Examination of Nutrient Sources and Transport in a Catchment with an Audubon Certiﬁed Golf Course

water

Article Examination of Nutrient Sources and Transport in a Catchment with an Audubon Certiﬁed Golf Course

1,2, , 2 3 1 Emilio Grande * † , Ate Visser , Pamela Beitz and Jean Moran 1 Department of Earth and Environmental Sciences, California State University East Bay, Hayward, CA 94542, USA; [email protected] 2 Nuclear and Chemical Science Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA; [email protected] 3 East Bay Regional Park District, Oakland, CA 94605, USA; [email protected] * Correspondence: [email protected] Now at the Department of Earth and Planetary Sciences, University of California Santa Cruz, † Santa Cruz, CA 95064, USA. Received: 5 August 2019; Accepted: 11 September 2019; Published: 14 September 2019

Abstract: Water bodies in the East Bay Regional Park District (EBRPD), California, United States, provide aesthetic value and critical ecosystem services, but are often adversely affected by the activities and infrastructure of the intensely urban environment that surrounds the parks. EBRPD leases a golf course (Tilden Golf Course (TGC)) in Tilden Regional Park, one of its most popular parks located in the Berkeley Hills, which was certified as an Audubon Cooperative Sanctuary in 2013. Nonetheless, application of nutrients and pesticides (fungicides, plant growth regulators and herbicides) are commonly used to maintain turf systems and may be transported via surface runoff or through subsurface drainage to surface waters, leading to the concern that golf courses are a major contributor to water pollution. We studied the possible contribution of nutrients (NO3-N and PO4-P) and pesticides transported via storm-generated surface runoff and via groundwater from TGC to the primary drainage in the watershed, Wildcat Creek. Lake Anza, a popular open water swimming lake, is located downstream from TGC and experiences occasional nutrient-driven algal blooms that have caused swim beach closures. Measured NO3-N and PO4-P in the stream, at times, exceeded concentration limits of 1 mg/L (as N) and 0.05 mg/L (as P), respectively, considered protective of aquatic ecosystems by the United States Environmental Protection Agency (1986). We found that phosphorous likely has a dominant natural source, but nitrogen is primarily derived from a golf course fertilizer source and its concentration increases in the stream during runoff events, while other soluble species decrease. Analyses of pesticides in water reveal the presence of Azoxystrobin in stream water at the golf course, but with concentrations well below the regulatory limit. These results indicate that all other pesticides applied on TGC are not likely transported to the stream, suggesting future reactive transport research must treat contaminant species independently based on their specific transport behaviors.

Keywords: golf course; nitrate; phosphate; nutrient transport; eutrophication; algal blooms; reactive transport; pesticides

1. Introduction An increased occurrence of eutrophication and cyanobacteria in surface waters has led to the need for a greater understanding of potential sources of nutrient contributions to water resources [1]. Recent studies in the water bodies of the East Bay Regional Park District (EBRPD) [2] have focused on cyanobacteria blooms in Lake Temescal, Oakland, California; such blooms have also occurred in Lake Anza, Berkeley, California. The latter is part of the Wildcat Creek hydrologic system and downstream

Water 2019, 11, 1923; doi:10.3390/w11091923 www.mdpi.com/journal/water Water 2019, 11, 1923 2 of 14 from the reach that is the subject of this research (Figure1). The Wildcat Creek catchment has not been studied extensively, but is of great importance due to the diverse land use occurring within its perimeter. For instance, Tilden Golf Course (TGC), run by a private concessionaire, is an important recreational resource that is traversed by Wildcat Creek, a perennial stream that provides essential ecosystemWater 2019 services., 11, x FOR PEER REVIEW 3 of 15

FigureFigure 1. Location1. Location of of Wildcat Wildcat Creek Creek in in the the East East Bay Bay Hills Hills of of the the San San Francisco Francisco Bay Bay Area. Area.Also Also shownshown in thein ﬁgure the figure Tilden Tilden Golf Golf Course Course and and Lake Lake Anza, Anza, the the section section of of Wildcat Wildcat CreekCreek that is is the the subject subject of of this this research,research, and and the the locations locations of of the the key key sampling sampling points,points, NorthNorth Bridge (NB), (NB), near near the the outlet outlet of of the the golf golf coursecourse and and Padre Padre (PD), (PD), upstream upstream of of the the golf golf course. course.

2. TheMaterials goal of and this Methods research was to quantify nutrients, specifically nitrate and phosphate (NO3-N, and PO4-P), transport in storm-generated surface runoff and groundwater from TGC into Wildcat Creek, and to evaluate2.1. Site Description the effect of pesticides and fungicides applied on TGC to surface water quality. TGC became a certified Audubon Cooperative Sanctuary in 2013. This management certification incorporates land Wildcat Creek is a perennial stream that bisects Tilden Golf Course (TGC), flows through Lake management goals that include chemical use reduction, water quality management and conservation, Anza and drains into northern San Francisco Bay, California. The subject of this study is the section andof habitat Wildcat enhancement Creek that intersects for wildlife TGC [3 ].upstream Specific from enhancements Lake Anza at (Figure TGC include 1). The conversionarea is characterized of fairways to naturalizedby a Mediterranean roughs climate (no irrigation, with the fertilizationwater year ex ortending pesticides) from adjacentlate October to creekto April. banks, The reductionaverage of chemicalannual precipitation inputs and within safer productthe watershed choices, is 720 reduction mm [16] of and irrigation the mean footprints, annual temperature irrigation upgrades,is 14 °C sediment[17]. Under pond these construction, climatic conditions, riparian restoration perennial andflow enhancements is sustained by in groundwater. downstream reaches. However, due to the frequent use of fertilizers and pesticides to maintain turf systems, golf courses can be contributors of water pollution [4,5], particularly, golf courses have been described to increase total phosphorous concentration in stream water [6]. Nitrogen and phosphorus are essential plant nutrients that are applied to golf courses, where they have been detected in leachate and surface runoff [1]. Surface water concentrations of phosphorus as low as 0.025 mg/L and nitrogen concentrations as low as 1 mg/L have been associated with increased algal growth [1,7]. Using these guidelines to control eutrophication, in 1986 the United States Environmental Protection Agency (EPA) established water quality criteria that limit total phosphorus concentration to 0.025 mg/L for lakes or reservoirs, 0.05 mg/L for streams draining into lakes or reservoirs [8]. Current recommended limits of total phosphorus (as P) concentration in lakes and reservoirs have been set for California and neighboring states (Ecoregion III) at 0.017 mg/L for most lakes and reservoirs [9]. Specific limits have also been listed by the EPA on California, ranging from 0.008 to 0.3 mg/L and 0.005 to 0.25 mg/L for specific lakes and streams, respectively, where most values are more stringent than 0.025 mg/L[10]. Phosphorus concentrations generally increase with

Water 2019, 11, 1923 3 of 14 increasing discharge and decrease with increasing event duration [11]; other studies have reported that transport of most of the phosphorus from agricultural areas to surface waters occurs during a few large runoff events [12], while another [7] found that most of the nutrient loading into surface water in urban environments comes from a small number of landscapes. The off-site transport of phosphorus in runoff from croplands tends to occur with soil particles, and the dissolved forms are more common in runoff from turf, because sediment loss is insignificant [7]. Shuman (2002) [13] observed that the mass of phosphorus in runoff from a golf course fairway turf was directly related to the fertilizer application rate, where the initial runoff event contained the highest concentration of transported phosphorus. Mallin and Wheeler (2000) [14], in a study in North Carolina, found that the nitrate concentration in stream water increased while passing through golf courses. Nutrient concentrations in golf course ponds and nearby streams have been found to be higher than in other surrounding areas [15]. Alternatively, King et al. (2007) [4], formulated the hypothesis that nutrient losses in storm event runoﬀ from a well-managed turfgrass golf course does not pose an environmental risk. Therefore, this research examines the possible contribution of nutrients to Wildcat Creek due to turf management of TGC and evaluates the total mass of nutrients transported by the creek for a signiﬁcant storm event.

