3.1 Yearly Concentrations in the Brazos River
Of the evaluated compounds, MeP, PrP, BuP, Cl2MeP, Cl2EtP, PHBA and DHBA were determined to be significantly different by site. Of the parent products, MeP was found in the highest concentrations at sites along the Brazos River that are upstream of the wastewater treatment and in the Tehuacana Creek tributary (Fig. 2). PrP was detected in higher concentrations than MeP at the sites following release. Most of the differences in BuP concentrations are due to higher concentrations at site 6 located Downstream (south) of the meeting point of Tehuacana Creek and the Brazos River. Concentrations of both dichlorinated paraben transformation products were higher at sites downstream of the effluent release location. While PHBA saw a significant decrease at the site just downstream of the release of wastewater effluent, DHBA did not significantly decrease. The oxidation of PHBA to DHBA could be responsible for the decrease in PHBA and could also explain the stability in DHBA concentrations.
The products that were both significantly different between sites and had notable spatial trends, were mapped to visualize differences in analyte concentrations at different points in the Brazos. Of the parent compounds MeP, PrP and BuP concentrations had spatial trends (Fig. 3). MeP concentrations decreased significantly from the site upstream of wastewater treatment to the site downstream of wastewater treatment. However, this decrease is unlikely to be solely due to dilution by the release of wastewater effluent, as the small outflow of the wastewater treatment with a flow of 32 ft3/sec is unlikely to have a large impact on the Brazos with flow rates ranging from an average of 220 ft3/sec in March 2021 to an average of 16,640 ft3/sec in June 2021. Flow rates remained above 2000 ft3/sec through August before decreasing in September. The higher water flows in the summer months would decrease concentrations due to dilution. Transformation of parabens in the environment is a major contributor to the removal of parabens in surface water. Biodegradation is the most likely route of transformation of parabens in surface water. The differences in microbial communities along the Brazos could be the cause for the sudden change in methyl paraben concentrations seen at the site downstream of wastewater treatment. Biodegradation occurs much more quickly under aerobic conditions than under anaerobic conditions. (Wu et al., 2017) A change to more aerobic conditions could cause more degradation to occur between site 3 and site 4. Neither hydrolysis nor photolysis are major pathways in paraben transformation in the environment. (Svobodova et al., 2006; Talrose et al., n.d.; Valkova et al., 2001; Xu et al., 2021) MeP concentrations in the Brazos increased from site 4 to site 6 (p = 0.041), which could be due to the input of MeP from other sources. Considering that site 6 is located near a residential neighborhood, input from recreational activities is one potential source of the observed increase. Desorption from solids released from the wastewater treatment plant is another possibility as parent parabens have been detected in wastewater sludge with MeP concentrations of 89.4 ng/g. (Chen et al., 2017) MeP has also been detected in river surface sediment, with a median concentration of 12.4 ng/g. (Feng et al., 2019) A study by Feng et al (2019) found that in the Huai River, mean MeP concentrations were 51.7 ng/L while mean MeP concentrations in the sediment were 11.6 ng/g giving a Kd of 0.224 L/g. However, the same study detected MeP concentrations of 8.84 ng/L in surface water and 13.0 ng/g in sediment in the Yellow River. The Kd for the Yellow River values was 1.47 L/g. Arfaeinia et al (2022) found that the median Kd between seawater and sediment for MeP was 3.52 g/mL, which is lower than the Kd values calculated from the MeP concentrations in the river samples, showing variation by environment. (Arfaeinia et al., 2022) Using an average of 0.847 L/g from the two Kd values determined from the Huai River and Yellow River, and with average total dissolved solids of 300 mg/L in effluent at WWTP1, the minimum concentration sorbed to sludge needed to cause the observed concentration change is 783 nmol/kg or 119 ng/g. MeP concentrations detected in wastewater sludge in previous studies show that this concentration change could possibly be due to desorption from wastewater solids after effluent release. While the tributary upstream of site 6 does have detectable MeP concentrations at site 5, a combination of the lower MeP concentration and small size of the tributary means that the tributary is unlikely to have a large impact on the paraben concentration in the Brazos. It is unlikely that the increase in MeP concentrations are due to direct input as the area of the Brazos River is not used for recreation. MeP concentrations decreased further downstream with significantly lower concentrations at site 8 when compared to site 6 (p = 0.045). The observed change in concentration is likely due to natural MeP degradation and dilution in the river. The tributary with the small wastewater treatment plant has a very low MeP concentration which is unlikely to have a large impact on MeP concentration in the Brazos. EtP concentrations were not significantly different between any site in the yearly analysis. This due to the low concentrations of EtP at all sites. The low EtP concentrations were expected due to EtP being used seldom in industry.
