Measuring hydraulic properties of peat through inversion of pumping test data
The hydraulic conductivity of peat affects flow patterns and rates that influence biogenic gas dynamics, vegetation patterns, and biomass accumulation rates in peatland ecosystems. Biogenic gasses, such as methane, trapped within the peat complicate this relationship by occluding pore space and changing the hydraulic conductivity of the peat. We have completed pumping tests at two locations, a forested bog and open lawn, within the Glacial Lake Agassiz Peatlands to characterize the hydraulic conductivity in this peatland. Traditional analysis of pumping tests provide bulk estimates of hydraulic conductivity. In contrast, the spatial distribution of hydraulic conductivity in our study area have been estimated using regularized inversion to calibrate a computer simulation with pumping test data. Computer models of pumping tests were developed using FiPy, a finite-volume modeling library. Parameter estimation software (PEST) was used to repeatedly run simulations while adjusting hydraulic properties assigned to cells within the computer model. Results from the regularized inversion of the pumping test data indicate that the hydraulic conductivity of the deeper peat at the bog and lawn sites typically ranged from 10-6 to 10-5 and 10-3 to 10-4 m/sec, respectively. Anomalous hydraulic conductivity bands were present in the deep peat and hydraulic conductivity did not systematically increase with depth. Lenses of low hydraulic conductivity peat are present in out calibrated models and may represent areas of trapped biogenic gas, but these lenses lie in regions of our computer model that are poorly constrained by the regularized inversion.
Ecohydrology of Minerotrophic Peatland of the La Grande River Watershed, Northern Quebec: Water Cycle, CO2 and CH4 Monitoring
Despite increasing precipitation over the last decades, inflows to the La Grande hydroelectric reservoir system have remained low. It is believed that further understanding of minerotrophic peatland hydrology may help explain this paradox. Recent research has shown that the water surface area (i.e., ponds, waterholes) characterizing these peatland has significantly increased over the last century in response to a changing climate. This process has been referred to as aqualysis, however, the hydrological stocks and fluxes of these northern boreal peatlands have not been thoroughly quantified so far. Furthermore, as the aqualysis takes place, there also exists a need to monitor the likely increases in CO2 and CH4 fluxes under changing climate conditions. All of the above observations have led to a multidisciplinary pilot, watershed-scale, ecohydrology study of minerotrophic peatland of the La Grande River watershed. This NSERC-CRD (Hydro- Quebec/Ouranos partnership) research project focuses on monitoring the water, CO2 and CH4 fluxes. Remote sensing (soil humidity status, land cover), hydrometeorological instrumentations (water fluxes), ground penetrating radar (water stocks), gas chambers (CO2 and CH4 fluxes) are the basis for this monitoring study. Also, on site vegetation and land cover characterizations, using remote sensing, will be used to develop ecological indices to help in understanding the manifestation of the aqualysis process over the last five decades. Finally, development of a process-based fen hydrology model and establishment of empirical relationships between ecohydrological parameters and greenhouse gas fluxes represent the key goals of this project
Expanding Peatlands in Alaska Caused by Accelerated Glacier Melting Under a Warming Climate
Most mountain glaciers worldwide have been retreating over the last century due to global warming. This is particularly true around the Gulf of Alaska, where glacier recession has further accelerated since 1988. It is well known that glacier meltwater plays a critical role in the global sea level rise, but its effects on structure and functioning of peatland ecosystems remain poorly understood. We have observed in the field that many peatlands in the Susitna Basin, south-central Alaska, are expanding. As high moisture conditions are needed to promote peatland development and expansion, a regional change toward wetter conditions is likely responsible for the ongoing paludification of these peatlands. However, instrumental climatic data from this region show no increase in precipitation but an increase in temperature (and presumably evaporation) over the last decades. We hypothesize that climatically-induced glacier melting is modifying the local/regional climate, especially air humidity during the growing season, promoting the expansion of peatlands. To document recent peatland vertical growth and lateral expansion, we collected two long peat cores and twelve 30-cm-long monoliths in 2008 along a 110-m transect from an expanding peatland margin toward the peatland center. Ecohydrologic changes were reconstructed from testate amoebae and plant macrofossils assemblages. Preliminary results from both long cores revealed a change in the vegetation assemblages from a mesotrophic fen dominated by sedges and brown mosses to a Sphagnum-dominated peat bog at 11 cm, suggesting a very recent modification of the local hydrologic regime. A simultaneous increase in moisture was reconstructed from testate amoebae records. These unusual shifts in peatland development and hydrology (e.g., wet conditions triggering the fen-bog transition) imply a recent increase of atmospheric water to these peatlands. Our ongoing lead-210 dating and additional proxy analysis will help us resolve the timing and nature of recent peatland changes. These data, together with glacier history and climate records, will allow us to further test our hypothesis that the increase in glacier meltwater is causing peatland expansion By acting as water sinks, peatlands located in glacierized watersheds may mediate the contribution of meltwater to present and future sea-level rise. Increases in peat accumulation rates due to favorable hydroclimatic conditions are also expected to promote carbon sequestration by these ecosystems. In contrast to the expected desiccation of peatlands under a warmer climate, enhanced growth due to glaciers-climate feedbacks in high-latitude regions may thus promote peatland expansion, even just temporally.
