Search

Tidal creeks play a vital role in influencing geospatial evolution and marsh ecological communities in coastal landscapes. However, evaluating the geospatial characteristics of numerous creeks across a site and understanding their ecological relationships pose significant challenges due to the labor-intensive nature of manual delineation from imagery. Traditional methods rely on manual annotation in GIS interfaces, which is slow and tedious. This study explores the application of Attention-based Dense U-Net (ADU-Net), a deep learning image segmentation model, for automatically classifying creek pixels in high-resolution (0.5 m) orthorectified aerial imagery in coastal Georgia, USA. We observed that ADU-Net achieved an outstanding F1 score of 0.98 in identifying creek pixels, demonstrating its ability in tidal creek mapping. The study highlights the potential of deep learning models for automated tidal creek mapping, opening avenues for future investigations into the role of creeks in marshes' response to environmental changes. [ABSTRACT FROM AUTHOR]

Bivalves are becoming an increasingly popular tool to counteract eutrophication, particularly in vegetated coastal ecosystems where synergistic interactions between bivalves and plants can govern important N sequestration pathways. In turn, new calls to evaluate how bivalve densities modify N pools and processes across multiple scales have surfaced.Ribbed mussels, Geukensia demissa, and their relationship with smooth cordgrass present a classic demonstration of positive bivalve‐plant interactions and offer a useful model for assessing density dependence. We measure porewater ammonium concentrations, N stable isotope signatures in cordgrass tissue, and sediment N fluxes in mussel aggregations and in cordgrass‐only plots across a southeastern U.S. salt marsh.In addition to measuring the effect of mussel presence, we evaluate mussel density dependence through a multiscale approach. At the patch scale, we quantify mussel density effects within their aggregations (individuals m−2) while at a larger landscape scale, we quantify mussel density effects on the cordgrass‐only areas they neighbour (individuals ~30 m−2).Porewater ammonium concentrations were halved in mussel biodeposits relative to sediments in cordgrass‐only areas and negatively related to mussel density within aggregations. Leaf clip ẟ15N signatures were nearly 2‰ higher in cordgrass growing among mussel aggregations and increased with increasing patch mussel density. Microcosm incubations showed that mussels enhanced N2 flux (i.e., nitrogen removal) and DIN flux (i.e., N regeneration) into the water column, where only nitrogen removal increased with increasing patch‐scale mussel density. Across the marsh landscape, mussel coverage drove ammonium accumulation and N2 flux in sediments.Synthesis. Our results suggest that, at the patch scale, mussels stimulate the microbial metabolism of N, the assimilation of this bioavailable N by cordgrass, and nitrogen removal in a positive, density‐dependent manner. Tidal currents redistribute mussel biodeposits from mussel aggregations to surrounding areas, influencing biogeochemical transformations at scales beyond their physical footprint. We emphasize that the N regeneration potential of bivalve populations is a significant metric contributing to their mitigation potential and that bivalve density effects may be non‐linear, vary across patch to ecosystem scales, and have differing implications for the plants with which they interact. [ABSTRACT FROM AUTHOR]

Coastal ecosystems provide critical storm and flood protection but are rapidly degrading worldwide, making their restoration urgent. We evaluated whether successional facilitation, where pioneers facilitate climax species, could be leveraged to accelerate coastal dune revegetation. A survey spanning 270 km of Southeast U.S. coastline revealed that Panicum amarum (bitter panicum) supported higher plant richness than Uniola paniculata (sea oats) and that sea oat cover was 230% greater on mature dunes than disturbed dunes, suggesting bitter panicum functions as a pioneer and sea oats as a climax species. A reciprocal transplant experiment confirmed this interpretation: bitter panicum stem production and height fell by 37% and >20 cm, respectively, when planted proximate to sea oats versus in isolation, whereas sea oats produced 38% more and >12 cm taller stems when planted proximate to bitter panicum versus in isolation. A second experiment evaluating the density‐dependence of this facilitative interaction revealed that sea oats transplanted into low densities of bitter panicum grew >15% taller than isolated and high‐density treatments. However, within 7 months, wave inundation eliminated over 60% of propagules in both experiments. To explore foredune inundation frequency and its implications for dune revegetation, we applied empirical wave runup models at 101 locations throughout Volusia County, Florida. While disturbance frequency varied seasonally and annually, sites with low dune crests and steep beach slopes experienced frequent inundation (>50 events/year). Given the interactions between geomorphology and vegetation success, we present a decision matrix to guide managers in determining optimal revegetation methods tailored to project goals and site conditions. [ABSTRACT FROM AUTHOR]