Loading...
01-1112�A�.��a� �d - �o� a$ a.OQ, Councii File # � �+'� �� Green Sheet # 1 \'j �. � y ORDINA OF SAINT P� Presented Referred To ..� Committee Date � �o 1 An ordinance amending Saint Paul Legislative Code Chapter 377 t� ��the use of fertilizers containing 2 phosphorus 3 THE COUNCIL OF THE CITY OF SA1NT PAUL DOES ORDAIN: Section 1 5 Chapter 377 of the Saint Paul Legislative Code is hereby amended to read as follows: 6 Sec. 377.01. Definitions. For the purposes of this chapter, the terms defined in this secrion have the meanings ascribed to them: 8 Person means any person, firm or corporation engaged in the business of lawn fertilizer or pesticide 9 applications and includes those persons licensed by the State of Minnesota pursuant to Minnesota Statutes, Secrion 10 18A-21 et seq. 11 Pesticide means any substance or mixture of substances intended for prevenring, destroying, repelling or 12 mifigating any pest, and any substance or mixture of substances intended for use as a plant regulator, defoliant or 13 desiccant. It also means any chemical or combination thereof registered as a pesticide with the U.S. Environmental 14 Protection Agency, any agency later assuming registration in the U.S. federal government, the State of Minnesota 15 Agricultural Deparhnent, or any other State of Minnesota government agency. 16 Sec. 377.02. License required; council approval. 17 (a) No person shall engage in the business of lawn fertilizer or lawn pesticide application in Saint Paul without 18 a license issued by the City of Saint Paul. 19 (b) All city programs for pesticide use shall be reviewed and approved by the city council prior to any application 20 upon city property. 21 Sec. 377.03. Fee. 22 The fee required for a license shall also be established by ordinance as specified in section 310.09(b) of the 23 Saint Paul Legislative Code. ; ► �' ;. . 1 Sec. 377.04. Employees licensed by state. O \ � �\\9� All ofiicensee's employees actually engaged in lawn pesticide applications shail be duly licensed by the State of Miunesota and shall be trained and qualified in the proper methods of handling and applications of pesticides. Satisfactory evidence that such employees are licensed by the state shall be maintained on file in the office of the license inspector. 6 Sec. 377.05. Division of health. 7 The � - DirectoroftheOfficeofLicense,Inspections 8 and Environmental Protection or his/her desi¢nee is directed to monitor the health and safety effects ofthe chemical 9 applications to lawns and to advise the ciTy council of any suspected hazards or violations. 10 Sec. 377.06. Class I license. 11 The license granted pursuant to the provisions of this chapter is designated as a Class � R license, subj ect to 12 the procedures applicable to Class � R licenses in Chapter 310. 13 Sec. 377.07. Pesticide applications; posting. 14 15 16 17 18 19 20 All persons who apply pesticides outdoors are required to post or affix warning signs on the street frontage ofthe properry so treated. The warning signs must protrude a minimum of eighteen (18) inches above the top ofthe grass line. The warning signs must be of a material rain-resistant for at least a forty-eight-hour period and must remain in place for at least a forty-eight-hour period or longer if the human re-enhy interval prescribed in the pesticide label specifies a longer hourly or daily interval. The information printed on the sign must be printed in contrasring colors and capitalized letters at least one-half inch or in another format approved by the coxnmissioner of agriculture, and shall provide the following information: 21 (1) 22 23 (2) 24 25 26 27 The name of the company applying the pesticide or, if not a company, the name of the person having done the application. The following language: "This area chemically treated. Keep children orpets offuntil (date of safe entry--at least forry- eight (48) hours after applicafion or longer if specified on pesricide label)" ar a universally accepted symbol and text approved by the commissioner of agriculture specifying a date of safe entry as specified herein. The warning sign may include the name of the pesticide used. 28 The sign shall be posted on the lawn or yard no closer than two (2) feet from the sidewalk ar right-of-way and no 29 further than five (5) feet from the sidewalk or right-of-way. 30 31 Sec. 377.08. Fertilizer Content. No person licensed under this chapter sha11 appl�y lawn fertilizer. Iiquid or 32 p_ranulaz. within the Citv of Saint Paul that is labeled to contain mare than 0% phosphate (P O ,�prohibition 33 shall not appplv to: 1 a. The naturallv occurrin� phosnhorus in unadulterated natural or organic fertilizing O l-1\�Y 2 products such as vard waste compost; 3 b. Use on newlv established or developed turf and lawn azeas durin¢ their first growing 4 season• 5 c. Turf and lawn azeas which soil tests taken according to Universit�of Minnesota 6 ¢uidelines and analyzed in a State of Minuesota certified laboratorv confirm are 7 below phosphorus levels established by the Universitv of Minnesota. In such cases, 8 lawn fertilizer application shall not exceed the Universitv ofMinnesota recommended 9 application rate for phosphorous. 10 11 12 13 14 15 16 17 Lawn fertilizers contaiuingphos hro orus applied pursuant to the above-listed exceptions shall be watered into the soil where the phosphorus can be nnmobilized and enerallv nrotected from loss bv runoff. Fertilizer applied to impervious services, such as sidewalks. drivewavs and streets is to be removed bv sweepins or other means immediately after fertilizer application is completed. Fertilizer is not to be apniied to frozen soil. saturated soil or under conditions ofim ep nding heaw rainfall. The Office ofLicense, Inspections and Environmental Protection shall be notified at least 24 hours prior to the application of anv lawn fertilizer containing_phosphorus that such fertilizer will be used. the amount to be used and the reason for its a�lication. Section 2 18 This ordinance sha11 take effect and be in force thiriy (30) days following its passage, approval and publication. Yeas Na s Absent Benanav i /' Blakey f Bostrom „/ Coleman ✓ Harris ,/ Lantry �/' Reiter �' Adopted by Council: Date - �. S� 1,oa � AdopUon Certified by Council Secretary By: 2 . � --Q- Approved by Mayor: Date By: � Requested by Deparnnent o£ � Form Approved by Ciry Attomey �� By: Approved by Mayor for Submission to Council � By: � 2 7 'A'� : �� °—�.- S a9�J�� V < �,.. , IST BE ON CIXRJCILAGB�IDA BY OCt. Z�F, ZOO1 a „��,m. ,_ _ ,. C 1- ►1\3 o/laio� GREEN SHEET No 113684 � TOTAL # OF SIGNATURE PAGES u��� u��— ❑ arvwnoa�r ❑ rnrcuu _ ❑ waxr,n�fEau¢ESO.