| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Natural Resources and Environmental Science, University of Nevada, Reno, NV 89557
b USDA Agricultural Research Service, 920 Valley Road, Reno, NV 89512
* Corresponding author (dwj{at}cabnr.unr.edu)
Received for publication May 2, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
0.05) losses of C, N, P, S, Ca, and Mg from the forest floor. There were no statistically significant effects on mineral soils aside from a decrease in total N in the surface (A11) horizon, an increase in pH in the A11 horizon, and increases in water-extractable SO42– in the A11 and A12 horizons. Burning caused consistent but nonsignificant increases in exchangeable Ca2+ in most horizons, but no consistent or statistically significant effects on exchangeable K+ or Mg2+, or on Bray-, bicarbonate-, or water-extractable P concentrations. Before the burn, there were no significant differences in leaching, but during the first winter after the fire, soil solution concentrations of NH4+, NO3–, ortho-P, and (especially) SO42– were elevated in the burned area, and resin lysimeters showed significant increases in the leaching of NH4+ and mineral N. The leaching losses of mineral N were much smaller than the losses from the forest floor and A11 horizons, however. We conclude that the major short-term effects of wildfire were on leaching whereas the major long-term effect was the loss of N from the forest floor and soil during the fire. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
While fire causes losses of total N from the forest floor and possibly surface soils, it often causes increases in soil mineral N due to the heat-induced degeneration of soil organic N (Neary et al., 1999; White et al., 1973). Similarly, fire can cause the release of inorganic P from soil organic matter. However, fire effects on soil P availability are complex: availability could be decreased by adsorption to newly exposed or newly created Fe and Al hydrous oxide surfaces or precipitation with Ca. On the other hand, P availability could be increased by degradation and partial combustion of soil organic matter and by increases in pH, causing desorption from Fe and Al hydrous oxide surfaces. Thus, it is not surprising that the effects of fire on soil P availability vary from study to study, with some showing increases (e.g., Dyrness et al., 1989; Hauer and Spencer, 1998; Saa et al., 1993; Romanyà et al., 1994), some showing decreases (e.g., Carreira et al., 1996), and some showing a varying relationship to fire intensity (Ketterings and Bigham, 2000). Fewer studies report on the effects of fire on soil SO42–, but most studies to date report substantial increases in SO42– availability in soils, soil solutions, and even streamwaters after fire (Bayley et al., 1992; Blank and Zamudio, 1998; Chorover et al., 1994; Williams and Melack, 1997). These increases in SO42– availability could be due to the combustion of S in soil organic matter and by pH-induced desorption from Fe and Al hydrous oxide surfaces.
The effects of fire on water quality are especially important in the Lake Tahoe Basin on the California–Nevada border (Stephens et al., 2005). Lake Tahoe is an ultra-oligotrophic freshwater lake renowned for its pristine conditions and extreme clarity. The 500-km2 lake is surrounded by a forested watershed approximately 800 km2 (Boardman, 1959). The relatively small watershed to surface water area ratio has produced pristine water conditions resulting from historically low levels of nutrient input (Goldman, 1988). Since 1968 lake clarity has diminished by an estimate of 0.37 m yr–1 from increased primary productivity due to accelerated N and P inputs (Goldman 1988). It is now believed that the N to P ratio of inputs into Lake Tahoe has shifted and currently reflects a P rather than N limiting system (Jassby et al., 1994). A substantial portion of the Lake Tahoe Basin has been categorized as a high-risk environment for catastrophic wildfire (Smith and Adams, 1991). Stephens et al. (2005) found that prescribed fire in the Lake Tahoe Basin had no effect on soluble reactive phosphate and minimal effects on nitrate in streamwaters. However, the effects of wildfire, which often kills vegetation and burns with greater severity than prescribed fire, remain largely unknown.
