Development of wetlands in aging reservoirs: Opportunities to enhance wetland capacity and improve water quality

Kansas Biological Survey Report No. 203

June 2018

By

Vahid Rahmani Department of Biological and Agricultural Engineering Kansas State University

Donald G. Huggins, Jude Kastens, Debbie Baker, and Kaitlyn Loeffler Central Plains Center for BioAssessment Kansas Biological Survey University of Kansas

For

Kansas Water Office CD-97751901 KUCR project KAN74759

1 Contents Introduction ...... 3 Project Outputs and Outcomes...... 5 Methods and Materials ...... 6 Reservoir Detention Time ...... 8 Precipitation and Inflow ...... 10 Flood Detention Zones ...... 10 Water sample collection ...... 14 Sediment collection ...... 15 Vegetation ...... 16 Birds and nuisance species ...... 17 Results ...... 17 Reservoir water elevation fluctuations ...... 17 NWI type and occurrence characteristics in the flood pool ...... 20 Discussion ...... 23 Box plots ...... 25 TN:TP ratios...... 39 Vegetation ...... 40 Shear stress...... 45 Conclusions and Future Studies ...... 46 References ...... 48 Appendix A...... 53 Appendix B...... 61 Appendix C...... 62

2 Introduction

Wetland ecosystems embrace a great variety of waterbodies that includes such things as swamps, marshes, bogs, fens, wet meadows and prairies, prairie potholes, playas, vernal pools and mudflats (Keddy 2010; Batzer and Baldwin 2012; USEPA 2015). The structure and function of wetlands is so varied and dependent upon soils, climate, broad-scale and local hydrology and other physical and biological factors that it remains difficult to adequately characterize and classify every type of wetland one may encounter in today’s environment. In addition to naturally occurring wetlands, modern landscapes often contain artificial or created wetlands that were constructed to provide designated goods or services such as storm-water treatment and mitigation of lost wetland acres (e.g. Moshiri 1993; Mitsch 2005; Vymazal 2010).

Wetlands that are created or later developed from impounding lotic ecosystems are different from naturally occurring wetlands but nevertheless are wetlands as they possess many of the basic properties and processes associated with wetlands (Smith et al. 2008). Impounded wetlands often have reduced temporal productivity because of more stable water levels and longer hydroperiods, shorter life spans due to high sediment infill rates, increased nutrient and contaminant loadings, and extended hydroperiods that can cause major shifts in biota (Smith et al. 2008). Despite their shortcomings, wetlands associated with reservoirs and impoundments represent a substantial portion of today’s wetland habitats in many regions of the country and provide unrecognized and underappreciated ecological goods and services to society. This project will, in part, identify, quantify and characterize some of the reservoir wetland systems found in Kansas. This synoptic survey focuses on quantifying wetland extent and location within reservoirs while assessing selected wetland elements that will help characterize the hydrology, biology, and physical and water quality properties of these systems to help determine future management opportunities. Wetlands connected to the reservoir waterbody at conservation pool levels are the focus of this project.

Legal protection is offered to natural wetlands by the US Army Corps of Engineers (USACE) under Section 404 (CWA) if they exhibit all of the following primary characteristics: hydrology, hydrophytes and hydric soils (see review by Strand and Rothschild 2010). It is important to understand that some areas that function ecologically as wetlands may possess only one or two of these three characteristics of jurisdictionally-defined wetlands. Such wetlands, however, perform valuable functions. US Fish and Wildlife Service (USFWS) identifies wetlands as ecosystems that are transitional between terrestrial and aquatic systems and are characterized by one or more of the following: 1) system periodically supports mostly hydrophytes; 2) bottom substrate is un-drained hydric soils; and 3) the substrate is non-soil and saturated or inundated by shallow water during the biological growing season (Cowardin et al. 1979).

While structural definitions (i.e. structural classification schemes) of wetlands remain somewhat elusive, wetland functions are less contentious and better accepted qualities of wetlands. USACE classified wetland functions into three broad categories (hydrological, biogeochemical, and habitat and food web support) with many specific functions and societal values listed (Smith et al. 1995). More recently, Tiner (2003) provided the rationale for the functional criteria used to identify wetlands of potential landscape significance in the Northeast. These functions include: 1) surface water detention, 2) coastal storm surge detention, 3) streamflow maintenance, 4)

3 nutrient transformation, 5) sediment and other particulate retention, 6) shoreline stabilization, 7) provision of fish and shellfish habitat, 8) provision of waterfowl and water bird habitat, 9) provision of other wildlife habitat, and 10) conservation of biodiversity.

Natural ecosystems and especially wetlands are in danger worldwide. Many anthropogenic and natural factors have contributed to wetland loss and degradation. These factors include land use change, irrigation, infrastructure development, hydrological modifications, water pollution, climate change, and sea level rise (Millennium Ecosystem Assessment 2005; Boelee et al. 2011).

Kansas has lost about half of its natural wetlands through cropland conversion and surface and groundwater depletions (Fretwell et al. 1996). Many of these wetlands are irreplaceable due to the permanency of prior conversions and reductions in available water in some areas of the state. This is similar to what has happened in California’s Central Valley, where more than 95 percent of the original wetland area has been lost (Gilmer et al. 1982) and increasing demands for water and land resources make it unrealistic to recover significant portions of the original wetland area. Wetland construction, restoration, and conservation are important considerations in both state- wide and nation-wide water resources management programs because of their role in preserving and enhancing water quantity and quality. Some of the many services wetlands can provide are: a) moderation of extreme precipitation and discharge events; b) sediment storage and load reduction into reservoirs and other water resources; and c) sequestration of pollutants and reduced downstream water resource contamination (United Nations WWAP 2015). These facts have prompted Kansans to focus on preserving wetland quality and identification of waterbodies and aquatic resources that provide possible wetland functions if not true wetlands by the USFWS definition. Kansas investigators assessed the potential of farm ponds and small impoundments in providing many wetland functions either through natural aging processes such as infilling and hydrophyte community development or through possible enhancement activities (EPA award CD 97746601-0, Huggins et al. 2017).

In 2016 and 2017 we performed similar assessment of the potential wetland functions of the riverine segments of 20 of the 24 federal reservoirs located throughout Kansas. These reservoirs are now decades old having mostly been constructed in the 1960s and 1970s. It appears that within the riverine segments of these reservoirs wetlands have evolved or are evolving as part of the reservoir aging process driven in part by sediment infilling and nutrient enrichment and resuspension (Jakubauskas et al. 2014; Dzialowski et al. 2008). For example, examination of the historical change of the Tuttle Creek Lake surface area shows that approximately 30% of the lake surface area has been lost since establishment of the reservoir in 1962. As the reservoir is filling with incoming sediments, the surface area converts to shallower areas where potential wetlands can exist. The shallow zone (depth < 2 m) covers approximately 8.5 km2 (19%) of the current Tuttle Creek Lake surface area. Water quality of these reservoirs varies spatially within the lake, changing temporally and probably with inflow variations, suggesting that the water components in the riverine segments will ultimately move down to deeper areas and impact the whole lake. This temporal and spatial variability can also be seen in chlorophyll-a gradients within reservoirs.

The understanding that riverine segments of reservoirs may function as wetland areas is not new as both Carney (2002) and Fram et al. (2001) included these areas as wetlands in water quality

4 studies in Kansas and California, respectively. Additionally, Nebraska identified two types of wetlands as occurring in their reservoirs – fringe wetlands and headwater wetlands (Pegg et al. 2015), while Mahlke (1996) identified five types of wetland communities in Oklahoma reservoirs. In addition to the riverine areas of reservoirs as developing wetland regions, reservoir deltas that have formed at the entry points of reservoir tributaries also seem likely areas for the creation of wetlands or systems with wetland functions. Potential biodiversity increases associated with aging reservoir deltas were addressed by Johnson (2002). Johnson gives several regional examples of aging reservoirs that have developed deltas (deltaic) and riverine and shoreline (fringe) systems that have become wetlands in both form and function as outlined by USFWS (Cowardin et al. 1979) and others (Levine and Willard 1990; Keddy and Fraser 2000; Kusler 1990; Lewis 1995; Kentula 2000).

Project Outputs and Outcomes

Project goals are presented in Table 1.

Table 1. Goals and actions performed to assess wetland potential in Kansas federal reservoirs. Goal Action Description Review NWI for listed wetlands either created by or hydrologically associated with 1 1 reservoirs in Kansas. Identify and compile other sources of wetlands associated with reservoir systems in 2 Kansas. Identify and map potential wetland area groups (2 depth groups) that have water at 2 1 historical median lake level. Identify and map of potential wetland pond areas and complexes that are predominately 2 dry or hydrologically outside the reservoir historical median water level. Using existing satellite imagery, aerial photography and group 1 and 2 coverages select 3 1 either group 1 or 2 coverages as best estimators of potential wetland areas for field studies and synoptic sampling. Perform field reconnaissance of small subset of reservoirs to validate group selection from 2 Action 1 as best GIS-based predictor of potential wetland areas. 4 1 Analyze reservoir inflow and outflow data to estimate lake retention time. Analyze reservoir storage level, precipitation, and evaporation information to estimate 2 depth in wet and dry season. Use LiDAR topography and bathymetric maps to estimate area under water for wet and dry 3 periods. 4 Examine frequency distribution of flooded/dewatered periods for potential wetland areas. Collect composite samples for both surface (≈ 0.25m) water and shallow sediment at each 5 1 riverine and main basin study areas. Water and sediment samples will be prepared and shipped to KSU service laboratory of 2 analysis of total nitrogen, total phosphorus and TOC for all samples. In situ water quality parameters will be will be obtained at one site (center) along a depth 3 profile for each riverine and basin study area. 4 Surface water composite samples will be collected for chlorophyll-a analysis as in Action 1. 5 Surface area estimates for macrophytes

5

6 List of all shorebirds and water birds observed and submitted to eBird website 7 Database construction and data population of the project database 6 1 QAPP 2 Write reports. 7 1 Meet annually with EPA project officers 2 Seek suggestions for project improvement from other experts

Methods and Materials

To be used in this study, a reservoir had to have both LiDAR elevation data to represent the topography around the reservoir, and bathymetry data to represent the reservoir bottom topography. It also had to have historic hydrologic data. Three reservoirs out of the 24 federal reservoirs in Kansas (Cedar Bluff, Keith Sebelius, and Waconda) did not have readily attainable, recently acquired bathymetry data, so these were excluded from study sites. We could not access the required hydrologic data for Lovewell and there were problems with the hydro-flattened reservoir extent depicted in the available LiDAR terrain model, so this reservoir was also excluded from the study. The remaining 20 reservoirs were sampled in the field and analyzed in this project (Figure 1). Methods for determining reservoir detention time, flood detention zones, water elevation history, slope, and potential wetland areas follow.

6

Figure 1. Location of study reservoirs, with Kansas average annual precipitation isotherms. Data source: PRISM Climate Group.

Table 2. Summary of characteristics for the 20 reservoirs examined in this study. Parameter Min Mean Median Max Surface area* (ac) 1151 6170 5474 15353 Volume* (ac-ft) 16528 114696 79037 373152 Age (years, as of 2017) 36 51 52 69 Maximum depth* (ft) 12.0 45.1 43.4 79.6 Shoreline length* (mi) 18.1 49.6 48.0 104.7 25th percentile elevation (ft) 793.8 1169.3 1057.6 1871.2 50th percentile elevation 795.0 1171.8 1059.0 1888.0 75th percentile elevation 796.9 1173.4 1061.1 1892.9 100th percentile elevation 829.7 1195.4 1103.3 1896.8 Top of flood pool elevation 825 1198.7 1102.0 1923.7 *At normal operating water levels.

7

Reservoir Detention Time

Retention time, or water residence time, is the average amount of time that a particle of water stays in a basin (Dingman 2015). This is an important parameter to study because the more time that water stays in a particular basin; the more pollutants get filtered out of the water (Ghanad 2014). For this study, the basins are the 20 federally regulated reservoirs in Kansas. To calculate the retention time for each of the reservoirs, the basin storage and total outflows were needed. Daily hydrology data were used from the United States Army Corps of Engineers Kansas City and Tulsa districts, the U.S. Geological Survey, and the U.S. Bureau of Reclamation (USACE 2016a, USACE 2016b, USBR 2016, USGS 2016). For nine of the reservoirs, data were utilized from the day that the reservoir first reached its regulation level through 2015. For ten of the reservoirs, data were utilized from the beginning of digital record keeping in 1995 through 2014 or 2015. For John Redmond, data were utilized from only 2013 and 2014 due to a recent change in regulation pool level (USACE 2013). Daily averages were calculated for the reservoir storage, precipitation, inflows, discharges, and evaporation. Although precipitation and inflows are not necessary to calculate retention time, they provide additional insight as to some underlying factors that indirectly impact the retention time of a basin, as will be discussed later. The average retention time for each of the reservoirs was calculated with two equations that were modified from Holden (2001) method:

푡표푡푎푙 푟푒푠푒푟푣표𝑖푟 푠푡표푟푎푔푒 (푓푡3) 푅푒푡푒푛푡푖표푛 푡푖푚푒 (푦푟) = (1) 푡표푡푎푙 푑𝑖푠푐ℎ푎푟푔푒 (푐푓푦)

푡표푡푎푙 푟푒푠푒푟푣표𝑖푟 푠푡표푟푎푔푒 (푓푡3) 푅푒푡푒푛푡푖표푛 푡푖푚푒 (푦푟) = (2) 푡표푡푎푙 푑𝑖푠푐ℎ푎푟푔푒 (ft3/yr)+ 푡표푡푎푙 푒푣푎푝표푟푎푡𝑖표푛 (ft3/yr)

Equation 1 and 2 differ in their calculated outflows. The retention time in equation 2 was calculated by using both the total discharge and total evaporation. The retention time in equation 1 was calculated by using just the total discharge, because most retention time calculations cited in literature do not take into account the average evaporation, possibly due to unavailable data. In this project, the retention times were calculated with both equations in order to evaluate the impact of evaporation on the average retention time for each reservoir.

