an integrated modeling approach to total watershed management

AN INTEGRATED MODELING APPROACH TO TOTAL WATERSHED MANAGEMENT: WATER QUALITY AND WATERSHED ASSESSMENT OF CHENEY RESERVOIR, KANSAS, USA STEVEN H. WANG1,∗ , DONALD G. HUGGINS2 , LYLE FREES3 , CHAD G. VOLKMAN3 , NIANG C. LIM4 , DEBRA S. BAKER5 , VAL SMITH6 and FRANK DENOYELLES, JR.4 1

Aquatic Ecotoxicology Lab, Kansas Biological Survey, University of Kansas, Lawrence, KS 66047; Aquatic Ecology, Kansas Biological Survey, University of Kansas, Lawrence, KS 66047; 3 Natural Resources Conservation Service, Department of Agriculture, Salina, KS 67401; 4 Kansas Biological Survey, University of Kansas, Lawrence, KS 66047; 5 Central Plains Center for BioAssessment, Kansas Biological Survey, University of Kansas, Lawrence, KS 66047; 6 Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence, KS 66045 (∗ author for correspondence, e-mail: [email protected], Tel: 785 864 1500) 2

(Received 14 November 2003; accepted 8 January 2005)

Abstract. Degradation of water quality is the major health concern for lakes and reservoirs in the central regions of the United States as a result of heavily devoted agricultural production. A vital key to the development of a reservoir management strategy is to identify nutrient loading that describes associated water quality conditions in reservoirs. This study integrated AnnAGNPS watershed and BATHTUB lake models to simulate actual lake water quality conditions of Cheney Reservoir, KS, and demonstrated the use of the coupled model for simulating lake response to changes in different watershed land use and management scenarios. The calibrated current-conditions model simulated in-lake reductions as much as 52% for TN, 48% for TP, and 70% for chlorophyll a due to conversion to native grass, and increases as much as 4% for TN, 9% for TP and 6% for chlorophyll a due to conversion of land from the Conservation Reserve Program (CRP) to cropland (15.5% of watershed). This model also demonstrated an increase in chlorophyll a (19%) as the lake sediment capacity was reached over the next century. Keywords: eutrophication, lake modeling, sedimentation, watershed management, watershed modeling

1. Introduction Modern American agriculture depends on fertilizers to maintain high production. Although the benefits of this chemically based system are obvious, the cost is greater than the benefits when the deterioration of the ecological integrity of aquatic ecosystems is significant and associated losses are taken into consideration (Duttweiler and Nicholson, 1983; Dahl and Johnson, 1991; CAST, 1992). In a recent 2000 water quality report to the United States Congress, agriculture was considered the leading source of impairment in the nation’s surface water (U.S. Environmental Protection Agency, 2002). Nutrients (e.g., nitrogen and phosphorus) were the most widespread pollutants identified as degrading to the quality of lakes and reservoirs Water, Air, and Soil Pollution (2005) 164: 1–19

C

Springer 2005

2

S. H. WANG ET AL.

(Carpenter et al., 1998; Downing et al., 1999; Saadoun, 2001; Smith et al., 2002; Smith, 2003). Cheney Reservoir, a federal multipurpose reservoir located in the agricultural heartland of the United States, serves as the major source of drinking water for the City of Wichita, Kansas. Due to increasing agricultural activities in its watershed, Cheney Reservoir has experienced degraded water quality problems (Christensen and Pope, 1997; Pope, 1998; Pope and Milligan, 2000; Milligan and Pope, 2001; Mau, 2001; Smith et al., 2001). The most common water quality problems are elevated nutrient levels, with concurrent elevations in algal biomass, and suspended solids and siltation that reduce light penetration, aesthetics, and lake depth and volume, leading to alteration of aquatic habitats. Currently, Cheney Reservoir is on the impaired waterbody list of the Clean Water Act as a result of eutrophication and siltation (KDHE, 2002). Reservoirs are relatively new and complex aquatic ecosystems that are distinguished from natural lakes in that reservoirs are constructed by the human impoundment of lotic ecosystems. A number of significant differences exist between reservoirs and natural lakes. First, construction of a reservoir greatly disturbs the abiotic and biotic environment. Second, reservoirs tend to have relatively short life spans. Coupled, these two factors create a highly unstable aquatic environment. Also, watershed conditions and stressors greatly influence the artificial ecosystems associated with reservoirs. This, in turn, directly affects reservoir health and stability (e.g., nutrient enrichment and sedimentation). Thus, it is reasonable to expect that the successful management of reservoirs, in regards to nutrient enrichment, must be based on a complete understanding of the inherent complexity and interactions of aquatic and terrestrial systems (Randtke and deNoyelles, 1985). A vital key to the development of a total reservoir management strategy is to model watershed nutrient loading and to link watershed-model estimates to models of lake water quality and health (Mankin et al., 2003). However, few published models directly identify nutrient loading that describes associated eutrophic conditions in reservoirs. The main objectives of this study were to use the integrated AnnAGNPS (Annualized Agricultural Nonpoint Source) watershed and BATHTUB (eutrophication) lake models to (1) provide watershed/lake managers and/or regulators estimates of background nutrient loadings and historic reservoir conditions by modeling of predevelopment and/or early land use development, and (2) investigate changes in the current eutrophication processes to the implementation of several watershed management practices. 2. Methods 2.1. S ITE

