ZAIDI Hachemi ¹*, AZZAOUI Mohamed Essalah², SEBTI Aissa ¹, BOUZENOUT Oussama ¹ KACHABIA El Khansa ³

¹ Higher School of Agronomy Mostaganem, Algeria; Biotechnology laboratory applied to agriculture and

environmental preservation

² Ibn Khaldoun University Tiaret, Algeria; Laboratoire d’Agrobiotechnologie et la Nutrition en Zones semi-Arides

³ Marine and coastal eco-biology laboratory, Badji Mokhtar University Annaba, Algeria

Received : 12/06/2025 ; Accepted : 26/10/2025

Abstract
Forest fires are a global issue with serious environmental, social, and economic consequences. In Algeria, more than 30,000 hectares of forest are burned annually. This study integrates data from the Chouachi Forest (W. Mostaganem) into a Geographic Information System (GIS) to analyze post-fire forest evolution following multiple fires between 2014 and 2020. The objective is to guide effective post-fire management strategies.

The methodology includes the creation of thematic maps using hypsometry, slope, exposure, topographic indices, NDVI (Normalized Difference Vegetation Index) for 2014, 2016, and 2020, NBR (Normalized Burn Ratio), normalized regeneration indices, and climatic data.

The resulting synthesis map serves as a decision-support tool for the management and regeneration of burned forests. This approach also enables the proposal of targeted strategies to promote post-fire forest recovery in Algeria.

Keywords: forests, post-fire evolution mapping, geographic information systems (GIS).

1. Introduction

Among all the threats to Mediterranean forests, fires are the most destructive. They degrade soil, alter landscapes, and jeopardize forest sustainability (Tir, 2016). In Algeria, forest fires are the leading cause of forest degradation, with 25,000 to 30,000 hectares affected annually (Bouregbi, 2014).

After a fire, new landscapes emerge, featuring surviving shrubs, damaged trees, and dense undergrowth rich in flammable species—conditions that can lead to further fires (Colin & Jappiot, 2001; Santiago Beltran, 2004).

In some regions, especially in western Algeria, vegetation regeneration has been nearly absent due to unfavorable conditions for natural recovery (Haddouche et al., 2004).

This study focuses on the Chouachi Forest, northwest of Mostaganem, which has experienced multiple fires in recent years. We analyzed forest stands affected in 2014, 2016, and 2020 to assess ecological response and vegetative recovery over time.

Thematic maps were produced using remote sensing and integrated into a GIS, allowing for the spatial analysis of fire impacts. Satellite imagery from pre- and post-fire periods, combined with field verification, enabled us to identify burned areas and calculate indices such as NDVI, NRI, and NBR. These changes reflect shifts in land cover and forest condition before and after fire events.

2. Study Area Description

The area studied in this research is located in the northeastern part of the Wilaya of Mostaganem, where the Chouachi Forest lies, approximately 42 km from the city center. Administratively, it falls under the jurisdiction of the Sidi Ali district. Covering an area of 623 hectares, the forest is dominated by Aleppo pine. It is crossed by National Road No. 59 and is divided into two main sectors: Dhar Chouachi and Hadjrat Merabitine. The administrative boundaries of the forest are as follows:

North: Sidi Ali municipality
East: Hadjadj municipality and Road No. 11
West: Sidi Ali municipality and Kerrada Dam
South: Belattar municipality

This research focuses on the dynamics of post-fire regeneration. Despite its relatively small size, the Chouachi Forest hosts substantial vegetative biomass, including large stands of Aleppo pine, mastic trees, and eucalyptus. The forest has been repeatedly affected by fires, especially between 2010 and 2020, resulting in several zones of post-fire regeneration. This allowed us to produce various maps illustrating changes and ecological evolution within the study area.

Figure 1. Geographic Location Map of the Chouachi Forest (Mostaganem).

2.1. Bioclimatology and Climate Summary
Several climatic parameters, such as precipitation and temperature, significantly influence post-fire regeneration. Therefore, it is essential to synthesize these elements in order to calculate various climatic indices.

The most commonly used indices for characterizing interannual variations in the Mediterranean hydro-climatic context include the continentality index, the summer drought index, the ombrothermic diagram, and the pluviothermic quotient developed by Emberger (Colmet-Daage, Malet et al., 2019).