In July of 2002, the "Gondola wildfire" ignited on the southeast side of Lake Tahoe near Stateline, Nevada, consuming approximately 280 ha, including 9 of 16 plots that had been previously established for a prescribed fire/mechanical treatment study. We had monitored leaching and sampled forest floors and soils in all but two of the plots that were to burn before the fire leaving us with the opportunity to study the effects of a wildfire with replicated, pre- and post-fire plots that included adjacent unburned controls. Such a circumstance is not unusual for prescribed fire, but is highly unusual for a wildfire. In this paper, we report the effects of this wildfire on forest floor loss, soil nutrient status, and soil leaching. We then compare our results with those typically expected following fire: losses of C and N from the forest floor, increases in pH and base cations in mineral soil, and increased N leaching (Certini, 2005; Neary et al., 1999; Raison et al., 1985).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Sixteen 400-m2 research plots were established in the fall of 2001 and baseline sampling was initiated the following spring. The wildfire occurred in July of 2002, completely burning five plots and partially burning four plots. All of the five completely burned plots and two of the four partially burned plots had been previously sampled for forest floor and soil nutrients. In this paper, we compare results from the five completely burned plots with the seven unburned plots that served as controls.
Forest Floor Nutrient Content
Before the fire, forest floor samples were collected from five random locations in each study plot from a litter ring of 0.07 m2 in area (Murphy, 2004). Samples were taken according to horizon (Oi, Oe, and Oa), and nonfoliar material up to 2.54 cm in diameter was placed into the category of "other." Post-burn samples were taken 1 m north of pre-burn sample locations within the burned plots. We did not resample the forest floor in the control area after the fire, assuming that net changes in it would be minor. Pre-burn Oe and Oa samples were floated in water to remove mineral material. Post-burn Oe and Oa samples were not floated in water due to the possible dissolution of any remaining ash and organic material that did not undergo complete combustion. Instead a 0.84-mm standard testing sieve was used to separate the mineral from the organic fraction in the post-burn samples. This size sieve proved to be sufficient in allowing smaller diameter material to pass through the sieve while retaining the organic fraction and larger diameter mineral material, which was subsequently removed by hand.
Large woody litter was inventoried by the methods of Pyne et al. (1996). For 100-h (>2.54- to <7.6-cm diameter) and 1000-h (>7.6-cm diameter) fuels, a single 4-m2 and a single 54-m2 subplot (respectively) were established in the center of each 400-m2 plot. The 100-h fuels within the 4-m2 plots were collected, dried to a constant weight, and later returned to the plot, minus a small subsample for nutrient analysis. For the 1000-h fuels, notation was made as to whether each piece was sound or decayed, and then lengths and mid-point diameters were recorded for calculation of the volume according to the Huber formula (Avery and Burkhart, 2002). Collection of 10 log sections from random locations, measuring their dimensions, and then drying and weighing them provided a density constant for use in converting volume to weight. Pre-burn samples were taken during May and June of 2002 and post-burn samples taken during September of 2002.
All forest floor samples were oven-dried at 55°C, weighed separately, and then bulked together in the laboratory by O-horizon category for each plot. Bulked Oi, Oe, and Oa horizon samples were ground in a Wiley Mini-Mill whereas samples from the "other" category were ground in a Wiley Mill-Model 4 (Thomas Scientific, Swedesboro, NJ). Ground samples were analyzed using a Jarrell Ash ion coupled plasma spectrophotometer (Thermo Jarrell Ash, Franklin, MA) after microwave digestion via a nitric acid hydrogen peroxide mixture at A&L Western Agricultural Laboratories (Modesto, CA). Total C and N were analyzed using a dry combustion C and N analyzer (LECO, St. Joseph, MI) at the Oklahoma State University Soil, Water, and Forage Analytical Laboratory (Stillwater, OK).