2. Materials and Methods

2.1. Site Description Wildcat Creek is a perennial stream that bisects Tilden Golf Course (TGC), ﬂows through Lake Anza and drains into northern San Francisco Bay, California. The subject of this study is the section of Wildcat Creek that intersects TGC upstream from Lake Anza (Figure1). The area is characterized by a Mediterranean climate with the water year extending from late October to April. The average annual precipitation within the watershed is 720 mm [16] and the mean annual temperature is 14 ◦C[17]. UnderWater these 2019 climatic, 11, x FOR PEER conditions, REVIEW perennial ﬂow is sustained by groundwater. 4 of 15

FigureFigure 2. Land 2. Land Cover Cover of theof the study study area area [ 18[18].]. NoticeNotice the the pr proximityoximity of ofTilden Tilden Regional Regional Park Park to a highly to a highly urbanizedurbanized area, area, the the city city of Berkeley,of Berkeley, California. California. DevelopedDeveloped open open space space (shown (shown in yellow) in yellow) consists consists of of areasareas with with a mixture a mixture of someof some constructed constructed materialsmaterials but but mostly mostly vegetati vegetationon in the in theform form of lawn of lawn grass, grass, includingincluding parks, parks, lawns, lawns, athletic athletic ﬁelds fields and and golf golf courses.courses.

The Wildcat Creek upper watershed exhibits diverse land use (Figure 2)[18]. The catchment is exclusively within Tilden Regional Park; nevertheless, Figures 1 and 2 show that the catchment is in close proximity to an urban area. In 2012, Landscape Architecture and Engineering, a consulting company contracted by the EBRPD, carried out a restoration project of Wildcat Creek within the golf course with the goal of enhancing habitat for fish and amphibians by way of a step and pool system [19]. The project consisted of building engineered in-stream steps to dissipate flow and create habitat for aquatic species, recontouring stream banks (Figure 3), and revegetation with native riparian plants [20]. Downstream of TGC, Wildcat Creek becomes channelized and concrete lined as it passes through the EBRPD botanical garden before reaching Lake Anza.

Water 2019, 11, 1923 4 of 14

The Wildcat Creek upper watershed exhibits diverse land use (Figure2)[ 18]. The catchment is exclusively within Tilden Regional Park; nevertheless, Figures1 and2 show that the catchment is in close proximity to an urban area. In 2012, Landscape Architecture and Engineering, a consulting company contracted by the EBRPD, carried out a restoration project of Wildcat Creek within the golf course with the goal of enhancing habitat for ﬁsh and amphibians by way of a step and pool system [19]. The project consisted of building engineered in-stream steps to dissipate ﬂow and create habitat for aquatic species, recontouring stream banks (Figure3), and revegetation with native riparian plants [ 20]. Downstream of TGC, Wildcat Creek becomes channelized and concrete lined as it passes through the EBRPDWater 2019 botanical, 11, x FOR garden PEER REVIEW before reaching Lake Anza. 5 of 15

FigureFigure 3. 3.Step Step and and Pool Pool system system builtbuilt onon WildcatWildcat Creek along the the golf golf course course to to dissipate dissipate flow ﬂow and and reducereduce stream stream turbidity. turbidity. Picture Picture from from Restoration Restoration DesignDesign Group [21]. [21].

TheThe Wildcat Wildcat Creek Creek Catchment Catchment isis characterizedcharacterized by the presence presence of of a asyncline, syncline, the the axis axis of ofwhich which roughlyroughly delineates delineates the the course course of of Wildcat Wildcat CreekCreek [22] [22],, and and which which also also likely likely controls controls groundwater groundwater flow. flow. TheThe limbs limbs of of the the syncline syncline dip dip at at 35 35◦°and and67 67◦° toto thethe south-west and and north-east north-east respectively respectively [22]. [22 ].These These conditionsconditions make make the the watershed watershed quite quite steep, steep, withwith TGC being the the only only relatively relatively flat flat topography topography of ofthe the areaarea (Figure (Figure1). 1). This This flattened flattened topography topography was was largelylargely createdcreated through alteration alteration for for the the development development of theof the golf golf course. course. A A summary summary of of the the catchmentcatchment characteristics is is listed listed in in Table Table 1.1 . Fertilizers are applied on Tilden Golf Course by both spray techniques and dry, granular broadcast Table 1. Summary of catchment characteristics. and include products with the following compounds: KNO3, FeSO4, MgSO4, and MnSO4. Pesticide productsArea usedof upper on thecatchment golf course largely consist of fungicides 3.5 and km plant2 growth regulators, followed by herbicides.Average annual These rainfall include: Podium (EPA #100-937), Chipco 720 mm Flo 26019 (EPA #432-888), Dorado (EPA #100-741),Elevation Proxy range (EPA #432-1230). TGC has been compliant 239–559m with the EBRPD policy of no phosphorous-basedGeology [22] fertilizer and Basalts slow-release from late Miocene nitrogen (Bald use. Peak KNO Basalt3 is and applied Moraga periodically Formation) at an applicationLand ratecover decided type [18] by the superintendent, based on weather conditions, season and soil moister. The highestMixed rates forest of application correspond to the summer months, 48% where on average 100 pounds of KNO3 perEvergreen month are forest applied. During winter, the rate of application 21% is 50 pounds per month. Scrub/Shrub 13% Developed open area 8% Grassland 7% Low intensity developed 3% Others 0%

Fertilizers are applied on Tilden Golf Course by both spray techniques and dry, granular broadcast and include products with the following compounds: KNO3, FeSO4, MgSO4, and MnSO4. Pesticide products used on the golf course largely consist of fungicides and plant growth regulators, followed by herbicides. These include: Podium (EPA #100-937), Chipco Flo 26019 (EPA #432-888),

Water 2019, 11, 1923 5 of 14

Table 1. Summary of catchment characteristics.

Area of upper catchment 3.5 km2 Average annual rainfall 720 mm Elevation range 239–559m Geology [22] Basalts from late Miocene (Bald Peak Basalt and Moraga Formation) Land cover type [18] Mixed forest 48% Evergreen forest 21% Scrub/Shrub 13% Developed open area 8% Grassland 7% Low intensity developed 3% Others 0%