DESCRIPTION

Cheney Reservoir is located in the south-central portion of Kansas near the City of Wichita. Although the reservoir is fed by a number of streams, it is mainly an

AN INTEGRATED MODELING APPROACH TO TOTAL WATERSHED MANAGEMENT

3

impoundment on the North Fork Ninnescah River. Silver Creek, Goose Creek, and Red Rock Creek are the three major streams that join the North Fork Ninnescah River to feed the reservoir. Cheney Reservoir (average depth 4.87 m) is a multipleuse reservoir (e.g. fishing, boating, swimming, and sightseeing) and has a normal surface area of 3,885 ha. The Cheney Reservoir watershed encompasses nearly 2,420 square kilometers (241,600 ha) of land located in Sedgwick, Reno, Kingman, Pratt, Stafford, and Kiowa Counties. Most of the watershed is underlain by consolidated rocks of Permian age covered by unconsolidated fluvial and windblown deposits of Pleistocene age (Christensen and Pope, 1997). Fewer than 4,000 people live on approximately 1,000 farms in the watershed. Four wastewater discharge point sources are located for the communities of Arlington, Staffors, Turon, and Preston, which together produce annual N and P loads of 3,350 and 1,320 kg, respectively, to the receiving streams and eventually to Cheney Reservoir (Koelliker and Bhuyan, 2000). Land use/land cover in the watershed is predominately agricultural, with 79.4% of the land in rangeland (such as pasture and hay field) and cultivated cropland. Corn (Zea mays [L.]), sorghum (Sorghum bicolor [L.]) and soybean (Glycine max [L.]) are the major crops planted in the watershed. Woodland occupies approximately 2.4% of the total area of the watershed. About 2% of the watershed is in residential or commercial uses.

2.2. D ATA

COLLECTION

AnnAGNPS requires more than 400 parameters in numerous categories. For example, the topographical parameters such as slope, slope length, slope-shape factor, and aspect/flow direction were derived from a digital elevation model (DEM) at a scale of 1:24,000. Soil Survey Geographic Database (SSURGO) was used to populate soil parameters. The 1997 LANDSAT image was used to generate the required land use data for the model whereas crop operation, fertilization and field-management data were obtained from the Cheney Reservoir Project Office and the Kansas Natural Resources Conservation Service (NRCS) field offices. Geomorphic data was parameterized from a comprehensive study of geomorphic assessment and classification of Kansas riparian system (Emmert et al., 2001). Monthly limnological data was obtained from the Cheney Reservoir study of Smith et al. (2001). Lake precipitation, evaporation, elevation, and total inflow data were attained from the Tulsa District of the Army Corps of Engineers. Streamflow information and associated water quality data were gathered from the USGS while atmospheric N and P inputs were complied from National Atmospheric Deposition Program/National Trend Network, a 1994 USGS report (Puckett, 1994), and a recent ARS study (Burkart and James, 2002).

4 2.3. MODEL

S. H. WANG ET AL.

SELECTION

The criteria used in the watershed and lake model selection process were 1) models had to be readily available to predict sediment and nutrient parameters for agricultural regions, 2) models needed to be implemented within a reasonable timeframe, and 3) the modeling efforts and tasks should not exceed the proposed budget. Of numerous watershed and reservoir models, AnnAGNPS and BATHTUB met these criteria and were selected. 2.3.1. AnnAGNPS Annualized Agricultural Nonpoint Source (AnnAGNPS) is a batch-process and continuous-simulation watershed model (Bosh et al., 1998). The model does distributed modeling that a target watershed is subdivided into homogenous cells (hydrologic units) to quantitatively estimate nutrient and sediment loading. The earlier versions of this model (e.g., AGNPS), which are single-event models, have been broadly and successfully used in the central regions of the United States (e.g., Koelliker and Bhuyan, 2000; Mankin and Kalita, 2000; Mankin and Koelliker, 2001). AnnAGNPS expand the original modeling capabilities of AGNPS by incorporating the Revised Universal Soil Loss Equation (RUSLE) and the Hydrogeomorphic Universal Soil Loss Equation (HUSLE) to predict soil and sediment loss from the field (Theurer and Clarke, 1991; Renard et al., 1997). 2.3.2. BATHTUB BATHTUB, an empirical model designed to assess eutrophication for morphometrically complex reservoirs (Walker, 1996), is an effective tool for water quality assessment and management (Ernst et al., 1994). BATHTUB is composed of three major components that include water balance, nutrient sedimentation, and eutrophication response models (expressed in terms of total N, total P, chlorophyll a, transparency, organic N, and organic P). One major advantage of BATHTUB over other models is its use of simple steady-state calculations to address eutrophication processes, which reduces data demands. In addition, the windy, relatively flat agricultural landscape of the Central Plains creates well mixed and turbid lentic waterbodies (O’Brien, 1975; Randtke and deNoyelles, 1985), thus rendering the comparatively simple BATHTUB model as more appropriate to use than the more complex two or three dimensional models such as CE–QUAL–W2 (Cole and Buchak, 1995). 2.4. MODEL

CALIBRATION AND VALIDATION

2.4.1. AnnAGNPS Two years (January 1997–December 1998) of streamflow and water quality data collected in the Red Rock Creek watershed were used to calibrate and validate AnnAGNPS for the Cheney Reservoir watershed (Figure 1). The daily volume of runoff from a storm was estimated using the USGS Hydrograph Separation


5

Figure 1. Location of Cheney Reservoir and its watershed, weather station locations (), lake sampling sites (•), and BATHTUB lake segments.