� ❑ qmcu�mnri�eero ❑ ❑ (CLJP ALL LOCATIONS FOR SICaNATURE) An ordinance amending Saint Paul Legislative Code Chapter 377 to prohibit the use of fertilizers containing phosphorus. PLANNING COMMISSION qB COMMITTEE CNIL SERVICE CAMMISSION IIy_\ �=,ti]qd�7 ' TOTAL AMOUNT OP TRANSACTION S �� ' FllNOING SOURCE FlNqNCIAL INFORMAl10N (EXPWf� H36 th15 p2fSMl�fi112R! VqIKIM UlMN 8 COIIffdC[ f0f fhi6 d2p3lIIT2M? VES NO Has thia Pe���m e<u been a dty employee9 ' YES NO Does tlds PersaVfim+ V� as4dN nM na�matHP� b'f anY cwrent dty emWoyee't YES MO Is ttiis peBOMfirtn a targeted �endoYt YES � COST/REVENUE BUDGEfED (GRCLE ON� ACTIVIiY NUMBER VES NO ❑ YINNESOTA BOARD OF WATEfl ANC SOIL NESOURCES NORTHERN FiEGION 394 S Lake Ave Room 403 Duluth, MN 55802-2325 PHONE (218)723-2350 FAX (218) 723-4794 �MINNESOTA BOAND OF WATEfl AND SOIL RESOURCES METRO REGION One W Water SL, Suite 200 St. Paul, MN 55107-2039 r� , Water Resources Education Date: November 12, 2001 To: Councilmember 7ay Benanav Councilmember Jerry Blakey Councilmember Dan Bostrom Councilmember Chris Coleman Councilmember Pat Harris Councilmember Kathy Lantry Councilmember Jun Reiter ! �� UNIVERSITY OF MINNESOTA Extension �� O I - ��l'�- � B � ,M,�,.,,,,,,. ' �CC�G�4�L n'YV!' 1 3 2 �Di �4TY ��ERK Mayor Norm Coleman From: Ron Struss, University of Minnesota Extension __'Sf Re: UM comments on proposed lawn fertilizer ordinances and regulations PHONE (651) 215-1950 FAx Attached are comments from a team of University of Minnesota specialists (651) 297-5615 developed to heip inform the City of St. Paul on their proposed lawn fertilizer [{ MiNNE50TA 80ARD OF ordinance and the Minnesota Senate on their November 15, 2001 hearing on W4TER AND SOIL 1aW11 fe1�.111ZEI$. * RESOUXCES SOU7HERN REGION 261 Highway 15 S New Ulm, MN 56073-8915 PHONE (507) 359-6090 If you would like clarification of these comments or fizrther information, please contact Dr. Carl Rosen at 612-625-811A or crosen(�a,soils.umn.edu. �12.i11C }�011. FAX (507) 359-6018 . � .. � << 0 � -I��a' Comments on proposed ordinances and legislation relating to lawn fertilizers University of Minnesota - November 9, 20U1 Carl Rosen, Extension Soil Scientist, Dept. of Soil, Water, and Climate, Univ. of Minnesota Brian Horgan, Extension Turf Specialist, Dept. of Horticulhu�al Science, Univ. of Minnesota Don White, Professor, Dept. of Horticulrisal Science, Univ. of Minuesota Robert Mugaas, Extension Educator, Hennepin County, Univ. of Minnesota Doug Foulk, Extension Bducator, Ramsey County, Bniv. of Minnesota Ron Struss, Extension Educator, Water Resources Center, Univ. of Minnesota Phosphoms is an essential element required by all forms of life. However, high phosphorus inputs have been linked to degradafion of lakes and rivers through promoting excessive algae gowth. The overall intent of the proposed ordinances and legislation is to reduce the amount of phosphorus entering surface waters and improve water quality. Proposed ordinances and legislation will restrict the use of phosphoms containing fertilizer for established lawns unless a need is indicated by a soil test. The rationale for this is supported by two sound premises: 1) Surveys conducted over the past 30 years have shown that 70% to 80% of the lawns in the Twin City Metropolitan Area have soil phosphorus levels in the very high range and would not require additional phosphorus for oprimal huf growth, and, 2) Applicafion of phosphorus to lawns not requiring phosphorus is a waste of a limited resource. An underlying assumprion is that reshicring the use of phosphorus on lawns will reduce the amount of phosphoms entering surface waters. Unfortunately, the scientific evidence to show that such a restricrion will improve water quality is lacldng. In fact, the pioneering studies conducted by the University of Minnesota in the 1970's suggest that in the short term, use of phosphorus fertilizer on lawns has little impact on phosphorus runoff compared to- the amounts of phosphorus in runoff resulting from breakdown of orgaxuc material such as leaves and grass clippings. Clearly, more quanfitanve research is needed to determine the fate of phosphorus in the lawn landscape and how resh-icting phosphoms fertilizer use for lawns will impact water quality. Reseazch proposals have been submitted by a team of turf gass and soil scientists at the University of Minnesota to deteinune the fate of phosphorus applied to lawns and to define management practices that will minimize movement of phosphorus into surface water runoff. Proposed ordinances and legisiarion may also raise expectarions that water quality will dramatically improve once lawns aze not fertilized with phosphorus. The problem is more complicated than simply restricting fertilizer use and will require a much more integrated approach to improve lake q_uality. In addition to reseazch, educational efforts should be implemented to address all pracrices that affect or contribute to phosphorus runoff in urban areas. UM comments on lawn fertilizer ordinances and regulations Page 7 of 2 �� � Y One fmal comment concerns a loophole in exisring and proposed ordinances that allows for application of organic fertilizer containing phosphorus. Most organic fertilizers have a nitrogen-to- phosphorus ratio that is mucfl lower than the common inorganic lawn fertilizers used today. Since the rate of lawn fertilizer applicarion is based on the amount of nitrogen applied, there will likely be more phosphoms applied when an organic fertiIizer is used tban when a more common inorganic lawn fertilizer is used. Since organic fertilizers with a 0% phosphorus label aze available in Mumesota, the reshiction should be for both inorganic and organic fertilizers. In summary, our comments on proposed ordinances and legislation are: • They aze based on a sound premise that regulaz app2icafion of phosphorus is not needed on most Twin City lawns. • Reseazch Yias not yet shown that restricting phosphorus fertilizer use on lawns will improve lake water quality. • They should be considered as one part of an overall phosphoms runoff management program. Lawn fertilizer ordinances or legislation will not solve the water quality problems of Twin City lakes on their own. • Educafian will be needed for successful compliance and reduction of phosphorus in urban nmof£ • The exemption provided for organic fertilizers is neither warranted nor advised. Thank you for the opportunity to comment. If you would like clarificarion of these comments or further information, please contact Carl Rosen at 612-625-8114 or crosen(c�soils.umn.edu. UM comments on lawn fertilizer ordinances and regulations Page 2 of 2 a De�n Vietor, 12:00 PM 11/6/01 -0600, Re: P in turf runoff X-From : dvietor@taexgw.tamu.edu Tue Nov 612:02:48 2001 X-Mailer: Novell GroupWise Intemet Agent 5.5.5.1 Date: Tue, 06 Nov 20�1 12:0037 -0600 From: "Don Vietor" <dvietor@taexgw.tamu.edu> To: Leslie A Everett <evere003@tc.umn.edu> Subject: Re: P in turf runoff Page 1 of Z O 1 — \\\1— I have attached Word97 files of a manuscript for which 1 wili submit revisions to Assoc. Editor of JEQ next week. The title page, text, and tables are in separate files. The manuscript and references represent the latest work we have (that is near publication) for runoff of P fertilizer and manure P from a relatively steep slope of turf. We applied P fertilizer rates that were refatively large, but comparabie to the farge P amounts observed in soi4 sampfes submitted from urban counties in Texas. If I can be of further help, please let me know. We are very interested in the new urban regulations being proposed for the twin cities. ls Ron Struss our best source for informafion related the new regulations? Don Vietor Donald M. Vietor Soil & Crop Sciences Texas A&M University College Station, TX 77843-2474 Tel. (979) 845-5357 FAX (979) 845-0456 email dvietor@tamu.edu »> "LesiieA. Everett" <evere003@tc.umn.edu> 11/05/01 03:41PM »> He44o Don, Now I've got a request for you! The cities of St. Paui and Minneapolis are in the middle of passing or implementing ordinances regarding phosphorus fertilizer use on lawns. The state legislature is also looking at the issue, with a hearing next week. Some people tell me there is no research data out there to support instituting controls on P fertilizer application to turf. My guess is that there must be some, and ihere should be some as weii regarding P fosses from inorganic fertilizer applied to pasture or hayland as an analogous system. Most current research focuses on manure applied to cropland and pastureland, which wouid not represent the lawn situation well. Ftave you got any references along this line? If so, p(ease send me a{ist, as well as an indication of who else 1 should contact. Thanks much, Les Everett � Gaumanu.doc � JEQtable.doc Printed for "Leslie A. Everett" <evere003@tc.umn.edu> 11/7/Ol O�-���� Response of Turf and Quality of Water Runoff to Manure and Fertilizer J.E. Gaudreau RH. White , D. M. �etor , T.L. Provin and C.L. Munster ' Soil & Crop Sciences Depattment and Z Agricultural Engineering Department, Texas A&M University, Coilege Station, Texas 77843-2474 ABBREVIATIONS DP, dissolved phosphorus; NO3 N, nitrate nitrogen; NHa-N, ammonium nitroDen; PP, particulate phosphorus; TKN, total Kjeldahl nitrogen. . < < 1 �\ _\\�.'