PrP had similar concentration changes in the river as MeP. PrP decreased at site 4 (site 3 and site 4: p = 0.030) downstream of effluent release before increasing at site 6 (site 4 and site 6: p = 0.015). However, the visual decrease in PrP is smaller than that of MeP, and the differences between site 6 and site 8 were determined to be insignificant (site 6 and site 8: p = 1.00) The lack of decrease shows that PrP degraded at a slower rate than MeP in the Brazos, this matches the transformation rates seen in many treatments and metabolic process, though PrP has been seen to degrade faster than MeP in activated sludge. (Abbas et al., 2010; Li et al., 2015; Lu et al., 2018) The Bull Hide Creek tributary also had a higher concentration of PrP but are still not in high enough concentration to explain the similarities in PrP concentrations in the main channel. Similarity of PrP concentrations in the Brazos River main channel demonstrates a slower degradation rate for PrP than for MeP.
BuP concentrations at upstream sites were lower in concentration than either MeP or PrP. This is expected as BuP is used much less often than MeP or PrP in either personal care products or foodstuff. BuP concentrations increased from site 3 to site 4 (p < 0.001), which could be due to release of BuP in wastewater effluent. BuP increased from site 4 to site 6 (p < 0.001), followed by a decrease from site 6 to site 8 (p < 0.001), as MeP had. No notable spatial trends were observed for BzP.
Of the transformation products, PHBA, DHBA, Cl2MeP and Cl2EtP concentrations showed spatial trends (Fig. 4). PHBA was the compound detected in highest concentrations at all sites with its highest concentration at 10.30 ng/L (4.850 ng/L to 21.88 ng/L) also having the largest variation amongst the compounds. As with MeP and PrP, PHBA concentrations significantly decreased from site 3 to site 4 (site 3-site 4: p = 0.031). This decrease in PHBA could be due to aerobic biodegradation occurring between site 3 and site 4. In the case of aerobic biodegradation, PHBA can be further degraded into phenol via carboxylases, which does not occur with anaerobic biodegradation. This change in environment is likely due to a dam less than a mile upstream of site 3 that causes aeration of the river water as the water traverses the dam. The short distances between the dam and site 3 could explain the changes are not yet seen at that site. PHBA concentrations increased visually from site 4 to site 6 and from site 6 to site 8. However, these increases were not significant (site 4 and site 6: p = 1.00, site 6 and site 8 p = 1.00). This was unexpected, as PHBA is the common degradation product of all parent parabens and would be expected increase in concentration as parabens degrade. Site 7 had a relatively high concentration of PHBA, when compared to the low concentration of the other analytes at that site with an average concentration of 4.465 ng/L (2.479 ng/L to 8.045 ng/L). This could be due to PHBA being a potential transformation product of UV disinfection. However, the higher concentration in the tributary did not seem to have significant effect on Brazos River PHBA concentrations. While DHBA concentrations showed overall significant difference between sites, the only significant differences were between the Brazos River sites and the tributaries and none of the sites along the river showed any significant differences between each other.
