Events

Christopher Holmes, Thorsten Schad, Sascha Bub, Magnus Wang, Klaus Hammel, Jane Tang, Gregor Ernst, Thomas G. Preuss

Natural and semi-natural habitats of soil living organisms in cultivated landscapes can be subject to unintended exposure by active substances of pesticides used in adjacent fields. Spray-drift deposition and runoff are considered major exposure routes into such off-field areas. In this work we develop a model (xOffFieldSoil) and associated scenarios to estimate exposure of off-field soil habitats at flexible levels of realism (https://github.com/xlandscape/xOffFieldSoilRisk). The modular approach consists of components each addressing a specific aspect of exposure processes, e.g., pesticide use, drift deposition, runoff generation and filtering, and PECsoil estimation. The approach is spatiotemporally explicit and operates at scales ranging from local edge-of-field to large landscapes. The outcome can be aggregated and presented as comprehensible endpoints to the risk assessor, considering ecologically relevant dimensions and scales for species protection. The approach also assesses the effect of mitigation options, e.g., field margins, in-field buffers, or drift-reducing technology. Presented scenarios start with a schematic edge-of-field situation and extend to real-world landscapes of up to 5 km x 5 km. A case study was conducted for two active substances of different environmental fate characteristics. Results are presented as a collection of percentiles over time and space, as contour plots and as maps. The results show that exposure patterns of off-field soil organisms are of a complex nature due to spatial and temporal variabilities combined with landscape structure and event-based processes. Our concepts and preliminary analysis demonstrate that more realistic exposure data can be meaningfully consolidated to serve in standard-tier risk assessments. The real-world landscape-scale scenarios indicate risk hot-spots which support the identification of efficient risk mitigation. As a next step, the spatiotemporally explicit exposure data can be directly coupled to ecological effect models (e.g., for earthworms or collembola) to conduct risk assessments at biological entity levels.

Christopher Holmes, Steve Kay, Matt Kern, Logan Insinga, Dana Christian, Jim Cowles, Kevin Henry

Estimation of potential aquatic exposure is an important component in the development of Biological Opinions (BOs) for pesticides within an endangered species assessment.  Species-specific and geospatially oriented information should be used to identify localized areas where predicted environmental concentrations may exceed a level of concern. These areas can then be addressed with mitigations to avoid jeopardy to the species or adverse modification to critical habitat. Due to their migratory nature, many salmonid species have expansive ranges covering multiple states, with various life stages present in the pacific ocean as well as inland waters. Recent BOs from NMFS utilize a 300-m pesticide use limitation area (PULA) around salmonid habitats where mitigations are intended to avoid Jeopardy to the species. Previously the Generic Endangered Species Task Force (GESTF) have developed and presented an NHDPlus-based aquatic modeling approach based on EPA’s Pesticides in Water Calculator (PWC) refined with spatial proximity and cropping density. This PWC+ model has been enhanced to focus specifically on the 300-m zone and provide additional refinement on localized stream segments as a species-centric approach with greater resolution than that available using the previous NHDPlus catchments. Within the refined PWC+ model, aquatic habitats have been sub-divided into species-relevant lengths (e.g., 100-m lengths) to further identify potential “hot spots” of exposure. In addition, the influence of cropping density has been limited to the area within the 300-m proximity zone. This focuses the analysis on the pesticide use sites most likely to contribute significant loadings and avoid possible ‘dilution’ effects of more distant cropping areas. Outputs identify the areas that may be suitable candidates for mitigations, while also identifying other areas that may not require additional mitigations beyond those already on the label due to lack of nearby use sites and/or low density of overall cropping proximate to habitat assessment units (e.g., 100-m segments).  The PWC+ approach can be efficiently applied across species and pesticides and represents a viable and practical option based on available resources. This presentation will briefly summarize the previously developed PWC+ model, discuss the rationale for the model modifications, and illustrate the results using real-world situations based on the recent Biological Opinion for carbamates.