�- 1 ABSTRACT 2 Manure applications can provide nutrients and other benefits to turfgrass produetion and 3 unused nutrients in manure residues can be exported through sod harvests. Yet, unused nutrients 4 near the soil surface could be transported in surface runoff and be detrimental to water quality. In 5 addition to measurements of bermudagrass (Cynodon dactydon var. Guymon) turf responses, 6 volumes and P and N concentrations of surface runoff were monitored during evaluations of 7 composted manure applications in turfgrass production. Manure rates that supplied 50 and 100 8 kg P ha' at the start of each of two monitoring periods were compared to P fertilizer rates of 25 9 and 50 kg ha' and an unfertilized control. Two applications of [NHa]zSOa (100 kg N ha" were 10 applied with the P fertilizer. Three replications of treatments were estabiished on a Boonviile 11 fine-sandy-loam (fine, smectitic, thermic Ruptic-vertic Albaqual� that was excavated to create 12 an 8.5% slope. Compared to initial soit tests, nitrate concentrations decreased to 2 mg kg 1 and P 13 concentrations increased aRer two manure and fertilizer applications and eight rain events over 14 the two monitoring periods. The fertilizer sources of N and P produced 19% more dry weight 15 and 21% lazger N concentrations in grass clippings than manure sources. Runoff volumes did not 16 differ between manure and fertilizer sources of P, but average volumes recorded for the 17 unfertilized control were 22% greater than either source or rate of P during the second 18 monitoring period. Dissolved P concentration (30 mg L in runoff was 5 times greater for 19 fertilizer than for manure when rain occuned 3 d after P applications at the same rate. Similarly, 20 total dissolved P losses in zunoff above those of the control were 1.4 times greater for fertilizer 21 than for manure when both were applied in two applications at equal P rates (100 ka P ha 1 y') 22 Under the relatively large P rates on a steep slope of turfgrass, P and N losses in runoff during 23 natural rain events were no greater for composted manure than for fertilizer sources of P. � m z p � _ � � �'�- 1 INTRODUCTION 2 Additions of organic amendments, inciuding composted sewage sludge, can reduce soil 3 bulk density and increase water infiitration rate and nutrient holdin� capacity of soil in turf�ass 4 production (An�le, 1994). In addition, the amendments can enhance turfgrass estabiishment and 5 quality compued to fertilizer sources of nutrients. Aithough costs of haulin? and handling 6 organic sources of nutrients are relatively large (Daniel et al., 1998), the high economic values of 7 turfgrass, including sod, can offset those costs. 8 Despite agronomic advantages of sludge and manure applications on turfgrass, nutrient 9 concentrations can increase near the soil surface (Vitosh et al. 1973, Kin�ery et al. 1994, and 10 Lund and Doss, 1980). After mineralization, accnmulations of manure sources of P neaz the soil 11 surface are transportable as both soluble and sediment-bound P in surface runoff (Vitosh et al. 12 1973, Kingery et al. 1994, Romkens et at., 1973). Similar increases of P concentration in surface 13 runoff were observed as rates of P fertilizer on grassland increased (Austin et al., 1996). 14 Large nitrate-N (NOs-I� concentrations in soil can similariy contribute to losses through 15 surface runof�' In addition, inorganic N in fertilizer appiications is soluble in water and readily 16 transported in water flow over and through soil. Linde and Watschke (1997) indicated NO3-N 17 losses in runoff were largest in initial runoff events after fertilizer applications. Runoff losses l8 decline as fertilizer N dissolves and infiltrates with water into soil (Schuman et al. 1973). Unlike 19 fertilizer N, organic N in manure is released slowly through mineralization and nitrification 20 processes. Slow release of the manure N could minimize the portion of N applied on turfgrass 21 that is transported in water, compared to inorganic N sources. 22 Sediment and associated nutrients aze transported with soluble nutrient forms in runoff. 23 In the case of turfgrass, Linde et. al (1995) reported an inverse relationship between plant density J 1 and sediment loss. Similarly, Gross et al. (1991) observed less sediment loss at dense compared 2 to sparse seeding rates of turfgrass. The relatively lazge plant densities of turfgrass could reduce 3 sediment and nutrient loss in runoff compazed to grasslands used for grazing and fora�e 4 production (Romkens et al. 1977). 5 The use and eaport of manure sources of nutrients through turfgrass sod production has 6 been proposed as a practice for reducing P loads on watersheds containing large densities of 7 animal feeding operations (Griffith, 2000). Sod harvest can remove and reduce P concentrations 8 near the soil surface, but potential losses of P and N after surface applications of manure on turf 9 need to be evaluated. The objectives of this study were: l.) Evaluate turf quality and P and N 10 concentrations of turFgrass clippings and soil in response to increasing rates of P and N in dairy 11 manure and inorganic fertilizer, 2.) Compare volumes and P and N concentrations of surface 12 runofF between manure and inorganic fertilizer treatments, and 3.) Relate the rate and source of 13 applied P and N to losses in surface runoff. 01-11 14 . < 4 O 1 _ 1 2 MATERTALS AND METHODS 3 Plot Design and Treatments. 4 5 Common bermudagrass was established on a slope of 8.5%. The ent'ue plot area was 6 hydro-seeded at a rate of 50 kg pure live seed ha 1 during October, 1997. Irrigation maintained 7 soil water content for seedling establishment and turfgrass growth without runoff. Plot 8 dimensions were 4 m x 1.5 m. Sheet metal strips (thickness = 1.9 mm) were inserted 5 cm into 9 soil around the perimeter of each plot to contain runoff. Runoff of each rain event was collected 10 through an H-flume at the base of each plot into an uncovered, 311-L container. 11 Applications of composted dairy manure and inorganic fertilizer comprised five 12 treatments on the slope of bermudagrass. Three replications of the treatments were distributed 13 along the slope in a randomized complete block design during monitoring periods in 1998 and 14 1999. The treatments were: control (no P), 100 and 200 kg P ha 1 y 1 as manure, and 50 and 100 15 kg P ha I y 1 as inorganic fertilizer. Experimental results were analyzed as a split-split plot 16 arrangement of the experimental design. Two monitoring periods (1998 and 1999) were main 17 plots, nutrient sources were sub-plots, and nutrient rates were sub-sub plots within three 18 replications. A single control plot (0 kg P and N ha i y 1 ) was included in each replication. 