Our calculations for retention time with (without) evaporation for the 20 study reservoirs varied from 21 (22) days for Toronto to just over 879 (1,419) days for Wilson Reservoir (Table 3). Retention time estimates (with vs. without evaporation) naturally varied between reservoirs due to both precipataiton and evaporation combinations unique to each spatial location within the state.

8

Table 3. Retention time and inflow for the study reservoirs. Mean retention Mean Mean retention Difference time w/o yearly Reservoir time w/ evaporation between retention evaporation inflow (yearly) times (w/o – w) (yearly) (cfs) Big Hill 1.032 1.390 0.358 33 Cheney -* 2.21 - 205 Clinton 0.697 0.791 0.094 244 Council 0.501 0.592 0.091 128 Grove El Dorado 1.164 1.772 0.608 174 Elk City 0.103 0.109 0.006 527 Fall River 0.077 0.079 0.002 429 Hillsdale 0.898 1.087 0.189 118 John - 0.049 - 2,080 Redmond Kanopolis - 0.393 - 240 Kirwin 1.637 2.460 0.823 55 Marion 0.956 1.845 0.889 95 Melvern 0.931 1.102 0.171 219 Milford 0.614 0.678 0.064 880 Perry - 0.518 - 718 Pomona 0.441 0.487 0.046 201 Toronto 0.058 0.060 0.002 530 Tuttle 0.182 0.188 0.006 2,294 Creek Webster 1.730 2.797 1.067 36 Wilson 2.407 3.886 1.479 135 * Evaporation data not available

Table 3 and Figure 1 illustrate that generally the lakes in the eastern half of Kansas have smaller retention times as compared to western reservoirs, with longer retention times occurring further west in the state. In a study by the Kansas Biological Survey conducted during 1999-2001, scientists calculated the retention time for 593 lakes and reservoirs in Iowa, Kansas, Missouri, and Nebraska that were all greater than ten acres. According to their data, the retention times calculated for the study reservoir in this study ranged from 15 to 708 days (Dzialowski 2008). Dzialowski et al.’s retention times are similar to retention times that were calculated in the Kansas Department of Health and Environment’s Total Maximum Daily Load reports for these Kansas reservoirs, which range from 22 to 876 days (KDHE).

Generally reservoirs in eastern Kansas indicated lower difference between the retention times calculated with and without evaporation, implying that average evaporation rates do not greatly impact the average retention times in this part (Table 3).

9

Precipitation and Inflow

Located in the middle of the United States, Kansas is highly variable spatially and temporally. For example, eastern Kansas has a humid climate while the southwestern part of the state has an arid to semi-arid climate. Kansas also exhibits a strong gradient of precipitation across the state, from approximately 950 mm/year in the east to approximately 500 mm/year in the west (Rahmani 2015). In order to potentially explain some of the underlying causes of the variation of retention times across the state, average precipitation and inflow data were calculated for each of the reservoirs. The average yearly precipitation data from 1998-2013 shows the same trend of more precipitation in the east and less precipitation in the west (Figure 1) (Goodin 1995, Rahmani 2015). A reason for the variation in precipitation across the state is location. The Rocky Mountains, located west of Kansas, block out moist air from the west coast, creating an area of low precipitation over western Kansas (rain shadow effect). However, the precipitation in eastern Kansas comes primarily from the Gulf of Mexico, where the air is moist. This creates frequent thunderstorms and precipitation that are common in the spring and summer in eastern Kansas (Goodin 1995).

The retention time results are expected because of the differences in precipitation and inflows that occur across the state of Kansas (Table 3 and Figure 1). For example, since the eastern half of Kansas has a relatively large amount of rainfall or surface flow, large volumes of water can be released more frequently without the concern of running the reservoir dry. This is not the case for the reservoirs in western Kansas. Most of the reservoirs in western Kansas hardly ever reach their regulation level, meaning that reservoir discharges need to be closely conserved. The variation in precipitation and inflows across Kansas suggests that in order to calculate an accurate retention time in a drier climate, such as western Kansas, it is important to take into account all of the possible discharges, including water lost due to evaporation.

Flood Detention Zones To determine wet (areas wet 50-75% of the time) and dry (areas wet 25-50% of the time) zones around each reservoir, we used historic water level elevation data and reservoir basin topography (a blend of LiDAR digital elevation data and bathymetry information, Figure 2 shows an example for Clinton Lake). The zone of 25th percentile water level to 75th percentile water level was termed the fluctuation zone. For 17 reservoirs, the 50th percentile was within on foot of regulation levels, while the median levels for three reservoirs located in the central part of the state (Kanopolis, Kirwin, and Webster) were considerably lower than their regulation level (by 1.5–14 ft). Water levels for all of these federal reservoirs are highly managed; however, the reservoirs in the central part of Kansas are less likely to reach their regulation level due to the lack of rainfall and insufficient inflow (Loeffler et al. 2018).

10

Figure 2. A) 2-m LiDAR bare earth elevation raster for Clinton Lake, B) bathymetry (rasterized and up-sampled to 2 m) for Clinton Lake, and C) merged LiDAR and bathymetry to create the empty basin topography for the flood pool.

Within the upper fluctuation zone (UFZ; the region between the 50th and 75th percentiles) and the upper flood pool (UFP; the region between the 75th percentile and the designed maximum flood pool level) we then examined median ground slope values summarized across wetland delineations from the NWI dataset. The historical reservoir water level data provided the range of typical water level elevations during wet and dry periods. Determining these elevation levels was performed as follows.

The 25th percentile lake level corresponds with the maximum area that is inundated at least 75% of the time (typical low-water condition), while the 75th percentile lake level corresponds with maximum area that is inundated at least 25% of the time (typical high-water condition). The 50th percentile water level elevation was compared to the regulation level for each reservoir to determine if the median observed lake water elevation was near the regulation level. Reservoir water level were used for the different periods for each reservoir depending on data availability and accuracy. More details are provided in section reservoir detention time. The reservoir regulation level also called the conservation pool level, is defined as the maximum water level elevation during normal operating conditions, without considering flood control storage (NOAA 2017; TWDB 2017). The 25th, 50th, 75th, and 100th percentile values were calculated by ranking the daily water levels in ascending order and analyzing them (Table 4).

Reservoir water elevations were analyzed to estimate the depth differences between the 50th and 25th percentiles and the 75th and 50th percentiles for each of the studied reservoirs (Table 4). The calculated depths not only describe the depth change during high and low water conditions, but can also describe the characteristics of the reservoir.

11

Table 4. Reservoir water elevation levels for the 25th, 50th, and 75th percentile lake extents and the depth changes between each of the lake extents. Reservoir 25th 50th 75th Difference Difference percentile percentile percentile between 50th between 75th water water water & 25th (ft) & 50th (ft) elevation (ft) elevation (ft) elevation (ft) Big Hill 857 858 858 0.9 0.4 Cheney 1420 1422 1422 1.4 0.3 Council Grove 1272 1274 1274 2.0 0.5 El Dorado 1337 1339 1339 1.5 0.5 Elk City 794 795 797 1.2 1.8 Fall River 948 949 951 0.5 1.6 John Redmond 1042 1043 1044 1.0 1.2 Marion 1349 1350 1350 0.7 0.6 Toronto 901 902 903 0.6 1.5 Clinton 875 876 877 1.1 1.6 Hillsdale 915 916 918 1.3 1.2 Kanopolis 1463 1464 1466 1.2 1.7 Kirwin 1715 1726 1730 11.0 3.3 Melvern 1034 1035 1037 1.4 1.3 Milford 1143 1144 1146 1.5 1.3 Perry 891 892 894 1.3 1.8 Pomona 973 974 976 1.3 1.7 Tuttle Creek 1074 1076 1078 1.9 2.9 Webster 1871 1888 1893 16.8 4.9 Wilson 1514 1515 1516 1.6 1.0

Field Methods

Site layout

Field work took place in Sept. – Nov. 2017 and June – Aug. 2017 with visits to 20 reservoirs and a total of 60 sites. In each lake, the main basin and one to three riverine segments were sampled at five locations each for in situ water quality, plant pigments and sediment (Figure 3). In each riverine segment, three points were sampled along a longitudinal transect following the mid-line of the study reach from the uppermost boat-accessible location (point 1), no closer than 1 m to the shore line and in at least 0.25 m of water, to the segment’s downstream end located at the 1 m depth level relative to the conservation pool (point 5). The center sampling point (3) on this transect was located approximately midway between points 1 and 5. The 1 m depth level was marked on a map prior to fieldwork. Two samples (points 2 and 4) were collected near the ends of the perpendicular sample transect line in at least 0.25 m of water and at least 1 m from each

12 shoreline when possible. Observational occurrences of wetland and aquatic plants along each transect were recorded to the lowest feasible taxonomic group (i.e. species, genus, family). In addition, 10 sampling points along each transect were sampled for submerged aquatic plants using a weighted garden rake. GPS coordinates and water depth were recorded at each sampling point. Sampling was set up similarly in the main basin, but with reversed longitudinal transect points with point 5 upstream and point 1 near the dam face. Transects were no longer than those in the riverine zones. The downstream point 1 was at least 2 m from the dam face.

Figure 3. Transect placement within a reservoir, with detail of Riverine 1 segment enlarged. Sample point 1 is near the shore or dam at >0.25m depth and riverine point 5 is at 1 m depth (blue line) at riverine sites. Sample points 2 and 4 were collected near the ends of the perpendicular sample transect line in at least 0.25 m of water and at least 1 m from each shoreline when possible. Vegetation is sampled at 10 locations (+) per transect.

In situ measurements were taken at each sampling point using a Horiba Model U-52 Water Quality Checker. Sufficient time was taken to allow sediment to settle and the instrument to acclimate prior to recording measurements of water temperature, pH, DO, turbidity, conductivity, total dissolved solids, salinity, and ORP (Table 1). Samples were taken at the

13 water’s surface (approx. 0.25m). The Horiba U-52 was calibrated prior to each sampling trip, or at minimum once a week during sampling periods, using the methods in the U-52 manual.

Table 5. Parameter precision (repeatability) and accuracy for the Horiba Model U-52 Water Quality Checker. PARAMETERS RANGE RESOLUTION REPEATABILITY ACCURACY Water -5 to 55 ± 0.10 (at calibration 0.01 ± 0.03 + 0.005 Temperature C° point) pH 0 to 14 0.01 ± 0.05 ± 0.1 Oxidation -2000 to + Reduction 1 mV ± 5 mV ± 15 mV 2000 mV Potential Dissolved 0 to 50.0 0 to 20 mg/L: ± 0.2 0.01 ± 0.1 oxygen mg/L mg/L 0 to 100 0 to 0.999 mS/cm: Conductivity ± 0.05% F.S. ± 1% F.S. mS/cm 0.001 Total Dissolved 0 to 100 0.1% F.S. ± 0 2 g/L ± 2 g/L Solids g/L 5% of reading or 0 to 800 5% of reading or 0.5 Turbidity 0.01 0.5 NTU whichever NTU NTU whichever is > is >

Water sample collection

See the project QAPP for detailed water collection field methods and equipment. 1. In each reservoir segment (main basin and riverine), at each of the same 5 sampling points at which in situ chemistry was measured, we collected a 400 ml sample of water from the upper 0.25 meters of the water column. 2. These 5 samples were composited into a single 2 L container, which was agitated prior to dispensing contents into subsample containers. 3. Subsamples were placed on ice and returned to the Central Plains Center for BioAssessment (CPCB) laboratory for storage, processing, and shipment (Table 6). 4. One 1 L subsample was filtered and processed for TSS, and VSS using the Gravimetric method, filtering 50 – 200 ml of sample water, drying the filter and sample at 103˚C, then igniting volatile solids at 500˚C, recording the filter weight before and after each step. 5. One 1 L sample was filtered for chlorophyll-a and phaeophyton-a analyses. Frozen filters were sent to the Iowa State Hygienic Lab for analyses. 6. The 50 ml subsample were sent to Kansas State University (KSU) soils lab for total nitrogen and total phosphorus content analyses by persulfate digestion (Alpken 1986a and b). 7. The 40 ml subsample was delivered to the KU Tertiary Oil Recovery Program (TORP) lab for TOC analysis.

