Forest Fire Situation in the State of Chiapas, Mexico
(IFFN No. 27 – July 2002, p. 21-33)
Fire Environment, fire regimes, and ecological role of fire in Chiapas
A forest fire is defined by the State law for fire prevention, fighting and control in the State of Chiapas (Poe, 1999), as: Aharmful event caused intentionally or fortuitously by fire, which occurs in areas covered by vegetation, trees, grasslands, scrubs, brushwoods and, in general, in every different plant association.
Chiapas is a southeastern Mexican state with a land area of 75,634 km2 (four percent of the national territory, the 8thbiggest state in Mexico (Inegi 1997) (Fig. 1). Sixty-eight percent of the State is forested. Sixty-four percent of the forested area is covered by woods (22percent) or jungle (42 percent) (Sarh 1994). Chiapas is divided into 111administrative units. During the period 1993-1999, 75 (68 percent) of them suffered from at least one forest fire (>0.5 ha), and 20 of them (18 percent)suffered from large forest fires ³500 ha. About 4.1 million hectares of forested land are subjected to forest fires (83 percent).
Forest fire regimes common to Chiapas can be summarized as follows:
Understory fire (applies to forests and woodlands)–fires are generally non-lethal to the dominant vegetation and do not substantially change the structure of the dominant vegetation. This situation would be the case forpine communities (and its mixed communities: pine-oak; pine-oak-liquid ambar) and oak communities. Fire does not always benefit genus Pinus, especially its most mesophyll species (in Chiapas case: Pinusayacahuite and P. pseudostrobus). Fire can be harmful to the establishment and the permanence of many pinewoods that, even if they are not destroyed, are prone to have regeneration difficulties (Miranda 1952, Rzedowski and McVaugh 1966 [in Rzedowski 1978]). Oaks are not pyric communities, especially in Chiapas, where they usually inhabit very humid places. This fire regime does not kill them, but it affects the quality of its habitat. Although it is not as frequent, when drought is a factor, there can be understory fires in wetter communities such as tropical deciduous forests, evergreen seasonal forests, montane rain forest, lower-montane rain forests, evergreen cloud forests, or even tropical rain forests (as in Breedlove, 1981). Repeated low-intensity fires in tropical communities, combined with low-intensity logging, increase future fire risk and drives to the savannization of the affected communities (Goldammer 1999, Cochrane 1999)
Stand replacement fire (applies to forests, woodlands, shrublands, and grasslands)–fires kill aboveground parts of the dominant vegetation, changing the aboveground structure substantially. Approximately 80percent or more of the aboveground dominant vegetation is either consumed ordies as a result of fires: This fire regime is not frequent in tree communities in the State, but when it occurs it usually has catastrophic connotations. Stand replacement fires are associated with severe drought conditions (which might be caused by global climatic phenomena like El Niño). Under these conditions, not only flammable forest types can burn (pines or its mixed communities), but even potentially non-flammable types like Abiescommunities, tropical rain forests, and evergreen cloud forests. This was the case with the 1998 forest fires. Stand replacement fires in tropical communities represent future fire-hazard (if there is no land use change) during the following decades, as unburned fuels remain there and dry out. Secondary succession can be initiated by opportunistic very flammable species, like ferns, that greatly increase fire-hazard.
Figure 1.Geographiclocation and administrative division of Chiapas, a southern Mexican state.
Mixed-severity fire (applies to forests and woodlands)– the severity of fire either causes selective mortality in dominant vegetation, depending on different tree species’ susceptibility to fire, or varies between understory and stand replacement: This regime can be typical forpine communities and their mixed variations (pine-oak; pine-oak-liquidambar).The result is a patchwork effect that rapidly turns green, but the habitatsquality is diminished. Moreover, those opened gaps are frequently occupied by very flammable, opportunistic species like ferns, so fire-hazard is increased during the decades after fire (Fig. 2).
Non-fire regime–little or no occurrence of natural fire: High humidity levels are responsible for this non-fire regime, under normal climatic conditions. In Chiapas, this generally would be the case for Abiescommunities, rain forests, and evergreen cloud forests.
