Fire Monitoring Program
Preliminary Results

Yosemite National Park
Fire Effects Monitoring Results in Yosemite's White Fir-Mixed Conifer Forest: Fuel Load and Tree Density Changes
Caroline Lansing, Prescribed Fire Specialist

Presented at Fire in California Ecosystems:
Integrating Ecology, Prevention, and Management
November 17-20, 1997,
Bahia Hotel, San Diego, CA

Abstract
Five years of Yosemite National Park fire effects monitoring data are analyzed to evaluate changes in fuel load and tree density following management ignited prescribed fire. Total fuel load in the white fir- mixed conifer forest was reduced by 57% immediately following prescribed fire. Total woody fuel load was reduced by 40%, with the smaller woody size classes showing the greatest reduction ($76%). Duff load was reduced by 92%. Total fuel load increased to 72% of prefire levels within 5-years postfire. Woody fuels increased to 91% of prefire loading, while duff accumulated to only 15% of prefire levels. Mean fuel accumulation rates were 2.1 tons/acre/yr (0.8 tonnes/ha/yr) for woody fuels, and 0.4 tons/acre/yr (0.1 tonnes/ha/yr) for duff. Overstory and pole tree densities were reduced by 38% and 80% respectively within 5 years following prescribed fire. The pole tree mortality resulted in a shift in stand structure with pole tree proportion of total tree density declining from 50% prefire to 24% five years postfire. Little change occurred in species composition. These results are used to assess achievement of Yosemite fire management objectives and to provide recommendations for objective refinement.

Methods

Site Description
The white fir-mixed conifer forest forms an almost continuous zone of dense forest between 5000 and 7500 ft (1500-2300 m). Overstory species consist primarily of mature white fir, sugar pine (Pinus lambertiana), incense-cedar and ponderosa pine (P. ponderosa), but also include red fir (A. magnifica), Jeffrey pine (P. jeffreyi), and giant sequoia (Sequoiadendron giganteum). Understory tree species are primarily white fir and incense-cedar, with some sugar pine and ponderosa pine. The ground layer is sparse with < 20% shrub cover. Fuels are best described by the short-needled conifer fuel model (Fuel Model 8) with inclusions of heavy dead and downed fuels resembling Fuel Model 10 (Albini 1976). All burns in this analysis were conducted between 1989 and 1995 within the Yosemite prescription parameters specified for Fuel Model 8 (van Wagtendonk 1974). None of the burn units had experienced fire within the last 30 years.

Burning Conditions
Burns were ignited utilizing a strip head-fire ignition pattern. Backing rates of spread ranged from 0-1 chain/hr (0-20 m/hour) with flame lengths of 0-1 ft (0-0.3 m), Head fire rates of spread ranged from 1-2 chains/hr (20-40 m/hour) with flame lengths of 0-2 ft (0-0.6 m). Ambient conditions were as follows: temperature 40-80EF (4-27EC), relative humidity 20-65%, and mid-flame windspeed 0-6 mph (0-10 km/h). Fuel moistures were: 1-hour time lag fuel moisture (TLFM), 6-13%; 10-hour TLFM, 8-10%; 100-hour TLFM, 10-16%; and 1000-hour TLFM, 15-30%.

Data Collection and Analysis
Plot locations were selected utilizing a stratified random sampling design within white fir-mixed conifer forest areas designated for management ignited prescribed fire. Data were collected within the 20 m x 50 m plots prefire, immediately postfire, and 1-, 2-, and 5-years postfire. Overstory trees $ 5.9 inches (15 cm) dbh were recorded within the entire plot area while pole trees between 1.0 - 5.9 inches (2.5-15 cm) dbh were sampled within one 10 m x 25 m quarter of the plot. All sampled trees were tagged, mapped, identified to species, and recorded as live or dead in accordance with Western Region Fire Monitoring Handbook protocols (USDI National Park Service 1992). Fuel load was measured along four 50 ft (15.2m) transects per plot using the planar intercept method (Brown and others 1982). Woody fuel load includes: 1-hour (0-0.24 inches or 0-0.63 cm in diameter), 10-hour (0.25-0.99 inches or 0.64-2.53 cm ), 100-hour (1.0-2.99 inches or 2.54-7.61 cm), and 1000-hr ($ 3 inches or $ 7.62 cm) TLFM fuels. Total fuel load also includes duff which consists of the layer of partially decomposed, consolidated organic matter below the litter layer. Litter, which is defined as the freshly cast organic matter still retaining its morphological characteristics, was measured but is not included in the total fuel load calculation.