2.2. Data Collection and Analysis Two locations were selected to collect water samples during this study (Figure1). North Bridge, located at the outlet of TGC, and Padre, a reference site located upstream and outside of the influence of TGC. Samples were collected at both locations during baseflow (Summer and Fall) and during storm events in 2017 and 2018. Stream samples collected during baseflow were collected as individual grab samples. One sample of imported irrigation water was also collected and analyzed for nutrient content. Irrigation water for TGC has an imported source, East Bay Municipal Utility District (EBMUD) supplies all water available for the park for public use and irrigation. This water source comes from a high elevation reservoir in the Sierra Nevada Mountains. In addition, EBMUD water is isotopically light with respect to stable isotopes of the water molecule of local water, has low-dissolved solids and is disinfected with chloramine. During precipitation events, stream samples were collected using a Teledyne ISCO automatic water sampler programmed to collect samples at specified time intervals. The ISCO was deployed at NB and the sampling intervals were determined based on the intensity and duration of the storm events. All anions, including NO3-N and PO4-P, analyses were carried out by Standard Method EPA 300.1, which employs ion chromatography, at the California state certified EBMUD water laboratory in Oakland, California. Pesticides and fungicides were analyzed by isotope dilution gas chromatography-mass spectrometry, after Cassada et al. (1994) [23] and by liquid chromatography/mass spectrometry, after Magalhaes et al. (2009) [24] at the Water Science Laboratory, University of Nebraska Lincoln. A total of 20 different pesticides were analyzed, including: Azoxystrobin, Thiacloprid, Thiamethoxam, Trifloxystrobin, Acetochlor, Alachlor, Atrazine, Chlorothalonil, DEA, DIA, Dimethenamid, EPTC, Metolachlor, Pendamethalin, Permethrin, Propachlor, Propazine, Simazine, Tefluthrin, Trifluralin. Stable isotopes of water samples (δ2H and δ18O) were analyzed by Cavity Ring Down Spectrometry with a Los Gatos Research liquid water stable isotope analyzer. Chloride, used as a tracer for hydrograph separation, was modeled for some stream water samples based on the linear relationship (R2 = 0.9895) between measured chloride and electrical conductivity (view Supplementary Figure S1, scatter plot and linear regression of Cl vs Conductivity). Spearman’s rank-order correlations were used to assess the relationship between the different solutes studied, including the association between stream discharge and nutrient concentrations. Hypothesis tests (t-test) were studied to determine the statistical significance of the differences observed on nutrient concentrations upstream and at the golf course. The cutoff for statistical significance (alpha) on our study was 0.05, all the statistical analyses were run using the functions available in base R (Vienna, Austria) [25].

3. Results and Discussion

3.1. Hydrograph Separation Storm hydrograph components were quantified using geochemical tracers (δ18O and chloride), and three end-member mixing analysis was carried out using data collected by the ISCO sampler Water 2019, 11, 1923 6 of 14 at North Bridge. Examination of stable isotopes of stream water samples at this location revealed a surprisingly large fraction of isotopically light irrigation water entering the stream at the golf course, even during precipitation events. Figure4 shows the δ18O and chloride values for the three end-members, and stream samples, for one of the runoff events analyzed during this study. All of the stream water samples plot inside the mixing triangle, supporting the notion that stream water is a mixtureWater of 2019 three, 11, sources x FOR PEER of REVIEW water in the stream (Baseflow/Groundwater, Event Precipitation water7 of 15 and Irrigation Water), the proportions of which vary over the event. The samples plotted in Figure4 were collectedand at Irrigation 1.5-hour Water), intervals the duringproportions the firstof which 24 h vary and atover 2-hour the event. intervals The samples for the remainingplotted in Figure 96 h of4 the were collected at 1.5-hour intervals during the first 24 hours and at 2-hour intervals for the remaining precipitation-runoff event. 96 hours of the precipitation-runoff event.

-3

-4

-5

-6

-7

-8

-9 O (permil) measured O (permil) 18 δ

-10 Irrigation Baseflow -11 Event Precip Stream -12

-13 02468101214161820 Cl− (mg/L) modeled FigureFigure 4. Graphical 4. Graphical representation representation of of end end members members andand samplessamples taken taken for for hydrograph hydrograph separation separation duringduring an event an event from from February February 28th 28th toto MarchMarch 6th, 6th, 2018. 2018. The The geochemical geochemical tracers tracers used for used the 3 for end- the 3 end-membermember mixingmixing model were were δ18δ18O O(vertical (vertical axis) axis) and andchloride chloride (horiz (horizontalontal axis). The axis). graph The shows graph that shows that thethe stream stream samples samples plotplot insideinside the mixing mixing triangle triangle,, agreeing agreeing with with the the thr threeee end end members members identified. identiﬁed. Error barsError ofbars end of membersend members represent represent standard standard deviation deviation ofof the mean mean of of measurements measurements for forthe the3 end 3 end members.members. Error barsError onbars stream on stream samples samples are standard are stan deviationdard deviation (analytical (analytical uncertainty) uncertainty) from individual from measurements.individual Horizontalmeasurements. error Horizontal bars are e withinrror bars the are symbol within the size. symbol size.

This sizableThis sizable input input of of irrigation irrigation waterwater to to the the stream stream is illustrated is illustrated in Figure in Figure5, which5, shows which three- shows component mixing during the storm hydrograph for the precipitation event showed in Figure 4. three-component mixing during the storm hydrograph for the precipitation event showed in Figure4. Figure5 shows that as the rainfall intensity increases (see Figure6 for the event hyetograph), there is an increase in the amount of irrigation water entering the stream. We hypothesize that irrigation water accumulates in the vadose zone during dry periods, and when there is a precipitation event, the irrigation water (pre-event water) is pushed out into the stream by the new event water that is infiltrating into the subsurface. The figure also illustrates that after the event nearly all of the irrigation water has been flushed out of the system and there is a two end-member mixture of “new” precipitation water and “old” baseflow/pre-event water. An analogous storm hydrograph, with similar component fractions, was observed for another runoff event, illustrating the importance of pre-event water, which includes a significant fraction of irrigation water, to stream flow generation in Wildcat Creek during runoff events (see Supplementary Figures S2–S4 for the storm hyetograph, graph of three end-member

Water 2019, 11, x FOR PEER REVIEW 8 of 15

100% 90% 80% 70% 60% 50%

Water 201940%, 11, 1923 7 of 14 30% 20% mixing and the hydrograph separation, respectively, for a precipitation event studied on April 5th to 10% theWater 9th, 2018).2019, 11, x FOR PEER REVIEW 8 of 15 0% 100% 0:30 3:30 6:30 9:30 2:00 6:00 2:00 6:00 2:00 6:00 2:00 6:00 12:30 15:30 18:30 21:30 12:30 15:30 18:30 21:30 10:00 14:00 18:00 22:00 10:00 14:00 18:00 22:00 10:00 14:00 18:00 22:00 10:00 14:00 18:00 22:00 90%2/28/2018 Fraction of irrigation Fraction of Baseflow Fraction of Rain 3/06/2018 80% 70% Figure 5. Storm Hydrograph for a precipitation event on February 28th to March 6th, 2018. Note the 60%gap (marked by a dashed line on the graph) between 23:00 on February 28th and 10:00 March 1st, 50%during which no samples were collected. 40% Figure 5 shows that as the rainfall intensity increases (see Figure 6 for the event hyetograph), 30% there is an increase in the amount of irrigation water entering the stream. We hypothesize that irrigation20% water accumulates in the vadose zone during dry periods, and when there is a precipitation event,10% the irrigation water (pre-event water) is pushed out into the stream by the new event water that0% is infiltrating into the subsurface. The figure also illustrates that after the event nearly all of the irrigation water has been flushed out of the system and there is a two end-member mixture of “new” 0:30 3:30 6:30 9:30 2:00 6:00 2:00 6:00 2:00 6:00 2:00 6:00 precipitation12:30 15:30 18:30 water21:30 and “old”12:30 15:30 baseflow/pre-event18:30 21:30 10:00 14:00 18:00 22:00 water.10:00 14:00 18:00 An22:00 analogous10:00 14:00 18:00 storm22:00 hydrograph,10:00 14:00 18:00 22:00 with 2/28/2018 similar component fractions,Fraction wasof irrigation observed forFraction anothe of Baseflowr runoff event,Fraction illustrating of Rain the importance3/06/2018 of pre-event water, which includes a significant fraction of irrigation water, to stream flow generation Figure 5. Storm Hydrograph for a precipitation event on February 28th to March 6th, 2018. Note the in WildcatFigure 5.Creek Storm duringHydrograph runoff for eventsa precipitation (see Suppl eventementary on February Figures 28th to S2, March S3 6th,and 2018. S4 for Note the the storm gap (marked by a dashed line on the graph) between 23:00 on February 28th and 10:00 March 1st, hyetograph,gap (marked graph by ofa dashed three end-memberline on the graph) mixing betw aneend the23:00 hydrograph on February separation, 28th and 10:00 respectively, March 1st, for a during which no samples were collected. precipitationduring which event no studiedsamples onwere April collected. 5th to the 9th, 2018). 0.2 0.00 Figure 5 shows that as the rainfall intensity increases (see Figure 6 for the event hyetograph), 0.18 there is an increase in the amount of irrigation water entering the stream. We hypothesize2.00 that irrigation0.16 water accumulates in the vadose zone during dry periods, and when there is a precipitation 4.00 event, 0.14the irrigation water (pre-event water) is pushed out into the stream by the new event water