14

Table 6. Parameter precision (repeatability) and accuracy for water samples analyzed in laboratories. Detection Hold Preserv Parameter Container Instrument Method Limit Time ation 21st Ed. Standard 1000 ml Benchtop Chlorophyll a Methods (APHA) 1.0 g L-1 28 days -20oC plastic fluorometer 10200-H Total Gravimetric; Suspended & Volatile 21st Ed. Standard 1000 ml Total Volatile Solids Methods (APHA) 1 mg L-1 7 days 4oC plastic Solids Ignited at 500 2540D&E (TSS & VSS) ˚C EPA Methods 415.1 & 40 ml 415.3 glass EPA Method 9060A amber vial 4oC Teledyne Standard Method 5310C with 20 H SO TOC Tekmar ASTM D4779 50 ppb 28 days 2 4 drops to Torch ASTM D4839 phos- pH<2 Cleaning Validation phoric USP TOC Method acid <643>/EP 2.2.44/JP Acid- 100 ml RFA Methodology no. TN* persulfate 0.2 ppm NA 4oC plastic A303-S170 digestion Acid- 100 ml RFA Methodology no. TP* persulfate 0.1 ppm NA 4oC plastic A303-S200-13 digestion *Typical preservation with H2SO4 and 28-day holding time is not applicable to this project.

Sediment collection

See the project QAPP for detailed sediment collection field methods and equipment. 1. In each reservoir segment (main basin and riverine segments), at each of the same 5 sampling points at which in situ chemistry was measured, sediment samples were collected from the upper 4 cm of sediment using a Wildco® liner-type Hand Corer or equivalent device. 2. Core diameter and length was recorded (all cores were 4.9 cm diameter). 3. Each core was placed in a separate labeled small plastic bag, the brand and size of which remained the same throughout the study as the average weight of these bags were used in calculations. 4. All 5 bagged cores from a site were placed into a larger bag labeled with site number, lake, and date which was sent to the University of Kansas (KU) Pedology Laboratory for analyses of three soil particle size classes, bulk density, total phosphorus and nitrogen, and percent organic matter (Table 7).

15

Table 7. Parameter precision (repeatability) and accuracy for sediment samples analyzed by the KU Pedology Laboratory. Detection Holding Parameter Container Instrument Method Preservation Limit Time Particle class Plastic Hydrometer 1% NA 60 d none size bag Plastic Core 0.1 g/mL Bulk density NA 30 d none bag Salicylic- ~0.1% Plastic Flow Sulfuric TN 30 d 4oC (on ice) bag analyzer Acid Digestion Salicylic- ~0.1% Plastic Flow Sulfuric TP 30 d 4oC (on ice) bag analyzer Acid Digestion % organic Plastic Carbon Coulometric 0.01% 60 d 4oC (on ice) matter bag Coulometer tritration % inorganic Plastic Carbon Coulometric 0.01% 60 d 4oC (on ice) matter bag Coulometer tritration

Vegetation

Because plant communities are the most recognized biological feature of wetlands and contribute greatly to habitat structure, biological sampling focused on measures of the phytoplankton and macrophyte communities. As mentioned, a subsample of the composite water was analyzed for phytoplankton chlorophyll-a and phaeophytin-a analyses as estimates of phytoplankton biomass.

Qualitative observational data on the occurrence of wetland and aquatic plants was used to generate a taxonomic list of plants occurring within the sampling site and not used to estimate vegetative coverages. These qualitative data was also used to generate a presence and absence plant community list for analysis and comparisons.

We used the macrophyte sampling method employed by the Kansas Department of Health and Environment to quantify vegetative cover along the two water sample transects. This KDHE method is based on the “point-quadrat method” (KDHE 2014). An outline of this method is presented below and more detail can be found in KDHE’s “Lake and wetland water quality monitoring program quality assurance management plan” (KDHE 2014).

Vegetation sampling methods 1. Ten vegetation sampling points were evenly distributed along each sampling transect. 2. At each sampling point the field crew made a visual inspection for macrophytes, and also noted terrestrial detritus and debris. 3. To sample submerged plants, a rake head on a rope was thrown over the boat, dragged over the lake bottom for 5 – 6 m, and retrieved.

16

4. Macrophytes snared on the rake were identified to the lowest feasible taxonomic level in the field if possible, or placed in labeled plastic bags, placed on ice, and returned to the McGregor Herbarium for identification by Dr. Craig Freeman. No plants were found that required ethanol preservation (i.e. Azolla, Ceratophyllum, Lemna, Najas, Spirodela, Wolffia). 5. Macrophyte abundance was reported as "percent of stations with macrophytes" (calculated as number of stations/20) and used as a surrogate of areal cover. The same metric was applied to each species as a measure of relative species abundance, while the number of species provides an estimation of species richness. 6. The extent of predominant macrophyte taxa that form extensive surface or subsurface colonies (≥ 100 m2) was noted on field maps. 7. All invasive and exotic aquatic plant species were noted on field maps. These include species listed in Midwest Aquatic Plant Management Society’s Plant Reference Chart (http://www.mapms.org/wp-content/uploads/2013/01/plantid.pdf) and by Kansas Department of Wildlife, Parks and Tourism (KDWPT 2014).

Birds and nuisance species

1. Bird species seen within or adjacent to each main basin and riverine site were recorded and submitted to the Cornell Ornithology Laboratory eBird website (http://ebird.org/content/ebird/). Unfortunately data retrieved from eBird was not specific for each site, but recorded for each reservoir as a whole. 2. Nuisance aquatic animal species observed in these study segments were recorded.

Results

Reservoir water elevation fluctuations

Kirwin, and Webster show a large change in depth between the 25th and 50th percentile lake extents, whereas the remaining lakes show a depth difference of less than 2.5 feet (Table 4). When paired with inflow data one can conclude that these three reservoirs typically have a low water level, which would lead to a greater depth change when big precipitation events occur. For example, a reservoir at a low water level will have a greater depth change during a precipitation event than the same precipitation event would have at the reservoir’s regulation level. Webster and Kirwin show a larger difference between the 75th and 50th percentile water elevations, while the other reservoirs all fall under a depth differences of less than 2.5 feet (Table 8).

The depth ranges seen between these lake extents describe some of the hydrological trends found at these reservoirs. For example, the larger the difference between these lake extents indicate the more varied extremes in reservoir water surface elevation. The reservoirs shown which have a larger depth difference are all found in the western part of the state, which has a drier climate.

The variability of the water elevation for each reservoir was assessed using box plots of their daily historical water elevations starting the day the reservoir first reached its regulation level (Figure 4). Box plots of historical daily water elevations can display a large amount of

17 information about how a specific reservoir is regulated. For example, within a boxplot, the top of the box is the 75th percentile elevation value, the bottom of the box is the 25th percentile elevation, and the line in the middle of the box is the median of the elevation values. The points that fall outside of the inner fences are said to be outliers and represent extreme elevations (Wilks 2001). Fifty percent of the dataset fall within the box, which displays the spread of the central part of the data. All of these federal reservoirs are highly regulated; however, the reservoirs in the western half of Kansas are less likely to reach their regulation level due to the lack of rainfall and inflows.

The reservoirs in the eastern half of Kansas are kept near their regulation level because they receive more inflow and therefore can better control the discharge and the elevation level of the reservoir, which can be shown using a boxplot with a small box covering a small range of elevations. This means that most of the time, the elevation of the lake is very close to the normal or regulated level. Big Hill, Clinton, Melvern, Pomona, and Toronto are able to be more regulated than some of the other reservoirs with different hydrological patterns. The boxplots of Big Hill, Clinton, Melvern, Pomona, and Toronto show outliers above the upper inner fence, indicating some extreme flooding events that raise the water elevation up higher than the normal spread of the data. Cheney, Council Grove, El Dorado, Elk City, Fall River, Hillsdale, Kanopolis, Marion, Milford, Perry, Tuttle Creek, and Wilson are similar in the fact that they are highly regulated with little variation in elevation fifty percent of the time, with extreme outliers indicating flooding events. However, unlike Big Hill, Clinton, Melvern, Pomona, and Toronto, these plots also have outliers below the 25th percentile, indicating extreme drought conditions where the reservoir water elevation falls below the normal spread of the data. Kirwin and Webster have a larger range of elevation values between the 75th and the 25th percentiles, indicating that these lakes aren’t as regulated as some of the others. The lack of outliers for Kirwin and Webster are due to the elevation being so varied. All the elevation values fall within either the percentiles or the inner fences. This corresponds to the climate differences found across Kansas with generally a smaller range of water elevations in the east and a wider range of elevations in the west, outliers of both high (flood conditions) and low (drought conditions) in the east, and less regulation of discharge in the western reservoirs.

18

Figure 4. Box plots of the reservoir water surface elevation (ft) for 22 federally owned reservoirs in Kansas using daily historical data. Outliers are shown in black. One study reservoir (John

19

Redmond) is not shown because our dataset contained only two years of records following the 2013 conservation pool level raise. Also, three non-study reservoirs (Cedar Bluff, Keith Sebelius, Waconda) are included. Note that the four reservoirs not showing outliers (Cedar Bluff, Keith Sebelius, Kirwin, Webster) are the westernmost federal reservoirs in the state, reflective of the difficulties of maintaining a consistent water level due to limited inflows.

Table 8. 25th, 50th, 75th, and 100th percentile water level elevations (ft) for the 20 study reservoirs. Adapted from Loeffler et al. 2018. Reservoir 25th percentile 50th percentile 75th percentile 100th percentile Big Hill 856.8 857.7 858.2 865.7 Cheney 1420.1 1421.6 1421.8 1429.4 Clinton 874.7 875.8 877.1 892.5 Council Grove 1271.6 1273.6 1274.1 1290.2 El Dorado 1337.2 1338.7 1339.1 1347.3 Elk City 793.8 795.0 796.9 829.7 Fall River 948.5 949.0 950.6 986.7 Hillsdale 915.0 916.3 917.6 928.5 John Redmond 1041.5 1042.5 1043.7 1068.8 Kanopolis 1463.1 1464.3 1466.0 1505.7 Kirwin 1715.2 1726.2 1729.5 1737.1 Marion 1349.2 1349.9 1350.5 1356.1 Melvern 1034.0 1035.4 1036.7 1053.5 Milford 1143.0 1144.5 1145.8 1181.9 Perry 890.5 891.8 893.6 920.9 Pomona 972.7 974.0 975.7 998.4 Toronto 901.3 901.9 903.4 933.2 Tuttle Creek 1073.6 1075.5 1078.4 1137.7 Webster 1871.2 1888.0 1892.9 1896.8 Wilson 1513.6 1515.2 1516.2 1548.2

NWI type and occurrence characteristics in the flood pool

To get a sense of wetland characteristics and propensities in the flood pool (area between 50th and 100th lake level percentiles), we examined NWI polygons within this zone from each reservoir. Specifically, we partitioned the flood pool into two subunits, the upper fluctuation zone (UFZ; between 50th and 75th percentiles) that is more frequently hydrologically connected to the reservoir, and the upper flood pool (UFP; between 75th and 100th percentiles) that is less frequently hydrologically connected to the reservoir. Methods and results, which we summarize here, are described in greater detail in Loeffler et al. 2018 (App. B). UFZ and UFP delineations

20 for Elk City and Webster can be seen in Figure 5 while Figure 6 illustrates NWI coverage for Clinton with respect to these zones.

Figure 5. 25th, 50th, and 75th percentile lake extents for (A) Elk City and (B) Webster reservoirs. Background is 2015 USDA-FSA National Agriculture Imagery Program data.

21

Figure 6. NWI wetlands around Clinton Lake. Background is 2015 USDA-FSA National Agriculture Imagery Program (NAIP) data.

Within the flood pool, the predominant NWI wetland type is forested/shrub wetlands, followed by emergent wetlands, riverine wetlands, and ponds. Descriptions for these four freshwater wetland categories along with reservoir-specific flood pool percentage compositions can be found in Loeffler et al. 2018 (App. B). Restricting the assessment to the less reservoir-impacted UFP portion of the flood pool resulted in an increase in the proportions of riverine wetlands and ponds compared to the flood pool as a whole. By contrast, in the UFZ portion of the flood pool, the occurrence rate of these two wetland types was reduced by about half compared to their proportional representation among NWI wetlands throughout the flood pool as a whole. The biggest fractional gains within the UFZs occurred in the emergent wetland class, which largely can be attributed to the appearance of “shoreline” wetland communities that tended to follow bathymetry/LiDAR elevation contours.

For the next assessment, we computed LiDAR slope within the flood pool and examined NWI occurrence tendencies with respect to ground slope in the UFZ compared to the UFP. Our hypothesis was that low-sloped ground in the more frequently inundated UFZ would have a greater propensity to be classified as an NWI wetland than in the UFP. To test this idea, we examined median slope values restricted to NWI polygons within each reservoir’s flood pool; if the median NWI slope in the UFZ was less than the median NWI slope in the UFP, this would be consistent with our hypothesis.