Rewetting an Abandoned Block-cut Peatland Using Spring Melt: How Much Water is Available?
Cacouna bog is an abandoned block-cut peatland which has recently (autumn 2006) received management intervention (ditches blocked to impede runoff) with the intent to accelerate spontaneous restoration. Because the peatland is a bog (ombrogenous) the spring snowmelt is a critical source of water that needs to be retained to offset the summer water deficit. A snow survey and snowmelt runoff study was conducted (spring 2005) to determine how much water was lost through the drainage ditches during this period. This will allow an estimate of how much water could be retained with the ditches blocked. A snow survey (depth and density) was conducted immediately before snow melt began on 24 March. At four contrasting sites in the peatland the daily snow ablation rate and snow surface density were recorded in addition to water table and discharge within and from the peatland (Upper Lagg, Down Lagg, Flume and Weir). Water samples were taken at these four sites from wells and from the snow pack (snow surface density), as well as all four stream discharge points, to determine electrical conductivity. Average snow depths ranged from 67 to 81 cm with densities of 0.30-0.34 g/cm3, representing a system-wide average water equivalence of 247 mm. Snow ablation averaged 17 to 27 mm day-1 or 5 to 10 mm day-1 water equivalent. Surface (0-3 cm) snow density remained between 0.3 and 0.4 g/cm3 for most of the melt period (with the exception of a snow event). Water tables fluctuated similarly across the peatland and rose between 8 cm and 20 cm over the study period. Discharge varied diurnally and peaked at all four locations on 14 April 2006 and at the Flume was 0.035m3/sec or 5.5 mm day-1. This corresponded roughly with when the snow at the ablation wires had melted entirely from 1 site, nearly gone (5/12 measurement locations along the wire had snow remaining) from a second site and only left in ditch (well shaded) at a third site (3/12). Electrical conductivity of runoff water varied diurnally (typically higher in the morning (lower diurnal snowmelt input) and lower in the afternoon (higher diurnal melt water inflow)) and generally decreased throughout the study period (from ∼80 μS/cm to ∼30 μS/cm). Total volume of water (from snow) for the entire site at the time of the snow survey was ∼136,000 m3 (247 mm) and discharge loses totalled ∼60,000 m3 (109 mm) for the snow melt period. The large difference is likely due to water still leaving the system once the snow had finished melting, as well as water loss due to sublimation of snow (and evapotranspiration of water in snow free areas).
Understanding Peat Bubbles: Biogeochemical-Hydrological Linkages
Decomposition of organic matter in peatland ecosystems produces gaseous end-products that can accumulate at depth and result in the build up of free-phase gas below the water table. This free-phase gas, or bubbles, reduces hydraulic conductivity, alters hydrologic and chemical gradients, and affects productivity surface vegetation through its role in peat buoyancy. In terms of greenhouse gas dynamics, these bubbles are likely the dominant subsurface stock of methane (CH4) and release of this CH4 to the atmosphere via ebullition may account for a significant portion of total efflux. Despite the importance of entrapped bubbles for peatland ecohydrological function there is still little known about how the quantity of bubbles varies between peatland types and at smaller scales within a peatland. Profiles of bubbles collected from several locations within four peatlands reveal that bubble volume varies significant among peatlands, between microforms and with depth. Previous studies also suggest that ebullition is spatially and temporally variable. This spatial variability may have important impacts on system ecohydrology and should be incorporated in models of peatland hydrology and development. This requires the difficult task of mapping bubble volume in three dimensions and over large areas. The potential for geophysical methods and the use of surface features to address this task will be discussed.