19 Dairy manure was analyzed before application usin� methods of the Texas A&M Soil, 20 Water and Forage Laboratory (Parkinson and Allen, 1975). Tatal P and N concentrations in 21 composted manure averaged 5.0 and 15.Sg kg i , respectively. The rates of total P, applied as 22 composted dairy manure, were two times those applied as inorganic fertilizer to compensate for 23 the slow release ofP from manure. The inorganic P in fertilizer was assumed completely soluble 24 after prills were applied on the plot surface. The rates of P applied as inorganic fertilizer 25 maintained or increased ea�tractable soil P concentrations above 40 mg kg' and similar to the P � • C 5 p � _\�\lr 1 levels in more than 70% of soil samples submitted from selected urban counties of Texas (T.L. 2 Provin, Personal Communication). 3 Composted dairy manure was applied at the start of monitoring periods in 7une, 1998 and 4 Mazch, 1999. Each application supplied 50% of the total P rate. Similazly, inorganic P was 5 broadcast at rates of 25 and 50 kg P ha I to the respective fertilizer treatmems before runoff 6 monitoring began during each period. In addition to P, inorganic N(100 kg N ha 1 as [NH4] 7 ZSOa) was applied to the two fertilizer treatments. The N rate for each period was split between 8 broadcast applications before runoff monitoring started and applications 61 and 40 days later 9 during the respective monitoring periods in 1998 and 1999. 10 Turl'grass Responses. 11 Plots were clipped 3.8-cm above the soil surface when turf reached a height of 5 to 7.5 12 cm. The first clipping date occurred 17 d after application of both P sources during the first 13 monitoring period. Piant uptake of nutrients was quantified through digestion, and analysis of 14 clipping samples taken during selected mowing dates. Clipping sampies were dried and analyzed 15 for total N and P by the Texas A&M University Soil, Water, and Forage Testing Laboratory 16 (Feagley et al., 1994, McGeehan and Naylor, 1988). 17 Color, density, and quality of turfgrass in plots were rated visually. The monthly ratings, 18 startin� 5 d after initial P applications, were based on a scale of 1 to 9. Brown turf was given a 19 color rating of 1 and dark green turf was rated 9. The density of an open turf canopy with 20 exposed soil was rated 1 and a closed canopy of tillers and leaves was rated 9. Quality ratings 21 integrated consistency, color, density and aesthetics into a single numerical value. Quality, 22 density, and color ratings near 5 represenied an average turfgrass that could be used for a home 23 lawn or sod production 24 Volume and Nutrient Concentration of Runoff. ��F � < � 1 Total runoff volume was determined by multiplying water depth, as a proportion of the 2 maximum, by the container volume. Daily rain amounts were recorded for natural events at an 3 onsite monitoring station. Rain depth for the 24-h period in which measurable runoff occurred 4 was subtracted from the depth of zunoff in containers. After each runoff event, SOOmL was 5 sampled after mixing the volume coilected in containers of each plot_ The samples were frozen 6 immediately to prevent microbial Meakdown of nutrients within the water sample. 7 The particulate fraction of N and P in the SOOmL samples was removed during filtration 8 through 1-µm glass microfiber filter. The 1-µm pore size permitted suction filtering of the 9 sample volume without plugging by organic and clay colloids and total dissolved P(DP) in the 10 fiitrate could be analyzed through Inductively Coupled Plasma optical emission spectroscopy 11 (ICP). In addition, the glass filter disk and particulate fraction were digested to detennine total P 12 and Total Kjeldahl Nitrogen (TKN) (Parkinson and Allen, 1975). Total P in digests of the 13 particulate fraction was analyzed through ICP. The TKN in the digests and the NO3-N and NHa- 14 N of the filtrate were measured in an auto analyzer. The NOs-N was analyzed using cadmium 15 reduction (Dorich and Nelson, 1984) and the NF3a-N was analyzed colorometrically (Dorich and 16 Nelson, 1983, Isaac and Jones, 1970). The NI-7a-N concentrations were measured in runoff of the 17 first three events in 1998 and the initial event in 1999. 18 A tea analysis was completed for three soil samples taken at random across the 3 19 replications of plots on the siope. Each sample comprised 12 to 15 cores, which were 2.5 cm in 20 diameter and 7.5 cm in depth. The soil is described as a U5DA sandy-loam or sandy-ciay-loam 21 containing 56% sand, 24% silt, and 20% clay. The native soil, a Boonville fine-sandy-loam (fine, 22 smectitic, fhermic Ruptic-vertic Albaqual fl, was excavated to construct the 8.5% slope. 23 Each plot was sampled and analyzed prior to the initial N and P applications and after 24 each monitoring period. Ten to 15 soil cores (2.5-cm diameter and depth of 7.5 cm) were �\����r ,� • 4 7 ��_\\\�' i randomly sampled and mixed to provide a plot composite. Bxtractable P and NO3-N of the 2 sampie from each plot were analyzed by the Texas A&M University Soil, Water, and Forage 3 Testing Laboratory. An acidified ammonium acetate - EDTA was used to estimate plant- 4 available P{Hons et al. 1990) and soil nitrate was extracted and analyzed using methods 5 described by Dorich and Nelson (1984). 6 Statistical Analysis. 7 The Statistical Analysis System (SAS, 1988) was used to analyze variation of turf 8 responses, runoff volumes, and P and N concentrations of runoff and soil among monitoring 9 periods, rain events, P sources, and P rates. Numerical ratings of turf and weights and nutrient 10 concentrations of clippings were pooled over the sampling dates of both monitoring periods for i l analysis. The Generalized Lineaz Models Procedure (SAS, 1988) was used to analyze variation 12 of soil nutrients and of volume and DP, NO3-N, and NH quantities for runoff filtrates. 13 Variation of total P and TKN in particulate fractions of runoff was similarly analyzed. When 14 interactions of effects of monitoring periods with P sources and rates were significant (P=0.05), 15 monitoring periods were analyzed separately. Similarly, when interactions between effects of 16 rain events and of P sources and rates were significant (P=0.05), rain events were analyzed 17 separately. The P rates were treated as class variables in the statistical model. � � 4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 � RESULTS AND DISCUSSION Turfgrass Responses. Applications of N totaling 200 kg N ha I with the inorganic P fertilizer contributed to significantly greater ratings (P = 0.05) of turf color, quality, and density than the other treatments during the two monitoring periods (Table 1). The mean color and quality ratings oftreatments fertilized with inorganic P and N were 20% greater than control or manure treatments. The slow release of available N from the composted manure could have limited N availability to turf compared to treatments feztilized with [NH Similar to differences in visual ratings, the treatments fertilized with inorganic P and N yielded 19% a eater dry weights and 21% larger 1V concentrations of clippings than treatments supplied manure sources of P and N(Table 1). The lack of treatment differences in P concentrations of clippings (3.9 mg g') indicated variation of inorganic N suppiy was the principal determinant of larger ratings and yield of turf fertilized with inorganic fertilizer rather than manure. Soil Analysis. Variation of extractable soil P between the two P rates applied as manure or fertilizer was statistically si�nificant (P= 0.05) on sampling dates in March and June, 1999 (Table 2). The soil tests reveal P accumulation neaz the soil surface after applications of dairy manure and inorganic fertilizer. Large amounts of manure residue near the soil surface facilitate removal of manure-P amounts in turf sod that are much greater than P removed in biomass harvests of other grass crops (Crriffitt�, 2000). The slow release of available P from manure was evident in smaller increases of extractable soil P for manure treatments on the two sampling dates in 1999, despite two-fold larger P rates in manure than in fertilizer applications. The relationship between applied �i -1\ lY k �� 9 01-\t�a 1 and ezctractable soil P is consistent with previous reports (Vitosh et al. 1973, Kingery et al. 1994, 2 and Lund and Doss, 1980). 