The mapping of ecosystem dynamics in African wetland landscapes within conservation areas typically relies solely on remote sensing approaches, potentially neglecting local perspectives and adaptations. This study presents an integrated socio-ecological systems (SES) framework applied to the endorheic Lake Chilwa Basin, Malawi, combining multi-temporal remote sensing analysis with participatory mapping methods to characterize wetland inundation dynamics and migratory fishing activity patterns.
Our methodology integrates Sentinel-1 InSAR processing with Landsat time series data (1994-2015) and community-based mapping approaches including key informant interviews, focus group discussions, and rapid participatory appraisals. The remote sensing component evaluates water-extraction indices (NDWI, MNDWI, AWEIsh) derived from multiple Landsat sensors (MSS, TM, ETM+, OLI) using spectral mixture analysis and soft classification techniques. Participatory methods capture social dimensions of landscape dynamics, particularly focusing on migrant fishing communities and seasonal resource use patterns.
Results reveal significant spatiotemporal variations in water levels and surface area, with major recession events documented in historical records (1879, 1900, 1914-15, 1922, 1931-32, 1934, 1954, 1960-61, 1967, 1973, 1995, 2012). The integrated approach enables refined mapping of ecosystem services at regional and local scales, revealing previously concealed spatiotemporal details of fishing regulations and enforcement conflicts within the lake’s political ecology.
This SES methodology provides a framework for future conservation initiatives in dynamic African wetland systems, advocating for locally grounded approaches that integrate biophysical and social dimensions for more responsive and effective conservation strategies.
This analysis presents an integrated remote sensing and participatory mapping approach to characterize lacustrine transgression-regression dynamics and associated socio-ecological patterns in the Lake Chilwa Basin, Malawi. The study addresses the limitations of purely technical remote sensing approaches by incorporating local knowledge and community perspectives into wetland ecosystem mapping.
Remote Sensing Analysis: Combine Sentinel-1 InSAR processing with Landsat time series data (1994-2015) to quantify hydroperiod fluctuations and littoral zone migration patterns using advanced spectral mixture analysis
Participatory Mapping Integration: Incorporate local ecological knowledge through key informant interviews, focus group discussions, and rapid participatory appraisals with migrant fishing communities
Socio-Ecological Systems Framework: Document spatiotemporal details of fishing regulations, population movements, trading routes, and resource use patterns previously concealed within the lake’s political ecology
Conservation Framework Development: Develop a transferable SES methodology for monitoring dynamic wetland ecosystems that balances technical precision with community-based knowledge systems
Historical records indicate major recession events occurred cyclically, providing crucial context for interpreting observed patterns within the lake’s documented century-scale variability. The 21-year observation period (1994-2015) represents a temporally constrained but methodologically comprehensive analysis that must be interpreted within longer-term ecological cycles.
tmap::tmap_mode("plot")
sf::sf_use_s2(FALSE)
aoi = read_sf("./inputs/chilwa_watershed_4326.shp")
bbox = terrainr::add_bbox_buffer(aoi, 10000, "meters")
# 'zoom' = resolution (higher than tm_basemap)
basemap = maptiles::get_tiles(
bbox,
zoom = 12,
crop = T,
provider = "OpenTopoMap"
)
tmap::tm_shape(basemap) + tm_rgb() +
tmap::tm_shape(aoi) +
tmap::tm_borders(lwd = 2, col = "red") +
tmap::tm_compass(position = c("right", "top")) +
tmap::tm_graticules(lines=T,labels.rot=c(0,90), lwd=0.2) +
tmap::tm_scalebar(c(0, 10, 20, 40), position = c("right", "bottom")) -> fieldmap
# tmap::tm_credits(maptiles::get_credit("OpenTopoMap"))
fieldmap
# width & height = res, dpi = size of add-ons
tmap::tmap_save(
fieldmap,
"./outputs/fieldmap-opentopo.png",
width=21600, height=21600, asp=0, dpi=2400)
aoi = read_sf("./inputs/chilwa_watershed_4326.shp")
