Prairie Pothole Region Wetlands and Subsurface Drainage Systems: Key Factors for Determining Drainage Setback Distances

Wetlands provide numerous ecological services to society and are a focal point of various conservation programs (e.g., U.S. Department of Agriculture Wetlands Reserve Program [NRCS 2018a]), nongovernment organizations (e.g., Ducks Unlimited, Inc.), and regulatory policies (e.g., Swampbuster provision of the 1985 Farm Bill [NRCS 2018b]) throughout the Prairie Pothole Region (PPR) of the United States. The PPR encompasses approximately 800,000 km² in the north-central United States and south-central Canada. It is one of the largest wetland ecosystems in the world and widely recognized for a suite of ecosystem services including provision of critical waterfowl habitats and sequestration of atmospheric carbon (Batt et al. 1989; Euliss et al. 2006; Badiou et al. 2011; Niemuth et al. 2014). Provisioning of these ecosystem services by PPR wetlands is dependent on hydrologic characteristics such as period of inundation, water depth, and ponded surface area (Zedler and Kercher 2005; Brinson and Eckles 2011; Gleason et al. 2011). Hence, anthropogenic disturbances that affect the water balance of these wetlands also have the potential to affect associated ecosystem services (Anteau 2012; Anteau et al. 2016).

Subsurface drainage systems (SDSs) are used throughout the United States to enhance agricultural production by facilitating water management. Historically, use of SDSs has been prevalent in the southeastern portions of the PPR (Iowa, Minnesota); however, use of these systems expanded rapidly into North Dakota and South Dakota (hereafter, Dakotas) during the 2000s in response to rising commodity prices, most notably corn, which peaked around 2012 (Roth and Capel 2012; Johnston 2013; Finocchiaro 2016; Werner et al. 2016). Subsurface drainage systems typically consist of perforated drainage pipe placed in various configurations below the soil surface to target specific areas (e.g., wetlands) or manage the water table of entire fields. Subsurface drainage systems situated beneath a wetland basin that often include a surface-inlet pipe can drain wetlands through direct removal of ponded surface water and subsurface soil pore water. Subsurface drainage systems placed in the surrounding contributing area of the catchment can indirectly affect wetland water balances through a reduction in water inputs from precipitation runoff (Tangen and Finocchiaro 2017) or by enhancing the downward or horizontal movement of water from a wetland through the lateral effect of the SDS (Werner et al. 2016). The effect of an SDS on a wetland's hydrology depends on factors such as topography, soils, and properties of the SDS such as size (pipe diameter, linear distance) and location (horizontal and vertical) of the drain pipe relative to the wetland (Werner et al. 2016; Tangen and Finocchiaro 2017). In general, national drainage policy requires the use of protective setback distances around wetlands when installing SDSs. The purpose of these buffers is to limit potential effects to the hydrology of wetlands (Tangen and Finocchiaro 2017).

Wetland easements are legal agreements by which private landowners are financially compensated by the U.S. Fish and Wildlife Service (USFWS) in exchange for not disrupting wetland hydrology, ecological processes, and functions. In the Dakotas, the USFWS oversees > 1,000,000 ha (ca. 2015) of wetland and grassland easements where private lands are managed to enhance wildlife habitat (USFWS 2014, 2015). Wetland easements account for greater than one-third of the conservation-easement lands in the Dakotas, and include hundreds of thousands of individual wetlands (USFWS 2014, 2015). Contract provisions of these easements prohibit wetlands from being drained, burned, filled, or leveled; however, when protected wetlands dry during their annual hydrologic cycle, landowners are allowed to farm, hay, or graze them. Likewise, grassland easements prohibit the conversion of grasslands to croplands but do not restrict grazing. The expanded use of SDSs in the Dakotas has drastically increased the number of drainage permit applications for privately owned agricultural lands, including those protected by USFWS wetland easements. Consequently, requests to install drainage systems on these easement lands also have increased. Subsurface drainage systems placed outside of the delineated wetland boundary, but within the local catchment, represent a potential form of indirect drainage; thus, for conservation easement lands SDSs typically are used in conjunction with prescribed drainage setback distances or in some cases are excluded from the catchment entirely.