To strengthen the assessment, we utilized only flood pool LiDAR data without including any bathymetry data, which are much less accurate and precise. We ignored pixels with zero slope, which were almost certain to have been part of a LiDAR hydro-flattened area and thus already

22 identified as a waterbody when the LiDAR raster data were originally developed. For example, this assured that any hydro-flattened portion of the main reservoir appearing in the flood pool could not bias the outcome.

Across the 20 study reservoirs, the average median percent slope underlying NWI polygons in the UFP was 3.2%, compared to 2.8% in the UFZ. Using the reservoir-specific UFP NWI slope as a maximum slope threshold resulted in an average capture of 54.6% (±11.7%) of the UFZ NWI area, an outcome that is weakly consistent with the hypothesis. However, when we spatially aggregated all of the flood pools together for a single (effectively area-weighted) evaluation, median percent slopes for UFP NWI and UFZ NWI were 2.1% and 1.4%, respectively. Applying the UFP NWI median slope as a threshold in this combined-reservoir framework resulted in the capture of 62.0% of the UFZ NWI area, a stronger result that better supports the hypothesis.

The implications from this flood pool assessment are two-fold. First, we found that regular fluctuations in the water level of managed reservoirs support the natural development of emergent wetlands near the reservoir perimeter. Second, with the increasing availability of high- quality LiDAR elevation data in Kansas and elsewhere, low-slope areas within the frequently inundated fluctuation zones around managed reservoirs have an increased likelihood for wetland development than do nearby, similarly sloped lands higher up in the flood pool. This outcome could allow for more optimal targeting of flood pool locations for wetland development or construction.

Discussion

The USEPA Region 7 RTAG water column benchmark values for lakes and reservoirs occurring in Region 7 (excluding Sand Hills lakes) are 0.70 ppm TN, 0.035 ppm for TP, and 0.008 ppm for chlorophyll a (USEPA RTAG7 2011, Table 9). Benchmark values for streams occurring in Region 7 are 0.9 ppm TN, 0.075 ppm TP, and 8 ug/l sestonic chlorophyll a (USEPA RTAG7 2016).

Table 9. Values from this study compared to USEPA Region 7 water column benchmark values for total nitrogen (TN), total phosphorus (TP) and chlorophyll a (chla). Location parameter Count Mean Median Minimum Maximum Main Basin TN ppm 20 0.92 0.76 0.49 2.78 Riverine TN ppm 40 1.24 1.09 0.60 4.27 RTAG lake TN ppm 0.70 RTAG stream TN ppm 0.90

Main Basin TP ppm 20 0.18 0.09 0.02 1.15 Riverine TP ppm 40 0.24 0.16 0.05 1.34 RTAG lake TP ppm 0.04 RTAG stream TP ppm 0.08

23

Location parameter Count Mean Median Minimum Maximum Main Basin chla ug/l 20 16.15 14.00 3.00 52.00 Riverine chla ug/l 40 26.55 25.00 6.00 61.00 RTAG lake chla ug/l 8.00 RTAG stream chla ug/l* 8.00

*Sestonic chlorophyll a.

All mean nutrient and chlorophyll a values for these study reservoirs were over the recommended nutrient and chlorophyll concentrations for regional lakes and rivers. Both main basin mean and median values for TN exceed the benchmark levels for this nutrient but not excessively. However study reservoir mean and median levels for TP and chlorophyll a were much higher (> 2X) than the suggested benchmark levels. This is especially true for chlorophyll a mean and median concentrations for riverine values even when compared to RTAG river and stream benchmarks. Mean riverine levels of chlorophyll a were three times higher than the river and stream benchmark value for this response variable, indicating that these reservoir regions were hyper-eutrophic even when compared to regional stream benchmark values that are much higher for TP but not chlorophyll a. It is not known if these riverine regions were responding directly to river inflow concentrations, but we did place all sampling sites well way from inflowing river water by selecting sites laterally and often above any incoming channel flow (Figure 7). This would suggest that both TP and to a lesser extent TN is being elevated in these upper riverine areas which in turn seems to be stimulating phytoplankton biomass. While main basin concentrations of nutrients were somewhat higher than lake benchmark values they were much lower than corresponding riverine levels suggesting these two areas of the reservoirs were responding to different in-reservoir factors such as water depth, distance from inflows and turbidity.

24

Figure 7. Riverine site TUB on the Blue River Riverine region of Tuttle Creek Reservoir. Riverine sample site was located outside the Blue River inflow channel and mean turbidity was 1,406 NTUs compared to the main basin mean turbidity of just under 18 NTU.

Box plots

Notched box plots were created to visibly examine the spread of the data, to note outliers, and to see differences between main basin and riverine data. Overlap of the notches of two populations indicate that their medians do not significantly differ. Outliers were removed from some analyses. The severe outliers are mostly in the riverine segments that are the most unstable water quality-wise due to river water inflows with minimal dilution and shallow depths creating mixing and turbulence. Main basin had significantly higher values than riverine sites (One-way ANOVA p <0.05) for total and inorganic carbon, TN and TP, and percent clay in sediment (Figure 8 - Figure 10). Riverine sites had significantly higher values of water column ORP, turbidity, TSS, NSS, NVSS, chlorophyll a, TN, and TP, and sediment percent silt and bulk density (Figure 11 -

25

Figure 17). Other in situ water quality parameters did not differ between main basin and riverine sites.

Inorganic carbon values in sediments were significantly different between main basin and riverine sites with both lower variance and values for the riverine areas. Carbon enters lakes and reservoirs from their watershed as inflows as DIC (dissolved inorganic carbon), DOC and POC (dissolved and particulate organic carbon, respectively). The dominant form of carbon is mostly determined by geography and hydrology with carbon also entering the system as atmospheric CO2 (Balmer and Downing 2011). Water bodies that are particularly eutrophic may actually show a net influx of atmospheric carbon into the system during summer months due to high levels of photosynthetic algae (Balmer and Downing 2011). Within a waterbody, CO2, which may comprise a large portion of total DIC, is both generated from respiration and utilized for photosynthesis. A portion of water DIC can settle out and become incorporated in DIC already found in lakes sediments. Our sediment findings indicate that riverine DIC levels are consistently lower than main basin values which tend to vary greatly across study reservoirs suggesting that in-reservoir processes are contributing to DIC levels perhaps through gains in atmospheric CO2. However, McDonald and coworkers (McDonald et al. 2011) found that on a national level DIC loading was a major determinant of CO2 concentrations and fluxes across the air-water interface in the majority of lakes and reservoirs studied. Our data is so limited that little can be said about DIC sources and fluxes in either water or sediment but it is clear that high levels of DIC are occurring in the more lake-like regions of the reservoir than the sediment deposited in the riverine reaches.

Of the three particle size groups measured in sediment samples, clay is the size group most likely to disperse across the reservoir system. This is because clay particles are very small, plate-like in shape and have both a positive and negative charge on opposite end of the particle. This alone often keeps clay particles in suspension nearly indefinitely.

26

Figure 8. Percent inorganic carbon in sediment, 11 outliers removed from box plots and one way ANOVA (F(1,47)=5.44, p=0.02). M = main basin, R = riverine samples.

Figure 9. Percent clay in sediment. One-way ANOVA (F(1,58)=14.18, p=0.00). M = main basin, R = riverine samples.

27

Figure 10. Percent silt in sediment, 4 outliers removed from one-way ANOVA (F(1,54)=17.81, p=0.00). M = main basin, R = riverine samples.

Figure 11. Total nitrogen ppm in sediment. One-way ANOVA (F(1,57)=10.40, p=0.00). M = main basin, R = riverine samples.

28

Figure 12. Total phosphorus ppm in sediment. One-way ANOVA (F(1,57)=17.94, p=0.00). M = main basin, R = riverine samples.

Figure 13. In situ water column turbidity (NTU), 3 outliers removed from boxplots and one-way ANOVA (F(1,55)=38.00, p=0.00). M = main basin, R = riverine samples.

29

Figure 14. Total suspended solids (TSS mg/l) in the water column, 6 outliers removed from box plots and one-way ANOVA (F(1,52)=35.09, p=0.00). M = main basin, R = riverine samples.

Figure 15. Volatile suspended solids (VSS mg/l) in the water column, 4 outliers removed from box plots and one-way ANOVA (F(1,54)=40.83, p=0.00). M = main basin, R = riverine samples.

30

Figure 16. Percent non-volatile suspended solids (NVSS) in the water column, 6 outliers removed from box plots and one-way ANOVA (F(1,52)=13.17, p=0.00). M = main basin, R = riverine samples.

Figure 17. Sediment bulk density (g/cm3). One-way ANOVA (F(1,58)=30.18, p=0.00). M = main basin, R = riverine samples.

31

Figure 18. Water column chlorophyll a (ug/L). One-way ANOVA (F(1,58)=9.53, p=0.00). M = main basin, R = riverine samples. Lake RTAG recommendation of 8 ug/l shown as dashed line.

Figure 19. Water column total nitrogen (ppm), 5 outliers removed from one-way ANOVA (F(1,53)=23.69, p=0.00). M = main basin, R = riverine samples. Lake RTAG recommendation of 0.70 ppm shown as dashed line.

32

Figure 20. Water column total phosphorus (ppm), 7 outliers removed from one-way ANOVA, (F(1,52)=10.07, p=0.00). M = main basin, R = riverine samples. Lake RTAG recommendation of 0.04 ppm shown as dashed line.

Across the entire dataset (minus outliers), regardless of site location, we looked at Pearson correlation coefficients (r) for water and sediment variables which were linear. Correlations may be impacted by the shallowness of riverine sites, as seen in scatter plots by site type (Figure 21 - Figure 22, all scatterplots are in App. A). Water depth significantly (p<0.05) correlated with all variables except % sand, salinity, TDS, conductivity, DO, ORP, and pH (all |r| > 0.34, (Table 10). Turbidity also correlated to many variables, including water TN (0.53) and TP (0.49), and sediment TN (-0.36) and TP (-0.30). Among nutrients, sediment TN and TP correlated (0.79) and water column TN and TP correlated (0.44). Sediment TP did not correlate with water TP, nor did sediment TN correlate with water TN. However, sediment TN negatively correlated (-0.39) with water TP. Other notable correlations are: % clay correlated with sediment TP (0.44) but not water TP; % clay correlated negatively with water TN (-0.31) but not sediment TN; % sand correlated negatively with both sediment TN (-0.36) and % total carbon (-0.39); % total carbon correlated with sediment TP (0.50) and TN (0.48). DO correlated with only pH (0.39). Chlorophyll a correlated with sediment TP (-0.32), % clay (-0.41), and water TN (0.59).

Bulk density negatively correlated (r = -0.65) with water depth, indicating that sediment in the main basin was less dense, which is evident in the box plot. Morris and Fan 1998 also reported higher sediment bulk density in the delta region (branches) of reservoirs and densities decreased in the direction of the dam. All bulk density values were less than 1.6 g/cm3, above which root growth tends to be restricted in soils (McKenzie et al. 2004). However, in terrestrial systems, other soil factors may also limit root growth (Hunt and Gilkes 1992). They also suggest that

33 sandy soils tend to have higher bulk density, however, in this study sediment bulk density showed low correlation with clay, (r = -0.26) and no correlation with % sand or silt. Given the lack of correlations between bulk densities and particle sizes in general it appears that the differences in riverine and main basin bulk densities were more attributable to the fluidized nature of the main basin sediment layer.

The scatter plot between depth and turbidity (Figure 21) suggests shallow depths may contribute to overall high turbidities but shallow depths in our study are autocorrelated with location within the reservoir systems (riverine vs. main basin). We suspect that both the inflowing river water is more turbid than the reservoir water and that the shallow depths of the riverine areas create greater opportunities for sediment resuspension that also added to total turbidity levels.

Figure 21. Scatter plot of site depth vs turbidity. Black circles = riverine sites, gray circles = main basin sites. Outliers removed.

Water column and sediment TN and TP were not highly correlated with each other except for water column TP and sediment TN which were negatively related. No explanation could be offered regarding this significant but modest relationship. Our analyses of water column TP and TN verses depth (Figure 22) all suggest that depth, location and turbidity or some combination of these variables are responsible for the observed differences in reservoir nutrients. Again inflowing river water could be higher in nutrients which are often the case when compared to downstream reservoir levels but these nutrients are also highly correlated with VSS. However VSS as well as TSS and NVSS are also strongly related to water column TP and TN, and these measures of suspended materials could be, in part, attributed to resuspension of sediments due to 34 shallow depths associated with riverine areas. Total nitrogen in the water column was also positively correlated to percent silt which was higher in the main basin areas suggesting that percent silt was not accounting for higher TN in the more turbid riverine areas.

Figure 22. Scatter plot of water column total phosphorus and total nitrogen and depth. Black circles = riverine sites, gray circles = main basin sites. Outliers removed.

35

Chlorophyll a concentrations were significantly higher in the riverine regions of the study reservoirs when compared to the main basin values. Water column levels of both TN and TP were also higher in the riverine areas but turbidity levels in the riverine regions were significantly higher. Chlorophyll a levels were also positively correlated to turbidity (r = 0.37) but more strongly related to VSS (r = 0.85). We believe that even in the more turbid riverine areas light was not a limiting factor, and the strong correlation between VSS and chlorophyll a was due to the high contribution of algal cells to the VSS concentrations in this region.

