Refined Burning Prescriptions for Yosemite National Park
NPS Occasional Paper No. 2
NPS Logo


Fire characteristics, fuel reduction, and vegetative changes were all affected by fuel type, method of burning, and fuel moisture level. Prefire fuel and vegetation measurements and caloric values differed by understory fuel types.

Fire Characteristics

The fire characteristics differed variously by fuel type, method of burning, and fuel moisture level. In general, the fires became hotter as they burned under drier fuel conditions (Table 1).

Table 1. Rate of spread, available fuel energy, and intensity are shown for 32 prescribed fires. Fires which did not ignite show no rate of spread while fires which ignited and subsequently went out have a rate of spread of 0.0 m/sec. Available fuel energy was not affected by fuel type or method of burning, and intensity was not affected by fuel type.

Bear Clover

HeadBack HeadBack HeadBack HeadBack HeadBack

16%.5.1.3 .1.0
.0 .0
13% . .01246.763.15.6
10%1.2.8 .5.1 1657.260.814.1

IGNITION. For a fire to ignite and to continue to burn, sufficient heat must be present to raise the temperature of the water in the fuel, to release bound water from the fuel, to vaporize the water in the fuel, and to heat the water vapor to flame temperature. At the higher fuel moisture levels, insufficient heat was produced to accomplish the above; consequently, no new fuel was ignited and the fires went out.

The method of burning also affects ignition. For low intensity backfires, the primary heat-transfer mechanism for preheating and drying fuels is radiation. A fire with the flames bending away from the unburned fuels radiates heat to those fuels and begins to dry them. A headfire, on the other hand, in addition to transferring heat by radiation, preheats fuels by convection. The flames bend toward the unburned fuels, thereby increasing the drying rate. This explains why some of the headfires were able to ignite while backfires at the same fuel moisture level were not.

Differences in the fuel-bed characteristics for the four fuel types resulted in different rates of ignition success. The bear clover plots had fuel beds consisting of ponderosa pine needles draped over the low bear clover plants. Such an arrangement not only allowed considerable drying of the fuel to occur but also provided adequate oxygen for combustion to continue. These plots ignited with a backfire when the fuel sticks indicated a moisture content of 19%, while plots in the other fuel types did not ignite with moisture contents above 13 or 16%. The fuel beds of the needle and valley plots did not differ in their general nature, although the needle plots tended to have more fuel present and a somewhat more porous arrangement. Ignition did not occur for backfires at 19% for the needle plots and at 19 and 16% for the valley plots. Beneath the incense-cedar understory, the fuel bed was more compact and had a higher ratio of incense-cedar to ponderosa pine needles. The incense-cedar needles are much smaller and form a tight fuel bed allowing little aeration. This characteristic affected drying and combustion and, consequently, ignition.

RATE OF SPREAD. Once a plot was ignited, its rate of spread was influenced by some of the same factors as ignition. The analysis of covariance of the differences in rates of spread of the various fires showed that they were significantly affected by fuel moisture level and method of burning. Fuel type and the covariate effect of wind speed were not significant. The rates of spread for the various fires are shown in Table 1. Some of the fires which ignited show no rate of spread because they went out before burning the entire plot.

As less energy was required to drive off the water in the fuel at lower moisture levels, more was available for preheating and drying new fuel. This sped the combustion process and increased the rate of spread. For the four fuel types there appeared to be a threshold between the 16% and 13% fuel moisture levels. On the moist side of this threshold, the fires did not spread at all or, at best, spread slowly. Rates of spread were considerably higher on the dry side of the threshold.

Headfire rates of spread in all types exceeded backfire rates. The higher rates for headfires were due to the preheating effect of the convection of hot gases and to new ignitions ahead of the fire by the tilted flames. All fires seemed to spread well once ignition occurred and there was sufficient dry fuel to carry the fire. The arrangement of the topmost needles, which were the driest, did not appreciably differ from type to type. Even on the incense-cedar plots, there were enough loose needles above the compact fuel bed to carry the fire well. The two fastest fires could not be explained within the design of the experiment. A possibility is that since both fires had some bear clover in the buffer strip in front of the plots, these fires were hotter when they reached the plots.