One of the objectives of national drainage policy for agricultural lands is to limit effects to the hydrology of wetlands (e.g., NRCS 2009); however, the definition of an acceptable effect has not been well defined. In the Dakotas, protective drainage setback distances are calculated using scope and effect equations along with site-specific soil and SDS properties (NRCS 1997, 2004, 2009). These equations were originally intended for determining optimal spacing for drainage lines, but scientific investigations examining their efficacy for protecting wetlands are lacking. Thus, data pertaining to the effectiveness of drainage setback distances for protecting wetlands across a range of site characteristics (e.g., low- or high-relief terrain) are an important information gap for USFWS personnel tasked with developing and implementing drainage policies for conservation programs. Relevant questions raised by various stakeholders include: 1) Does an SDS located within a wetland's catchment have the potential to affect wetland hydrology through the disruption of precipitation runoff? 2) What are the primary factors to consider when determining potential effects of an SDS on wetland hydrology? and 3) Do the standard, prescribed drainage setback distances minimize potential effects of an SDS on wetland hydrology?

Although relevant research is sparse, a recent modeling study suggested that farmed prairie wetlands could have shortened periods of inundation and lower water depths depending on where SDSs are situated within a wetland's catchment (Werner et al. 2016). Moreover, a recent field study of prairie wetlands demonstrated that precipitation runoff from the catchment can be an important component of a wetland's seasonal water balance that could be affected by an SDS (Tangen and Finocchiaro 2017). Tangen and Finocchiaro (2017) also suggested that prescribed drainage setback distances could reduce the effects of an SDS on wetland hydrology under certain conditions. Overall, these studies suggest that SDSs located adjacent to wetlands can affect their hydrology by varying degrees, depending on factors such as depth of drainage pipe, distance of the SDS from wetland, slope gradient (e.g., high-, low-relief terrain), soil properties, and groundwater connection (recharge, discharge). In fact, period of inundation, water depth, or ponded surface area could be reduced by greater than 50% under certain scenarios (Werner et al. 2016; Tangen and Finocchiaro 2017). For this study, we assessed information used for prescribing lateral setback distances to identify important SDS and catchment characteristics to consider when establishing drainage-related policies for conservation lands. Moreover, we assessed the effectiveness of prescribed setback distances for excluding SDSs from wetland catchments and the potential effects of SDSs on wetlands located on lands protected by conservation easements.

Studies have shown statistical relations between wetland area and other physical features such as water depth, water volume, and catchment or contributing area (Hayashi and van der Kamp 2000; Gleason and Tangen 2008; Tangen and Finocchiaro 2017). For this study, we developed regression models to predict 1) contributing area of each catchment on the basis of wetland area, and 2) maximum length of slope on the basis of the predicted contributing area (Table 1). Application of these models for estimating wetland catchment areas and length of upland slopes is detailed in subsequent paragraphs.

We defined a wetland catchment as the wetland basin and its adjacent upland contributing area (Finocchiaro et al. 2014). Here, wetland area represents the maximum wetland extent based on the spill point and the maximum length of slope represents the greatest distance between the wetland and catchment boundaries (i.e., longest slope segment). We performed linear regression on natural log-transformed data (catchment areas, slopes) using PROC REG in SAS (version 9.4; SAS Institute Inc., Cary, NC). We developed independent models for the Glaciated Plains and Missouri Coteau physiographic regions of the PPR because of regional differences in topography (Gleason et al. 2008). We used a comprehensive topographic database, consisting of data from a range of PPR wetland studies, to develop these regional models (Tangen and Bansal 2018). This database included variables such as wetland area, contributing area, maximum slope length, and physiographic region (Table S1, Supplemental Material).

