Pesticides in streams are increasingly becoming a global concern, but there is little information on the safe concentration of aquatic ecosystems. In a 30-day mesocosmic experiment, the native benthic aquatic invertebrates were exposed to the common insecticide fipronil and four kinds of degradation products. The fipronil compound caused changes in the emergence and trophic cascade. The effective concentration (EC50) at which fipronil and its sulfide, sulfone and desulfinyl degradation products cause a 50% response has been developed. Taxanes are not sensitive to fipronil. The hazard concentration of 5% of the affected species from 15 mesocosmic EC50 values ​​is used to convert the compound concentration of fipronil in the field sample into the sum of toxic units (∑TUFipronils). In 16% of streams drawn from five regional studies, the average ∑TUFipronil exceeded 1 (indicating toxicity). Invertebrate indicators of species at risk are negatively correlated with TUTUipronil in four of the five sampling areas. This ecological risk assessment shows that low concentrations of fipronil compounds will reduce stream communities in many parts of the United States.
Although the production of synthetic chemicals has greatly increased in recent decades, the impact of these chemicals on non-target ecosystems has not been fully understood (1). In surface water where 90% of global farmland is lost, there is no data on agricultural pesticides, but where there are data, the time for pesticides to exceed regulatory thresholds is half (2). A meta-analysis of agricultural pesticides in surface waters in the United States found that in 70% of sampling locations, at least one pesticide exceeded the regulatory threshold (3). However, these meta-analyses (2, 3) only focus on surface water affected by agricultural land use, and are a summary of discrete studies. Pesticides, especially insecticides, also exist in high concentrations in urban landscape drainage (4). It is rare to conduct a comprehensive assessment of pesticides in surface water discharged from agriculture and urban landscapes; therefore, it is not known whether pesticides pose a large-scale threat to surface water resources and their ecological integrity.
Benzopyrazoles and neonicotinoids accounted for one third of the global pesticide market in 2010 (5). In surface waters in the United States, fipronil and its degradation products (phenylpyrazoles) are the most common pesticide compounds, and their concentrations usually exceed the aquatic standards (6-8). Although neonicotinoids have attracted attention due to their effects on bees and birds and their prevalence (9), fipronil is more toxic to fish and birds (10), while other phenylpyrazoles Class compounds have herbicidal effects (5). Fipronil is a systemic insecticide used to control pests in urban and agricultural environments. Since fipronil entered the world market in 1993, the use of fipronil in the United States, Japan and the United Kingdom has greatly increased (5). In the United States, fipronil is used to control ants and termites, and is used in crops including corn (including seed treatment), potatoes and orchards (11, 12). Agricultural use of fipronil in the United States peaked in 2002 (13). Although no national urban use data is available, urban use in California peaked in 2006 and 2015 (https://calpip.cdpr.ca) .gov/main .cfm, accessed December 2, 2019). Although high concentrations of fipronil (6.41μg/L) are found in streams in some agricultural areas with high application rates (14), compared with agricultural streams, urban streams in the United States generally have more detection and higher High concentrations, positive for the occurrence of storms are associated with the test (6, 7, 14-17).
Fipronil enters the aquatic ecosystem of runoff or leaches from the soil into the stream (7, 14, 18). Fipronil has low volatility (Henry’s law constant 2.31×10-4 Pa m3 mol-1), low to moderate water solubility (3.78 mg/l at 20°C), and moderate hydrophobicity (log Kow is 3.9 to 4.1)), the mobility in the soil is very small (log Koc is 2.6 to 3.1) (12, 19), and it exhibits low-to-medium persistence in the environment (20). Finazepril is degraded by photolysis, oxidation, pH-dependent hydrolysis and reduction, forming four main degradation products: dessulfoxyphenapril (nor sulfoxide), phenaprenip sulfone (sulfone), Filofenamide (amide) and filofenib sulfide (sulfide). Fipronil degradation products tend to be more stable and durable than the parent compound (21, 22).
The toxicity of fipronil and its degradation into non-target species (such as aquatic invertebrates) has been well documented (14, 15). Fipronil is a neurotoxic compound that interferes with chloride ion passage through the chloride channel regulated by gamma-aminobutyric acid in insects, resulting in sufficient concentration to cause excessive excitement and death (20). Fipronil is selectively toxic, so it has a greater receptor binding affinity for insects than mammals (23). The insecticidal activity of fipronil degradation products is different. The toxicity of sulfone and sulfide to freshwater invertebrates is similar or higher than that of the parent compound. Desulfinyl has moderate toxicity but is less toxic than the parent compound. Relatively non-toxic (23, 24). The susceptibility of aquatic invertebrates to fipronil and fipronil degradation varies greatly within and between taxa (15), and in some cases even exceeds an order of magnitude (25). Finally, there is evidence that phenylpyrazoles are more toxic to the ecosystem than previously thought (3).
Aquatic biological benchmarks based on laboratory toxicity testing may underestimate the risk of field populations (26-28). Aquatic standards are usually established by single-species laboratory toxicity testing using one or several aquatic invertebrate species (for example, Diptera: Chironomidae: Chironomus and Crustacea: Daphnia magna and Hyalella azteca). These test organisms are generally easier to cultivate than other benthic macroinvertebrates (for example, phe genus::), and in some cases are less sensitive to pollutants. For example, D. Magna is less sensitive to many metals than certain insects, while A. zteca is less sensitive to the pyrethroid insecticide bifenthrin than its sensitivity to worms (29, 30). Another limitation of existing benchmarks is the endpoints used in the calculations. Acute benchmarks are based on mortality (or fixed for crustaceans), while chronic benchmarks are usually based on sublethal endpoints (such as growth and reproduction) (if any). However, there are widespread sublethal effects, such as growth, emergence, paralysis, and developmental delay, which may affect the success of taxa and community dynamics. As a result, although the benchmark provides a background for the biological importance of the effect, the ecological relevance as a threshold for toxicity is uncertain.