that is0.12 infiltrating into the subsurface. The figure also illustrates that afterP (mm) the event nearly all6.00 of the irrigation water has been flushed out of the system and there is a two Qend-member (m3/s) mixture of “new” 0.1 mm precipitationm3/s water and “old” baseflow/pre-event water. An analogous storm hydrograph,8.00 with similar0.08 component fractions, was observed for another runoff event, illustrating the importance of 10.00 pre-event0.06 water, which includes a significant fraction of irrigation water, to stream flow generation in Wildcat Creek during runoff events (see Supplementary Figures S2, S3 and S4 for the storm 0.04 12.00 hyetograph, graph of three end-member mixing and the hydrograph separation, respectively, for a 0.02 precipitation event studied on April 5th to the 9th, 2018). 14.00 0.20 0.00

0.18 2.00 Figure0.16 6. Wildcat Creek’s hyetograph and hydrograph for the precipitation event shown in Figure5. Figure 6. Wildcat Creek’s hyetograph and hydrograph for the precipitation event shown in Figure 5. The shaded area denotes a period of missing data, between 23:00 on February 28th and 10:00 March4.00 1st, The0.14 shaded area denotes a period of missing data, between 23:00 on February 28th and 10:00 March marked1st, marked by the dashedby the dashed line in line Figure in Figure5. 5. 0.12 P (mm) 6.00 Q (m3/s) 3.2. Nutrient0.1 Concentrations in Wildcat Creek mm m3/s 8.00 Nutrient0.08 samples collected upstream and at the outlet of the golf course during baseflow showed 10.00 an increase0.06 in nutrient concentrations for both NO3-N and PO4-P at the downstream location (Figure7). Paired t-test analysis, carried out for samples collected on the same day upstream and downstream of 0.04 12.00 TGC were t (4) = 2.79, p = 0.049 for NO3-N and t (4) = 3.87, p = 0.018 for PO4-P. With mean differences 0.02 of 0.9 mg/L and 0.04 mg/L for NO3-N and PO4-P, respectively. The observed p-values14.00 were less than the alpha0 level of 0.05, indicating that the increase in nutrients concentration at the golf course was statistically significant. NO3-N concentrations observed at North Bridge at times exceeded the 1 mg/L concentration that the EPA [8] recommends as a maximum to avoid eutrophication for a stream Figure 6. Wildcat Creek’s hyetograph and hydrograph for the precipitation event shown in Figure 5. drainingThe into shaded a lake. area POdenotes4-P concentrations a period of missing measured data, between in the 23:00 stream on February always 28th exceeded and 10:00 March the 0.05 mg/L maximum1st, valuesmarked by of the PO dashed4-P considered line in Figure protective 5. for aquatic ecosystems [8]. These results indicate

Water 2019, 11, x FOR PEER REVIEW 9 of 15

3.2. Nutrient Concentrations in Wildcat Creek Nutrient samples collected upstream and at the outlet of the golf course during baseflow showed an increase in nutrient concentrations for both NO3-N and PO4-P at the downstream location (Figure 7). Paired t-test analysis, carried out for samples collected on the same day upstream and downstream of TGC were t (4) = 2.79, p = 0.049 for NO3-N and t (4) = 3.87, p = 0.018 for PO4-P. With mean differences of 0.9 mg/L and 0.04 mg/L for NO3-N and PO4-P, respectively. The observed p-values were less than the alpha level of 0.05, indicating that the increase in nutrients concentration at the golf course was Water 2019statistically, 11, 1923 significant. NO3-N concentrations observed at North Bridge at times exceeded the 1 mg/L8 of 14 concentration that the EPA [8] recommends as a maximum to avoid eutrophication for a stream draining into a lake. PO4-P concentrations measured in the stream always exceeded the 0.05 mg/L that althoughmaximum there values is of an PO increase4-P considered of PO4 protective-P concentration for aquatic downstream ecosystems of[8]. TGC, These even results under indicate natural that although there is an increase of PO4-P concentration downstream of TGC, even under natural conditions (at Padre), the system is naturally rich in PO4, supporting eutrophication. conditions (at Padre), the system is naturally rich in PO4, supporting eutrophication.

FigureFigure 7. Boxplots 7. Boxplots of nutrientsof nutrients concentrations concentrations forfor samples collected collected upstream upstream and and downstream downstream of of Water 2019, 11, x FOR PEER REVIEW 10 of 15 TGC.TGC. The graph The graph shows shows an increase an increase in concentrations in concentrations at the at golfthe coursegolf course respect respect to the to upstreamthe upstream samples for bothsamples solutes for both analyzed, solutes NOanalyzed,-N ( leftNO)3-N and (left PO) and-P PO (right4-P ().rightp-values). p-values obtained obtained from from paired t-t-test however, increased, while chloride3 decreased. For4 the same event, the Spearman’s correlation were 0.048test were and 0.048 0.018 and for 0.018 NO3-N for NO3-N and PO4-P, and PO4-P, respectively, respectively, confirming confirming the statisticalthe statistical significance significance of the coefficient between NO3-N and flow rate is 0.54 with a p-value of 0.01; revealing the presence of an observedof the increases. observed increases. additional source of NO3-N into Wildcat Creek. The observed concentrations of NO3-N in the stream during precipitation events are likely lessened by the periodic addition of irrigation water (with very The eTheffects effects of the of the golf golf course course on on Wildcat Wildcat Creek Creek duringduring runoff runoff eventsevents were were considered considered using using time time low NO3-N concentration) to the stream, and possibly by denitrification in turf soils that may become series data from the ISCO sampler. Figure 8 illustrates the results obtained for one of the precipitation seriesanaerobic data from during the ISCO intense sampler. precipitation. Figure 8All illustrates the runoff the events results analyzed obtained during for one this of study the precipitationyielded events studied (66.5 mm [26]) on February 28th, 2018. Samples were collected at 1.5-hour intervals. eventssimilar studied patterns (66.5 of mm increased [26]) on NO February3-N with increased 28th, 2018. runoff Samples as illustrated were collected in Figure at8. 1.5-hour intervals. Figure 8 shows an increase in the nutrient concentrations with increase in stream discharge. NO3- N100 concentrations were below 1 mg/L at the beginning of the event but exceeded that value0.14 as the precipitation event advanced. This result contrasts with results from most studies on rainfall-runoff relationships [27,28], where NO3-N concentrations in streams decrease by dilution during Cl- (mg/L) 0.12 precipitation events. In Wildcat Creek, NO3-N increased with rainfall intensity, likely because the golf10 course is a source of readily soluble NO3-N. This is further evidenced by the fact that conservative 0.10

chloride and electrical conductivity decreased with increasing discharge (Figure 8), showing a typical/s) 𝑇𝑢𝑟𝑏 (𝑁𝑇𝑈) 3 dilution pattern. NO3-N, a soluble100 anion, is expected to behave similarly to chloride and decrease in 0.08 concentration in a natural system with no anthropogenic inputs. The relationship between chloride 1mg/L NO3-N and1 stream discharge during the precipitation event (shown in Figure 8) was examined using a Spearman’s rank order correlation, yielding a correlation coefficient (ρ) of −0.92 and a p-value <0.06 0.001, Cond. (mS/cm) illustrating the strong, negative correlation between this solute and stream discharge. NO3-N, Stream Stream Discharge (m 0.04 0.1