Of the 32 fires, only two had wind speeds in excess of 3 mph. Wind speeds were within the acceptable range of 0-10 mph for all fires. Since the wind speeds were so low, they did not affect rate of spread. The three two-way interactions between fuel type, fuel moisture level, and burning method were not significant.

AVAILABLE FUEL ENERGY. The available fuel energy of a fire is the amount of energy actually released by the fire. This value was determined by multiplying the amount of fuel consumed by its caloric value. Available energy is therefore related to fuel loss.

When each fuel layer was analyzed separately, significant differences in available energy between the fuel moisture levels were recognized (Table 1). No significant differences were found between the weathered or decomposed needle layers.

INTENSITY. Fire intensity, the product of available fuel energy and rate of spread, was influenced most by those factors affecting rate of spread. The rate of energy released is directly related to the speed of the fire and, consequently, the method of burning. Fuel moisture, acting through available fuel energy, also affects intensity. The more fuel which is available and the faster it burns, the higher the intensity. A significant contrast was present between fires which burned at 19 and 16% fuel moisture, and those which burned at 13 and 10%. The former fires burned slowly if at all, while the latter fires spread well.

As additional moisture is lost below the 10% level, fuel moisture becomes less important. At that point, so little water is left in the fuel that the total amount of fuel available becomes more important. Fire intensity increased as fuel moisture decreased and was greater for headfires than for backfires (Table 1).

SCORCH HEIGHT. The scorch height of the fires differed significantly between the fuel moisture levels and the methods of burning (Table 2). However, neither wind speed nor the two-way interactions showed significance.

Table 2. Scorch height in meters is shown for 32 prescribed fires.

Bear Clover
HeadBack HeadBack HeadBack HeadBack

19% 1.7.9 .0
16% 1.61.2 1.1.6 .6
13% 2.41.9 6.91.1 1.31.0 1.4.0
10% 3.02.6 9.92.1 2.41.2 2.21.8

Since rate of spread is an indicator of the rate of energy release, the effects on scorch height were related to rate of spread. As a fire spreads faster, it releases heat energy at an increasing rate. A high rate of energy release has high temperatures associated with it. If the temperature in the crown reached the lethal level of 60°C, scorch occurred. Since the headfires and the backfires burning at the lower moisture levels had greater rates of spread, scorch height was also greater. Some of the fires scorched crowns before they went out.

Fuel Measurements

HEAVY FUEL. The amount of heavy fuel on each plot differed significantly by fuel type (Table 3). The valley plots, which had large concentrations of downed trees, had larger accumulations of heavy fuels.

Table 3. Prefire measurements for fine and heavy fuels are shown in grams per meter.2

Bear CloverNeedle ValleyIncense-cedar

Vegetative166.422.6 28.218.0
Fresh318.0422.3 383.8292.0
Weathered565.7604.5 532.5431.5
Decomposed4127.43565.7 4615.12830.9
Total Fine5177.52971.0 3915.53572.4
Heavy Fuel508.61726.5 5062.41261.3
Total5686.16341.6 8977.94833.7

2California Wildland Danger Rating System (USFS 1962).

In general, the heavy fuels were not affected by the fires. In the spring these fuels are wet internally, not drying until mid-summer. The greatest amount of reduction occurred in the needle plots where the heavy fuels were relatively small in size. The valley plots, which had concentrations of large heavy fuels, had no reduction of those fuels.

FINE FUEL. The amount of fine fuel (Table 3) differed by fuel type. There were significantly more fine vegetative fuels on the bear clover plots due to the bear clover ground cover. The incense-cedar plots had less fuel in each layer than the others because there were fewer ponderosa pine trees and hence fewer pine needles on these plots. Other comparisons showed that the bear clover plots differed in the amount of fuel in the fresh, weathered, and decomposed layers from the needle and valley plots, and that the valley and needle plots did not differ in the amount of vegetative fuel present.