Soil
Pre- and post-burn mineral soil samples were taken immediately beneath the forest floor samples (five random locations within each plot). Pre-burn samples were collected during May and June of 2002 (2 mo before the wildfire); post-burn samples were collected during July of 2003 (1 yr after the wildfire). Soil was collected from four depths (0–10, 10–30, 30–60, and 60–100 cm) corresponding to the A11, A12, AC, and C horizons, respectively (Rogers, 1974), in each sampling. All soil samples were field bulked and homogenized by plot and depth. Bulked samples were oven-dried at 55°C and passed through a 2-mm sieve. The 2-mm fraction was analyzed for exchangeable Ca2+, Mg2+, K+, Na2+ (10 g soil in 50 mL 1 N ammonium acetate), bicarbonate-extractable P (2 g soil in 50 mL 0.05 M NaHCO3–), and Bray-extractable P (2 g soil in 0.5 M HCl plus 1 M NH4F). Extracts were analyzed using a Jarrell Ash ion coupled plasma spectrophotometer (Thermo Jarrell Ash) at A&L Western Agricultural Laboratories. Total C and N were analyzed using a dry combustion C and N analyzer (LECO) at the Oklahoma State University Soil, Water, and Forage Analytical Laboratory. Water extractions were performed using a mechanical vacuum extraction (Centurion International, Lincoln, NE) from 2.5 g of soil and 50 mL of deionized water through approximately 1.1 g of high-grade filter pulp (Schleicher and Schuell, Dassel, Germany). Extractions were analyzed for water-soluble ortho-P and SO42– using high performance ion exchange chromatography (Dionex, Sunnyvale, CA) at the Agricultural Research Service Soils Lab in Reno, NV. Two soil pH readings (deionized water and 0.01 M CaCl2) were measured from a 1:1 soil to solution ratio using a glass electrode at the University of Nevada, Reno.
Post-fire Erosion
Post-fire soil and forest floor sampling were complicated by erosional losses from the site during a high intensity (15.2 mm), short duration hail- and rainstorm event that we witnessed just 3 wk after the wildfire and before full post-burn access. This event caused visible rilling and losses of ash and soil from the burned area. The material that eroded from the fire collected in a riparian area immediately below the fire and we were able to subsequently estimate its mass. Points along the perimeter of the ash flow were delineated using a handheld GPS unit. GPS points on eight transects were also taken (2-m intervals) with depth and bulk density samples at each transect point. Data points were created as a theme in ArcView 3.3 (ESRI, 2000). Contours were created from these points of varying ash depth and bulk density. The two sets of contours were intersected to create a polygon theme of areas with different ash depth and bulk density. The area for each polygon was calculated and then multiplied by the ash depth associated with the polygon to get the ash volume. The resulting ash volume was then multiplied by the bulk density associated with the same polygon to obtain the ash mass located within the said polygon. This was done for all polygons within the ash deposition theme. The ash masses for all polygons were then summed to find the total ash mass within the ash deposition, 378.6 metric tons of ash. The mass of ash was then related back to the estimated source area. The outline of a minimum and maximum possible source area was created from GPS points within Arc View 3.3. The total deposition ash mass was divided by the minimum and maximum areas to get a minimum and maximum ash mass per ha of source area. These numbers were then divided by the sampled bulk density of the source area before the fire to obtain a range of possible depths of soil removed from the source area in the debris flow. This gave a maximum of 0.6 to 1.4 cm of surface soil that could have been lost from the site. Analyses of the nutrient movement associated with this event are still underway and will be reported in a later paper that addresses the complete nutrient budgets associated with this fire, including gaseous losses from vegetation, forest floor, and soil as well as post-fire erosional and leaching losses.
Inorganic Nitrogen and Phosphorus Flux
Soil inorganic N and P fluxes were estimated using resin lysimeters containing Rexyn 300 (H-OH) ion exchange resin according to the design of Susfalk and Johnson (2002). Resin lysimeters were placed in four randomly selected points within each control and burn plot at a depth of 10 cm from the surface of the mineral soil. Resin lysimeters were installed during the winter and spring months before and after the burn. Once removed from the field, resins from the lysimeters were extracted in 100 mL of 1 M KCl for 1 h and analyzed for NH4+ and NO3– using high performance ion exchange chromatography (Dionex). Ortho-P on the extract was analyzed using a Quick Chem 800 flow injection auto analyzer (Lachat, Milwaukee, WI) at the Oklahoma State University Soil, Water, and Forage Analytical Laboratory. Because of high blank values, it was not possible to use resin lysimeters to estimate SO42– flux.
Soil Solution
One soil water suction lysimeter (Soilmoisture Equipment, Santa Barbara, CA) was installed in each 400-m2 plot center at a depth of 30 cm to monitor soil solution pH, NO3–, NH4+, ortho-P, and SO42– during the winter and spring months following the wildfire event. Soil solution pH was determined using a glass electrode. Ortho-P was analyzed with a Quick Chem 800 flow injection auto analyzer. Ammonium, NO3–, and SO42– were analyzed using high performance ion exchange chromatography (Dionex) at the Oklahoma State University Soil, Water, and Forage Analytical Laboratory.