0.05 mg/L PO4-P 0.02

0.01 0.00

Q (m3/s) Nitrate-N (mg/L) Phosphate-P (mg/L) Conductivity (mS/cm) Turbidity (ntu/100) Chloride (mg/L)

Figure 8. NO3-N, PO4-P, Cl−, conductivity and turbidity sampled at TGC during a storm event from Figure 8. NO3-N, PO4-P, Cl−, conductivity and turbidity sampled at TGC during a storm event from February 28th to March 1st, 2018 in Wildcat Creek. Samples were taken at 1.5 h intervals. Primary February 28th to March 1st, 2018 in Wildcat Creek. Samples were taken at 1.5 h intervals. Primary verticalvertical axis axis is a is logarithmic a logarithmic scale. scale. The The horizontal horizontal dasheddashed lines mark mark the the maximum maximum NO NO3-N3 -Nand and PO4 PO- 4-P recommendedP recommended values values by the by EPA the EPA to avoid to avoid eutrophication eutrophication for for a stream a stream that that drains drains into into a a lake. lake. TheThe error barserror are 3 barsthe are minimum 3× the minimum detection detection limit. limit. ×

PO4-P concentrations in Figure 8 exceed the 0.05 mg/L limit throughout the event, with concentration values increasing as stream discharge increases; confirmed by the Spearman’s rank- order correlation between these two variables, yielding ρ = 0.45 and a p-value of 0.01. Similar observations were reported in the literature [29,30], and are explained by the fact that PO4 is associated with particulate matter. Figure 8 also illustrates the close association of PO4 with turbidity ( = 0.67 and a p-value < 0.001), showing a strong relationship between the two parameters. During a precipitation events, samples were collected upstream and at the golf course, before and soon after the precipitation event. The results showed an increase in nutrient concentration in the stream after the events at both locations (Table 2). These findings are common for PO4 due to the strong association of PO4 with stream turbidity. PO4-P concentrations always exceeded the allowed limit, even at Padre, likely due to a decomposition of organic matter, or a lithologic or sedimentary source of this solute within the watershed. Similar observations have been made in an adjacent catchment [2], where excess phosphorous was attributed, in part, to the erosion of apatite-rich geologic formations. Our results, however, showed an increase in PO4-P concentration at the golf course with respect to the upstream reference location, revealing that although the system naturally has excess PO4, the effects of land use (e.g., removal of natural vegetation during construction of the golf course), favoring runoff over infiltration, could have an influence on turbidity, and therefore, potentially, on elevated values of PO4-P levels in Wildcat Creek.

Water 2019, 11, 1923 9 of 14

Figure8 shows an increase in the nutrient concentrations with increase in stream discharge. NO3-N concentrations were below 1 mg/L at the beginning of the event but exceeded that value as the precipitation event advanced. This result contrasts with results from most studies on rainfall-runoff relationships [27,28], where NO3-N concentrations in streams decrease by dilution during precipitation events. In Wildcat Creek, NO3-N increased with rainfall intensity, likely because the golf course is a source of readily soluble NO3-N. This is further evidenced by the fact that conservative chloride and electrical conductivity decreased with increasing discharge (Figure8), showing a typical dilution pattern. NO3-N, a soluble anion, is expected to behave similarly to chloride and decrease in concentration in a natural system with no anthropogenic inputs. The relationship between chloride and stream discharge during the precipitation event (shown in Figure8) was examined using a Spearman’s rank order correlation, yielding a correlation coefficient (ρ) of 0.92 and a p-value < 0.001, illustrating the − strong, negative correlation between this solute and stream discharge. NO3-N, however, increased, while chloride decreased. For the same event, the Spearman’s correlation coefficient between NO3-N and flow rate is 0.54 with a p-value of 0.01; revealing the presence of an additional source of NO3-N into Wildcat Creek. The observed concentrations of NO3-N in the stream during precipitation events are likely lessened by the periodic addition of irrigation water (with very low NO3-N concentration) to the stream, and possibly by denitrification in turf soils that may become anaerobic during intense precipitation. All the runoff events analyzed during this study yielded similar patterns of increased NO3-N with increased runoff as illustrated in Figure8. PO4-P concentrations in Figure8 exceed the 0.05 mg /L limit throughout the event, with concentration values increasing as stream discharge increases; confirmed by the Spearman’s rank-order correlation between these two variables, yielding ρ = 0.45 and a p-value of 0.01. Similar observations were reported in the literature [29,30], and are explained by the fact that PO4 is associated with particulate matter. Figure8 also illustrates the close association of PO 4 with turbidity (ρ = 0.67 and a p-value < 0.001), showing a strong relationship between the two parameters. During a precipitation events, samples were collected upstream and at the golf course, before and soon after the precipitation event. The results showed an increase in nutrient concentration in the stream after the events at both locations (Table2). These findings are common for PO 4 due to the strong association of PO4 with stream turbidity. PO4-P concentrations always exceeded the allowed limit, even at Padre, likely due to a decomposition of organic matter, or a lithologic or sedimentary source of this solute within the watershed. Similar observations have been made in an adjacent catchment [2], where excess phosphorous was attributed, in part, to the erosion of apatite-rich geologic formations. Our results, however, showed an increase in PO4-P concentration at the golf course with respect to the upstream reference location, revealing that although the system naturally has excess PO4, the effects of land use (e.g., removal of natural vegetation during construction of the golf course), favoring runoff over infiltration, could have an influence on turbidity, and therefore, potentially, on elevated values of PO4-P levels in Wildcat Creek.

Table 2. Results of chemical analysis carried out upstream and at the golf course before (11/8/2017) and soon after (11/9/2017) a precipitation event studied.

Location Date NO3-N N (mg/L) PO4-P (mg/L) K (mg/L) Na (mg/L) NB 11/8/17 0.48 0.091 0.911 27.4 PD 11/8/17 0.19 0.075 0.501 22.7 NB 11/9/17 2.2 0.13 3.69 23.4 PD 11/9/17 1.2 0.1 0.826 19.9

Concentrations of NO3-N were also observed to increase during the precipitation event sampled at Padre, the upstream reference site, exceeding the 1 mg/L limit (Table2). Nonetheless, as shown in Table2, potassium was observed to increase at a much higher rate at North Bridge than at Padre (65% and 305% increase at Padre and North Bridge, respectively), while sodium did not change considerably Water 2019, 11, x FOR PEER REVIEW 11 of 15

Table 2. Results of chemical analysis carried out upstream and at the golf course before (11/8/2017) and soon after (11/9/2017) a precipitation event studied.

