersal capabilities. Plastic-cover tiles had ca. 3.5x less cell density and biomass than control tiles overall, and motile taxa showed higher dispersal capabilities. The remaining guilds showed results inconsistent with theoretical expectations. These results could be explained mainly by high dispersal rates from nearby sources, and only a secondary role for algal drift from upstream.

In Chapter 4 (The twist), we assessed the role of the periphyton biofilm mat in driving community structure, by carrying out a translocation experiment. We exposed ceramic tiles for 4, 11 and 21 days in a source river, generating a biofilm gradient (thin, medium, thick). These tiles were translocated into 10 receiver rivers and exposed there for 25 days before collection for laboratory analyses. We also exposed clean tiles in all rivers and considered these as our reference communities. After translocation, tiles exposed for 11 and 21 days in the source river were more similar to the source river than to their receiving river, indicating that thick biofilms slowed community turnover. This pattern was modulated by pollution levels: translocated tiles in low-phosphorus rivers exhibited greater similarity to both source and receiver rivers than tiles translocated into high-phosphorus rivers. We suggest that, in low-pollution rivers, communities were dominated by taxa from the receiver river and the source river that were adapted to local conditions. By contrast, in high-pollution rivers, high dispersal rates determined a faster shift to the new community.

To synthesize these findings, the evidence from Chapter 2 suggests that periphyton communities are highly connected at the reach scale. Further, in just one dispersal event an algal taxon can reach a new river, indicating high dispersal rates and complex biogeographical scenarios. However, results from Chapter 3 indicate that the role of drifting algae in community assemblages may be secondary, and that the principal propagule source is from nearby streambed patches instead of upstream. Finally, in Chapter 4 we observed that polluted rivers pushed dispersal processes through thicker biofilms, exhibiting faster taxa turnover and presence of rare taxa. Our experiments highlight the role of dispersal processes in periphyton metacommunities. Pollution levels interacted with periphyton productivity, playing a role in habitat selection but also in modulating dispersal rates. Thus, biofilm thickness shifted periphyton metacommunity dynamics by reducing environmental constraints and increasing dispersal rate, indicating that metacommunity processes change with time in stream periphyton.