The average amount of vegetative and fresh needle fuel reduction is shown in Table 4. Prescribed fires within the range of intensities produced in this study had little effect on the reduction of fuels in the weathered and decomposed needle layers. The inability to show significance was due to the small quantities of fuel consumed in these layers and the large variance between before and after subplots in the decomposed layer. The amount of fuel burned in the lower layers is small when burning is done during the spring. These layers have high moisture contents and dry slowly in comparison to the fresh needle layer. As the season progresses, differences in moisture content become smaller and more of the weathered and decomposed layers are consumed.

Table 4. Fuel losses in grams per meter2 are shown for the vegetative and fresh needle layers. Losses in the weathered and decomposed needle layers were not significant.

Bear Clover
Loss% Loss% Loss% Loss%

19% Vegetative 167.088.2 0.00.0 0.00.0 79.044.1
Fresh 158.576.2 118.025.7 98.531.0 49.518.8
16% Vegetative 140.5100.0 41.5100.0 0.00.0 60.028.1
Fresh 234.088.0 378.087.4 101.525.6 55.022.1
13% Vegetative 186.5100.0 31.5100.0 101.598.1 115.551.3
Fresh 341.5100.0 500.598.9 299.568.2 88.542.6
10% Vegetative 106.5100.0 40.0100.0 0.00.0 169.578.9
Fresh 470.592.2 385.598.6 329.083.7 440.097.4

The amount of vegetative fuel reduction was affected by fuel type and fuel moisture level, but not by the method of burning. The bear clover and incense-cedar plots, which had distinct vegetative fuel layers, had greater losses than the valley and needle plots. The latter two types had practically no vegetative layer and did not differ from each other.

Fuel moisture effects interacted with fuel type to produce differences between fuel moisture levels. The bear clover plots burned at higher moisture levels due to aerated fuel-bed characteristics which were influenced by the plants. As a consequence, the bear clover plants were also consumed at those levels. Beneath the incense-cedar understory, the compact fuel bed did not carry a fire until the fuel moisture level reached 10%. Since the bear clover plots burned at all moisture levels, the effect due to moisture level was produced by the incense-cedar plots. These combined effects also made the fuel type-moisture level interaction significant.

Practically all the vegetative fuels on the bear clover plots were burned by headfires and backfires alike. Incense-cedar fuel reduction was also not affected by the method of burning since the small amounts consumed were almost equal. The strong influence of the type effect made the type-method interaction significant.

The loss of fresh needle fuels was significant for fuel type and fuel moisture level. Method of burning did not affect the amount of loss. The incense-cedar plots differed significantly from the rest, and there was a significant contrast between the valley and needle plots. Each fuel moisture level contrast was also significant.

Differences in fuel reduction due to fuel type are a result of the fuel-bed characteristics of the various types. The incense-cedar plots, with their compact fuel beds, did not burn over much of the fuel moisture range and, consequently, had little reduction in the fresh needle layer. Although the bear clover plots had fuel beds which were quite porous, there was no significant difference between them and the combined valley and needle plots. This could be attributed to the fact that the needle plot fuel beds, while not quite as open, were of such a nature that most of the fresh needle layer was consumed at all moisture levels except the 19% level. The contrast between the needle and valley plots, on the other hand, was significant, indicating some difference in fuel beds. One obvious difference was the presence of some herbaceous material on the valley plots. Their added moisture content could have been enough to cause differences in fresh needle reduction.

The fresh needle layer was most sensitive to changes in fuel moisture content. Consequently, as more water was present in the fuel layer, more energy was required to heat and drive out the water. This energy could not combust new fuel until the energy requirements of the water had been met.

Method of burning was not significant because, if a fire burned, both headfires and backfires consumed all the available fuel at any particular moisture level. The strengths of the type and moisture effects combined to make that interaction significant.