Statistics
For statistical analysis, plots were considered treatment replicates. Seven control plots and five completely burned plots were compared. Treatment effects on the forest floor nutrient content within the burned plots were determined using Student's t tests with Microsoft Excel software (Microsoft, 2003). As noted above, we did not sample the forest floor within the unburned plots, assuming that changes in it would be minimal over one year's time. For resin lysimeters and many soil parameters, both seasonal and inter-annual variations can be expected and thus simple comparisons of pre- and post-burn values within the fire could be confounded by both temporal variations and the effects of the fire itself. Thus, changes in soils and soil leaching measured with resin lysimeters as a result of fire were assessed by horizon using ANOVA with treatment (burned and unburned) and time (before and after the fire) on samples taken both before and after the fire. A significant effect of fire was indicated by significance (P 0.05) in the ANOVA treatment x time interaction term and Student's t tests. The ANOVA tests on soils were performed using DataDesk Version 6.0 software (Velleman, 1997) and Student's t tests were performed by Microsoft Excel software.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
0.05, Student's t test) significant decreases in forest floor contents of all but K (the changes in forest floor K were significant at the 0.10 level). Compared to pre-burn forest floor estimates, average forest floor C, N, P, S, K, Ca, and Mg losses were 94, 92, 76, 84, 53, 90, and 90%, respectively, following the wildfire (Fig. 1).
|
|
|
|
|
0.05 in the treatment x time interaction term) of the fire on soils were a decrease in N concentration in the A11 horizon; increase in pH in the A11; increases in water-extractable SO42– in the A11 and A12 horizons; and an apparent relative decrease in C to N ratio in the A12 horizon (Table 1). We suspect that the latter is a spurious result, however, given that actual changes in the C to N ratio in the burned plots were very small and the significance of the interaction term was mostly the result of an apparent (and probably spurious) increase in C to N ratio in the unburned plots (Fig. 2). Burning seemed to cause consistent but nonsignificant increases in exchangeable Ca2+ in most horizons, but no consistent or statistically significant effects on exchangeable K+ or Mg2+, or on Bray-, bicarbonate-, or water-extractable P concentrations (Fig. 2, 3, and 4).
Soil Leaching and Soil Solution
The fire caused statistically significant increases in NH4+ and mineral N (NH4+ + NO3–) leaching as measured by resin lysimeters (Fig. 5 and Table 2). The fire also appeared to cause increases in NO3– and ortho-P leaching, but these changes were not quite statistically significant (P = 0.07 for NO3– and P = 0.06 for ortho-P).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The losses of soil C and N with wildfire in this study agree with the findings of Baird et al. (1999) who reported significant C and N losses in the A horizon 1 yr following wildfire and Grier (1975) who calculated large losses of soil N during a wildfire in eastern Washington. On the other hand, Dyrness et al. (1989) reported both increases and decreases in soil N following fire in Alaska. Rashid (1987) measured increased soil C and N 5 mo after wildfire, but concentrations decreased to near control levels 16 mo later. Fernández et al. (1999) reported decreases in soil C initially after wildfire, but concentrations increased to levels comparable to unburned control samples within 4 mo due to the decomposition and disintegration of dead roots.
Our results conflict with previous research by Dyrness et al. (1989) and Blank et al. (1994) who found increases in extractable P immediately following wildfire. These authors did not report longer-term results (e.g., 1 yr following wildfire), however. If there were any initial differences in soil P status as a result of the Gondola wildfire, they were apparently temporary, returning to levels comparable to the unburned controls within 1 yr. DeBano and Klopatek (1988) reported that significant increases in NaHCO3––P lasted only 45 d following prescribed burning, and Adams et al. (1994) noted that it is common for fire-induced extractable-P increases to decline to pre-burn levels within 1 yr. The effects of the fire on soil solution ortho-P suggest that available P was higher in the early stages after the fire; however, cumulative P leaching as measured by resin lysimeters was not significantly affected by the fire. Thus, we can draw no firm conclusions about the short-term effects of fire on P availability in this study.