2.2. Continentality Index
Also referred to as the mean extreme thermal amplitude, this index plays a critical role in vegetation distribution and climatological studies (Benabadji and Bouazza, 2000).

The thermal classification of climates proposed by Debrach (1953) is based on the difference between the average maximum and minimum monthly temperatures (M–m):

Insular climate: M – m < 15°C
Coastal climate: 15°C < M – m < 25°C
Semi-continental climate: 25°C < M – m < 35°C
Continental climate: M – m > 35°C

The recorded values for the Chouachi Forest over the period 1990–2020 are as follows:

This classifies the area as having a coastal climate.

2.3. Summer Drought Index

This simple index characterizes summer drought by calculating the ratio between summer precipitation (P.E) and the average maximum temperature of the hottest summer month.

In the Mediterranean region, the summer is considered dry when the index value (I) is less than 5 — which is the case in our study area (Lebourgeois and Piedallu, 2005).

I.e= P.E /M

Forest	Period	Summer rainfall « mm »	M°C	I,e
Chouachi	1990-2020	19,08	28,94	0,65

Summer rainfall (P.E): 19.08 mm

2.4. Ombrothermic Diagram

The ombrothermic diagram developed by Bagnouls and Gaussen offers an accessible and effective method for representing and comparing climatic patterns (Charre, 1997). The diagram indicates that the dry period extends from early May to mid-October — lasting at least five months — while the wet season spans from late October to late April, lasting approximately seven months.

During the dry season, vegetation regeneration patches often wither due to insufficient moisture.

Figure 2. Ombrothermic Diagram of the Study Area (1990–2020), Based on the Method of Bagnouls and Gaussen (1953)

2.5. Emberger’s Pluviothermic and Climatic Quotient

Emberger’s index is used to characterize and classify climates across various bioclimatic zones. Each zone is defined based on the minimum average temperature of the coldest month and its corresponding pluviothermic coefficient (Emberger, 1955; Daget, 1977).
The study area currently falls within the semi-arid bioclimatic zone.

Figure 3. Location of the forest in the study area on the EMBERGER climagram.

3. Materials and Methods

The approach adopted to meet the stated objectives—namely, assessing the condition of the forest after the fire using remote sensing techniques—can be summarized as follows:

The selection of satellite images best suited to our objectives. The satellite images used in this study come from the LANDSAT 8 OLI/TIRS and SENTINEL-2 satellites. The chosen images aim to observe the condition and dynamics of the Chouachi Forest and to gather information about the area prior to field visits, such as the location of regeneration patches, forest cover types, and vegetation density.
The analysis of these images also allows for the calculation of key vegetation indices.

Figure 4. Study area extracted from the 2014 Figure 5. Study area extracted from the 2021

LANDSAT 8 OLI image . SENTINEL-2A image .

3.1. Fire Severity Dynamics and Post-Fire Regeneration

Burned forest areas do not necessarily exhibit the same degree of degradation, due to the heterogeneity of environmental factors such as topography, vegetation, and weather conditions. Using satellite imagery, the physical changes caused by fire can be detected. In fact, several other useful pieces of information can also be extracted, such as the different severity classes and the vegetation recovery rate after the fire (Cherki and Gmira, 2013).

3.2. Fire Severity Degree Map – NBR

The Normalized Burn Ratio (NBR) is a modified version of the NDVI that uses bands 8 and 12, instead of bands 5 and 8 on the Sentinel-2 satellite. It is based on the mid-infrared (MIR) band, which highlights reflectance changes caused by fire.

The formula used to calculate NBR is:

In this study, the differenced NBR (dNBR) is calculated by subtracting the post-fire NBR from the pre-fire NBR. This allows for the measurement of burn severity and the observation of changes or the evolution of burned areas between two images acquired over the same zone.

The formula used to calculate the dNBR is: dNBR = NBR _(Pre-fire) – NBR _(Post-fire)

The dNBR values allow for determining the degree of severity for each fire event. Four severity classes were defined: None, Low, Moderate, and High

Table 1. Fire severity classes.