Forest fires regimes in Chiapas have an important human component that can be summarized as the combination of different causes: structural and immediate causes (Vélez 1990):
1. Structural causes(permanent social and ecological conditions that cause the fire problem,
– Climaticand topographic characteristics: Chiapas, as other tropical states in Mexico, has a natural seasonal variation in rainfall, which is divided into a rainy season (May to October) and a dry season (November till April). Dry seasons also are characterized by the highest temperatures. Forest fires in Chiapas follow, more or less, this rainy scheme and they are concentrated in a part of the dry period (from January to May, or June, when there are special conditions of drought, affected by global climatic phenomena like ENSO). So it seems theres a tendency to extend the fire season to the first months of the rainy season, either because rain comes later, or because accumulated drought stress requires some time to recover. There are no recorded forest fires outside of these months. This implies a concentration of resources(human and material) in a specific part of the year. On the other hand, the complex topography of the state makes it difficult to attack fire, and steep topography makes fires spread faster (Deeming et al. 1977, Rothermel 1984).
Figure 2. Widespread establishment of bracken fern after a fire,
in Lagunas de Montebello NationalPark.
– Fire is habitually used as a traditional tool. Fire and agricultural/farming activities have always been very common in this region.
– A socio-economic framework that does not help conservation measures. Sixty percent of the population are rural inhabitants (INEGI 1997). There is high demographic pressure and the cattle sector is evolving.
– Existence of natural resources potentially transformable into agricultural fields. This creates migration of farmers from high demographic pressure regions to lower demographic pressure regions; as well as migration from neighboring states with a lower amount of potential lands for agricultural.
– Inefficientfire and land use change legal measures. (See section on Public Policies Affecting Wildfire Impacts).
Among the structural causes, El Niño must be taken into account as one of the major causes for the depletion of Chiapas forests. Figure 3 shows the arboreal burned surface affected from 1984 until 1999, versus the variation of the SOI index. The SOI is a measure of the disturbance produced by the climatic phenomenon El Niño, and in Chiapas, negative peaks of the SOI indicate severe affectations and hence, severe drought. This figure shows the strong relationship between severe El Niño and arboreal burned surfaces, which indicate the need for developing special campaigns for these years.
2. Immediate causes: they refer to human activities that provoke forest fires, in a direct or indirect way (Velez, 1990).
– 52 percent of the fires are due to negligence(agriculture or farming activities), while 29 percent are related to arson. Over17 percent of the fires are from unknown causes.
In terms of the affected area:
Pine-oak communities represent 74 percent of the fires, and 61 percent of the total area burned in the state. Fifty-four percent of the total area affecting pine-oak communities was due to surface fires, forty-two percent was affected by crown fires, and the rest were ground fires. Most common pines in the State are (Miranda 1975): Pinusoocarpa, P. pseudostrobus, P.montezumae, P. teocote, and P.tenuifolia (these live between 750-3000 meters and are the ones to suffer more frequent fires). P.hartwegii and P. rudis occur between 2800-4000 meters and dont frequently suffer from fires due to the cold, wet conditions.
Tropical rain forest and montane rain forest represented 7 percent of the incidences, and 18 percent of the total burned area in the State. Fifty-five percent of it’s total burned area was due to fires that affected crowns. Twenty-seven percent of its total burned area was due to ground fires and 17percent was related to surface fires.
Evergreen cloud forests represented 0.3 percent of the fires; and less than 1 percent of the total burned area in the State. Ninety-nine percent of the total area burned in the cloud forest was due to surface fires.
In general terms, forest fires in Chiapas are mainly surface fires (80 percent of the fires and 51 percent of the total burned area)(Román et al. 2000). High-intensity crown fires that severely affect the forests occur mainly in years of extreme drought, affected by the global phenomenon El Niño. This was the case of 1997-1998, or 1986-1987.
Figure 3. Variation of the arboreal burned area versus the oscillations of SOI from 1984 to 1999. In 1987, 72.255 ha were affected and 45.771ha in 1998. Medium point of the x-axis separations represent the years, starting in 1984. SOI: Southern Oscillation Index (Tahiti-Darwin) for the dry season. (November to April). SOI Resource: NOAA web, 2000 Note that the period analyzed in this graphic is longer than the one used for the statistics.
