Algeria: Forest Fires in Algeria and the Case of the Domanial Forest of Bou-Taleb, Setif (IFFN No. 22

Among the factors that threaten the forests of Algeria fire constitutes the most dangerous factor and causes severe ecological, economic and, sometimes, human losses. Within some hours fire destroys what has naturally grown over years and centuries.

In Algeria, the fire statistics established for nine years (1979-1987) provide an average of 37,000 ha of burned surface (Tab.1) (Madoui 1988). Seen its negative aspects, fire is considered like a curse of which we must face periodically. The forest law (forest code) clearly state the struggle against forest fires being a duty of all citizen.

Tab.1. Statistics of fires in forest and non-forest lands (scrub, brushwood, Alfa) during the period 1979-87.
Source: Ministry for Water Resources and Forestry, Algeria (1987).

Year	Number of Fires	Burned surface (ha)		Total burned area (ha)	Average area burned (ha)
Year	Number of Fires	Forest Land	Non-Forest Land	Total burned area (ha)	Average area burned (ha)
1979	361	15,663	95	15,758	43.65
1980	730	19,731	7,215	26,946	36.91
1981	796	8,982	7,378	16,360	20.55
1982	638	7,354	2,027	9,381	14.70
1983	990	13,3042	87,528	220,570	222.80
1984	562	1,398	2,981	4,379	7.79
1985	747	1,757	3,274	5,031	6.73
1986	1,170	7,920	3,689	11,609	9.92
1987	1,321	10,439	12,862	23,301	17.64
Total	7,315	206,286	127,049	333,335	45.57

Every year a considerable budget is assigned for the active control of forest fires. However, fires continue to affect large vegetated areas. The data of 1983 are unforgettable in the memories of foresters and firemen. This situation was the reason why we conducted the survey presented in this paper to contribute data on forest fires in Algeria and particularly in mountains of Bou-Taleb. From an ecological point of view Bou-Taleb is a primordial importance because it acts as a natural barrier between the influences of the desert and the city of Setif. The mountains are occupied by the most important plant formations in Algeria with tree species such as cedar (Cedrus libanotica ssp. atlantica (Manetti)Holm.), oak (Quercus rotundifolia Lamk.), and pine (Pinus halepensis Mill.).

Being located in a distance of about 50 km inland from the coastline the massif of Bou-Taleb is situated between the high plains of Setif in the north and the basin of Hodna in the south. It is part of the mountain chain of Hodna and constitutes an important mountain link but is well fragmented in its oriental part (Fig.1).

The existence many chaabet (chaâba = ravine) and the relatively elevated altitude of these djebels give the massif of Bou-Taleb an extremely intersected relief in its whole. Prevailing steep slopes in its southern part and the sparse vegetative cover encouraged significant hydric erosion after summer downpours. From the climatic viewpoint the lower altitudes are characterized by a semi-arid bioclimate with cool winters in the north, cold winters in the south, and very cold winters in the sub-humid high altitudes. The average annual precipitation ranges between 300 and 600 mm. The dry season is longer at low altitudes and can last up to five months; in the higher altitudes a short dry season usually does not exceed three months.

The number of forest fires recorded in Bou-Taleb, compared to the data noted by Boudy (1955), is in rise since the independence, but irregularly. To highlight this observation, two statistical overviews on fire occurrence have been established for this region. The first concerns the period before the independence and covers the period 1907 to 1957; the second concerns the period after the independence for the period 1971 to 1991 (Tab.2).

During the first period of 51 years, 77 fires have been recorded in Bou-Taleb which destroyed a surface of more than 14,15O hectares, with an average of 1.5 fires per year and an annual mean surface burned of 277 ha. During this period half of the forest vegetation of Bou-Taleb has been devastated (49.76%).

