NO3 hotspots in pristine watershed of the Boreal Plain: interactions of local landforms and regional hydrogeology.
Surface and ground water nitrate (NO3-N) concentrations of 3 mg/L, and often greater than 10 mg/L, are frequently observed in pristine watersheds of the Western Boreal Plain, Alberta, Canada. Such concentrations are greater then expected in non-disturbed locations, and in some cases are well above the WHO recommended levels. Nitrous oxide (N2O) fluxes from soil surfaces in these nitrate hot spots range from 10- 100, and on occasion greater then 500 ug/m2/h and may contribute significantly to global green house gas emissions. Nitrate hot spots are frequently observed in topographic low riparian areas with large water table fluctuations within humic organics sediments. However, their occurrence is not consistent in riparian zones across the landscape making it difficult to generalize N dynamics based solely on local influences of soil temperature, moisture and carbon dynamics. The Boreal Plain is characterized by sub-humid climate and deep heterogeneous glacial deposits resulting in complex surface - groundwater interactions with diverse geochemistry due to variable geologic strata that may also influence initial source and dynamics of N in individual riparian locations. We present representative distributions of surface and groundwater NO3-N concentration in the Boreal Plain, Alberta and put forward a series of hypothesis to assess the relative role and interaction of local soil and vegetation characteristics with hydrogeology on regional scale nutrient source, flow path and soil moisture regimes. Understanding the regional and local controls of nitrate hotspots in pristine boreal forest are essential to provide background references and allow for accurate assessment of the impacts of climate change and intensive land use currently affecting the Boreal Plain.
Quantifying the Role of Water Table Dynamics on Net Ecosystem Exchange of CO2 in a Northern Temperate Shrub Wetland
Wetlands represent up to a third of the global soil carbon, and so they are a large component of an uncertain terrestrial carbon flux. In the northern temperate forests around the upper Great Lakes of North America forest and shrub stature wetlands cover about one-third of the total land area. In northern Wisconsin lateral subsurface water redistribution associated with hummocky glaciated terrain drives lowland water table heights. A multi-year trend of declining water table height in these areas has been observed throughout this region since year 2000. We examined the effect of this declining water table on net ecosystem exchange (NEE) of carbon measured at the Lost Creek AmeriFlux tower site (46 deg. 49 min. N, 89 deg. 58.7 min. W), which lies within a shrub stature alder and willow wetland. On an inter-annual basis from 2001-2007 there was no significant correlation between annual total NEE and annual mean water table height. There were offsetting increases in both respiration and gross ecosystem production (GEP) as the water table fell through a level of about 20-30 cm below the surface. During an exceptionally dry growing season in 2007 GEP was anomalously low switching the site from a carbon sink to a source for the year. To quantify the role of water table dynamics at shorter timescales and to examine the relative responses of GEP and respiration in more detail we used a coupled water and carbon transport model along with Bayesian analysis. Short-term water table dynamics significantly improved the prediction of NEE, reflecting the importance of seasonal distribution of precipitation on the coupled water and carbon exchanges.
The Catena Concept Revisited: Spatial Optimization of Ecohydrologic Form and Function
Over the past two decades, empirical evidence and theory have been developed that suggest that plot scale ecosystem properties such as canopy density and root depth evolve towards a state that maximizes resource use and net primary productivity. We generalize this concept from the plot scale to the catchment by examining canopy density as a function of available energy, water and nutrients connected along hydrologic flowpaths. We use a combination of field measurement, signal processing and distributed simulation to identify emergent optimal ecohydrologic patterns in a set of Long Term Ecological Research sites reflecting the interactions between catchment geomorphic, soil, climate and ecosystem processes. Results to date reveal interesting adjustments of above and below ground canopy structure and physiologic function with water and nutrient availability that indicate the tendency to develop landscape scale optimization, beyond that achieved at individual plots, of net primary productivity and water use efficiency.
A Spatially Explicit Modeling Approach to Capture the Hydrological Effects on Biogeochemical Processes in a Boreal Watershed
Current estimates of terrestrial carbon (C) fluxes overlook hydrological controls. A modeling study was
conducted to explore the hydrological, ecophysiological and biogeochemical interactions in a humid boreal
ecosystem. Several hydro-ecological processes were simulated and validated using field measurements for
two years. After gaining confidence in the model's ability and having understood that topographically driven
sub-surface baseflow is the main process determining the soil moisture regime in humid boreal ecosystem,
its influence on ecophysiological and biogeochemical processes were investigated. Three modeling
scenarios were designed that represent strategies that are commonly used in ecological models to represent
hydrological controls. These scenarios were: 1) Explicit, where realistic lateral water routing was
considered; 2) Implicit, where calculations were based on a bucket-modeling approach; and 3) NoFlow,
where the lateral sub-surface flow was turned off in the model. In general, the Implicit scenario
overestimated GPP, ET and NEP, as opposed to the Explicit scenario. The NoFlow scenario
underestimated GPP and ET but overestimated NEP. The key processes controlling the differences were due
to the combined effects of variations in plant physiology, photosynthesis, heterotrophic respiration, autotrophic
respiration and nitrogen mineralization; all of which occurred simultaneously in different directions, at different
rates, affecting the spatio-temporal distribution of terrestrial C-sources or sinks (NEP). The scientific
implication of this work is that regional or global scale terrestrial C estimates could have significant errors if
proper hydrological constraints are not considered for modeling ecological and biogeochemical processes
due to large topographic variations of the Earth's surface and also because of the non-linear interactions
between these processes.
