Examining plant composition to predict methane emissions from flooded sediments

Background/Question/Methods In anoxic sediments, wetland vegetation is a dominant pathway for methane release into the atmosphere. Plant composition plays an important role in determining methane emissions in wetlands, but few studies examine how methane dynamics differ across a range of species. We hypothesized that methane emissions differ across plant functional groups, with patterns distinctive among forbs, clonal dominants, and tussock-forming graminoids. Additionally, we assessed the role of soil microbial communities in influencing methane flux across plant species.

Results/Conclusions Plant composition influenced methane emissions, such that the tussock- or clump-forming species (Carex stricta and Scirpus atrovirens) showed the highest emission levels. The two clonal dominant, invasive plants (Typha angustifolia and Phalaris arundinacea) had the lowest methane emissions of all seven graminoid species, and the forbs (Mimulus ringens and Verbena hastata) showed no detectable methane flux from leaves. Plant biomass was useful in predicting methane flux rates only for graminoid plants, showing decreased methane emissions with greater plant biomass. Analysis of soil archaeal methanogen (methane producer) communities (as revealed by 16S rRNA T-RFLP analysis) showed that methanogen community structure had no effect on methane flux. Although the relative proportions of the different terminal fragments differed across plant species, they were not correlated with methane emissions or plant biomass. Similarly, the phospholipid fatty acid (PLFA) profiles indicative of bacterial methanotrophs (methane consumers) showed no relation with methane flux. Analysis of signature ¹³C-PLFAs suggests no Type II (18:1ω7c) methanotrophic utilization of recently derived carbon from the plants. This study indicates that plant functional groups and plant biomass may be useful predictors of methane emissions from flooded sediments, while microbial community structure showed no predictive trends.