Cl2MeP concentrations remained consistently low upstream of wastewater effluent release, with no significant differences between sites 1 through 3 (site 1 and site 2: p = 1.00, site 2 and site 3: p = 1.00). However, at site 4, downstream of wastewater treatment, concentrations were notably higher and Cl2MeP concentrations were significantly different from site 3 (site 3 and site 4 p = 0.032). Dichlorinated parabens are not used in industry and are not likely to be introduced from other sources between site 3 and site 4 and dichlorinated parabens quantified at upstream sites are likely remaining from release in effluents further upstream. A waterpark with high concentrations of chlorine located along the Bosque River 2 km upstream of the where the Bosque and Brazos meet could introduce chlorine or chlorinated parabens to the upstream sites. The dichlorinated paraben species are released in higher concentrations than the parents in effluent at the major wastewater treatment plant. (Penrose and Cobb, 2023) Cl2MeP concentrations remained consistent at sites downstream of effluent release, with only small visual decreases and no significant differences between sites 4, 6 and 8 (site 4 and site 6: p = 1.00, site 6 and site 8: p = 1.00). The stability in concentrations at downstream sites shows that Cl2MeP degraded at a much slower rate than the parent parabens. Given the low concentrations of Cl2MeP in the tributary and water quantity differences between the tributaries and the Brazos River, it is unlikely that either of the tributaries played a significant role in the lack of decreases in Cl2MeP at sites 6 and 8.
Cl2EtP concentrations had similar trends to Cl2MeP, with a stable concentration upstream of effluent, an increase at release (site 3 and site 4: p = 0.036), and a stable concentration downstream. Cl2EtP concentrations were lower than Cl2MeP concentrations due to the higher use of MeP in products resulting in higher concentrations of MeP transformation products in the river. Both dichlorinated paraben products were released from wastewater effluent and entered the Brazos at concentrations high enough to cause a change in the Cl2MeP concentrations in the Brazos River. This combined with persistence seen with both compounds show that the dichlorinated species, will be more relevant than their respective parent compounds in the environment. Dichlorinated PHBA is an important transformation product due to the high concentrations in the river. Data has been obtained for both mono and dichlorinated PHBA. However, they were not included in spring quality control due to the standards not being available until after spring analysis and neither are included in this study. While the dichlorinated species were in detectable concentration changes along the river, ClMeP and ClEtP did not have noticeable spatial trends during the yearly analysis, which is due to the low concentrations of ClMeP and ClEtP at all sites. The dichlorinated species are the only compounds evaluated that increase in the Brazos River directly as due to the release of wastewater effluent. Longer chained chlorinated species have not been evaluated due to a lack of available standards.
3.2 Seasonal Differences in Analyte Concentrations Along the Brazos River.
3.2.1 Differences in Analyte Concentrations Between Seasons
Of the quantified compounds, PrP BzP, DHBA and ClEtP concentrations were determined to be significantly different between seasons. (Table 1). For PrP most of the differences were due to lower concentrations in the summer. Variation in concentrations were generally highest in the summer for multiple compounds, which is likely due to the low concentrations during June and early July and high concentrations in early August. Summer was expected to have the highest concentrations due to increased use of personal care products, particularly sunscreen. However, the study area has a large student population that would be absent during summer. The low concentrations in early summer could also be explained by high water levels during that period due to a large amount of rainfall in late May and early June. High flow rates in May would also cause low analyte concentrations in late Spring. March and April flow rates were lower at an average of 183 ft3/sec and 176 ft3/sec respectively before increasing due to rainfall in May with average flow in May being 7617 ft3/sec. Flow rates were highest in June at 16640 ft3/sec and remained higher throughout July and in early August with average flows of 5870 ft3/sec in July and 1674 ft3/sec in August. August had higher analyte concentrations than the other summer months, the change in analyte concentrations are due to the decreasing water level and return of the student population. The change from low analyte concentrations in early and mid-summer and higher analyte concentrations in late summer caused higher variation in the summer. By September the flow rates were near flows in April at 375 ft3/sec, though there were days in September with flow rates up to 1840 ft3/sec. High water levels impact the concentrations detected in late spring, but overall did not cause noticeable differences between spring and winter or spring and fall. Summer PrP concentrations were significantly different than winter but not significantly different from fall and spring, with summer having generally lower concentrations than winter. This is likely due to both the high-water levels that persisted from late spring throughout most of summer and the decreased population in the general area during the summer. While ClEtP concentrations showed significant differences by season, adjusted p-values showed no differences between ClEtP concentration between any two specific seasons (Table 2). Seasonal differences in DHBA concentration were determined to be caused by low DHBA concentrations in the winter.