Severity / NBR	Code
None	1
Low	2
Moderate	3
High	4

3.3. Normalized Regeneration Index (NRI) Map

Disturbances affecting forest ecosystems after a fire can trigger several advanced degradation processes. Therefore, it is necessary to develop tools to monitor vegetation recovery and landscape health after fire (Marchetti, Ricotta et al., 1995).

The Normalized Regeneration Index (NRI), like the NDVI, is used to study the difference in chlorophyll activity before and after a fire. It is calculated using the following formula:

Where NDVI (Pre-fire) and NDVI (Post-fire) are the normalized vegetation indices measured before and after the fire.

The NRI ranges from 0 to 1; a value close to 1 indicates a very high rate of post-fire vegetation recovery. Both NDVI and NRI were used to study the spatiotemporal variation of post-fire regeneration (Cherki and Gmira, 2013). The following table shows the vegetation recovery rate after the fire.

Table 2. Post-fire vegetation regeneration rate.

3.4. Process for Creating the Change Detection Map

The approach adopted to meet the stated objectives aimed to assess the condition of the forest after the fire by using remote sensing techniques to obtain spatial knowledge of forest cover and burned areas (land use, vegetation cover, etc.). In addition, Geographic Information Systems (GIS) were used for analysis and modeling of various polygons. These tools made it possible to map the spatial distribution of different land cover classes for the final change detection map.

For this purpose, three satellite images were used:

LANDSAT 8 OLI, acquired on October 3, 2014
LANDSAT 8 OLI, acquired on November 25, 2016
SENTINEL-2, acquired on August 15, 2020

3.5. Class Identification

The ROIs (Regions of Interest) were identified and coded on the different image scenes, based on the following land cover classes:

Dense forest
Open forest
Dense matorral
Open matorral
Bare soil
Burned area

For each land cover class, ROIs (training plots) were selected as representative sites of the spectral characteristics of that class, allowing for the definition of spectral signatures.

3.6. Field Survey Mission and Floristic Data Collection

Vegetation survey sheets were prepared to gather specific information that satellite imagery cannot provide and to validate the pixel classes resulting from the classification process. The pixel classes to be verified in the field were identified using a subjective sampling method, based on environmental factors such as exposure, slope, and altitude, as well as the structure of the vegetation cover (dense or open).

Survey locations were selected based on floristic, structural, and site homogeneity. Each plot covered an area of 100 m², with the objective of characterizing each vegetation unit by identifying the dominant tall and low woody species. The number of sampling plots established in each forest section depended on the burned area and thefire severity,as indicated by theNormalized Burn Ratio (NBR).

Field missions were organized for both reconnaissance and validation of the remote sensing results. Each vegetation survey included:

A list of plant species (floristic inventory)
Ecological and structural parameters of the site

These field missions made it possible to:

Evaluate the quality of results obtained from unsupervised classifications
Assess the impact of human activities on observed changes and environmental degradation
Identify and define various vegetation formations and other thematic units based on their spectral responses in color composites
Characterize vegetation types based on their physiognomy, floristic composition, and post-fire evolution

A total of 24 sampling plots were carried out for the floristic inventory: 16 located in burned areas, and the remainder in unaffected or peripheral zones. The geographic coordinates of each site were recorded using Google Earth Pro and Android mobile applications, and were documented in the survey sheets.

4. Results and Discussion

4.1. description of the study area

The altitude values in the Chouachi Forest range from 220.8 m to 432.88 m. Based on the hypsometric map analysis, three main altitudinal classes were identified and distributed as follows:

A. 300–400 m: This class covers the majority of the study area, representing 71.1% of the total surface.

B. Below 300 m (<300 m): This class occupies 21.55% of the total area.

C. Above 400 m (>400 m):Largely concentrated in the central part of the forest, this class accounts for 7.35% of the total surface.

Figure 6. Hypsometric Map . Figure 7. Slope Map.

Figure 8. Exposure Map. Figure 9. Topomorphological Map.

The slope values in the Chouachi Forest range from 0.05% to 30.25% (equivalent to 0° to 74°). Based on the slope map analysis, the slope categories are distributed as follows:

Gentle slopes (0%–6%): Cover 31.95% of the study area.
Moderate slopes (6%–12%): Represent the largest portion, occupying 49.6% of the area.
Steep slopes (12%–24%): Account for 19.15% of the study area.
Very steep slopes (>24%): Found in highly rugged terrain, covering only 0.28% of the total area.

