Factors controlling spatiotemporal variability of soil carbon accumulation and stock estimates in a tidal salt marsh

Abstract

Abstract. Tidal salt marshes are important contributors to soil carbon (C) stocks despite their relatively small land surface area. Although it is well understood that salt marshes have soil C burial rates orders of magnitude greater than those of terrestrial ecosystems, there is a wide range in accrual rates among spatially distributed marshes. In addition, wide ranges in C accrual rates also exist within a single marsh ecosystem. Tidal marshes often contain multiple species of cordgrass due to variations in hydrology and soil biogeochemistry caused by microtopography and distance from tidal creeks, creating distinct subsites. Our overarching objective was to observe how soil C concentration and dissolved organic carbon (DOC) vary across four plant phenophases and across three subsites categorized by unique vegetation and hydrology. We also investigated the dominant biogeochemical controls on the spatiotemporal variability of soil C and DOC concentrations. We hypothesized that subsite biogeochemistry drives spatial heterogeneity in soil C concentration, and this causes variability in total soil C and DOC concentrations at the marsh scale. In addition, we hypothesized that soil C concentration and porewater biogeochemistry vary temporally across the four plant phenophases (i.e., senescence, dormancy, green-up, maturity). To test these interrelated hypotheses, we quantified soil C and DOC concentrations in 12 cm sections of soil cores (0–48 cm depth) across time (i.e., phenophase) and space (i.e., subsite), alongside several other porewater biogeochemical variables. Soil C concentration varied significantly (p < 0.05) among the three subsites and was significantly greater during plant dormancy. Soil S, porewater sulfide, redox potential, and depth predicted 44 % of the variability in soil C concentration. There were also significant spatial differences in the optical characterization properties of DOC across subsites. Our results show that soil C varied spatially across a marsh ecosystem by up to 63 % and across plant phenophase by 26 %, causing variability in soil C accrual rates and stocks depending on where and when samples are taken. This shows that hydrology, biogeochemistry, and plant phenology are major controls on salt marsh C content. It is critical to consider spatiotemporal heterogeneity in soil C concentration and porewater biogeochemistry to account for these sources of uncertainty in C stock estimates. We recommend that multiple locations and sampling time points are sampled when conducting blue C assessments to account for ecosystem-scale variability.