We obtained geographic information system (GIS) data layers during 2015 from the USFWS Habitat and Population Evaluation Team (Bismarck, ND) that included boundaries for USFWS wetland and grassland easements (easement parcels) in North Dakota. We combined easement data layers with soils (Soil Survey Geographic Database; NRCS 2018c), wetland (National Wetlands Inventory [NWI]; USFWS 2018), and land-use (2015 Cropland Data Layer; USDA 2015) layers to extract data for site characterization and analyses. Additional layers required to apply previously described regression models and calculate lateral setback distances (see below) included county and Level III Ecoregion (USEPA 2013). We used the NWI data to approximate the extent (wetland area, distribution) of palustrine emergent wetlands within each easement tract (Cowardin et al. 1979). We accomplished this by selecting wetland features completely within the easement boundary and simply dissolving (merging) adjacent features (i.e., wetland zones) into distinct wetlands. This process resulted in identification of nearly 470,000 and 45,200 wetlands located within wetland and grassland easement boundaries, respectively. These wetlands represent all wetlands within the easement boundaries, not just the wetlands that were specified for protection in the wetland easement contracts. Specific protected wetlands were not identified in the USFWS spatial data. All subsequent analyses were constrained to include only wetlands located within the PPR (Level III Ecoregions 42, 46, 48) that were between 0.1 and 10 ha, as wetlands outside of this range often are not targeted for drainage or beyond the data range of the regression models described previously (Table 1).

We used wetland area from the NWI in conjunction with the previously described regression models to estimate the contributing area of each catchment associated with wetland easements. This estimate simply represents total contributing area, and is not spatially explicit. Subsequently, we used contributing area to estimate maximum length of slope for each catchment (Table 1). Within a GIS environment, concentric wetland and contributing-area polygons were overlain with Soil Survey Geographic Database data to extract soil map units and associated attributes (e.g., slope gradient). We determined land use for each conservation easement tract using the 2015 Cropland Data Layer data. We determined total area of cropland, noncropland, and other lands for each easement tract by summing the areas of each land-cover type (e.g., corn, soybeans, grassland/pasture; Table S2, Supplemental Material). When the area classified as croplands exceeded the area classified as noncroplands the entire tract was categorized as cropland, and vice versa. For application of region-specific regression models, we considered catchments located in Level III Ecoregions 42 and 46/48 (USEPA 2013) part of the Missouri Coteau and Glaciated Plains physiographic regions, respectively (Gleason et al. 2008). We also used wetland areas on the basis of the NWI data to determine total wetland area associated with USFWS wetland and grassland easements. We considered these areas the nondrained baseline for conservation easements to be used to examine the potential landscape-level (i.e., USFWS easements) implications of SDSs (e.g., percent reduction in wetland surface area).

We determined the lateral effect, or drainage setback, distance for each wetland catchment within an easement area using data from the Natural Resources Conservation Service's lateral effects calculator for North Dakota (https://www.nrcs.usda.gov/wps/portal/nrcs/detail/nd/technical/engineering/?cid=stelprdb1269614, accessed January, 2018). The calculator provides a lateral effect distance on the basis of county, soil map unit from the Soil Survey Geographic Database, average depth below ground of the SDS (drainage pipe), and diameter of the drainage pipe. For determination of the lateral setback distance, we selected the dominant soil map unit for each contributing area, in terms of aerial coverage, as the representative soil. For soil map units with multiple components (most map units), we selected the largest lateral setback distance as the setback. There were 38 soil map units with no prescribed setback distance; thus, we excluded the wetlands associated with these soils (< 1% of total wetlands) from analyses.

We calculated a mean lateral setback distance, consisting of all county-soil map unit combinations, for the range of drainage pipe depths and diameters specified in the NRCS drainage calculator for North Dakota. Specified depths and diameters ranged from 0.6 to 2.7 m (2 to 9 ft.) and 7.6 to 30.5 cm (3 to 12 in.), respectively. We examined these mean lateral setback distances by drainage pipe depth and diameter with the purpose of identifying relations between the key SDS characteristics and prescribed lateral setback distances.