Chlorophyll a concentrations were positively related to TN and not TP concentrations which is something we see in many of our studies of Kansas reservoirs (Dzialowski et al. 2005, 2008, 2009; Wang 2005). This apparent relationship may, in part, be because most of our chlorophyll a concentrations are due to high cyanobacteria levels that tend to relate to TN levels (Yuan et al. 2014,), however it has been noted that high TN can also have a negative effect of chlorophyll a levels (see Filstrup et al. 2018).

Table 10. Pearson correlations (outliers removed) with r values given in the variable name row, and probability (p) and count (n). Significant correlations are shaded, and these also had significant linear regressions. sed. sed. bulk % % water water Water TP TN den. inorg. total % % TP TN TN:TP ppm ppm g/cm3 C C sand % silt clay ppm ppm depth m 0.52 0.56 0.52 -0.65 0.35 0.46 -0.13 -0.34 0.39 -0.46 -0.47 p 0.00 0.00 0.00 0.00 0.01 0.00 0.33 0.01 0.00 0.00 0.00 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 temp. C 0.16 -0.04 -0.01 -0.33 0.05 -0.05 0.08 0.19 -0.31 -0.31 -0.08 p 0.27 0.77 0.92 0.01 0.75 0.72 0.55 0.16 0.02 0.02 0.58 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 pH -0.21 -0.30 -0.25 -0.02 0.07 -0.10 0.21 0.10 -0.27 0.21 0.12 p 0.15 0.02 0.06 0.87 0.63 0.46 0.10 0.47 0.03 0.12 0.38 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 ORP 0.05 0.12 0.10 0.33 -0.01 0.06 -0.10 0.01 0.11 0.06 0.14 p 0.74 0.38 0.48 0.01 0.95 0.68 0.46 0.92 0.40 0.67 0.31 n 46.00 56.00 56.00 57.00 47.00 50.00 57.00 53.00 57.00 50.00 52.00 cond. mS/cm -0.08 -0.03 0.22 -0.16 0.84 0.13 0.29 0.02 -0.41 -0.18 0.03 p 0.58 0.81 0.09 0.21 0.00 0.35 0.02 0.87 0.00 0.21 0.84 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 turb. NTU -0.36 -0.30 -0.36 0.63 -0.28 -0.27 0.02 0.33 -0.23 0.49 0.53 p 0.01 0.02 0.01 0.00 0.05 0.06 0.86 0.02 0.08 0.00 0.00 n 48.00 56.00 56.00 57.00 49.00 50.00 57.00 53.00 57.00 52.00 54.00 Ave. DO mg/L 0.18 -0.16 -0.15 0.11 0.05 -0.02 0.12 -0.11 0.12 0.00 0.15 p 0.23 0.22 0.27 0.39 0.74 0.87 0.38 0.41 0.35 1.00 0.28 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 Ave. TDS -0.08 -0.03 0.22 -0.16 0.84 0.13 0.29 0.02 -0.40 -0.17 0.03 p 0.57 0.82 0.09 0.23 0.00 0.34 0.03 0.87 0.00 0.21 0.82 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00

36

sed. sed. bulk % % water water Water TP TN den. inorg. total % % TP TN TN:TP ppm ppm g/cm3 C C sand % silt clay ppm ppm Ave. salinity % -0.08 -0.05 0.21 -0.15 0.84 0.12 0.32 0.00 -0.41 -0.17 0.04 p 0.60 0.72 0.11 0.25 0.00 0.40 0.01 0.97 0.00 0.22 0.78 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 TSS mg/L -0.40 -0.36 -0.36 0.53 -0.26 -0.22 -0.11 0.32 -0.17 0.50 0.48 p 0.01 0.01 0.01 0.00 0.08 0.13 0.43 0.02 0.22 0.00 0.00 n 45.00 53.00 53.00 54.00 47.00 47.00 54.00 51.00 54.00 49.00 51.00 VSS mg/L -0.35 -0.42 -0.35 0.50 -0.23 -0.32 0.14 0.38 -0.36 0.46 0.57 p 0.02 0.00 0.01 0.00 0.12 0.02 0.31 0.01 0.01 0.00 0.00 n 47.00 55.00 55.00 56.00 48.00 49.00 56.00 52.00 56.00 51.00 53.00 NVSS % -0.21 -0.05 -0.30 0.38 -0.57 -0.18 -0.10 -0.01 0.14 0.35 0.09 p 0.18 0.74 0.03 0.00 0.00 0.20 0.47 0.92 0.30 0.02 0.53 n 44.00 53.00 53.00 54.00 45.00 50.00 54.00 50.00 54.00 47.00 49.00 chlal a ug/L -0.10 -0.32 -0.17 0.20 -0.15 -0.13 0.23 0.17 -0.41 0.26 0.59 p 0.50 0.01 0.19 0.13 0.32 0.34 0.07 0.22 0.00 0.06 0.00 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 60.00 53.00 55.00 water TN ppm -0.22 -0.25 -0.16 0.49 -0.09 0.02 0.05 0.33 -0.31 0.44 p 0.13 0.06 0.24 0.00 0.54 0.89 0.69 0.02 0.02 0.00 n 49.00 54.00 54.00 55.00 48.00 50.00 55.00 51.00 55.00 51.00 water TP ppm -0.79 -0.21 -0.39 0.31 -0.29 -0.14 -0.01 0.08 0.05 p 0.00 0.14 0.00 0.02 0.06 0.35 0.97 0.61 0.73 n 49.00 52.00 52.00 53.00 45.00 46.00 53.00 49.00 53.00 % clay 0.13 0.44 0.12 -0.26 -0.22 0.33 -0.55 -0.41 p 0.36 0.00 0.36 0.04 0.14 0.02 0.00 0.00 n 49.00 59.00 59.00 60.00 49.00 53.00 60.00 56.00 % silt -0.16 -0.26 -0.17 0.21 -0.06 -0.11 -0.52 p 0.30 0.05 0.21 0.12 0.70 0.43 0.00 n 45.00 55.00 55.00 56.00 45.00 49.00 56.00 % sand -0.06 -0.36 -0.12 0.24 0.17 -0.39 p 0.68 0.01 0.35 0.07 0.23 0.00 n 49.00 59.00 59.00 60.00 49.00 53.00 % total C 0.20 0.50 0.48 -0.42 0.13 p 0.18 0.00 0.00 0.00 0.38 n 45.00 53.00 53.00 53.00 47.00 % inorganic C 0.00 -0.14 -0.16 -0.03 p 0.99 0.33 0.28 0.85 n 44.00 49.00 49.00 49.00 bulk den. g/cm3 -0.33 -0.54 -0.58 p 0.02 0.00 0.00 n 49.00 59.00 59.00 sed. TN ppm 0.58 0.79 p 0.00 0.00 n 48.00 59.00

37

sed. sed. bulk % % water water Water TP TN den. inorg. total % % TP TN TN:TP ppm ppm g/cm3 C C sand % silt clay ppm ppm sed. TP ppm 0.34 p 0.02 n 48.00

Table 11 continued. Pearson correlations (outliers removed) with r values given in the variable name row, and probability (p) and count (n). Significant correlations are shaded, and these also had significant linear regressions. chla NVSS VSS TSS salinity DO turb. cond. temp. ug/L % mg/L mg/L % TDS mg/L NTU mS/cm ORP pH C depth m -0.40 -0.49 -0.61 -0.63 0.14 0.14 0.08 -0.62 0.14 -0.25 -0.08 0.01 p 0.00 0.00 0.00 0.00 0.28 0.29 0.54 0.00 0.28 0.06 0.54 0.93 n 60.00 54.00 56.00 54.00 60.00 60.00 60.00 57.00 60.00 57.00 60.00 60.00 temp. C 0.26 -0.11 0.29 -0.05 0.08 0.08 0.01 0.07 0.09 -0.35 0.28 p 0.04 0.42 0.03 0.72 0.54 0.55 0.93 0.59 0.51 0.01 0.03 n 60.00 54.00 56.00 54.00 60.00 60.00 60.00 57.00 60.00 57.00 60.00 pH 0.31 -0.27 0.26 0.06 0.11 0.09 0.39 0.11 0.09 -0.42 p 0.02 0.05 0.06 0.66 0.41 0.50 0.00 0.43 0.49 0.00 n 60.00 54.00 56.00 54.00 60.00 60.00 60.00 57.00 60.00 57.00 ORP -0.16 0.32 0.18 0.26 -0.11 -0.11 -0.01 0.32 -0.12 p 0.23 0.02 0.20 0.07 0.40 0.40 0.96 0.02 0.38 n 57.00 51.00 54.00 52.00 57.00 57.00 57.00 55.00 57.00 cond. mS/cm -0.06 -0.52 -0.17 -0.32 1.00 1.00 -0.01 -0.27 p 0.65 0.00 0.21 0.02 0.00 0.00 0.92 0.04 n 60.00 54.00 56.00 54.00 60.00 60.00 60.00 57.00 turb. NTU 0.37 0.54 0.85 0.85 -0.27 -0.27 0.13 p 0.00 0.00 0.00 0.00 0.04 0.04 0.35 n 57.00 51.00 56.00 54.00 57.00 57.00 57.00 Ave. DO mg/L 0.14 -0.13 0.15 -0.03 0.01 -0.01 p 0.27 0.36 0.26 0.84 0.97 0.92 n 60.00 54.00 56.00 54.00 60.00 60.00 Ave. TDS -0.06 -0.52 -0.17 -0.32 1.00 p 0.65 0.00 0.21 0.02 0.00 n 60.00 54.00 56.00 54.00 60.00 Ave. salinity % -0.05 -0.51 -0.16 -0.32 p 0.72 0.00 0.24 0.02 n 60.00 54.00 56.00 54.00 TSS mg/L 0.31 0.55 0.84 p 0.02 0.00 0.00 n 54.00 48.00 53.00 VSS mg/L 0.58 0.29 p 0.00 0.04 n 56.00 50.00 NVSS % -0.14

38

chla NVSS VSS TSS salinity DO turb. cond. temp. ug/L % mg/L mg/L % TDS mg/L NTU mS/cm ORP pH C p 0.32 n 54.00

TN:TP ratios

Individual TN:TP ratios were calculated for all riverine and main basin water samples. One-way ANOVA results suggested that there are significant differences between riverine and main basin TN:TP ratios (F(1,47)=5.64, p=0.02, Figure 23). Dzialowski and co-workers (2005) found that in comparing bioassay results to calculated TN:TP ratios for 19 small Kansas reservoirs, N-limited reservoirs had water column TN:TP ratios <18 (molar); reservoirs that were co-limited by N and P had water column TN:TP ratios between 20 and 46; and reservoirs that were P limited had water column TN:TP ratios >65. It appears that all our study sites are co-limited or TN-limited regardless of reservoir region, however main basin ratios tended to be higher and vary over a broader range of values. Some of the main basin ratios did approach the co-limiting ratios of 20- 46 observed by Dzialowski and co-workers. Filstrup and co-authors (2018) noted that nitrogen and phosphorus commonly co-limit primary productivity in lakes with higher levels of chlorophyll a associated with lakes having very high nutrient levels. Paerl et al. (2016) also found that co-limitations were more common than previously suspected and both TN and TP should be reduced to limit primary production in lakes and reservoirs. Dzialowski and others have noted that the majority of Kansas reservoirs in their studies had TP and chlorophyll concentrations that were indicative of eutrophic or hypereutrophic conditions (Dodds et al. 2006, Dzialowski et al. 2005, 2008, 2009; Wang 2005, USEPA 2011).

39

Figure 23. TN:TP in water (TN and TP outliers removed, 2 ratio outliers removed). One-way ANOVA (F(1,47)=5.64, p=0.02). M = main basin, R = riverine samples. Dashed line at TN:TP = 18 represents point of N-limitation found by Dzialowski et al. (2005).

Vegetation

Vegetation surveys were performed in the shallow water zones (i.e. average  1 meter) of riverine regions and the deep water zones of the reservoirs’ main basins adjacent the dam areas. These surveys were designed to collect and identify macrophytes that may be present as part of the aquatic bed class (AB) of Cowardin et al. (1979). Few macrophytes and wetland plants were found in these surveys and none were associated with main basin regions of any reservoir. This was somewhat expected based on previous experiences and considering the depths associated with these sample areas. In addition, to identifying macrophytes and wetland plant abundances as percent cover the occurrences of vegetation detritus and debris were tallied. Dead and decomposing plant material (e.g. leaves, twigs, stem fragments) were recorded as terrestrial to distinguish this vegetation material from living macrophytes. The only plant material found in main basin zones was this detritus/debris class and only in Marion, John Redmond and Toronto Reservoirs in southeastern Kansas. Macrophytes and wetland plants were recorded in the riverine regions of nine reservoirs – three in the drier western part of the state and the rest in the wetter southeastern portion of Kansas (Table 12). We used Spearman Rank correlation coefficients (r) for vegetation which is ranked data (0, 5, 10, 15, 20%, etc.). Analysis including both riverine and

40 main basin sites (minus outliers) showed only terrestrial vegetation significantly (p<0.05) correlated with depth, turbidity, TSS, VSS, NVSS, chlorophyll, % clay, % silt, % inorganic carbon, bulk density, and % macrophytes. Due to paucity of vegetation in the main basin we re- ran correlations for only the riverine sites.

