Carbon Movement and Sequestration in SMC Type I Installations

s and Publications 8 Student Bio 11 2nd Qtr 2004 Carbon Movement and Sequestration in SMC Type I Installations A.B. Adams, R.B. Harrison Ecosystem Science Division, University of Washington Seattle, Washington Introduction An information source for Douglas-fir growth response has been the Regional Forest Nutrition Research Project (RFNRP) initiated in 1969. Information for soils in the installations is more limited. Earlier soil carbon (C) research in these Type 1 installations (Phase 13) was done by Peterson, Ryan and Gessel (1984). In the Pacific Northwest, soil carbon (C) is significant because boreal forest have the potential to mitigate the increase in atmospheric concentration of CO2 (Lal, 2003) while remaining highly productive forests. In this study, we considered carbon sequestration ca. 30 years after the initiation of repeated urea applications on second growth Douglas-fir stands in RFNRP installations in the foothills of the Central Puget Trough. Urea and soil type were treatment variables. Initially, our sampling was hindered by a lack of precipitation during the winter of 2000-01 (Fig. 1); however, the summer of 2001 and winter of 2001-02 had normal to above average precipitation. This provided us with a simple means of comparing the relation of water and DOC flux in our study region. Figure 1: Precipitation patterns for Site 1 (Port Gamble) showing the unusually dry winter rain pattern for winter 2000-01. In addition, the subsequent summer was unusually wet. In the winter of 2002-03 rainfall at our 4 sites was above average. Sample Collection and Methods Methods are described in detail in Adams, et al. (manuscript in review). Site Description and Treatments: Soils at two sites (Tables 1 and 2) were derived from glacial material. Site 1 was coarse outwash, and Site 2 was deep sandy outwash. Of the two volcanic soil types, Site 3 was coarse loamy ash above glacial outwash material, and Site 4 was a deep (>1 m) ashy loam, tephra mix. Control and urea (fertilized with 448 kg ha-1 and then 224 kg ha-1 every four years to a total application of 1142 kg ha-1 urea) plots were chosen from each installation. 2 A.B. Adams with a 20 centimeter Lysimeter at the UW Soilslab Field Soil Collections and Carbon of the <2mm Soil Fraction: Soil collections were made according to horizons located in the soil profile. All soil <25 mm and all rocks >25 mm were weighed as two separate components in the field. Sub-samples of the <25 mm component were taken for each layer identified in the field and used to determine field moisture and the particle size distribution. Bulk density (Db), was determined by displacement methods and hammer corer and measurements based on the location and nature of the soil (Blake and Hartge, 1986; Harrison, et al., 2003). The <2 mm material was finely ground with a mortar and pestle and analyzed for total C using a CHN analyzer. Lysimeter Installations, Data Collection and C Flux: A zero tension pan lysimeter was placed under the forest floor. In addition, three negative tension lysimeters at depths of 15, 50 and 100 cm were installed (Titus and Mahendrappa, 1996) at each plot. Soil solutions were collected every 4-6 weeks. Hydrologic flux was determined using the Thornwaite method (Dunne and Leopold, 1978) to estimate potential evapo-transpiration rates for each site, and then by weighing potential evapo-transpiration against precipitation to determine net soil-water flux. Coupled with C concentration data, we were then able to estimate C fluxes by depth (Table 3). Weather data for each site was extrapolated using local weather stations nearest to our sites (National Climatic Data Center 2003 and Washington Annual Precipitation 1998). For Site 1 we used the Landsburg Station, for Site 2 the Chimacum Station and for Sites 3 and 4 the Mud Mountain Dam Station. Results and Summary Characterization of Soil Physical Properties (Tables 1 and 2): The glacial soils are both coarse-textured, but Site 1 has a high proportion of material >25 mm, and only 10% of the material is <2 mm. The other glacial soil (Site 2) was predominantly (>70%) sand and had no material >25 mm. The volcanic soils were both silty loam. Site 3 is 40-50% < 2 mm and about 10% rocks >25 mm. Site 4 is 65-90% <2 mm and has no rocks >25 mm. Bulk density values ranged from as low as 0.13 for forest floor samples to 1.65 for compacted sand in the Bm horizon of Site 2. Bulk densities were low (<0.4) for O horizons, and Db increased with depth. Forest floors (O horizons) were variable between sites ranging from scarcely present at Sites 2 and 4 to conspicuously present at Sites 1 and 3. Percent C was >30% in all forest floors except for Site 4 (the site with the most mineral soil C) (Fig. 2). At Sites 2, 3 and 4, forest floors were much smaller in treated plots relative to controls. This was probably because the urea fertilization increased decomposition rates. Soil Carbon: At Site 1 the fertilized plot had 10% less C than did the control, but more C was found in treated pits at Sites 2, 3 and 4 (46%, 135% and 37% more C, respectively) (Fig. 2). The four glacial pits had significantly less carbon than the four volcanic pits (p= .008 for parent material effect in ANOVA). Most (>80%) C in coarse outwash was in the forest floor (Fig. 3). In the sandy glacial outwash soil C was distributed equally throughout the control, but yielded proportionately more in the treated Bs horizon (39% versus 24% or less in the other three horizons). Table 1 Site characteristics with installation tree sample dates. Parameters Site 1 Site 2 Site 3 Site 4 Cedar River Port Gamble Radio Hill Mud Mountain 1992 1991 200