The Use of Fire in Silviculture1 Pierre Delabraze and Jean Ch. Valette2 During the last century, the maintenance of chestnut plantations to make harvest easier, of pinelands to reduce fire risk and damage, and of pastureland to get rid of ungrazed grass, was carried out by means of small winter prescribed burns. This method was nearly completely given up as a consequence of the exodus from rural areas. In forests under the authority of the state and the departments, prescribed burning was not accepted as a forestry technique. Currently the use and even the introduction of fire in forest areas have been banned for fear of fire outbreaks. Owners are given impairments from October 15 to April 1. The decrease in fire danger (outbreaks and especially spread) is related to the clearing of underbrush. The four techniques: clearing by hand or machine, use of weed-killers, use as pasturelands, and prescribed burning are scientifically compared in order to present a set of efficient and often complementary methods to the managers. To deal with prescribed burnings, the Mediterranean Sylviculture Center of the I.N.R.A., Forest Research Department, has developed research in order of urgency established after consultation with the forest managers. Dealing first with forecasting of fire danger, the research has led to study of the silvicultural possibilities of fire along these lines: Abstract: In the French mediterranean area the use of fire, practiced in the past by the farmers, is now being reactivated by the forest managers. It relies on the flammabilities of the most important species and on the combustibilities of the main vegetal associations. The first prescribed burns demonstrate the sensitivity of forest trees and the role of dead material and wind. The results of these tests lead also to a better evaluation of fire risk and to organization of the clearings. - use of data gathered through these investigations, in order to perfect the prescribed burning methods. FLAMMABILITY OF FOREST SPECIES Principle of Measurement The flammability of a vegetal sample is calculated according to the time necessary for appearance of flame when the sample is subjected to a fierce heat radiation. Apparatus and Method Experimental Apparatus A radiator sends out a flux of about 7 watts per square centimeter and 3 microns wavelength. The pilot flame allows the ignition of the air-gas mixture resulting from the thermic decomposition of the sample, but it does not play any part in this decomposition. − flammability of the main species of forest trees, in order to determine the risks of fire outbreak and to understand their development − combustibility of the main forest associations, to define the risks of fire spread − combustibility of the forest litter responsible for both the outbreak and spread of fires 1 Presented at the symposium on Dynamics and Management of Mediterranean-type Ecosystems June 22-26, 1981, San Diego, California. 2 I.N.R.A. Station de Sylviculture Mediterraneenne Avenue A. Vivaldi - 84000 AVIGNON. Gen. Tech. Rep. PSW-58. Berkeley, CA: Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture; 1982. Figure 1--Flammability measurement apparatus. 475 The Fuel (table 1) Only the flammability of the fine elements of the main forest species is measured, as these elements cause the fires. Experimental Checklist and Significant Data For each studied species, the monthly periodicity of testing is reduced to 10 days when the fire danger increases, generally in the summer time. Picked according to very precise criteria on homogeneity, the 200 or 300 grams of green matter required for a specific test are put into tight bags under partial vacuum and then placed in a cooling-box; thus the water-loss is reduced to a minimum during transport. A specific test is composed of 100 basic tests made in three running series. Each 1-gram sample is put on the heating disk. The time between the moment when the sample is set in place to the appearance of the first flame is recorded. The test is positive if the time is less than 60 seconds. Then for each specific test these values are calculated: − the percentage of positive tests − the mean flammation time, averaging the flammation times of the positive tests, expressed in seconds − the mean moisture content, averaging the moisture contents of the four samplings, expressed as a percent of the ovendry weight (24 hours at 60ºC). The close relation between the percentage of positive tests and the mean flammation time have made a scale of flammation possible. (table 2) Results Species can be classified at regular intervals according to their flammability. Table 1 makes clear that the flammability is: − grossly linked to plant physiologic activity; very low or close to null at the time of bud burst, it increases with the time of new tissue lignification and reaches a maximum at the time of summer dormancy − temporarily influenced by rainfall or the air moisture content; specific