Within riverine sites, percent terrestrial vegetation significantly (p<0.05) correlated with (r given) DO (-0.37), water TP (-0.33), water TN:TP (0.38), % total C (0.39), sediment TN (0.63) and TP (0.41), and percent macrophytes (0.42). Macrophytes additionally correlated with only water pH (-0.33).

Table 12. Percent cover and number of taxa of macrophytes in riverine sites in study reservoirs. Mean percent Number of Number of riverine Reservoir macrophytes/wetland taxa branches sampled plant cover Wilson 40.0 3 2 Kirwin 30.0 1 1 Big Hill 20.0 1 1 Webster 23.3 1 3 El Dorado 5.0 1* 1 Pomona 3.3 1* 3 Toronto 2.5 1* 2 Council Groves 2.5 1* 2 Melvern 2.5 1* 2 Milford 0.0 0 1 Perry 0.0 0 3 Tuttle Creek 0.0 0 1 Cheney 0.0 0 2 Clinton 0.0 0 3 Elk City 0.0 0 3 Fall River 0.0 0 2 Hillsdale 0.0 0 3 John Redmond 0.0 0 1 Kanopolis 0.0 0 2 Marion 0.0 0 2

Table 13. Spearman Rank (pair-wise missing value removal) correlation matrix for percent macrophyte and terrestrial vegetation and water and sediment variables in all data (minus outliers). Significant (p<0.05) r values are highlighted. variable % macrophytes % terrestrial % macrophytes % terrestrial Ave. depth m -0.14 -0.72 water TN ppm 0.13 0.55 p 0.28 0.00 p 0.34 0.00 n 60.00 60.00 n 55.00 55.00 Ave. temp. C -0.03 0.10 water TP ppm -0.05 0.28

41

variable % macrophytes % terrestrial % macrophytes % terrestrial p 0.85 0.46 p 0.74 0.04 n 60.00 60.00 n 53.00 53.00 Ave. pH -0.22 -0.06 Water TN:TP 0.08 -0.04 p 0.09 0.66 p 0.59 0.80 n 60.00 60.00 n 49.00 49.00 Ave. ORP 0.20 0.22 % clay -0.25 -0.39 p 0.13 0.10 p 0.05 0.00 n 57.00 57.00 n 60.00 60.00 Ave. conductivity mS/cm -0.03 0.09 % silt 0.05 0.43 p 0.83 0.48 p 0.74 0.00 n 60.00 60.00 n 56.00 56.00 Ave. turbidity NTU 0.09 0.65 % sand 0.11 0.04 p 0.53 0.00 p 0.38 0.75 n 57.00 57.00 n 60.00 60.00 Ave. DO mg/L -0.14 -0.03 % total C -0.07 -0.10 p 0.29 0.81 p 0.60 0.50 n 60.00 60.00 n 53.00 53.00 Ave. TDS -0.02 0.09 % inorganic C -0.17 -0.36 p 0.90 0.50 p 0.26 0.01 n 60.00 60.00 n 49.00 49.00 Ave. salinity % 0.05 0.04 ave. bulk density g/cm3 0.01 0.48 p 0.69 0.78 p 0.97 0.00 n 60.00 60.00 n 60.00 60.00 TSS mg/L 0.08 0.61 sediment TN ppm 0.20 -0.05 p 0.55 0.00 p 0.13 0.70 n 54.00 54.00 n 59.00 59.00 VSS mg/L 0.17 0.65 sediment TP ppm -0.10 -0.20 p 0.22 0.00 p 0.47 0.13 n 56.00 56.00 n 59.00 59.00 NVSS % 0.17 0.43 percent macrophytes 1.00 0.47 p 0.22 0.00 p 0.00 0.00 n 54.00 54.00 n 60.00 60.00 chlorophyll a ug/L 0.04 0.39 percent terrestrial veg 0.47 1.00 p 0.75 0.00 p 0.00 0.00 n 60.00 60.00 n 60.00 60.00

Table 14. For only riverine sites, Spearman Rank (pair-wise missing value removal) correlation matrix for percent macrophyte and terrestrial vegetation and water and sediment variables in all data (minus outliers). Significant (p<0.05) r values are highlighted. % macrophytes % terrestrial % macrophytes % terrestrial Ave. depth m 0.22 -0.12 water TN ppm -0.09 0.17

42

% macrophytes % terrestrial % macrophytes % terrestrial p 0.17 0.46 p 0.59 0.30 n 40.00 40.00 n 37.00 37.00 Ave. temp. C -0.07 0.02 water TP ppm -0.31 -0.33 p 0.65 0.89 p 0.07 0.05 n 40.00 40.00 n 36.00 36.00 Ave. pH -0.33 -0.29 Water TN:TP 0.28 0.38 p 0.04 0.07 p 0.10 0.02 n 40.00 40.00 n 35.00 35.00 Ave. ORP 0.16 0.02 % clay -0.18 -0.23 p 0.35 0.92 p 0.26 0.15 n 37.00 37.00 n 40.00 40.00 Ave. conductivity mS/cm -0.05 0.17 % silt -0.14 0.15 p 0.75 0.30 p 0.40 0.37 n 40.00 40.00 n 36.00 36.00 Ave. turbidity NTU -0.20 0.18 % sand 0.12 0.01 p 0.23 0.29 p 0.48 0.97 n 37.00 37.00 n 40.00 40.00 Ave. DO mg/L -0.24 -0.37 % total C 0.00 0.39 p 0.13 0.02 p 0.98 0.02 n 40.00 40.00 n 35.00 35.00 Ave. TDS -0.04 0.16 % inorganic C -0.14 -0.20 p 0.81 0.31 p 0.46 0.28 n 40.00 40.00 n 32.00 32.00 ave. bulk Ave. salinity % 0.09 0.17 density g/cm3 -0.28 -0.07 p 0.57 0.28 p 0.08 0.67 n 40.00 40.00 n 40.00 40.00 sediment TN TSS mg/L -0.14 0.11 ppm 0.43 0.63 p 0.42 0.55 p 0.01 0.00 n 34.00 34.00 n 40.00 40.00 sediment TP VSS mg/L -0.14 0.04 ppm 0.05 0.41 p 0.42 0.82 p 0.75 0.01 n 37.00 37.00 n 40.00 40.00 percent NVSS % 0.11 0.25 macrophytes 1.00 0.42 p 0.53 0.14 p 0.00 0.01 n 36.00 36.00 n 40.00 40.00 percent chlorophyll a ug/L -0.14 0.12 terrestrial veg 0.42 1.00 p 0.40 0.47 p 0.01 0.00 n 40.00 40.00 n 40.00 40.00

43

Figure 24. Box plot of macrophytes and non-shoreline wetland plants in riverine and main basin. No wetland or macrophytes were taken at any main basin site. M = main basin, R = riverine samples.

Figure 25. Box plots of vegetation detritus and debris (= terrestrial vegetation category). M = main basin, R = riverine samples.

44

Shear stress

Logistic regression was used explore the potential of bottom shear stress as a limiting factor in wetland plant and macrophyte establishment in the nutrient rich, shallow-water riverine segments of our study sites. We were able to obtain daily wind speed and fetch for all study reservoirs as well as mean riverine depths over a period of 10 or more years of operation. These data were then used to calculate bottom shear stress (tau). We followed the approach used by Laenen and LeTourneau (1996) in our calculations of bottom shear stress. In order to calculate bottom shear stress one needs to compute wave period (T), wave length in shallow water (Lo), and wave height (H). Forecasting equations for waves in shallow water were used to develop wave geometry (USACE 1984).

Using the logistic regression program in NCSS 9 statistic software (NCSS 2013) we did a single run using presence and absence of wetland plant occurrence in each riverine sampling area and bottom shear stress measures (using mean wind and depth variables). This model was deemed good with a modest model R2 of 0.22. This model appeared to correctly estimate about 64% of the cases (i.e. presence or absences of plants) which is considered good to fair for a model outcome. A model with high discrimination ability will have high sensitivity and specificity simultaneously, leading to an ROC curve (Receiver Operating Characteristic Curve) which goes close to the top left corner of the plot. A model with no discrimination ability will have an ROC curve which is the 45 degree diagonal line. Our model’s ROC curves for both conditions showed indicated a moderate amount of model discrimination using only the mean value for bottom shear stress (Figure 26). Our model may have lost some of its discriminatory power because bottom shear stress values are a mean value calculated over years and many not represent current and prevailing conditions influencing our survey results during the time of our study.

We believe that further research needs to be done on other aspects and measures of wind and shear stress that are impacting wave, sediment resuspension and bottom stress and scour that could be limiting wetland development in these shallow riverine regions of reservoirs. Unfortunately we did not have the time or resources to fully explore the various potential impacts of wind on wetland development in our study reservoirs. New efforts need to better identify both open water and shoreline plant community development and quantify wind and water elevation influences on establishment and maintenance of wetland communities in aging reservoir systems.

45

Figure 26. Receiver Operating Characteristic Curves (ROC) from logistic regression of a single run using presence and absence of wetland plant occurrence in each riverine sampling area and bottom shear stress measures (using mean wind and depth variables). Red line (0) indicates sites with no wetland plants, Blue line (1) indicates sites with wetland plants, including Salix.

Conclusions and Future Studies

Several significant water quality differences were noted between riverine and main basin study areas. Most significant and important ecological distinctions between these reservoir regions was the typically high water column nutrients and various measures of turbidity and suspended materials. Total nitrogen and total phosphorus were higher in the waters of the riverine regions as were turbidity, TSS, VSS and NVSS. However, sediment nutrient concentrations and inorganic carbon were significantly higher in the main basin study areas when compared to the riverine areas. These significant differences in sediment chemistry were accompanied by significant differences in the physical nature of the sediment itself. Main basin sediment was composed of significantly higher percentage of clay but lower amounts of silt and main basin bulk densities were much lower than the more compacted riverine sediments.

Chlorophyll a concentrations were significantly higher in the riverine regions of our study reservoirs when compared to main basin regions despite much higher turbidity. The concentrations of water column TN and TP were very high in both reservoir regions and neither were considered to be limiting in nature. Total nitrogen to total phosphorus ratios were significantly higher in the main basin but varied greatly between reservoir main basin regions as 46 shown in Figure 23. In Texas reservoirs the strongest influence on phytoplankton concentrations along their longitudinal axis was depth with highest concentrations occurring in less than 0.9 m of water (Forbes et al. 2008). Similar longitudinal patterns of phytoplankton concentrations have been found in other reservoirs (Rychtecký and Znachor 2011, Wang et. al. 2011).

As mentioned, the higher turbidity of the riverine zones measured during the times of our study did not seem to negatively affect chlorophyll a concentrations in these regions. The correlation coefficient for turbidity vs. chlorophyll a was small (r = 0.37) but indicated a positive relationship between these two variables. While turbidity did not seem to limit algal production, given the near absence of vascular wetland plants and macrophytes it is unknown if high turbidities were controlling or eliminating occurrences of higher plants in these regions. There was no significant relationships noted between measures of turbidity and suspended solids as indicated in Table 13. The correlation coefficient values for all turbidity and suspended solids variables except for NVSS were small negative r values. While these r values were both small and non-significant their consistent negative relationships with the macrophyte class variable might suggest some possible relationship. These interpretations are limited by the fact that few reservoirs had any measurable amounts of wetland plants or macrophytes present in either the main basin or riverine zones sampled.

What was notable but unmeasured in this study was the influence of wind conditions on turbidity and surface and wave conditions. Our observational qualitative assessment is that high winds along with fetch length seem to highly affect wave intensity and duration which in term increase turbidities in the shallow riverine regions of our study reservoirs. In addition it is hypnotized that bottom shear stress increased with increased wave action thus generating sediment resuspension loads that increased turbidity levels and created unstable bottom sediment conditions that limited successful growth and maintenance of rooted vascular macrophytes and wetland plants. Analytical analyses of observed relationships between wind, waves, bottom shear stress measures and occurrences of open water wetland plants was not practical due to the lack of data for most of these variables. Even with measures of wind, wind fetch and bottom shear stress the very limited occurrences notes for wetland plants in our surveys limits their use to perhaps presence and absence data and logistic regression assessment with the potential and continuous measured causal variables such as wind velocities.

Further work with in identifying the relationships between reservoirs, their operation and wetland development within these systems needs to include measures of the impact of wind on reservoir water conditions. In addition, our use of existing NWI data and its spatial assessment with various water level variables was considered only partial effective. Examination of the old NWI data and their polygon locations suggested that these data need to be re-examined for both spatial and taxonomic accuracy. Much of the NWI data used in this study was mapped in the mid- to late 1980s making these data about 30 year old. Thus many NWI surveys were taken during the very early years of reservoir construction and operation. Changes in these now aging reservoir systems and wetland plant communities is suspected to have negatively influenced our assessment of current reservoir data with these 30-year old NWI data. In order to correctly assess current relationships between these reservoir systems and possible wetland development within them, current shoreline wetland surveys within each reservoir needs to be performed so that temporally concurrent data can be used in future assessments.