Program, HYSEP (Sloto and Crouse, 1996). To determine sediment loading under typical runoff conditions, the separated runoff and average monthly total suspended solids (TSS) concentrations were evaluated graphically. Based on the intensity and duration of runoff events in 1997 and 1998, the TSS concentration of 90 mg L−1 was selected as the upper limit for baseflow conditions and the lower limit for runoff conditions (Figure 2). The initial procedure for the calibration of AnnAGNPS was to estimate the model runoff using individual and average precipitation data from two weather stations located in the watershed. However, due to the presence of localized storms (rainfall variability) at these weather stations, these simulation results exhibited

6

S. H. WANG ET AL.

Figure 2. Separated runoff hydrograph and TSS for storm flow events for the Red Rock Creek during 1997–1998.

monthly values that were significantly different from the observed values. Therefore, adjustments were made to the average rainfall values in an attempt to better fit the volume estimates of individual runoff events based on the USGS stream gauging data. Additionally, slightly different runoff curve numbers (+3CN) of hydrologic soil groups A and B for row crops of no conservation practices were assigned to AnnAGNPS at antecedent moisture content II (U.S. Department of Agriculture, 1972). Figure 3 shows the validation results of the calibrated AnnAGNPS. The average prediction error of the calibrated model was 6% for water yield, 45% sediment yield, 48% for TN and 20% for TP. 2.4.2. BATHTUB In order to ensure that BATHTUB was locally applicable, the water, nutrient and eutrophication components needed to be adjusted to characterize the current reservoir condition. While runoff and nutrient loads to Cheney Reservoir were estimated by AnnAGNPS, the baseflow volume and its nutrient loads were computed using the HYSEP and water quality data. Streamflow and nutrient loads for the ungauged area were proportionally estimated according to its size. In this study, the second-order decay rate was selected to represent the in-lake nutrient responses as suggested by Walker (1996) while chlorophyll a (characterized by N-limited) and Secchi depth (in relation with chlorophyll a and turbidity) functions used in the eutrophication model was determined using the algal bioassay experiments and in-lake water quality data.


7

Figure 3. Cumulative plots of water, sediment, TN, and TP yields in model (AnnAGNPS) validation.

The elemental ratios of TN to TP (TN:TP) have been commonly used to infer nutrient limitation in terms of which of these nutrients is most likely limiting algal growth in the water. This is based on the relative requirement for each nutrient by different types of algae. The TN:TP ratios for algae tend to be 10:1 by weight. Higher ratios, particularly above 17:1, infer phosphorus limitation for algae and lower ratios, particularly below 5:1 infer nitrogen limitation and favor the nitrogen-fixing cyanobacteria (Smith 1982, 1998; Smith and Bennett, 1999). The average ratio of TN and TP for Cheney Reservoir was 6:1, indicating a strong potential for N-limitation of algal growth (Smith et al., 2001). For a validation purpose, laboratory algal bioassays were conducted at the Kansas Biological Survey in July 2002 to corroborate the above conclusion. Raw surface water with naturally occurring algae was collected from the main basin of Cheney Reservoir and placed in bottles spiked with various combinations of potassium nitrate (KNO3 ) and potassium phosphate (KH2 PO4 ) to achieve the concentrations of 800 µg N L−1 and 200 µg P L−1 , respectively. Additional sample water was placed in bottles without nutrient spikes and exposed to different levels of light to determine light limitation. Conditions causing increased growth in the bottle provided some support for identifying the conditions regulating algal growth in the reservoir. As indicated in Figure 4, NO3 was

8

S. H. WANG ET AL.

Figure 4. Mean fluorescence values for light and nutrient addition treatments for lake water collected from main basin of Cheney reservoir on July 2, 2002. Fluorescence was measured daily during an eight-day period. HL and LL indicate light treatment at 620 and 400 microeinsteins, respectively. An error bar represents one standard deviation.

required to support growth, suggesting that the availability of N was necessary for any acceleration of surface algal growth in the main basin. BATHTUB was calibrated basin-wide because the Cheney Reservoir is well mixed due to of its size, orientation, and an open landscape (O’Brien, 1975; Randtke and deNoyelles, 1985). Residual TN and/or TP values were used to examine if


9

internal nutrient cycling existed; predicted nutrient values less than measured nutrient values reflected the potential internal nutrient cycling effect as a result of anoxic conditions and sediment resuspension. The initial calibration results indicated that a discrepancy existed between the predicted and measured data for the Mud Creek Cove. This is likely due to the fact that a large amount of water flowed into the Mud Creek Cove from the main basin of the reservoir, causing turbulence to re-suspend the bottom sediment, which resulted in a release of nutrients. The internal load was therefore added to the section in BATHTUB. Figure 5 shows the calibration results of BATHTUB for total and organic P, total and organic N, chlorophyll a, and Secchi depth. The model prediction corresponded

Figure 5. Limnological parameters in model calibration for water year 2000: observed values (•) from Cheney Reservoir, Kansas; predicted values (◦) generated from BATHTUB. Mean indicates area-weighted mean for the whole reservoir (error bars are standard errors).