Logan Insinga¹, Steve Kay², Dean Desmarteau³

1. Applied Analysis Solutions LLC
2. Pyxis Regulatory Consulting
3. Waterborne Environmental Inc.

The USEPA’s Pesticide in Water Calculator (PWC) simulates pesticide applications to land surfaces and subsequent transport to and fate in waterbodies.  When using PWC, one significant issue is accurately modeling pesticides as they typically have diverse uses and restrictions that vary based on the use site, region, and time of year.  These uses and restrictions are codified in the product label and can be highly interrelated with many permutations of potential applications. Therefore, defining conservative yet label-compliant use-patterns can be difficult and time consuming.  Failure to account for all labeled agronomic restrictions including total pesticide amount applied and number of application limitations on an annual and interval specific (e.g., pre-emergence or post-emergence) basis, as well as multiple application rates with rate-specific limitations (i.e., number of applications and temporal windows), minimum reapplication intervals, and pre-harvest intervals can lead to modeling of non-label-compliant or unrealistic application use patterns.  In addition, efforts to ensure modeling is conservative (e.g., simulation of applications during wettest months of the year or maximizing total amount applied in a year) can significantly complicate application date assignment logic.  To address these difficulties, the Generic Endangered Species Task Force (GESTF) has developed the PWC-PREP Tool to facilitate batch file preparation for PWC modeling.  Specifically, the PWC-PREP Tool features a robust algorithm that generates label-compliant application dates and rates for a wide variety of use sites, regions, and chemicals.  Leveraging a graphical user interface, the tool is highly configurable and allows the user to control modeling parameters related to application methods, drift factors, agronomic restrictions, and date-assignment. The PWC-PREP Tool to has been successfully used to efficiently parameterize a national set of PWC runs (n > 30,000) that follow label instructions in a manner that is transparent and can be repeated.

Christopher Holmes¹, Thorsten Schad², Sascha Bub³, Thomas Preuss²

1. Applied Analysis Solutions, Winchester, VA, US
2. Bayer AG, Crop Science Division, Environmental Safety, D-40789 Monheim, Germany
3. University of Koblenz-Landau, Germany

Abstract:

Bee effect modelling has become a core instrument in bee risk assessment (EFSA 2013 'Guidance Document on the risk assessment (RA) of plant protection products (PPP) on bees', EFSA 2018 'A systems-based approach to the environmental RA of multiple stressors in honey bees', EFSA 2021 'Outline of the revision of the Guidance on the RA of PPPs on bees'). Pesticide RA using such models is based on scenarios. In this project we (i) identified key conceptual elements of scenarios in regulatory RA, (ii) developed proposals for these elements, and (iii) generated scenarios for the BEEHAVE model to demonstrate these elements. The example assumed a RA for honeybees related to the use of a PPP in apples located in France. As apple cultivation density increases so does attractiveness of sites for honeybee keeping (occurrence) and exposure potential. Therefore scenario representativeness (level of conservatism) was mainly driven by regional apple cultivation density. Three regions were selected located in different climatic regions to account for weather variability. In these regions, local site selection (i.e., placing the beehive) was done in a combination of bee forage mapping at medium resolution and beekeeper preference. For each site, scenarios were constructed for a 9 km radius around the beehive. In view of a scenario development framework, we propose a tiered scenario development scheme to implement a well-defined level of complexity together with a targeted certainty evaluation. We developed structured bee forage information layers for transparent bee forage modelling: (i) Land use/cover, (ii) Vegetation, (iii) BeeForage. All information layers are spatially and temporally explicit. The transition from one information layer to the next was done by explicit modules, e.g., the BeeForage module represents an approach on how nectar and pollen provision is modelled for a given vegetation and its phenology. According to the tiered scenario scheme, simple modules can be replaced with more sophisticated ones if needed and available. In our example we start with simple representations of processes, e.g., bee forage (nectar and pollen) provision is modelled in five categories (0-4, 4=mass forage) represented in a lookup table defined from literature, by vegetation patches with a monthly resolution vegetation phenology. Due to current technical limitations of BEEHAVE, the spatiotemporally explicit BeeForage(x,t) information is aggregated into <500 units.   View Poster

Jillian LaRoe¹

1. Applied Analysis Solutions LLC, Winchester, VA, USA

Abstract:

The invasive species habitat tool, INHABIT, is a publicly available web tool developed by the U.S. Geological Survey (Fort Collins Science Center, CO) to model habitat suitability for terrestrial invasive vegetation species across the continental United States. Due to increasing challenges of managing invasive species as well as their ecological and economic impacts, best practices have called for proactive management to prevent spread. The primary challenge land managers are often presented with is that broad-scale assessments require comprehensive landscape surveys of species locations, and often little is known about the distribution of new invaders. INHABIT aims to fill the knowledge gap between known distributions of more than 140 invasive species and the potential risk they pose by modeling broad-scale habitat suitability (90-meter spatial resolution) at four different thresholds, from more conservative estimates to more prolific potential ranges. Using species distribution modeling techniques within a semiautomated workflow, the INHABIT maps ensemble five different modeling algorithms and two different background point selection methods using ecologically relevant predictors specific to each target species. The web tool provides users with the ability to investigate species relevant to different management boundaries, including distance metrics to known species occurrences and summaries of the suitable habitat within a given management area. The habitat suitability maps for each species are also publicly available for download to be used in other analyses. INHABIT development efforts were aimed to support the management objectives of the National Park Service; however, the results have the potential to support broad-scale proactive management of many invasive vegetation species (https://gis.usgs.gov/inhabit/).

View Poster

Jillian LaRoe¹, Christopher M. Holmes¹ and Thorsten Schad²

1. Applied Analysis Solutions LLC, Winchester, VA, USA
2. Bayer AG, Crop Science Division, Environmental Safety, D-40789 Monheim, Germany

Abstract:

Grasslands support essential biodiversity and ecosystem services, and their resilience is threatened by climate change and degradation through anthropogenic land use changes. Remotely sensed satellite imagery have previously been utilized to characterize the intensity and usage of grasslands over time and broad geographic areas.

We used synthetic aperture radar (Sentinel-1) and optical multispectral imagery (Sentinel-2) with repeat images captured every six and five days (respectively) to identify grassland changes at a high temporal resolution, as highly managed grasslands may be cut multiple times per season (March – October). Timeseries images were used to characterize grassland type, cutting frequency, and use intensity for over 6,000 grassland parcels in Germany between 2018 and 2021. Spectral indices related to vegetation were derived across grassland patches to create annual metrics. Thresholding methods were applied to annual time series of Sentinel-1 and Sentinel-2 to detect cutting events within each grassland patch. Results from the two methods were combined to improve cut detection confidence. Finally, a random forest modeling approach was used to classify the grassland type and use intensity. Comparable approaches have been used to develop landscape scenarios for honey bee risk assessment and health monitoring, select and characterize habitat for wood mouse (Apodemus sylvaticus) population models for regulatory pesticide risk assessment, and within other studies to characterize insect habitat and decline.

The final statistics provide a suite of metrics that may be useful for assessing the quality of insect habitat, habitat connectivity, and investigating which grassland management practices best support biodiversity. Furthermore, these methods could be refined for monitoring grassland use and intensity of use over time to inform management strategies or assess restoration effectiveness.

View Poster

Christopher Holmes, Steve Kay, Matt Kern, Logan Insinga, Dana Christian, Jim Cowles, Kevin Henry

Abstract:

Prospective aquatic exposure modeling is one of several tools used to assess the potential for impacts in Biological Opinions (BO) for pesticides. The Biological Evaluation (BE) prepared by the EPA for the insecticide carbaryl provides aquatic exposure estimates by using well-established pesticide fate and transport models (e.g., PRZM, VVWM, AgDRIFT) applied to a standardized set of crop/soil/weather scenarios with reported aquatic concentrations generally limited to 30 annual maxima values. Variability in aspects such as spatial proximity, the incorporation of Percent Crop Area, distribution of pesticide treated acres, and temporal aspects of exposure have been identified as areas for possible refinement in recent BEs and BOs and are employed to better inform the risk evaluation. This paper describes a highly efficient, structured approach that builds on the EPA’s aquatic modeling to increase the spatial/temporal context and resolution of exposure estimates to listed species and produces well-defined and reproducible species-specific estimated aquatic concentrations. Utilizing data for each species, the set of NHD+ catchments, landscape-based crop proximity and density, pesticide usage (e.g., Percent Crop Treated), and additional user-selected options (e.g., PCT multiplier, PCA multiplier, alternate application rates), inputs were processed to produce refined exposure concentrations suitable for aggregation at multiple spatial scales appropriate for the species being examined (e.g., range, critical habitat, spawning areas, recovery areas, existing conservation programs). Temporal aspects were examined using multiple endpoints of the 30-year dataset (e.g., 1-in-15 year, median of annual maxima, 95th percentile of daily) as well as daily distributions. Individual PWC output files (labeled use/HUC/aquatic bin/proximity/application method) were scanned using Python scripts to develop summary concentration profiles for differing time periods and distribution points to provide inputs for the refined modeling. The landscape-scale aquatic modeling results provide valuable information to help inform localized mitigations for individual species and across specific use patterns.

