3.1 Occurrence, concentrations, and composition of PFAS in surface waters from Biscayne Bay, Tampa Bay, ENP, and Key West in Central and South Florida
PFAS were detected in samples from 33 sampling sites that cover 13 sites surrounding the Biscayne Bay area, 8 sites from the Tampa Bay area, 6 sites from the ENP adjacent canals, and 6 sites from Key West. The concentration of individual PFAS congeners and total PFAS concentrations (the sum of all the PFAS congeners detected for the same sites) are shown in Table S6. For the sampling sites collected multiple times (≥ 2), the average (Min, Max) concentrations were reported. The spatial distribution of total PFAS in the sampling sites along Central and South Florida is displayed in Fig. 2. Overall, total PFAS concentrations ranged from 6.50 ng L− 1 (TWP, Key West) to 169 ng L− 1 (HB, Miami) with 6 locations above 60 ng L− 1 and 28 locations below 60 ng L− 1.
To better understand spatial variations in Biscayne Bay and canals, we have combined data from samples collected in Oct 2020 (N = 13), Jan 2021 (N = 14), which were previously analyzed and presented (Li et al., 2022), with recently collected samples in Aug 2021 (N = 13). As a result, a total of 40 samples were averaged from 3 sampling trips during 2020–2021, with the average of total PFAS concentrations ranging from 11.4 ng L− 1 (MB17th) to 91.0 ng L− 1 (LR2). Relatively higher concentrations were identified at two sites from Little River (81.0, 91.0 ng L− 1), two sites from Miami River (72.4, 72.1 ng L− 1), Little Arch Creek (54.5 ng L− 1), and Biscayne Bay canal C-8 (53.9 ng L− 1), whereas sites from Miami Beach (MB10th, MB14th, and MB17th) were all below 20 ng L− 1. The variability of the three sets of data is presented as error bars in Fig. 2. Results from multiple sampling events conducted during rainy and dry seasons suggested that PFAS concentration from the same location remains relatively stable with small variations throughout the whole period of study. Little River, Miami River, and Biscayne Bay canal BC-8 are identified as “hot spots”, where the highest levels were observed in this study and the previous one (Li et al., 2022).
From 10 samples collected at ENP adjacent canals in two sampling trips, the average of total PFAS concentrations ranged from 35.0 ng L− 1 (C103) to 52.1 ng L− 1 (C113), as shown in Fig. 2, Table S6. The surface water at site HB was also from canal C-103 closer to its east opening to Biscayne Bay passing Homestead Air Reserve Base. The sampling site HB is around 16 km away from site C103 adjacent to ENP and showed a total concentration of 169 ng L− 1, which was the highest total PFAS concentration detected throughout this study. The 6 samples spreading along the coast of Key West presented total PFAS concentrations in the range of 6.50 ng L− 1 (TWP) to 19.1 ng L− 1 (EKP).
Total PFAS concentrations ranged from 17.4 (TC) to 60.6 ng L− 1 (TB) from the 7 samples collected from Tampa Bay and its adjacent area. Three sites from the North region of Tampa Bay (TB, CB, BPP) and two sites from central Tampa Bay (VP, CTB) falls into a range of 37.8–60.6 ng L− 1, whereas the surface water sample from Terra Ceia aquatic preserve located in South Tampa Bay (TC) had the lowest total PFAS concentration of 17.4 ng L− 1. Two samples collected along the west coast in Port Charlotte (PC) and Marco Island (MI) showed concentrations of 26.11, and 38.23 ng L− 1, respectively.
Overall, among 30 PFAS congeners covered in this study, all were detected in one or more sites. The PFAS composition in each location is presented in Fig. 3. It can be observed that compositions vary from location to location, however, predominant congeners of each defined area can be identified based on detection rates and mean concentrations.
In Biscayne Bay samples, among 28 PFAS congeners detected (Adona and PFONS were not detected), PFOS was the predominant PFAS, with concentrations ranging from 1.16 ng L− 1 to 24.1 ng L− 1 (mean: 10.1 ng L− 1), followed by PFPeA ( 1.50–12.5 ng L− 1; mean: 5.75 ng L− 1), PFHxA (1.15–10.9 ng L− 1; mean: 4.72 ng L− 1), 6 − 2 FTS (0.313–21.8 ng L− 1; mean: 4.32 ng L− 1), PFBA (1.15–10.2 ng L− 1; mean: 4.26 ng L− 1), PFBS (0.671–8.58 ng L− 1; mean: 3.46 ng L− 1), PFOA (0.660–6.90 ng L− 1; mean: 2.84 ng L− 1), PFHpA (0.766–5.01 ng L− 1; mean: 2.69 ng L− 1), and PFHxS (0.520–5.58 ng L− 1; mean: 2.43 ng L− 1). PFOS, PFPeA, PFHxA, PFBA, PFBS, PFOA, PFHpA PFHxS had detection rates of 100%, and 6 − 2 FTS had a detection rate of 92.5%.