10

S. H. WANG ET AL.

well with the measured values for each segmented region of Cheney Reservoir. No significant differences were noticed at the 95% confident level between the predicted and observed values. 2.5. MODEL

SCENARIOS DESCRIPTIONS

Three watershed management scenarios were evaluated to assess their hypothetical impacts on water quality of Cheney Reservoir. Scenario 1 converted the entire watershed to native-grass prairie to approximate the predevelopment and/or early land use development. All land use was changed to grassland, no fertilizer was added, and feedlots and point sources were removed to estimate background nutrient loadings and historic reservoir conditions. Scenario 2 simulated the effect of changing continuous and conventional (mulch-till) wheat and/or milo (grain sorghum) cultivation to no-till wheat and milo rotation. Scenario 3 simulated the effect of converting the CRP filter strip corridors back to conventional tillage. The Conservation Reserve Program is an U.S. Department of Agriculture’s environmental program that provides farmers with technical and financial assistance to ease the highly erodible lands from agricultural production to protect and/or improve water quality. The Cheney Watershed Program encourages CRP contract holders to leave a 60 m grass filter strip along the stream banks as designated by the U.S. Geological Survey (Conservation Technology Information Center, 2002).

3. Results The Generation of weather Elements for Multiple applications (GEM) weather generator model was used to estimate the rainfall distribution code and energy intensity using each watershed county’s monthly precipitation. The seed value in GEM was adjusted until it reached a reasonable fit to the observation values. The GEM is a stochastic model that generates a time series of daily weather elements (e.g., precipitation, maximum and minimum temperature, and dewpoint) for a given location based on weather stations positioned in the region (Johnson et al., 2000). The GEM data used in this study was generated through an evaluation of the historical weather records of the 11 weather stations surrounding the watershed (Koelliker and Bhuyan, 2000). In this study, 30-year simulation data was used to represent a typical historic and future weather pattern for scenarios analysis (Figure 6). The AnnAGNPS results of the current condition generated by the GEM data are summarized in Table I, with a comparison to the averaged runoff data of 1997 and 1998 at the USGS gauging station (07144780) located on the North Fork Ninnescah River above Cheney Reservoir. BATHTUB estimated Cheney Reservior’s average depth and hydraulic residence time to be 4.90 m and 2.56 yr, respectively. Based on these estimates, there were 344,155 kg of TN and 63,789 kg of TP entering, and 68,768 kg of TN and 11,092 kg


11

TABLE I Comparisons of water, sediment, TN, and TP yield estimated by AnnAGNPS (30-yr GEM climate data) and the USGS measured data using averaged values of 1997–1998 at the USGS gauging station (07144780) Parameter

AnnAGNPS

USGS (1997–1998)

Water (million m3 ) Sediment (Mg) TN (Mg) TP (Mg)

61.96 22884.52 158.58 30.80

56.19 23396.72 112.58 40.62

Figure 6. A histogram of 30-year GEM data. A density trace is constructed by computing the data density for each simulated year. The average rainfall of the GEM data is 752 mm.

of TP existing the reservoir annually, assuming nutrient values exported to be the same as the main basin nutrient concentrations. Approximately, 80% of TN (288,604 kg yr−1 ) and 83% of TP (54,829 kg yr−1 ) were retained during the water year 2000 (Table III). The predicted area-weighted average concentrations of TN, TP, chlorophyll a, Secchi, organic N, and organic P were 0.64 mg L−1 , 105 µg L−1 , 16 µg L−1 , 0.68 m, 0.46 mg L−1 , and 66 µg L−1 , respectively. Using an algal (chlorophyll a) nuisance threshold of 10 µg L−1 , Cheney Reservoir’s water quality was impacted by nuisance levels of algae for 69% of days in the growing season. 3.1. SCENARIO 1 As indicated in Table II, the presumed native conditions increased the ability of the watershed to retain water, as reflected in a 31% reduction in annual runoff. The

12

S. H. WANG ET AL.

TABLE II Summary of AnnAGNPS results for watershed management scenarios Description:

Current

Native

Wheat-milo rotation

CRP to cropland

Management scenario: Land use conversion, % Average surface inflow, mm yr−1 Sediment in surface inflow kg ha−1 yr−1 Sediment N in surface inflow kg ha−1 yr−1 Dissolved N in surface inflow kg ha−1 yr−1 Sediment P in surface inflow kg ha−1 yr−1 Dissolved P in surface inflow kg ha−1 yr−1

– 26.92 113.42 0.34 0.18 0.09 0.10

1 76.40 18.54 3.62 0.00a 0.04 0.01 0.04

2 – 25.91 84.43 0.25 0.17 0.07 0.08

3 15.50 25.40 140.95 0.43 0.17 0.11 0.11

a

Indicates 0.004. TABLE III BATHTUB estimated nutrient loads for Cheney Reservoir during the water year 2000 TN Source Input Atmospheric deposition Internal load Runoff at Station 07144780 Baseflow at Station 07144780 Red Rock Creek Ungauged (near-lake) area Total Output Release from Dam Water Withdrawal from Wichita Total