For each catchment within a wetland easement boundary, we calculated a difference by subtracting the estimated length of slope from the prescribed lateral setback distance. Consequently, positive differences indicate where SDSs would be excluded from the catchment and negative differences indicate where SDSs would be located within the catchment. We summarized results of these calculations and compared them to determine the extent to which setback distances effectively excluded SDSs from the wetland catchments.

For only those catchments where length of slope exceeded the lateral setback distance, we calculated an “at-risk length of slope” by subtracting the lateral setback distance from the length of slope. This distance represents the portion of the catchment with the potential to be affected by an SDS when relying on prescribed setback distances for protective buffers around wetlands. We also calculated the percentage of the slope (length) represented by this at-risk slope.

Constraining data to PPR wetland basins between 0.1 and 10 ha resulted in the inclusion of 253,396 wetlands associated with wetland easements and 26,092 wetlands associated with grassland easements (Table S3, Supplemental Material). Mean wetland areas from the NWI were 7,344 (SE = 24) and 7,638 (SE = 76) m² for the wetland and grassland easements, respectively. The total area of wetlands within wetland and grassland easements in the PPR of North Dakota was 2,060 km². Although the ponded water area of individual wetlands varies temporally, we considered this area the baseline for wetlands associated with conservation easement lands. Wetland and grassland easements accounted for 90 and 10% of the total wetland area in easement tracts, respectively. On the basis of the dominant land cover of the entire easement parcel, 48.7% of catchments associated with wetland easements were categorized as nonagricultural, whereas 99.8% of catchments associated with grassland easements were nonagricultural. On the basis of the representative slope gradient of the dominant Soil Survey Geographic Database soil mapping unit, 43% of catchments associated with wetland easements were located within low-relief terrain (< 3% slopes), whereas 18% were located within relatively high-relief terrain (> 9% slopes); slopes of the remaining catchments were intermediate (Table 2). For grassland easements, 15% of catchments were located within low-relief terrain, whereas 56% were located within high-relief terrain (Table 2).

On the basis of all county-soil mapping unit combinations for North Dakota, there was a distinct curvilinear relation between depth of drainage pipe and mean lateral setback distance, regardless of the drainage pipe diameter (Figure 1; Table S4, Supplemental Material). Drainage pipe diameter had little effect on lateral setback distance compared with drainage pipe depth. In general, SDSs placed closer to the surface (shallower) were characterized by shorter lateral setback distances compared with those placed deeper (Figure 1). In other words, the “affected area” around a drainage pipe, which is determined by the scope and effect equations, increases as the depth of the pipe increases (NRCS 1997).

For catchments associated with wetland easements, mean differences between the prescribed lateral setback distance and maximum length of slope (setback − slope) were highly variable and ranged from − 6 (SD = 51) to 48 (SD = 73) depending on drainage pipe depth and diameter (Figure 2; Table S5, Supplemental Material). These differences suggest that, on average, SDSs would be excluded from these catchments once the drainage pipe depth approaches 0.75 m (Figure 2). However, smaller wetlands tended to have smaller catchments and shorter slopes (Table 1), and greater than 90% of these wetlands were relatively small (< 2 ha). Thus, it is important to consider the size distribution of wetlands when interpreting overall means such as these. In fact, it is apparent that the minimum drainage pipe depth required to exclude SDSs from wetland catchments would vary greatly depending on wetland area, a rough surrogate for length of slope (Figure 3). In general, larger wetlands that are typically associated with longer slopes would require greater drainage pipe depths to result in setback distances that would exclude SDSs from the catchment (Figure 3).

For the set of wetland easements, the percentage of catchments where an SDS would be located within the catchment generally decreased as the depth of the drainage pipe increased, regardless of pipe diameter (Figure 4; Table S5). The percentage of catchments where the length of slope exceeded the lateral setback distance ranged from 44 to 59% for the deepest (2.7 m) and shallowest (0.6 m) drainage pipe depths, respectively (Figure 4). For the 44–59% of catchments where the lateral setback distance was less than the length of slope (i.e., an SDS would be within the catchment), the mean distance between the drainage setback distance and catchment boundary (based on the maximum slope length) ranged from approximately 22 to 40 m depending on drainage pipe depth (Figure 5). These estimates represent the distance within a catchment that is upslope of an SDS and equate, on average, to roughly one-quarter to one-half of the maximum slope distance (Figure 6).

