Lands use dynamics and the role of war in the forest cover change in Marawi City, Philippines

J. Bio. & Env. Sci. 20 2 5 27 | Abdurrahman et al. RE RERE RESEARCH SEARCHSEARCH SEARCH PAPER PAPERPAPER PAPER OPEN ACCESS OPEN ACCESSOPEN ACCESS OPEN ACCESS Lands use dynamics and the role of war in the forest cover change in Marawi City, Philippines Abdulmadjid A. Abdurrahman *1 , Jaime Q. Guihawan 1 , Rylle Adriane Galvez 1 , Peter D. Suson 1 , Olive A. Amparado 2 , Wella Tiu-Tatil 1 1 Department of Environmental Science, School of Interdisciplinary Studi es, Mindanao State University, Iligan Institute of Technology, Iligan City, Lanao del Norte, Philippines 2Department of Biological Sciences, College of Science and Mathematics, Mindanao State University, Iligan Institute of Technology, Iligan City, Lanao del Norte, Philippines Article published on May 05, 2025 Key words: Land cover change, Armed conflict, Remote sensing, Post - war recovery, Deforestation, Environmental neglect Abstract The 2017 Marawi City Siege caused widespread destruction and displacement, along with significant environmental impacts, particularly on the forest ecosystems. This study assessed land cover changes in Marawi City, Philippines, associated with the 2017 armed conflict using Landsat satellite imagery from 2010, 2017, and 2024. Remote sensing analysis utilizing QGIS and Maximum Likelihood Classification generated six distinct land cover classes: forests, grassland, cropland, bare land, water bodies, and built-up areas. Results revealed significant post-conflict landscape transformations, with forest areas decreasing by 185.47 ha (-5.35%) between 2017 and 2024, while grassland declined by -778.43 ha (-21.72%). Simultaneously, agricultural lands and builtup areas increased substantially, with built-up areas expanding by 440.09 (+62.55%) by 2024. Accuracy assessment of the 2024 classification demonstrated 89.93% overall accuracy with a kappa coefficient of 0.86. Interviews with local environmental offices confirmed that the forced relocation of commercial and residential infrastructure due to the destruction of the city's economic center and increased illegal logging activities for livelihood purposes like charcoal production significantly contributed to these changes. The accelerated postconflict forest loss pattern in Marawi mirrors experiences documented in other post-conflict regions globally, where environmental concerns are often sidelined in favor of infrastructure development and economic recovery. These findings highlight the environmental vulnerabilities in conflict-affected areas and the necessity for sustained conservation efforts to promote ecological resilience and support post-conflict recovery. * Corresponding Author: Abdulmadjid A. Abdurrahman  abdulmadjid.abdurra[email protected]du.ph Journal of Biodiversity and Environmental Sciences (JBES) ISSN: 2220-6663 (Print) 2222-3045 (Online) Vol. 26, No. 5, p. 27-33, 2025 http://www.innspub.net J. Bio. & Env. Sci. 20 2 5 28 | Abdurrahman et al. Introduction Land-use decisions can often undergo sudden alteration in response to shocks, with armed conflicts being the most violent and globally prevalent (Baumann and Kuemmerle, 2016). While armed conflicts can lead to significant destruction of infrastructure and displacement of populations, their environmental impact is also a mounting global concern and can be long-lasting (Daiyoub et al., 2023). Studies have confirmed that armed conflicts and wars have a causative link to environmental degradation (Ebert, 2020). In conflict zones, environmental damage can be direct or indirect (Daiyoub et al., 2023), with deforestation as one primary representation of the environmental suffering caused by armed conflicts (Baumann and Kuemmerle, 2016; Lindholm and Kropp, 2019). Armed conflicts have also been shown to maintain complex relationships with ecosystem and land use change and have been responsible for some cases of increased deforestation rates (Baumann and Kuemmerle, 2016; Ebert, 2020). These conflicts lead to land cover change, a decline