In the ENP canal samples, 25 PFAS congeners were detected (6 − 2 FTS, Adona, FHxSA, PFTeDA, FOSA were not detected). Since HB is apart from the ENP adjacent canals, thus it was not included in the following composition analysis represented by ENP samples. PFBA (5.33–12.2 ng L− 1; mean: 9.98 ng L− 1; detection frequency- DF:100%), PFOS (3.42-13.0 ng L− 1; mean: 7.48 ng L− 1; DF: 100%), PFBS (2.01–5.54 ng L− 1; mean: 3.45 ng L− 1; DF:100%), PFPeA (1.85–3.56 ng L− 1; mean: 2.99 ng L− 1; DF:100%), PFHxA (1.14–2.62 ng L− 1; mean: 2.06 ng L− 1; DF:100%), PFOA (0.985-2.40 ng L− 1; mean: 1.74 ng L− 1; DF: 100%), and PFHpA (0.73–1.53 ng L− 1; mean: 1.23 ng L− 1; DF: 90%), were identified as the predominant congeners in the ENP canals. Different from Biscayne Bay samples, some long-chain PFAS such as PFOUDS, PFUdA, N-EtFOSAA, PFDOA, PFTrDA were also predominantly present in the samples.
In Key West samples, 23 PFAS congeners were detected (8 − 2 FTS, PFNS, PFDS, FHxSA, PFUdA, PFTrD, PFTeDA were not detected), whereas PFOS (0.675-3.70 ng L− 1; mean:1.98 ng L− 1), PFBA (1.18–2.15 ng L− 1; mean: 1.64 ng L− 1), PFHpA (0.568–2.62 ng L− 1 mean: 1.48 ng L− 1), and PFPeA (0.494–2.09 ng L− 1; mean: 1.23 ng L− 1) were identified as the predominant congeners at 100% detection rates.
In samples from Tampa Bay area, 19 PFAS were detected (4 − 2 FTS, GenX, Adona, PFONS, PFOUDS, FHxSA, N-MeFOSAA, N-EtFOSAA, PFDoA, PFTrDA, and PFTeDA were not detected), and PFOS was the predominant PFAS, with a concentration ranging from 2.40 ng L− 1 to 25.7 ng L− 1 (mean: 10.0 ng L− 1), followed by PFHpA (1.01–23.1 ng L− 1; mean: 8.36 ng L− 1), PFHxS (0.564–5.48 ng L− 1; mean: 2.83 ng L− 1), PFPeA (1.92–6.10 ng L− 1; mean: 3.72 ng L− 1 ), PFBS (1.19–4.49 ng L− 1; mean: 3.08 ng L− 1), PFBA (1.53–4.49 ng L− 1; mean: 3.08 ng L− 1), and PFHxA (1.33–4.38 ng L− 1; mean: 2.6 ng L− 1), with detection rates of 100% for all the congeners.
3.2 Occurrence, concentrations, and composition of PFAS in tap waters from West Coast and Central Florida
Tap water samples were collected along the West coast of Florida (Sarasota, Tampa City, St. Petersburg, Port Charlotte, Ft. Meyers, and Naples), as well as from cities in Central Florida, (Port St. Lucie, Fort Drum, St. Cloud, and Orlando). Total PFAS concentrations ranged from 1.61 ng L− 1 (Naples) to 45.2 ng L− 1 (Tampa City) in tap water as shown in Fig. 4, and Table S7. Most of the samples fell below 20 ng L− 1 with the exception of Tampa City, where was found a concentration of 45.2 ng L− 1. Tap waters from Naples, Ft Meyers, Sarasota, Orlando, St. Cloud showed total PFAS concentrations below 6 ng L− 1.
