Methane emissions from created and restored freshwater and brackish marshes in southwest Florida, USA

It is important to estimate greenhouse gas emissions from newly created and restored wetlands so that we learn how to design them to minimize these emissions. Spatial and temporal patterns of methane emissions were measured in three wetland marsh complexes in southwest Florida: a created freshwater marsh on a university campus, a restored brackish/salt marsh on the fringe of a mangrove (Rhizophora mangle) coastal swamp; and a natural freshwater marsh adjacent to a cypress (Taxodium distichum) strand. Non steady -state rigid chambers were used twice a day (morning and afternoon) on a monthly schedule for 13 months. The mean +/- standard error (median) of methane emissions from permanently and intermittently flooded areas were 124 +/- 47 (0) and 0.8 +/- 0.7 (0) g-CH4-C m(-2) y(-1) in the created freshwater marsh, and 58 +/- 14 (9.4) and 0.1 +/- 0.3 (0) g-CH4-C m(-2) y(-1) in the restored brackish marsh. Methane missions were 0.8 +/- 0.5 (0) g-CH4-C M-2 y(-1) in the intermittently flooded natural freshwater marsh that served as our control. In general, average methane emissions were higher from permanently than intermittently flooding areas (P<0.05); our data suggested that ebullition may have occurred in permanently flooding areas and caused up to 6.3 g-CH4-C m(-2) d(-1) in pulses in the created marsh and 1.7 g-CH4-C m(-2) d(-1) in the restored marsh. Removing ebullition outliers, methane emission estimates decreased by onefourth to one-half the annual rates given above to 30 g-CH4-C m(-2) y(-1) in permanent created freshwater marshes and 27 g-CH4-C m(-2) y(-1) in the restored brackish marshes. There were no significant differences in methane emissions from created, restored or natural marshes in intermittently flooding areas (P > 0.05). CH4 emissions exhibited strong quadratic and linear relationships with water depth when all sites were compared (P< 0.05), but no significant relationships with soil temperature or salinity. These results give useful information on ways to minimize methane emissions when creating and restoring wetlands. But these results should be only one of the considerations for wetland design as other ecosystem services should also be considered. (C) 2016 Elsevier B.V. All rights reserved.