Among the total of all fires recorded half of the number of fires (38 fire locations) devastated six main districts located all on the South flanks of the massif: Kef Haoumars (8 fires); Bou-Ričche (7 fires); Afghan, Thniet Sefra, and Chaabet Khrouf with 6 fires each and finally the Grouve canton with 5 fires. The surfaces affected, however, are relatively small as compared to the total fire occurrence.

Tab.2. Distribution of fires by canton (district) in the domanial forest of Bou-Taleb

Period	1907-1957 (51 years)		1971-1991 (21 years)
Cantons (Districts)	Number	Burned surface (ha)	Number	Burned surface (ha)
Kef Haoumar	8	836	1	5
Bou-Ričche	7	17	5	4
Afghane	6	8	3	11
Thniet Sefra	6	8	3	30
Chaabet Khrouf	6	181	3	9
Grouve	5	3	3	11
Bou-Ich	1	1,004	4	24
Tinzert	1	167	5	10
Total	40	2,224	27	104

It must be noted that in the years 1956 and 1957 the largest burned areas was recorded (10,300 ha) which corresponds to 73% of the total area burned; they corresponded to the war of Algeria (Madoui 1995).

On the other hand the fire statistics for the second period (21 years) show that 63 fires have been recorded in Bou-Taleb which destroyed more than 511 ha, with an average of three fires by year and a yearly mean surface of 24 ha.

The distribution of these fires in the massif, by canton, is irregular and have affected almost all mountains (Fig.1). Table 2 shows that the Bou-Ričche cantons, Tinzert, and Bou-Ich are most affected. In the period of 21 years, these cantons recorded 5 fires and 5 and 4 fires respectively, with the burned surfaces distinctly lower than the mean, except for the Bou-Ich canton.

The canton which recorded the largest burned surface is Chaabet Said and a small part of canton Bou-Rhioul with 225 ha devastated by only one fire. In the massif these two cantons are colonised by a green oak scrub (Quercus rotundifolia Lamk.) mixed with juniper (Juniperus oxycedrus L.). In some places juniper is dominating. This type of population is not as vulnerable to fires as pine populations but the fire has been encouraged by two favourable factors, the violent winds (sirocco) and a rich and very abundant herbaceous stratum (notably Aegylops ovata Eig., Cynosurus elegans Desf., Echinaria capitata (L.) Desf.).

Come then cantons, Arrhas, Afghan, Chaabet Khrouf, Hadjar Labiod, Thniet Sefra, Oued-guebala, Grouve, Makhrouze with three fires each. Among them, the Oued-Guebala canton is most sensible. Within 21 years more than 72 ha have been affected by fires. One of the major reason was the lack of accessibility (lack of forest roads) which made fire suppression extremely difficult.

All reports are in agreement concerning species affected by fire. As indicated by Le Houerou (1973) and Quezel (1980) Aleppo Pine (Pinus halepensis Mill.) is mainly affected, followed mostly by green oak (Quercus rotundifolia Lamk.) and (or) Juniper (Juniperus oxycedrus L.). According to Boudy (1955), the Algerian pine populations attract fire.

Vegetation fires are rarely ignited by natural causes. In Bou-Taleb, for instance, no single fire has been caused by lightning, the only possible natural fire cause. The activities of humans, either directly or indirectly, are exclusively causes of wildfires (Boudy 1952).

The statistical data for the 15-year period 1977 to 1991, for which we have almost complete information, show the importance of the unknown origin of fires. Unknown caused fires represent 82% of all cases; they consumed more than 409 ha, equivalent to 94% of the burned surface. Ranking net are fires caused by carelessness, mainly by smokers who throw glowing cigarettes or matches which act as embers in the wind. This category of causes is weak in Bou-Taleb by only representing 6% of all cases.

Another underlying cause of fire are conflicts between residents and forest authorities. Those residents and forest owners who had been penalised by forest rangers due to violation of laws, such as illegal grazing or wood cutting, start fires as an act of revenge. This category of causes is hidden in category unknown and most likely represent a considerable part.