FUEL CALORIC VALUES. There were significant differences in fuel caloric values due to fuel type, fuel layer, and the type-layer interaction (Table 5). Orthogonal contrasts for fuels with ash showed that the bear clover fuels differed from the needle and incense-cedar fuels combined and the needle fuels singularly. Fuel-layer contrasts were significant between the decomposed layer and the rest, between the vegetative layer and the fresh and weathered layers, and between the fresh layer and the weathered layers. The drop in caloric value with ash is primarily due to increasing incorporated inorganic matter in the lower litter layers.

Table 5. Caloric values for the various fuels layers are shown in kilocalories per gram with and without inorganic ash.

Bear Clover
w/ashw/o ash w/ashw/o ash w/ashw/o ash w/ashw/o ash


With the inorganic ash removed, fuel type, fuel layer, and interaction effects were still significant. The contrast between the valley fuels and the rest was also significant. The caloric values for the bear clover fuels and the needle and incense-cedar fuels did not differ although all the other contrasts remained significant.

Vegetation Measurements

OVERSTORY BASAL AREA. The overstory basal area did not differ significantly between fuel types since the areas were selected to minimize overstory differences. The average total basal area on all plots was 80.01 m2/ha. Ponderosa pine dominated the stands with 61.2% of the total basal area, consisting primarily of trees with diameters greater than 20 inches. Scattered, larger incense-cedar trees and numerous smaller trees made up 33.9% of the overstory. The remainder of the basal area was distributed over California black oak, Douglas-fir, and white fir trees.

UNDERSTORY BASAL AREA. Prefire understory basal area varied between types. This was expected since the understory was the criterion for selection of fuel types. The incense-cedar plots had 90% of the total basal area of 39.36 m2/ha. Incense-cedar reproduction made up 95% of the basal area on the incense-cedar plots, the remainder being ponderosa pine. The sparse understory on the bear clover, needle, and valley plots consisted of scattered incense-cedar trees.

The change in understory basal area after burning was 2.34 m2/ha (100%) for the bear clover plots and 7.17 m2/ha (21.4%) for the incense-cedar plots. Ponderosa pine understory basal area was reduced 0.29 m2/acre (14.8%) on the incense-cedar plots. There was no significant reduction on valley and needle plots.

UNDERSTORY DENSITY. The understory density prior to burning differed significantly between fuel types. Table 6 shows the number of stems per hectare of the various species by height class for the four fuel types. Fuel type and fuel moisture level, as well as their interactions, significantly affected density losses in the 0-0.3 m height class for bear clover, ponderosa pine, and all species combined.

Table 6. Prefire understory density in stems per 0.1 ha is shown for each species. Losses in stems per 0.1 ha and percent loss are also shown.

Species Size
Bear Clover
Bef.Loss% Bef.Loss% Bef.Loss% Bef.Loss%

Bear Clover 0-.39780 790080.7300 28394.0 00 0.01148 423.7
Incense-cedar 0-.3 800.0 2000.0 2000.0 4000.0
.3-1 000.0 1300.0 000.0 9000.0
1-3 33100.0 1000.0 300.0 1784826.9
Tot 11327.2 4300.0 2300.0 3084815.6
Ponderosa Pine 0-.3 000.0 000.0 985758.2 200.0
.3-1 000.0 300.0 000.0 300.0
1-3 000.0 000.0 000.0 27311.1
Tot 000.0 300.0 985758.2 3239.4
All Species 0-.3 9788790080.7 32028388.4 1185748.3 1190423.5
.3-1 000.0 1600.0 000.0 9300.0
1-3 33100.0 1000.0 300.0 2055124.9
Tot 9791790380.7 34628381.8 1215747.1 1488936.3

Density losses in the 0.3-1 m height class were not significant. To a large extent, this size class was absent on all plots except the incense-cedar plots on which the losses were confined to some incense-cedar trees on the 10% plots.

A greater number of trees were in the 1-3 m height class, primarily on the incense-cedar plots. Density losses for incense-cedar and all species combined were attributed to fuel type and moisture level, and to the type-moisture interaction. The incense-cedar plots differed from the plots in the other type and the 10% plots differed from those which did not burn.

<<< Previous <<< Contents >>> Next >>>

Last Updated: 01-Mar-2005