In order to better understand the effects of fipronil compounds on benthic aquatic ecosystems (invertebrates and algae), natural benthic communities were brought into the laboratory and exposed to concentration gradients during the 30-day flow Fipronil or one of the four fipronil degradation experiments. The research goal is to produce a species-specific 50% effect concentration (EC50 value) for each fipronil compound representing a broad taxa of a river community, and to determine the impact of pollutants on community structure and function [ie, hazard concentration] 5 % Of affected species (HC5) and indirect effects such as altered emergence and trophic dynamics]. Then the threshold (compound-specific HC5 value) obtained from the mesoscopic experiment was applied to the field collected by the United States Geological Survey (USGS) from five regions of the United States (Northeast, Southeast, Midwest, Northwest Pacific, and Central California Coastal Zone) Data) as part of the USGS regional stream quality assessment (https://webapps.usgs.gov/rsqa/#!/). As far as we know, this is the first ecological risk assessment. It comprehensively investigates the effects of fipronil compounds on benthic organisms in a controlled meso-environment, and then applies these results to continental-scale field assessments.
The 30-day mesocosmic experiment was conducted at the USGS Aquatic Laboratory (AXL) in Fort Collins, Colorado, USA from October 18th to November 17th, 2017, for 1 day of domestication and 30 days of experimentation. The method has been previously described (29, 31) and detailed in the supplementary material. The meso space setting contains 36 circulating flows in the four active flows (circulating water tanks). Each living stream is equipped with a cooler to keep the water temperature and is illuminated with a 16:8 light-dark cycle. The meso-level flow is stainless steel, which is suitable for the hydrophobicity of fipronil (log Kow = 4.0) and suitable for organic cleaning solvents (Figure S1). The water used for the meso-scale experiment was collected from the Cache La Poudre River (upstream sources including Rocky Mountain National Park, National Forest and Continental Divide) and stored in AXL’s four polyethylene storage tanks. Previous assessments of sediment and water samples collected from the site did not find any pesticides (29).
The meso-scale experiment design consists of 30 processing streams and 6 control streams. The treatment stream receives treated water, each of which contains unreplicated constant concentrations of fipronil compounds: fipronil (fipronil (Sigma-Aldrich, CAS 120068-37-3), amide (Sigma-Aldrich, CAS 205650-69-7), desulfurization group [US Environmental Protection Agency (EPA) Pesticide Library, CAS 205650-65-3], sulfone (Sigma-Aldrich, CAS 120068-37-2) and sulfide (Sigma-Aldrich, CAS 120067-83-6); all purity ≥97.8%. According to published response values ​​(7, 15, 16, 18, 21, 23, 25, 32, 33). By dissolving fipronil compound in methanol ( Thermo Fisher Scientific, American Chemical Society certification level), and dilute with deionized water to the required volume to prepare a concentrated stock solution. Because the amount of methanol in a dose is different, it is necessary to add methanol to all treatment streams as needed. In the three controls, to ensure the same methanol concentration (0.05 ml/L) in the streams. The middle view of the other three control streams received river water without methanol, otherwise they were treated as all other streams.
On the 8th day, the 16th day and the 26th day, the temperature, pH value, electrical conductivity and the degradation of fipronil and fipronil were measured in the flow membrane. In order to track the degradation of the parent compound fipronil during the media test, fipronil (parents) was used to treat the fluid intestinal mucosa for another three days [days 5, 12 and 21 (n = 6)] for temperature, pH , Conductivity, fipronil and fipronil degradation sampling. The pesticide analysis samples were collected by filtering 10 ml of flowing water into a 20 ml amber glass vial through a Whatman 0.7-μm GF/F syringe filter equipped with a large diameter needle. The samples were immediately frozen and sent to the USGS National Water Quality Laboratory (NWQL) in Lakewood, Colorado, USA for analysis. Using an improved method of the previously published method, Fipronil and 4 degradation products in water samples were determined by direct aqueous injection (DAI) liquid chromatography-tandem mass spectrometry (LC-MS / MS; Agilent 6495). The instrument detection level (IDL) is estimated to be the minimum calibration standard that meets the qualitative identification standard; the IDL of fipronil is 0.005 μg/L, and the IDL of the other four fipronil is 0.001 μg/L. The supplementary material provides a complete description of the methods used to measure fipronil compounds, including quality control and assurance procedures (for example, sample recovery, spikes, third party inspections, and blanks).
At the end of the 30-day Mesocosmic experiment, the enumeration and identification of adult and larval invertebrates were completed (the main data collection endpoint). The emerging adults are collected from the net every day and frozen in a clean 15 ml Falcon centrifuge tube. At the end of the experiment (day 30), the contents of the membrane in each stream were scrubbed to remove any invertebrates, and sieved (250 μm) and stored in 80% ethanol. Timberline Aquatics (Fort Collins, CO) has completed the taxonomic identification of larvae and adult invertebrates to the lowest taxonomic level possible, usually species. On days 9, 19 and 29, chlorophyll a was measured in triplicate in the mesoscopic membrane of each stream. All chemical and biological data as part of the mesoscopic experiment are provided in the accompanying data release (35).