The exposure map classifies the terrain based on cardinal directions, revealing a nearly equal distribution among the four categories:

• Southern exposure dominates, covering 26.34% of the total area.
• Western exposure matches the southern exposure, also accounting for 26.34%.
• Eastern exposure covers 25.16% of the forest.

• Northern exposure is the least prevalent, representing 23.30% of the area.

The topomorphological map shows the following terrain distribution:
• Lower piedmont dominates the area, covering 79.21% of the total surface.

• Plains represent 11.21% of the study area.

• Upper piedmont covers 9.21%.

• Mountain areas account for only 0.01% of the total surface.

These variations reflect the geomorphological structure of the study area, which influences both fire spread and post-fire regeneration.

The topomorphological index map provides insight into the spatial distribution of terrain conditions that influence fire propagation potential across the study area.

The analysis indicates that:

• 42.88% of the total area exhibits favorable conditions for fire spread, making it the most dominant category.
• This is followed by the moderately favorable class, which accounts for 29.84% of the study area.
• In contrast, the less favorable and highly favorable zones combined represent only 27.27% of the total surface, indicating a relatively limited extent of areas at the extreme ends of the fire risk spectrum.

When describing the study area, we can see underscore the significant role of topographic features, elevation highlight, Exposure and slope distribution in modulating wildfire risk, with a large portion of the forest falling into terrain categories that potentially facilitate fire propagation.

NDVI Vegetation Indices Map

The NDVI values calculated for the Chouachi Forest range between -1 and +1. The lowest values, typically below 0 and shown in red, indicate a complete absence of chlorophyll activity. These areas correspond to zones affected by fire, national roads, or firebreak trenches.
The yellow color, which represents NDVI values near zero, generally corresponds to the network of forest tracks.

In contrast, green areas indicate vegetation, including stands unaffected by fire or regeneration patches, while darker green tones represent dense matorrals and maquis vegetation.

Figure 10. NDVI Vegetation Indices Map (2014-2016-2021).

The final change detection map was created by combining field data with NDVI values from different time periods, which allowed for the classification of various forest formations based on their presence and degree of vegetative cover.

The difference between pre-fire and post-fire NDVI values is essential for evaluating vegetation behavior after a fire event. This response depends on the floral composition and the density of the understory.

4.3. Fire Severity Dynamics and Post-Fire Regeneration Results

Satellite image analysis made it possible, on one hand, to assess fire severity through the Normalized Burn Ratio (NBR), and on the other hand, to estimate the vegetation recovery rate after the fire using the Normalized Regeneration Index (NRI).

The dNBR Severity Degree Map

Figure 11. dNBR Severity Degree Map.

According to the map, fire severity is particularly high in the southwestern part of the forest. Areas with low or no severity correspond to zones that were not affected by the 2020 fire.

The map is used during field monitoring missions to assess the different levels of vegetation degradation following the fire. It supports the selection of plot locations and helps determine the number of samples to collect for each severity class. Furthermore, it enables analysis of the fire’s impact on post-fire vegetation succession and assists in the decision-making process for restoring the affected areas.

4.5. Classification Results

4.8.1. Classification of Pre-Fire Imagery

A classification of forest cover—primarily based on area and density—was carried out on pre-fire imagery from 2014. This classification allowed for a clear interpretation and mapping of land use, revealing the dominance of woody vegetation classes, particularly dense forest and open forest.

An area of 228.83 ha, or 29.93% of the total Chouachi forest area, is classified as clear forest (10.61% of the total, covering 81.13 ha) and dense forest (19.31% of the total, covering 147.69 ha). In comparison, matorral vegetation represents about 44.19% of the total forest area. This indicates that the Chouachi forest is predominantly covered by both clear and dense matorrals, occupying 224.62 ha and 113.30 ha, respectively. The remaining land consists of bare soil, including the network of forest tracks and firebreaks, which accounts for 197.78 ha, or 25.86% of the total area. (See Table 3).

Figure 12. Land Use Map of 2014.

Table 3. Land Use Area Distribution in 2014.