Examination of relationships between lateral setback distances and SDS characteristics revealed that a major determinant of the lateral setback distance was the depth of the SDS. This analysis suggests that instead of simply relying on a prescribed drainage setback distance, conservation personnel could also consider the depth of the SDS that was used to calculate the setback distance, especially when the goal is to limit effects to wetland hydrology. Essentially, on the basis of prescribed lateral setback distances, a shallower SDS would be located closer to a wetland compared with a system situated at greater depths. Thus, shallower systems would have greater potential to affect wetland hydrology when solely relying on the setback distance to limit such effects. Werner et al. (2016) concluded that changes to wetland period of inundation in a low-relief terrain were more sensitive to the depth of an SDS than to the distance (i.e., setback) from a wetland, substantiating the importance of considering SDS characteristics such as depth when assessing potential effects to wetland hydrology.

Subsurface drainage systems located within wetland catchments have the potential to affect wetland hydrology to varying degrees (Werner et al. 2016; Tangen and Finocchiaro 2017). Thus, the exclusion of SDSs from wetland catchments would be optimal for the conservation of wetland hydrology. Our results suggest that setback distances can be effective at excluding SDSs from wetland catchments depending on catchment size, or more specifically length of slope. Setback distances generally would protect wetland hydrology of smaller wetlands (< 2 ha) by excluding SDSs from the wetland catchments. Conversely, setback distances for the larger wetlands (> 8 ha) typically would allow for placement of SDSs within wetland catchments. Setback distances for wetlands of intermediate size would afford various levels of protection depending on previously mentioned factors. Roughly 50% of the catchments within wetland easement areas that we considered for this study were characterized by setback distances that would allow placement of an SDS within the catchment. The at-risk upslope portion of these catchments represented approximately one-quarter to one-half of the length of the dominant slope. Thus, there would be potential for SDSs to affect the hydrology of these wetlands when relying on prescribed setback distances for protective buffers.

Wetland easements in North Dakota protect a large number of wetlands from direct drainage (USFWS 2014, 2015). The 46% of catchments within wetland easement areas identified in this study to be associated with agricultural lands, however, represent wetlands that may be vulnerable to indirect drainage from SDSs located in the adjacent contributing area. A high proportion of these wetlands also were associated with relatively low-relief terrain, which often is optimal for agriculture and targeted for drainage because of characteristic areas of saturated soils or ponded water. In addition to these low-relief areas, soil data suggested that wetland easements encompassed a wide variety of terrains, including high-relief areas. Consequently, wetland hydrology could be affected in a variety of ways including lateral drainage, interception of groundwater, and reduced precipitation runoff (Werner et al. 2016; Tangen and Finocchiaro 2017). The magnitude of the potential effect on these catchments would depend on whether drainage setback distances were required, as well as factors relating to the catchment (e.g., soils, slope length and gradient) and SDS (drainage pipe depth). Any potential effects also would be contingent on the intent of a given landowner to install an SDS.

Stipulations within wetland easement contracts explicitly protect wetlands from direct drainage. Although not specified in the easement contract, grassland easements indirectly protect wetlands from drainage and other effects by restricting agricultural activities in the catchment. Wetlands and grasslands provide numerous ecosystem services to society, as well as critical habitats to a wide variety of wetland-dependent wildlife (Batt et al. 1989; Zedler and Kercher 2005; Niemuth et al. 2006; Gleason et al. 2011; Mushet et al. 2014). Consequently, wetland and grassland easements combined provide significant ecological benefits in terms of providing and protecting wetlands. Information provided by this and other studies suggest that drainage policy could have a meaningful effect on the ecological benefits provided by easement lands (Werner et al. 2016; Tangen and Finocchiaro 2017). As an example, drainage resulting in 10 and 50% losses in total ponded wetland area (2,060 km²) could reduce wetland area by approximately 200 and 1,000 km², respectively. Although these are extreme scenarios that assume pervasive use of SDSs, they do demonstrate the value of conservation easements for protecting important wetland systems, and illustrate potential effects of alternative drainage policies that could range from no drainage within a catchment to drainage with or without setback distances.

