Watching the Saltmarsh Grow: A High-Resolution Remote Sensing Approach to Quantify the Effects of Wetland Restoration

Abstract

Coastal wetlands are restored to regenerate lost ecosystem services. Accurate and frequent representations of the distribution and area of coastal wetland communities are critical for evaluating restoration success. Typically, such data are acquired through laborious, intensive and expensive field surveys or traditional remote sensing methods that can be erroneous. Recent advances in remote sensing techniques such as high-resolution sensors (<2 m resolution), object-based image analysis and shallow learning classifiers provide promising alternatives but have rarely been applied in a restoration context. We measured the changes to wetland communities at a 200 ha restoring coastal wetland in eastern Australia, using remotely sensed Worldview-2 imagery, object-based image analysis and random forest classification. Our approach used structural rasters (digital elevation and canopy height models) and a multi-temporal technique to distinguish between spectrally similar land cover. The accuracy of our land cover maps was high, with overall accuracies ranging between 91 and 95%, and this supported early detection of increases in the area of key ecosystems, including mixed she-oak and paperbark (10 ha), mangroves (0.91 ha) and saltmarsh (4.31 ha), over a 5-year monitoring period. Our approach provides coastal managers with an accurate and frequent method for quantifying early responses of coastal wetlands to restoration, which is essential for informing adaptive management in the regeneration of ecosystem services.

Restoration promotes ecological functioning through greater complementarity

“…Longfin eels and mud crabs dominated carrion consumption, and changes in the rate of this function were positively correlated with increases in their abundance. Longfin eels were dominant scavengers in areas where the area of mangroves and saltmarsh increased over time, whereas mud crabs were the primary consumers of carrion in locations with high connectivity to adjoining wetland habitats (Rummell et al 2022). These findings demonstrate the importance of functional complementarity and habitat diversity in recovering wetlands.…”

“…LCPA was calculated as the accumulative cost of moving through each cell of the RS from each sampling location at the Yandina Creek wetland to the nearest creek outside of the restoration site (see Supplement S1). The spatial extent of estuarine habitats at each site were quantified by adopting a highly accurate approach to measure the change in the area of mangrove, saltmarsh, and reed communities that include a high-resolution (0.5-2 m) Worldview 2 satellite imagery, object-based image analysis, and a random forest classifier (sensu Rummell et al 2022). The three-dimensional properties of each site were also quantified using digital elevation models (Geoscience Australia) by calculating the slope and elevation of each site (Table S2) (Borland et al 2021).…”

Fish and crustaceans provide early indicators of success in wetland restoration

“…2) Remote sensing approaches are commonly used methods to detect large‐scale shifts in the coverage and distribution of coastal and particularly blue carbon ecosystems (Pham et al 2019, 2019). We followed the remote sensing approach by Rummell et al (2022), which is designed to track the response of target ecosystems to wetland restoration. Briefly, this approach used a combination of multispectral and high spatial resolution Worldview‐2 imagery, an object‐based image analysis and random forest algorithm to classify estuarine land cover including, mangroves, saltmarsh, mangrove fern ( Acrostichum speciosum ), common reed ( Phragmites australis ), and water bodies.…”

“…These classes encompassed target ecosystems and areas of open water, at the restoration site and control wetlands, and variations in their distribution and area, which are often associated indicators of restoration success and with changes in the habitat function of wetlands for estuarine fish and crustaceans (Olds et al 2012; Zhao et al 2016; Gilby et al 2021). These maps were found to be highly accurate consistently achieving overall accuracies between 90 and 95% accuracy and were generated in QGIS after random forest classification via confusion matrices (Rummell et al 2022).…”

Application of Artificial Intelligence in the Evaluation of Positional Accuracy and Statistical Validation of Digital Elevation Models for Hydrological Studies

“…Digital elevation models (DEMs) are crucial for hydrology and water resources management. They provide a three-dimensional representation of the Earth's surface, which is essential for understanding and managing the flow and distribution of water, presenting several critical applications of DEMs in hydrology such as in watershed delineation [7][8] [9][8] [10] [11], hydrological modeling [12], flood risk analysis [13] [14], soil erosion and sediment transport [15] [16], river and stream network analysis [17] [18], rainfall-runoff analysis [19] [20], wetland mapping and analysis [21] [22], groundwater flow and recharge analysis [23] [24], irrigation planning and management [25], climate change impact assessment [10] [26].…”