Forest fragmentation has been recognized as an important problem for forest managers interested in maintaining ecosystem health. As further fragmentation takes place, the size of plots decrease and edge to volume ratios continue to climb (Franklin and Forman, 1987; Hunter, 1997). As the total length of edges increases with continued forest harvests, the possibility of significant sifts in soil processes over the landscape also increases. There is a long-standing interest in how forest edges influence plant and animal distributions (Harris, 1988). Typically, shade-intolerant and weedy species proliferate along edges, there are increased wind-shear forces resulting in mature tree mortality, changes in bird, mammal and insect distributions and distinctive microclimatic gradients along edges (Laurence and Yensen, 1991).
Recent studies by Chen et al. (1993, 1995) have shown that microclimate effects can be detected as far as 240 m into a Douglas-fir OG forest from edges of an adjacent 15 year stand (15YS) stand. Near the edge, greater extremes in relative humidity, wind velocity, solar radiation, and soil temperature were observed when compared to values observed deep within the OG forest or in the middle of the clear-cuts (Chen et al., 1995). Shifts in vegetation have also been observed up to 120 m from an edge (Chen et al., 1992). Much less is known about how forest edges influence biological and chemical characteristics of forest soils. This lack of basic information should be a major concern as forest managers attempt to assess the effects of forest harvesting patterns on biogeochemical processes over large watersheds.
Soil samples were collected along 27 transects with one transect taken at each site. These transects were 150 m long running along the same contour from OG forests into harvested stands. The transect extended 75 m either side of the edge with soil samples taken every 5 m. Fifteen m on either side of the edge, the sampling spacing was decreased to 1 m intervals to better define patterns closest to the edge. It also allowed us to test for spatial variability at both 1 and 5 m intervals. This dataset contains all of the 1 meter data to give the finest spatial definition 15 m either side of the edge between the oldgrowth segment of the transect and the harvested stand.
Stands used in this study had been commercially harvested which means that in each case, OG stands were cut, the slash and debris burned leaving burnt stumps and some old decayed logs. They were typically replanted with Douglas-fir seedlings within 2 years of harvest. The edge was defined as the line formed by the main stems of uncut old-growth trees along the perimeter of the harvested site. In most cases, the edge was well defined.
The experimental design consisted of 3 stand ages and 3 locations with each replicated 3 times for a total of 27 transects. The three stand ages were those harvested approximately 5, 15, and 40 years prior to the study. Nine transects of each age class were sampled. Of these 9 transects, 3 transects were in low-flat sites (646 to 800 m in elevation, 3 were higher (1000 to 1262 m) south-facing and 3 in higher (800 to 1338 m) north-facing sites.
The 5 year stands (YS) generally had early successional herbs and shrubs with large gaps between young trees which were usually < 2 m height. Much of the ground had little vegetation. Stands approximately 15 years still had gaps between trees and contained only small areas that were not covered with trees or shrubs; few herbs are present. Most of the trees were Douglas-firs which were generally less than 5 m in height. The 40 year stands had enclosed Douglas-fir canopies (most greater than 15 m) with relatively little understory vegetation.
Within OG adjacent to 5 YS, understory vegetation was starting to grow in response to increased solar radiation. The extend of this zone varied with elevation and aspect. This zone of understory growth was generally wider and better established in OG adjacent to 15 YS and persisted in a reduced form in OG stands adjacent to 40 YS.
1. Sample collection and preparation
Mineral soil samples were collected with a trowel to a depth of 10 cm and transported to the laboratory in an ice chest. Soils were stored at 15°C until the initiation of the analyses which was within 16 h of their receipt. All variables except field respiration rates were measured during Aug. 1995.
2. Field studies
Field (forest floor) respiration rates were measured using a non-dispersive, infrared CO2 analyzer (Li-cor, LI-6200). Measurements were made over a period of 1 min after the chamber gas had reached ambient CO2 concentration. The instrument was calibrated on-site against a known standard at each of the locations. A Q10 adjustment was made for ambient soil temperature. Soil temperatures were measured by electronic thermometers that had been calibrated at 0°C with ice water. The probes were inserted 10 cm into the mineral soil.
The occurrence of ectomycorrhizal mats in cores was determined visually in the field by inspecting the relative abundance of mats in 4.5 x 10 cm cores. Two distinctive mat types were scored separately; as mats similar to those of the genus Hysterangium and mats similar to those of the genus Gautieria. This approach has been used successfully in the past to document ectomycorrhizal mat distribution patterns in coniferous forests of the Pacific Northwest (Griffiths et al., 1996)
In preparation for laboratory analysis, all soils were sieved through a 2-mm sieve. Soil pH was measured in 1:10 (soil: distilled water) slurries of oven-dried (100°C) soil. These slurries were shaken for 1 h prior to reading pH values with Sigma model E4753 electrode. Soil organic matter (SOM) was measured by loss-on-ignition at 550°C for 6 h following oven drying at 100°C.
Denitrification potential (DENIT) was measured using a method similar to that used by Groffman and Tiedje (1989). Each reaction vessel (25 mL Erlenmeyer flask) contained 5 g of less than 2mm, field-moist soil. The flask was sealed with a rubber serum bottle stopper and purged with Ar to displace O2 in the headspace. After purging with argon, 2 mL of a 1 mM solution of glucose and NO3- was added each flask which was subsequently incubated at 25°C for one hr. This preincubation period was used because time series experiments on representative soils showed a lag in N2O production during this period. The same experiments have shown linear N2O production rates during the following 2-4 hr (data not shown). After the preincubation period, 0.5 mL of headspace gas was removed from the reaction vessel and injected into a gas chromatograph fitted with an electron capture detector (Hewlett Packard model 5890 GC fitted with Hewlett Packard model 3396 integrator). The integrator was calibrated using known gas standards and the external calibration method.
Another headspace N2O analysis was made after an additional two hr incubation at 25°C. The net N2O released over this 2 h period was used to estimate N2O production rates. Acetylene was not routinely added to the headspace to prevent the conversion of N2O to N2 because randomly selected samples (10% of the total) were also assayed with a 10% acetylene atmosphere. There were no significant differences between N2O production rates with and without acetylene.
Long-term respiration measurements (LTR) were made on field moist sieved soils (4 g dry wt.) which were brought to 75% moisture. Enough sterile deionized water was added to equal 3 g water per flask to give a final soil moisture of 75%. As was the case with denitrification potential measurements, LTR were measured in 25 mL Erlenmeyer flasks. Once the flasks were sealed with serum bottle stoppers, they were incubated for 14 days at 24°C and analyzed for CO2 concentrations in the headspace by gas chromatography. CO2 concentrations were measured on the same gas chromatograph and integrator as that used for measuring N2O. A flame ionization detector and a methanizer in series was used to measure CO2.
Exchangeable NH4+ (AMMONIUM) concentration was determined by shaking 10 g field-moist soil with 50 mL 2 M KCl for 1 h (Keeney and Nelson, 1982). After adding 0.3 mL 10 M NaOH to the slurry, NH4+ concentration was measured with a Orion model 95-12 ammonium electrode (Orion Research Inc. Boston, Ma, USA). Mineralizable N (MINN) was measured using the waterlogged technique of Keeney and Bremner (1966). Ten g of field-moist soil was added to 53 mL of distilled water in a 20 x 125 mm screw cap test tube and incubating at 40°C. After 7 days, 53 ml of 4 M KCl was added to the slurry and NH4+ concentration was determined with the ammonium electrode. Mineralizable N was calculated from the difference between initial and final NH4+ concentrations.