View Presentation

Logan Insinga¹, Steve Kay², Dean Desmarteau³

1. Applied Analysis Solutions, LLC
2. Pyxis Regulatory Consulting, Inc.
3. Waterborne Environmental, Inc.

Abstract:

Endangered species risk assessments for pesticides use prospective models to simulate exposure and aquatic environmental fate. The USEPA’s Pesticide in Water Calculator (PWC) simulates pesticide applications to land surfaces using the Pesticide Root Zone Model (PRZM) and the subsequent transport to and fate in surface water bodies using the Variable Volume Water Model (VVWM). Pesticides typically have diverse uses and restrictions that vary based on the use site, reapplication intervals, region, and time of year. Although PWC does accept a batch file to assist in automation, manual parameterization of numerous model runs is time-prohibitive and prone to human error. Specifically, defining application information (i.e., application dates and rates) is cumbersome due to the agronomic restrictions that are unique to use site and region. Agronomic restrictions include total pesticide amount and number of application limitations on an annual and interval specific (e.g., pre-emergence or post-emergence) basis, as well as multiple application rates with rate-specific limitations (i.e., number of applications and temporal windows), minimum reapplication intervals, and pre-harvest intervals. In addition, efforts to ensure modeling is conservative (e.g., simulation of applications during wettest months of the year and maximizing application rates) further complicate date assignment logic. The Generic Endangered Species Task Force (GESTF) have developed the PWC Tool to automate batch file preparation for PWC modeling. Specifically, the PWC Tool features a robust algorithm that generates label-compliant application dates and rates for a wide variety of use sites and regions. Landscape scale refinements can be implemented via the tool’s graphical user interface where the user defines drift factors that correspond to various transport mechanisms (e.g., aerial, airblast, ground), distances, and receiving water body. The App Dates Tool to has been successfully used to parameterize a national set of PWC runs (n > 30,000) in a manner that was efficient, transparent, and repeatable.

GitHub

View Poster

Christopher Holmes, Steve Kay, Matt Kern, Logan Insinga, Dana Christian, Dean Desmarteau, Jim Cowles, Kevin Henry

Abstract:

Prospective aquatic exposure modeling is one of several tools used to assess the potential for impacts in Biological Opinions (BO) for pesticides. The Biological Evaluation (BE) prepared by the EPA for the insecticide carbaryl provides aquatic exposure estimates by using well-established pesticide fate and transport models (e.g., PRZM, VVWM, AgDRIFT) applied to a standardized set of crop/soil/weather scenarios with reported aquatic concentrations generally limited to 30 annual maxima values. Variability in aspects such as spatial proximity, the incorporation of Percent Crop Area, distribution of pesticide treated acres, and temporal aspects of exposure have been identified as areas for possible refinement in recent BEs and BOs and are employed to better inform the risk evaluation. This paper describes a highly efficient, structured approach that builds on the EPA’s aquatic modeling to increase the spatial/temporal context and resolution of exposure estimates to listed species and produces well-defined and reproducible species-specific estimated aquatic concentrations. Utilizing data for each species, the set of NHD+ catchments, landscape-based crop proximity and density, pesticide usage (e.g., Percent Crop Treated), and additional user-selected options (e.g., PCT multiplier, alternate application rates), inputs were processed to produce refined exposure concentrations suitable for aggregation at multiple spatial scales appropriate for the species being examined. Temporal aspects were examined using multiple endpoints of the 30-year dataset (e.g., 1-in-15 year, median of annual maxima, 95th percentile of daily) as well as daily distributions. Individual PWC output files (labeled use/HUC/aquatic bin/proximity/application method) were scanned using Python scripts to develop summary concentration profiles for differing time periods and distribution points to provide inputs for refined modeling. Options for results presentation integrated into existing R-Plots from Services’ BOs will be illustrated. The spatially explicit aquatic modeling results provide valuable information to help inform localized mitigations for individual species and across specific use patterns. Aspects of efficiency, transparency and documentation were all addressed in the generation of this framework. This presentation describes the purpose, design decisions, and results of these post-screening refinements as related to the recent carbaryl BE. 