Ecological surveys were conducted in small (wading) streams in five major areas of the United States, and pesticides were monitored during the previous index period. In short, based on agricultural and urban land use (36-40), 77 to 100 locations were selected in each region (444 locations in total). During the spring and summer of one year (2013-2017), water samples are collected once a week in each region for 4 to 12 weeks. The specific time depends on the region and development intensity. However, the 11 stations in the northeast region are almost in the watershed. No development, except that only one sample was collected. Since the monitoring periods for pesticides in regional studies are different, for comparison, only the last four samples collected at each site are considered here. It is assumed that a single sample collected at the undeveloped Northeast site (n = 11) can represent the 4-week sampling period. This method leads to the same number of observations on pesticides (except for the 11 locations in the Northeast) and the same duration of observation; it is believed that 4 weeks is long enough for long-term exposure to the biota, but short enough that the ecological community does not Should recover from these contacts.
In the case of sufficient flow, the water sample is collected by means of constant velocity and constant width increments (41). When the flow is not enough to use this method, you can collect samples by deep integration of samples or grabbing from the center of gravity of the flow. Use a large-bore syringe and disc filter (0.7μm) to collect 10 ml of filtered sample (42). Through DAI LC-MS/MS/MS/MS, water samples were analyzed at NWQL for 225 pesticides and pesticide degradation products, including fipronil and 7 degradation products (dessulfinyl fipronil, fipronil) Sulfides, fipronil sulfone, deschlorofipronil, desthiol fipronil, amide, fipronil and fipronil). ). Typical minimum reporting levels for field studies are: fipronil, desmethylthio fluorobenzonitrile, fipronil sulfide, fipronil sulfone, and deschlorofipronil 0.004 μg/L; dessulfinyl fluorfenamide and The concentration of fipronil amide is 0.009 μg/liter; the concentration of fipronil sulfonate is 0.096 μg/liter.
The invertebrate communities are sampled at the end of each area study (spring/summer), usually at the same time as the last pesticide sampling event. After the growing season and the heavy use of pesticides, the sampling time should be consistent with the low flow conditions, and should coincide with the time when the river invertebrate community matures and is mainly in the larval life stage. Using a Surber sampler with a 500μm mesh or a D-frame net, invertebrate community sampling was completed in 437 out of 444 sites. The sampling method is described in detail in the supplementary material. On NWQL, all invertebrates are usually identified and listed at the genus or species level. All chemical and biological data collected in this field and used in this manuscript can be found in the accompanying data release (35).
For the five fipronil compounds used in the mesoscopic experiment, the concentration of the larval invertebrates reduced by 20% or 50% was calculated relative to the control (ie EC20 and EC50). The data [x = time-weighted fipronil concentration (see supplementary material for details), y = larval abundance or other metrics] were fitted to the R(43) extended package using a three-parameter logarithmic regression method” drc”. The curve fits all species (larvae) with sufficient abundance and meets other metrics of interest (for example, taxa richness, total mayfly abundance, and total abundance) to further understand the community effect. The Nash-Sutcliff coefficient (45) is used to evaluate the model fit, where a poor model fit can receive infinite negative values, and the value of a perfect fit is 1.
To explore the effects of fipronil compounds on the emergence of insects in the experiment, the data were evaluated in two ways. First, by subtracting the average appearance of the control flow meso from the appearance of each treatment flow meso, the cumulative daily occurrence of insects from each flow meso (the total number of all individuals) was normalized to the control. Plot these values ​​against time to understand the deviation of the treatment fluid mediator from the control fluid mediator in the 30-day experiment. Second, calculate the total occurrence percentage of each flow mesophyll, which is defined as the ratio of the total number of mesophylls in a given flow to the average number of larvae and adults in the control group, and is suitable for three-parameter logarithmic regression. All the germination insects collected were from two subfamilies of the Chironomidae family, so a combined analysis was performed.
Changes in community structure, such as the loss of taxa, may ultimately depend on the direct and indirect effects of toxic substances, and may lead to changes in community function (for example, trophic cascade). To test the trophic cascade, a simple causal network was evaluated using the path analysis method (R package “piecewiseSEM”) (46). For mesoscopic experiments, it is assumed that fipronil, desulfinyl, sulfide and sulfone (not tested amide) in the water to reduce the biomass of the scraper, indirectly lead to an increase in the biomass of chlorophyll a (47). The compound concentration is the predictor variable, and the scraper and chlorophyll a biomass are the response variables. Fisher’s C statistic is used to evaluate model fit, so that a P value <0.05 indicates a good model fit (46).
In order to develop a risk-based eco-community threshold protection agent, each compound has obtained 95% of the affected species (HC5) chronic species sensitivity distribution (SSD) and hazard concentration protection. Three SSD data sets were generated: (i) only meso data set, (ii) a data set containing all meso data and data collected from EPA ECOTOX database query (https://cfpub.epa.gov/ecotox) /, accessed on March 14, 2019), the study duration is 4 days or longer, and (iii) a data set containing all mesoscopic data and ECOTOX data, in which ECOTOX data (acute exposure) divided by acute to The ratio of chronic D. magna (19.39) to explain the difference in exposure duration and approximate the chronic EC50 value (12). Our purpose of generating multiple SSD models is to (i) develop HC5 values ​​for comparison with field data (only for SSDs for media), and (ii) assess that media data are more widely accepted than regulatory agencies for inclusion in aquaculture The robustness of life benchmarks and standard setting of data resources, and therefore the practicability of using mesoscopic studies for the adjustment process.