in biodiversity, and changes to the overall ecosystem services. In the context of Marawi City, Philippines, the siege in 2017 depicts an important narrative of the city's history. It caused widespread damage and displacement of its local population, significantly affecting the urban landscape and the surrounding forest ecosystems, potentially altering forest ecosystem dynamics in ways yet to be fully understood. Hence, this study aims to assess the changes in Marawi City's land cover associated with the siege. Materials and methods The study used land satellite images for land classification downloaded from United States Geological (USGS) Earth Explorer (https://earthexplorer.usgs.gov/) (Rwanga and Ndambuki, 2017). The satellite images taken from the website across 2010, 2017, and 2024 served as input to QGIS v. 3.34 software to generate land cover maps. The study also utilized interviews with concerned offices to further support the analysis. Description of study area Marawi City, perched at an estimated elevation of 716.1 m (~2,350 ft.) above sea level, is nestled in the region of the Bangsamoro Autonomous Region in Muslim Mindanao (BARMM), Southern Philippines (Fig. 1). The city has a land area of 87.55 sq km. Its population, as determined by the 2020 Census, was 207,010, which represents 17.32% of the total population of Lanao del Sur province or 4.70% of the overall population of the Bangsamoro Autonomous Region in Muslim Mindanao (PSA, 2020). Fig. 1. Location map of the study area - Marawi City, Lanao del Sur, Philippines generated using QGIS 3.34 J. Bio. & Env. Sci. 20 2 5 29 | Abdurrahman et al. Results Land cover classification Land cover maps for the years before the siege (2010), during the siege (2017), and after the siege (2024) were generated using satellite data. This includes Landsat 5 Thematic Mapper (TM) for 2010 and Landsat 8 Operational Land Imager/ Thermal Infrared Sensor (OLI/TIRS) for 2017 and 2024 via the USGS Earth Explorer website (Jia et al., 2014). Utilizing these data sets were essential to provide a foundation for robust analysis of the land cover changes for the years before, during, and after the siege. Table 1. Land cover classification categories and their descriptive characteristic Land cover class Description Forest Primarily covered by trees and woody vegetation Grassland Areas dominated by shrubs and grasses Cropland Includes areas utilized for cultivating crops Bare land Exposed soils that are devoid of vegetation Water bodies Streams, rivers, and lakes Built - up Areas characterized by artificial structures, such as buildings and roads Table 2. Confusion matrix and accuracy assessment of land cover classification – 2024 Matrix Forest Built - up Cropland Water bodies Grassland Bare land Total User’s accuracy (%) Producer’s accuracy (%) Forest 574 0 3 0 0 0 577 99.48 93.55 Built - up 0 81 0 0 0 0 81 100.00 100.00 Cropland 0 0 17 0 5 0 22 77.27 78.63 Water bodies 1 0 1 4 0 0 6 66.67 100.00 Grassland 6 0 4 0 66 0 76 86.84 87.41 Bare land 0 0 3 0 0 4 7 57.14 100.00 Total 581 81 28 4 71 4 769 Over all accuracy (%) 89.93 Kappa 0.86 After extracting and projecting the research region, a supervised classification procedure based on the Maximum Likelihood Classification (MLC) algorithm using the Semi-Automatic Plugin (SCP) in QGIS was utilized to generate six distinct classes: forests, grassland, cropland, bare land, water bodies, and built-up (Table 1). During the image classification process, training samples using SCP were generated for each class by delineating polygons that contain pixels representing a particular class. Accuracy assessment Ground truthing and Google Earth are the most commonly adopted accuracy assessment techniques as they relate the classified image data with accurate on-field data sources for the affirmation process (Tilahun, 2015). After the satellite images were classified, 50 random points were assigned. These random points were validated using on-site verifications and Google Earth to adapt validation approaches. The accuracy assessment evaluates the