reactions take shape-the pluviometry, and more precisely the amount of water stored in the soil, bring about wide variations in flammability. Use of the Results Knowledge of the specific flammabilities leads to: 476 - defining the fire outbreak risks and above all their sudden aggravation and their territorial distribution - making seasonal maps of fire danger, from the forest associations maps, which should give a sufficiently accurate account of the flora composition of the various stories. Table 1--Specific flammabilities during the fire danger period Studied species Flammability July August September 4 4 4 4 4 4 5 5 5 4 4 4 5 4 4 5 5 5 4 4 4 4 4 4 5 5 5 0 1 2 2 4 4 5 5 5 5 4 5 3 2 4 Calcareous Provence Tree story Pinus halepensis Quercus ilex Quercus pubescens Shrub story Quercus coccifera Phillyrea angustifolia Rosmarinus officinalis Cistus albidus Thymus vulgaris Herbaceous story Brachypodium ramosum Crystalline Provence Tree story Pinus pinaster Quercus suber Shrub story Erica arborea Erica scoparia Arbutus unedo Cistus monspelliensis Calluna vulgaris 2 4 3 3 3 3 4 4 3 5 3 5 5 4 5 4 5 5 2 4 4 4 3 3 4 3 4 3 2 5 - 4 5 5 5 5 5 5 5 3 3 3 4 5 4 1 2 2 5 4 5 3 3 3 5 5 4 2 4 0 2 3 4 4 2 3 4 4 4 1 2 3 4 4 1 3 4 4 4 3 3 3 4 4 2 3 4 4 4 3 2 3 4 4 1 1 2 3 4 1 0 3 Table 2--Flammability marks levels as a function of the percent of positive tests and the mean flammation time Mean flammation (time seconds) Percent of positive tests 98 to 95 to 90 to 85 to 80 to 100 97 94 89 84 80 less from from from from more 5 4 3 2 1 1 than 12.5 17.5 22.5 27.5 than 12.5 to 17.5 to 22.5 to 27.5 to 32.5 32.5 4 3 2 1 1 0 3 2 1 1 0 0 2 1 1 0 0 0 1 1 0 0 0 0 The forester in charge of forests threatened by fires is given the decision elements to rank his intervention: - on areas of species which are highly flammable or dangerous owing to sensitive surrounding plantations 1 0 0 0 0 0 − on bushy and low-branching plants such as Pinus halepensis, with its dead lower verticils, covered with dry needles, which come nearly in contact with the high calorific potential associations of Ulex parviflorus, Quercus coccifera, and Brachypodium ramosum − last but not least, when he uses prescribed burning. COMBUSTIBILITY OF FOREST ASSOCIATIONS Principle First restricted to the low bush associations, the study of combustibility is based on a replanting of the various vegetal stories, from the litter to the bush, on a combustibility measurement apparatus, and consists in recording the data of the combustion. Test Phases and Data Collected The vegetation and litter are both gathered from homogeneous and representative associations on eight aligned and continuous plots, each 1 square meter. Before each sampling a sketch of the plant layout on the soil is drawn, along with a description of the phenologic stages and of respective measurements and weight. The crop and all the following processes are preceded by measurements of the air temperature and moisture content and of the wind speed and bearing. Transport is carried out under cover. The plants are planted out on the apparatus according to the sketches. A detailed checklist states the fire ignition conditions. During the burning, the fire rate of spread and the flame front characteristics are recorded. At the end of the test, the unburnt plants are measured and weighed. Apparatus and Method The Combustibility Measurement Apparatus (figures 2 and 3). The litter and collected vegetation are laid on identical surfaces on eight trucks, each 1 square meter by 25 centimeters. The moving walls act as the surrounding vegetation by protecting the fire from external agents (wind) and by reflecting the heat; a top screen simulates the tree canopy. The variable thread blade fan allows creation of wind at speeds ranging from 5 to 40 meters per second. Fuel The study deals with these main forest associations: Calcareous Provence Quercus coccifera Ulex parviflorus Rosmarinus officinalis Quercus ilex Crystalline Provence Arbutus unedo Erica raborea Cistus monspelliensis Calluna vulgaris Figure 3-- Combustibility measurement apparatus (inside). Results (table 3) Figure 2--Combustibility measurement apparatus (outside). The litter plays a determining part because of the surface dead fuel. Fire has great difficulty in spreading when the litter is missing, even if the living fuel is at its lowest moisture content. Only the surface of dead fuel burns when it is bulky; when it is abundant and light it encourages the fire spread. The fire rate of spread is negatively linked with the litter moisture content--hence the propitious consequences of a light rainfall. 