An assessment and characterization of pharmaceuticals and personal care products (PPCPs) within the Great Lakes Basin: Mussel Watch Program (2013-2018)

Abstract

Defining the environmental occurrence and distribution of chemicals of emerging concern (CECs), including pharmaceuticals and personal care products (PPCPs) in coastal aquatic systems, is often difficult and complex. In this study, 70 compounds representing several classes of pharmaceuticals, including antibiotics, anti-inflammatories, insect repellant, antibacterial, antidepressants, chemotherapy drugs, and X-ray contrast media compounds, were found in dreissenid mussel (zebra/quagga; Dreissena spp.) tissue samples. Overall concentration and detection frequencies varied significantly among sampling locations, site land-use categories, and sites sampled proximate and downstream of point source discharge. Verapamil, triclocarban, etoposide, citalopram, diphenhydramine, sertraline, amitriptyline, and DEET (N,N-diethyl-meta-toluamide) comprised the most ubiquitous PPCPs (> 50%) detected in dreissenid mussels. Among those compounds quantified in mussel tissue, sertraline, metformin, methylprednisolone, hydrocortisone, 1,7-dimethylxanthine, theophylline, zidovudine, prednisone, clonidine, 2-hydroxy-ibuprofen, iopamidol, and melphalan were detected at concentrations up to 475 ng/g (wet weight). Antihypertensives, antibiotics, and antidepressants accounted for the majority of the compounds quantified in mussel tissue. The results showed that PPCPs quantified in dreissenid mussels are occurring as complex mixtures, with 4 to 28 compounds detected at one or more sampling locations. The magnitude and composition of PPCPs detected were highest for sites not influenced by either WWTP or CSO discharge (i.e., non-WWTPs), strongly supporting non-point sources as important drivers and pathways for PPCPs detected in this study. As these compounds are detected at inshore and offshore locations, the findings of this study indicate that their persistence and potential risks are largely unknown, thus warranting further assessment and prioritization of these emerging contaminants in the Great Lakes Basin.

Keywords: Dreissenid mussels; Great Lakes; Land-use; PPCPs; Point source; WWTPs.

Fig. 1

Great Lakes Mussel Watch Program (MWP) inshore (rivers, harbors, embayment, and tributaries) and offshore (nearshore lake and deep-water lake) dreissenid mussel 2013–2018 sampling locations. Some sampling locations have multiple sites. Most sampling locations have 1–3 sites, except Maumee River (8 sites), Muskegon (9 temporal sites), Milwaukee Estuary (13 sites), and Niagara River (20 sites). Additional information on MWP sampling locations description is provided in Table S1 and Table S10 (Supplementary Information)

Fig. 2

Great Lakes dreissenid mussel sampling location associated land-use categories and land cover estimates (%) found proximate (within 3-km buffer) to each MWP site. Sampling locations land cover estimates (%) are presented in order from greatest to least developed land cover category. Land-use information is presented for most MW sites, as some sites were removed due to space constraints. Table S10 (Supplementary Information) provides more details on the land-use categories and land-cover estimations at MWP sites. Individual sites are listed by their general location (associated river/lake region) and state which corresponds with mussel study sites provided in Table S1

Fig. 3

Box and whisker plots showing PPCP concentrations (log ng/g (wet weight)) detected in dreissenid mussels sampled between 2013 and 2018. Figure A depicts PPCP concentrations summarized by compounds in descending order based on highest to lowest mean concentration, while figure B groups the same compounds and depicts PPCP concentration summarized by compound class in descending order based on highest to lowest mean concentration. The x axis (log scaled) represents several orders of magnitude difference in PPCP compounds and compound class concentrations quantified in dreissenid mussel tissue samples. Additional information is provided in Table S4

Fig. 4

Detection frequency (%) for PPCP compounds detected at least once in dreissenid mussels sampled at Great Lakes study sites between 2013 and 2018. The total number of compounds detected per site is depicted in adjoining parentheses. Compounds are ordered by those most frequently to least detected

Fig. 5

Spatial grouping and clusters for Great Lakes MW PPCP sampling locations resulting from random forest (RF) unsupervised classification and cluster analyses. Ellipses represents the 95% confidence intervals for each RF cluster. RF classification results represent three major clusters (clusters 1–3) based on PPCP presence/absence (1/0) and composition profile across mussel sampling locations. Planted/cultivated land cover category was dominant (> 30%) at one study site (NRYT-INMU-CH-10.18; see Fig. 2). Additional information on study sites that overlap clusters and fall within the 95% confidence interval is provided in Table S10

Fig. 6

Box and whisker plots showing PPCP concentration (log ng/g (wet weight)) profile for compounds observed in individual clusters derived from unsupervised random forest (RF) classification. PPCP compounds are ordered by lowest to highest concentration in each RF cluster. Cluster 3 represents the group with the most predominant PPCP mixtures and contained the greatest chemical load and composition observed between clusters. Additional information is provided in Table S5

Fig. 7

Box and whisker plots depicting detected concentrations (log ng/g (wet weight)) for individual PPCPs detected in dreissenid mussels sampled at sites proximate to point source discharge (WWTPs and WWTPs/CSOs), sites downstream and along gradients of wastewater discharge (WWTP Gradient), and non-WWTP sites (sites not influenced by WWTPs or CSOs) during 2013–2018. PPCP compounds are ordered by lowest to highest concentration in each discharge category. Additional information is provided in Table S8

Fig. 8

Box and whisker plots showing PPCP concentration (log ng/g (wet weight)) measured in dreissenid mussels from designated A developed (combined developed medium intensity; developed high intensity; developed, open space; developed, low intensity categories) and B open-water sites. PPCP concentrations summarized by compounds in descending order based on highest to lowest mean concentration. The x axis (log scaled) represents several orders of magnitude difference between PPCP concentrations quantified in dreissenid mussel tissue sampled at developed and open-water sites. Additional information is provided in Table S11