Land Use Class	Area (ha)
*Clear Forest*	81,13
*Dense Forest*	147,69
*Clear Matorral*	224,62
*Dense Matorral*	113,30
*Bare Soil*	197,78

4.8.2. Classification of Post-Fire Imagery

Image Classification After Fire Impact

The interpretation of the 2021 land use map shows that forest cover was significantly altered following the fire. While the same vegetation classes as before are still identifiable, new classes have also emerged. There is a marked decrease in the area of each original class, indicating a substantial reduction in woody cover due to the fire’s impact on forest structure and composition.

Figure 13. Land Use Map of 2021.

The analysis of the 2021 land use map reveals the following distribution:

Dense forest covers 79.87 ha, representing 10.43% of the total forest area.
Open (clear) forest occupies 68.34 ha, or 8.92% of the total area.
Matorrals span 209.32 ha, divided into clear matorral (18.09%) and dense matorral (9.24%).
A new class, young dense forest, has appeared, covering 186.64 ha (or 24.38%). This emergence is attributed to the complete exposure of certain areas to the 2014 and 2016 fires, which triggered significant regeneration patches.
The final two classes are bare soil, covering 139.64 ha (18.24%), and the area affected by the 2020 fire, which extends across 81.67 ha.

Table 4. Land Use Area Distribution in 2021.

Land Use Class	Area (ha)
*Clear Forest*	68,34
*Dense Forest*	79,87
*Young Dense Forest*	186,64
*Clear Matorral*	138,52
*Dense Matorral*	70,79
*Bare Soil*	139,64
*Fire-Affected Area*	81,67

Detection and Creation of the Final Change Map

Based on the analysis of land use changes, the final change map was produced to highlight the differences in vegetation cover and land use between the pre-fire and post-fire conditions.

Figure 14. Final Change Map (2014–2021).

The results derived from the classifications make it possible to quantify land use and highlight the dynamics of forest cover across the two time points.
By comparing the data from the two land use maps, significant changes in vegetation have been identified within the study area. All land use classes recorded in 2014 are still present in 2021, with two additional classes appearing. Overall, due to the recurrence of multiple fires over the years, most natural forest formations have shown signs of regression.

Table 5. Vegetation Class Statistics for 2014 and 2021.

Class	Class 2014		Class 2021
Class	Area (ha)	Percentage (%)	Area (ha)	Percentage (%)
Clear Forest	81,13	10,61	68,34	8,92
Dense Forest	147,69	19,31	79,87	10,43
Young Dense Forest	0	0	186,64	24,38
Clear Matorral	224,62	29,37	138,52	18,09
Dense Matorral	113,30	17,85	70,79	9,24
Bare Soil	197,78	25,86	139,64	18,24
Fire-Affected Area	0	0	81,67	10,7

Figure 15. Distribution of Land Use Classes in 2014 and 2021.

From 2014 to 2021, vegetation classes underwent significant degradation, with the total area decreasing from 764.54 ha to 613.04 ha, as detailed below:

Dense forests were transformed into young dense forests, matorrals, bare soil, and fire-affected areas, with a decrease of 67.82 ha (a reduction of 45.92%). Dense forests represented the majority of the forest area in 2014.
Clear forests saw a reduction of approximately 12.78 ha (or 15.76%), with most of this loss attributed to the 2021 fire.
Matorrals also experienced a considerable reduction, from 337.92 ha to 209.32 ha, representing a decline of 38.06%.
A new class, young dense forest, emerged, covering 186.64 ha in 2021. This reflects natural regeneration in areas affected by the fires of 2014 and 2016, with species like Pinus halepensis and Phillyrea angustifolia now present.
According to the statistics (Table 5), the total area impacted by the 2021 fire is approximately 81.67 ha, including 26.58 ha of clear forest, 11.69 ha of dense forest, 22.39 ha of clear matorral, and 11.73 ha of dense matorral. The bare soil area also decreased, from 197.78 ha to 139.64 ha.

In the Chouachi forest, fire disturbances have disrupted the natural vegetative system, allowing for regeneration of the forest cover, though not necessarily identical to its pre-fire state. From a floristic perspective, the post-fire community is richer in species and differs significantly from the original one.

