Fuel management plays a crucial role in preventing, fighting and modeling forest fires, both wild and prescribed. One of the main variables to measure within it is fuel load, so the estimation of it should be considered seriously.

Fuel load is used in fire effects assessments, useful as a measure of fuel reduction, critical in fire behavior simulation and prediction, and specially important in prescribed fire planning.

Several researchers have determined the best methods to estimate plant biomass, as described by Catchpole and Wheeler (1992). However, and under the RX-fire management requirements, the selected method should match the following requirements: it should be fast (based in visual techniques); it should be non destructive, as much as possible, to not alter the future RX-fire behavior; cost efficient; not sophisticated; and as accurate as possible. The method should also allow the differentiation between different strata, providing helpful information for modeling fire behavior.

The project “Using RX-fire to maintain shaded fuelbreak areas in Southern Spain” was developed in the Province of Malaga (Spain) by the Andalusia State Forest Service during year 1999 and 2000. Under the project, and from the necessity of estimating the fuel load for all the units burned, a visual-weight method, named the “Rank Method”, was defined and developed. The Rank Method is mainly based in the “weight estimation” method, introduced by (Pechanec, 1937). This technique has been proved to work very well for grass and low shrubs, especially in homogeneous vegetation with few species. The shaded fuelbreak areas, purpose of the project, are characterized by these type of vegetation complex. However, the Rank Method also improves the efficiency by adapting other techniques as the “relative weight estimation” (Hutchings and Schmautz, 1969), “comparative yield method” (Freidel and Bastin, 1988) or the “photo keys” method.

The accuracy of the Rank Method was tested by comparison with a destructive systematic sampling, considered as accurate. The results showed that the difference between the fuel load estimation from the Rank Method and the destructive systematic sampling were always under the 15% of the overall fuel load. Thus the most common difference between both methods was about 1 Ton/ha, considered as a low error if we compare the cost of the systematic sampling, as well as the relative influence of this “1 ton/ha” in the prescribed fire management.

However, the use of the Rank Method, as well as other visual techniques, is limited to those vegetation complexes relatively homogeneous, with predominance of grasses, litter and small shrubs. The taller the vegetation the greater the need for increasing the sample size, with the increase in cost and time resources. On the other hand, the decrease of accuracy would be very important in heterogeneous vegetation, specially if the weight differs importantly from one species to the others.

Due to shaded fuelbreak areas are usually characterized by homogeneous small shrubs, the method resulted very appropriate, fast and because of this, inexpensive. Thus it is very adequate for estimating the fuel load for prescribed fire management purposes.