Semester of Graduation

Summer 2021

Degree

Master of Science (MS)

Department

Environmental Sciences

Document Type

Thesis

Abstract

Microplastics have quickly emerged as a concerning pollutant in both freshwater and marine environments. Their recent discovery means that their impacts are still being studied, however, it is important to continuously monitor their concentrations. An important conduit of plastic pollution to marine environments are rivers and streams. Previous models have estimated the transport of plastics from land to sea in many parts of the world. Most of these models, however, have been conducted at coarse spatial resolutions that make it difficult to establish tractable management programs to minimize this impact. Here, a previously existing model was applied to model for microplastic emissions from rivers to sea (Lebreton et al., 2017) in the Amite Watershed using population density, watershed distribution, and dams as inputs. To validate this model, six different streams were selected and sampled at along the Amite River with a gradient of predicted microplastic exports. For this in-situ sampling, a simple and economically viable sampling method was chosen for three months of sampling (October, November, and December). The expected result was that the observed values from the in-situ sampling would correlate with the predicted model’s output. The model’s predicted waste outputs for the Amite watershed ranged from 3.33E-6 to 4.89E+3 kg/day. In general, predicted values were at a discrepancy from the observed values. Two sites sampled at (Mill Creek and Clay Cut Bayou), showed higher (observed being 8.78E-3 kg/day while the model predicted 8.28E-7 kg/day) and lower (observed being at 1.22E-2 kg/day while the model predicted the output to be 2.60E-2 kg/day) than expected values respectively, potentially due to the stream conditions and the sampling method chosen which did not account for microplastics caught in sediments. These discrepancies suggest that the sampling method chosen may not be adequate for microplastic analysis in the Amite River watershed. The modeling approach applied is easy to replicate and offers a first glimpse of potential plastic pollution hotspots at the sub watershed level. However, the results of the in-situ study suggest a need for further model validation efforts using alternative methods of plastic sampling that include sediment sampling for better characterizing slow moving waters in Louisiana. This study also highlights the potential need for incorporating other important spatial predictors of plastic pollution, aside from population and runoff. Recommendations for future studies include improving modelling and sampling accuracy for microplastics.

Committee Chair

De Jesus Crespo, Rebeca

DOI

10.31390/gradschool_theses.5371

Share

COinS