Table 1
General differences between seasons. Bolded values are significant.
Compound | P-Values Between Seasons |
MeP | 0.610 |
EtP | 0.803 |
PrP | 0.022 |
BuP | 0.076 |
BzP | 0.019 |
PHBA | 0.746 |
DHBA | 0.029 |
MePOH | 0.184 |
EtPOH | 0.070 |
ClMeP | 0.809 |
Cl2MeP | 0.506 |
ClEtP | 0.027 |
Cl2EtP | 0.359 |
Table 2
Significant p-values from seasonal analyses. A lager table that includes non-significant p-values is included as supplemental information. (Table S5)
Compound | Seasons | P-Value |
MeP | Fall and Summer | 0.0381 |
Spring and Summer | 0.0088 |
BzP | Fall and Summer | 0.0004 |
Summer and Winter | 0.0361 |
DHBA | Fall and Winter | 0.0485 |
Spring and Winter | 0.0492 |
Summer and Winter | 0.0453 |
MePOH | Spring and Winter | 0.0363 |
EtPOH | Spring and Winter | 0.0272 |
Summer and Winter | 0.0487 |
Cl2EtP | Fall and Summer | 0.0497 |
Spring and Summer | 0.0264 |
Summer and Winter | 0.0076 |
General differences in concentration by season were determinable for multiple compounds. When evaluating differences in seasonal concentrations at each individual site, differences were much more sporadic only having a few notable differences randomly distributed across sites. (Table 3). None of the three upstream sites had any significant differences in MeP, PrP, or PHBA concentrations between seasons, despite those three compounds being found in the highest concentrations at the three sites. Reasons for the lack of differences could be consistent release of these compounds from upstream, though differences would still be expected due to changes in water level. Sites 4, 6 and 8 were expected to have different concentrations of Cl2MeP and Cl2EtP, as the concentrations released in wastewater effluent upstream of site 4 would vary by season and flow rate through the plant. However, Cl2MeP concentrations were only seasonally different at site 3, while Cl2EtP was significantly different at site 3 and the two tributaries. No specific compound was seasonally different across all upstream sites or all downstream sites and so no trends could be identified. P-values from Dunn’s tests showing which seasonal concentrations were significantly different at each site are included as supplemental information.
Table 3
Seasonal concentration differences at individual sites along the Brazos River. Bolded values are significant.
| Seasonal P-Values by Site |
Compound | Site 1 | Site 2 | Site 3 | Site 4a | Site 5 | Site 6 | Site 7b | Site 8 |
MeP | 0.576 | 0.587 | 0.752 | 0.236 | 0.035 | 0.564 | 0.043 | 0.459 |
EtP | 0.358 | 0.009 | 0.196 | 0.586 | 0.220 | 0.224 | 0.108 | 0.845 |
PrP | 0.364 | 0.217 | 0.089 | 0.190 | 0.791 | 0.438 | 0.009 | 0.152 |
BuP | 0.665 | 0.325 | 0.141 | 0.392 | 0.460 | 0.137 | 0.021 | 0.638 |
BzP | 0.021 | 0.308 | 0.227 | 0.002 | 0.002 | 0.144 | 0.046 | 0.182 |
PHBA | 0.711 | 0.068 | 0.407 | 0.160 | 0.579 | 0.325 | 0.3372 | 0.138 |
DHBA | 0.199 | 0.919 | 0.499 | 0.346 | 0.753 | 0.768 | 0.023 | 0.566 |
MePOH | 0.026 | 0.562 | 0.367 | 0.386 | 0.649 | 0.513 | 0.027 | 0.580 |
EtPOH | 0.578 | 0.914 | 0.612 | 0.893 | 0.015 | 0.118 | 0.002 | 0.700 |
ClMeP | 0.124 | 0.009 | 0.165 | 0.501 | 0.819 | 0.554 | 0.456 | 0.597 |
Cl2MeP | 0.809 | 0.011 | 0.164 | 0.118 | 0.028 | 0.050 | 0.560 | 0.501 |
ClEtP | 0.640 | 0.015 | 0.056 | 0.022 | 0.226 | 0.496 | 0.025 | 0.780 |
Cl2EtP | 0.079 | 0.161 | 0.074 | 0.828 | 0.086 | 0.024 | 0.112 | 0.486 |
a. Site 4 is just downstream of WWTP1 b. Site 7 is downstream of WWTP2.
