Lower Columbia Estuary Partnership

Abstract Extreme weather events increasingly threaten coastal water quality, yet the mechanisms by which tropical cyclones impair microbial water quality remain poorly quantified. We developed a Large-Language-Model-Assisted Microbial Source Tracking (LAMST) framework to trace the origins of microbial threats and applied it to coastal waters impacted by Hurricane Milton along Florida’s Gulf Coast. LAMST integrates 16S rRNA sequencing with species-level metadata from the NCBI BioSample database to classify taxa as marine, terrestrial, or wastewater in origin. Across 30 sites and three time points (1 week to 7 months post-storm), twelve microbial analytes, including fecal indicator bacteria (FIB), pathogens, and antimicrobial resistance genes (ARGs), were quantified. Colored dissolved organic matter as well as carbon and nitrogen stable isotopes in particulate material were characterized for the same samples and independently corroborated LAMST’s terrestrial and marine microbial classifications. Increases in terrestrial and marine bacterial counts in coastal waters indicated concurrent mobilization of land-derived inputs and marine sources following the hurricane. Site fixed-effects regressions showed terrestrial and wastewater bacteria were strongly associated with enterococci and two ARGs (i.e., sul2 and tetA ), whereas marine bacteria correlated with Vibrio parahaemolyticus . LAMST provides a quantitative, generalizable framework for source-resolved microbial tracking without reliance on regional reference sequences.

Abstract Changes in fish assemblages were examined before and after a culvert was modified to improve the fish passage at the Mirror Lake Complex (MLC), located along the Columbia River Gorge, Oregon. Conditions at the culvert limited water flow between the Columbia River and the MLC during certain portions of the year; thus, the outlet and interior of the culvert were modified to improve fish passage. Prior to the culvert modification, three sites were sampled monthly between April and August 2008, 5.0 km and 0.5 km upstream of the culvert and immediately downstream of the culvert. Following the culvert modification in the late summer of 2008, the same sites were sampled from 2009 to 2012, with two additional sites added in 2010. Prior to the culvert modification, the lower sites (i.e., the sites closest to the Columbia River) supported native and nonnative fish species, while the upper sites were dominated by native species. During the 4 years of monitoring after culvert modification, these distinctions between the upper and lower sites remained. A significant increase in water temperature and species richness was observed at the site just upstream of the culvert, but other changes in fish composition (density, diversity, percent of nonnative species) were not observed. However, at the upper sites, while nonnative species were absent before culvert modification, they were present after modification. Modifications made at the culvert, in combination with seasonal variation in water level and water temperature, may have influenced fish communities in the MLC. Given predicted regional and global climatic changes, our study indicates the importance of long-term monitoring of restoration sites for the presence of nonnative species and the effects of environmental variables, such as water temperature.