Narrative summary of major wildfire impacts on people, property, and natural resources during the 1990s
Forest fires in Chiapas do represent a problem in the state (Valera 1994). In 1987, Chiapas was considered, together with Madagascar, one of the hottest points in the world (State Government 1992, in Villafuerte et al. 1997). If we use standardized indexes that represent fire frequency and area affected by fires in Chiapas, compared with those indexes for the whole Republic, well see that fire frequency in Chiapas(Fig.4) used to have much higher levels during the 1980s. But during this last decade, it has decreased until it is lower than the national level. However, the area index has always been much higher in Chiapas than in the rest of Mexico. It was 20 times higher in 1986 and 8 times higher in 1998 (Fig.5). Mean burned area in Chiapas is more than 10 times the national level.
During 1993-1999, there were 1362 forest fires in Chiapas, affecting an area of 327,534 ha (SEMARNAP 1999). This represents 6 percent of Chiapas forested land. Twenty-six percent of that area included trees with different severities of burning. Forty-six percent of the area affected herbaceous layers. The mean area of each fire during these7 years was 210 ha.
The importance of forest fires in this southern State refers to the value of its vegetation. Mexico ranks third in the world in terms of megadiversity (Mittermeier 1988. Toledo and Eccardi 1989, Mittermeier and Mittermeier 1992, in Challenger,1998) . Among the States, Chiapas has, together with Oaxaca, the most diverse flora of the Republic (Miranda 1975). It presents 8,248 flora species, 80 percent of the tropical tree species, 33 percent of the reptiles (with important endemisms), 33percent of the amphibians, and 80 percent of the known butterflies in the Mexican tropical rain forests (State Government 1992).
Although forest fires mostly affect non-arboreal layers(74 percent of the total burned area during 1993-1999), 79,011 ha of arboreal communities were affected by several classes of intensity. Fifty-eight percent of the total arboreal layer affected by fires during these 7 years was due to 1998s forest fires (Román et al. 2000):
Surface fires were responsible for 50 percent of the arboreal layer affected. This type of fire (that reaches the lower part of the crown at some points) doesn’t necessarily kill the tree. But the trees can be weakened and become more vulnerable to pests. For example, Dendroctonus frontalis pests, in the National Park of Las Lagunas de Montebello, were enhanced by forest fires that happened there in 1997 and 1998. On the other hand, surface fires can severely affect trees that have not evolved in a fire environment (evergreen cloud forests, or tropical rain forests in their several categories).
The other 50 percent of the fires affecting the arboreal layer were crown fires, which were especially important in the years 1997 and 1998. In 1998, the drought conditions were severe due to El Nino. During this period, crown fires affected pine-oak communities, as well as potentially non-flammable communities like tropical rain forest, montane or lower montane rain forest, and evergreen cloud forests (as in Breedlove, 1981). Taking into account the State area occupied by these communities, pine-oak burned the most (3 percent of the State pine-oak communities, or 176,548 ha); tropical rain forest and montane/lower montane rain forest occupied the second position (1.5 percent of the State area, or 53,005 ha); and the evergreen cloud forests represented the smallest area affected (0.05% of the State area affected, or 2384 ha). Of special interest are two tropical rain forest areas that burned in the 1998 fires: Los Chimalapas and The Biosphere Reserve of El Ocote (corridor of Chimalapas, Ishiki, pers.communication, 1999).
Figure 4. Fire frequency index for the State of Chiapas and the Republic of Mexico, 1983-1999.
Figure 5. Burned area index for the State of Chiapas and the Republic of Mexico,1984-1999.
Chimalapas fires case example (García 1999, Figures 6 and 7): La Selva de los Chimalapas is 600,000 hectares located in the Tehuantepec isthmus. Part of its area is placed in Oaxaca and part in Chiapas. Its northern edge is bounded by Uxpanapa and Veracruz and its western boundary contacts the transisthmic road (carretera transístmica). In 1996, almost 80 percent of its area was in excellent conservation condition, consisting of 9 natural ecosystems (Breedlove 1981) (tropical rain forests, montane rain forests, evergreen seasonal forests, evergreen cloud forests, temperate forests (pine-oak), and even Abiesformations (oyameles). This biological diversity has given Chimalapas an exceptional ecological value. Moreover, this tropical formation, together with la Selva Lacandona and La Selva del Ocote, represent the northern location of this type of tropical community, which makes them even more valuable. These areas are found in the confluence of different biogeographical regions, which increases the number of endemisms.