Among 30 PFAS, 17 compounds were detected in one or more samples, with FHxSA, PFONS, FOSA, PFNS, PFDS, PFUdA, PFTeDA, 4 − 2 FTS, 6 − 2 FTS, N-EtFOSAA, PFONDS, GenX, and Adona not being detected. The predominant congeners were PFHpA (< MDL-8.57 ng L− 1; mean: 2.45 ng L− 1), PFPeA (< MDL-8.76 ng L− 1; mean: 1.95 ng L− 1), PFOS (< MDL-6.50 ng L− 1; mean: 1.63 ng L− 1), PFHxA (< MDL-3.70 ng L− 1; mean: 1.61 ng L− 1), PFBS (< MDL-4.12ng L− 1; mean: 1.42 ng L− 1), PFOA (< MDL-2.94 ng L− 1, mean: 1.36 ng L− 1), and PFBA (< MDL-4.02 ng L− 1; mean: 1.84 ng L− 1). The detection rates of PFBS, PFPeA, PFPeS, PFHxA, PFHxS, PFHpA, PFOA, and PFOS, ranged from 70 to 90%, while the detection rates of PFBA, FBSA, PFHpS, PFNA, PFDA, PFDoA, and PFTrDA ranged from 10–50%. The composition of PFAS congeners of each sample is shown in Figure S1. Overall, PFHpA accounts for 16.6% of total PFAS detected, followed by PFPeA (14.8%), PFBA, PFBS, FBSA, PFHxA, PFHxS, PFOA, and PFOS ranged from 6.38–9.67%, while other PFAS contributed with less than 4.12%.
3.3 Spatial distribution and potential sources of PFAS in surface waters
Higher concentrations of PFAS are mostly observed in samples from polluted rivers, or samples near coastal estuaries, and point sources, such as military airbases, WWTPs, airports, etc. These water bodies pass through highly urbanized areas with large populations, businesses, in which various chemicals and waste discharges, drainage and runoffs might all end up to. The highest concentration in surface water was observed in the C103 canal (HB: 169 ng L− 1) at 2 km from the Homestead Air Reserve Base, where the historical use of AFFF containing PFAS is well known, and whereas the sample from the same canal on the west side close to ENP is nearly 5-fold lower, suggesting the potential impact of this point source on PFAS levels. The highest levels in Biscayne Bay surface water were found in Miami River (98.9 ng L− 1; Sep 2021), Little River (91.7 ng L− 1; Sep 2021), and BC-8 canal (103 ng L− 1; Aug 2020), which were previously also reported as polluted waterways in South Florida with high levels of wastewater tracers, pharmaceutical and personal care products (PPCPs), and steroid hormones (Blair & Kemp, 2004; Ng et al., 2021).
The ENP adjacent canals presented PFAS levels ranging from 30 to 60 ng L− 1. These freshwater canals on the eastern boundaries serve as a buffer zone that separates the wetlands of ENP from highly productive subtropical agricultural lands and urban development areas (Quinete et al., 2013). Therefore, potential sources are likely to be a mixture of rainfall and runoffs from urban and agricultural areas of southeast Florida, such as from the usage of PFAS containing insecticides and fertilizers (Savvaides et al., n.d.) (Borthakur et al., 2022). Since there are no studies to date performed inside the ENP preserved area, the impact of PFAS in the Everglades water quality is still uncertain.
Key West and Miami beach surface waters showed concentrations below 20 ng L− 1. Though the sample locations include tourist beaches, marinas, drainage openings from apartment buildings, which could have contributed with PFAS input, it still presented a relatively low PFAS pollution level. These samples are associated with the highest salinities observed in this study, one possible reason is that PFAS levels can be substantially lowered by the dilution effect in seawater (Wang et al., 2019).
North Tampa Bay (where Tampa city is located) surface water showed slightly higher concentration than samples from central Tampa Bay followed by samples from South Tampa Bay, which coincides with the population in this area decreasing from North to South, for example, North Tampa Bay (Tampa city) has a population of 383,959, followed by Central Tampa Bay (St. Peterburg) with a population of 258,308, and South Tampa Bay (Palmetto City) with a population of 13,323. (cencus.gov, April 1, 2020). In addition, airports (Tampa International Airport, Clearwater International Airport, and Peter O. Knight Airports), military bases (MacDill Air Force Base), landfills (Pinellas County Solid Waste Disposal), wastewater treatment plants (St Petersburg Wastewater Treatment and Howard F. Curren Advanced Wastewater Treatment Plant) are all found concentrated in North Tampa Bay, which could have contributed to PFAS input in this area. In April 2021, 814 million liters of legacy phosphate mining wastewater and marine dredge water from Piney Point Mining Phosphate Facility were discharged into South Tampa Bay and Port Manatee. Our sampling trip was conducted about one month after the incident, whereas the wastewater input and runoff it carried over might have contribute to the water quality deterioration in this area, potentially affecting the PFAS levels as well.
