Soil Iron Content as a Predictor of Carbon and Nutrient Mobilization in Rewetted Fens

Abstract

Rewetted, previously drained fens often remain sources rather than sinks for carbon and nutrients. To date, it is poorly understood which soil characteristics stimulate carbon and nutrient mobilization upon rewetting. Here, we assess the hypothesis that a large pool of iron in the soil negatively affects fen restoration success, as flooding-induced iron reduction (Fe3+ to Fe2+) causes a disproportionate breakdown of organic matter that is coupled with a release of inorganic compounds. We collected intact soil cores in two iron-poor and two iron-rich drained fens, half of which were subjected to a rewetting treatment while the other half was kept drained. Prolonged drainage led to the mobilization of nitrate (NO3-, > 1 mmol L-1) in all cores, regardless of soil iron content. In the rewetted iron-rich cores, a sharp increase in pore water iron (Fe) concentrations correlated with concentrations of inorganic carbon (TIC, > 13 mmol L-1) and dissolved organic carbon (DOC, > 16 mmol L-1). Additionally, ammonium (NH4+) accumulated up to phytotoxic concentrations of 1 mmol L-1 in the pore water of the rewetted iron-rich cores. Disproportionate mobilization of Fe, TIC, DOC and NH4+ was absent in the rewetted iron-poor cores, indicating a strong interaction between waterlogging and iron-mediated breakdown of organic matter. Concentrations of dissolved phosphorus (P) rose slightly in all cores upon rewetting, but remained low throughout the experiment. Our results suggest that large pools of iron in the top soil of drained fens can hamper the restoration of the fen's sink-service for ammonium and carbon upon rewetting. We argue that negative effects of iron should be most apparent in fens with fluctuating water levels, as temporary oxygenation allows frequent regeneration of Fe3+. We conclude that rewetting of iron-poor fens may be more feasible for restoration.

Fig 1. Experimental set-up.

40 intact vertical soil cores were collected in 4 drained fens using sharpened PVC tubes (45 x 12.5 cm), and were then placed in individual containers filled with stagnant de-oxygenized artificial groundwater. Tubes were perforated at the bottom to allow water inflow. Rhizons were placed at 5, 15 and 25 cm below the soil surface, and connected to vacuum-syringes. Half of the cores were rewetted to peat surface level, while the other half was kept moderately drained (water level 27 cm below peat surface level).

Fig 2. Changes in pore water pH and electrical conductivity (EC) in 40 soil cores.

The cores differ in experimental water level treatment (rewetted or desiccated) and initial soil iron content (high or low). Soil cores were classified into 4 groups: rewetted iron-poor fens (n = 10 cores from 2 sites), desiccated iron-poor fens (n = 10 cores from 2 sites), rewetted iron-rich fens (n = 10 cores from 2 sites), and desiccated iron-rich fens (n = 10 cores from 2 sites). Dots represent group means ± SE.

Fig 3. Iron, nutrient and carbon mobilization.

Mobilization of (a) dissolved iron, (b) total inorganic carbon, (c) dissolved organic carbon, (d) ammonium, (e) methane and (f) total dissolved phosphorus over time (t = 0, 30 and 127 days) in the pore water of 40 soil cores that differ in experimental water level treatment (rewetted or desiccated) and initial soil iron content (high or low). Soil cores were classified into 4 groups: rewetted iron-poor fens (n = 10 cores from 2 sites), drained iron-poor fens (n = 10 cores from 2 sites), rewetted iron-rich fens (n = 10 cores from 2 sites), and drained iron-rich fens (n = 10 cores from 2 sites). Dots represent group means ± SE.

Fig 4. Nitrate mobilization.

Mobilization of nitrate (NO₃^-) over time (t = 0, 30 and 127 days) in the pore water of 40 soil cores that differ in experimental water level treatment (rewetted or drained) and initial soil iron content (high or low). Soil cores were classified into 4 groups: rewetted iron-poor fens (n = 10 cores from 2 sites), drained iron-poor fens (n = 10 cores from 2 sites), rewetted iron-rich fens (n = 10 cores from 2 sites), and drained iron-rich fens (n = 10 cores from 2 sites). Dots represent group means ± SE.

Fig 5. Relationship between iron, TIC, DOC and NH4+.

Correlations between the change in pore water Fe concentrations (ΔFe) and the change in concentrations of (a) total inorganic carbon (ΔTIC), (b) dissolved organic carbon (ΔDOC) and (c) ammonium (ΔNH₄⁺) (in μmol L^-1) in 20 rewetted and 20 drained soil cores over 127 days (n = 4 sites).