Scott D. Dyer^{1 2}, Raghu Vamshi¹, Christopher M. Holmes³, Nicholas S. Greena,⁴, Brenna Kent¹, Iain Davies⁵

1. Waterborne Environmental, Inc.
2. LeTourneau University
3. Applied Analysis Solutions
4. Kennesaw State University
5. Personal Care Products Council

Abstract:

Recently questions have been raised regarding the environmental safety of some UV filters used in personal, skin care and beach products to corals.  In some cases (Hawaii, Key West, Palau) regulatory actions have been precautionary, leading to bans.  Eco-epidemiology is a methodology that considers impairments to environmental species and communities from complex combinations of multiple physical and chemical factors with the intent of developing weights of evidence for potential causal relationships.  In this study a large set of natural and human influenced factors (including potential risks of measured UV filters and beach visitors) along with coral cover data were assembled to assess the potential adverse effects of UV filters on corals surrounding Oahu, Hawaii within the context of other factors.  Principal component analyses simplified the coral data into two components:  PC1 representing species diversity and PC2 representing abundance.  These components along with all other factors were correlated against each other to determine if some factors could act as proxies for other factors via multiple linear regression and boosted regression tree analyses.  Overall, there were good agreements between the two regression methods.  The boosted regression tree for PC1 (diversity) showed that 90 percent of the variance were addressed by 3 factors:  wave power, temperature (long-term mean) and benthic turf algae.  The remaining 10 percent included 13 other factors including beach visitors and UV filters.  The regression tree for PC2 (abundance) illustrated 3 factors (temperature long-term mean, latitude, and temperature long-term standard deviation) addressed 75% of the variance.  Twelve other factors, including beach visitors and UV filters) were associated with the remaining 25 percent.  Hence, it appears that UV filter hazards do not significantly address reduced coral diversity and abundance, suggesting that precautionary bans may not achieve their intended results.

Christopher Holmes¹, Thorsten Schad², Sascha Bub³, Thomas Preuss²

1. Applied Analysis Solutions, Berryville, VA, US
2. Bayer AG, Crop Science Division, Environmental Safety, D-40789 Monheim, Germany
3. University of Koblenz-Landau, Germany

Abstract:

Bee effect modelling has become a core instrument in bee risk assessment (EFSA 2013 'Guidance Document on the risk assessment (RA) of plant protection products (PPP) on bees', EFSA 2018 'A systems-based approach to the environmental RA of multiple stressors in honey bees', EFSA 2021 'Outline of the revision of the Guidance on the RA of PPPs on bees'). Pesticide RA using such models is based on scenarios. In this project we (i) identified key conceptual elements of scenarios in regulatory RA, (ii) developed proposals for these elements, and (iii) generated scenarios for the BEEHAVE model to demonstrate these elements. The example assumed a RA for honeybees related to the use of a PPP in apples located in France. As apple cultivation density increases so does attractiveness of sites for honeybee keeping (occurrence) and exposure potential. Therefore scenario representativeness (level of conservatism) was mainly driven by regional apple cultivation density. Three regions were selected located in different climatic regions to account for weather variability. In these regions, local site selection (i.e., placing the beehive) was done in a combination of bee forage mapping at medium resolution and beekeeper preference. For each site, scenarios were constructed for a 9 km radius around the beehive. In view of a scenario development framework, we propose a tiered scenario development scheme to implement a well-defined level of complexity together with a targeted certainty evaluation. We developed structured bee forage information layers for transparent bee forage modelling: (i) Land use/cover, (ii) Vegetation, (iii) BeeForage. All information layers are spatially and temporally explicit. The transition from one information layer to the next was done by explicit modules, e.g., the BeeForage module represents an approach on how nectar and pollen provision is modelled for a given vegetation and its phenology. According to the tiered scenario scheme, simple modules can be replaced with more sophisticated ones if needed and available. In our example we start with simple representations of processes, e.g., bee forage (nectar and pollen) provision is modelled in five categories (0-4, 4=mass forage) represented in a lookup table defined from literature, by vegetation patches with a monthly resolution vegetation phenology. Due to current technical limitations of BEEHAVE, the spatiotemporally explicit BeeForage(x,t) information is aggregated into <500 units.