reliability of land use classification results. The minimum acceptable interpretation accuracy for land use and land cover classification should not fall below 80% (Anderson et al., 1976; Islami et al., 2022). The 2024 classification (Table 2) demonstrated 89.93% overall accuracy with a kappa coefficient of 0.86, achieving an "Excellent" rating showing high agreement (Dorhi et al., 2022), validating the reliability of the land cover classification. Only the 2024 land cover map was evaluated in terms of accuracy. The validation of the most recent classification was valid for the earlier dates when working with historical data series when data for the earlier dates is not available (Cabral et al., 2010; Vasconcelos et al., 2015; Cabral and Costa, 2017; Temudo et al., 2020). J. Bio. & Env. Sci. 20 2 5 30 | Abdurrahman et al. Fig. 2. Land use land cover map of Marawi City – (a) 2010, (b) 2017, (c) 2024 Land cover dynamics The data shows the land cover changes from each land cover class from the years before (2010) the siege, the year (2017) of the siege, and the postconflict year (2024). The 2010, 2017, ad 2024 land cover maps (Fig. 2. a, b and c) revealed that forest had the largest land cover area (Table 2), while water bodies had the smallest land cover area among all land cover classes. The result shows most of the land cover classes are steadily increasing from 2010 to 2017 (Table 3), except for grassland and bare land. The city’s central livelihood and major government and commercial infrastructures fueling the city’s economy are located in the southern part of the city, which can also explain why built-up areas and cropland areas are mostly concentrated and increasing in the southern part of the city, although an evident increase of cropland areas are starting to advance in the central part of the city during 2017. In 2024, however (Table 4), the forest have experienced a decrease, losing a total of 197.92 ha (-5.86%), while grassland areas also declined by 1,278.93 ha (-44.06%). This pattern of post-conflict forest loss mirrors the experience of the Zagros forest after the Iran-Iraq war, where deforestation intensified in the post-war periods, with losses of 2,066 ha during 1993-1998 (Heidarlou et al., 2020). Conversely, both cropland and built-up areas had increased by 1,179.24 ha (+110.83%) and 298.87 ha (+32.39%) respectively, which can be attributed from the population shift after the Marawi Siege. Table 3. Area changes in land cover areas in hectares from 2010 to 2024 Land cover 2010 2017 2024 Forest 3,069.26 3,378.16 3,180.24 Grassland 3,450.77 2,902.71 1,623.78 Cropland 921.62 1,064.04 2,243.28 Water bodies 65.4 60.86 68.33 Built - up 765.16 922.85 1,221.72 Bare land 132.75 76.34 67.61 a b c J. Bio. & Env. Sci. 20 2 5 31 | Abdurrahman et al. Table 4. Total Area changes in land cover areas (ha) and (%) from 2010 to 2024 Land cover class Change 2010 - 2017 (ha) Change 2010 - 2017 (%) Change 2017 - 2024 (ha) Change 2017 - 2024 (%) Forest 308.90 10.06% - 197.92 - 5.86% Grassland - 548.06 - 15.88% - 1,278.93 - 44.06% Cropland 142.42 15.45% 1,179.24 110.83% Water bodies - 4.54 - 6.94% 7.47 12.27% Built - up 157.69 20.61% 298.87 32.39% Bare land - 56.41 - 42.50% - 8.73 - 11.44% Discussion The changing trend in land covers shows a decreasing trend in grassland and forest areas during the post-war period from 2017 to 2024 while agricultural lands and built-up areas increase significantly during the post conflict. The rapid urban expansion and shifting cultivation can mainly be attributed with the population shift in the study area (Kisangga et al., 2024), recognizing the contribution of forced migration (Gbanie et al., 2018). Forkuor andand Cofie (2011) and Mansary et al. (2016) further support this result of the study. In most cases, post-conflict recovery prioritizes infrastructure development while environmental concerns are sidelined (Chan et al., 2019). The study's result agrees with this argument as the interview with the city’s Environment and Natural Resource Office (CENRO) states that one of the