TP

kg yr−1

%

kg yr−1

%

34021.3 7122.4 120293.0 99763.5 24074.6 58880.1 344154.9

10 2 35 29 7 17 100

1749.2 3276.3 34993.4 4961.9 5458.3 13349.5 63788.6

3 5 55 7 8 21 100

37559.2 31208.3 68767.5

55 45 100

6058.0 5033.7 11091.7

55 45 100

native conditions also reduced watershed sediment load 97% to 755 Mg per year (3.62 kg ha−1 yr−1 ), sediment N load 99% to 1.02 Mg per year (0.004 kg ha−1 yr−1 ), dissolved N load 78% to 8.20 Mg per year (0.04 kg ha−1 yr−1 ), sediment P load 90% to 1.90 Mg per year (0.01 kg ha−1 yr−1 ), and dissolved P load 54% to 9.65 Mg per year (0.04 kg ha−1 yr−1 ). Converting the entire watershed to the native-grass prairies resulted in substantial reductions in chlorophyll a (70%), total N (52%), and total P (48%), and a great improvement in Secchi depth (22%). The area-weighted trophic conditions for the Cheney Reservoir were 0.31 mg L−1 , 54 µg L−1 , 5 µg L−1 , and 0.83 m for TN, TP,


13

chlorophyll a, and Secchi depth, respectively. With this watershed management, the frequency of algal nuisance occurrence showed a reduction from 69% to 7% of the growing season. 3.2. S CENARIO 2 With no-till wheat-milo rotation, there was a 3% decrease in runoff accompanied by a 25% reduction in sediment load to 17,617 Mg per year (84.43 kg ha−1 yr−1 ), 28% reduction in sediment N load to 51.00 Mg per year (0.25 kg ha−1 yr−1 ), 2% reduction in dissolved N load to 35.99 Mg per year (0.17 kg ha−1 yr−1 ), 27% reduction in sediment P load to 14.30 Mg per year (0.07 kg ha−1 yr−1 ), and a 23% decrease in dissolved P load to 16.11 Mg per year (0.08 kg ha−1 yr−1 ). BATHTUB indicated that under the Scenario 2 management, there was a 7% decrease in chlorophyll aaccompanied by a 5% reduction in TN concentrations, and a 13% reduction in TP concentrations. In addition, Secchi disc reading increased from 0.68 m to 0.69 m and the frequency of algal nuisance occurrence decreased from 69% to 64%. This implies that crop rotation can be a good nutrient management practice that could provide Cheney Reservoir with substantial sediment or nutrient yield reductions. It is likely that crop rotations improve soil structure by incorporating crop residue into soil after harvest, which can increase soil fertility and water infiltration rate (Iowa State University, 2002). 3.3. SCENARIO 3 Converting all CRP filter strips to cropland (Scenario 3) though resulted in a 5% reduction in runoff, produced a 25% increase in sediment, 25% increase in sediment N, and 24% increase in sediment P loads to 29,409 Mg per year (140.94 kg ha−1 yr−1 ), 88.99 Mg per year (0.43 kg ha−1 yr−1 ), and 23.84 Mg per year (0.11 kg ha−1 yr−1 ), respectively. The lake model revealed that converting the 91,778 acres of CRP to cropland resulted in a 6% increase in chlorophyll a concentration from 16 to 17 µg L−1 along with a 4% increase in TN concentrations from 0.64 to 0.67 mg L−1 , 9% increase in TP concentrations from 105 to 114 µg L−1 , and a 1% decrease in Secchi disc reading from 0.68 to 0.67 m. This demonstrates the positive impacts that CRP has on nutrient load reductions in the receiving waters and lends support to the furthering of this program. 4. Discussion 4.1. TASTE

AND ODORS

Cheney Reservoir is a eutrophic and N-limited lake. Frequently a massive build up of cyanobacterial biomass (often referred to as blue-green algae) has occurred,

14

S. H. WANG ET AL.

resulting in offensive tastes and odors of drinking water. To control taste and odor problems, Smith et al. (2001) suggested that total P and chlorophyll a need to be maintained at concentrations below 110 µg L−1 and 11 µg L−1 , respectively, in all parts of the reservoir. Though the wheat-milo rotation could improve water quality, this management alone may not be sufficient to reduce taste and odor problems resulting from the presence of elevated chlorophyll a concentrations. Concentrations of chlorophyll a in Scenario 2 ranged from 24 µg L−1 in the riverine to 11 µg L−1 in the main basin. The results of BATHTUB suggest that to reduce chlorophyll a concentrations below the target level for all parts of Cheney Reservoir, watershed TN and TP loads need to be reduced by approximately 70% from the current loading (Figure 7). However, this practice may not be the best possible approach because it maintains the low nutrient condition (TN:TP = 6.3) favorable for cyanobacterial growth (Levich, 1996; Havens et al., 2003). Reducing the TP load alone shows only a slight reduction in chlorophyll a concentrations. However, this P management strategy may create the lake conditions that would not be conducive for cyanobacteria. For example, if P load is reduced by 70%, TN:TP ratio increases to 10.7 (from 6.1) at the current condition, which is not as conducive to cyanobacterial growth and the associated taste and odor problems.

Figure 7. Effects of scaled reduction in inflow nutrient loading on chlorophyll a concentrations in Cheney Reservoir for current, scenario, and targeted conditions.