477 The herbaceous story transmits the fire from the litter to the shrubby story. Its phenologic stage and its abundance affect fire spread more by its high flammability than by the released energy. The graminaceae that dry out in summer are the more dangerous; some of these turn green again after the first storms of late summer. The rain brings about a quick change in combustibility. It increases the moisture content of the litter and leads to a new start of the vegetative activity. In summer, 50 percent of the tests in which fire covers the 8 meters of the apparatus are in samples taken more than 10 days after the last rain; on the other hand, 85 percent of the tests whose combustion stops within the first meters of the apparatus have been on samples taken less than 10 days after a rain. Wind speed and rate of fire spread are positively linked to one another. The increase of the flame length and the width of the flaming area the flaming front incline on the combustible explain it. Fire spread is linked with the plant structure: for the same biomass, an effective low bulk density increases combustibility owing to the large heat-absorbing surface and to the possible ventilation inside the fuel. Table 3--Combustibilities of the main forest associations Forest Associations Combustibility Calcareous Provence Quercus coccifera Ulex parviflorus Rosmarinus officinalis Quercus ilex Crystalline Provence Arbutus unedo Erica arbora Cistus monspelliensis Calluna vulgaris Good in summer, poor in autumn Excellent in summer, existent in winter Poor Similar to Q. coccifera Poor Excellent in summer, Excellent in case of summer drought Good in summer A fuel bed (86 x 58 x 6 centimeters) rests on scales (15.00 + 1.5 grams). Eight thermocouples (Nc-Na), one above the other, and a fluxmeter are connected to recorders. A photographic apparatus records the combustion. Studied Species Use of the Results The combustion of the low story vegetation is correctly modelled on the apparatus: − the flame front specific energies are calculated according to Byram's formula from the vanished combustible mass, its moisture content, fire rate of spread, and knowledge of the heat release − the energy radiated by the flame front is inferred from study of its development and of the measured temperatures (adaptation of the StephanBlotzmann formula) − also, the biomasses and the laws of their development are drawn from these tests. Material is harvested in autumn from branches of Pinus nigra Arn. ssp laricio, Pinus pinaster Soland, Pinus halepensis Mill, Cedrus atlantica Manetti. Fuel It has these characteristics: − surface/volume ratio of the leaf or the needle, expressed as = 2 e or = 2 (+2) e where e = thickness of leaf or needle − initial mass DEAD MATERIAL COMBUSTIBILITY − moisture content, expressed as a percent of ovendry weight Principle of Investigation − dry material initial mass Currently, dead fuel combustibility is only studied on leaves or needles picked on trees, desiccated, and then laid out by hand in regular layers. − bulk density, , in grams per centimeter Results Apparatus and Method Experimental Apparatus (figure 4) 478 Dried needles of Pinus nigra laricio = 4.7 mm -1 = 1.8 10 -2 g.cm -1 All the combustion data of the tests made in horizontal position are positively related to the initial ovendry mass from a threshold of steady rate of spread on 3.6 t/ha. The most accurate fittings are linear and the correlation coefficients are superior to 0.9. When the fire moves upwards, from a 20º slope, the fire spread laws are totally altered. The fire spreads up the slope very fast, burning only the upper layer of the dead fuel. The tests made with fires moving downwards give results which can be compared with those made on a flat position, yet the slopes of the regression lines are lower. Dried needles of Pinus halepensis = 6 mm -1 = 3.3 .10 -2 g.cm -3 The fire rate of spread threshold is close to 1.8 t/ha. All the data are related to the initial ovendry mass, yet the slopes of the regression lines are significantly lower than those of the two aforementioned pine trees. Dried needles of Cedrus atlantica = 33.10 -2 g.cm -3 Achieving a continuous carpet requires the equivalent of 5 t/ha. Even for the highest values the fire does not spread. Dried needles of Pinus pinaster HIGH CALORIFIC VALUES, ASH AND MINERAL CONTENTS = 2.7 mm -1 = 2.10 -2 g.cm -3 The characteristics of the combustion of this dead fuel are akin to those obtained with needles of Pinus nigra laricio. The slopes of the regression lines are not significantly different whereas the ordinates at the origin are slightly superior. The data allow an accurate estimation of the energy released during the combustion. The High Calorific Value Determining this allows us to calculate the potential energy of forest combustibles and the maximum heat release. Measured every month according to the classical calorimetry method, the H.C.V. are comparatively lower during growing time than after lignifying time. The H.C.V. of Erica arborea is higher than the values given for most of the tree species (table 4). The H.C.V. of mediterranean species are generally superior to those of more northern species. As for the building material, no link between the H.C.V. and the specific flammabilities has been clearly shown. Ash