Coupling of Water and Carbon Cycles in Boreal Ecosystems at Watershed and National Scales
The boreal landscapes is relatively flat giving the impression of spatial homogeneity. However, glacial activities have left distinct fingerprints on the vegetation distribution on moderately rolling terrains over the boreal landscape. Upland or lowland forests types or wetlands having various degrees of hydrological connectivitiy to the surrounding terrain are typical of the boreal landscape. The nature of the terrain creates unique hydrological conditions affecting the local-scale ecophysiological and biogeochemical processes. As part of the Canadian Carbon Program, we investigated the importance of lateral water redistribution through surface and subsurface flows in the spatial distribution of the vertical fluxes of water and carbon. A spatially explicit hydroecological model (BEPS-TerrainLab) has been developed and tested in forested and wetland watersheds . Remotely sensed vegetation parameters along with other spatial datasets are used to run this model, and tower flux data are used for partial validation. It is demonstrated in both forest and wetland watersheds that ignoring the lateral water redistribution over the landscape, commonly done in 1-dimensional bucket models, can cause considerable biases in the vertical carbon and water flux estimation, in addition to the distortion of the spatial patterns of these fluxes. The biases in the carbon flux are considerably larger than those in the water flux. The significance of these findings in national carbon budget estimation is demonstrated by separate modeling of 2015 watersheds over the Canadian landmass.
Modeling Mass Balances Of Dissolved Organic Carbon For Lakes Within Regional Watersheds
The fate of carbon flows is important for understanding and predicting the function of both land and lake systems within regional watersheds. In this study, we create a meta-model of empirical relationships for dissolved organic carbon (DOC) fluxes to predict DOC concentrations in lakes of regional watersheds within eastern Canada. We apply an empirical model to predict DOC fluxes from land to streams and lakes. This empirical model is based on digital terrain analysis techniques for deriving the proportion of wetlands within catchments combined with indexes of climatic conditions to predict annual DOC export (g C/m2/yr). This approach has been used previously to explain almost 90% of DOC exports ranging from 0.90 to 13.74 g C/m2/yr from 1st order catchment on glaciated landscapes across eastern Canada; it is used herein to predict DOC exports in both 1st and higher order catchments. We apply a steady-state model to predict DOC retention in lakes using measured DOC inputs and outputs. The meta-model, combines the two empirical models and uses a spatially explicit tracking system to monitor DOC movement from its source areas within the land, to its partial retention within lakes, and its export to rivers. The meta-model predictions of concentrations of DOC in lakes were reasonable (according to model fit and residuals). Further research is needed to improve estimation of model coefficients and to incorporate processes not currently represented.
Eco-hydrologic responses of mountain forested watersheds to climate warming: the role of variability and uncertainty in subsurface drainage characteristics.
In the mountainous Western US, spatial variation in eco-hydrologic processes is a complex function of geology, soil, topography, climate and vegetation patterns. Understanding how these different controls vary and interact within a regional landscape across a range of scales is a key challenge in understanding impacts of climate change. Much of the current research focuses on spatial-temporal patterns of snow accumulation and melt as important drivers of summer streamflow, and points to dramatic changes in water resources with reduced snowpacks. Recent work has also shown evidence of increases in vegetation mortality associated with summer drought. We show that modelling these responses requires estimation not only of snowmelt response to warming, but also of spatial variation in subsurface geologic controls on drainage. Using a coupled process-based model of ecosystem hydrologic and carbon cycling processes, we demonstrate that soil moisture drainage characteristics exert a significant control on how these systems respond to earlier snowmet. We utilize uncertainty analysis of soil hydrologic parameters to provide insight into the extent to which improvements in our ability to estimate soil parameter will impact predictions of carbon and hydrologic fluxes under a warming climate. These modeling studies provide an expanded perspective on landscape level sensitivities to climate warming, and can provide guidance for a more sophisticated and spatially distributed approach to monitoring climate effects.