PresentationsCrop Protection, Water/Wastewater Assessments2018
Interpreting water quality monitoring observations through modeling: PRZM/SWAT and SEAWAVE-Q
Room: Ballroom East – Theater 4
Date: Wednesday August 22, 2018
Start Time: 10:35AM
Presentation Code: AGRO 231
Water quality monitoring data, specifically pesticides, can represent best-available exposure profiles related to ecological risk assessment; however, there are challenges in synthesizing these data toward making strong conclusions about the nature of the potential range of risk. Some of these challenges include: data collection frequency, monitoring period duration, interpreting exposure profiles from one location to another, and monitoring system scale. To address these challenges, statistical approaches may be applied to characterize empirical trends. Additionally, process-based numerical systems modeling approaches can offer a different perspective on synthesis and interpretation of monitoring data. A comparison of a statistical and process-based numerical model was conducted to evaluate strengths and weaknesses of representing, synthesizing, and conclusions from monitoring data. This comparison was developed from pesticide measurements of six intensively-monitored HUC12 headwater watersheds in the Midwest. SEAWAVE-Q is a regression model that incorporates a linear trend term, covariates accounting for seasonality, and a transformation of flow to represent a long-term pesticide trend at a specific monitoring location. PRZM/SWAT (pesticide root zone model/soil water assessment tool) is a spatially-distributed hydrologic and chemical transport numerical model that combines upland chemical and hydrologic processes from PRZM and stream flow and chemical transport processes from SWAT. The predictive quality and limitations of these two models was assessed against observed, daily concentration measurements as well as hypothetical data collection frequency and timing (derived from sub-sampling the same data sets). Results suggest that a process-based modeling approach, such as PRZM/SWAT, may be more advantageous when calibration data are available.
Daniel Perkins, Andy Jacobson, Colleen Roy, Farah Abi-Akar (Waterborne Environmental), Wenlin Chen (Syngenta Crop Protection). Interpreting water quality monitoring observations through modeling: PRZM/SWAT and SEAWAVE-Q. ACS 2018. Presentation. Boston, MA.
Evaluation of applied, cross-sector vegetative best management practices in rights-of-way on pollinators
Transmission and pipeline rights-of-way (ROWs) in the U.S. are estimated to occupy approximately 21 million acres, roughly equivalent to the average harvested corn acres in Missouri and Illinois combined. Additionally, there are nearly 4 million miles of roadside ROW and 233 thousand miles of railroad ROW in the U.S. These ROWs dissect agricultural land, urban areas, and natural areas like forests and grasslands. Managed ROWs represent an opportunity to provide habitats for numerous species of plants and animals, including pollinators that provide critical ecosystem services for farms, natural areas, and private homeowners. Strategic creation of pollinator habitat in ROW with respect to cross-sector land management offers large potential benefits to agricultural production, conservation, education, and research. In this work, we hypothesize that the success of managed ROWs to provide quality pollinator habitat is based on a combination of integrated vegetation management (IVM) practices within ROWs, physical habitat characteristics, surrounding land use composition specific to geography/ecoregion characteristics, and pollinators of interest. We evaluate an experimental design aimed at identifying the effects of different IVM practices in ROWs on pollinators. We rely on variabilities in IVM practices and physical characteristics of ROWs reported in literature and other sources to statistically determine which design elements are important to discern potential effects on pollinator habitat quality. The findings will promote informed evaluation of conservation management strategies. Appropriate statistical and field study designs aimed to characterize pollinator success based on IVM practices would allow ROW managers to gain quantitative information to better understand where and when to establish IVMs across spatially explicit, complex and diverse landscapes. Additionally, agricultural land managers may benefit from implementation of certain IVM practices in ROWs that will positively impact crop production while promoting conservation of pollinators and other species.
Farah Abi-Akar, Daniel Perkins, Joshua Amos, Amelie Schmolke (Waterborne Environmental), Stan Vera-Art (Grow With Trees), Iris Caldwell (University of Chicago). Evaluation of applied, cross-sector vegetative best management practices in rights-of-way on pollinators. ACS 2018. Presentation. Boston, MA.