SSD was developed for each data set using the R package “ssdtools” (48). Use the bootstrap (n = 10,000) to estimate the HC5 average and confidence interval (CI) from the SSD. Forty-nine taxa responses (all taxa that have been identified as genus or species) developed through this research are combined with 32 taxa responses compiled from six published studies in the ECOTOX database, for a total of 81 Taxon response can be used for SSD development. Since no data was found in the ECOTOX database of amides, no SSD was developed for amides and only one EC50 response was obtained from the current study. Although the EC50 value of only one sulfide group was found in the ECOTOX database, the current graduate student has 12 EC50 values. Therefore, SSDs for sulfinyl groups have been developed.
The specific HC5 values ​​of fipronil compounds obtained from the SSD data set of Mesocosmos only were combined with field data to assess the exposure and potential toxicity of fipronil compounds in 444 streams from five regions in the United States. In the last 4-week sampling window, each concentration of fipronil compounds detected (undetected concentrations are zero) is divided by its respective HC5, and the compound ratio of each sample is summed to obtain The total toxicity unit of fipronil (ΣTUFipronils), where ΣTUFipronils> 1 means toxicity.
By comparing the hazard concentration of 50% of the affected species (HC50) with the EC50 value of taxa richness derived from the medium membrane experiment, the SSD obtained from the medium membrane data was evaluated to reflect the sensitivity of the wider ecological community to fipronil degree. . Through this comparison, the consistency between the SSD method (including only those taxa with a dose-response relationship) and the EC50 method (including all unique taxa observed in the middle space) using the EC50 method of measuring taxa richness can be evaluated Sex. Dose response relationship.
A pesticide risk species (SPEARpesticides) indicator was calculated to investigate the relationship between the health status of invertebrate communities and ΣTUFipronil in 437 invertebrate-collecting streams. The SPEARpesticides metric converts the composition of invertebrates into an abundance metric for biological taxonomy with physiological and ecological characteristics, thereby imparting sensitivity to pesticides. The SPEARpesticides indicator is not sensitive to natural covariates (49, 50), although its performance will be affected by severe habitat degradation (51). The abundance data collected on-site for each taxon is coordinated with the key value of the taxon related to the ASTERICS software to assess the ecological quality of the river (https://gewaesser-bewertung-berechnung.de/index.php/home. html). Then import the data into Indicate (http://systemecology.eu/indicate/) software (version 18.05). In this software, the European trait database and the database with physiological sensitivity to pesticides are used to convert the data of each site into SPEARpesticides indicator. Each of the five regional studies used the General Additive Model (GAM) ["mgcv" package in R(52)) to explore the relationship between the SPEARpesticides metric and ΣTUFipronils [log10(X + 1) conversion] Associated. For more detailed information on SPEARpesticides metrics and for data analysis, please see the Supplementary Materials.
The water quality index is consistent in each flow mesoscopic and the whole mesoscopic experiment period. The average temperature, pH and conductivity were 13.1°C (±0.27°C), 7.8 (±0.12) and 54.1 (±2.1) μS/cm (35), respectively. The measured dissolved organic carbon in clean river water is 3.1 mg/L. In the meso-view of the river where the MiniDOT recorder is deployed, the dissolved oxygen is close to saturation (average> 8.0 mg/L), indicating that the stream is fully circulated.
Quality control and quality assurance data on fipronil are provided in the accompanying data release (35). In short, the recovery rates of laboratory matrix spikes and mesoscopic samples are usually within acceptable ranges (recoveries of 70% to 130%), IDL standards confirm the quantitative method, and laboratory and instrument blanks are usually clean There are very few exceptions other than these generalizations discussed in the supplementary material. .
Due to system design, the measured concentration of fipronil is usually lower than the target value (Figure S2) (because it takes 4 to 10 days to reach a steady state under ideal conditions) (30). Compared with other fipronil compounds, the concentration of desulfinyl and amide changes little over time, and the variability of the concentration within the treatment is smaller than the difference between treatments except for the low concentration treatment of sulfone and sulfide. The time-weighted average measured concentration range for each treatment group is as follows: Fipronil, IDL to 9.07μg/L; Desulfinyl, IDL to 2.15μg/L; Amide, IDL to 4.17μg/L; Sulfide, IDL To 0.57μg/liter; and sulfone, IDL is 1.13μg/liter (35). In some streams, non-target fipronil compounds were detected, that is, compounds that were not spiked into a specific treatment, but were known to be degradation products of the treatment compound. The mesoscopic membranes treated with the parent compound fipronil have the highest number of non-target degradation products detected (when not used as a processing compound, they are sulfinyl, amide, sulfide and sulfone); these may be due to the production process Compound impurities and/or degradation processes that occur during the storage of the stock solution and (or) in the mesoscopic experiment rather than the result of cross-contamination. No trend of degradation concentration was observed in fipronil treatment. Non-target degradation compounds are most commonly detected in the body with the highest treatment concentration, but the concentration is less than the concentration of these non-target compounds (see the next section for the concentration). Therefore, since non-target degradation compounds are usually not detected in the lowest fipronil treatment, and because the detected concentration is lower than the effect concentration in the highest treatment, it is concluded that these non-target compounds have minimal impact on the analysis.