Our results for SO42– were consistent with previous research showing increases in both soil (Blank et al., 1994; Blank and Zamudio, 1998; Khanna and Raison, 1986) and soil solution (Chorover et al., 1994) SO42– after fire. These effects are most often attributed to the oxidation of S in organic material and the release of sequestered SO42– from plant litter tissue, but may also be augmented by pH-induced desorption of SO42– from soils.
The lack of significant increases in pH and exchangeable base cations in this study contrasts with most studies, which show increases in base cations and pH following fire due to the influx of ash into the soil (Certini, 2005; Khanna et al., 1994; Raison and McGarity, 1980). Fire can cause increases in pH both by the combustion of some carboxylic (acid) groups in the soil organic matter as well as through the release of base cations from the ash (Certini, 2005). We observed consistent increases in exchangeable Ca2+ and pH in all horizons after the fire (Fig. 3 and 4); the lack of statistical significance in these changes may well be an artifact of inherently large soil exchangeable base cation pools and high variability. Thus, while the expected patterns in pH and base cations in soils were not statistically confirmed in this study, the observed trends in exchangeable Ca2+ were consistent with the literature. For exchangeable Mg2+ there was an inexplicable downward trend in both burned and unburned plots after the fire (Fig. 4). Because these changes occurred in both burned and unburned plots, the effects of the fire were not statistically significant, even though some of the apparent changes over time were significant in both cases. While we cannot explain this apparent temporal variation in exchangeable Mg2+, we note that this result illustrates the importance of having control plots: had we monitored changes in only the burned plots, as is so often the case in wildfire studies, we would have drawn a very different and erroneous conclusion about the effects of fire on exchangeable Mg2+.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Ammonium and mineral N leaching as measured by resin lysimeters were greater in the burned plots after the fire but not before, and soil solution SO42–, NH4+, NO3–, and ortho-P concentrations were elevated during the winter and spring after the fire. Ortho-P leaching as measured by resin lysimeters was elevated in the burned plots after the fire, but the changes were not quite significantly different (P = 0.06). The leaching losses of N were considerably smaller than losses of N from the forest floor.
We conclude that from the biogeochemical cycling point of view, the most significant short-term effects of the wildfire were the increases in the soil solution concentrations and/or leaching of mineral forms of N, S, and P. The potential limitations of S to aquatic ecosystems in the Tahoe Basin are unknown, but the increases in N and P mobility could have significant impacts on water quality in Lake Tahoe. Over the longer term, the losses of ecosystem N capital from the forest floor and soil during the fire are probably most significant, unless such losses are replaced by post-fire N fixation.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Qian, S. L. Miao, B. Gu, and Y. C. Li Estimation of Postfire Nutrient Loss in the Florida Everglades J. Environ. Qual., July 23, 2009; 38(5): 1812 - 1820. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Qian, S. L. Miao, B. Gu, and Y. C. Li Effects of Burn Temperature on Ash Nutrient Forms and Availability from Cattail (Typha domingensis) and Sawgrass (Cladium jamaicense) in the Florida Everglades J. Environ. Qual., February 6, 2009; 38(2): 451 - 464. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. M. Carroll, W. W. Miller, D. W. Johnson, L. Saito, R. G. Qualls, and R. F. Walker Spatial Analysis of a Large Magnitude Erosion Event Following a Sierran Wildfire J. Environ. Qual., May 25, 2007; 36(4): 1105 - 1111. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. M. Loupe, W. W. Miller, D. W. Johnson, E. M. Carroll, D. Hanseder, D. Glass, and R. F. Walker Inorganic Nitrogen and Phosphorus in Sierran Forest O Horizon Leachate J. Environ. Qual., March 1, 2007; 36(2): 498 - 507. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Saito, W. W. Miller, D. W. Johnson, R. G. Qualls, L. Provencher, E. Carroll, and P. Szameitat Fire Effects on Stable Isotopes in a Sierran Forested Watershed J. Environ. Qual., January 9, 2007; 36(1): 91 - 100. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