Data were analyzed utilizing the Fire Monitoring Software version 3.0 (USDI National Park Service 1997). This software provides a platform for data entry and storage while also performing functions including minimum plot calculations and analyses of change over time. Using fuel load as the primary variable, calculations indicated a need for a minimum of 6 plots to achieve an 80% confidence level with a precision value of 20. In this analysis, immediate postfire results are based on 11 plots that burned in 4 prescribed fires. Postfire fuel accumulation and tree density results are based on 6 plots that reached the 5-year postfire stage. Minimum plot recommendations for the secondary variable of tree density were not met. Therefore while trends can be observed, additional data collection will be necessary before generalizations can be made regarding tree density.

Introduction
Yosemite National Park began implementing a management ignited prescribed fire program in 1970. Goals of the program include the utilization of prescribed fire as a tool both to simulate natural fire regimes and to reduce hazardous fuel loadings. While these goals can be complementary, a century of effective fire exclusion has resulted in unnaturally high fuel loads, thereby precluding burning under natural conditions in many areas. Management ignited prescribed fires in heavily fueled areas have fuel reduction as their primary objective. These fires are intended to restore fuel loadings and forest structure to levels within the range of natural variability. Once this has been achieved, subsequent burns more closely resembling fires occurring under natural conditions can be implemented.

The white fir (Abies concolor)-mixed conifer forest occupies 31% of the area in which Yosemite conducts management ignited prescribed fire. The estimated natural fire return interval in this community is 8 to18 years (USDI National Park Service 1990). Between 1931 and 1995, 947 lightning fires were suppressed in Yosemite white fir-mixed conifer forests, leaving 77% (88,000 acres) of this community with unnaturally high fuel loadings. Fire exclusion has favored the invasion of white fir and incense-cedar (Calocedrus decurrens) understory trees (van Wagtendonk 1985). Dense thickets of these trees, coupled with higher downed woody fuel loadings, have created conditions conducive to unnaturally severe wildland fires. The Ackerson Wildfire of 1996 burned an additional 19,000 acres in Yosemite white fir-mixed conifer forest, serving as a poignant example of increased flammability due to fire exclusion.

The primary objective for the initial prescribed fire in the white fir-mixed conifer forest is $50% reduction in total fuel load to be achieved through a universal $50% reduction across all size classes of woody fuels, litter and duff. A secondary objective for the initial prescribed fire is to achieve mortality on $40% of pole size (1.0- 5.9 inches (2.5-15cm) dbh) white fir and incense-cedar, thereby reducing the aerial fuels that contribute to development of crown fire.

Recognizing the need for a method of evaluating achievement of objectives and over-all program success, fire managers began implementing a long-term fire effects monitoring program in 1989. The monitoring program follows the methodologies outlined in the National Park Service Western Region Fire Monitoring Handbook (USDI National Park Service 1992). In this analysis, fire effects monitoring plots are used to assess fuel reduction and accumulation, as well as to examine short-term changes in pole and overstory tree density following prescribed fire. The results are then incorporated into recommended program refinements.

Results and Discussion

Immediate Fuel Reduction
Total fuel load (n=11 plots) decreased 57% from 52.1 tons/acre (18.8 tonnes/ha) prefire to 22.3 tons/acre (8.0 tonnes/ha) immediately postfire. Duff was reduced the greatest, decreasing 92% from 17.5 tons/acre (6.3 tonnes/ha) to 1.4 tons/acre (0.5 tonnes/ha). Woody fuels decreased from 35.0 tons/acre (12.6 tonnes/ha) prefire to 21.0 tons/acre ( 7.6 tonnes/ha) immediately postfire representing a 40% reduction (Figure 1).

Within the woody fuel classification, the 1000-hr size class fuels comprised the majority of the total loading (84%) prefire, yet they showed the least immediate postfire percent reduction (28.9 tons/acre prefire, 19.8 tons/acre postfire) leaving 68% of their original loading. This comparatively low reduction is the result of higher fuel moistures (>20%) in this size class. Smaller fuels generally have lower moisture contents and require less heat to reach ignition temperature. These smaller woody size classes (1-hour, 10-hour, 100-hour) all showed much greater immediate postfire percent reductions ($76%), decreasing as a group from 5.7 tons/acre (2.0 tonnes/ha) prefire to 1.2 tons/acre (0.4 tonnes/ha) postfire (Figure 2). Smaller fuels occupy a small proportion of total woody fuel loading (16%) yet are important targets in hazard fuel reduction burns due to their significant contribution to accelerated rates of spread. Although fuel reduction rates were not universal across all fuel size classes, prescribed fires under these conditions were successful in accomplishing the overall objective of $50% total fuel reduction.