Since our figures established for the domanial forest of Bou-Taleb for a period of 15 years are significant, we considered it useful to compare them with other, more ancient statistics.

The comparison of our present data to those recorded before the independence allows to reconstruct the long-term evolution of fire causes (Tab.3).

Tab.3. Comparison of fire causes (%) during the two observation periods1907-1957 and 1977-1991.

The first observation reveals that there is a big difference between the two periods. We have an increase at the level of percentages of the unknown origin fires and reduction at the level of those of fires caused by carelessness. Following explanations are possible:

The comparison of present fire cause statistics with earlier statistics reveal several tendencies (Tab.4). It seems that before independence the share of unknown cause is lower than at present: 37.7% versus 82%. On the other side the share of 52% of carelessly caused fires were much higher than at present (17.8%). However, the large fires are caused by arson. This category represents 7.8% of all fire causes before the independence versus 0% after. However, actually this category of causes nowadays would be more important if investigations had been conducted more rigorously by the concerned services; this fact explains the elevated percentage of unknown causes.

Tab.4. Vegetation fires in the forest massif of Bou-Taleb: causes, frequencies, and burned surfaces.
Source: Local fire reports.

Period	1907-1957 (51 years)				1977-1991 (15 years)
Fire Causes	Number	%	Surface burned		Number	%	Surface burned
Fire Causes	Number	%	(ha)	%	Number	%	(ha)	%
Negligence	40	51.9	1115.1241	7.9	8	19.8	25.4700	5.9
Unknown	29	37.7	543.8502	3.8	37	82.2	409.0831	94.1
Arson	6	7.8	12200.0000	86.2	-	-	-	-
Lightning	2	2.6	290.4500	2.1	-	-	-	-
Total *	77	100	13949.4243	100	45	100	434.5531	100

It is noteworthy to mention lighting-caused fires. The share before the independence was 2.6% of the total number of fires whereas they are absent to our time. One of the possible explanations is that climate changed from more humid to dry.

Before independence the occurrence of fires was distributed fairly balanced all over the year. Among the 77 fires recorded in Bou-Taleb only 36 were recorded in summer (46.7%), 31 into spring (40.3%) and only 10 (13%) in winter (Tab.5).

This distribution of fire occurrences all year round maybe explained by the role the domanial forest of Bou-Taleb played to meet the needs of the local population, especially during the extreme exploitation phase during World War II. According to Boudy (1955) 140,000 steers of green oak and pine and 8,300 m³ of pine wood were harvested at that time.

Period	1907-1957 (51 years)		1971-1991 (21 years)
Seasons	Number of fires	Percentage %	Number of fires	Percentage %
Summer	36	46.7	56	88.9
Winter	10	13.0	3	4.8
Spring	31	40.3	34	6.3
Total	77	100	63	100

Currently most reported fires in Bou-Taleb (88.9%) take place in the hot summer months June to September, a period during which the forest is frequently visited for grazing, collecting honey, and picking fruits. The remainder of fires takes place at the end of the spring and the beginning of the winter with 6.4% and 4.8% respectively of all cases.

Most fires start in July (41%), mainly during daytime. The concentration of fires is between 09:00h and 18:00h with 50 fires equalling 92.6% of the total number. The majority of fires (43, equalling 79.6%) occur between 11:00h and 17:00h. Only 4 fires (7.4%) occur at night between 21:00h and 08:00h in the morning. It must be assumed that fires started by carelessness start during daytime while night fires can be mainly arson fires (Ministry for Water Resources and Forestry, Algeria 1987).

Fires in forest and other vegetation are a common feature in Mediterranean ecosystems, including in the biogeographical region of Algeria, and have contributed to shape vegetation composition and model landscapes. With the modern increase of wildfire frequency a catastrophic dimension has been reached and it is time to think seriously about taking action to reduce the negative impacts of fire. The statistical approach presented in this paper contributes towards a better understanding of the role and causes of fires and will facilitate the search for solutions. As Effenterre (1990) points out, an analysis of the vegetation types affected by fire allows to determine the need for reforestation including the selection of species which are best adapted to a fire environment. Action must be taken.

