Global mangrove soil organic carbon stocks dataset at 30 m resolution for the year 2020 based on spatiotemporal predictive machine learning.

This dataset presents global soil organic carbon stocks in mangrove forests at 30 m resolution, predicted for 2020. We used spatiotemporal ensemble machine learning to produce predictions of soil organic carbon content and bulk density (BD) to 1 m soil depth, which were then aggregated to calculate soil organic carbon stocks. This was done by using training data points of both SOC (%) and BD in mangroves from a global dataset and from recently published studies, and globally consistent predictive covariate layers. A total of 10,331 soil samples were validated to have SOC (%) measurements and were used for predictive soil mapping. We used time-series remote sensing data specific to time periods when the training data were sampled, as well as long-term (static) layers to train an ensemble of machine learning model. Ensemble models were used to improve performance, robustness and unbiasedness as opposed to just using one learner. In addition, we performed spatial cross-validation by using spatial blocking of training data points to assess model performance. We predicted SOC stocks for the 2020 time period and applied them to a 2020 mangrove extent map, presenting both mean predictions and prediction intervals to represent the uncertainty around our predictions. Predictions are available for download under CC-BY license from 10.5281/zenodo.7729491 and also as Cloud-Optimized GeoTIFFs (global mosaics).

Bird community shifts associated with saltwater exposure in coastal forests at the leading edge of rising sea level.

Rising sea levels dramatically alter the vegetation composition and structure of coastal ecosystems. However, the implications of these changes for coastal wildlife are poorly understood. We aimed to quantify responses of avian communities to forest change (i.e., ghost forests) in a low-lying coastal region highly vulnerable to rising sea level. We conducted point counts to sample avian communities at 156 forested points in eastern North Carolina, USA in 2013-2015. We modelled avian community composition using a multi-species hierarchical occupancy model and used metrics of vegetation structure derived from Light Detection and Ranging (LiDAR) data as covariates related to variation in bird responses. We used this model to predict occupancy for each bird species in 2001 (using an analogous 2001 LiDAR dataset) and 2014 and used the change in occupancy probability to estimate habitat losses and gains at 3 spatial extents: 1) the entire study area, 2) burned forests only, and 3) unburned, low-lying coastal forests only. Of the 56 bird species we investigated, we observed parameter estimates corresponding to a higher likelihood of occurring in ghost forest for 34 species, but only 9 of those had 95% posterior intervals that did not overlap 0, thus having strong support. Despite the high vulnerability of forests in the region to sea level rise, habitat losses and gains associated with rising sea level were small relative to those resulting from wildfire. Though the extent of habitat changes associated with the development of ghost forest was limited, these changes likely are more permanent and may compound over time as sea level rises at an increasing rate. As such, the proliferation of ghost forests from rising sea level has potential to become an important driver of forest bird habitat change in coastal regions.

Leveraging Big Data Towards Functionally-Based, Catchment Scale Restoration Prioritization.

The persistence of freshwater degradation has necessitated the growth of an expansive stream and wetland restoration industry, yet restoration prioritization at broad spatial extents is still limited and ad-hoc restoration prevails. The River Basin Restoration Prioritization tool has been developed to incorporate vetted, distributed data models into a catchment scale restoration prioritization framework. Catchment baseline condition and potential improvement with restoration activity is calculated for all National Hydrography Dataset stream reaches and catchments in North Carolina and compared to other catchments within the river subbasin to assess where restoration efforts may best be focused. Hydrologic, water quality, and aquatic habitat quality conditions are assessed with peak flood flow, nitrogen and phosphorus loading, and aquatic species distribution models. The modular nature of the tool leaves ample opportunity for future incorporation of novel and improved datasets to better represent the holistic health of a watershed, and the nature of the datasets used herein allow this framework to be applied at much broader scales than North Carolina.