Flooding does not always constrain riparian species composition: Evidence of environmental filtering and limiting similarity in floodplain forests

Background/Question/Methods

Flooding dynamics are expected to drive compositional patterns of woody overstory species in bottomland ecosystems, yet there is some evidence that flood variability may differentially constrain assemblage composition. Numerous studies have related distributions of plant species to floodplain landforms and hydrogeomorphic processes, documenting shifts in community composition along hydrologic and hydraulic gradients. Despite clear associations between plant composition and physical conditions, the relative importance of environmental filtering has not been explicitly evaluated in floodplain forests. The objective of this study was to assess phenomenological evidence consistent with ecological theories of community assembly in floodplain forests using ecologically-relevant functional traits. We quantified patterns of functional similarity and trait space occupancy for a regional sample of woody overstory plant communities that represented four hydro-physical settings (ecotypes), and evaluated how landforms may interact with broader flooding patterns to influence functional diversity. We hypothesized that the strongest evidence of environmental filtering would be detected in floodplain ecotypes associated with frequent, prolonged, or high-intensity flood events. We predicted that evidence for competitive interactions would be strongest in ecotypes with relatively infrequent, short-duration, and/or low-energy flood events. 

Results/Conclusions

We found evidence consistent with environmental filtering in two of the four ecotypes, evidence of competitive interactions in one ecotype, and inconsistent response patterns in the fourth ecotype. As assemblage richness increased in the two ecotypes associated with prolonged and frequent flooding, species were more likely to be functionally similar to each other, consistent with species packing, though the strength of this pattern varied. Rich, mesic conditions found in a third ecotype appeared to support diverse assemblages that were not products of environmental filtering. Instead, competition for light rather than plentiful nutrients may drive community patterns in this ecotype. We also found evidence that landforms interact with reach-scale flooding characteristics to influence forest assembly, but these interactions vary among ecotypes. Our results demonstrate that 1) flooding is not always an environmental constraint on floodplain forest ecosystems, and 2) that assemblage-landform associations are not reliably transferrable across different hydro-physical systems. This study highlights the critical need to quantify flooding dynamics in ecologically meaningful ways.