The Delmarva Peninsula in the eastern United States is partially characterized by thousands of small, forested, depressional wetlands that are highly sensitive to weather variability and climate change, but provide critical ecosystem services. Due to the relatively small size of these depressional wetlands and their occurrence under forest canopy cover, it is very challenging to map their inundation status based on existing remote sensing data and traditional classification approaches. In this study, we applied a state-of-the-art U-Net semantic segmentation network to map forested wetland inundation in the Delmarva area by integrating leaf-off WorldView-3 (WV3) multispectral data with fine spatial resolution light detection and ranging (lidar) intensity and topographic data, including a digital elevation model (DEM) and topographic wetness index (TWI). Wetland inundation labels generated from lidar intensity were used for model training and validation. The wetland inundation map results were also validated using field data, and compared to the U.S. Fish and Wildlife Service National Wetlands Inventory (NWI) geospatial dataset and a random forest output from a previous study. Our results demonstrate that our deep learning model can accurately determine inundation status with an overall accuracy of 95% (Kappa = 0.90) compared to field data and high overlap (IoU = 70%) with lidar intensity-derived inundation labels. The integration of topographic metrics in deep learning models can improve the classification accuracy for depressional wetlands. This study highlights the great potential of deep learning models to improve the accuracy of wetland inundation maps through use of high-resolution optical and lidar remote sensing datasets.

Thermokarst lakes are a ubiquitous and important landscape feature in cold regions and are changing tremendously due to the accelerated climate change. In thermokarst lakes, sediment and water are two distinct but highly interconnected habitats, harboring different bacterial communities in terms of taxonomic composition. However, the co-occurrence networks of these bacterial communities remain unclear. Here, we investigate the co-occurrence ecological networks of sediment and water bacterial communities for thermokarst lakes in the Yellow River Source Area on the Qinghai-Tibet Plateau. The results show that the bacterial communities construct distinct co-occurrence networks in sediment and water. The metacommunity network was parsed into four major modules formed by the operational taxonomic units (OTUs) enriched in sediment or water independently, and water-enriched OTUs exhibited much closer interconnections than sediment-enriched OTUs. When considering the sediment and water bacterial networks separately, different topological properties and modular patterns present: the sediment bacterial network was more clustered while the modules less responded to the environmental variables. On the contrary, the water bacterial network was more complex with the OTUs more interconnected and its modules more responded to the environmental variables. Moreover, the results of the structural equation model suggest that, by the influence of environmental variations on individual modules, the water bacterial communities would be more vulnerable under the fact of accelerating climate change. This study provides insights beyond a conventional taxonomic perspective, adding our knowledge of the potential mechanisms structuring bacterial community assembly and improving our prediction of the responses of this fast-changing ecosystem to future climate change.

Thermokarst lakes are forming from permafrost thaw and are severely affected by accelerating climate change. Sediment and water in these lakes are distinct habitats but closely connected. However, our understanding of the differences and linkages between sediment and water in thermokarst lakes remains largely unknown, especially from the perspective of community assembly mechanisms. Here, we examined bacterial communities in sediment and water in thermokarst lakes in the Yellow River Source area, Qinghai-Tibet Plateau. Bacterial taxa were divided into abundant and rare according to their relative abundance, and the Sorensen dissimilarity (β_sor) was partitioned into turnover (β_turn) and nestedness (β_nest). The whole bacterial communities and the abundant and rare subcommunities differed substantially between sediment and water in taxonomical composition, α-diversity, and β-diversity. Sediment had significantly lower α-diversity indexes but higher β-diversity than water. In general, bacterial communities are predominantly governed by strong turnover processes (β_turn/β_sor ratio of 0.925). Bacterial communities in sediment had a significantly higher β_turn/β_sor ratio than in water. Abundant subcommunities were significantly lower in the β_turn/β_sor ratio compared with rare subcommunities. The results suggest that the bacterial communities of thermokarst lakes, especially rare subcommunities or particularly in sediment, might be strongly structured by heterogeneity in the source material, environmental filtering, and geographical isolation, leading to compositionally distinct communities. This integral study increased our current knowledge of thermokarst lakes, enhancing our understanding of the community assembly rules and ecosystem structures and processes of these rapidly changing and vulnerable ecosystems.