3 In contrast to P, soil NOs-N concentrations of all treatments decreased from beginning to 4 end ofthe study (Table 1). The decrease in soil IvT03-N is consistent with amounts ofN removed 5 in clippings (up to 36 kg ha and estimates of equal or greater amounts of N in grass parts 6 below the cutting height (Schuman et al., 1973). Although not measured in this study, leaching 7 and volatilization losses could have contributed to losses of appiied N and small NOs-N 8 concentrations in soil (7ohnson et al., 1995, Tennan, 1979). 9 Runoff Votume. 10 Runoff volumes differed significantly (P=0.01) among four rain events during each 1 I monitoring period and the first event did not occur unti160 d after the P applications during 1998 12 (Table 2). Asynchrony between rainfall and runoff measurements during 2 days of a pzolonged 13 rain event resulted in a runoff depth greater than the 24-h rain total for event D in 1998. A 14 portion of the rain recorded for event C contributed to runoff measured for D. In addition, the 15 antecedent rainfall of event C saturated the soil and maximized the portion af rain lost as runoff 16 during event D. 17 In addition to event differences, runoff volumes of the unfertilized control were 18 significantly (P=0.05) greater than treatments that received either manure or fertilizer P in 1999. 19 The average volume of the control was 22°lo greater than volumes recorded for either rate or 20 source of P. Relatively large clipping dry weights and density ratings for the two inorganic P 21 rates, which included 100 kg ha 1 of inorganic I�i (Table 1), were consistent with observed 22 differences in runoff volume. Runoff volume was expected to decrease as the plant density 23 ratings of the turf increased (I.inde and Watschke, 1997). . . io O�-ttta- 1 Nutrient Concentrations in Runoff. 2 An interaction between rain events and P rates was significant (P=0.01) for DP in runoff 3 during each monitoring period (1998 and 1999). In contrast to the initial rain event in 1999, DP 4 concentrarions in runoff for the latter three events in 1999 and all four events during 1998 were 5 relatively small (Table 4). Irrigation during the 60-d period between P applications on 6 bermuda�rass turf and the first rain event in 1998 reduced DP concentrations at the soil surface 7 and limited DP concentrations in runoff compared to 1999. The lazge reduction of DP 8 concentrations in runoff a8er the initial event in 1999 was similaz to previous studies of turf and 9 pasture (Edwazds and Daniel, 1994, McLeod and Hegg, 1984, Austin et al., 1996, Linde and 10 Watschke, 1997). 11 The vaziation of runoff concentrations of DP among P rates, including the control, and 12 between P sources was significant (P=0.05) on seven of the rain dates during the two monitoring 13 periods (Table 4). Significant interactions (P=0.001) between P rate and sources revealed greater 14 differences in DP of runoff between P rates of manure than between P rates of fertilizer for five 15 rain events. During seven rain events, vaziation of DP concentrations in runoff corresponded 16 with relative differences in P rate between fertilizer and manure P sources (Table 4). The DP 17 concentrations ofthe 100-kg rate of manure P averaged 2 times greater than the 50-kg rate of 18 fertilizer P for all events in 1998 and the latter three events in 1999. 19 The differences in runoff concentrations of DP between P rates were largest during the 20 initial rain event 3 d after manure and fertilizer applications in 1999 (Tables 3 and 4). In contrast 21 to seven other rain events, differences in mean DP concentrations of runoff of this initial rain 22 between the two P rates of fertilizer were 3 times greater than differences between the two P 23 rates of manure (Table 4). Similarly, differences in DP of runoff between each rate of P fertilizer 24 and the control were 3 times greater than DP differences between respective smaller and larger ! '., ll ��-��,� 1 rates of manure P and the control. The DP concentrations in runoff from the 50-kg rate of 2 fertilizer P were 206% larger than runoff from the 100-kg rate of manure P for first rain in 1999. 3 In a previous comparison between fertilizer and poultry (Gallus gallus domestieus) litter, 4 differences in DP concentration of runoff were greatest between P sources during the first 5 simulated rain event after application on tall fescue (Festuca arunciinacea, Schreber) (Edwards 6 and Daniel, 1994). Dissolved P concentration in runoff from fertilized tall fescue was 2 times 7 greater than runoff concentrarions after the same P rate was applied as poultry litter. The clipping 8 height of tall fescue was 2.4 times taller than that of bermudagrass in the present study. Yet, DP 9 concentrations in the initial runoff after application of comparable P rates were similar between 10 the studies of ta11 fescue and bermudagrass (Table 4). ll Similaz to DP, NO3-N and NH concentrations were largest in the first runoff event 3 d 12 after the manure and fertilizer applications in 1999. In addition, the N source by rate interaction 13 was significant (P=0.01) for NO3 N and NHa-N in runoff of this initial event during 1999. The 14 large NO N concentration in the initial rwioff of the lazger manure rate in 1999 was consistent IS with 5.3 times more total N in the manure than in the initial applicatian of 50 kg N ha as 16 [NHa}ZSOa (Table 5). Previous evaluations of plant uptake of N during the first year after dairy 17 manure application indicated 21% of the I�i in manure was equivalent to N appiied as fertilizer 18 (Klausner et al., 1994). The relatively lazge I�O concentrations in nznoff 3 d after application 19 of the two manure rates during 1999 indicated more than 21% of the total N in composted 20 manure was in nitrate form. Uniike the firsY rain event, the NO concentrations in runoff of 21 fertilized treatments were significantly greater (P=0.05) than manure treatments during rain 22 events B, C, and D of 1999 (Table 5). Larger NO3 N concentrations and losses in runoff from 23 fertillzer compared to manure or organic sources of N have previously been reported (Edwards 24 and Daniel, 1994, McLeod and Hegg, 1984). � 1 12 01-t11� 1 The variation of NO3-N concentrarion in runoff between the total N rates in manure or 2 fertilizer was significant (P=0.05) for five of eight events during both monitoring periods (Table 3 5). The NO concentration in runoff of fertilized piots was 2 to 10 times greater than controls. 4 Concentrations of NOs-N in runoff from the larger manure rate were greater than the control plot 5 for 7 of the 8 runoff events. Austin et ai. (1996) observed comparable increases in NO3-N losses 6 as fertilizer rate was increased. 7 The initia] application of [NHa]zSO4 with P fertilizer in 1999 contributed to 32 mg L� of 8 NHa-N in runoff 3 d later. Similaz NH4-N concentrations were observed in runoff of simulated 9 rainfall shortiy after N feRilizer was applied to tall fescue stands (Edwards and Daniel, 1994). 