Refined land cover for improving the confidence of pesticide risk assessments
Today, field-scale crop location data are available at higher spatial resolution and classification accuracy than ever before. These data represent “best available scientific data” and can improve the confidence of human health and ecological risk assessments by providing more realistic, site-specific exposure estimates. For example, state-wide agricultural-focused land cover on an annual basis with 56 crop classes and a spatial resolution of ~2 meters is now available in California. Given that California ranks first in total crop receipts in the US and bounds nearly 20% of all federally listed species, increased spatial land use data resolution has the potential for measurable impact in risk assessment. Other novel opportunities exist for incorporating better data by cooperating with specialty crop grower groups who often have field surveys locating their member’s farms (e.g., Florida green beans). High-confidence data can be used to improve Percent Crop Area (PCA) adjustment factors required for estimating drinking water concentrations in human health risk assessments. From a FIFRA/ESA perspective, it is proving difficult to make determinations for many crop-species combinations with any degree of confidence using the spatial information presently available. Although inconvenient, there is arguably a need to incorporate better data as it becomes available.
To demonstrate the impact of the newly-available California land use data, a case study was conducted to contrast the effects of refined crop cover data on spatial proximity to non-target terrestrial species’ habitat. From an aquatic perspective, watershed scale (HUC-12) PCAs calculated using the EPA crop use sites, the 2014 CWS PCA guidance document approach, and the CA DWR land cover will be covered. These results can be used to gauge the impact that higher spatial resolution and classification accuracy can have on pesticide risk assessments relying on crop location information.
Daniel Perkins, Joshua Amos. Refined land cover for improving the confidence of pesticide risk assessments. ACS 2018. Presentation. Boston, MA.
Assessing the environmental risk of pesticides, biopesticides, and anthelmintics used in managing vector-borne diseases
Insecticides, biopesticides, and parasitical agents are among the arsenal of tools used to control the transmission of vector-borne diseases. In the United States, these products are regulated by the U.S. Environmental Protection Agency under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) or by the Food and Drug Administration under the Federal Food, Drug, and Cosmetic Act (FD&C Act). Chemical and biological releases to the environment can occur from a variety of mechanisms including aerial or ground applications of pesticides to the landscape, wash-off of ectoparasiticides or excretion of anthelmintics from livestock or pets, and down-the-drain discharges from washing treated clothing. Approaches to evaluating the different delivery mechanisms and their potential adverse impacts to non-target organisms are presented through case studies.
W. Martin Williams, Joshua Amos, Megan White Guevara, Amy M. Ritter. Assessing the environmental risk of pesticides, biopesticides, and anthelmintics used in managing vector-borne diseases. ACS 2018. Poster. Boston, MA.
Using geospatial techniques for effective product stewardship
The goal of this project was to identify agricultural soils reflected in the acetochlor label use restriction, which overlap with shallow ground water. Acetochlor-based products are labeled for use within the United States to control annual grasses and certain broadleaf weeds and can be applied preplant, at-planting, preemergence and/or postemergence to labeled crops. Acetochlor product labels restrict applications within 50 feet of any well where depth to ground water is ≤ 30 ft. The Acetochlor Registration Partnership (ARP – Dow AgroSciences and Monsanto Company) developed voluntary Best Management Practices (BMPs) for acetochlor to reduce the potential for the active substance and its major environmental metabolites in ground water and surface water, following application to agricultural fields. The ARP offers a website (www.arpinfo.com) with resources for end-users to help ensure the effective use and stewardship of products containing acetochlor. Depth to ground water information is often not readily available, which makes it challenging to depict the spatial distribution of areas where the use restriction should be observed. One can point to privacy concerns as a reason for the lack of ground water depth information available to the public. An assessment was conducted to identify crop lands with potential acetochlor use restrictions in Arizona, using geospatial techniques. Arizona’s Department of Water Resources provides historical ground water depth data for point locations via a public monitoring database (http://www.azwater.gov/azdwr/GIS/). Groundwater depth information was extracted from the monitoring database for the period 1995 to 2015. As a data handling procedure, outliers were removed, then cluster and hotspot analyses were performed. The final dataset contained over 12,000-point observations of ground water depth for a 20-year period. Long-term average depths were calculated for each location. To generate a state-wide average groundwater depth map, a spatial interpolation technique was applied to the GIS vector or point dataset. The final ground water map was overlaid with National Agricultural Statistics Service (NASS) Cropland Data Layer (CDL) data to identify agricultural soils with potential concerns. For cotton, the assessment showed that just 3% of the agricultural areas overlap with shallow ground water.