Chimalapas, as is true of the rest of the tropical rain forests in Chiapas, suffers from several pressures: 1) agrarian conflicts and cattle invasion, 2) increase of the cattle area, 3) State legal border conflicts, 4) increase of narco-traffic settlements, and 5) new population settlements (high immigration rates and demographic pressure concentrated at some points). These pressures at areas of higher fire risk (associated with the higher flammability of temperate forest fuels) are the reason for several smaller forest fires at the eastern side.
Fires in Chimalapas started between the 4th and 6th of May in 1998 and burned for 45 days. These fires were characterized by surface, crown, and ground types of fires (SEMARNAP 1999). During the first 3 days there were 17 fire fronts in a horizontal line of 100 km, almost at the same latitude, affecting the heart of the tropical rain forest and the evergreen cloud forests. This is a very inaccessible location, where there are no villages, nor agrarian or cattle activities. The extreme drought associated with the El Niño event of 1997-1998 helped establish the conditions for these fires to spread. On the 7th of June when rain started, there had been 68 forest fires in the area; 17 of them in the heart of the jungle. Only 9 of the fires could be fought. The rest were inaccessible due to topographic reasons and were obscured by smoke. Four hundred farmers helped fight the fires initially, and the number grew to 1000 civil persons later. National and international help also assisted with the firefighting. Finally, the burned area was considered to be more than 100,000 ha, although SEMARNAP reports refer to 25,000 ha.
The SEMARNAP fire field register indicates that the fires burned 15,600 ha in the arboreal layers, 1400 ha in the shrub layers, and 5000 ha in the herbaceous communities (Román et al. 2000). After the fires, land use changes and new population settlements of “unlanded people” have put obstacles in the way of restoration processes (Ishiki, pers. Communication, 1999).
The fires of 1998 created a high “environmental”concern for the Chiapas population and its authorities due to several reasons:
Smoke produced by the fires covered the sky for almost one month (spreading to several areas of southern USA). Visibility was reduced, the environment was very warm, and combustion particles were responsible for severe breathing diseases in the whole State.
Several airports had to close their facilities as visibility didn’t allow the landing of planes.
Newspapers and radios were constantly communicating the occurrence of new fires in the State. So the chaotic use of fire and its consequences were more apparent during that year.
Potentially non-flammable communities were affected by fire, so the conviction of “ecological disaster” was a constant.
Severe drought occurred during the end of 1997 and the first 6 months of 1998, not only in forested areas, but also in agriculture lands. Farmers were concerned about lowered production.
Chiapas is a State with important tourism due to its Mayan archaeological sites and its natural resources. Fires that affect national parks and other tourism attractions can produce negative consequences on that sector’s income.
Figure 6. SEMARNAPs staff in the Chimalapas fire of98. Fire spread to the surrounding vegetation from a forested plot destinated to be transformed into agrarian land, which didnt use correct security measures. Courtesy of SEMARNAP Tuxla, 1998
The cause of fires is mainly associated (considering SEMARNAP’s fire field registers) with negligence arising from agrarian and farming activities. These two activities are responsible for 46 percent  of the fires and 66.2 percent of the total area affected by fires (Román et al. 2000). However, it is difficult to determine where negligence ends and intentional fires start. In Central America burns are sometimes done with deliberate negligence, letting fires affect much higher areas than the areas that will be cultivated. The objective is to weaken woods in order to obtain forestry cut authorizations or land use changes. This is due to the fact that legislation allows cuts to improve forest health, but restricts commercial cuts, for which several studies have to be presented, as well as paying a fee (Vélez 1986).
Figure 7.1998 forest fire in the Chimalapas jungle, a biological corridor of the Ocotejungle. 22,500 ha burned in this fire, among the three border states: Chiapas, Oaxaca and Veracruz. Courtesy of SEMARNAP Tuxla.1998
Figure 7. 1998 forest fire in the Chimalapas jungle, biological corridor of the Ocote jungle. 22,500 ha burned in this fire, among the three border states: Chiapas, Oaxaca and Veracruz. Courtesy of SEMARNAP Tuxla.1998
Public policies affecting wildfire impacts
There are several laws related to the problem of forest fires in the State. We can divide them into federal and State legislation:
Forestry and hunting legislation. December of 1992, reformed in May of 1997.
General law of ecological equilibrium and environmental conservation. January of 1988, reformed in December of 1996.
State (there are 9 documents. Some of the most important are):
Law for the prevention, fighting, and control of forest fires in the State of Chiapas. February of 1999.