3.2.2 Differences Between Sites Across Seasons
In general, concentration differences between sites at each season was consistent with the differences seen in the yearly analysis with some variation between seasons. (Table 4) MeP concentrations were generally the highest of the parent parabens at upstream sites with averages of 1.814 ng/L (0.4826 ng/L to 6.817 ng/L) during fall, 1.162 ng/L (0.3133 ng/L to 3.145 ng/L) during spring, 0.8018 ng/L (0.2031 ng/L to 3.165 ng/L) during summer and 0.9434 ng/L (0.1849 ng/L to 4.814 ng/L) during the winter at site 1. MeP concentrations varied by site during every season except summer (Table 4) with Dunn’s Tests (Table S6 to Table S10) showing that the differences are specifically between sites 1 and 4 (spring: p = 0.009, fall: p = 0.028, winter: p = 0.022), sites 2 and 4 (spring: p = 0.011, fall: p = 0.009, winter: p = 0.038) and sites 3 and 4 (spring: p = 0.006, fall: p = 0.018, winter p = 0.018). Therefore, concentrations of MeP dropped significantly during three out of four of the seasons moving from site 3 to site 4, which matches the change seen in the yearly analysis. Maps showing spatial changes and bar graphs showing analyte concentrations by site are included as supplemental information. The increase in MeP at site 4 to site 6 seen in the yearly analysis was visually seen across seasons. However, winter (site 4 to site 6: p = 0.041) was the only season with a significant increase, going from an average MeP concentration of 0.0421 ng/L (0.0155 ng/L to 0.1049 ng/L) at site 4 to 0.8998 ng/L (0.3292 ng/L to 2.459 ng/L) at site 6. No significant trends were identified for PrP in any individual season. Visual changes in PrP were similar to that of changes seen in MeP. However, the only significant differences were due to low PrP concentrations in the tributaries with no differences between sites in the main channel, sites and the tributaries, and not between two sites along the Brazos. BuP was the only compound with significant differences between sites in every season (spring: p < 0.001, summer: p < 0.001, fall p < 0.001, winter: p = 0.031). However, BuP concentrations are noticeably high at site 6 with average concentrations of 0.4685 ng/L (0.3928 ng/L to 0.5588 ng/L) in spring, 0.6120 ng/L (0.4716 ng/L to 0.7943 ng/L) in summer, 0.5617 ng/L (0.5023 ng/L to 0.6280 ng/L) in the fall and 0.5239 ng/L (0.4714 ng/L to 0.5822 ng/L) in the winter. The high concentrations of BuP at site 6 are likely responsible for the observed differences, as the only differences in BuP concentrations are between site 6 and other sites. BzP concentrations were significantly different between sites in the fall and winter. However, differences in the fall were between sites 1 and 4 (p = 0.035), sites 4 and 5 (p = 0.009) and sites 5 and 6 (p = 0.0122), while differences in winter were only between site 4 and 5 (p = 0.036). While there are differences, no trends were identified.