Amy M. Ritter (Waterborne Environmental), Cornelis Hoogeweg (Waterborne Environmental), Mark Anthony Thomas (Monsanto Company), Annette Kirk (Monsanto Company). Using geospatial techniques for effective product stewardship. ACS 2018. Poster. Boston, MA.
Pesticides in Flooded Applications Model (PFAM) ecological modeling sensitivity and the impact of a receiving water body on ecological estimated environmental concentrations
The Pesticide in Flooded Application Model (PFAM) is used to estimate surface water concentrations primarily for pesticide applications to rice paddies. PFAM (version 2.0) has the potential to assess pesticide concentrations in rice paddy water and a receiving water body. However, the Environmental Fate and Effects Division in the EPA currently uses only the in-paddy concentration from the PFAM model for ecological risk assessments. A receiving waterbody such as a canal would be appropriate as a representative aquatic environment for ecological risk assessment of species (e.g. fish) not found in a typical US rice paddy. An assessment was performed using a hypothetical pesticide to conduct PFAM ecological sensitivity runs. The “ECO CA Winter No Turnover” and “ECO MS Winter No Turnover” scenarios were used in the modeling exercise. The simulations were performed with a single application per year on a standard 10-ha paddy. Pesticide concentrations in the paddy were compared with concentrations in two receiving waterbodies (canal and pond). The presentation will show the impact on the estimated environmental concentration (EEC) due to changes in the baseflow, surrounding watershed size and curve number, holding periods, and drift factors. Concentrations in the pond waterbody and canal were significantly lower than the in-paddy concentrations. This presentation highlights refinement options for appropriate aquatic environments that may receive outflow from a rice paddy.
Amy M. Ritter, W. Martin Williams. Pesticides in Flooded Applications Model (PFAM) ecological modeling sensitivity and the impact of a receiving water body on ecological estimated environmental concentrations. ACS 2018. Presentation. Boston, MA.
Modeling chemical partitioning at the water-sediment interface
The varying composition of bed sediments, combined with hydrodynamic and biological perturbations, have created challenges in modeling the partitioning of chemicals at the water-sediment interface in natural waters. A variety of approaches have been developed to predict chemical mass balance between the water column and benthic sediment. These approaches often involve some empirically-derived component to account for the many physical, chemical, biological, and temporarily varying processes that may affect chemical exchanges between water and sediment. This presentation looks at the different deterministic and empirical approaches, and commonly used assumptions, in several water quality models used in regulatory risk assessments and the establishments of TMDLs. Results of several approaches are compared for a variety of water depths, water chemistries, and hydraulic conditions.
W. Martin Williams, Amy M. Ritter. Modeling chemical partitioning at the water-sediment interface. ACS 2018. Presentation. Boston, MA.