Several years after the fires, the burned areas show no signs of returning to their initial condition, either in terms of flora or vegetation strata. The appearance of species absent in unburned areas indicates active natural regeneration processes.

The fire did not entirely destroy the forest but caused a shift in its structure. For instance, Eucalyptus stands (Eucalyptus sp.), arranged in parallel strips, help prevent the spread of fire. Due to their fire resistance, they have remained largely unchanged. Likewise, the dense scrubland of Thuya (Tetraclinis articulata) in the southeastern part of the forest has evolved over time, as it was never affected by fire.

In contrast, most of the Pinus halepensis (Aleppo pine) forests that were impacted by the fire have been replaced by a greater richness of herbaceous species, including Pistacia lentiscus, Phillyrea angustifolia, Cistus monspeliensis, and Calicotome villosa.

Table 6. Results of Forest Change.

Land Use Change	Area (ha)
Degradation of Clear Forest to Fire-Affected Area	26,58
Degradation of Clear Forest to Young Dense Forest	14,88
Progression of Clear Forest to Dense Forest	12,84
Degradation of Clear Forest to Clear Matorral	14,10
Degradation of Clear Forest to Dense Matorral	3,89
No Change in Clear Forest	6,75
Degradation of Dense Forest to Bare Soil	3,09
Degradation of Dense Forest to Fire-Affected Area	11,69
Degradation of Dense Forest to Young Dense Forest	59,34
No Change in Dense Forest	35,83
Degradation of Dense Forest to Clear Matorral	32,43
Degradation of Dense Forest to Dense Matorral	3,51
No Change in Bare Soil	106,36
Degradation of Bare Soil to Fire-Affected Area	9,03
Progression of Bare Soil to Young Dense Forest	19,43
Progression of Bare Soil to Dense Forest	5,56
Progression of Bare Soil to Clear Matorral	28,47
Progression of Bare Soil to Dense Matorral	11,45
Progression of Bare Soil to Clear Forest	13,11
Degradation of Clear Matorral to Bare Soil	19,76
Degradation of Clear Matorral to Fire-Affected Area	22,39
Degradation of Clear Matorral to Young Dense Forest	55,92
Progression of Clear Matorral to Dense Forest	12,10
No Change in Clear Matorral	36,64
Progression of Clear Matorral to Dense Matorral	40,10
Progression of Clear Matorral to Clear Forest	37,17
Degradation of Dense Matorral to Bare Soil	4,35
Degradation of Dense Matorral to Fire-Affected Area	11,74
Degradation of Dense Matorral to Young Dense Forest	36,91
Progression of Dense Matorral to Dense Forest	13,49
Degradation of Dense Matorral to Clear Matorral	26,10
No Change in Dense Matorral	11,66

Diachronic Study Between 2014 and 2021
A detailed mapping based on satellite data, supported by a Geographic Information System (GIS), enables the evaluation of the state of the Chouachi forest. This approach clearly reveals whether the forest heritage has undergone progressive changes, regression, or remained stable following the major fires of 2014, 2016, and 2020.

Figure 16. Map of the Evolution of the Chouachi Forest.

that the vegetation cover has evolved positively, with slow and modest progression, particularly in the southwestern parts of the forest that have never been affected by fires. The total area of progression is 217.99 ha, representing 28.68% of the total forest area.

Causes of Progression

Among the causes of this progression are the natural post-fire regeneration of certain species, the reforestation efforts implemented by the General Directorate of Forests (DGF) in some areas of the forest, and the development of road networks and firebreaks.

Areas of Regression

The results of our diachronic study between 2014 and 2020 show that the vegetation cover has evolved negatively, with significant degradation, particularly evident in certain zones. The total regressed area is 357.52 ha, representing 47.19% of the total forest area.

Following the cartographic analysis, field exploration, and satellite image processing, we identified the following factors contributing to forest degradation:

Figure 17. Map of Regression of the Chouachi ForestLes Incendies.

Table 7. Fire Incidents and Burned Areas in the Chouachi Forest (2009–2020).