In media experiments, benthic macroinvertebrates were sensitive to fipronil, desulfinyl, sulfone, and sulfide [Table S1; original abundance data is provided in accompanying data version (35)]. Fipronil amide is only for the fly Rhithrogena sp. Toxic (fatal), its EC50 is 2.05μg/L [±10.8(SE)]. Dose-response curves of 15 unique taxa were generated. These taxa showed mortality within the tested concentration range (Table S1), and targeted clustered taxa (such as flies) (Figure S3) and rich taxa (Figure 1) A dose response curve was generated. The concentration (EC50) of fipronil, desulfinyl, sulfone and sulfide on the unique taxa of the most sensitive taxa range from 0.005-0.364, 0.002-0.252, 0.002-0.061 and 0.005-0.043μg/L, respectively . Rhithrogena sp. And Sweltsa sp.; Figure S4) are lower than the more tolerated taxa (such as Micropsectra / Tanytarsus and Lepidostoma sp.) (Table S1). According to the average EC50 of each compound in Table S1, sulfones and sulfides are the most effective compounds, while invertebrates are generally the least sensitive to desulfinyl (excluding amides). Metrics of the overall ecological status, such as taxa richness, total abundance, total pentaploid and total stone fly, including taxa and the abundance of some taxa, these are very rare in meso and cannot be calculated Draw a separate dose response curve. Therefore, these ecological indicators include taxon responses not included in the SSD.
Taxa richness (larva) with a three-level logistic function of (A) fipronil, (B) desulfinyl, (C) sulfone, and (D) sulfide concentration. Each data point represents larvae from a single stream at the end of the 30-day meso experiment. Taxon richness is the count of unique taxa in each stream. The concentration value is the time-weighted average of the observed concentration of each stream measured at the end of the 30-day experiment. Fipronil amide (not shown) has no relationship with rich taxa. Please note that the x-axis is on a logarithmic scale. EC20 and EC50 with SE are reported in Table S1.
At the highest concentration of all five fipronil compounds, the emergence rate of Uetridae declined. The percentage of germination (EC50) of sulfide, sulfone, fipronil, amide and desulfinyl was observed to decrease by 50% at the concentrations of 0.03, 0.06, 0.11, 0.78 and 0.97μg/L respectively (Figure 2 and Figure S5) . In most of the 30-day experiments, all treatments of fipronil, desulfinyl, sulfone and sulfide were delayed, except for some low-concentration treatments (Figure 2), and their appearance was inhibited. In the amide treatment, the accumulated effluent during the entire experiment was higher than that of the control, with a concentration of 0.286μg/liter. The highest concentration (4.164μg/liter) during the entire experiment inhibited the effluent, and the effluent rate of the intermediate treatment was similar to that of the control group. (figure 2).
Cumulative emergence is the average daily average emergence of each treatment minus (A) fipronil, (B) desulfinyl, (C) sulfone, (D) sulfide and (E) amide in the control stream The average daily average emergence of the membrane. Except for control (n = 6), n = 1. The concentration value is the time-weighted average of the observed concentration in each flow.
The dose-response curve shows that, in addition to taxonomic losses, structural changes at the community level. Specifically, within the test concentration range, the abundance of may (Figure S3) and taxa abundance (Figure 1) showed significant dose-response relationships with fipronil, desulfinyl, sulfone, and sulfide. Therefore, we explored how these structural changes lead to changes in community function by testing the nutritional cascade. Exposure of aquatic invertebrates to fipronil, desulfinyl, sulfide and sulfone has a direct negative impact on the biomass of the scraper (Figure 3). In order to control the negative impact of fipronil on the biomass of the scraper, the scraper also negatively affected the chlorophyll a biomass (Figure 3). The result of these negative path coefficients is a net increase in chlorophyll a as the concentration of fipronil and degradants increases. These fully mediated pathway models indicate that increased degradation of fipronil or fipronil leads to an increase in the proportion of chlorophyll a (Figure 3). It is assumed in advance that the direct effect between fipronil or degradation concentration and chlorophyll a biomass is zero, because fipronil compounds are pesticides and have low direct toxicity to algae (for example, the EPA acute non-vascular plant baseline concentration is 100μg / L fipronil, disulfoxide group, sulfone and sulfide; https://epa.gov/pesticide-science-and-assessing-pesticide-risks/aquatic-life-benchmarks-and-ecological-risk), All results (valid models) support this hypothesis.
Fipronil can significantly reduce the biomass (direct effect) of grazing (scraper group is larvae), but has no direct effect on the biomass of chlorophyll a. However, the strong indirect effect of fipronil is to increase the biomass of chlorophyll a in response to less grazing. The arrow indicates the standardized path coefficient, and the minus sign (-) indicates the direction of association. * Indicates the degree of importance.
The three SSDs (middle layer only, middle layer plus ECOTOX data, and middle layer plus ECOTOX data corrected for differences in exposure duration) produced nominally different HC5 values ​​(Table S3), but the results were within the SE range. In the rest of this study, we will focus on the data SSD with only the meso universe and the related HC5 value. For a more complete description of these three SSD evaluations, please refer to the supplementary materials (Tables S2 to S5 and Figures S6 and S7). The best-fitting data distribution (lowest Akaike information standard score) of the four fipronil compounds (Figure 4) used only in the meso-solid SSD map is the log-gumbel of fipronil and sulfone, and the weibull of sulfide And desulfurized γ (Table S3). The HC5 values ​​obtained for each compound are reported in Figure 4 for the meso universe only, and in Table S3 the HC5 values ​​from all three SSD data sets are reported. The HC50 values ​​of fipronil, sulfide, sulfone and desulfinyl groups [22.1±8.78 ng/L (95% CI, 11.4 to 46.2), 16.9±3.38 ng/L (95% CI, 11.2 to 24.0), 8 80±2.66 ng/L (95% CI, 5.44 to 15.8) and 83.4±32.9 ng/L (95% CI, 36.4 to 163)] These compounds are significantly lower than the EC50 taxa richness (total number of unique taxa) (Table S1; the notes in the supplementary material table are micrograms per liter).
