Occurrence and Fate of Agriculturally Derived Natural Estrogens in the Environment

Abstract

It is alarming that hormonally active substances are regularly detected in surface waters worldwide. Until now, few studies have been conducted on the role of agriculture as a potential source for endocrine disrupting substances, such as the natural estrogens 17α-estradiol (E2α), 17β-estradiol (E2β), estrone (E1), and estriol (E3), in surface waters. This could be of particular concern because female livestock animals excrete natural estrogens during different stages of the estrous cycle and pregnancy through urine and feces. The majority of urinary and fecal livestock excretions are collected in the form of slurry. The fate of these estrogens in slurry is not yet fully understood. Slurries are applied in large amounts on agricultural areas as fertilizer. The export of slurry derived natural estrogens from agricultural areas is a potential source for estrogenic pollution in surface waters. Even at low nanograms per liter concentrations in surface waters, estrogens exert endocrine disrupting effects in fish – the most important, but not the only, recipient class of organisms. The aim of this doctoral thesis was to better understand the occurrence and fate of agriculturally derived natural estrogens in the environment and to evaluate their risk as pollutants for the aquatic environment. Specifically, the three main objectives were: 1) to determine concentrations of natural estrogens in slurry and to investigate their fate during slurry storage, 2) to quantify natural estrogen emissions to subsurface tile drains from agricultural areas fertilized with livestock slurry, and 3) to assess the prevalence of natural estrogens in surface waters with agriculturally dominated catchments. The occurrence and fate of natural estrogens was studied in cattle and pig slurries. The effect of on farm slurry storage on natural estrogen concentrations in cattle slurry was investigated first by collecting and analysing slurry samples in different compartments of a dairy housing over four months. This study was complemented with a nationwide cattle and pig slurry pit monitoring, which aimed to determine average natural estrogen concentrations in Swiss livestock slurries. Natural estrogens in slurry were extracted with an adapted QuEChERS method and quantified with liquid chromatography and tandem mass spectrometry (LC-MS/MS). During storage, natural estrogen concentrations remained relatively constant in different compartments of the dairy housing. Mean natural concentrations ranged from 138 to 861 ng/L and from 54 to 244 ng/L in Swiss cattle and pig slurries, respectively. 17α-Estradiol predominated in cattle slurries and E3 was most abundant in pig slurries. In Switzerland, the estimated annual load of total natural estrogens applied on agricultural areas through slurry applications was 36 mg/ha. The majority of the total natural estrogens (E2α+E2β+E1+E3) is emitted to the environment through wastewater treatment plant effluents (i.e., 68 %) and not through slurry applications. Natural estrogens were applied on tile-drained experimental fields in the form of cattle slurry, pig slurry or dissolved in water, and quantified in flow-proportionally collected drainage water. A liquid-liquid extraction followed by derivatization with dansyl chloride and the same LC-MS/MS method as for estrogens in slurry extracts were used for extraction and quantification of natural estrogens in drainage water. After pig and cattle slurry applications, concentration maxima in drainage water occurred immediately after rain events, with 73, 8, 37, and 60 ng/L for E2α, E2β, E1, and E3, respectively. The fractions of individual natural estrogens exported to tile drains after cattle and pig slurry applications were on average 0.26% and 0.18%, respectively. Lower exported fractions for estrogens applied in aqueous solution point at particle facilitated transport when applied in slurry. Preferential flow was proposed as main natural estrogen transport pathway to tile drains. Our estimation shows that in Switzerland annually 6.3 g of E2α, 0.6 g of E2β, 7.4 g of E1, and 7.7 g of E3 are emitted to surface waters through subsurface tile drains. Natural estrogen emissions from WWTPs were estimated to be three to four orders of magnitude higher than those from tile drains, i.e., 1.5 kg of E2β, 9.7 kg of E1, and 51.1 kg of E3. The prevalence of natural estrogens (derived from farming livestock) in streams and ponds was studied in the catchment of Lake Baldegg, which is a livestock farming dominated catchment in Switzerland. Passive samplers were deployed in one tributary, daily time-proportional water samples were collected in five tributaries for 30 days, and grab samples were taken in 12 ponds in the catchment. All monitoring campaigns were conducted at the beginning of the vegetation period, when slurry applications are most frequent. Aqueous samples were chemically analysed in the same way as drainage water samples and stream water samples additionally with ERα-CALUX – a bioassay for assessing total estrogenic activity. Certain natural estrogens were frequently detected and their mean concentrations were between below limit of detection (0.10 ng/L) and 0.55 ng/L for E2β and E1, respectively. Both passive sampling and bioassay results largely supported these findings. A Monte Carlo simulation underestimated measured mean natural estrogen concentrations by a factor of three to 11. Temporal estrogen occurrence patterns point to preferential flow from agricultural areas as the main emission process to surface waters. In ponds, natural estrogens were occasionally detected up to 8.6 ng/L for E2α. Overall, our results showed that agriculture is a source for natural estrogens in surface waters. Slurry derived estrogens caused temporarily elevated estrogen concentrations in drainage and stream water, but in streams the European Union environmental quality standards (EU EQS) (0.4 ng/L for E2β and 3.6 ng/L for E1) were never exceeded for longer than a day. Our estimation demonstrated however, that in a stream impacted by agriculture and an effluent of a WWTP estrogen concentrations that exceed the threshold values might occur and human derived estrogens were the predominating estrogen source. Critical E1 concentrations for aquatic organisms occurred in one pond. Consequently, the risk for estrogen concentrations above the EU EQS seems to be elevated in ponds surrounded by agricultural landscapes as well as in small and medium sized streams impacted by both agriculture and WWTP effluents.