PFAS concentrations identified in each sampling area are compared and presented in Fig. 5. The identified predominant PFAS included PFOS, PFPeA, PFHxA, PFBA, PFOA, PFHxA, PFHpA, PFHxS, PFNA, PFHpS, PFPes, and 6 − 2 FTS. As seen in Fig. 5, in general, highest concentration of PFAS were found in Biscayne Bay and canals waters followed by Tampa Bay, ENP canals, and Key West for most of the congeners, except for PFBA, which was higher in ENP canals, and PFHpA, which showed the highest level in Tampa Bay.
3.4 Spatial distribution and potential sources of PFAS in tap water from Florida
To better address PFAS spatial distribution, tap water samples collected in this study from Central Florida and the West coast of South Florida are compared to samples collected from the metropolitan area on the East coast of South Florida previously published in Li et al., 2022 using the same method. The average PFAS concentration of the samples from the same region were calculated for comparison and displayed in Fig. 6. The regions in Florida were divided into three groups: Central Florida (St. Lucie, Okeechobee, Osceola, Orange counties; N = 4); West coast of South Florida (Sarasota, Hillsborough, Pinellas, Charlotte, Lee, Collier counties; N = 6), and East coast of South Florida (Palm Beach, Broward, Miami Dade counties; N = 22, data from Li et al. 2022). The total PFAS concentrations were the highest in the East coast of South Florida (mean: 83.0 ng L− 1), followed by the West coast of South Florida (mean: 14.4 ng L− 1), and Central Florida (mean: 8.00 ng L− 1) as shown in Figure S2. This trend coincides with the increased population in the defined groups: East coast of South Florida with a population of 5.6 million followed by West coast of South Florida with 3.6 million, and Central Florida with 1.7 million. Though PFAS levels could be associated with demographics factors (higher population number) and related human activities (higher production and discharge of industrial and domestic wastewater, landfills disposals, among others), the number of samples assessed in other regions and counties were low and further studies including a larger number of samples are needed to allow a more comprehensive and better understanding on the occurrence, distribution, and fate of PFAS in Florida.
The concentrations of most PFAS congeners were higher on the East coast of South Florida, followed by West coast of South Florida, and Central Florida as shown in Fig. 6, except for PFPeS, GenX, PFHpA, and Adona. Adona is an emerging PFAS substitute of PFOA and PFOS which was not detected in any East coast samples but showed concentration up to 6.50 ng L− 1 in the West Coast and Central Florida tap waters. PFBA, PFOS, PFPeA, and PFHxA levels on the East coast showed higher concentration trend compared to the other regions in Florida.
The source of tap water in most of the cities in South and Central Florida such as Miami, Orlando, Fort Mayers, Naples, and Port St. Lucie is primarily from the Floridian aquifer, whereas the sources of tap water in Tampa region which include Tampa city, St. Peterburg, Sarasota, Port Charlotte in this study, are diverse coming from surface water from rivers and canals, groundwater from Floridian aquifer, and desalinated seawater (Tampa.gov, April 1, 2020). However, studies on PFAS occurrence and levels in Floridan aquifers and surface water sources used for drinking purposes are still lacking to be able to draw any conclusion on the contamination source of PFAS in drinking water. It was found that the PFAS level in drinking water is higher than that of surface water in the east coast samples, which could arise from precursors breakdown processes during the water treatment and contamination during distribution processes (Li et al., 2022), but in the West coast and Central Florida water, more samples are needed to evaluate the difference on PFAS levels in surface water and tap waters.
3.5 Principal Component Analysis (PCA)
The Principal Component Analysis (PCA) determined the total variance of the dataset explained by the principal components (PCs) and their eigenvalues in both surface and tap waters. The number of PCs was determined using the Kaiser criterion (eigenvalues > 1; (Mooi & Sarstedt, 2011). Eight PCs were extracted from the surface waters’ dataset, displaying a cumulative variance of 84.71%. As for the tap waters’ dataset, eleven PCs were extracted, exhibiting a cumulative variance of 83.55%.
