BG2-3: Climate warming will cause shoulder seasons to exert primary controls over C and N mineralization in boreal forests.

We will measure net nitrogen mineralization using the buried bag method. Soils will be sampled to a depth of 20 cm (excluding live moss and recent litterfall) with a 6.5 cm diameter stainless steel corer at six random intervals along randomly oriented 50 m transects in each of the five successional stages replicated three times to avoid pseudoreplication (Hurlbert 1984). To estimate the relative importance of early vs. late winter processes, we plan to start winter incubations during freeze-up in October and retrieve incubated cores in early January, early March, and the remaining samples one week after breakup in late May. (Logistic constraints pertaining to travel by river boat to our field sites may prevent us from retrieving the soil cores immediately after snowmelt). This design allows us to examine nutrient cycling processes during times of potential increased activity due to climate change that have previously received very little attention.

Potential protease activity will be measured on soil from all successional stages on the floodplain. Soil samples will be solubilized in a 50 mmol/l sodium citrate buffer (pH adjusted to match the natural pH of the soil) at a ratio of 1g soil to 5 g buffer. To this we will add 200 mL per g soil toluene to prevent microbial activity. Soil solutions will be incubated on a shaker table at 20^o C. At intervals of 1.5 hours, from T₀ to six hours (T₄), TCA (trichloroacetic acid, 0.11 mol/l, sodium acetate, 0.22 mol/l, acetic acid, 0.33 mol/l) are added to subsets of the samples to stop the reaction (Watanabe and Hayano, 1995). Samples are then vacuum filtered using Type A/E glass fiber filters. Samples will be refrigerated and analyzed the next day for total amino acid concentrations by fluorescence spectroscopy on a FL-600 microplate mutidetection reader (Bio-Tek, Wisconsin, USA), using the OPA reagent. Amino acid concentrations are determined by comparison to a leucine standard analyzed at the same time. After subtraction of T₀ values protease rates are estimated as leucine equivalent amino acid produced per hour per gram soil.

Soil temperature will be monitored at two types of sites. The sites of the first type are the intensive-measurements permafrost observatories. At these sites, vertical temperature profiles in the air, ground surface, active layer and near-surface permafrost (at depths between zero and 1 meter with 0.1 m intervals) will be recorded at hourly intervals, using an automatic data logger (Campbell CR-10X). Thermoresistor sensors (Campbell L107 and thermistor rods of MRC) will be used for temperature measurements. Soil moisture will be also continuously monitored at hourly interval at several depths within the active layer and upper permafrost. TDR-type sensors (VITEL) will be used in these measurements. At some locations, deeper permafrost temperatures in the drill holes will be measured seasonally. There are 25 sites of the second type (extensive permafrost temperature measurements network). These sites distributed within 100x100 km area around Fairbanks, which includes Bonanza Creek and CPCRW research areas. Each of the extensive temperature measurement sites include measurements (using Onset StowAway single-sensor automatic data mini-loggers) of the ground surface temperatures (0.02 m depth) and soil temperatures within the active layer at 0.4 to 0.5 m. Air temperature also will be recorded at the selected extensive sites. The locations of these extensive sites were chosen to reflect surface and microclimate conditions at different topographic and vegetation settings to determine the effects of snow cover, vegetation, and surface morphology on the ground surface, active layer, and permafrost temperatures.

Coupled to our soil temperature measurements, from which we can estimate the length of the active season for biogeochemical processes, we will measure vertical heat flow through the soil profile in early successional shrub communities using self-calibrating soil heat flux plates (Campbell Scientific, Inc). The heat flux plates will be installed at 10 cm depth and data will be collected on an hourly basis.

Laboratory methods

In the laboratory soil sample will be homogenized by depth, and coarse debris and stones were removed. Sub-samples will be saved for moisture determinations, after which 10 g fresh weight soil were extracted with 0.5 M K₂SO₄ over-night, filtered under vacuum through 50 mm glass fiber filters (Gelman Laboratory A/E Glass Fiber Filters), and frozen prior to analysis. Filtered extracts were analyzed for NH₄⁺ using a phenol hypochlorite assay and NO₃^- using the Griess-Illosway procedure in combination with a Cadmium-reduction column on a modified Technicon autoanalyzer II (Whitledge et al. 1981). Mineralization rates will be calculated as the change in inorganic N (NH₄⁺ + NO₃^-) over the incubation period.