Insinga, L.¹; Holmes, C.M.¹; Schad, T.²

1. Applied Analysis Solutions
2. Bayer CropScience

View Abstract

Insinga, L.¹; Holmes, C.M.¹; Schad, T.²

1. Applied Analysis Solutions
2. Bayer CropScience

View Abstract

Holmes, C.M¹; Kern, M.²; Kay, S.³; Cowles, J.⁴; Henry, K.⁴

1. Applied Analysis Solutions
2. Balance EcoSolutions
3. Pyxis Regulatory Consulting
4. NovaSource

Abstract:
Prospective aquatic exposure modeling is a key aspect in the assessment of potential jeopardy of endangered species during the preparation of a Biological Opinion (BO) related to the federal action of pesticide (re)registration.  The Biological Evaluation (BE) prepared by the EPA provides the foundation of aquatic exposure estimates using well-established pesticide fate and transport models (e.g., PWC, PRZM, VVWM, AgDRIFT) applied to a standardized set of crop/soil/weather scenarios. Exposure estimates are based on 30 years of model simulation implementing labeled uses of the pesticide specifying maximum application rates, number of applications, re-application intervals and assuming treated crop is adjacent to the aquatic environment.  This screening-level approach forms the basis from which subsequent refinements incorporating spatial, temporal, and pesticide usage variability can be applied to better inform the weight of evidence process. This presentation will describe the purpose, design decisions, assumptions and uncertainties, along with results of these post-screening refinements as related to the recent carbaryl BE. Variability in such aspects as spatial proximity, the incorporation of Percent Crop Area, distribution of pesticide treated acres, and temporal aspects of exposure have been identified as areas for possible refinement in recent BEs and BOs as well as agency guidance documents. The implementation of these refinements shows that the 1-in15 year EECs based on screening level assumptions of pesticide use and hypothetical water body scenarios may occur at some locations at limited times, but they are far less likely to occur (within species range/habitat) than is assumed in screening-level risk assessments. The methodologies applied within these refinements allow for a quantitative approach for incorporating variations in landscape and agronomic practices near endangered species locations.

Session:
Advancing Endangered Species Risk Assessment and Mitigation from National-scale to Species-centric Assessments Using 'Best Available' Data

Christian, D.¹; Holmes, C.M.¹; Kay, S.²; Kern, M.³; Cowles, J.⁴; Henry, K.⁴

1. Applied Analysis Solutions
2. Pyxis Regulatory Consulting
3. Balance EcoSolutions
4. NovaSource

Abstract:
The spatial relationship between potential pesticide use sites and Endangered Species locations (e.g., range, critical habitat, suitable habitat, known locations) determines exposure and is an influential factor in evaluating potential risk to these species.  We developed models and processing approaches to characterize spatial relationships in an efficient and documented manner. As a first step, we characterized potential pesticide use sites (i.e., Use Data Layers, UDLs) in relation to species aquatic habitats. As a second step, we defined the spatial extent of species habitats using the NHD+ framework of flowlines, catchments and static water bodies which formed the basis for our spatial definition of habitat. Each water catchment was characterized in terms of total UDL area forming the basis for Percent Crop Area (PCA) refinements applied to baseline PWC Estimated Environmental Concentrations (EECs) concentrations. Further, eight proximity zones (PZs) around the water bodies were spatially defined in order to quantify UDL areas within each PZ. The use of proximity zones is significant for refinement of ESA assessments as PZ cropping density can incorporate relevant landscape-based drift deposition and runoff/erosion information.  We also quantified, for each catchment and UDL, the proportion of crop within the catchment that was greater than 300m (~1000’) or 900m (~2600’) beyond which spray drift deposition no longer contributes to the water body EEC (from ground and aerial applications, respectively). The ability to characterize variability in potential exposure within and between catchments enabled a quantitative and visual distribution of potential exposure across a species range or critical habitat. This landscape-specific information supports spatially-informed, species-specific exposure analyses as an essential line of evidence for the Biological Opinion. This presentation will describe the design and practical implementation of spatial proximity and cropping density approaches for listed species.