Modelling Microplastics in Rivers in the US (339)
Pollution with nano- and microplastics (MPs; particles < 5 mm) is a topic of emerging concern and as such receives growing interest. Although measurement and monitoring data are indispensable, there also is a need for estimated concentrations to enable prospective assessments and to guide analysis of retrospective ecological analyses. Besseling et al (2017) provided the NanoDUFLOW model, a detailed MP aggregation-sedimentation model integrated in a hydrological and particle transport model. A much larger scale model potentially suitable to simulate MPs originating from WWTPs is the iSTREEM® model, which has been developed to estimate chemical concentration distributions for all rivers and streams of the USA receiving WWTP discharges. Here we merge these two riverine modeling worlds: NanoDUFLOW with iSTREEM for MPs, to simulate spreading of MPs from WWTP point sources in US waterways and to assess export to the Great Lakes for a range of particle sizes. This combines the mechanistic realism of NanoDUFLOW, accounting for formation and settling of heteroaggregates, with the US well-established iSTREEM implementation. We modeled floating as well as non-buoyant MP, for diverse sizes, from 100 nm to 10 mm, a range that incorporates the theoretical parabolic size-settling relationship reported by Besseling et al (2017). Depth dependent in-stream first order removal rate constants simulated with NanoDUFLOW were combined with standard iSTREEM output (which was used to simulate the emission, transport and water column concentrations of MP) in an Excel-based post-processing phase, without modifing the iSTREEM model directly. Simulations were spatially explicit with MP concentrations being modeled for the Sandusky River watershed in Ohio (~3500 km2). Emissions were based on per capita usage and population served for each of the 20 WWTPs within the watershed. Modelling results show the effects of population density, MP size and density on riverine concentrations and export to Lake Erie. Buoyant as well as the smallest non-buoyant MP fractions can be transported over long distances, reaching receiving waters such as the Great Lakes. In contrast, larger non-buoyant MPs settle more locally in the vicinity of the WWTPs. Simulating depth-dependent removal as demonstrated here could be incorporated into the core iSTREEM code in order to efficiently process all US waterways impacted by WWTPs, as well as examining ultimate marine discharge proportions by particle size.
A. Koelmans (Wageningen University); C.M. Holmes (Waterborne Environmental). Modelling Microplastics in Rivers in the US. SETAC EU 2018. Presentation.
Implications of Dataset Selection and GIS Processing on Modelling (MO143)
Groundwater assessment guidelines provided by the FOCUS groundwater working group (2009) and EFSA (2014) describe succinctly a multi-tiered modelling framework that includes spatiotemporal assessments in the higher tiers; e.g., tier 3a and 3b. As part of the spatio-temporal assessment several GIS and daily climate datasets were recommended. These recommended datasets, however, have been superseded by new datasets in the past few years. Specifically, daily weather and soils data have undergone significant updates, which are reflective of the considerable effort in Europe to update this spatial information. Not only does dataset choice, but also how datasets are being processed in a geographic information system, impact modeling results. Basic assumptions regarding aggregation of data, data slicing for determining climatic zones and data resolution impact our modelling results. In this poster, we will show the implications of data selection and data processing on a distributed modelling framework centered around GeoPEARL 4R. Specifically we will focus on differences between datasets, data set resolution, capturing variability and ones ability to model at the pan-European level within EFSA’s tier 3 guidelines.
G. Hoogeweg, M. Geuvara (Waterborne Environmental). Implications of Dataset Selection and GIS Processing on Modelling. SETAC EU 2018. Poster.
Development of an European Tier 3+ Spatially Distributed Modelling Framework (MO141)
Higher tier groundwater assessment in the European Union (EU28) allow the use of spatially distributed modeling approaches for the assessment of groundwater and exposure of soil organisms. An advantage of a distributed model is that model inputs can reflect local conditions and capture the spatial variability of the landscape and weather patterns. An advanced modelling framework, based on the GeoPEARL 4R model was developed for the EU28. This model fills the niche for higher Tier assessments needs. This modelling framework represents over 1.340.000 km2 of arable agricultural lands in Europe. Nearly 382.000 unique soil, weather, FOCUS zone combinations represent the variability of the landscape and climate. Datasets to populate the model, included CORINE land cover, soils data (ESDB, ESDB Derived Data for Modelling and HYPRES, EFSA organic matter) and the JRC MARS 25km gridded daily weather data. Agricultural management practices, irrigation, and cropping scenarios are gleaned from the standard FOCUS modelling scenario, but can be updated as needed. This European modeling framework (EMF2014) can be used for EU28, member state, FOCUS zones or crop specific groundwater vulnerability assessments, screening of existing and new plant protection products, context setting of standard scenarios, test sites, and lysimeter, site selection. In this presentation we will show how we developed the framework and several example outputs as well as discuss the implications of conducting largescale distributed modelling assessment.
G. Hoogeweg (Waterborne Environmental); P. Sweeney (Syngenta). Development of an European Tier 3+ Spatially Distributed Modelling Framework. SETAC EU 2018. Poster.