Water flow in Sphagnum hummocks: Mesocosm measurements and modelling

• Published in: Journal of hydrology (Amsterdam) 2010-02, Vol.381 (3), p.333-340
• Main Authors: Price, Jonathan S., Whittington, Peter N.
• Format: Article
• Language: English
• Subjects: Earth sciences; Earth, ocean, space; Evaporation; evaporation rate; Exact sciences and technology; Fluid flow; Fluxes; Hydraulics; Hydrology. Hydrogeology; Hydrus 1D; mathematical models; Moss; Mosses; peatlands; RETC; saturated hydraulic conductivity; simulation models; Sphagnum; Staged; Unsaturated hydraulic conductivity; water content; water flow; Water retention; water table; Water tables
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2009.12.006

The internal water fluxes within Sphagnum mosses critically affect the rate of evaporation and the wetness of the living upper few centimetres of moss (capitula) and the physiological processes (e.g. photosynthesis) that support them. To quantify water fluxes and stores in Sphagnum rubellum hummocks we used a 30 cm high column (mesocosm) of undisturbed hummock moss including the capitula, and applied a number of experiments to investigate (1) staged lowering (and raising) of the water table ( wt) with a manometer tube; (2) pumped seepage of about 0.7 cm d −1 to produce a wt drop of 1.5 cm day −1; and (3) evaporation averaging 3.2 mm d −1. Water content ( θ) at saturation ( θ s ) was ∼0.9 cm 3 cm −3 for all depths. Residual water content ( θ r ) was 0.2 cm 3 cm −3 at 5 cm depth, increasing to 0.47 cm 3 cm −3 at 25 cm depth. Hydraulic conductivity ( K) of the same top 5 cm layer ranged from 1.8 × 10 −3 m s −1 at θ s to 4 × 10 −8 m s −1 at θ r . By comparison K at 25 cm depth had a much more limited range from 2.3 × 10 −4 m s −1 at θ s to 1.1 × 10 −5 m s −1 at θ r . Staged wt lowering from −10 cm to −30 cm (no evaporation allowed) resulted in an abrupt change in θ that reached a stable value generally within an hour, indicating the responsiveness of moss to drainage. Staged increases also resulted in an abrupt rise in θ, but in some cases several days were required for θ to equilibrate. Pumped seepage resulted in a sequential decline of θ, requiring about 10 days for each layer to reach θ r after the water table dropped below the sensor at the respective depths. Evaporation resulted in a similar pattern of decline but took almost three times as long. The computer simulation Hydrus 1D was used to model the fluxes and provided a good fit for the staged lowering and pumped seepage experiments, but overestimated the water loss by evaporation. We believe the reason for this is that over the longer evaporation experiment, the monolith underwent consolidation and shrinkage which reduced saturated hydraulic conductivity, thus reducing the rate of upward water flux – not accounted for in the simulation. Declining θ s in lower layers (i.e., before pore drainage) was evidence of consolidation. The study confirms that the hydraulic structure results in a rapid transition to a low but stable water content in upper mosses when the water table falls, a low unsaturated hydraulic conductivity in such circumstances that constrains upward water flux caused by evaporation when θ