47

References

Antonius L. and A.P. LeTourneau. 1996. Upper Klamath Basin nutrient-loading study: Estimate of wind-induced resuspension of bed sediment during periods of low lake elevation. U.S. Geological Survey, Open-file Report 95-414, Portland, OR. APHA, AWWA, WEF. 2005. Standard methods for the examination of water and wastewater, 21st Ed. American Public Health Association, American Water Works Association, and Water Environment Federation. Washington, D.C. Balmer, M.B. and J.A. Downing. 2011. Carbon dioxide concentrations in eutrophic lakes: Undersaturation implies atmospheric uptake. Inland Waters 1:125-132. Batzer, D.P. and A.H. Baldwin. 2012. Wetland habitats of North America: ecology and conservation concerns. Univ. California Press: Oakland, CA. Boelee, E., S. Atapattu and T. Amede. 2011. Ecosystems for water and food security. UNEP: nairobi/colombo. Brown, R. 1984. Relationships between suspended solids, turbidity, light attenuation, and algal productivity. Lake and Reservoir Management, 1:198-205. Carney, E. 2002. Kansas Wetland Survey: Water Quality and Functional Potential of Public Wetland Areas. CD-997243-01-0, Kansas Department of Health and Environment: Topeka, KS. 46pp. Cowardin, L.M., V. Carter, F.C. Golet and E.T. LaRoe. 1979. Classification of wetlands and deepwater habitats of the United States. FWS/OBS-79/31. Fish and Wildlife Service, US Department of the Interior: Washington, DC. 142pp. Dingman, L. 2015. Physical Hydrology. Waveland Press, Inc., Long Grove, IL. Dodds W.K, C.E. Carney and R.T. Angelo. 2006. Determining ecoregional reference conditions for nutrients, secchi depth and chlorophyll a in Kansas lakes and reservoirs. Lake and Reservoir Management 22(2):151-159. Dzialowski, A.R., S. Wang, N. Lim, W.W. Spotts, D.G. Huggins. 2005. Nutrient limitation of phytoplankton growth in central plains reservoirs, USA. Journal of Plankton Research, 27(6):587–595. https://doi.org/10.1093/plankt/fbi034. Dzialowski, A.R., S. Wang, N. Lim, J.H. Beury and D.G. Huggins. 2008. Effects of sediment resuspension on nutrient concentrations and algal biomass in reservoirs of the Central Plains. Lake and Reservoir Management 24(4):313-320. Dzialowski, A. R., Smith, V. H., Huggins, D. G., Lim, N. C., Baker, D. S., & Beury, J. H. 2009. Development of predictive models for geosmin-related taste and odor in Kansas, USA, drinking water reservoirs. Water Research 43(11):2829-2840. Filstrup, C. T., Wagner, T., Oliver, S. K., Stow, C. A., Webster, K. E., Stanley, E. H., & Downing, J. A. 2018. Evidence for regional nitrogen stress on chlorophyll a in lakes across large landscape and climate gradients. Limnology and Oceanography, 63(S1), S324-S339. Forbes, M.G., R.D. Doyle, J.T. Scott, J. K. Stanley, H. Huangand, B.W. Brooks. 2008. Physical Factors Control Phytoplankton Production and Nitrogen Fixation in Eight Texas Reservoirs. Ecosystems 11:1181-1197. Fram, M.S., B.A. Bergamaschi and R. Fujii. 2001. Improving water quality in Sweetwater Reservoir, San Diego County, California; sources and mitigation strategies for trihalomethane (THM)-forming carbon. USGS Fact Sheet 112-01. 4pp. Freeman, C.C. 2012. Coefficients of conservatism for Kansas vascular plants (2012) and selected life history attributes. Kansas Biological Survey and R.L. McGregor Herbarium, Lawrence,

48

KS. http://biosurvey.ku.edu/sites/kbs.drupal.ku.edu/files/docs/Coefficients%20of%20Conservati sm%20for%20Kansas%20Vascular%20Plants%20%282012%29.pdf Fretwell, J.D., J.S. Williams and P.J. Redman. 1996. National water summary on wetland resources. Water-Supply Paper 2425, U.S. Government Printing Office: Washington, D.C. Ghanad, S., H. Moazed, S. Nasab, and N. Jafarzadeh. 2014. Effect of Reed and Hydraulic Retention Time on the Lead Removal in Horizontal Subsurface Flow of Constructed Wetlands. J. of Environmental Studies 40(4):937-948. Gilmer, D.S., M.R. Miller, R.D. Bauer and J.R. LeDonne. 1982. California's Central Valley wintering waterfowl: concerns and challenges. US Fish & Wildlife Publications 41. Goodin, D.G. J.E. Mitchell, M.C. Knapp and R.E. Bivens. 1995. Climate and Weather Atlas of Kansas: An Introduction, Kansas Geological Survey. Hintze, J.L. 2013. Number Cruncher Statistical System: Dr. Jerry L. Hintze, Kaysville, Utah. Hoffman, D.K., Wilhelm, S.W. and Wurtsbaugh, W.A. 2016. It takes two to tango: When and where dual nutrient (N & P) reductions are needed to protect lakes and downstream ecosystems. Environmental Science and Technology, 50(20), 10805-10813. Holdren, C., W. Jones, and J. Taggart. 2001. Managing Lakes and Reservoirs, Madison, WI. Houts, M., J. Neel, F.J. Norman, D. Baker, and J. Kastens. 2011a. Developing a methodology to identify existing and potential wetlands areas for protection, enhancement, and restoration. In: EPA Wetland Program Development Grant CD 97701101 Kansas Water Office Final Report. Kansas Water Office. Topeka, KS. Houts, M, J. Neel, F. Norman, and D. Baker. 2011b. Methodology to detect potential wetland areas, a user guide for the custom ArcGIS toolbox "Wetland Detection_v2. In: EPA Wetland Program Development Grant CD 97701101 Kansas Water Office Final Report. Kansas Water Office. Topeka, KS. Huggins, D., J. Kastens, D. Baker, and C. Freeman. 2017. Conversion of existing farm ponds to wetlands in agricultural landscapes for mitigation, land use treatment and conservation with a perspective toward climate change. Kansas Biological Survey Report No. 189. 91pp Hunt N. and R. Gilkes. 1992. Farm Monitoring Handbook. The University of Western Australia: Nedlands, WA. Jakubauskas, M., J. deNoyelles and E. Martinko. 2014. Bathymetric survey of John Redmond Reservoir, Coffey County, Kansas. 2014-01, Kansas Biological Survey, Applied Science and Technology for Reservoir Assessment (ASTRA) Program: Lawrence, KS. Johnson, W.C. 2002. Riparian vegetation diversity along regulated rivers: contribution of novel and relict habitats. Freshwater Biology 47(4): 749-759. Kansas Water Office. 2011. EPA Wetland Program Development Grant CD 97701101 Kansas Water Office Final Report. Topeka, KS. Kansas Water Office. 2013. Demonstration and Application of the Kansas Wetland Assessment Method. In: EPA Wetland Program Development Grant CD 97714501 Kansas Water Office Final Report. Topeka, KS. KDHE. 2014. Lake and wetland water quality monitoring program quality assurance management plan. Revision 4, 03/07/14, Kansas Department of Health and Environment, Division of Environment: Topeka, KS. KDWPT. 2014. Kansas Waters with Aquatic Nuisance Species. Kansas Department of Wildlife, Parks, and Tourism: Topeka, KS.

49

Keddy, P. and L.H. Fraser. 2000. Four general principles for the management and conservation of wetlands in large lakes: the role of water levels, nutrients, competitive hierarchies and centrifugal organization. Lakes & Reservoirs: Research & Management 5(3): 177-185. Keddy, P.A. 2010. Wetland ecology: principles and conservation. 2nd ed. Cambridge University Press: Cambridge, United Kingdom. Kentula, M.E. 2000. Perspectives on setting success criteria for wetland restoration. Ecological Engineering 15(3):199-209. Kusler, J.A. 1990. Wetland creation and restoration: the status of the science. Island press: Corvallis, OR. Levine, D.A. and D.E. Willard. 1990. Regional analysis of fringe wetlands in the Midwest: creation and restoration. In Wetland creation and restoration: the status of the science. Island Press, Washington, DC. 299-325. Lewis, W.M. 1995. Wetlands: characteristics and boundaries. The National Academies Press: Washington, D.C. Loeffler, K.M., V. Rahmani, J.H. Kastens, and D.H. Huggins. 2018. An examination of the potential wetland development landscape around managed reservoirs in the central U.S. Great Plains. J. Appl. Geography. 93:16-24. Mahlke, J. 1996. Characterization of Oklahoma Reservoir wetlands for preliminary change detection mapping using IRS-1B Satellite imagery. Geoscience and Remote Sensing Symposium, 1996. IGARSS '96. ‘Remote Sensing for a Sustainable Future.’ International, Lincoln, NE 3:1769-1771. McDonald, C.P., E.G. Stets, R.G. Striegl and D. Butman. 2013. Inorganic carbon loading as a primary driver of dissolved carbon dioxide concentrations in the lakes and reservoirs of the contiguous United States. Global Biochemical Cycles 27; 1–11. McKenzie N, K. Coughlan, and H. Cresswell. 2002. Soil Physical Measurement and Interpretation for Land Evaluation. CSIRO Publishing, Collingwood, Victoria. 379pp. Millennium Ecosystem Assessment. 2005. Ecosystems and Human Well-being: Wetlands and Water: Synthesis. World Resources Institute: Washington, DC. 80pp. Mitsch, W. J. 2005. Wetland creation, restoration, and conservation: the state of science. Newnes: Amsterdam, Netherlands. Morris, G.L., and J. Fan. 1998. Reservoir Sedimentation Handbook: Design and Management of Dams, Reservoirs, and Watersheds for Sustainable Use. McGraw Hill, 848pp. Moshiri, G. A. 1993. Constructed wetlands for water quality improvement. CRC Press: Boca Raton, FL. Multi Water Quality Checker U-50 Series Instruction Manual. 2009. Horiba Instruments Inc., Irvine, CA. April 2009. 129pp. NCSS 9. 2013. NCSS, LLC. Kaysville, UT, USA, ncss.com/software/ncss NOAA. 2017. Glossary of Hydrologic Terms. National Weather Service, National Oceanic and Atmospheric Administration. Available at: http://www.nws.noaa.gov/om/hod/SHManual/SHMan014_glossary.htm Paerl, H.W., J.T. Scott, M.J. McCarthy, S.E. Newell, W.S. Gardner, K.E. Havens, D.K. Hoffman, S.W. Wilhelm, W.A. Wurtsbaugh. 2016. It Takes Two to Tango: When and Where Dual Nutrient (N & P) Reductions Are Needed to Protect Lakes and Downstream Ecosystems. Environ. Sci. Technol. 50(20):10805-10813.

50

Pegg, M.A., K.L. Pope, L.A. Powell, K.C. Turek, J.J. Spurgeon, N.T. Stewart, N.P. Hogberg and M.T. Porath. 2015. Reservoir rehabilitations: seeking the fountain of youth. Fisheries 40(4):177-181. Rahmani, V., S.L. Hutchinson, J.A. Harrington Jr, J. Hutchinson and A. Anandhi. 2015. Analysis of temporal and spatial distribution and change-points for annual precipitation in Kansas, USA. International Journal of Climatology 35(13):3879-3887. Rychtecký, P. and P. Znachor. 2011. Spatial heterogeneity and seasonal succession of phytoplankton along the longitudinal gradient in a eutrophic reservoir. Hydrobiologia 663:175-186. Smith, L.M., N.H. Euliss, D.A. Wilcox and M.M. Brinson. 2008. Application of a geomorphic and temporal perspective to wetland management in North America. Wetlands 28(3):563- 577. Smith, R.D., A. Ammann, C. Bartoldus and M.M. Brinson. 1995. WRP-DE-9, DTIC Document: Washington, DC. Strand, M. and L. Rothschild. 2010. What Wetlands are Regulated? Jurisdiction of the §404 Program. The Environmental Law Reporter 40(4):10372-10393. Tiner, R.W. and N. Region. 2003. Correlating enhanced National Wetlands Inventory data with wetland functions for watershed assessments: a rationale for northeastern US wetlands. US Fish and Wildlife Service, National Wetlands Inventory Program, Region 5. TWDB. 2017. Glossary of Terms. Texas Water Development Board. Available at: https://waterdatafortexas.org/reservoirs/glossary. United Nations WWAP. 2015. The United Nations World Water Development Report 2015: Water for a Sustainable World. UNESCO: Paris. USACE. 2016a. United States Army Corps of Engineers. Kansas City District. Available at: http://www.nwk.usace.army.mil. USACE. 2016b. United States Army Corps of Engineers. Tulsa District. Available at: http://www.swt.usace.army.mil. USACE. 2015. Cheney Lake Current Condition. Available at: http://www.swt- wc.usace.army.mil/CHEN.lakepage.html. Accessed May, 15 2015. USACE. 2013. Water Supply Storage Reallocation John Redmond Dam and Reservoir, Kansas. U.S. Army Corps of Engineers: Tulsa, OK. USACE. 1984. Wave and Water Level Predictions, Ch. 3, In Shore Protection Manual, vol.1. Department of the Army, Waterways Experiment Station, Vicksburg, MS. USBR. 2016. United States Bureau of Reclamation. Available at: https://www.usbr.gov. USEPA. 2015. Wetlands. EPA. Available at: http://water.epa.gov/type/wetlands/. Accessed May 8 2015. USEPA. 2011a. 2012 National Lakes Assessment. Field Operations Manual. EPA 841-B-11-003, U.S. Environmental Protection Agency: Washington, DC. USEPA. 2011b. National Wetland Condition Assessment: Field Operations Manual. EPA-843- R-10-001, U.S. Environmental Protection Agency: Washington, DC. USEPA RTAG7. 2011. Nutrient Reference Condition Identification and Ambient Water Quality Benchmark Development Process, Freshwater Lakes and Reservoirs within USEPA Region 7. Kansas Biological Survey, Lawrence, KS 42pp. USEPA RTAG7. 2016. Nutrient Reference Condition Identification and Ambient Water Quality Benchmark Development Process, Rivers and Streams within USEPA Region 7. US

