Quantifying mangrove carbon assimilation rates using UAV imagery.

Mangrove forests are recognized as one of the most effective ecosystems for storing carbon. In drylands, mangroves operate at the extremes of environmental gradients and, in many instances, offer one of the few opportunities for vegetation-based sequestering of carbon. Developing accurate and reproducible methods to map carbon assimilation in mangroves not only serves to inform efforts related to natural capital accounting, but can help to motivate their protection and preservation. Remote sensing offers a means to retrieve numerous vegetation traits, many of which can be related to plant biophysical or biochemical responses. The leaf area index (LAI) is routinely employed as a biophysical indicator of health and condition. Here, we apply a linear regression model to UAV-derived multispectral data to retrieve LAI across three mangrove sites located along the coastline of the Red Sea, with estimates producing an R2 of 0.72 when compared against ground-sampled LiCOR LAI-2200C LAI data. To explore the potential of monitoring carbon assimilation within these mangrove stands, the UAV-derived LAI estimates were combined with field-measured net photosynthesis rates from a LiCOR 6400/XT, providing a first estimate of carbon assimilation in dryland mangrove systems of approximately 3000 ton C km-2 yr-1. Overall, these results advance our understanding of carbon assimilation in dryland mangroves and provide a mechanism to quantify the carbon mitigation potential of mangrove reforestation efforts.

Mangrove distribution and afforestation potential in the Red Sea.

Mangrove ecosystems represent one of the most effective natural environments for fixing and storing carbon (C). Mangroves also offer significant co-benefits, serving as nurseries for marine species, providing nutrients and food to support marine ecosystems, and stabilizing coastlines from erosion and extreme events. Given these considerations, mangrove afforestation and associated C sequestration has gained considerable attention as a nature-based solution to climate adaptation (e.g., protect against more frequent storm surges) and mitigation (e.g. offsetting other C-producing activities). To advance our understanding and description of these important ecosystems, we leverage Landsat-8 and Sentinel-2 satellite data to provide a current assessment of mangrove extent within the Red Sea region and also explore the effect of spatial resolution on mapping accuracy. We establish that Sentinel-2 provides a more precise spatial record of extent and subsequently use these data together with a maximum entropy (MaxEnt) modeling approach to: i) map the distribution of Red Sea mangrove systems, and ii) identify potential areas for future afforestation. From these current and potential mangrove distribution maps, we then estimate the carbon sequestration rate for the Red Sea (as well as for each bordering country) using a meta-analysis of sequestration values surveyed from the available literature. For the mangrove classification, we obtained mapping accuracies of 98 %, with a total Red Sea mangrove extent estimated at approximately 175 km2. Based on the MaxEnt approach, which used soil physical and environmental variables to identify the key factors limiting mangrove growth and distribution, an area of nearly 410 km2 was identified for potential mangrove afforestation expansion. The factors constraining the potential distribution of mangroves were related to soil physical properties, likely reflecting the low sediment load and limited nutrient input of the Red Sea. The current rate of carbon sequestration was calculated as 1034.09 ± 180.53 Mg C yr-1, and the potential sequestration rate as 2424.49 ± 423.26 Mg C yr-1. While our results confirm the maintenance of a positive trend in mangrove growth over the last few decades, they also provide the upper bounds on above ground carbon sequestration potential for the Red Sea mangroves.

UAV RGB, thermal infrared and multispectral imagery used to investigate the control of terrain on the spatial distribution of dryland biocrust.

Biocrusts (topsoil communities formed by mosses, lichens, bacteria, fungi, algae, and cyanobacteria) are a key biotic component of dryland ecosystems. Whilst climate patterns control the distribution of biocrusts in drylands worldwide, terrain and soil attributes can influence biocrust distribution at landscape scale. Multi-source unmanned aerial vehicle (UAV) imagery was used to map and study biocrust ecology in a typical dryland ecosystem in central Spain. Red, green and blue (RGB) imagery was processed using structure-from-motion techniques to map terrain attributes related to microclimate and terrain stability. Multispectral imagery was used to produce accurate maps (accuracy > 80%) of dryland ecosystem components (vegetation, bare soil and biocrust composition). Finally, thermal infrared (TIR) and multispectral imagery was used to calculate the apparent thermal inertia (ATI) of soil and to evaluate how ATI was related to soil moisture (r 2 = 0.83). The relationship between soil properties and UAV-derived variables was first evaluated at the field plot level. Then, the maps obtained were used to explore the relationship between biocrusts and terrain attributes at ecosystem level through a redundancy analysis. The most significant variables that explain biocrust distribution are: ATI (34.4% of variance, F = 130.75; p < 0.001), Elevation (25.8%, F = 97.6; p < 0.001), and potential solar incoming radiation (PSIR) (52.9%, F = 200.1; p < 0.001). Differences were found between areas dominated by lichens and mosses. Lichen-dominated biocrusts were associated with areas with high slopes and low values of ATI, with soil characterized by a higher amount of soluble salts, and lower amount of organic carbon, total phosphorus (Ptot) and total nitrogen (Ntot). Biocrust-forming mosses dominated lower and moister areas, characterized by gentler slopes and higher values of ATI with soils with higher contents of organic carbon, Ptot and Ntot. This study shows the potential to use UAVs to improve our understanding of drylands and to evaluate the control that the terrain has on biocrust distribution.

Mask R-CNN and OBIA Fusion Improves the Segmentation of Scattered Vegetation in Very High-Resolution Optical Sensors.

Vegetation generally appears scattered in drylands. Its structure, composition and spatial patterns are key controls of biotic interactions, water, and nutrient cycles. Applying segmentation methods to very high-resolution images for monitoring changes in vegetation cover can provide relevant information for dryland conservation ecology. For this reason, improving segmentation methods and understanding the effect of spatial resolution on segmentation results is key to improve dryland vegetation monitoring. We explored and analyzed the accuracy of Object-Based Image Analysis (OBIA) and Mask Region-based Convolutional Neural Networks (Mask R-CNN) and the fusion of both methods in the segmentation of scattered vegetation in a dryland ecosystem. As a case study, we mapped Ziziphus lotus, the dominant shrub of a habitat of conservation priority in one of the driest areas of Europe. Our results show for the first time that the fusion of the results from OBIA and Mask R-CNN increases the accuracy of the segmentation of scattered shrubs up to 25% compared to both methods separately. Hence, by fusing OBIA and Mask R-CNNs on very high-resolution images, the improved segmentation accuracy of vegetation mapping would lead to more precise and sensitive monitoring of changes in biodiversity and ecosystem services in drylands.