Complex invader-ecosystem interactions and seasonality mediate the impact of non-native Phragmites on CH 4 emissions

Invasive plants can influence ecosystem processes such as greenhouse gas (GHG) emissions from wetland systems directly through plant-mediated transfer of GHGs to the atmosphere or through indirect modification of the environment. However, patterns of plant invasion often co-vary with other environmental gradients, so attributing ecosystem effects to invasion can be difficult in observational studies. Here, we assessed the impact of Phragmites australis invasion into native shortgrass communities on methane (CH 4 ) emissions by conducting field measurements of CH 4 emissions along transects of invasion by Phragmites in two neighboring brackish marsh sites and compared these findings to those from a field-based mesocosm experiment. We found remarkable differences in CH 4 emissions and the influence of Phragmites on CH 4 emissions between the two neighboring marsh sites. While Phragmites consistently increased CH 4 emissions dramatically by 10.4 ± 3.7 µmol m −2 min −1 (mean ± SE) in our high-porewater CH 4 site, increases in CH 4 emissions were much smaller (1.4 ± 0.5 µmol m −2 min −1 ) and rarely significant in our low-porewater CH 4 site. While CH 4 emissions in Phragmites -invaded zones of both marsh sites increased significantly, the presence of Phragmites did not alter emissions in a complementary mesocosm experiment. Seasonality and changes in temperature and light availability caused contrasting responses of CH 4 emissions from Phragmites - versus native zones. Our data suggest that Phragmites -mediated CH 4 emissions are particularly profound in soils with innately high rates of CH 4 production. We demonstrate that the effects of invasive species on ecosystem processes such as GHG emissions may be predictable qualitatively but highly variable quantitatively. Therefore, generalizations cannot be made with respect to invader-ecosystem processes, as interactions between the invader and local abiotic conditions that vary both spatially and temporally on the order of meters and hours, respectively, can have a stronger impact on GHG emissions than the invader itself.