Location Date NO3-N N (mg/L) PO4-P (mg/L) K (mg/L) Na (mg/L) NB 11/8/17 0.48 0.091 0.911 27.4 PD 11/8/17 0.19 0.075 0.501 22.7 NB 11/9/17 2.2 0.13 3.69 23.4 PD 11/9/17 1.2 0.1 0.826 19.9

Concentrations of NO3-N were also observed to increase during the precipitation event sampled

Waterat Padre,2019, 11the, 1923 upstream reference site, exceeding the 1 mg/L limit (Table 2). Nonetheless, as shown10 of 14in Table 2, potassium was observed to increase at a much higher rate at North Bridge than at Padre (65% and 305% increase at Padre and North Bridge, respectively), while sodium did not change beforeconsiderably and after before the runoand afterﬀ event the (12%runoff and event 15% (1 decrease2% and 15% at Padre decrease and Northat Padre Bridge and respectively),North Bridge indicatingrespectively), that indicating potassium that nitrate potassium (KNO3), nitrate a fertilizer (KNO used3), a onfertilizer the golf used course, on the is agolf likely course, source is ofa likely NO3 tosource the stream. of NO3 to the stream. AdditionalAdditional NONO33 sampling in a spring upstream ofof TGCTGC alsoalso showedshowed aa slightslight increaseincrease inin NONO33-N-N concentrationconcentration afterafter aa rainrain eventevent studiedstudied (Figure(Figure9 9),), suggesting suggesting a a natural natural source source of of NO NO33.. However, thethe increaseincrease inin NONO33-N at this springspring waswas substantiallysubstantially damped with respect to thethe golfgolf coursecourse increaseincrease (190%(190% andand 420%420% increaseincrease upstreamupstream andand atat thethe golfgolf course,course, respectively,respectively, forfor aa precipitationprecipitation eventevent studied).studied). ThisThis indicatesindicates that, that, although although natural natural sources sources of of NO NO3 are3 are present present in thein the basin basin (runo (runoffﬀ from from leaf litterleaf /litter/forestedforested land, land, atmospheric), atmospheric), the golf the course golf course supplies supplies an additional an additional source ofsource NO3-N of toNO Wildcat3-N to CreekWildcat that Creek raises that inputs raises above inputs recommended above recommended levels. levels.

2.7

2.4

2.1

1.8

1.5 -N (mg/L) 3 1.2

NO 0.9

0.6

0.3

0 Before Precipitation Event After Precipitation Event

Spring Water Surface Water (NB)

Figure 9. NO3-N sampled in a spring upstream of TGC and at NB before and after a precipitation event,Figure on 9. NovemberNO3-N sampled 15th and in a 16th, spring 2017, upstream respectively. of TGC and at NB before and after a precipitation event, on November 15th and 16th, 2017, respectively. 3.3. Concentration Discharge Analysis 3.3. ConcentrationConcentration-discharge Discharge Analysis (C-Q) graphs were analyzed for two precipitation events to examine the transportConcentration-discharge of nutrients during runo(C-Q)ff graphsevents. were C-Q relationshipsanalyzed for cantwo revealprecipitation hydrochemical events to processes examine occurringthe transport in theof nutrients stream, and during therefore runoff aidevents. in determining C-Q relationships the origin can reveal and fate hydrochemical of nutrients inprocesses surface wateroccurring [29]. in This the relationshipstream, and therefore usually describes aid in determ a hysteresisining the loop, origin especially and fate during of nutrients large dischargein surface events,water [29]. with This scouring relationship that mobilizes usually nutrientsdescribes [a31 hy].steresis Hysteresis loop, loops especially have been during observed large discharge between dischargeevents, with and scouring many water that qualitymobilizes parameters, nutrients including[31]. Hysteresis NO3 and loops PO have4 [28, 31been]. C-Q observed hysteresis between loop direction is controlled by changes in stream discharge and temperature, while the magnitude is discharge and many water quality parameters, including NO3 and PO4 [28,31]. C-Q hysteresis loop controlleddirection is by controlled antecedent by rainfall changes [31 in]. stream The slope discharge and direction and temperature, of the loop providewhile the information magnitude on is flowcontrolled pathways by antecedent and dominant rainfall sources [31]. duringThe slope runo andff events. direction Clockwise of the loop rotational provide direction information typically on occursflow pathways when higher and dominant concentrations source ares during observed runoff on theevents. rising Clockwis limb ofe therotational hyetograph direction compared typically to the falling limb, indicating flushing of nutrients from near sources [29]. Counterclockwise directional rotation commonly occurs in the presence of delayed solutes from deeper subsurface zones or distant tributaries [31]. Positive slopes represent solute flushing, while negative slopes indicate solute dilution [32]. Both events analyzed showed similar patterns in the slope and rotation direction. Figure 10 shows the C-Q relationship for the same runoff event as shown in Figure8. The graph illustrates a complex hysteresis loop and figure-eight pattern, as described by Lloyd at al. (2016) [28]. These complex loops are produced by a change in the form of the relationship between discharge and suspended sediment concentration during the precipitation event. Interestingly, at the end of the event, the concentrations of both NO3-N and PO4-P are higher than their initial concentrations even though the last sample plotted was collected 40 hours after the precipitation event, highlighting the hysteresis process. Water 2019, 11, x FOR PEER REVIEW 12 of 15

occurs when higher concentrations are observed on the rising limb of the hyetograph compared to the falling limb, indicating flushing of nutrients from near sources [29]. Counterclockwise directional rotation commonly occurs in the presence of delayed solutes from deeper subsurface zones or distant tributaries [31]. Positive slopes represent solute flushing, while negative slopes indicate solute Water 2019, 11, 1923 11 of 14 dilution [32]. Both events analyzed showed similar patterns in the slope and rotation direction.

3.0 0.16 3/01/18 05:00 0.14 2.5 3/01/18 02:00 3/01/18 11:00 0.12

2.0 3/01/18 23:00 0.1 3/06/18 08:00

1.5 0.08 -N (mg/L) -P (mg/L) 3 4 PO

NO 3/01/18 14:00 2/28/18 23:00 3/02/18 14:00 0.06 1.0

0.04 3/01/18 03:30 3/01/18 11:00 0.5 3/06/18 08:00 0.02

2/28/18 12:30 0.0 0 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Stream Discharge (m3/s) Nitrate-N (mg/L) Phosphate-P (mg/L)