The expansion of subsurface drainage in the PPR has generated concerns over the direct and indirect effects of this practice on regional hydrology and wetland systems (Schilling and Libra 2003; Schilling and Helmers 2008; Blann et al. 2009; Schilling et al. 2012; Werner et al. 2016). Although relevant research is sparse, studies have suggested that an SDS could affect wetland hydrology when located within a wetland's catchment, even when drainage setback distances are used (Werner et al. 2016; Tangen and Finocchiaro 2017). U.S. Fish and Wildlife Service wetland easement contracts stipulate that wetlands cannot be drained; thus, exclusion of SDSs from protected wetland catchments might be ideal owing to uncertainties associated with drainage setback distances and potential indirect effects of SDSs. Our results suggest that drainage setback distances would exclude SDSs from catchments of smaller PPR wetlands with shorter slopes in the adjacent upland contributing area. For larger wetlands, however, considerable areas of the catchment would be vulnerable to drainage that may affect wetland hydrology (Werner et al. 2016; Tangen and Finocchiaro 2017). Moreover, we identified the depth of an SDS as an important factor to consider when prescribing drainage setback distances to protect wetlands. Thus far, limited research has identified important factors related to the effects of SDSs on wetland hydrology and only provided rough estimates of these effects (Werner et al. 2016; Tangen and Finocchiaro 2017). Future research, including the collection of relevant field data, is required to assess potential effects of SDSs on wetland hydrology across a range of site and meteorological conditions.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Data and metadata associated with this paper are available online through the U.S. Geological Survey (USGS) ScienceBase Catalog (see Tangen 2018).

Table S1 (TOPOGRAPHY.csv). Topographic data (catchment areas and slopes) used for developing models presented in Table 1. These data were obtained from a comprehensive database, consisting of data from a range of Prairie Pothole Region wetland studies conducted between 1997 and 2013 (Tangen and Bansal 2018).

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S1 (16 KB XLSX); also available at http://dx.doi.org/10.5066/F72806H6 (16 KB CSV).

Table S2. Land-cover types from the 2015 Cropland Data Layer used to determine dominant land use associated with U.S. Fish and Wildlife Service conservation easements: cropland, noncropland, or other. We obtained the data layer from the U.S. Department of Agriculture, National Agricultural Statistics Service data repository web site (USDA 2015).

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S2 (14 KB DOCX).

Table S3 (SITES.csv). Characteristics (area, slope, land use) of wetland catchments associated with wetland and grassland easements in the Glaciated Plains and Missouri Coteau physiographic regions of the Prairie Pothole Region of North Dakota. We obtained easement data (ca. 2015) from the U.S. Fish and Wildlife Service, Habitat and Population Evaluation Team (Bismarck, ND). We based wetland areas on National Wetlands Inventory data (USFWS 2018), catchment slopes on the Soil Survey Geographic Database (NRCS 2018c), dominant land uses on the 2015 Cropland Data Layer (USDA 2015), and physiographic regions on Level III Ecoregions (Gleason et al. 2008; USEPA 2013).

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S3 (10.4 MB XLS); also available at http://dx.doi.org/10.5066/F72806H6 (10.2 MB CSV).

Table S4 (LSD.csv). Mean lateral setback distance, consisting of all county-soil map unit combinations, calculated for the range of drainage pipe depths and diameters specified in the Natural Resources Conservation Service's drainage calculator for North Dakota (ca. 2017; https://www.nrcs.usda.gov/wps/portal/nrcs/detail/nd/technical/engineering/?cid=stelprdb1269614, accessed January, 2018).

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S4 (1 KB XLS); also available at http://dx.doi.org/10.5066/F72806H6 (1 KB CSV).