51

Environmental Protection Agency Regional Technical Advisory Group, Region 7, Kansas City, KS. 64pp. USGS. 2016. United States Geological Survey. Kansas Office. Available at: https://www.usgs.gov. Vymazal, J. 2010. Water and nutrient management in natural and constructed wetlands. Springer Verlag: London, UK. 330pp. Wang, L., Q. Cai, M. Zhang, L. Tan and L. Kong. 2011. Longitudinal patterns of phytoplankton distribution in a tributary bay under reservoir operation. Quaternary International 224(2):280-288. Wang, S. H., Dzialowski, A. R., Meyer, J. O., Lim, N. C., Spotts, W. W., & Huggins, D. G. 2005. Relationships between cyanobacterial production and the physical and chemical properties of a Midwestern Reservoir, USA. Hydrobiologia, 541(1):29-43. Wilks, D.S. (2011) Statistical Methods in the Atmospheric Sciences, Academic Press, Amsterdam, The Netherlands. Yuan, L.L., A.I. Pollard, S. Pather, J.L. Oliver, and L. D'Anglada. 2014. Managing microcystin: identifying national‐scale thresholds for total nitrogen and chlorophyll a. Freshwater Biology 59(9):1970-1981.

52

Appendix A. Scatter plots for select variables verses water TN and TP, sediment TN and TP, and sediment bulk density. For all, black circles = riverine sites, gray circles = main basin sites, and all outliers were removed.

53

54

55

56

57

58

59

60

Appendix B. See Loeffler, K.M., V. Rahmani, J.H. Kastens, and D.H. Huggins. 2018. An examination of the potential wetland development landscape around managed reservoirs in the central U.S. Great Plains. J. Appl. Geography. 93:16-24.

61

Appendix C. Vegetation and birds noted at each site.

Table 1. Presence or absence of macrophytes and terrestrial vegetation at each site. Macro sp. rich = species richness of macrophytes. Main basins are shaded gray. Ceratophyllum Perse Phragmites Potamo Schoeno Macro code demersum Lemna caria australis geton Salix plectus sp rich Terrestrial BIB C. demersum 1 Terrestrial BIM CGM CGMU Terrestrial CGN Salix 1 Terrestrial CHE Terrestrial CHM CHN Terrestrial CLD Terrestrial CLM CLR Terrestrial CLW Terrestrial EDC Terrestrial EDH Lemna Salix 2 Terrestrial EDM EDS Terrestrial EKC Terrestrial EKE Terrestrial EKM EKS Terrestrial FAB Terrestrial FAF Terrestrial FAM HIB Terrestrial HIM HIN Terrestrial HIR Terrestrial JOM Terrestrial JON Terrestrial KAB Terrestrial KAM KAS Terrestrial Perse 1 KIB caria Terrestrial KIM MAC Terrestrial Ceratophyllum Perse Phragmites Potamo Schoeno Macro code demersum Lemna caria australis geton Salix plectus sp rich Terrestrial MAF Terrestrial MAM Terrestrial MEC Terrestrial MEM MET Salix 1 Terrestrial MIM MIR Terrestrial PED Terrestrial PER Terrestrial PES Terrestrial POC Terrestrial POH Salix 1 Terrestrial POM POV Terrestrial TOE Salix 1 Terrestrial TOM Terrestrial TOV Terrestrial TUB Terrestrial TUM WEM Perse 1 WES caria Terrestrial WIH P. australis 1 Terrestrial WIM Potamo Schoeno 3 WIS P. australis geton plectus Terrestrial

Table 2. Birds observed at each site in 2016, formatted for upload to eBird.org. Big Hill Council Council Council Cheney Cheney Big Hill Cheney Clinton Clinton Clinton Reservoir at Grove Grove Grove Reservoir at Reservoir at Clinton Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Big Hill Reservoir at Reservoir at Reservoir at Eastern North Fork Reservoir at Main Main Main Rock Creek Wakarusa Creek Main Munker Neosho Branch Ninnescah Deer branch branch branch branch branch branch branch branch branch branch branch branch Latitude 37.32478 37.26966 38.68275 38.72536 38.71301 37.78366 37.72896 37.80947 38.95270 38.92776 38.88385 38.90508 Longitude ‐95.44531 ‐95.46751 ‐96.50335 ‐96.50565 ‐96.55065 ‐97.80610 ‐97.78999 ‐97.89202 ‐95.42908 ‐95.33617 ‐95.40837 ‐95.43343 Date 9/26/2016 9/26/2016 10/11/2016 10/11/2016 10/11/2016 9/28/2016 9/28/2016 9/28/2016 9/8/2016 9/8/2016 9/21/2016 9/8/2016 Start Time 11:48 13:45 10:12 15:11 12:45 14:13 15:48 12:12 13:00 10:09 13:01 14:56 StateKSKSKSKSKS KS KS KS KS KS KS KS Country US US US US US US US US US US US US Protocol stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary Num Observers222222222222 Duration (min)6166734855 41 57 60 65 109 59 53 All Obs Reported (Y/ YYYYYYYYYYYY Dist Traveled (Miles) 000000000000 Area Covered (Acres 4 14 20 5 7 4 10 15 15 86 10 10 Notes BIB BIM CGM CGMU CGN CHE CHM CHN CLD CLM CLR CLW great blue heron x xx x Laridae sp. x 2 x x xxxxxx osprey x x|or bald eagle American Coot 2 Double‐crested Cormorant x xx x x American White Pelican x xxxx great egret x xx x turkey vulture xx bald eagle x Red‐tailed Hawk Red‐headed Woodpecker canada goose Table 2 cont'd. Birds observed at each site in 2016, formatted for upload to eBird.org. Toronto Hillsdale Hillsdale Marion Marion Marion Perry Perry Perry Toronto Toronto Hillsdale Hillsdale Perry Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Eastern Main Rock Creek Cottonwood French Main Deleware Main Slough Main Verdigris Bull branch Niles branch Rock branch Branch branch branch branch branch branch branch branch branch branch branch branch 38.70461 38.65503 38.71495 38.66208 38.45398 38.38753 38.36422 39.23943 39.11838 39.16304 39.17391 37.77556 37.74391 37.77102 ‐94.97774 ‐94.90643 ‐94.87727 ‐94.94479 ‐97.17344 ‐97.14689 ‐97.09267 ‐95.45444 ‐95.41827 ‐95.51311 ‐95.40315 ‐95.90786 ‐95.92658 ‐95.94749 9/22/2016 9/20/2016 9/20/2016 9/22/2016 10/5/2016 10/5/2016 10/5/2016 11/2/2016 11/2/2016 11/2/2016 11/2/2016 10/26/2016 10/26/2016 10/26/2016 12:22 12:07 14:06 14:38 15:15 13:20 11:15 14:18 8:57 12:18 10:19 13:05 10:49 14:45 KS KS KS KS KS KS KS KS KS KS KS KS KS KS US US US US US US US US US US US US US US stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary 22222222222222 77 88 53 53 44 47 69 39 66 41 45 55 51 47 YYYYYYYYYYYYYY 00000000000000 12353224105258732018 HIB IM|very wind HIN HIR MAC MAF MAM PED PEM PER PES TOE|no birds TOM TOV x xxxx xx xxx x

xx xxxxxxx xx xxxx xx

x x x x x Table 3. Birds observed at each site in 2017, formatted for upload to eBird.org. John John Kanopolis El Dorado El Dorado El Dorado El Dorado Elk City Elk City Elk City Elk City Fall River Fall River Fall River Kanopolis Kanopolis Redmond Redmond Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Smoky Hill Cole Creek Harrison Main Satchel Chetopa Elk River Main Squaw Badger Fall River Main Bluff Creek Main Main Neosho River branch branch branch branch branch branch branch branch branch branch branch branch branch branch branch branch Latitude 37.93459 37.84451 37.84603 37.88175 37.23180 37.25693 37.27488 37.23948 37.67352 37.66762 37.65565 38.23829 38.26892 38.61481 38.61439 38.66767 Longitude ‐96.77896 ‐96.74156 ‐96.82028 ‐96.75111 ‐95.80944 ‐95.82491 ‐95.78697 ‐95.78143 ‐96.06579 ‐96.10474 ‐96.05940 ‐95.76463 ‐95.84253 ‐98.01462 ‐97.96980 ‐98.00635 Date 6/27/2017 6/27/2017 6/27/2017 6/27/2017 8/8/2017 8/8/2017 8/8/2017 8/8/2017 6/21/2017 6/21/2017 6/21/2017 7/6/2017 7/6/2017 8/24/2017 8/24/2017 8/24/2017 Start Time 16:25 13:24 11:35 14:53 13:51 12:44 11:41 14:56 15:24 14:09 12:02 12:17 10:04 12:20 11:10 9:11 State KS KS KS KS KS KS KS KS KS KS KS KS KS KS KS KS Country US US US US US US US US US US US US US US US US Protocol stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary Num Observers 234567891011121314151617 Duration (min) 37 40 54 36 39 37 49 33 41 46 59 39 63 34 55 52 All Obs Reported (Y/N) YYYYYYYYYYYYYYYY Dist Traveled (Miles) 0000000000000000 Area Covered (Acres) 2 7 18 2 1 5 31 2 2 2 19 20 2 3 20 3 Notes EDC EDH EDM EDS EKC EKE EKM EKS FAB FAF FAM JOM JON KAB KAM KAS great blue heron 2 1 1 1 1 3 4 Larinae sp. x22 osprey American Coot Double‐crested Cormorant 1 131 American White Pelican 40 great egret 3 1 5 2 turkey vulture bald eagle 1 Red‐tailed Hawk Red‐headed Woodpecker canada goose yellow‐headed blackbird Anatinae sp. 22 Sterninae sp. 8 Killdeer 2 Table 3 cont'd. Birds observed at each site in 2017, formatted for upload to eBird.org. Melvern Pomona Pomona Webster Kirwin Kirwin Melvern Melvern Milford Milford Pomona Pomona Tuttle Creek Tuttle Creek Webster Wilson Wilson Wilson Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Reservoir at Marais Des Coyote Hundred N. Fork Bow Creek Main Main Turkey Main Republican Main Valley Brook Blue River Main Main Hell Creek Main Saline Cygne Creek and Ten Solomon branch branch branch branch branch branch branch branch branch branch branch branch branch branch branch branch Mile branch branch 39.63465 39.66403 38.49790 38.51356 38.50597 39.08350 39.23116 38.67465 38.69424 38.65094 38.68988 39.41380 39.26074 39.40664 39.38506 38.90974 38.96394 38.94506 ‐99.16846 ‐99.13140 ‐95.87987 ‐95.71541 ‐95.81421 ‐96.89975 ‐97.00664 ‐95.65083 ‐95.61823 ‐95.56071 ‐95.57722 ‐96.72483 ‐96.59518 ‐99.42593 ‐99.50788 ‐98.47877 ‐98.49725 ‐98.65421 8/23/2017 8/23/2017 6/19/2017 6/19/2017 6/19/2017 7/25/2017 7/25/2017 7/28/2017 7/28/2017 7/28/2017 7/28/2017 7/18/2017 7/18/2017 8/2/2017 8/2/2017 8/1/2017 8/1/2017 8/1/2017 14:37 15:51 15:51 11:36 13:50 12:44 10:50 15:18 13:58 9:43 11:33 14:18 10:55 9:19 11:10 15:07 16:41 12:43 KS KS KS KS KS KS KS KS KS KS KS KS KS KS KS KS KS KS US US US US US US US US US US US US US US US US US US stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary stationary 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 51 52 54 48 54 73 44 46 41 59 45 49 74 60 39 48 86 64 YYYYYYYYYYYYYYYYYY 000000000000000000 2389122352 2 11923138202 1123 KIB KIM MEC MEM MET MIM MIR POC POH POM POV TUB TUM WEM WES WIH WIM WIS 312 222 x2 x 3

1 xx x 5 x 1 4 11

45 2 4