Boudy, P. 1952. Guide forestier en Afrique du Nord. Ed. La maison rustique., 505 p. Paris.
Boudy, P. 1955. Economie forestičre nord-africaine. Tome IV. Description forestičre de l'Algérie et de la Tunisie. 483 p., Ed. Larose, Paris.
Le Houerou, H.N. 1973. Fire and vegetation in the Mediterranean basin. Proceedings Annual Tall Timbers Fire Ecology Conference 13, 237-277. Tall Timbers Research Station, Tallahassee, Florida.
Madoui, A. 1987. Rapport entre les facteurs du milieu et le risque d'incendie. Cas de la foręt domaniale de Bou-Taleb (Sétif). Mem. D.E.S., Inst. Biol., Univ. Sétif, 51 p. + annexes, notice explicative et 06 cartes au 1:20,000.
Madoui, A. 1988. History of the last ten years of fire in Algeria. Post-graduate report, 10p.+ App.
Madoui, A. 1995. Contribution ŕ l'étude de l'impact écologique des feux de foręts sur la végétation du massif forestier de Bou-Taleb (Sétif). Thčse de Magister, Inst. Bio. Univ. Sétif. 281 p.+Ann..
Ministry for Water Resources and Forestry, Algeria. 1987. Bilan des incendies de foręts en Algérie, 220 p.
Quezel, P. 1980. L'homme et la dégradation récente des foręts au Maghreb et au proche orient. Naturalia monspeliensia, N? Hors Série, 147-152.
Van Effenterre, C. 1990. Prévention des incendies de foręts. Statistique et politique. Rev. For. Fr. XLII, 20-32.

Year	Number of Fires	Burnt Surface (x 1000 ha)	Year	Number of Fires	Burnt Surface (x 1000 ha)
1876	120	55,172	1896	179	14,091
1877	134	40,538	1897	396	79,203
1878	164	8,156	1898	150	12,384
1879	218	17,663	1899	272	16,099
1880	137	20,881	1900	162	2,937
1881	244	169,056	1901	135	9,687
1882	130	4,018	1902	475	141,141
1883	148	2,464	1903	388	94,398
1884	147	3,232	1904	244	2,759
1885	285	51,569	1905	255	7,676
1886	288	14,043	1906	219	9,126
1887	395	53,714	1907	211	4,457
1888	311	14,788	1908	344	6,54
1889	309	17,807	1909	278	9,751
1890	202	23,165	1910	482	24,294
1891	393	45,924	1911	322	16,309
1892	409	135,574	1912	338	26,505
1893	398	47,757	1913	696	138,191
1894	308	100,89	1914	362	43,305
1895	250	32,907	1915	237	19,35

Year	Number of Fires	Burnt Surface (x 1000 ha)	Year	Number of Fires	Burnt Surface (x 1000 ha)
1963		3,9233	1978		41,551672
1964		9,38572	1979		15,66256
1965		50,624281	1980		26,944609
1966		2,50343	1981		17,361391
1967		49,56148	1982	638	9,38176
1968		14,54981	1983	990	221,367
1969		13,31432	1984	562	4,73184
1970		30,43867	1985	747	4,6683
1971		57,83518	1986	1170	21,53775
1972		4,09776	1987	1321	23,300359
1973		34,53025	1988	1146	27,757801
1974		11,0025	1989	595	3,23666
1975		37,331	1990	911	28,046529
1976		19,945289	1991	1189	13,17615
1977		43,947

Causes of forest fires	Period 1907-1957	Period 1977-1991
Natural causes Lightning	2.60 %
Negligence Smokers Hunters of honey Other	51.95 %	17.78 %
Arson	7.79 %
Unknown causes	37.66 %	82.22 %