drivers of this transformation in the city is the forced relocation of commercial and residential infrastructures to the different parts of the city due to the destruction of the city's economic heart, resulting in a 298.87 ha (+32.39%) increase in built-up areas. This pattern is comparable to the increase of built-up areas observed in Sardasht from 18 km² to 34 km² during the postconflict period (Heidarlou et al., 2020) and during the post-war in the western area of Sierra Leone from 1976 to 2011, as their analysis shows a 54% decline in forest cover and 241% in the urban expansion (2342.0 ha) due to the increase of population in the area (Gbanie et al., 2018). Further, the data from the Community Environment, Natural Resources, and Energy Office, District 1 (CENREO-01) shows that there is an increase in illegal logging activities during the post-conflict period, mainly for livelihoods like charcoal making and conversion for agricultural purposes. A crucial generalization emerging from this and similar studies is that post-conflict recovery creates a critical "environmental vulnerability window" where accelerated resource exploitation coincides with weakened institutional capacity. Post-conflict recovery strategies therefore should include forest protection, conservation, and restoration while strengthening local institutional capacity. Conclusion The study documented significant land cover transformations in Marawi City following the 2017 siege, with substantial forest and grassland decline and an increase in built-up and agricultural areas. The 5.86% and 44.06% decrease in forest cover and grassland along with the 32.39% increase in built-up areas between 2017 and 2024 demonstrate the rapid land conversion occurring during the post-conflict period, driven by forced relocation of commercial and residential infrastructure due to displaced populations which caused increase illegal logging activities mainly for livelihood, and agricultural expansion. These findings align with global patterns observed in other post-conflict regions. Recommendation(s) Based on the findings, the study recommends that local government units establish dedicated environmental monitoring teams to assess forest resources regularly. Integrating community-based forest management approaches would empower local stakeholders in conservation efforts while providing alternative livelihood options to reduce pressure on forest resources, as recommended by the Community Environment, Natural Resources, and Energy Office, District 1 (CENREO-01). The city also needs localized expertise in sustainable forest management rather than depending solely on regional offices. Finally, a J. Bio. & Env. Sci. 20 2 5 32 | Abdurrahman et al. balance between post-conflict infrastructure development and environmental conservation should be established through strategic land use planning incorporating ecosystem services valuation in rehabilitation programs. Incorporate forest conservation and green space development into the city's long-term urban development plan. This would help mitigate the observed pattern of uncontrolled urban expansion and improve urban climate resilience. Future land cover studies should prioritize acquiring cloud-free satellite imagery to overcome the classification challenges encountered with cloud cover in Landsat data. Additionally, while Landsat imagery provides accessible data, supplementing with higher-resolution satellite imagery such as Sentinel-2 (10m) or PlanetScope (3m) would enable more accurate detection of small-scale forest changes and fragmentation patterns, ultimately improving land classification accuracy. References Anderson J, Hardy E, Roach J, Witmer R. 1976. A land use and land cover classification system for use with remote sensor data. U.S. Geological Survey Professional Paper 964, Washington: Government Printing Office. Baumann M, Kuemmerle T. 2016. The impacts of warfare and armed conflict on land systems. Journal of Land Use Science 11, 672–688. Cabral AIR, Costa FL. 2017. Land cover changes and landscape pattern