10 During the monitoring period in 1998, NH concentrations in runoff (3.2 mg L 11 days after 11 the second [NHa]zSOa application were smaller than the initial event in 1999. Irrigation during 12 the 11 d before the rain event couid have dissolved and transported the NHa-N into soii. In 13 contrast to observations after fertilizer appiications, NHa-N concentrations in runoff shortly after 14 composted manure applications in 1998 and 1999 were < 1 mg L (data not shown). Near-zero 15 NHc-N concentrations were observed in simulated runoff 14 d or more after poultry Iitter was 16 applied to tall fescue (Edwazds and Daniel, 1994). 17 Nutrient iosses in runoSf. 18 The potential for removing and exporting lazge amounts of manure P and N through sod 19 is an incentive for lazge manure rates that exceed P and N amounts needed for turf growth 20 (Crriffith, 2004). The volumes and P and N concentrations of runoff on the steep slope of 21 bermudagrass provide estimates of potential P and N losses and environmental impacts of the 22 large manure and fertilizer rates on turf. The DP amounts in runoff differed significantly 23 (P=0.05) between rates and between manure and fertilizer sources during 1999. During eight rain 24 events, the 200 kg of P in two manuze applications contributed 7.1 kg ha 1 more DP to runoff � , li 01-1\\� 1 than the control. A similar loss of DP in runoff was observed after two fertilizer applications 2 totaling 100 kg P ha i . The DP losses during eight rain events following applications of lower 3 manure rates totaling 100 kg P ha' were 3.0 kg ha 1 greater than the control and similar to two 4 fertilizer applications totaling 50 kg P ha 1 . 5 The portion of total P in turf cligpings and runoff attributed to manure (controi amounts 6 were subtracted) was only 2.8 to 3.8% of P applied during both monitoring periods (Table 6). 7 Comparable percentages of P in pouitry litter applications were collected in runoff during four 8 simulated rain events on a 5% slope of perennial grass (Edwards and Daniel, 1994). The small 9 amounts collected in clippings and runoff and extracted from soil (Table 2) indicate most of the 10 P in applied manure remained on or in soil and available for harvest with sod. 11 Similar to DP, NO losses in runoff differed significantly (P=0.05) between rates and 12 between manwe and fertilizer sources during 1999. During eight rain events of both monitoring 13 periods, 3.9 kg ha' more NOs-N was lost in runoff from the higher manure rate (two applications 14 of 267 kg N ha �) than from the control. The NOs-N losses in runoff of the larger manure rate 15 were 2 times greater than the lower manure rate and treatments fertilized with 200 kg N as 16 [NHa]zSOa. 17 Losses of NT3 in runoff soon after N applications revealed an advantage of manure 18 over fertilizer applications on tur£ The largest loss comprised 10.3 kg NHa N ha 1 in runoff 3 d 19 after 50 kg N was applied as [NH in 1999. The total NH4-N losses in runoff during two 20 rain events following N fertilizer applications were 2.9 times greater than total NO3-N losses 21 from the larger manure rate during all eight rain events in 1998 and 1999. The NHa-N losses 22 made up 40 to 42 % of total N amounts in clippings and runoff (Table 6). Similar to the NHa-N 23 loss from fertilizer, DP losses in runoff above those of the control were 2.7 times greater for the 24 50-kg rate of fertilizer-P than for the 100-kg rate of manure-P during the first rain event in 1999. � 14 Ol �111�- 1 An advantage of turf in a system for eacporting manure P and N was evident in negligible 2 losses of particulate forms of P and N after surface application of composted manure. Caiculated 3 total losses of PP and TKN after manure or fertilizer applications during the eight rain events in 4 1998 and 1999 did not differ from the control (Table 7). In addition, amounts of PP and TKN in 5 runoff decreased significanfly (P =0.05} in both years after the first rainfall event (Table 7). 6 Reductions in PP and TKN after the initial runoff event of each monitoring period could be 7 attributed to increases in turfgrass plant density over time (Linde and Watschke, 1997, McLeod 8 and Hegg, 1984). The density ratings and clipping dry weights (Table 1) indicate additions ofN 9 fertilizer with manure P could increase plant density and minimize losses of particulate forms of 10 P from turf. Yet, large runoff losses of fertilizer N compared to manure alone could be 11 problematic (Table 6). 12 CONCLUSIONS 13 The slow release of P and I3 from composted manure can limit turf growth and 14 quality compared to timely applications of soluble fertilizers. Yet, the slow release of manure P 15 and N resulted in smaller DP, NOs-N, and NHa-N concentrations in runoffthan fertilizer P and N 16 during rain events after both were applied. At equal P rates, runoff losses of DP attributed to a 17 recent application of P was 58% less for manure than for fertilizer P. Similariy, runoff losses of 18 DP totaled over eight rain events were A4% less for manure than for fertilizer applied at equal P 19 rates. Applications of N fertilizer with manure could increase turf quality and P and N amounts 20 in clippings, but timing of applications in relation to rain events will be critical to prevent lazge 21 runoff losses of N on steep slopes. 22 The surface application manure on turf optimizes potential removal and export of excess 23 P and N during harvest of the sod layer. One disadvantage of the large manure rates was evident 24 in the relatively large DP concentrations and losses observed in runoff, which could raise ,x , 15 0�-��1Y 1 concentrations of DP and accelerate eutrophication in surface waters (Daniel et al., 1998). The 2 observations of runoff losses on the steep slope did represent a worst-case situarion for turfgrass 3 sod production, but it is clear than manure rates need to be managed on a site-specific basis to 4 prevent edge-of-field losses of P and N in runoff. 5 REFERENCES 6 Angle, 7.S. 1994. Sewage sludge compost for estabiishmern and maintenance of turfgrass. p. 45- 7 52, In Anne R. Leslie (ed), Handbook of integrated pest management for turf and 8 omamentals. Lewis Publishers, Boca Rotan. 9 Austin, N.R., J.B. Prendergast, and M.D. Collins. 1996. Phosphorous losses in irrigation runoff � 10 from fertilized pasture. J. Environmental Quality. 25:63-68. 11 Daniei, T.C., A.N. Sharpley, and 7.L. Leymunyon. 1998. Agricukural phosphorous and 12 eutrophication: A symposium overview. J. Environmental Quality. 27251-257. 13 Dorich, R.A., and D.W. Nelson. 1984. Evaluation of manual cadmium reduction methods for 14 determination of nitrate in potassium chloride extracts of soils. Soil Sci. Soc. Am. J. 15 48:72-75. 16 Dorich, R.A., and D.W. Nelson. 1983. Direct colorometric measurement of ammonium in 17 potassium chloride extracts of soil. Soil Sci. Soc. Am. J. 47:833-836. 18 Edwazds, D.R, and T.C. Daniel. 1994. Quality of runoff from fescuegrass plots treated with ✓ 19 poultry litter and inorganic fertilizer. J. Environ. Qual. 23:579-584. 20 Feagley, SB., and M.S. Valdez, and W.H. Hudnall. 1994. Papermill sludge, phosphorous, 21 potassium, and lime e£fect on clover grown on a mine soil. 7. Environmental Quality. 22 23:759-765. 23 Griffith, E.N. 2000. Export of manure sources of phosphorus and nitrogen throu�h turfgrass sod_ 24 M.S. Thesis. Texas A&M University, College Station, Texas. 43 pages. F r. 16 0�-Il I� 1 Gross, C.M., J.S. Angle, RL. Hill, and M.S. Welterlen. 1991. Runoff and sediment losses from � 2 tall fescue under simulated rainfall. 7. Environmental Quality. 20:604-607. 3 Hons, F_M., L.A_ Lazson-Vollmer, and MA. Locke. 1990. NH 4 phosphorous as a soil test procedure. Soil Sci. 149249-256. 5 Isaac, RA., and J.B. Jones, 7r. 1970. Auto-analysis for the analysis of soil and plant rissue 6 extracts. P. 57-64. In Advances in Automated A.nalysis, Technicon Congr. Proc., 7 Technicon Corp., Tanytown, N.Y. 8 7ohnson, A.F., D.M. Vietor, F.M. Rouquette, 7r., V.A. Haby, and M,L. Wolfe. 