These immediate postfire results correspond well with comparable data collected at Sequoia and Kings Canyon National Parks (Keifer 1995). Duff loading and consumption results were similar, while Yosemite woody and total fuel consumption values were slightly lower than those found in Sequoia and Kings Canyon. The differences in consumption can largely be explained by higher 1000-hr fuel moisture during the Yosemite burns.

Postfire Fuel Accumulation
Total fuel load (n=6 plots) increased to 42.9 tons/acre (15.4 tonnes/ha) 5-years postfire, reaching 72% of prefire levels. Woody fuels achieved 91 % of prefire loadings, increasing from 30.0 tons/acre (10.8 tonnes/ha) immediately postfire to 40.6 tons/acre (14.6 tonnes/ha) 5 years postfire. This equates to a mean woody fuel accumulation rate of 2.1 tons/acre/year (0.8 tonnes/ha/year). Duff accumulated at a much slower rate, reaching only 15% of prefire loading. Duff increased from 0.6 tons/acre (0.2 tonnes/ha) immediately postfire to 2.4 tons/acre (0.9 tonnes/ha) 5-years postfire, resulting in a mean duff accumulation rate of 0.4 tons/acre/year (0.1 tonnes/ha/year) (Figure 3).

The mean woody fuel accumulation rate of 2.1 tons/acre/yr corresponds favorably with results of a similar study in Yosemite by van Wagtendonk and Sydoriak (1987) who found rates of 2.0 tons/acre/yr in white fir. However, their values for duff loadings and duff accumulation rates were significantly higher than those in this analysis.

Woody fuel accumulation after fire is generally attributed to fire-killed branches and small trees dropping to the forest floor. Although the initial prescribed fire resulted in significantly reduced total fuel loading, postfire fuel accumulation, in addition to the residual fuel loading, results in a fuel condition still exceeding levels considered to be within the range of natural variability. While these ranges of natural fuel loadings are not exactly known, estimates can be made based on fire return interval and natural fuel accumulation rates data (van Wagtendonk 1986). A second prescribed burn within 10 years, targeting greater reduction in the larger size class woody fuels, would move toward bringing fuel loadings to more natural levels. Reduced stand density as a result of the first burn should result in slower fuel accumulation rates after the second burn. Once a natural fuel load has been achieved, successive fires could be implemented under conditions and at frequencies more closely resembling natural fires.

Tree Density
Density of overstory trees (n=6) was reduced from 189 trees/acre (468 trees/ha) to 116 trees/acre (288 trees/ha) 5 years following prescribed fire, for a 38% reduction. Further analysis of overstory mortality by diameter class revealed that mortality was evenly distributed throughout the larger diameter trees both by size class and by species. Density of pole trees was reduced from 186 trees/acre (460 trees/hectare) to 40 trees/acre (93 trees/hectare), for an 80% reduction . White fir and incense-cedar poles, which together comprise 91% of the pole tree diameter class population, decreased 81% from pre-burn to 5 years postfire and showed little change in relative density between species. The remaining pole tree population consists of sugar pine which experienced similar mortality rates.

Current park objectives for burning in the white fir-mixed conifer forest do not specify desired or acceptable levels of mortality for the overstory diameter class. Therefore, while a 38% reduction in overstory tree density seems high, objective refinement and a limited sample size preclude such a conclusion. For the pole diameter class, the objective of $40% mortality on white fir and incense-cedar was achieved with a rate of 81%. This high mortality in the pole trees can be attributed to lethal cambial temperatures and crown scorch and has resulted in a change in stand structure. The percentage of trees in the pole category of 1.0-5.9 inches (2.5-15cm) dbh shifted from 50% prefire to 24% 5 years postfire (Figure 4). Repeated burns may continue this structural shift toward a diameter class distribution more closely resembling stand structure believed to occur under natural conditions.

Conclusions
Five years of fire effects data collected in Yosemite's white fir-mixed conifer forest provide sufficient information to enable park managers to evaluate success of short-term fire program objectives. The primary objective of $50% fuel load reduction was achieved in total fuel load, duff, and in all woody size class fuels except the 1000-hr fuels, which were reduced by only 32%. The secondary objective of $40% mortality on pole size white fir and incense-cedar was also achieved with a mortality rate of 81% within the plots sampled.