Figure 10. Hysteresis loop observed from Concentration-Discharge relationship during runoff event fromFigure February 10. Hysteresis 28th to March loop observed 6th, 2018. from Concentration-Discharge relationship during runoff event from February 28th to March 6th, 2018. The NO3-N loop has a positive slope (lower left to top right), which translates as flushing of NO3-NFigure to the stream10 shows during the C-Q the relationship event. The for NO the3 -Nsame C-Q runoff loop event follows as shown a counterclockwise in Figure 8. The direction,graph indicatingillustrates a delayeda complex arrival hysteresis of this loop solute, and figure-eight likely entering pattern, the as described stream dissolved by Lloyd inat al. pre-event (2016) [28]. water. TheseThese findings complex agree loops with are produced results obtained by a change by Aguilerain the form and of the Melack relationship (2018) between [29] in theirdischarge research and on suspended sediment concentration during the precipitation event. Interestingly, at the end of the streams in coastal California, where NO3-N enrichment/flushing with anticlockwise direction was observedevent, the on undevelopedconcentrations sites.of both NO3-N and PO4-P are higher than their initial concentrations even though the last sample plotted was collected 40 hours after the precipitation event, highlighting the PO -P values followed a complex hysteresis loop, with a clockwise direction and the PO -P peak hysteresis4 process. 4 occurring early in the runoff event, which is evidence for quick flushing of particulate matter that is The NO3-N loop has a positive slope (lower left to top right), which translates as flushing of NO3- exhausted by the end of the precipitation event. Of note is the fact that the PO4-P concentration at the N to the stream during the event. The NO3-N C-Q loop follows a counterclockwise direction, beginningindicating of thea delayed event wasarrival below of this the solute, 0.05 mg likely/L limit entering but later the stream surpassed dissolved that value in pre-event and stayed water. above it afterThese the findings event, fromagree Februarywith results 28th obtained at 12:30 by PM Aguile untilra March and Melack 6th at (2018) 8:00 AM, [29] 139.5in their h. research The hysteresis on loopstreams for PO in4 di coastalffers from California, C-Q graphs where described NO3-N enrichment/flushing by Lloyd et al. (2016) with [28 anticlockwise], where 35 out direction of 41 analyzed was runoobservedff events on exhibit undeveloped anti-clockwise sites. behavior for PO4. The clockwise loop shown in Figure 10 may be related toPO a4-P source values of followed PO4 close a complex to the stream, hysteresis such loop, as particulate with a clockwise phosphorous direction from and golfthe PO course4-P peak runo ff, becauseoccurring its concentration early in the runoff peaks event, before which the is peak eviden ofce the for hyetograph. quick flushing A of distant particulate source matter of PO that4 would is causeexhausted a delayed by the arrival end of the this precipitation solute to the event. stream, Of note and is an the anticlockwise fact that the PO loop.4-P concentration at the beginning of the event was below the 0.05 mg/L limit but later surpassed that value and stayed above 3.4.it Other after Waterthe event, Quality from Issues February 28th at 12:30 PM until March 6th at 8:00 AM, 139.5 hours. The hysteresis loop for PO4 differs from C-Q graphs described by Lloyd et al. (2016) [28], where 35 out of Fungicides are the primary pesticide applied to support grass growth and avoid the development 41 analyzed runoff events exhibit anti-clockwise behavior for PO4. The clockwise loop shown in of disease on the golf course turf. Fungicide application occurs only on greens, constituting 3 acres of Figure 10 may be related to a source of PO4 close to the stream, such as particulate phosphorous from the overall 120 acres occupied by the golf course. We studied the possible presence of pesticides in in Wildcat Creek due to land use by sampling upstream and downstream from the golf course. Sampling was carried out before and after precipitation events, and during baseflow. While several pesticides were analyzed, only Azoxystrobin, a broad-spectrum fungicide, was found in the stream. This chemical was observed only at North Bridge, not upstream from the golf course, and its concentration increased after precipitation events. However, per the U.S. EPA’s Pesticide Fact Sheet of Azoxystrobin (1997) [33], the maximum allowable toxicity concentration of Azoxystrobin in water is 168 µg/L, protective for freshwater fish and aquatic invertebrates. The highest value of this chemical observed in Wildcat Creek Water 2019, 11, 1923 12 of 14 was 0.282 µg/L (minimum detection limit of 0.01 µg/L). Therefore, the presence of this fungicide in the stream is not expected to pose environmental risk to Wildcat Creek’s ecosystem or any aquatic organism. The irrigation water used on TGC is disinfected by addition of chloramine, which could be a water quality issue, especially considering its toxicity to fish [34] and to soil microbes that can play an important role in nutrient management [35]. However, contact of chloramine with soil and sediments with organic matter should lead to sorption and rapid degradation of chloramine, and consequently, much lower (or zero) concentrations of chloramine should be expected in Wildcat Creek. On the other hand, Song et al. (2019) [35] observed a reduction in the microbial community diversity in soils exposed to chlorine by drip irrigation, which could inhibit its transformation and adsorption by nutrients in the soil. If chloramine breaks down to chlorine and ammonia in the stream, then volatilization will decrease its concentration. While sampling for chloramine or its degradates was not performed, it is unlikely that chloramine is transported by Wildcat Creek and thus is not expected to be a direct water quality issue for the ecosystem.

4. Conclusions

Measured NO3-N and PO4-P in Wildcat Creek, at times, exceeded concentration limits of 1 mg/L and 0.05 mg/L respectively, considered protective of aquatic ecosystems. Observations upstream and downstream from the golf course illustrated that phosphorous likely has a natural source, and that the system is naturally rich in PO4. Nevertheless, an increase in PO4-P downstream from TGC was observed in every sample analyzed. No phosphate-based fertilizer is currently used at the golf course; however, TGC may have contributed to a historical legacy of PO4 that affects water quality in the stream. Indirectly, the presence of the golf course may be influencing the increase of PO4-P in Wildcat Creek because removal of native vegetation and changes in soil types may reduce infiltration and favor runoff, which would manifest as an increase in PO4-Pconcentration in the creek. PO4 transport to the stream may have also been accelerated by the reduction of residence time of water within the golf course, caused by the continuous application of irrigation water to the system. NO3, unlike PO4, likely has a golf course fertilizer source, as potassium nitrate (KNO3) fertilizer is periodically applied on TGC. Analysis of stream samples revealed that NO3-N concentrations downstream from the golf course were elevated with respect to background levels. Of note was the increase in NO3 concentration in the stream during runoff events. When stream discharge increased, conservative chloride decreased by dilution, but NO3-N concentration increased. Concentration- discharge hysteresis loops studied for NO3-N during precipitation events suggest a delayed arrival of NO3-N, likely entering the stream dissolved in pre-event water. Despite significant improvements to the ecological functions of the TGC and the subsequent Audubon Sanctuary Certification, integrated geochemical analyses suggest that Tilden Golf Course has a somewhat negative effect on water quality to downstream receptors and may be an additional source of excess nutrients to Lake Anza, where eutrophication is a concern. PO4 from historic fertilizer use and natural sources may have also accumulated in bottom sediments of Lake Anza after runoff events. In anoxic conditions, bound PO4 is released into the water column, promoting algal growth. NO3 in the system also appears to have some natural provenance, but was observed to increase substantially downstream from golf course. It is important to note that Wildcat Creek, between the golf course and Lake Anza, is channelized and concrete lined. This disconnection between the creek and its bed, bank, flood plain and riparian community results in little opportunity for nutrient uptake before deposition in the Lake Anza impoundment. Therefore, excess nutrients accumulate in the lake at a greater rate than they would under natural conditions, possibly contributing to algal growth.

Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4441/11/9/1923/s1, Figure S1: Scatter plot and linear regression of chloride vs conductivity, Figure S2: Storm hyetograph for precipitation event on April 5th to the 9th, 2018, Figure S3: graph of three end-member mixing and Figure S4: hydrograph separation for the event illustrated in Figure S2. Water 2019, 11, 1923 13 of 14