1445. Estimating 9 probabilities of nitrogen and phosphorus loss from animal waste application. P. 411-418, 10 In K. Steele (ed), Animal waste and the land-water interface. Lewis Publishers, Boca 11 Raton. 12 Kingery, W.L., C.W. Wood, D.P. DeLaney, J.C. Williams, and G.L. Mullins. 1994. Impact of �/ 13 long-term land application of broiler litter on environmentaliy related soil properties. J. 14 Environmental Quality. 23:139-147. 15 Klausner, S.D., V.R. Kanneganti, and D.R. Bouldin. 1994. An approach for estimating a decay 16 series for organic nitrogen in animal manure. Agron. J. 86:897-903. 17 Linde, D.T., and T.L. Watschke. 1997. Nutrients and sediment in runoff from creeping � 18 bentgrass and perennial ryegrass turfs. J. Environmental Quality. 26:1248-1254. 19 Linde, D.T., T.L. Watschke, A.R. 7arrett, J.A. Borger. 1995. Surface mnoff assessment from ✓ e 1 � �- 20 creeping bentgrass and perennial ryegrass turfs. J. En�ironmental Quality. 87:176-182. 21 Lund, Z.F. and B.D. Doss. 1980. Coastal bermudagrass yield and soil properties as affected by 22 surface-applied dairy manure and its residue. 7. Environmental Quality. 9:157-162. 23 McLeod, R.V. and R.O. Hegg. 1984. Pasture runoffwater quality fromapplication of inorganic . 24 and organic nitrogen sources. 3. Environmental Quality. li:122-126. n, � 1� o1-11t�- 1 Parkinsoq 7.A, and S.E. Allen. 1975. A wet o�cidation procedure for detemunation of nitrogen 2 and mineral nutrients in bioloaical material. Comm. Soii Sci. and Plant Anal. 6:1-11. 3 Romkens, J.M.M., D.W, Nelson, and 7.V. Mannering. 1973. Nitrogen and phosphorous 4 composition of surface runoff as affected by tillage method. J. Environmental Quality. 5 2292-295. 6 Schuman, G.E., R.E.Burwell, R.F. Piest, and R.G. Spomer. 1973. Nitrogen losses in surFace 7 runoff from agricultural watersheds on Missouri Valley Loess. 8 J. Environmental Quality. 2:299-302. 9 SAS Iastitute. 1988. SAS/STAT user's guide: Statistics, version 6.03 ed. SAS Institute, Cary, 10 N.C. 11 Terman, G.L. 1979. Volatilization losses of nitrogen as ammonia from surface-applied fertilizers, 12 organic amendments, and crop residues. Advances in Agronomy 31: 189-222. 13 Vitosh, M.L., J.F. Davis, and B.D. Knezek 1973. Long-term effects of manure, fertilizer, and 14 plow depth on chemical properties of soils and nutrient movement in a monoculture corn 15 system. J. Environmental Quality. 2:296-299. 16 Y 4 6�-t��� y 'O O �L d ' M C �i O a+ O � � � � .-+ � G C. � � L d > O 'Q ... O O c N C 0 L d U c 0 V C d .; � C. � a "a A W A �c R • U � _ Q y b0 G Ri 1 � � i Gi.n� i 0 F C.1 � H b0 G . 4 � ; ; 1 �i C � 00 rl � M 7 � M 7 i .,� b0 � �I .�• oo r '» o M M M M M i � ��C ...r 1(� �O M 7 � � T � � � � i N M � N C vi �o �o Vi vi M 00 �O C1 N �A N N 7 �A 00 O� �O N O 7 N V� 7 N � ^� � � � N N V�] � � i � d < i QI L d d � � '� O y � U W � �� o a � L �%J Vl UI 'L^ w C G .. � �3 x � s � � Y � � z x N � � u i, � � O � � Q O •= U � O � L V h 'K N i � �% .� . V � L K II ` w � 7 � C C V Q � L d � � � � � ^^ i # V � U O L 4 O i a� p "�S a�i V ./+ sC � .,I�.�' h � i ��' � � � �" � x • � c° 3 >, K II � o� a a > d 'O b. � i.+ C � C c� p Q Q U R C. C d °i � 3 �� � d z�. U L ~ i! O � d � y ..' ¢ s, ..�.� Il p * p � ,� II .,^. p � O ic x � � t � v i �' � i V1 Qw O ""i '� Q„ y�, 'C �j� d d y � 6� Li �s, � � 3 ri y d 0 L y O � = U O ... c cC .> +' v v' iA � O y � X p � � � � y � R ..- C 'C 'r �y y R a r R� C dD CC y � �a d r. w n K a• c W i •"� h d �a (in i � •� i Fi �C � 'C � 3 �D � 3 w v�i -.'i'+ y w � � � 7 � +..� > 3 � . ..� K K a� r 'O d > Ti CC � � Q� � � � % R Q� a' E; C. G�i �' � �e w � a� � k � � ° .?'s � � °...' a � a �, = � i L O'3 r�i� 6� � 8� O �y �..� E"� � A �p 'X V1 wc �a+ � v �/. 'K^ K ti �, � 1\ 1 }� � d L � � � � U K o � b0 � C � G C � • y Q � � � .� Y � d � d � "a C .� £ � :.+ � C � O O £. w d � +.+ Q L � r. R: �' N C N � O � .� a+ � � a � �. U O 4 � w' 0. pD s�, r a +.+ a •O � L'.. N CJ U :-. O w £ a � � L p r C � R' i Z i y � Y � _ � � � N �C '�^ � d K s. � � d � F� a .. � H � !'r� !"I .�-i N N � � � � � rl �i 00 F, � N � � i i i � � � � i � � i � � � � 1 + .r L bD � N V�J a, x t. d O � L �. v y r" i O c� w � � � � Y�1 .-i N � � N N �' � � 1�l1 O .-� � L '3 C R' � y � h .ri 'X ri w * x x ,*� x .Y �' y .K ,�. o' x ri' C C y N H � � �i H N y ri {�i Zi d � r. L y e " o `� ;; a Q � ' L �a � '} � '3 U1 CSr V� I� d 7 � � � � `O a C y O L �..� � � � O � O ':, O L G:: C U R �C. �-1 R� O L � � � G O L p da L ,� � R G �„� � c U d .��' �+ p y Q �M E � U V � r, +� � d a �a, � a x � t ,.�+ � � �. F. � a � �- � 1 .� • _l. � � �,,, � � a � c r, � a '-I � � o � � r � ta � r'�.i '� 1 N 7 7 O� � � M �-+ N tV � � � 7 A � ai ' � � � 7 a � O�0 7 r N o, o � ,.r .-� ,� � � � � L � �. � � �.y w U � N M � � a i R � a „ , £ � M N 7 � H � w � y a = � N � N � � � ' +-' � r � �o 0 o A V eV O tV M C � fY. 6 � N in r e� � '3 y � �O N VJ DO M � ; e�+ .�-� �-1 � � � a i .. ,_ � �"' N � �--I V1 d £ e}' �-+ V� N p A i � d i E 'O O N � f H i � r � � w d � � o W �.� � Q� C� U A A � � V1 O N * � Q � � a c � � � on > c :r CC U �i R 'C y d L � w+ ' .j � O �° p O � � C y � � � �� �o d F° — � o � � O C Py % Vl 3.' p 'p'� _ �i ,�j,. � � •� W+ f.' a � U a � y t;, � L U � C c d �,� O p So�„ � C�C i � U ` R G' m 4 y,� v �. C.^ L: w � � p �*.. K L � x e� y �,,, C s�. +°�,' C >' c y x" o '�' > A = �" �^ vs P, W .a n.,,, 0 5 — \� \'+� � ri �i O 'G C'i � � C .� i. 3 W i d 4� 4 � O C i d w R p� 3 � e a; y a C � � i � .� o` G � d U V C R U � P�V � 'C a v O > •� h � �, c A °D C � 'i. Y d � C'' O H � � � � H � � � rl t � O� N t� �O lV N t7 'Q V er M C� 7 .ti �i ri .-i N R � � i�O N M C h .��i � N M 7 � �C m � � ' N � � M V � � N N M M �(. a bC �i' c�'i i c� a a v: ci H i ti f`1 N M h i m � �` ; v: '+ 'r t : ao e�-i � ey e�i ti ey N c, C� � H � x c. c+ rl 1 .; m r � � . Y r �o o v+ a? � �G C h C1 e-1 C�f ao a v v a ti ti tV 1V 'S � N V�i h O ey U O � C � U W � � . t x s .x �k x K X ; � i % • i x � � x • x + + x � * x • �' C � a : � � s x * W � � L = x • x « � x K # i . « � : . � x K � � � Y ,. L � L � ri C1. r d � Y Y �i d i "3 C � � � O a) > r. O v 7 q O y " i L d '-I N p � O � O � � y � R � ri � � �Q O R V� CC O H � e A; i y 0. � d � C � o � � �i d Q Q •� � � •y y C 1 �L .� � � C. � � C � O � V � �' R' + � vs u a�i � .�" � '^- x � y a+ Y� L Q �% E"� a+ +i �n %� � " • .4 b1-V\��- a 0 .� d a b� C . O w .� � � bD G .� � "'a w fJ d d w C. .� L L � w 6. O � O O O i C C O � G i w C d U L' O U Q� v G1 L ti z G R � � V � � ti h .__ � c� � � x a c� .�r x x r � e� o .-7 .+ o .-i .-i v, v� o e� O O O O C a � � I� N M V � ti C N N O ti i m ; ..^' 1 � � b d' � rl j C H 'rl C C a bL � c� � � 7 .y N N ef H � p ti eti C C i � � � N 'eT M M �'1 .ti � C N tV C O � � c� ti x a .ti �O M N 00 �p o v .-i Q .-i '. � M M N t7 C C o C C � i t� S t � � � O C M � z 'Y 'r ey '-1 N L h y L +" ' � ^j = • i � v w � y * � 2 % * * d y K * q X i' y � � G � � � e C.� • � � • s ,* x x x • • � . • � � y � C G a U d U L i L � z � � � � •� � a on '" � d i �+ ° � •� L O '-i �T. p U � d L' L K w ri C O O O i+ � V O � � � w , � r. '3 Li C R. � � N C C bD O 'y L � O � i CC N � ... a+ � ^ � � O O �� � � N 'X � x *� � z x� r ,, 4 � \ �\��}� L I OI N � C �L � � w L � « N N i OI � � � � L N m � O w O C 7 L W 0 N C O U l�0 w d Y � 7 Ci F. A C 'a C R � > 0 N N � C � R N � C . G V C � w t� Q j d � O � V � a � a R C pp R � Z � R C F +" C � � N N Q C R F � x N N J � e C d V d a N y N /0 y � L C t3 � Y R z m N 3 O a. L y Q R O J ' L O1 Q. w Y O H R r L R � � Y X N N O J � e C d i a N 41 R J L Y � d Y Z N N � O N J � � r Y � O m ~ � N � � O �„ R Z R' L +�-� Y � d L 3 O � Q ? tn a0 a0 6i o� N M � r � W � � c0 N I� � � � ti r � � � � � O O O O � O O N c'1 7 0 N � � � N � � N � � � � u'i N N I� N N N rn N � �1 N N 0 0 0 � c N N N O N O y U tL d 3 C N � O � v o � a � �o. U � �" . � R t6 � c 6f � C � O � � C � •L � L W � � C � � � a � N y C O C O � � .