While short-term burn objectives of reduced fuel loading and pole tree mortality were generally met, fuel reduction and accumulation rates as well as shifts in stand structure already indicate a need for prescription and objective refinement. The fuel reduction data suggest that universal $50% reduction across all woody fuel size classes may not be achievable without an unacceptably high level of overstory mortality. Using the current prescription, 1000-hr fuels were reduced by only 32% yet the heat generated resulted in an overstory mortality rate of 38%. Objectives targeting greater fuel load reduction in the smaller woody size classes and less in the 1000-hr would still result in significantly reduced hazardous fuels while potentially limiting overstory mortality. Similarly, objectives defining acceptable levels of overstory mortality need to be developed.

The fuel accumulation data demonstrate that a second burn will be necessary in white-fir mixed conifer forests within approximately 10 years of the first burn. Fuel reduction objectives for this second burn should target ranges of absolute values for fuel loadings rather than targeting percent reduction values. Such ranges would enable fire managers to evaluate when an area could be considered sufficiently devoid of hazardous fuels and would also provide direction as to timing of subsequent burns.

The overstory and pole tree density reduction data indicate a positive trend toward a more natural size class distribution. As with fuel loading, target ranges of absolute values for density by species and diameter class should be specified in the objectives of the second burn. These target values should be based on estimates of what the stand structure and composition would be under natural conditions, while also considering desired density for sufficient hazard fuel reduction.

These recommended changes for improving current objectives to better address over-all programmatic goals are based on five years of fire effects monitoring plot data. While this is a small data set over a relatively short time frame, the recommended changes seek more to clarify program direction than to change it. Continued data collection and analysis must occur in order to provide information for further refinement of prescriptions and objectives to incorporate desired long-term results.

Acknowledgements
I would like to thank Dr. Jan van Wagtendonk for providing not only inspiration, but also tireless patience and advice. In addition, I would like to thank Leslie Uhr and the prescribed fire crews of Yosemite over the past several years. This paper is the fruit of their labor.

Literature Cited

Albini, F. A. 1976. Estimating wildfire behavior and effects. USDA For. Serv. Gen. Tech. Rep. INT-30. 92 p. Intermountain Forest and Range Experiment Station, Ogden, Utah.

Brown, J. K., R. D. Oberhue, and C. M. Johnston. 1982. Inventorying surface fuels and biomass in the Interior West. USDA For. Serv. Gen. Tech. Rep. INT-129. 48 p.

Keifer, M. B. 1995. Fuel load and tree density changes following prescribed fire in giant sequoia-mixed conifer forest: The first 14 years of fire effects monitoring. (In Review)

Rothermel, R. C. 1972. A mathematical model for fire spread predictions in wildland fuels. USDA For. Serv. Res. Pap. INT-115. 40 p.

Swetnam, T. W. 1993. Fire history and climate change in giant sequoia groves. Science 262: 885-889.

USDI National Park Service. 1990. Yosemite National Park Fire Management Plan. Yosemite National Park. Yosemite, CA. 131 p.

USDI National Park Service. 1992. Western Region Fire Monitoring Handbook. Western Regional Office. San Francisco, CA. 134 p.

USDI National Park Service. 1997. Western Region Fire Monitoring Handbook- Fire Effects Monitoring Software version 3.0

van Wagtendonk, J. W. 1974. Refined burning prescriptions for Yosemite National Park. USDI Nat'l Park Serv. Occ. Pap. 21. 21p.

van Wagtendonk , J. W. 1985. Fire suppression effects on fuels and succession in short-fire-interval wilderness ecosystems. p.119-126 In: Proceedings — symposium and workshop on wilderness fire. USDA For. Serv. Gen. Tech. Rep. INT-182. 434 p.

van Wagtendonk, J. W. 1986. The Role of Fire in the Yosemite Wilderness. P. 2-9 in: Proc. Nat=l Wilderness Res. Conf.: Current Res. USDA, For. Serv. Gen. Tech. Rep. INT-212. 553 p.

van Wagtendonk, J. W. and C. A. Sydoriak. 1987. Fuel accumulation rates after prescribed fires in Yosemite National Park. Proc. 9th Conf. Fire For. Meteor. 9:101-105.

van Wagtendonk, J.W. 1997. Estimating Fuel Bed Characteristics of Sierra Nevada Conifers. (In Review) - submitted to the Western Journal of Applied Forestry.