and Mineral Contents Determined every month by decomposition at 450°C in a muffle furnace, the ash contents are low during growing time and higher after lignifying time. The ash content of the Arbutus unedo leaves is double that of the ash content of the Erica arborea leaves, whose Ca content is very similar to the Ca content of the Ulex parviflorus shoots (table 4). The study did not show any clear link between these contents and the characteristics of combustion. THE PRESCRIBED BURNINGS IN FRENCH MEDITERRANEAN AREA Widely used in the past, the prescribed burning method is nowadays only used by a few shepherds in Corsica and some peasants in the Maures and Cevennes Mountains. Figure 4--Diagram of the apparatus for measuring the dead material combustibility. 479 Table 4--Specific high calorific values, ash and mineral contents Fuel Contents Leaves of Erica arborea Needles of Pinus halepensis Shoots of Ulex parviflorus Leaves of Arbutus unedo Leaves of Quercus ilex Leaves of Quercus coccifera H.C.V. (kJ.g-1) Ash 24.0 22.2 20.9 21.0 20.3 20.0 2.42 3.20 2.34 4.15 4.05 4.20 K Mg P (percent ovendry weight) 0.07 0.11 0.08 0.10 0.09 0.07 0.49 0.50 0.62 0.61 0.50 0.53 0.23 0.15 0.13 0.27 0.15 0.14 Ca 0.30 0.63 0.37 0.99 0.96 1.22 Aims Results (table 6) The tests on prescribed burnings are intended to define the rules of use according to the climatic, relief, and soil conditions and to improve knowledge of the reaction of forest tree species to the various types of prescribed burning. These tests being on the whole quite recent, the results are only provisory. − large damages to the tree story in tests 3, 8 and 9, which were made at the bottom of the slope on account of an adverse wind Plotting Out and Method Each test area is delimited by a surrounding cleaned area, gridded with posts to make marking and measurement easier to determine. − importance, nature, and qualities of the dead fuel and of the various vegetation stories − measurements of the trees where thermosensitive plates are set − microclimatic conditions − aspect of the ground and conditions prevailing there after the test. − if the wind direction is well established and if its speed is low and steady − The prescribed burns are generally made against the wind or at counterslope. Table 5--Scale for obvious fire damages 480 − the high combustibility of Quercus ilex and of Pinus halepensis foliage and the relative endurance of needles of Pinus pinaster. − if the dead fuel or herbaceous layer is continuous enough to guarantee a steady rate of fire spread Later, the restoration and new setting up of vegetation will be observed with interest (table 5). Obvious damages − satisfactory aspect of tests 4 and 7, ignited at counterslope, and of tests 1, 2 and 6, which were made against the wind: the low stories are well burnt back and trees are spared Prescribed burning can only be correctly carried out − fire ignition conditions, flame front spread, development of flames on trees, and reactions of the various species No damage Over 2/3 of the initial leaf mass are green From 1/3 to 2/3 of the initial leaf mass are green Less than 1/3 of the initial leaf mass is green Some green leaves No green leaves From the first investigations, the following points can be noted: if the fuel on the ground is dry enough − if the low verticils of trees are more than 2 to 4 meters above the low story. Conclusions Scale 0 1 2 3 4 5 These first tests show that planted firebreaks or trimmed bushes of Quercus pubescens covering a graminaceous grassland can be maintained, that there are maintenance problems in planted firebreaks where there are no resinaceous species, owing to a scarce litter, and that there are limits to the use of prescribed burning in forested areas as soon as the mattoral (maqui and garrigue) are fully grown. PUBLICATIONS OF THE CENTER ON THESES SUBJECTS Caramelle Ph., Clement A., 1978 - Inflammabilite et combustibilite de la vegetation mediterraneenne. Flammability and combustibility of the mediterranean vegetation. (3rd year thesis E.N.I.T.E.F., July 1978, 158 p.) Delaveaud, P., 1981 - Le feu, outil sylvicole. Utilisation pratique des donnees de combustibilite. The fire as a sylvicultural technique; practical use of combustibility data. (3rd year thesis E.N.I.T.E.F., July 1981, to be published.) Doat, J.; Valette, J. Ch., 1981 - Le pouvoir calorifique superieur d'especes forestieres mediterraneennes. High calorific value of mediterranean forest species. Ann. Sci. For., to be published. Valette, J. Ch.; Clement, A.; Delabraze, P., 1979 - Inflammabilite d'especes mediterraneennes. Flammability of some mediterranean species. (Research note 79-3, Mediterranean Sylviculture Center, 39 p.) Valette, J. Ch,; Vannier, G., 1981 - Inflammabilite estivale des principales especes forestieres de Provence calcaire. Summer flammability of the main forest species of calcareous Provence. To be published. 481 Table 6 --First tests of prescribed burnings 482