Author Contributions: Formal analysis, E.G. and J.M.; Funding acquisition, P.B.; Investigation, E.G.; Methodology, A.V. and J.M.; Supervision, J.M.; Validation, P.B.; Writing–original draft, E.G.; Writing–review & editing, E.G., A.V., P.B. and J.M. Funding: This research received no external funding. Acknowledgments: The authors sincerely thank the East Bay Regional Park District and the Tilden Golf Course staff for the support and the data provided. We would also like to thank the three anonymous reviewers for their useful and targeted suggestions and comments that have improved the quality of the manuscript. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Rice, P.; Horgan, B. Nutrient loss with runoff from fairway turf: An evaluation of core cultivation practices and their environmental impact. Environ. Toxicol. Chem. 2011, 30, 2473–2480. [CrossRef][PubMed] 2. Lovelace, F. Sources of Excess Phosphate Leading to Cyanobacteria Blooms at Lake Temescal, Oakland, CA. Master’s Thesis, California State University East Bay, Hayward, CA, USA, 2017. 3. Mullen, M.W.; Bell, M.A. Sanctuary Program for Golf Courses to Improve Water Quality: Experience by Audubon International and other States. Audubon Int. 2004, 2, 4. 4. King, K.W.; Balogh, J.C.; Hughes, K.L.; Harmel, R.D. Nutrient Load Generated by Storm Event Runoff from a Golf Course Watershed. J. Environ. Qual. 2007, 36, 1021–1030. [CrossRef][PubMed] 5. Baris, R.D.; Cohen, S.Z.; Barnes, N.L.; Lam, J.; Ma, Q. A quantitative analysis of over twenty years of golf course monitoring studies. Environ. Toxicol. Chem. 2010, 29, 1224–1236. [CrossRef][PubMed] 6. Mackey, M.J.; Connette, G.M.; Peterman, W.E.; Semlitsch, R.D. Do golf courses reduce the ecological value of headwater streams for salamanders in the southern Appalachian Mountains? Landsc. Urban Plan. 2014, 125, 17–27. [CrossRef] 7. Walker, W.J.; Branham, B. Environmental impacts of turf-grass fertilization. In Golf Course Management and Construction: Environmental Isues; Balogh, J.C., Walker, W.J., Eds.; Lewis Publisher Inc.: Chelsea, MI, USA, 1992; pp. 105–219. 8. United States Environmental Protection Agency. Quality Criteria for Water; US. EPA 440/5-86-001; U.S. EPA: Washington, DC, USA, 1986. 9. U.S. Environmental Protection Agency. Ambient Water Quality Criteria Recommendations: Information Supporting the Development of State and Tribal Nutrient Criteria for Lakes and Reservoirs in Nutrient Ecoregion III 2001; US EPA: Washington, DC, USA, 2001. 10. US EPA State Progress toward Developing Numeric Nutrient Water Quality Criteria for Nitrogen and Phosphorus. Available online: https://www.epa.gov/nutrient-policy-data/state-progress-toward-developing- numeric-nutrient-water-quality-criteria (accessed on 3 September 2019). 11. Quinton, J.; Catt, J.; Hess, T. The removal of phosphorus from soil: Is event size important? J. Environ. Qual. 2001, 30, 538–545. [CrossRef][PubMed] 12. Edwards, W.M.; Owens, L.B. Large storm effects on total soil erosion. J. Soil Water Conserv. 1991, 46, 75–77. 13. Shuman, L.M. Phosphorus and Nitrate Nitrogen in Runoff Following Fertilizer Application to Turfgrass. J. Environ. Qual. 2002, 31, 1710–1715. [CrossRef][PubMed] 14. Mallin, M.; Wheeler, T. Nutrient and fecal coliform discharge from coastal North Carolina golf courses. J. Environ. Qual. 2000, 29, 979–986. [CrossRef] 15. Lewis, M.A.; Boustany, R.G.; Dantin, D.D.; Quarles, R.L.; Moore, J.C.; Stanley, R.S. Effects of a Coastal Golf Complex on Water Quality, Periphyton, and Seagrass. Ecotoxicol. Environ. Saf. 2002, 53, 154–162. [CrossRef] [PubMed] 16. PRISM Climate Group, Oregon State U. Available online: http://www.prism.oregonstate.edu/normals/ (accessed on 7 May 2019). 17. PWS Dashboard|Weather Underground. Available online: https://www.wunderground.com/dashboard/pws/ KCABERKE150 (accessed on 7 May 2019). 18. NOAA: Data Access Viewer. Available online: https://coast.noaa.gov/dataviewer/#/landcover/search/ (accessed on 7 May 2019). 19. Stream Alteration Agreement, Wildcat Creek, East Bay Regional Park District; EBRPD: Oakland, CA, USA, 2012. Water 2019, 11, 1923 14 of 14

20. Tilden Golf Course Wildcat Creek Restoration Design; Landscape Architecture and Engineering: Berkeley, CA, USA, 2012. 21. Restoration Design Group. Available online: http://www.restorationdesigngroup.com/ (accessed on 20 May 2019). 22. Graymer, R.W. Geologic Map and Map Database of the Oakland Metropolitan Area, Alameda, Contra Costa, and San Francisco Counties, California. U. S. Geological Survey, Miscellaneous Field Studies MF-2342, scale 1:50,000. Available online: https://pubs.usgs.gov/mf/2000/2342/ (accessed on 20 February 2018). 23. Cassada, D.A.; Spalding, R.F.; Cai, Z.; Gross, M.L. Determination of atrazine, deethylatrazine and deisopropylatrazine in water and sediment by isotope dilution gas chromatography-mass spectrometry. Anal. Chim. Acta 1994, 287, 7–15. [CrossRef] 24. Magalhaes, L.C.; Hunt, T.E.; Siegfried, B.D. Efficacy of neonicotinoid seed treatments to reduce soybean aphid populations under field and controlled conditions in Nebraska. J. Econ. Entomol. 2009, 102, 187–195. [CrossRef][PubMed] 25. R Core Team. R: A Language and Environment for Statistical Computing; R Fundation for Statistical Computing: Vienna, Austria, 2019. 26. California Data Exchange Center-Query Tools. Available online: https://cdec.water.ca.gov/ (accessed on 9 May 2019). 27. Moravcová, J.; Pavlíˇcek,T.; Ondr, P.; Koupilová, M.; Kvítek, T. Comparison of parameters influencing the behavior of concentration of nitrates and phosphates during different extreme rainfall-runoff events in small watersheds. Hydrol. Earth Syst. Sci. Discuss. 2013, 10, 12105–12151. [CrossRef] 28. Lloyd, C.E.M.; Freer, J.E.; Johnes, P.J.; Collins, A.L. Using hysteresis analysis of high-resolution water quality monitoring data, including uncertainty, to infer controls on nutrient and sediment transfer in catchments. Sci. Total Environ. 2016, 543, 388–404. [CrossRef] 29. Aguilera, R.; Melack, J.M. Concentration-Discharge Responses to Storm Events in Coastal California Watersheds: C-Q Storm Responses Coastal California. Water Resour. Res. 2018, 54, 407–424. [CrossRef] 30. USGS Cooperative Matching Funds. Available online: https://www.usgs.gov/mission-areas/water-resources/ science/usgs-cooperative-matching-funds?qt-science_center_objects=0#qt-science_center_objects (accessed on 9 May 2019). 31. Bieroza, M.Z.; Heathwaite, A.L. Seasonal variation in phosphorus concentration–discharge hysteresis inferred from high-frequency in situ monitoring. J. Hydrol. 2015, 524, 333–347. [CrossRef] 32. Andrea, B.; Francesc, G.; Jérôme, L.; Eusebi, V.; Francesc, S. Cross-site Comparison of Variability of DOC and Nitrate c–q Hysteresis during the Autumn–winter Period in Three Mediterranean Headwater Streams: A Synthetic Approach. Biogeochemistry 2006, 77, 327–349. [CrossRef] 33. United States Environmental Protection Agency. Office of Prevention, Pesticides and Toxic Substances. In Pesticide Fact Sheet: Azoxystrobin; US EPA: Washington, DC, USA, 1997. 34. United States Environmental Protection Agency. Chloramines-Related Research; US EPA: Washington, DC, USA, 2009. 35. Song, P.; Feng, G.; Brooks, J.; Zhou, B.; Zhou, H.; Zhao, Z.; Li, Y. Environmental risk of chlorine-controlled clogging in drip irrigation system using reclaimed water: The perspective of soil health. J. Clean. Prod. 2019, 232, 1452–1464. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).