Data Package Metadata   View Summary

Experimental Design

The study site at the Eight Mile Lake watershed is located in the northern foothills of the Alaska Range. Permafrost temperature has been monitored annually on a gentle north-facing slope at this site, in a 30 m deep borehole that was installed in 1985 before the permafrost started to thaw. While permafrost thaw can sometimes result in water ponding depending on local topography, this landscape consists largely of relatively well-drained uplands. In this watershed, we monitored three sites that represented differing amounts of change as a result of permafrost thaw based on observations of ground subsidence and depth of thaw. Our minimal thaw site had the shallowest summer thaw depth and the least ground surface subsidence, and the vegetation was typical moist acidic tundra, dominated by the tussock-forming sedge, Eriophorum vaginatum, with coexisting deciduous and evergreen shrubs, and an understory of mosses and lichens. A second site, located adjacent to the borehole, had moderate summer thaw depths and increased ground subsidence. Here the vegetation composition was similar to our minimal thaw site, but with increased biomass of all plant groups. Lastly, an extensive thaw site was located in an area where permafrost degradation had occurred at least several decades prior to the establishment of the borehole; thermokarst can already be observed in air photos of the area taken in 1954. This site had the deepest summer thaw depth and the most ground subsidence, and the plant biomass has shifted to dominance by shrubs, including blueberry (Vaccinium uliginosum) and cloudberry (Rubus chamaemorus), at the expense of sedges. River water samples were collected from Panguine Creek which drains Panguine Catchment. Panguine Catchment lies adjacent to Gradient site, EML and shares similar permafrost and geological characteristics. Panguine Creek was sampled at the junction with the highway. The river bed is approximately 3 meters wide.

Methods

FIELD: Before river ice break-up, river waters were collected by drilling a hole into the ice with an ice auger and sampling water from underneath the ice using a pre-rinsed plastic container attached to a non-metal pole. After river ice break-up, river waters were collected using a bucket and string. Temperature, conductivity and pH were measured with a combined temperature, conductivity and pH probe (Hanna Instruments). The electrode was calibrated daily using pH 4 and 7 buffers (NIST). Redox potential was measured in perched waters using a Thermo Scientific Orian Star electrode. Conductivity and pH were measured on soil porewaters in the lab using a pH and conductivity meter (Mettler Toledo Seven Compact Duo S213) with a conductivity electrode (Inlab 715-4 mm) and pH electrode (Inlab micro). LAB: River waters were filtered within 5 hours of sampling using a Mityvac hand pump. Samples for major element concentrations were collected by filtration through pre-cleaned 0.22 micrometre nitrocellulose filters (Millipore). Samples for Dissolved Organic Carbon (DOC) were collected using 0.7 micrometre pre-combusted glass fiber filters (GF/F) (Whatman). All waters were stored in acid-washed polypropylene bottles and refrigerated immediately at 8 deg C. Unfiltered river waters underwent an additional size separation to determine the proportion of elements transported as colloids (0.22 micrometre-1 kDa size fraction) and the portion of elements transported as truly dissolved species (less-than 1 kDa fraction). Size separation was performed using tubular dialysis membranes (Spectrapore 7) with a molecular weight cut-off size of 1 kDa (approximately 1.3 nm). Before use, the membranes were washed in 2 % HNO3 and MilliQ water. The clean membranes were filled with 50 ml of MilliQ water, sealed with clips (Spectrapore 7) and stored in MilliQ in 2 L containers. In the field, a membrane was transferred into an acid-washed 2 L bottle containing unfiltered river water or perched waters. The container was left to equilibrate for 72 hours. All membranes were handled using acid-washed pincettes and gloves to minimize contamination. All filtered and dialysis waters were stored in acid-washed polypropylene bottles and refrigerated immediately at approximately 8 deg C. Alkalinity was measured using an automatic titrator (Metrohm 916 Ti-Touch) by titrating over the pH range 4.2 to 4.5 with 0.0025 M HCl.