Fire Dates	Burned Area
25/05/2009	0.03 ha
09/11/2009	0.02 ha
17/08/2011	0.0003 ha
05/09/2011	0.07 ha
25/09/2013	0.02 ha
29/10/2013	0.03 ha
12/07/2014	0.03 ha
25/07/2014	0.04 ha
29/07/2014	0.01 ha
11/08/2014	2 ha
12/08/2014	2 ha
22/08/2014	0.02 ha
26/08/2014	0.04 ha
26/08/2014 to 30/08/2014	70 ha
05/09/2014	0.0004 ha
16/09/2014	0.005 ha
26/06/2015	0.02 ha
16/07/2015	0.02 ha
02/08/2015	0.04 ha
30/08/2015	0.05 ha
10/08/2016	0.50 ha
29/08/2016	0.015 ha
14/09/2016	0.0008 ha
08/10/2016	0.003 ha
24/10/2016	60 ha
27/06/2017	0.02 ha
23/07/2017	0.0001 ha
09/08/2017	0.40 ha
05/11/2019	0.10 ha
06/08/2019	0.30 ha
07/07/2020	0.02 ha
13/07/2020 to 14/07/2020	3 ha
10/08/2020 to 12/08/2020	40 ha
14/08/2020	0.01 ha

The total area burned during the 2014–2020 decade amounted to 176.81 hectares, resulting from 34 fires, with an average of 16.07 hectares per year. The distribution of these fires was uneven across the forest. The year 2014 recorded the highest burned area, with 70.12 hectares, while 2009 had the smallest, with only 5 ares.

Of the 34 fires, 14 occurred in August, representing 41.17% of all incidents. This is understandable, as August follows a relatively long dry season (according to the ombrothermic curve), making the vegetation more flammable due to dryness.

Overgrazing

Overgrazing remains one of the most critical degrading factors. In the case of the Chouachi Forest, it has led to progressive degradation over several decades, affecting a variety of plant species and worsening soil erosion.

Poor Management

The absence of an effective strategy to adapt to new ecological conditions and to combat anthropogenic (human-caused) pressures continues to hinder forest preservation.

Areas with No Change

Most of the unchanged areas are either unaffected by fires or consist of roads, tracks, and firebreaks. This class covers 182.02 hectares, representing 24.13% of the total forest area.

4.4. Post-Fire Regeneration Dynamics

Figure 18. Normalized Regeneration Index Map.

We observed that regeneration is heterogeneous across the entire study area. Some locations show strong regeneration, while others remain weak, with very high regeneration immediately after the fire that gradually diminishes over time.

Vegetation regrowth after the fire is more pronounced in the southwestern part of the forest, which suffered severe degradation. This suggests a direct correlation between regeneration rate and fire severity. Areas with moderate or slightly weak regeneration correspond to the previously burned zones from 2014 and 2016.

Overall, sampling results show that the natural regeneration of Aleppo pine stands after fire in the study area is not satisfactory. The density of regenerated Aleppo pine in the Chouachi forest is low and spatially uneven throughout the burned area. It varies depending on soil type and is hindered by the presence of floristic communities that block sunlight from reaching the seedlings, thereby inhibiting their growth.

We also observed that the presence of litter (plant debris) on the soil surface promotes better vegetative regrowth in areas where the fire only affected the crowns. This litter acts as thermal protection, unlike in areas that experienced two fires in the last six years, where no Aleppo pine seedlings are present due to severe soil degradation.

The regenerated Aleppo pine seedlings often dry out because of competition from other species that develop quickly and take over the original vegetation.

It is also important to note that the presence of burned Aleppo pine stumps indicates multiple regeneration patches, meaning that post-fire recovery depends on the pre-fire pine density.

Figure 19. a) Area Affected by Fire Twice; b) Burned Aleppo Pine Stump

Among the species that proliferated after the fire were sage-leaved cistus (Cistus salvifolius) and Montpellier cistus (Cistus monspeliensis). These are indicator species of forest degradation and regenerate easily through seed dispersal.

The Eucalyptus forest was severely affected by the fire. However, Eucalyptus possesses a strong regenerative capacity after fire by producing suckers. One year after the fire, the Eucalyptus crowns had turned green again due to the development of aerial shoots.

Several species reappeared, including basal suckers from Phillyrea angustifolia, Phillyrea latifolia, Erica arborea, Pistacia lentiscus, and Quercus coccifera. These suckers reached heights of 50 cm to one meter. This rapid vertical growth is related to the increased light availability in the forest following the destruction of the entire reforested stand.