dynamics in Senegal and Guinea Bissau borderland. Applied Geography 82, 115–128. Cabral AIR, Vasconcelos MJ, Oom D, Sardinha R. 2010. Spatial dynamics and quantification of deforestation in the central-plateau woodlands of Angola (1990–2009). Applied Geography 31, 1185–1193. Chan L, Ruwanpura K, Brown B. 2019. Environmental neglect: Other casualties of post-war infrastructure development. Geoforum 105, 63–66. Daiyoub A, Gelabert P, Saura-Mas S, VegaGarcia C. 2023. War and deforestation: Using remote sensing and machine learning to identify the war-induced deforestation in Syria 2010–2019. Land 12, 1509. Dorji D, Shukla SK, Wangchuk D, Dorji KT, Gyeltshen K, Chettri N, Aryal KR, Al Abdouli KM. 2022. Evaluation of land use/cover change and urban sprawling pattern using remote sensing and GIS: A case study in Thimphu, Bhutan. Civil Engineering and Architecture 10, 2572–2579. Ebert D. 2020. Effects of armed conflict on forest cover and vegetation seasonality in Cambodia. Ph.D. thesis, Department of Physical Geography and Ecosystem Services, Lund University, Sweden. Forkuor G, Cofie O. 2011. Dynamics of land-use and land-cover change in Freetown, Sierra Leone and its effects on urban and peri-urban agriculture—A remote sensing approach. International Journal of Remote Sensing 32, 1017–1037. Gbanie SP, Griffin AL, Thornton A. 2018. Impacts on the urban environment: Land cover change trajectories and landscape fragmentation in post-war Western Area, Sierra Leone. Remote Sensing 10, 129. Heidarlou HB, Shafiei AB, Erfanian M, Tayyebi A, Alijanpour A. 2020. Armed conflict and land-use changes: Insights from Iraq-Iran war in Zagros forests. Forest Policy and Economics 118, 102246. Islami, Tarigan S, Wahjunie E, Dasanto B. 2022. Accuracy assessment of land use change analysis using Google Earth in Sadar Watershed Mojokerto Regency. IOP Conference Series: Earth and Environmental Science 1055, 012016. J. Bio. & Env. Sci. 20 2 5 33 | Abdurrahman et al. Jia K, Wei X, Gu X, Yao Y, Xie X, Li B. 2014. Land cover classification using Landsat 8 Operational Land Imager data in Beijing, China. Geocarto International 29, 941–951. Kissanga R, Catarino L, Máguas C, Cabral AIR. 2024. Dynamics of land-cover change and characterization of charcoal production and trade in southwestern Angola. Remote Sensing Applications: Society and Environment 34, 101162. Landholm DM, Pradhan P, Kropp JP. 2019. Diverging forest land use dynamics induced by armed conflict across the tropics. Global Environmental Change 56, 86–94. Mansaray LR, Huang J, Kamara AA. 2016. Mapping deforestation and urban expansion in Freetown, Sierra Leone, from preto post-war economic recovery. Environmental Monitoring and Assessment 188, 470. Pawlik Ł, Šamonil P, Malik I, Kroh P, Ślęzak A, Daněk P. 2019. Geomorphic edge effects in response to abiotic and anthropogenic disturbances in forest ecosystems of the Gorce Mountains, Western Carpathians. CATENA 177, 134–148. PhilAtlas. Geographic information of Marawi City. Online available from https://www.philatlas.com/mindanao/barmm/lanaodel-sur/marawi.html Rwanga SS, Ndambuki JM. 2017. Accuracy assessment of land use/land cover classification using remote sensing and GIS. International Journal of Geosciences 8, 611–622. Temudo MP, Cabral AIR, Talhinhas P. 2020. Urban and rural household energy consumption and deforestation patterns in Zaire province, Northern Angola: A landscape approach. Applied Geography 119, 102207. Tilahun A. 2015. Accuracy assessment of land use land cover classification using Google Earth. American Journal of Environmental Protection 4, 193–198. Vasconcelos MJ, Cabral AIR, Melo JB, Pearson TRH, Pereira HA, Cassamá V, Yudelman T. 2015. Can blue carbon contribute to clean development in West Africa? The case of Guinea-Bissau. Mitigation and Adaptation Strategies for Global Change 20, 1361–1383. Wilson C. 2014. Spectral analysis of civil conflictinduced forced migration on land-use/land-cover change: The case of a primate and lower-ranked cities in Sierra Leone. International Journal of Remote Sensing 35, 1094–1125.