No limits to peat bog growth? Transport and thermodynamic constraints on anaerobic organic matter decomposition
In diffusion dominated systems, for which many thick peat deposits provide a model, slowness of transport and lack of free energy may pose a limit to methanogenic decomposition of organic matter and ultimately to closing the carbon cycle. To test this hypothesis we (I) conducted controlled column experiments with homogenized peat over an 18 month period, (II) investigated transport, in situ respiration pathways, rates and thermodynamic conditions in a nothern peatland, and (III) modelled depth profiles of CO2 and CH4 in the deposits. Vertical transport in the peatland was dominated by diffusion leading to the buildup of DIC and CH4 with depth (5500 µmol L 1 DIC, 500 µmol L 1 CH4). Highest DIC and CH4 production rates occurred close to the water table (decomposition constant kd ~10-3 to 10 4 a-1) and decreased to about kd = 10-7 a-1. The accumulation of metabolic end-products diminished in situ energy yields of acetoclastic methanogenesis to the threshold for microbially mediated processes (-20 to -25 kJ mol-1 CH4). The methanogenic precursor acetate also accumulated (150 µmol L 1). In line with these findings, CH4 was formed by hydrogenotrophic methanogenesis at Gibbs free energies of 35 to 40 kJ mol-1 CH4. This was indicated by an isotopic fractionation αCO2-CH4 of 1.069 to 1.079. Fermentative degradation of acetate, propionate and butyrate attained Gibbs free energies close to 0 kJ mol-1 substrate. In peat columns with homogenenous peat-sand mixtures of 50%, 15% and 5% dry weight, steady state CO2 production also decreased from about 10 to 0 nmol cm-3 d-1 and of CH4 from 1 to 0 nmol cm-3 d-1 with depth. Very similar depth profiles of concentrations and volumetric rates developed near endproduct thresholds of 600µmol CH4 and 10 mmol L-1 CO2, despite the differences in organic matter content. The modeling exercise showed that a consistent development of CH4 concentration profiles over time in the columns could only be accomplished with rates of acetoclastic methanogenesis decreasing to 0 near a critical Gibbs free energy of about -27 KJ mol-1. The results thus suggest that, even in absence of inorganic electron acceptors, respiration rates in peat bogs are likely higher near the redox interface to the atmosphere due to lower respiration endproduct concentrations. Similar effects ensue when rates of transport are elevated or pools of CO2 and CH4 are eliminated. With decomposition being constrained, peat bog growth may occur longer than previously thought.
Solute Transport in Unsaturated Sphagnum Mosses
Natural Sphagnum cushions develop an upwardly increasing concentration of dissolved solutes during periods of sustained upward capillary flow of solutes, and become enriched by evaporative loss of water. The transport process is poorly documented as a consequence of poor parameterization of unsaturated flow parameters, and the lack of transport parameters such as dispersivity and solute retardation coefficients for flow in unsaturated mosses. Sphagnum mosses contain hyaline cells and dead-end pores that can store but not transmit water and solute. Since these spaces do not drain at moderate (negative) pressures (ψ), the ratio of fluid actively flowing in films in the unsaturated moss to that which is stored decreases as the moss drains. Solutes can pass by diffusion from the film of flowing water into these closed spaces resulting in increased dispersion of the flowing solute, and retardation of even conservative solutes like chloride. These processes were demonstrated in unsaturated Sphagnum mosses using a step input solute (NaCl) source from a constant head device for undecomposed near-surface moss (~5 cm depth), and slightly more decomposed deeper moss (~25 cm depth). Smaller water retention in the undecomposed upper moss sample resulted in lower unsaturated hydraulic conductivity thus lower flow rates. When the sample was initially drained (ψ = ~ 4 cm of water) it was determined that the solute breakthrough expressed as relative concentration (C/C0 = 0.5) occurred at a cumulative discharge of 91.5 ml and at 5.8 minutes in the upper moss, compared to 233.2 ml after 2.8 minutes in the lower (more decomposed) sample. In a drier state (ψ = ~ 16 cm of water), C/C0 = 0.5 was reached after 67.9 ml of discharge at 37.9 minutes in the upper moss compared to 109.2 ml and at 22.4 minutes in the lower sample. Thus less solute flow is required for breakthrough in less decomposed mosses, and in mosses that are relatively dry. Dispersivity was determined on the basis of best-fit of the Ogata-Banks model of advection and dispersion. The results demonstrate the importance of, and necessity to include, the role of hyaline cells in solute transport in unsaturated mosses.