Table 4
General differences between sites by season. Bolded values are significant.
| P-Values by Season |
Compound | Spring | Summer | Fall | Winter |
MeP | < 0.001 | 0.079 | < 0.001 | < 0.001 |
EtP | 0.016 | 0.814 | < 0.001 | 0.064 |
PrP | 0.073 | 0.009 | 0.001 | 0.003 |
BuP | < 0.001 | < 0.001 | < 0.001 | 0.031 |
BzP | 0.979 | 0.021 | < 0.001 | < 0.001 |
PHBA | 0.256 | 0.052 | 0.048 | 0.013 |
DHBA | 0.015 | 0.001 | 0.100 | 0.481 |
MePOH | 0.830 | 0.086 | 0.009 | 0.792 |
EtPOH | 0.811 | 0.006 | 0.037 | 0.050 |
ClMeP | 0.005 | 0.502 | < 0.001 | 0.006 |
Cl2MeP | 0.261 | 0.008 | < 0.001 | 0.018 |
ClEtP | 0.004 | 0.284 | < 0.001 | 0.006 |
Cl2EtP | 0.169 | 0.039 | < 0.001 | 0.043 |
During most seasons, PHBA and DHBA were present in higher concentrations than any of the other analytes, likely due to being natural degradation products of all parent parabens. PHBA concentrations were significantly different between site during the fall and winter while DHBA concentrations were significantly different in spring and summer. However, in all cases, the differences were between non-tributary sites and site 5, meaning the only notable trend is that there are significantly lower concentrations of PHBA at site 5 during fall and winter and DHBA during spring and summer. It is important to note that PHBA was only significantly different during seasons where DHBA was not significantly different, and vice versa. This could be due to DHBA being a hydroxylated product of PHBA. Increases in DHBA concentration would be related to decreases in PHBA concentration, which could be due to transformation of PHBA into DHBA by aerobic microbes using hydroxylases. Transformation from PHBA to DHBA upstream of site 5 during the fall in winter would decrease PHBA concentrations, causing PHBA concentrations at that site to be significantly lower than other sites. Cl2MeP and Cl2EtP concentrations were significantly different between sites during all seasons except spring (Table 5). Further investigation using adjusted p-values from Dunn’s tests (Table S4) did not show significant differences in Cl2MeP concentrations between any individual sites during the summer. Differences in winter followed trends seen in the yearly analysis with a significant decrease from site 3 and site 4 (p = 0.024) and no significant changes at sites downstream of site 4. As with the yearly concentrations, Cl2MeP remained consistent from site 4 to site 6 and from site 6 to site 8. Given the lack of statistical differences, the numerical increases are due to the high variation of Cl2MeP concentrations during all seasons at all downstream sites. One explanation for the high variation in Cl2MeP is inconsistent release in Cl2MeP concentrations in wastewater effluent between weeks. During fall there was a numeric change between site 3 and site 4 but no significant change, though there were differences between other upstream and downstream sites showing that there was a change between upstream and downstream sites (site 2 and 4: p < 0.001. site 1 and 6: p = 0.004, site 2 and 6: p < 0.001, site 2 and 8: p = 0.008). Cl2EtP changes were spatially similar to Cl2MeP. While Cl2EtP concentrations were determined to be different during the summer, the only differences were between site 6 and site 7 (p = 0.020) and between site 3 and site 7 (p = 0.010). The lack of differences between upstream and downstream sites of both Cl2MeP and Cl2EtP in the summer, means that the release of effluent did not have the same effect on river water concentrations that it had in other seasons. This could be due to the high rainfall in early summer that would not only increase water level in the Brazos, thereby increasing dilution of effluent, but also cause high flow rates and reduced retention times in wastewater treatment, reducing transformation from parent to dichlorinated species. As with Cl2MeP, Cl2EtP also was not significantly different between site 3 and 4 during the fall but were different between other upstream and downstream sites (site 2 and 6: p = 0.034). However, unlike Cl2MeP, Cl2EtP concentrations were significantly different between individual upstream sites and site 4 during spring (site 2 and 4: p = 0.023, site 3 and 4: p = 0.034). Cl2MeP concentrations would be expected to have greater increases due to the higher use of MeP that would result in higher concentrations released in effluent.