COS 39-4
A comparison of functional trait distributions among riparian floodplain landforms

Tuesday, August 6, 2013: 2:30 PM
L100C, Minneapolis Convention Center
Molly Van Appledorn, Geography & Environmental Systems, University of Maryland Baltimore County, Baltimore, MD
Matthew E. Baker, Geography & Environmental Systems, University of Maryland Baltimore County, Baltimore, MD
Background/Question/Methods

Riparian floodplain ecosystems are biological hotspots of diversity for a variety of organisms, but are highly sensitive to environmental changes occurring throughout the watershed. Measures of functional diversity offer predictive insight into how such changes could alter plant distributions within riparian ecosystems. In this study we ask: 1) Does functional diversity of overstory woody plant species vary among 3 floodplain landforms (levees, 1st bottoms, and backswamps), and 2) what functional traits are more strongly associated with each landform? Because each landform is associated with distinct hydrologic and hydraulic characteristics, we expected functional diversity and trait importance to vary in accordance with environmental conditions. We analyzed a unique, spatially extensive dataset that inventoried 170 overstory riparian plant communities on landforms along river-valley transects in Michigan's lower peninsula. Species abundance data were combined with functional trait data from publically-available databases to calculate Functional Dispersion (FDis) and abundance-weighted mean trait values by sampling plot. Because most traits were correlated, we used Principle Components Analysis to reduce dimensionality. We compared FDis, abundance-weighted trait, and PCA axis loading distributions across landform types using Kruskal-Wallis non-parametric tests. 

Results/Conclusions

We found that there was no difference in FDis values among levees, 1st bottoms and backswamp landforms (X2=1.5, p=0.47). The greatest range of FDis variability was found in levees; FDis variation may reflect the wide range of hydrologic and hydraulic conditions these systems experience. FDis values appeared to either decline or increase with distance from stream channel across transects, possibly indicative of a spatial process affecting trait diversity not captured in this study. However, trait importance did vary among landform types. Species inhabiting levees were significantly less fecund (X2=5.47, p=0.064 ) and less tolerant of anaerobic conditions than species from either 1st bottoms and backswamps (X2=5.40, p=0.067). Although not statistically significant, levee species were more able to re-sprout, and backswamp species appeared to be the most tolerant of anaerobic conditions in accordance; these results correspond with expectations about how trait-environment relationships vary throughout floodplain ecosystems. This study is the first to describe linkages among plant phenotypes and environmental conditions across multiple river-valley transects and a variety of hydrologic settings. Further analyses of trait-environment relationships have the potential to improve our understanding of controls on riparian plant diversity and predictions about the effects of environmental alteration.