15

4.2. SEDIMENTATION Sediment deposition or siltation in reservoirs is a major concern for aquatic resource managers because it not only reduces the availability of water storage space over time (Thornton, 1990), but it also contributes to decreased water quality as a result of increased nutrients and turbidity, decreased lake depth, and other morphological changes that accelerate lake aging and eutrophication (Presing, 1996; Hyenstrand et al., 1998; deNoyelles et al., 1999). Cheney Reservoir, which was constructed between 1962 and 1965 by the Bureau of Reclamation, has a designed 100-year sediment storage capacity of 18.86 hm3 . The recent sediment study conducted by the USGS indicated that the sediment had filled 27% of the capacity, which is less than the originally designed rate (Mau, 2001). At the current rate, the 100-year sediment storage capacity would be filled by 2090. The effect of sedimentation on water quality of Cheney Reservoir was modeled by reducing reservoir depth. As indicated in Figure 8, the area-weighted average chlorophyll a concentration increased as the reservoir became shallower. A 19% increase in the chlorophyll a concentration from the current condition was noted as the sediment capacity was filled up in year 2090. As expected, other accompanying conditions were elevated TN (6%), TP (6%), organic N (12%) and P (11%), and algal nuisance (13%) as well as reduced Secchi disc reading (4%).

Figure 8. Effect of sedimentation on chlorophyll a and algal nuisance frequency for current conditions.

16

S. H. WANG ET AL.

5. Conclusion Cheney Reservoir currently experiences eutrophication and siltation problems and is named on the state’s impaired waterbody list. Previous studies have shown that nonpoint sources closely associated with increasing agricultural activities in the watershed are the major contributor to the degraded water quality. The calibrated AnnAGNPS shows that at the pre-settlement or historic watershed condition Cheney Reservoir would have better water quality conditions, with chlorophyll a, Secchi disc reading, and algal nuisance at 5 µg L−1 , 0.83 m, and 7%, respectively. Total N and TP concentrations would be approximately 50% lower than current levels. The eutrophication model (BATHTUB) suggests that chlorophyll a levels would decrease in the future as nutrient loads, particularly N, decrease. To slow the eutrophication processes, nutrient loads need to be controlled. The model indicates that at the present deposition rate sedimentation will adversely affect water quality. There is a need for the project partners, watershed stakeholders, and resource management agencies to work together to develop a creative, proactive, and voluntary water quality management strategy that meshes with modeling tools to address watershed, riparian, and stream management (Seda et al., 2000; Gonzalez, 2000; Smith et al., 2002; Mankin et al., 2003). Cheney Reservoir is likely a N-limited lake in nature (TN:TP = 5.7 at the pre-settlement condition) in which cyanobacterial growth could lead to taste and odor problems. However, these problems can be managed through an ecological approach as suggested by Smith et al. (2001): to shift the phytoplankton community from cyanobacteria to other desirable species (e.g., green algae) that are edible to Daphnia and/or other aquatic organisms, P concentrations need to be reduced more than N concentrations to create a lake condition that is not suitable for cyanobacteria. To monitor this biomanagement method, an ecological model needs to be introduced to carefully evaluate natural interactions between nutrients and phytoplankton, zooplankton, and fish communities. The coupled modeling or total watershed approach shown in this study is demonstrated to be useful to provide lake and watershed managers the baseline water quality condition to develop sustainable management goals. This linkage of AnnAGNPS to BATHTUB is particularly helpful for lake total maximum daily loads (TMDL) development because it offers quantitative assessments load reduction and the resultant lake water quality for decision makes to implement a target water quality standard.

Acknowledgement Special thanks are given to Mr. Gerald Blain of the City of Wichita, Mr. Lyle Frees of the Natural Resources Conservation Service (Salina, KS), Dr. Larry Pope of the U.S. Geological Survey (Lawrence, KS), and Dr. Val Smith of the Department