Insinga, L.¹; Holmes, C.M.¹; Desmarteau, D.²; Kay, S.³; Kern, M.⁴; Cowles, J.⁵; Henry, K.⁵

1. Applied Analysis Solutions
2. Waterborne Environmental
3. Pyxis Regulatory Consulting
4. Balance EcoSolutions
5. NovaSource

Abstract:
The challenge of evaluating exposure potential from multiple labeled uses for over 1600 Endangered Species nationally can be intimidating. It is important to be as spatially detailed as possible in order to identify potential product use areas and regional factors that are most relevant to each species being evaluated. In addition, with over 1600 species being evaluated for a multitude of labeled uses modeled on 13 agricultural Use Data Layers (UDLs), efficiency and reproducibility are critical.  We developed a highly efficient and structured approach that incorporates best available information in order to increase the reliability and relevance of exposure estimates to listed species.  The refinement framework incorporates additional dimensions of aquatic exposure modeling such as cropping density and proximity, pesticide treated acre distribution, and consideration of temporal exposure. A structured and programmatic framework was implemented to build upon the baseline scenario concentrations, and apply spatial, temporal and agronomic aspects to produce well-defined and reproducible species-specific aquatic concentrations.  In this presentation we will describe how daily data from thousands of individual PWC output files (labeled use / HUC / aquatic bin / proximity / application method) were scanned using Python scripts to develop summary concentration profiles for differing time periods and distribution points.  Utilizing those data for each species, the set of NHD+ catchments, landscape-based crop proximity and density, pesticide usage (e.g., Percent Crop Treated), and user options (e.g., PCT multiplier, alternate application rates) were processed to produce refined exposure concentrations suitable for aggregation at multiple spatial scales appropriate for the species being examined.  Transparency of input data, modeling run options and availability of raw and formatted results was paramount in this process and resulted in the ability to rapidly process and document alternative cropped areas, application rates, intervals and other variables.

Kern, M.¹; Kay, S.²; Holmes, C.M.³; Cowles, J.⁴; Henry, K.⁴

1. Balance EcoSolutions
2. Pyxis Regulatory Consulting
3. Applied Analysis Solutions
4. NovaSource

Abstract:
The protection of federally listed threatened and endangered species requires a multidisciplinary approach that leverages best available data to accurately identify and address risk to species and their habitats. Assessment and management uncertainty can be significantly reduced by applying localized spatial and temporal landscape information along with information about potential stressors. Several pieces of information should be collected at the onset of each assessment to address potential risk including 1) an understanding of where the species and its habitats are most likely to occur in the landscape, 2) identification of factors or conditions required to support the species, 3) knowledge and prioritization of stressors already identified as influential to species status, 4) a determination of where the stressor being evaluated is likely to occur on the landscape relative to species and its habitats, 5) an assessment of the potential degree to which the stressor may influence the species. This information can be used for targeted management strategies regarding the unique relationship between the species and stressors being evaluated. State and federal agencies often collect extensive information on species status, requirements, habitats, influencing stressors and recovery goals which can be used to develop the information set described above. In cases where pesticides are being evaluated as a potential stressor, registrants often have extensive product knowledge that can be applied to further reduce uncertainty in these assessments. We present an approach using a model insecticide that integrates species and product information at a local, species-centric scale. Several examples are presented that outline a pragmatic approach to reducing assessment uncertainty so that species management strategies and resources can be effectively applied. These general approaches can be used for many pesticides and will facilitate a more efficient process for evaluating risk that may, or may not, be reasonably certain to occur for species.

Session:
Advancing Endangered Species Risk Assessment and Mitigation from National-scale to Species-centric Assessments Using 'Best Available' Data

Kay, S.¹; Kern, M.²; Holmes, C.M.³; Cowles, J.⁴; Henry, K.⁴

1. Pyxis Regulatory Consulting
2. Balance EcoSolutions
3. Applied Analysis Solutions
4. NovaSource

Abstract:
The process to assess the potential of a pesticide to jeopardize endangered species continues to evolve with input from regulatory agencies, industry, non-governmental organizations, and other stake holders. The current approach does not consider possible measures (or “avoidance”) to help address potential jeopardy until the very end of the process when the Services might issue Reasonable and Prudent Alternatives (RPAs) intended to reduce exposure and/or Reasonable and Prudent Measures (RPMs) intended to minimize “incidental take”. However, proposed avoidance measures could be integrated earlier in the assessment process to develop targeted, species-specific solutions that address potential jeopardy concerns while avoiding unnecessary impacts to growers. This presentation will describe a proposed strategy for identifying the potential need for avoidance in a targeted manner, possible ways to incorporate that information into the ESA assessment as early in the process as possible, and how the avoidance can be implemented.