Table S5 (SETBACK.csv). Relations (difference, at-risk length of slope) between lateral setback distances and calculated length of slope for wetland catchments associated with wetland easements in the Glaciated Plains and Missouri Coteau physiographic regions of the Prairie Pothole Region of North Dakota. We obtained easement data (ca. 2015) from the U.S. Fish and Wildlife Service, Habitat and Population Evaluation Team (Bismarck, ND). We based wetland areas on National Wetlands Inventory data (USFWS 2018) and physiographic regions on Level III Ecoregions (Gleason et al. 2008, USEPA 2013). We determined lateral setback distances (including pipe depth and diameter) using data from the Natural Resources Conservation Service's drainage calculator for North Dakota (ca. 2017; https://www.nrcs.usda.gov/wps/portal/nrcs/detail/nd/technical/engineering/?cid=stelprdb1269614, accessed January, 2018) along with the dominant soil mapping unit of the catchment from the Soil Survey Geographic Database (NRCS 2018c).

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S5 (866 MB XLXS); also available at http://dx.doi.org/10.5066/F72806H6 (846 MB CSV).

Reference S1. Cowardin LM, Carter V, Golet FC, LaRoe ET. 1979. Classification of wetlands and deepwater habitats of the United States. Washington, D.C.: U.S. Department of the Interior. Fish and Wildlife Service, Office of Biological Services, FWS/OBS-79/31.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S6 (17.4 MB PDF).

Reference S2. Gleason RA, Laubhan MK, Euliss NH Jr, editors. 2008. Ecosystem services derived from wetland conservation practices in the U.S. Prairie Pothole Region with an emphasis on the U.S. Department of Agriculture Conservation Reserve and Wetlands Reserve Programs. Reston, Virginia: U.S. Geological Survey. Professional Paper 1745.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S7; also available at http://pubs.usgs.gov/pp/1745/ (4 MB PDF).

Reference S3. Natural Resources Conservation Service. 1997. Hydrology tools for wetland determination. Washington, D.C.: U.S. Department of Agriculture. Natural Resources Conservation Service, Engineering Field Handbook, National Engineering Handbook, Part 650, Chapter 19.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S8; also available at http://directives.sc.egov.usda.gov/OpenNonWebContent.aspx?content=17556.wba (206 KB PDF).

Reference S4. Natural Resources Conservation Service. 2004. Hydrology tools for wetland determination: North Dakota Supplement. Washington, D.C.: U.S. Department of Agriculture. Natural Resources Conservation Service, Engineering Field Handbook, National Engineering Handbook, Part 650, Chapter 19.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S9; also available at http://www.nrcs.usda.gov/wps/portal/nrcs/detail/nd/technical/engineering/?cid=stelprdb1269614 (85 KB PDF).

Reference S5. Natural Resources Conservation Service. 2009. Tile guidelines for protecting wetland hydrology. Washington, D.C.: U.S. Department of Agriculture. Natural Resources Conservation Service, South Dakota Hydrology Technical Note #4.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S10; also available at http://www.nrcs.usda.gov/wps/portal/nrcs/detail/sd/technical/engineering/?cid=nrcs141p2_036576 (193 KB PDF).

Reference S6. U.S. Fish and Wildlife Service. 2014. U.S. Fish and Wildlife Service in North Dakota.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S11; also available at https://www.fws.gov/uploadedFiles/ND%20USFWS%20Office%20Profiles%20FINAL%2012_2014.pdf (1 MB PDF).

Reference S7. U.S. Fish and Wildlife Service. 2015. U.S. Fish and Wildlife Service Annual Report of Lands as of September 30, 2015.

Found at DOI: http://dx.doi.org/10.3996/092017-JFWM-076.S12; also available at https://www.fws.gov/refuges/land/LandReport.html (22 MB PDF).

Funding for this study was provided by the USFWS, Plains and Prairie Potholes Landscape Conservation Cooperative, and the USGS. Special thanks are given to Alec Boyd for providing GIS support. We also thank Sheel Bansal, David Haukos (Associate Editor), and the anonymous reviewers for their insight and helpful comments on this manuscript.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.