State law of ecological equilibrium and environmental protection. July of 1991. Reformed by the decree law number 26, in February of 1992.
Decree law number 35: Declaration of municipalities in Chiapas with high fire risk. July of 1990.
Cattle law for the state of Chiapas.
Main points of these laws can be summarized as follows:
State legislation did not forbid burning for traditional agriculture and/or cattle activities (decree law number 26)
State laws did regulate the use of fire, forcing the establishment of several precautionary measures, in order to prevent forest fires when fire was used for traditional activities:
Decree law number 35:
– Creation of security corridors around the area meant to burn: minimum sizes, periods of the year when these corridors must be done, etc.
– Correct location of fuels meant to be burned, as well as several measures referred to fire behaviour in the presence of steep slopes and wind.
– Obligation of notifying the intention to burn to the local authorities and neighbours.
– Number of persons that must take care of the burning process, as well as the prohibition of starting fires in neighbouring field plots at the same time.
Federal legislation did consider and restrict land use changes, without prohibiting them in those areas that fitted the requirements established by the law (Forestry and hunting law).
Federal legislation did force civil participation in fighting and extinguishing fires, with administrative and/or penal sanctions:
Forestry and hunting law (administrative sanctions):
To whom, having the obligation, doesn’t prevent nor fight forest fires.
To whom, by negligence, provokes forest fires.
To those owners of burned areas that don’t notify the authorities of the presence of fire.
To whom, without authorization, changes land uses in forested land.
To whom, intentionally provokes forest fires (administrative or penal sanction).
Severity of forest fires in 1998 lead to the publication of the State law for the prevention, fighting, and control of forest fires in Chiapas, which:
Forbids “slash and burn” activities in those municipalities considered at high risk in the decree law 35 (46 percent of the municipalities in Chiapas), but it does not forbid the use of fire in any areas if it is for agriculture and/or cattle activities. (There are contradictions in those measures.)
Includes new type of sanctions, in addition to administrative and penal sanctions:
Authorities can decide to eliminate state “field support”, as federal support to any producer that does not take into account the legislation referred to forest fire prevention, fighting, and control.
Includes several incentives (preference support, prizes, and federal financing) to those who demonstrate several measures of environmental protection, among them:
Having avoided fire, traditional practices of “slash and burn”, or any other environmental impact.
Having reduced environmental impacts (deforestation, rational water use…).
Having promoted the culture of “no burning”.
Having introduced sustainable measures in their agriculture and/or cattle activities.
Creates actions destined to enhance the culture of “no burning” (educational measures).
Increases the services related to fire prevention and fire warning.
There are in Chiapas, several laws related to the problem of forest fires, but there are few measures to control their applicability. Moreover, measures proposed in some of this legislation depend on federal financing and there is the possibility this might not occur. On the other hand, fire has always been a traditional tool in Mayan culture and it fits into their natural and social conditions. Fire is not, “per se”, the problem, but it is the negligent use of fire together with a socio-economic framework that enhances environmental impacts. It’s impossible to control the use of fire in a State where approximately 60 percent of the inhabitants are rural population (Inegi 1997). This is especially true if alternatives are not proposed or are not useful. In this sense, fire legislation is quite removed from social reality in the State of Chiapas.
Together with the legislation, there are other activities in the state established to create some environmental concern about forest fires:
-SEMARNAP’s informational activities like speeches and courses.
-Advertising related to prevention of forest fires (wall paintings, radio messages, institutional calls, etc.).
Referring to civil participation in the State, there is a very satisfactory response to firefighting and detection activities. During 1993-1999 civil participation provided, on the average, 65 percent of the total number of firefighters. In several years (1993, 1994), civil firefighters comprised three times the number of official staff (Román et al. 2000). So, although negligence by the civil sector is mainly responsible for starting forest fires in Chiapas, it’s also the civil sector that mainly fights the fires.
Table 2. Wildfire statistics of Mexico, 1990-1999.
** This year includes 22,000 ha of Los Chimalapas forest fire, that should be divided into two parts, as this fire was located between Oaxaca and Chiapas.