Session:
Advancing Endangered Species Risk Assessment and Mitigation from National-scale to Species-centric Assessments Using 'Best Available' Data

Dyer, S.D.¹ ²; Vamshi, R.¹; Kent, B.¹; Abi-Akar, F.¹; Holmes, C.M.³; Green, N.S.¹ ⁴; Davies, I.⁵

1. Waterborne Environmental, Inc.
2. LeTourneau University
3. Applied Analysis Solutions
4. Kennesaw State University
5. Personal Care Products Council

View Abstract

Holmes, C.M.¹; Insinga, L.¹; Sweeney, P.²; Loiseau, L.³; Johnson, D.²

1. Applied Analysis Solutions, Berryville, VA, USA
2. Syngenta, Jeallot’s Hill, Bracknell, UK
3. Syngenta AG, Basel, CH

View Presentation Slides

View Extended Abstract

Holmes, C.M.¹; Marshall, S.²; Maltby, L.³; Sweeny, P.⁴; Thorbek, P.⁵; Otte, L.C.⁶

1. Applied Analysis Solutions LLC, VA, USA
2. Consultant, Bedford, UK
3. University of Sheffield, UK
4. Syngenta, Bracknell, UK
5. BASF, Limburgerhof, Germany
6. BASF, Ludwigshafen, German

View Presentation Slides

View Extended Abstract

Surfactant case study: Increasing the ecological relevance of chemical risk assessments using geospatial approaches

View Surfactant Poster

View Surfactant Case Study Abstract

PPP case study: Increasing the ecological relevance of chemical risk assessments using geospatial approaches

View PPP Poster

View PPP Case Study Abstract

Holmes, C.E.¹ ²; Löw, F.; Schad, T.³; Holmes, C.M.²

1. Virginia Polytechnic Institute and State University, Blacksburg, VA, USA
2. Applied Analysis Solutions, Berryville, VA, USA
3. Bayer CropScience, Monheim, Germany

View Presentation Slides

View Extended Abstract

Holmes, C.M.¹; Tandy, E.¹; Schleicher, Z.²; Kilgallon, J.³ ; Hodges, J.³

1. Applied Analysis Solutions, Berryville, VA, USA
2. Southern Illinois University, Edwardsville, IL, USA
3. Unilever Safety & Environmental Assurance Centre, Sharnbrook, UK

Abstract:

Domestic wastewater discharged to the environment may contain chemical substances that are the result of everyday human activity such as personal care, home cleaning and using medications. While most discharge is to freshwater rivers and lakes, some facilities in coastal areas discharge into the marine environment. As population densities continue to grow in coastal areas, understanding where these marine emissions occur is increasingly important for assessing both marine and freshwater ecological risk. Making this type of evaluation across large spatial scales is challenging though because information on wastewater treatment plant (WWTP) location and discharge type is often lacking. To address this data gap, we developed a global-scale dataset focused on marine emissions with the goal of understanding what fraction of the population in coastal areas may be discharging to a marine environment via a WWTP. Information on WWTP location, population served, and effluent flow were collected for 44,000 WWTPs across 34 countries serving one billion people. Using these training data, we developed a spatial extrapolation method utilizing urban and rural populations, their likelihood of purchasing products based on economic productivity, and how far they live from the coast. Using global spatial datasets, extrapolation was possible from our training data (e.g., an entire country or only a selected portion thereof) to any target country that lacks source data for WWTPs. The extrapolation results in aggregated estimates (e.g., administrative units, hydrologic basins) of the fraction of population connected to WWTPs discharging directly to the marine environment. This poster will present information on the source WWTP data collected, global spatial data utilized, extrapolation methods examined with results, along with an evaluation of results using data outside the training set. A spatial understanding of the WWTP effluent fraction discharged to the marine environment not only increases our ability to assess and protect near shore marine habitat, it also refines emission estimates for freshwater ecological risk scenarios used in regulatory assessments.

View Presentation Slides

Holmes, C.M.¹; Pritsolas, J.²; Pearson, R.²; Butts-Wilmsmeyer, C.²; Schad, T.³

1. Applied Analysis Solutions, Berryville, VA, USA
2. Southern Illinois University, Edwardsville, IL, USA
3. Bayer CropScience, Monheim, Germany

View Presentation Slides

View Extended Abstract