sigma0_wet = raster::raster("./outputs/sar/Sigma_VV_db_slv_WET.tif")
sigma0_dry = raster::raster("./outputs/sar/Sigma_VV_db_mst_DRY.tif")
multi_temp = raster::stack(sigma0_wet, sigma0_dry)
bbox = terrainr::add_bbox_buffer(aoi, 40000, distance_unit = "meters")
vbox = ext(vect(bbox))
sites_locator <- st_as_sf(data.frame(
longitude = c(32, 36), latitude = c(-8,-16)),
coords = c("longitude", "latitude"), crs = 4326, agr = "constant")
malawi = ggplot(data = world) +
geom_sf(fill = "antiquewhite1") +
annotate(geom = "text", x = 35.5, y = -8.5, label = "Malawi",
color = "grey22", size = 4.5) +
coord_sf(xlim = c(32, 36.5), ylim = c(-8.5, -18)) +
xlab("Longitude")+ ylab("Latitude") +
theme(panel.grid.major = element_line(
colour = gray(0.5), linetype = "dashed", size = 0.5),
panel.background = element_rect(fill = "aliceblue"),
panel.border = element_rect(fill = NA))
chilwa <- ggplot(aoi) +
geom_sf() +
theme_void() +
theme(
panel.border = element_rect(fill = NA, colour = "black"),
plot.background = element_rect(fill = "antiquewhite1")
)
lake = ggplot(aoi) +
theme_void() +
geom_sf(lwd = 20, color = "red")
ggdraw() +
draw_plot(malawi) +
draw_plot(chilwa, height = 0.15, x = -0.05, y = 0.15) +
draw_plot(lake, height = 0.02, x = 0.1, y = 0.4)
plotRGB(multi_temp, r=2, g=1, b=1, stretch="lin") #stretch="hist"
tmap::tmap_mode("plot")
sf::sf_use_s2(FALSE)
# Load study area boundaries
aoi = read_sf("./inputs/chilwa_watershed_4326.shp")
bbox = terrainr::add_bbox_buffer(aoi, 10000, "meters")
# Generate basemap with topographic context
basemap = maptiles::get_tiles(
bbox,
zoom = 12,
crop = T,
provider = "OpenTopoMap"
)
# Create comprehensive study area map
fieldmap <- tmap::tm_shape(basemap) +
tmap::tm_rgb() +
tmap::tm_shape(aoi) +
tmap::tm_borders(lwd = 2, col = "red") +
tmap::tm_compass(position = c("right", "top")) +
tmap::tm_graticules(lines=T, labels.rot=c(0,90), lwd=0.2) +
tmap::tm_scalebar(c(0, 10, 20, 40), position = c("right", "bottom")) +
tmap::tm_layout(
title = "Lake Chilwa Basin Study Area",
title.position = c("left", "top"),
legend.position = c("left", "bottom")
)
fieldmap
# Generate regional context map
world <- ne_countries(scale = "medium", returnclass = "sf")
malawi_context <- ggplot(data = world) +
geom_sf(fill = "antiquewhite1") +
geom_sf(data = aoi, fill = "red", alpha = 0.7) +
annotate(geom = "text", x = 35.5, y = -8.5, label = "Malawi",
color = "grey22", size = 4.5) +
coord_sf(xlim = c(32, 36.5), ylim = c(-8.5, -18)) +
labs(
title = "Regional Context: Lake Chilwa Basin",
x = "Longitude",
y = "Latitude"
) +
theme_minimal() +
theme(
panel.grid.major = element_line(colour = gray(0.5), linetype = "dashed", size = 0.5),
panel.background = element_rect(fill = "aliceblue"),
panel.border = element_rect(fill = NA, colour = "black")
)
malawi_context
Lake Chilwa Basin represents one of Africa’s most productive yet dynamic endorheic ecosystems, characterized by extreme seasonal and inter-annual variability. Located in southern Malawi, this shallow terminal basin supports one of the continent’s most densely populated regions and exhibits remarkable ecological productivity during wet periods.
The basin features highly dynamic ecological, economic, and social landscapes driven by unimodal rainfall patterns (November-April) and sporadic “chiperone” rains (May-August). Annual fluctuations follow closely linked precipitation patterns, while longer-term cycles of approximately 15 years produce dramatic lake recessions and varying degrees of complete desiccation.
During recession periods, aquatic species take refuge in surrounding residual swamps dominated by salt-hardy vegetation (Typha domingensis Pers.). The subsequent refilling process initiates complex succession dynamics, with emergent food webs driven by detritus and bacterial processes in alkaline, nutrient-rich sediments.
Lake Chilwa’s remarkable productivity has been documented in peak years: 159kg ha⁻¹ (1979) and 113kg ha⁻¹ (1990), surpassing productivity levels of major African lakes (Lake Malawi: 40kg ha⁻¹; Lake Tanganyika: 90kg ha⁻¹; Lake Victoria: 116kg ha⁻¹). This boom-and-bust productivity pattern sustains complex socio-economic systems including permanent lakeshore communities and seasonal migrant fishing populations.