INTRODUCTION 



The relationship of fire to soil nutrient cycling and 

 availability has been the subject of intensive study for many 

 years. Numerous reviews cover this and other facets of fire 

 effects on soil (such as Viro 1974; Raison 1979; Wells and 

 others 1979). Of the many essential plant nutrients, 

 nitrogen (N) is the most affected by burning because 

 nearly all N in forest soils is present in the organic form. 

 Volatilization of soil and plant N occurs during a fire, with 

 the actual amounts lost dependent on fire intensity (Knight 

 1966). In most soils, N-containing rocks and minerals 

 are lacking to replace fire-related N losses. Nitrogen 

 additions to the soil come from: (1) small amounts of N 

 present in precipitation and dust; (2) conversion or "fixa- 

 tion" of atmospheric N2 gas into usable forms by soil and 

 root-inhabiting microorganisms; or (3) application of 

 mineral or organic fertilizers. 



In contrast to total N changes, soil N availability may 

 be improved following a fire, with increases in N mineral- 

 ization rates frequently reported (Wells and others 1979). 

 The growth of postfire regeneration may be favored by 

 such higher levels of soil ammonium (NH4) or nitrate 

 (NO3), but these greater amounts of available N would 

 also be susceptible to leaching losses, especially as NO3. 



When evaluating the potential environmental impact of 

 various harvesting treatments on site quality, some 

 reduction of soil N levels from burning may be acceptable 

 as compared to the N losses that may occur by use of 

 mechanical site preparation equipment (Wells and others 

 1979). In order to determine the suitability of fire as a 

 postharvest site treatment, information must be obtained 

 on how fire affects the soil N status. This paper reports the 

 results of a prescribed broadcast fire on soil N levels and 

 other soil properties following the clearcutting of a 

 mature Douglas-fir/western larch forest in western 

 Montana. 



STUDY AREA AND TREATMENT 



This study was conducted on the Coram Experimental 

 Forest, approximately 10 miles (16 km) south of Glacier 

 National Park in western Montana, as part of a comprehen- 

 sive residue utilization research program. The experi- 

 mental site was an undisturbed 250-year-old forest 

 typical of the Douglas-fir/western larch timber type. 

 Douglas-fir {Pseudotsuga menziesii), western larch 

 (Larix occidentalis), subalpine fir {Abies lasiocarpa), 

 and Engelmann spruce (Picea engelmannii) were the 

 dominant tree species in the study area. The vegetation 

 represents an Abies lasiocarpa/Clintonia habitat type 

 (Pfister and others 1977). 



Elevation of the study plots was nearly 4,538 ft (1 375 m). 

 The plots were located on a steep (55 to 60 percent) 

 east-facing slope. The soils were quite stony (>50 

 percent) and derived from weathered argillite and impure 

 limestone material. Soil fine materials (<0.08 inch, or 

 <2 mm) were silt loam in texture (Klages and others 

 1976). 



The area used in this study was a 7.5-acre (3-ha) clearcut 

 harvested in the fall of 1974 and broadcast burned in early 

 September 1975. As part of the residue utilization study. 



all live and dead material (standing and down) of 8.0-ft 

 (2.5-m) length, 3-inch (7.6-cm) diameter, and at least one- 

 third sound was removed prior to burning. The fuels on the 

 site were relatively moist at the time of burning, which 

 resulted in a generally low-intensity fire. This was 

 shown by a duff depth reduction of only 25 percent from 

 the original 2.7-inch (6.8-cm) thick surface organic layer. 

 A more detailed description of the burn conditions 

 and fuel volumes have been given by Artley and others 

 (1978) and Benson and Schlieter (1980). An adjacent uncut 

 stand was used as a control. 



METHODS 



Sampling 



We took 30 soil cores (4x12 inches, or 10 x 30 cm) 

 randomly throughout the cut area 1 day prior to burning, 

 2 days following the burn, and 6 weeks after the burn for 

 determination of: total and available N, organic matter 

 content, acidity (pH), populations of nitrifying bacteria, 

 and the occurrence of water repellent layers. In addition, 

 we took 10 cores periodically from September 1975 to 

 October 1976, both in the burned area and the adjacent 

 control stand, to more closely monitor changes in available 

 N levels and soil acidity. The bimonthly samplings in 

 December, February, and April were limited to six cores at 

 midplot due to deep snow on the site. 



Each soil core was separated in the field into the follow- 

 ing fractions: (1) surface litter (0-| horizon); (2) humus 

 (O2 horizon); (3) decayed wood in the soil (referred to here 

 as the O3 layer); (4) surface 0-2 inches (0-5 cm) of 

 mineral soil; and (5) remaining mineral soil to a total 

 core depth of 12 inches (30 cm). Approximate soil volumes 

 occupied by each fraction were determined by measuring 

 its depth in the undisturbed core. The occurrence of 

 water repellency in the mineral soil was determined using 

 the water-drop-penetration test on the soil surface and 

 at each 2-inch (5-cm) soil depth (Adams and others 

 1970). 



Soil Analysis 



In the laboratory each soil fraction was shaken for 5 

 minutes in a standard 0.08 inch (2 mm) soil sieve. The 

 decayed wood and humus aggregates were gently 

 crumbled before sieving. Material less than 0.08 inch 

 (2 mm) was retained for chemical analysis. Determinations 

 of NH4 and NO3 content, acidity, and autotrophic nitrifying 

 bacteria populations were performed on undried soil within 

 24 hours after collection. 



Ammonium content of each soil fraction was measured 

 in a 2N KCI extract by a specific-ion electrode (Banwart 

 and others 1972). Nitrate content was determined on a 

 distilled water leachate by specific-ion electrode, 

 according to Bremner and others (1968). Acidity was 

 measured electrometrically using a 1 :2 mineral soil to water 

 ratio, or a 1:5 organic matter to water ratio. Nitrifying 

 bacteria numbers (Nitrosomonas) were estimated in the 

 O2 and 0-2 inches (0-5 cm) of mineral soil fractions using 

 the Most-Probable-Number Technique (Alexander and 

 Clark 1965). 



Soil for total N and organic content analyses was dried 

 in a forced-draft oven at 140° F (60° C). Total N values 



1 