Conclusion

The forests of the Mostaganem region are also threatened by wildfires, due to the geographical location and the type of climate, both of which favor forest fire occurrences.

The use of various GIS tools, remote sensing techniques, and field observations led to the creation of a change detection map of the forested environment before and after the fire. This map was integrated with other datasets such as slope maps, aspect maps, and vegetation index maps within a Geographic Information System, in order to assess the vegetation recovery status in burned areas.

The results obtained from this natural disturbance (fire) indicate a major transformation of large portions of the Chouachi forest into open shrubland, young dense forest, and burned zones. The fire had a severe impact on vegetation, causing a deep alteration in floristic composition and structure. A full recovery to pre-fire conditions was not observed. However, some pre-fire species reappeared in the first year post-fire, as they possess adaptive traits to fire disturbances.

The regeneration of Aleppo pine was not homogeneous across the entire area. Some zones exhibited strong regeneration, while others showed weak vegetative recovery. The Normalized Regeneration Index (NRI) helped identify these different zones along with their respective regeneration rates.

Looking ahead, it is essential to protect the environment from recurring fires and other destructive factors, and to encourage natural regeneration by thinning seedlings (to reduce negative competition from understory vegetation) or by reforesting areas with low post-fire regeneration. Both planting density and species selection are key. According to vegetation surveys conducted during fieldwork, the current density is 1,000 trees per hectare. For species selection, a mixture of broadleaf and coniferous species—such as Aleppo pine and Eucalyptus—is less vulnerable to fire than pure coniferous stands.

Bibliographical References

Bagnouls, F., & Gaussen, H. (1953). Biological climates and their classification. Annales de Géographie. JSTOR.

Benabadji, N., & Bouazza, M. J. R. E. R. (2000). Some climatic changes observed in the southwest of Oran (Western Algeria), 3, 117–125.

Bouregbi, I. (2014). Causes and consequences of forest fires on cork production in the cork oak forests of northeastern Algeria.

Charre, J. (1997). Draw me a climate. What to think of the ombrothermic diagram, 2(97), 29–31.

Cherki, K., & Gmira, N. J. R. d. é. (2013). Post-fire regeneration dynamics and fire severity in Mediterranean forests: Case of Maâmora forest, northern Morocco.

Colin, P.-Y., & Jappiot, M. (2001). Forest fire protection: Technical sheets for Mediterranean basin countries. Food & Agriculture Organization.

Colmet-Daage, R., et al. (2019). Climate change and sub-regional meteorological data in the Western Mediterranean.

Floret, C., & Pontanier, R. (1982). Aridity in presaharan Tunisia: Climate, soil, vegetation, and land management. Office de la Recherche Scientifique et Technique Outre-Mer (ORSTOM).

Haddouche, I., et al. (2004). Use of remote sensing to study deforestation: Case of the Djelfa region. In Mediterranean Colloquium on Sustainable Management of Mountainous Areas. Dept. of Earth Sciences and Agronomy, University of Tlemcen.

Imberger, L. (1955). A biogeographical classification of climates.

Le Houérou, H. N. (1975). The vegetation of North Africa. Centre National de la Recherche Scientifique (CNRS).

Lebourgeois, F., & Piedallu, C. J. R. f. f. (2005). Understanding drought levels in site studies and forest management using bioclimatic indices.

Marchetti, M., et al. (1995). A qualitative approach to mapping post-fire regrowth in Mediterranean vegetation using Landsat TM data, 16(13), 2487–2494.

Monjauze, A., & Le Floc’h, E. (1983). Sahelian vegetation. Centre National de la Recherche Scientifique (CNRS).

Quézel, P. (1985). Degradation of North African forests. Options Méditerranéennes.

Santiago Beltran, R. J. L. (2004). Silvicultural recommendations for cork oak forests affected by fire, 15–27.

Tir, E. (2016). Spatial analysis and mapping of post-fire forest regeneration in the Wilaya of Tissemsilt.

Tracking Forest Evolution After Fires Using GIS Adaptationthe case of the forest Sidi Ali in the Mostaganem province, Algeria