Landscape topography is an important driver of landscape distributions of soil properties and processes due to its impacts on gravity-driven overland and intrasoil lateral transport of water and nutrients. Rapid advancements in aerial, space, and geographic technologies have led to large scale availability of digital elevation models (DEMs), which have proven beneficial in a wide range of applications by providing detailed topographic information. In this report, we presented a summary of recent topography-based soil studies and reviewed five main groups of topographic models in geospatial analyses widely used for soil sciences. We then compared performances of two types of topography-based models—topographic principal component regression (TPCR) and TPCR-kriging (TPCR-Kr)—to ordinary kriging (OKr) models in mapping spatial patterns of soil organic carbon (SOC) density and redistribution (SR) rate. The TPCR and OKr models were calibrated at an agricultural field site that has been intensively sampled, and the TPCR and TPCR-Kr models were evaluated at another field of interest with two sampling transects. High-resolution topographic variables generated from light detection and ranging (LiDAR)-derived DEMs were used as inputs for the TPCR model building. Both TPCR and OKr models provided satisfactory results on SOC density and SR rate estimations during model calibration. The TPCR models successfully extrapolated soil parameters outside of the area in which the model was developed but tended to underestimate the range of observations. The TPCR-Kr models increased the accuracies of estimations due to the inclusion of residual kriging calculated from observations of transects for local correction. The results suggest that even with low sample intensives, the TPCR-Kr models can reduce estimation variances and provide higher accuracy than the TPCR models. The case study demonstrated the feasibility of using a combination of linear regression and spatial correlation analysis to localize a topographic model and to improve the accuracy of soil property predictions in different regions.

The water discharge and sediment load have been increasingly altered by climate change and human activities in recent decades. For the Pearl River, however, long-term variations in the sediment regime, especially in the last decade, remain poorly known. Here we updated knowledge of the temporal trends in the sediment regime of the Pearl River at annual, seasonal and monthly time scales from the 1950s to 2020. Results show that the annual sediment load and suspended sediment concentration (SSC) exhibited drastically decreased, regardless of water discharge. Compared with previous studies, we also found that sediment load and SSC reached a conspicuous peak in the 1980s, and showed a significant decline starting in the 2000s and 1990s, respectively. In the last decade, however, water discharge and sediment load showed slightly increasing trends. At the seasonal scale, the wet-season water discharge displays a decreasing trend, while the dry-season water discharge is increasing. At the monthly scale, the flood seasons in the North and East Rivers typically occur one month earlier than that in the West River due to the different precipitation regimes. Precipitation was responsible for the long-term change of discharge, while human activities (e.g. dam construction and land use change) exerted different effects on the variations in sediment load among different periods. Changes in the sediment regime have exerted substantial influences on downstream channel morphology and saltwater intrusion in the Greater Bay Area. Our study proposes a watershed-based solution, and provides scientific guidelines for the sustainable development of the Greater Bay Area.

Soil redistribution (erosion and deposition) can greatly affect the fate of soil organic carbon (SOC) in agroecosystems. Landscape topography is one of the key factors controlling erosion processes and creating spatial variability in SOC. We combined carbon (C) isoscape (isotopic landscape) analysis, historic orthophoto interpretation, cesium ( ¹³⁷ Cs) inventory measurement, and digital terrain analysis to quantify SOC dynamics and soil redistribution relationship and their responses to landscape topography in an Iowa cropland field with soybean/maize (C3/C4) rotation. The historic orthophotos and ¹³⁷ Cs were used to reflect soil redistribution before and after the 1960s, respectively. Topography‐based models were developed to simulate ¹³⁷ Cs inventory, SOC density, and C isotopes using stepwise principal component regression. Spatial patterns of SOC were similar to soil erosion/deposition patterns with high SOC density in depositional areas and low SOC density in eroded areas. Soil redistribution, SOC density, and isotopic signature of SOC (δ ¹³ C) were highly correlated with topographic metrics, suggesting that topographic heterogeneity drove the spatial variability in erosion and SOC dynamics. Considering the isotopic composition of SOC, C3‐derived SOC density was strongly controlled by topographic metrics, but C4‐derived SOC density showed weaker expression of spatial pattern and poor correlation to topographic parameters. The resulting topography‐based stepwise principal component regression models captured more than 60% of the variability in SOC density, δ ¹³ C _, and C3‐derived SOC density but could not reliably predict C4‐derived SOC density. Our results indicate that exploring C isotopes in response to soil erosion is important to understand the fate of eroded SOC within croplands under C3/C4 cultivation.