TY - JOUR
T1 - A Ramsar site catchment undergoing major land use/land cover change
T2 - Scenarios from elephant marsh, Malawi
AU - Makwinja, Rodgers
AU - Tesfamichael, Solomon G.
AU - Curtis, Christopher J.
N1 - Publisher Copyright:
© 2025 The Authors
PY - 2025/1
Y1 - 2025/1
N2 - Sustainable landscape planning in Ramsar site catchments requires data on the historical land use and land cover change (LULCC) and its drivers. Thus, this paper quantifies LULCC from 2002, 2011 and 2020 and its associated drivers. The Maximum Likelihood supervised classification was used to produce land use/land cover (LULC) categories for each year. The Bayesian hierarchical clustering approach (BHCA) and binary logistic regression (BLR) identified and ranked drivers. From 2002 to 2020, forest land exhibited a dramatic decline (25.83%/year) while bare/barren land increased noticeably (3.59%/year) during the same period. Other land use classes, such as marshes and water bodies declined at 8.23%/year and 2.64%/year, respectively. Shrub land, eroded land, irrigated land, cultivated land, and settlement increased at 1.76%/year, 1.15%/year, 1.72%/year, 1.9%/year, and 2.21%/year, suggesting that LULCC is indeed an emerging threat to the Elephant Marsh Ramsar site. Dynamic drivers such as browsing effects, forest fire, dry season farming, irrigation effects, deforestation, policy conflicts, weak governance systems, weak enforcement of registration, unplanned settlements, population growth, economic development, illiteracy levels, poverty levels and food insecurity had a coefficient mean score of >2.0 and were significant (P < 0.01), classifying them as higher risk. The coefficient mean scores for expansion of invasive species, charcoal production, infrastructure development, low household incomes, animal grazing, heat waves, drought, cyclone storms, unsustainable farming systems, lack of awareness, land tenure systems, brick making and sand mining, weak enforcement of registration, mining, water abstraction, strong winds and wetland vegetation-over-exploitation were <2.04 and significant at P < 0.05, suggesting risky drivers. However, all drivers interacted with LULCC at multiple scales to produce complex feedback effects, influencing landscape management policies. These findings are critical, because they act as a baseline for managers and policy makers to prioritize integrated land-use planning to safeguard Ramsar site integrity, and account for future environmental and economic uncertainties.
AB - Sustainable landscape planning in Ramsar site catchments requires data on the historical land use and land cover change (LULCC) and its drivers. Thus, this paper quantifies LULCC from 2002, 2011 and 2020 and its associated drivers. The Maximum Likelihood supervised classification was used to produce land use/land cover (LULC) categories for each year. The Bayesian hierarchical clustering approach (BHCA) and binary logistic regression (BLR) identified and ranked drivers. From 2002 to 2020, forest land exhibited a dramatic decline (25.83%/year) while bare/barren land increased noticeably (3.59%/year) during the same period. Other land use classes, such as marshes and water bodies declined at 8.23%/year and 2.64%/year, respectively. Shrub land, eroded land, irrigated land, cultivated land, and settlement increased at 1.76%/year, 1.15%/year, 1.72%/year, 1.9%/year, and 2.21%/year, suggesting that LULCC is indeed an emerging threat to the Elephant Marsh Ramsar site. Dynamic drivers such as browsing effects, forest fire, dry season farming, irrigation effects, deforestation, policy conflicts, weak governance systems, weak enforcement of registration, unplanned settlements, population growth, economic development, illiteracy levels, poverty levels and food insecurity had a coefficient mean score of >2.0 and were significant (P < 0.01), classifying them as higher risk. The coefficient mean scores for expansion of invasive species, charcoal production, infrastructure development, low household incomes, animal grazing, heat waves, drought, cyclone storms, unsustainable farming systems, lack of awareness, land tenure systems, brick making and sand mining, weak enforcement of registration, mining, water abstraction, strong winds and wetland vegetation-over-exploitation were <2.04 and significant at P < 0.05, suggesting risky drivers. However, all drivers interacted with LULCC at multiple scales to produce complex feedback effects, influencing landscape management policies. These findings are critical, because they act as a baseline for managers and policy makers to prioritize integrated land-use planning to safeguard Ramsar site integrity, and account for future environmental and economic uncertainties.
KW - Drivers
KW - Landscape planning
KW - LULCC
KW - Ramsar site
KW - Remote sensing
UR - http://www.scopus.com/inward/record.url?scp=86000735692&partnerID=8YFLogxK
U2 - 10.1016/j.rsase.2025.101508
DO - 10.1016/j.rsase.2025.101508
M3 - Article
AN - SCOPUS:86000735692
SN - 2352-9385
VL - 37
JO - Remote Sensing Applications: Society and Environment
JF - Remote Sensing Applications: Society and Environment
M1 - 101508
ER -