In the meso-scale experiment, when exposed to (A) fipronil, (B) dessulfinyl fipronil, (C) fipronil sulfone, (D) fipronil sulfide for 30 days, the species sensitivity is described It is the EC50 value of taxon. The blue dashed line represents 95% CI. The horizontal dashed line represents HC5. The HC5 value (ng/L) of each compound is as follows: Fipronil, 4.56 ng/L (95% CI, 2.59 to 10.2); Sulfide, 3.52 ng/L (1.36 to 9.20); Sulfone, 2.86 ng/ Liter (1.93 to 5.29); and sulfinyl, 3.55 ng/liter (0.35 to 28.4). Please note that the x-axis is on a logarithmic scale.
In the five regional studies, Fipronil (parents) was detected in 22% of the 444 field sampling points (Table 1). The detection frequency of florfenib, sulfone and amide is similar (18% to 22% of the sample), the detection frequency of sulfide and desulfinyl is lower (11% to 13%), while the remaining degradation products are very high. Few (1% or less) or never detected (Table 1). . Fipronil is most frequently detected in the southeast (52% of the sites) and least frequently in the northwest (9% of the sites), which highlights the variability of benzopyrazole use and potential stream vulnerability across the country. Degradants usually show similar regional patterns, with the highest detection frequency in the southeast and the lowest in northwest or coastal California. The measured concentration of fipronil was the highest, followed by the parent compound fipronil (90% percentage of 10.8 and 6.3 ng/L, respectively) (Table 1) (35). The highest concentration of fipronil (61.4 ng/L), disulfinyl (10.6 ng/L) and sulfide (8.0 ng/L) was determined in the southeast (in the last four weeks of the sample). The highest concentration of sulfone was determined in the west. (15.7 ng/L), amide (42.7 ng/L), dessulfinyl flupirnamide (14 ng/L) and fipronil sulfonate (8.1 ng/L) (35). Florfenide sulfone was the only compound that was observed to exceed HC5 (Table 1). The average ΣTUFipronils between the various regions vary greatly (Table 1). The national average ΣTUFipronils is 0.62 (all locations, all regions), and 71 sites (16%) have ΣTUFipronils> 1, indicating that it may be toxic to benthic macroinvertebrates. In four of the five regions studied (except the Midwest), there is a significant relationship between SPEARpesticides and ΣTUFipronil, with adjusted R2 ranging from 0.07 along the coast of California to 0.34 in the southeast (Figure 5).
*Compounds used in mesoscopic experiments. †ΣTUFipronils, the median of the sum of toxin units [observed field concentration of four fipronil compounds/hazard concentration of each compound from the fifth percentile of the SSD-infected species (Figure 4)] For the weekly samples of fipronil, the last 4 weeks of pesticide samples collected at each site were calculated. ‡The number of locations where pesticides are measured. §The 90th percentile is based on the maximum concentration observed on site during the last 4 weeks of pesticide sampling. with the percentage of samples tested. ¶ Use the 95% CI of the HC5 value (Figure 4 and Table S3, only meso) to calculate the CI. Dechloroflupinib has been analyzed in all regions and has never been found. ND, not detected.
The Fipronil toxic unit is the measured fipronil concentration divided by the compound-specific HC5 value, which is determined by the SSD obtained from the media experiment (see Figure 4). Black line, generalized additive model (GAM). The red dashed line has a CI of 95% for GAM. ΣTUFipronils is converted to log10 (ΣTUFipronils+1).
The adverse effects of fipronil on non-target aquatic species have been well documented (15, 21, 24, 25, 32, 33), but this is the first study in which it is sensitive in a controlled laboratory environment. The communities of the taxa were exposed to fipronil compounds, and the results were extrapolated on a continental scale. The results of the 30-day mesocosmic experiment can produce 15 discrete aquatic insect groups (Table S1) with unreported concentration in the literature, among which the aquatic insects in the toxicity database are underrepresented (53, 54). Taxa-specific dose-response curves (such as EC50) are reflected in community-level changes (such as taxa richness and may fly abundance loss) and functional changes (such as nutritional cascades and changes in appearance). The effect of the mesoscopic universe was extrapolated to the field. In four of the five research areas in the United States, the field-measured fipronil concentration was correlated with the decline of the aquatic ecosystem in the flowable water.