17

of Ecology and Evolutionary Biology, University of Kansas for permission to use watershed, stream, and lake data. Appreciation is extended to Dr. Marios Sophocleous of the Kansas Geological Survey for sharing his expertise in streamflow (baseflow/runoff) separation techniques. Many thanks are given to Scott Campbell, Clint Goodrich, Connie Tra, and Andrew Hwang for their assistance in bioassay experiments. In addition, we appreciate helpful comments from Dr. J. Stamm and anonymous reviewers, which have greatly improved the quality of this manuscript. This study was funded by the U.S. Environmental Protection Agency through award number X-99797001-0 to the Kansas Biological Survey. References Bosch, D. D., Bingner, R. L., Theurer, F. G., Felton, G. and Chaubey, I.: 1998, ‘Evaluation of the AnnAGNPS water quality model’, ASAE Paper No. 98–2195, St Joseph, Michigan, p.12. Burkart, M. R. and James, D. E.: 2002, ‘Geographic distribution of excess agricultural nitrogen in the Gulf of Mexico’, USDA-ARS, National Soil Tilth Laboratory. Ames, Iowa. Retrieved from http://www.nstl.gov/pubs/burkart/nia/hypoxia3.htm. Carpenter, S. R., Caraco, N. F., Correll, D. L., Howarth, R. W., Sharpley, A. N. and Smith, V. H.: 1998, ‘Non-point pollution of surface waters with phosphorus and nitrogen’, Ecological Applications 8, 559–568. Council for Agricultural Science and Technology (CAST): 1992, ‘Water quality: Agriculture’s role’, Task Force Report No. 120, Ames, IA, p. 103. Christensen, V. G. and Pope, L. M.: 2001, ‘Occurrence of dissolved solids, nutrients, atrazine, and fecal coliform bacteria during low flow in the Cheney Reservoir Watershed, south-central Kansas’, U.S. Geological Survey, Water-Resources Investigations Report 97–4153, Lawrence, KS, p. 13. Conservation Technology Information Center: 2002, ‘Community Award Winner: Cheney Watershed’, Retrieved from http://www.ctic.purdue.edu/KYW/newsreleases/wswinnercheney.html. Cole, R. W. and Buchak, E. M.: 1995, ‘CE–QUAL–W2: A two dimensional, laterally averaged, hydrodynamic and water quality model’, Version 2.0, Instruction Report EL–95–1, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. Dahl, T. E. and Johnson, C. E.: 1991, Status and Trends of Wetlands in the Conterminous United States, mid-1970’s to mid-1980’s, U.S. Department of the Interior, Fish and Wildlife Service, Washington, D.C., p. 28. deNoyelles, F., Jr., Wang, S. H., Meyer, J. O., Huggins, D. G., Lennon, J. T., Kolln, W. S. and Randtke, S. J.: 1999, ‘Water quality issues in reservoirs: Some considerations from a study of a large reservoir in Kansas’, in Proceedings of the 49th Annual Conference of Environmental Engineering, Department of Civil and Environmental Engineering and Division of Continuing Education, The University of Kansas, Lawrence, KS, U.S.A., February 1999, pp. 83– 119. Downing, J. A., Watson, S. B. and McCauley, E.: 1999, ‘Predicting cyanobacteria dominance in lakes’, Can. J. Fish. Aquat. Sci. 46, 1905–1908. Duttweiler, D. W. and Nicholson, H. P.: 1983, ‘Environmental problems and issues of agricultural nonpoint source pollution’, in F. W. Schaller and G. W. Bailey (eds.), Agricultural management and water quality, Iowa State University Press, Ames, IA, pp. 3–16. Emmert, B., Hase, K. and Rajala, T.: 2001, Geomorphic Assessment and Classification of Kansas Riparian Systems, Kansas Water Office, p. 166. Ernst, M. R., Frossard, W. and Mancini, J. L.: 1994, ‘Two eutrophication models make the grade’, Water Environ. Technol. November, 15–16.

18

S. H. WANG ET AL.

Gonzalez, E. J.: 2000, ‘Nutrient enrichment and zooplankton effects on the phytoplankton community in microcosms from El Andino reservoir (Venezuela)’, Hydrobiologia 434, 81–96. Havens, K. E., Phlips, E. J., Cichra, M. F. and Li, B. L.: 1998, ‘Light availability as a possible regulator of cyanobacteria species composition in a shallow subtropical lake’, J. Freshwat. Biol. 39, 547–556. Havens, K. E., James, R. T., East, T. L. and Smith, V. H.: 2003, ‘N:P ratios, light limitation, and cyanobacterial dominance in a subtropical lake impacted by non-point source nutrient pollution’, Environ. Pollut. 122, 379–390. Iowa State University: 2002, ‘Nutrient management: Crop rotation’, Continuing Education and Communication Services, Water Resour. Water Quality Publications NMEP5, Iowa State University, Ames, IA, p. 4. Johnson, G. L., Daly, C., Taylor, G. H. and Hanson, C. L.: 2000, ‘Spatial variability and interpolation of stochastic weather simulation model parameters’, J. Appl. Meteorol. 39, 778–796. Kansas Department of Health and Environment (KDHE): 2002, Lower Arkansas River Total Maximum Load: Cheney Lake, Bureau of Water, KDHE, Topeka, KS, p. 7. Koelliker, J. K. and Bhuyan, S. J.: 2000, ‘Evaluation of AGNPS modeling in Cheney Lake NPS management’, Final Report, Department of Biological and Agricultural Engineering, Kansas State University, Manhattan, KS, p. 99. Levich, A. P.: 1996, ‘The role of nitrogen-phosphorus ratio in selecting for dominance of phytoplankton by cyanobacteria or green algae and its application to reservoir management’, J. Aquat. Ecosyst. Health 5, 55–61. Mankin, K. R. and Kalita, P. K.: 2000, ‘Horseshoe Creek watershed water quality assessment’, Final Report. KDHE Contract NPS 97–149, Bureau of Water, KDHE, Topeka, KS. Mankin, K. R. and Koelliker, J. K.: 2001, ‘Clinton Lake water quality assessment project’, Final Report, KDHE Contract NPS 98–059, Bureau of Water, KDHE. Topeka, KS. Mankin, K. R., Wang, S. H., Koelliker, J. K., Huggins, D. G. and deNoyelles, F., Jr.: 2003, ‘Watershed-lake quality modeling: Verification and application’, J. Soil Water Conserv. 58, 188– 197. Mau, D. P.: 2001, ‘Sediment deposition and trends and transport of phosphorus and other chemical constituents, Cheney Reservoir watershed, south-central Kansas’, U.S. Geological Survey, WaterResources Investigations Report 01–4085, Lawrence, KS, p. 40. Milligan, C. R. and Pope, L. M.: 2001, ‘Occurrence of phosphorus, nitrate, and suspended solids in streams of the Cheney Reservoir Watershed, south-central Kansas, 1997–2000’, U.S. Geological Survey, Water-Resources Investigations Report 01–4199, Lawrence, KS, p. 18. O’Brien, W. J.: 1975, ‘Factor limiting primary productivity in turbid Kansas reservoirs’, Kansas Water Resources Research Institute, Project Completion Report 156, University of Kansas, Lawrence, KS. Pope, L. M.: 1998, ‘Watershed trend analysis and water-quality assessment using bottom-sediment cores from Cheney Reservoir, south-central Kansas, 1997–1998’, U.S. Geological Survey, WaterResources Investigations Report 98–4227, Lawrence, KS, p. 24. Pope, L. M. and Milligan, C. R.: 2000, ‘Preliminary assessment of phosphorus transport in the Cheney Reservoir Watershed, south-central Kansas, 1997–1998’, U.S. Geological Survey, WaterResources Investigations Report 00–4023, Lawrence, KS, p. 29. Presing, M., Herodek, S., Voros, L. and Kobor, I.: 1996, ‘Nitrogen fixation, ammonium and nitrate uptake during a bloom of Cylindrospermopsis raciborskii in Lake Balaton’, Archiv f¨ur Hydrobiologie 136, 553–562. Puckett, L. J.: 1994, ‘Nonpoint and point sources of nitrogen in major watersheds of the United States’, U.S. Geological Survey, Water-Resources Investigations Report 94–4001, p. 9. Randtke, S. J. and deNoyelles, F., Jr.: 1985, ‘A critical assessment of the management practices on water quality, water treatment, and sport fishing in multipurpose reservoir in Kansas’, Project Completion