IFFN/GFMC contribution submittedby:
Rosa Maria Roman Cuesta
CREAF. Unidad de Ecologia. Facultad de Ciencias
Universidad Autonoma de Barcelona
08193 Bellaterra, Barcelona
Alemán, T. 1997. La explotación del bosque en las regiones indígenas: sus aportes y perspectivas en la regeneración de alternativas de uso sostenido de los recursos naturales. In: Los Altos de Chiapas: Agricultura y Crisis rural. Tomo 1: Los recursos naturales. ECOSUR San Cristóbal, departamento de difución. Ed por Manuel Parra y Blanca Díaz
Breedlove, D. 1981. Introducción a la flora de Chiapas. Academia de Ciencias de California. Departamento de Botánica. Academia de Ciencias de California. Editor, Dennis Breedlove.
Challenger, A. 1998. Utilización y Conservación de los Ecosistemas Terrestres de México. Pasado, presente y futuro. Comisión Nacional para el Conocimiento y el Uso de la Biodiversidad, 813 p.
Cochrane et al. 1999. Positive feedbacks in the fire dynamic of closed canopy tropical forests. Science 284, 1832-1835.
Deeming, J. et al. 1977. The national fire danger rating system, 1978. Gen. Tech. Report. INT-39. USDA, FS, IFRES. Odgen, Utah, 63 p.
SEMARNAP. 1999. Informe final de la campaña de prevención y combate de incendios forestales en el estado de Chiapas. Temporada 1998-1999. SEMARNAP Tuxla.
Statal Government. 1992. Nueva legislación ecológica del estado de Chiapas. Talleres gráficos del Estado, Tuxla Gutiérrez, Chiapas.
Valera, J. 1994. Cifras sobre la deforestación de México. Inventario Nacional Forestal. SEMARNAP web.
Vélez, R. 1986. Incendios forestales y su relación con el medio rural. Revista de Estudios Agro-Sociales. Num. 136, (Julio-Septiembre 1986)
Vélez, R. 1990. Los incendios forestales en España. Ecología, Fuera de Serie nº 1, 213-221.
Villafuerte et al. 1997. La cuestión ganadera y la deforestación. Viejos y nuevos problemas en el trópico y en Chiapas. Universidad de ciencias y artes del estado de Chiapas & Gobierno del estado de Chiapas, 215 p.
This research was supported by the EU under the INCO-DC programme (framework4) as part of the SUCRE project (ERBIC-18-Oct97-0146)  We have to remark that this value corresponds to the total forested area of the administrative units that had fires, but overweighted, as fires are recurrent in many places.
Fire results are based on SEMARNAPs fire field registers.
Another consideration refers to the fact that, during these years (1993-1999), there has been an unusual succession of ENSOs: 1991-1992; 1993; 1994; 1997-1998 (NOAA website 2000). The last one (1997-1998) had a global repercussion and also affected Chiapas fire dynamic, by leading to the occurrence of fires in jungle communities, so area available to fire is overweighted compared to normal conditions.
 The relationship between fire and slash and burn activities is often confused in Chiapas. Slash and burn activities are agricultural systems which respect the natural cycles of vegetation in order to maintain productivity and land regeneration. The problem is that these systems require low demographic pressure (< 25 inhab/km2), as well as extensive forest resources that can be left in fallow long enough to recover (Sanders and Price 1968; Wolf 1983 [in Alemán 1997]). If not enough time is allowed for fallow in the agricultural fields, then slash and burn activities are being deprived of an important feature. There are very few places in the world, as well as in Mexico, were there are still the natural and socio-economic conditions that allow the preservation of slash and burn techniques. Instead we have what is known as milpa que camina, which is an irreversible, unidirectional, and rapid system for transforming woods into agricultural lands, with the exploitation of a few agricultural species (Alemán 1997).  Fire frequency index= Number of fires/ [Forested area (ha)/10.000]
Fire area index=[Total area affected by fires (ha)/ Forested area (ha)]*100
 SEMARNAP official data of fires during 1993-1999 were a bit higher than the data obtained with the fire field registers. (Published data were: 1362 forest fires and 327,534 affected hectares versus 1438 forest fires and 301,759 affected hectares reported in the fire field registers).  There are seven Megadiverse countries in the world and they support (altogether) between 50 and 80 percent of the world’s species diversity (Mittermeier 1988, The Economist 1988 [in Challenger 1998]).  Negligent acts are responsible of 52 percent of the fires. Negligent acts include several causes. The main ones are agriculture and farming activities, which represent 46 percent of the total fires. The other 6 percent is due to forestry activities, mountaineering, campsites, etc