The HC5 value of 95% of the species in the medium membrane experiment has a protective effect, indicating that overall aquatic invertebrate communities are more sensitive to fipronil compounds than previously understood. The obtained HC5 value (florfenib, 4.56 ng/liter; desulfoxirane, 3.55 ng/liter; sulfone, 2.86 ng/liter; sulfide, 3.52 ng/liter) is several times (florfenib) to three times More than an order of magnitude (desulfinyl) below the current EPA chronic invertebrate benchmark [fipronil, 11 ng/liter; desulfinyl, 10,310 ng/liter; sulfone, 37 ng/liter; and sulfide, for 110 ng/liter (8)]. Mesoscopic experiments identified many groups that are sensitive to fipronil instead of those indicated by the EPA chronic invertebrate benchmark (4 groups that are more sensitive to fipronil, 13 pairs of desulfinyl, 11 pairs of sulfone and 13 pairs) Sulfide sensitivity) (Figure 4 and table) S1). This shows that benchmarks cannot protect several species that are also observed in the middle world, which are also widespread in aquatic ecosystems. The difference between our results and the current benchmark is mainly due to the lack of fipronil toxicity test data applicable to a range of aquatic insect taxa, especially when the exposure time exceeds 4 days and fipronil degrades. During the 30-day mesocosmic experiment, most insects in the invertebrate community were more sensitive to fipronil than the common test organism Aztec (crustacean), even after correcting the Aztec The EC50 of Teike makes it the same after acute transformation. (Usually 96 hours) to chronic exposure time (Figure S7). A better consensus was reached between the medium membrane experiment and the study reported in ECOTOX using the standard test organism Chironomus dilutus (an insect). It is not surprising that aquatic insects are particularly sensitive to pesticides. Without adjusting the exposure time, the meso-scale experiment and the comprehensive data of the ECOTOX database showed that many taxa were observed to be more sensitive to fipronil compounds than diluted Clostridium (Figure S6). However, by adjusting the exposure time, Dilution Clostridium is the most sensitive organism to fipronil (parent) and sulfide, although it is not sensitive to sulfone (Figure S7). These results illustrate the importance of including multiple types of aquatic organisms (including multiple insects) to produce actual pesticide concentrations that can protect aquatic organisms.
The SSD method can protect rare or insensitive taxa whose EC50 cannot be determined, such as Cinygmula sp. , Isoperla fulva and Brachycentrus americanus. The EC50 values ​​of taxa abundance and may fly abundance reflecting changes in community composition are consistent with the HC50 values ​​of the SSD of fipronil, sulfone and sulfide. The protocol supports the following idea: The SSD method used to derive thresholds can protect the entire community, including rare or insensitive taxa in the community. The threshold of aquatic organisms determined from SSDs based on only a few taxa or insensitive taxa may be greatly insufficient in protecting aquatic ecosystems. This is the case for desulfinyl (Figure S6B). Due to the lack of data in the ECOTOX database, the EPA chronic invertebrate baseline concentration is 10,310 ng/L, which is four orders of magnitude higher than the 3.55 ng/L of HC5. The results of different taxon response sets produced in mesoscopic experiments. The lack of toxicity data is particularly problematic for degradable compounds (Figure S6), which may explain why the existing aquatic biological benchmarks for sulfone and sulfide are about 15 to 30 times less sensitive than the SSD HC5 value based on China Universe. The advantage of the medium membrane method is that multiple EC50 values ​​can be determined in a single experiment, which is sufficient to form a complete SSD (for example, desulfinyl; Figure 4B and Figures S6B and S7B), and have a significant impact on the natural taxa of the protected ecosystem Many responses.
Mesoscopic experiments show that fipronil and its degradation products may have obvious sublethal and indirect adverse effects on community function. In the mesoscopic experiment, all five fipronil compounds appeared to affect the emergence of insects. The results of the comparison between the highest and lowest concentrations (inhibition and stimulation of individual emergence or changes in emergence time) are consistent with the previously reported results of meso experiments using the insecticide bifenthrin (29). The emergence of adults provides important ecological functions and can be altered by pollutants such as fipronil (55, 56). Simultaneous emergence is not only critical for insect reproduction and population persistence, but also for the supply of mature insects, which can be used as food for aquatic and terrestrial animals (56). Preventing the emergence of seedlings may adversely affect the food exchange between aquatic ecosystems and riparian ecosystems, and spread the effects of aquatic pollutants into terrestrial ecosystems (55, 56). The decrease in the abundance of scrapers (algae-eating insects) observed in the meso-scale experiment resulted in a decrease in algae consumption, which resulted in an increase in chlorophyll a (Figure 3). This trophic cascade changes the carbon and nitrogen fluxes in the liquid food web, similar to a study that evaluated the effects of pyrethroid bifenthrin on benthic communities (29). Therefore, phenylpyrazoles, such as fipronil and its degradation products, pyrethroids, and perhaps other types of insecticides, may indirectly promote the increase in algal biomass and the perturbation of carbon and nitrogen in small streams. Other impacts may extend to the destruction of carbon and nitrogen cycles between aquatic and terrestrial ecosystems.
The information obtained from the medium membrane test allowed us to evaluate the ecological relevance of fipronil compound concentrations measured in large-scale field studies conducted in five regions of the United States. In 444 small streams, 17% of the average concentration of one or more fipronil compounds (average over 4 weeks) exceeded the HC5 value obtained from the media test. Use the SSD from the meso-scale experiment to convert the measured fipronil compound concentration into a toxicity-related index, that is, the sum of toxicity units (ΣTUFipronils). The value of 1 indicates toxicity or the cumulative exposure of fipronil compound exceeds the known protection Species worth 95%. The significant relationship between ΣTUFipronil in four of the five regions and the SPEARpesticides indicator of invertebrate community health indicates that fipronil may adversely affect benthic invertebrate communities in rivers in multiple regions of the United States. These results support the hypothesis of Wolfram et al. (3) The risk of phenpyrazole insecticides to surface waters in the United States is not fully understood because the impact on aquatic insects occurs below the current regulatory threshold.