19

Report 252, Kansas Water Resources Research Institute, University of Kansas, Lawrence KS, p. 171. Renard, K. G., Foster, G. R., Weesies, G. A., McCool, D. K. and Yoder, D. C.: 1997, Predicting Soil Erosion by Water: A Guide to Conservation Planning with the Revised Universal Soil Loss Equation (RUSLE), U.S. Department of Agriculture, Agriculture Handbook No 703, pp 404. Saadoun, I. M. K., Schrader, K. K. and Blevins, W. T.: 2001, ‘Environmental and nutritional factors affecting geosmin synthesis by Anabaena sp’, Water Res. 35, 1209–1218. Seda, J., Hejzlar, J. and Kubecka, J.: 2000, ‘Trophic structure of nine Czech reservoirs regularly stocked with piscivorous fish’, Hydrobiologia 429, 141–149. Smith, V. H.: 1982, ‘The nitrogen and phosphorus dependence of algal biomass in lakes: An empirical and theoretical analysis’, Limnol. Oceanogr. 27, 1101–1112. Smith, V. H.: 1998, ‘Cultural eutrophication of inland, estuarine, and coastal waters’, in M. L. Pace and P. M. Groffman (eds.), Limitation and frontiers in ecosystem science, Springer-Verlag, New York, NY, pp. 7–49. Smith, V. H. and Bennett, S. J.: 1999, ‘Nitrogen:phosphorus supply ratios and phytoplankton community structure in lakes’, Archiv fur Hydrobiologie 146, 37–53. Smith, V. H., deNoyelles, F., Jr., Graham, D. W. and Randtke, S. J.: 2001, ‘A comparative water quality study of Cheney Reservoir, Kansas’, Final Report, Department of Ecology and Evolutionary Biology, University of Kansas. Lawrence, KS. Smith, V. H., Sieber-Denlinger, J., deNoyelles, F., Jr., Campbell, S., Pan, S., Randtke, S. J., Blain, G. and Strasser, V. A.: 2002, ‘Managing taste and odor problems in a eutrophic drinking water reservoir’, Lake Reservior Manage. 18, 319–323. Smith, V. H.: 2003, ‘Eutrophication of freshwater and coastal marine ecosystems: A global problem’, Environmental Science and Pollution Research International 10, 126–139. Sloto, R. A. and Crouse, M. Y.: 1996, ‘HYSEP: A computer program for streamflow hydrograph separation and analysis’, U.S. Geological Survey, Water-Resources Investigations Report 96– 4040, p. 46. Theurer, F. G. and Clarke, C. D.: 1991, ‘Wash load component for sediment yield modeling’, in Proceedings of the Fifth Federal Interagency Sedimentation Conference, Las Vegas, Nevada, March 1991, pp. 7-1 to 7-8. U.S. Department of Agriculture: 1972, ‘Hydrologic Soil Complexes’, National Engineering Handbook, Soil Conservation Service, U.S. Department of Agriculture, Washington, D.C. U.S. Environmental Protection Agency (US EPA): 2002, National Water Quality Inventory: 2000 Report to Congress, EPA-841-R-02-001, Office of Water, Washington, D.C. Walker, W. W., Jr.: 1996, ‘Simplified procedures for eutrophication assessment and prediction: User manual’, Instructional Report W–96–2 (updated April 1999), U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. Wang, S. H., Huggins, D. G., Lim, N. C., Baker, D. S., Spotts, W. W., Goodrich C. A., deNoyelles, F., Jr., Campbell, S. W., Frees, L. and Volkman, C.: 2003, ‘Cheney Reservoir water quality and its watershed assessment’, Kansas Biological Survey, Report No. 112, Lawrence, KS, p. 37.