A representation of the results of the PCA is given by the PCA biplots in Fig. 7. These plots display loadings of the variables (i.e., vectors), determining how strongly each of the variables influence a PC. The further away these vectors are from a PC origin, the higher the influence they have on that PC. Small angles in these loadings indicate positive correlations, while large angles indicate negative correlations, and a 90° angle indicates no correlation. In Fig. 7A, which represents the PCA results for surface waters, a higher loading of the congeners 4 − 2 FTS, 6 − 2 FTS, and 8 − 2 FTS was noted, showing a high influence of these congeners, especially in the Biscayne Bay area. These congeners displayed a strong positive correlation, suggesting they have similar sources. The PCA biplot also showed clusters of samples based on their similarities. Samples from the same areas clustered in groups suggest shared similarities in compounds’ composition, which was especially true in Key West.
In Fig. 7B, which represents the PCA results for tap waters, multiple high loadings were noted, including total PFAS (ΣPFAS), and the congeners PFOA, PFHpA, PFHxA, PFBA, PFNA, PFHxS, and PFOS. This shows a higher influence of these congeners in the data variability, especially in the East coast of South Florida region. Interestingly, all of these congeners are from the same two categories: perfluoroalkyl carboxylic acid (PFCA, including PFOA, PFHpA, PFHxA, PFBA, and PFNA) and perfluoroalkyl sulfonic acid (PFSA; including PFHxS, and PFOS), indicating similar composition and potential sources.
3.6 Ecological and human health risk assessment
PFAS have been identified as Contaminants of Emerging Concern (CEC), which their ubiquitous presence may pose ecological and public health risks. Levels of PFOA and PFOS found here in tap water are below the health advisory guideline from the U.S EPA, which the sum of PFOA and PFOS should not be above 70 ng L− 1. Nevertheless, this level is being currently re-evaluated considering more recent scientific data and new analyses which indicate that negative effects might occur in levels lower than the established advisory value (Mccarthy et al., 2017; Post, 2021). Seafood consumption is another important route of dietary exposure to humans, as PFAS were found in fish tissues including Striped Mullets (Mugil Cephalus) which is a native Floridian fish (Bangma et al., 2018; Denys et al., 2014). Though there are no federal established guidelines that monitor surface water contamination for PFAS in the U.S., currently, the Florida Department of Environmental Protection (FDEP) has developed provisional surface water screening values of 1300 µg L− 1 of PFOA and 37 µg L− 1 of PFOS for fresh-water systems, and 13 µg L− 1 of PFOS in saltwater systems, considering the protection of human health for the consumption of freshwater and estuarine finfish and shellfish (FDEP, 2021). The levels of PFOA ranged from 0.265 ng L− 1 to 10.2 ng L− 1, and PFOS ranged from 0.68 ng L− 1 to 25.7 ng L− 1 in surface water samples from Biscayne Bay, Tampa Bay, ENP, and Key West covered in this study, which are all below these screening levels. However, considering that coastal Florida supports heavy seafood production and consumption, PFAS monitoring on these areas is needed for further human health risk assessment.
In addition, these aquatic ecosystems that support indispensable biomes are incessantly stressed due to these anthropogenic pollutants. As PFAS were identified in Florida coastal water samples, previous study evaluated West Indian manatees inhabiting three coastal sites in Florida (Brevard County, Crystal River, and ENP), where FPOS was detected in the plasma of every Manatee (N = 69) with concentrations up to 166 ng/g ww. Coastal area covered in this study are natural habitats for manatees (Deutsch et al., 2003), and PFOS was also found to be the most predominant PFAS determined in our surface water samples, which suggest the potential environmental impact on these vulnerable and endangered species. Moreover, another study showed that corals, a crucial component to wildlife, tourism, and storm control, can rapidly bioconcentrate and eliminate PFOS, and exposure to PFOS (100 ng L− 1) was associated with increased oxidative stress (Bednarz et al., 2022). When combined with elevated temperature, PFOS can exacerbate the oxidative stress response leading to impaired photosynthesis in corals, which indicates that interactive effects of PFOS exposure with other environmental stressors can induce additional biological effects (Bednarz et al., 2022).
The levels of PFOS found in this study are above most of strict thresholds recommended in Europe, Australia, and New Zealand (0.23 to 23 ng L− 1) for the purpose of protecting aquatic biota. Though PFOS is the most prevalent PFAS detected in our study, PFBA, PFBS, and PFPeA were also predominant in the aquatic environment, therefore, it is important to take into consideration the potential PFAS synergetic effects on aquatic wildlife, which are currently not well understood and demands further assessments.