Most streams with fipronil content above the toxic level are located in the relatively urbanized southeast region (https://webapps.usgs.gov/rsqa/#!/region/SESQA). The previous assessment of the area not only concluded that fipronil is the main stressor affecting the invertebrate community structure in the creek, but also that low dissolved oxygen, increased nutrients, flow changes, habitat degradation, and other pesticides and The pollutant category is an important source of stress (57). This mixture of stressors is consistent with the “urban river syndrome”, which is the degradation of river ecosystems commonly observed in relation to urban land use (58, 59). Urban land use signs in the Southeast region are growing and are expected to increase as the population of the region grows. The impact of future urban development and pesticides on urban runoff is expected to increase (4). If urbanization and the use of fipronil continue to grow, the use of this pesticide in cities may increasingly affect stream communities. Although the meta-analysis concludes that the use of agricultural pesticides threatens global stream ecosystems (2, 60), we assume that these assessments underestimate the overall global impact of pesticides by excluding urban uses.
Various stressors, including pesticides, can affect macroinvertebrate communities in developed watersheds (urban, agricultural and mixed land use) and may be related to land use (58, 59, 61). Although this study used the SPEARpesticides indicator and aquatic organism-specific fipronil toxicity characteristics to minimize the impact of confounding factors, the performance of the SPEARpesticides indicator may be affected by habitat degradation, and fipronil can be compared with other Pesticide related (4, 17, 51, 57). However, a multiple stressor model developed using field measurements from the first two regional studies (Midwestern and Southeastern) showed that pesticides are an important upstream stressor for macroinvertebrate community conditions in wading rivers. In these models, important explanatory variables include pesticides (especially bifenthrin), nutrients and habitat characteristics in most agricultural streams in the Midwest, and pesticides (especially fipronil) in most cities in the southeast. Changes in oxygen, nutrients and flow (61, 62). Therefore, although regional studies attempt to address the impact of non-pesticide stressors on response indicators and adjust the predictive indicators to describe the impact of fipronil, the field results of this survey support fipronil’s view. ) Should be considered one of the most influential sources of pressure in American rivers, especially in the southeastern United States.
The occurrence of pesticide degradation in the environment is rarely documented, but the threat to aquatic organisms may be more harmful than the parent body. In the case of fipronil, field studies and meso-scale experiments have shown that degradation products are as common as the parent body in the sampled streams and have the same or higher toxicity (Table 1). In the medium membrane experiment, fluorobenzonitrile sulfone was the most toxic of the pesticide degradation products studied, and it was more toxic than the parent compound, and was also detected at a frequency similar to that of the parent compound. If only the parent pesticides are measured, potential toxicity events may not be noticed, and the relative lack of toxicity information during pesticide degradation means that their occurrence and consequences may be ignored. For example, due to lack of information on the toxicity of degradation products, a comprehensive assessment of pesticides in Swiss streams was carried out, including 134 pesticide degradation products, and only the parent compound was considered as the parent compound in its ecotoxicological risk assessment.
The results of this ecological risk assessment indicate that fipronil compounds have adverse effects on river health, so it can be reasonably inferred that adverse effects can be observed anywhere where fipronil compounds exceed the HC5 level. The results of mesoscopic experiments are independent of location, indicating that the concentration of fipronil and its degradation products in many stream taxa is much lower than previously recorded. We believe that this discovery is likely to be extended to the protobiota in pristine streams anywhere. The results of the meso-scale experiment were applied to large-scale field studies (444 small streams composed of urban, agricultural, and land mixed uses across five major regions in the United States), and it was found that the concentration of many streams where fipronil was detected is expected to be The resulting toxicity suggests that these results may extend to other countries where fipronil is used. According to reports, the number of people using Fipronil is increasing in Japan, the UK and the US (7). Fipronil is present on almost every continent, including Australia, South America and Africa (https://coherentmarketinsights.com/market-insight/fipronil-market-2208). The results of the meso-to-field studies presented here indicate that the use of fipronil may have ecological significance on a global scale.
For supplementary materials for this article, please see http://advances.sciencemag.org/cgi/content/full/6/43/eabc1299/DC1
This is an open access article distributed under the terms of the Creative Commons Attribution-Non-Commercial License, which allows the use, distribution and reproduction in any medium, as long as the final use is not for commercial gain and the premise is that the original work is correct. Reference.
Note: We only ask you to provide your email address so that the person you recommend to the page knows that you want them to see the email and that it is not spam. We will not capture any email addresses.
This question is used to test whether you are a visitor and prevent automatic spam submission.
Janet L. Miller, Travis S. Schmidt, Peter C. Van Metre, Barbara Mahler ( Barbara J. Mahler, Mark W. Sandstrom, Lisa H. Nowell, Daren M. Carlisle, Patrick W. Moran
Studies have shown that common pesticides that are frequently detected in American streams are more toxic than previously thought.
Janet L. Miller, Travis S. Schmidt, Peter C. Van Metre, Barbara Mahler ( Barbara J. Mahler, Mark W. Sandstrom, Lisa H. Nowell, Daren M. Carlisle, Patrick W. Moran
Studies have shown that common pesticides that are frequently detected in American streams are more toxic than previously thought.
©2021 American Association for the Advancement of Science. all rights reserved. AAAS is a partner of HINARI, AGORA, OARE, CHORUS, CLOCKSS, CrossRef and COUNTER. ScienceAdvances ISSN 2375-2548.