� O. y c6 V _ � C� O Q Z U 'Q m w a O � N C V � N N O � ."' C � N U C E O_ U • Q •O - O � �p_ Q N N � � N N N +- 'C p_ O O . N � � C 3 p N N O O - ' w � C C 3 3 � _ � C C � � � O O � . Q a Q (� � Z a �J « m . r-�... 0�-\�\s- Table 7. Total P and total Kjeldahl nitrogen (TKN) of particulate fraction of runoff for four rainfall events during monitoripg periods in each 1998 and 1999. Runoff was collected from Yreatments comprising two rates of either composted dairy manure or inorganic fertilizer before runoff monitoring began. 1998 1999 Event Total P TKN Total P TKN ----------- �g Plot 1 ---- ---------- mg plo£I ----- � 57 B 14 C 28 D 41 # MSDo.os 15 265 46 73 152 43 62 34 37 24 16 349 165 180 124 89 f Miniumum significant difference within columns using Tukey's Studentized Range, P=0.05. `' — _ Nutrients and Sediment in Runoff from Creeping Bentgrass and Perennial R` Turfs Douelas T. Linde" and Thomas L. �Vatschke ABSTRACT Althoueh scientist> fia}�e found 4+tde tcanspoct of nutrienu to date in runofF (rum turF_rasses. more research is needed on a x�ider range oEsuil conditions and mans�ement scenar�os. Thii studp ��as designed to as>e>s nutrient and sediment tr�nsport from creepin� bentarass (4grostis pa(usiris HudsJ and perennial rpegrrss (Lolium perersne LJ turG and to asse>s the influence that .�ertical mo..ing had on sediment tmnsport Sloped p�ots of ben[�rass and n e�mss. maintained simitnr to a golf f�irr�a�, ��cre ircisated to forte mnoff Cor the venera- tion oC mnoPF and leachatc �+�ater samples. About 1? h before each runoff e.�ent, iaigation •�as used ro equilibza�e soil moisture for ail plo[s. Foc four erents, pbts ���ere tre�red xith (crtilizer ai a rate of }.9 g N m'. 0.3 g P m and 4.1 g iC m'' about 4 h aFter pre-erent �r��gstion and S h before runotF. For another Four e�cnts, plots »�ere verticut 6 h bePure runoll- «�ter ssmples »ere anskzed fur NO tot�l I:je�dahl-� (T6ti), ph�sphate, and sediment. p]ean \0-�: wn- centrations rarelr ecceeAed 1 mg L Phosphete and TF1 concentra• tions and losses sieniticnnd} incre�sed �+hen runoll �+�s Forced 8 h after fertilization. On a�ersee fur these c��ents, ll % oC �pp��ed P and 2 applied ti»as delectvd in runull and IS % applfed P and 3 applied A' ��'as detected in Ie�chate. Fur all other e�ents nutrient concentrations and los>es e'ere consistentl} lo��er. ��ertical mowine h�d little afFect on sediment transporL 5ediment transport Crom both turfs a�eraged O.S k� ha On golF Eairwu��s, oll-site mo�'ement oF nutrients mar happen if rvnoff occurs soun aEter granular fertilizer is applied tu a neady saturated soil. I � THE RESEAFCH CO d3i2, S :izntisti hace found that nutrient transporc in runoft from turfgrass is small (blortor. et aL. 19SS; Gross et al.. 1990: Harrison et ai., 1993: Linde et al.. 199-'.)• Ho«'e�'er. additional research is needzd on a l��ider ran�e of soil conditions and man- a�ement scenarios beEore �ny generalities can bz made about surtace transporc of nutrients from golf courses. Harrison et al. (1993) found that conczntra[ions of NO ti and phosphatz in runoff and leachatz from turf maincained like a home lawn rarely excreded � and 2 mQ L respectively. Using crzzpme bent�rass and perennial ry'e�rass turG maintainzd Iike a QoIE fainv'ay. Linde et aL (1994) rzported that concentrations of \p—\, phosphate, and TK� in runotif and leachate rarely exceeded 7, �, and 2 mo L respecti�'aly. In fact. nutrient concentrations and losses (presentzd as loading rates) usually reflected those Eound in the «'ater used for irri�ation. For rcnoff ecents wichin ?=� h follo«�ing ierti(ization. the runoft contaiaed an avzrage of 1% oE the applied � and 3?% of thz apptied P for both turfs. The leachate contained an averagz of � 2% of applizd iv and 1�3°io of appliz:l P for both turfi. Linde et al. (199-i) concluded that for similar condi[ions on a eolf D.T. Lindz. Dep. oC A?-�onomc and Emiron. Scizn�'.. Dtlawara �'af- Icc Cufk?e. l0U E. Buder Ave.. Doy(zsto��a PA 1S90t: and T.L. \\�atichk:. Dep. o[ Ao•000my, Prnnsyl�a�ia S:ata Univ.. LL6 ASI Bld,_ (:ni��ervtc Park. PA 1650?. Receivcd I1 Oct. 1996. *Cor.z- sponding au[hor. � Publi>hzd in J. Enciron Qual. 26:12�1S-t?i (1997). 0 � -�\\d— fair«'ay�, it «ould be reasonable to assume that little off- site tran�port of nutrients from the fair�ca�' ��ould occur as a result of fertilization. Linde ec aL (199<} had ro limit s[atistical anat}'sis to individual dates bzcause major soil moisture difYerences esisted bzn�zzn e�ents and be- t«•ezn turf species. In the Simixed studies on erosion from turfgrasszs, scientists have found that turfgrasses �rzatic rzducz ero- sion compared to bare soil (�Vauchop� et a1..1990; Gross et a1.,1990,1991). Usin� a mised stand of bermuda�rass (Cynodon dacrylon L. Pers.) and bahiaarasi (Paspalum notnnun Flug�e vac. suare Parodi). «'auchope et al. (1990) found that thz a��era�z soil loss for simutated rainfalts a[ 69 mm h was 28 k¢ ha ` for bare p(ots and 3 k� ha ' for grassed plots. Using slopzd plots of tall fescue (Fesu�ca arundinacea SchrebJ. Gross et al. (1991) reportzd the a� erage soil loss for a 30-min, 120-mm h intensity scorm was 519 k� ha for bare soil and 54 kg ha`` for mature tall fzscue seeded at 4SS ka ha"'. Gross et al. (1991) concluded that even lo«' dznsity turf stands could si�nificantly reduce erosion and a«'zll-maintainzd stand should not bz a si�nificant sourcz of sediment. The studias conducted by `h'auchopz et al. (1990) and Gross et al. (1990,1991) used turfs maintained at heights >S cm. No published studies wzre found that included inEocmation concernin� soil loss from turfs maintained similar to a �olf fair« ay (about 13 cm heieht). In addi- tion, no studies were found that pro�ided inEormation on the influence that vertica! mowing for thatch manaee- ment had on soil loss from turfgras�. Since �'ertical mo«'- ine for thatcn management typically resul[s in the physi- cal removal of oraanic matcer and shatlo«' grooves in the suil. it is possiblz that soi( loss mac increasz. Crezping bentgrass and perennial rcearass are t«•o turf�rasses commonly used for �olf couriz fairways in the tamperate ciimate regions of thz li.S. Perennia! ry'egrass is a medium-tezturzd, bunch-ttipz species that does not form a dzfinite thatch layzr. ��'hen closel}' mowed, perennial rye�rass forms a turf «'«h a shoot densicy' bzt��'een 100 to 200 shoots dm (Beard, 1973). Crzepin� bent�rass is a fine-tzxtured, sroloniferous spe- cies that forms a definite thatch la}�zr. �Vhen closely mowed, creeping bentgrass forms a turf �cith a shoot densin >200 shoots dm '- (Bzard, 19i_). 7'he objec[ives of this rzsearch «'zre (i) to assess the transport of KO N, TKN, phosphace, and szdiment from the turfs and (ii) ro determine the influencz that vertical mo«�ino for thatch mana�ement had on sedi- ment transport from the turfs. DiETHODS AND DIATERI�LS Sic established tucf runoEE ptots ustd b}' Lir.de et al. (199-' anS 199�) werz used for this smdy. In 1991, tF.ree piots (each 6� m«•idz by 19 m lona) «'erz establishzd ro'Pznneaslz' crzzpin� bentgrass and thrzz to a perennial r}'e�rass blznd izas n � _���� LI�DE & w��.TSCHKE: RC�OFF FRO>t SE�"IGRASS A�D RY'EGRASS Tl'RFS 1249 ('Cita:ion II'. 'Commar.dzr'. 'Ome�a II'). Hacrison et al. .-eTticat mo�cing. Flots rzce:� thz usual pre-even: irri�ation (1993} characczrized thz sue as having a•�ariablz slopz be- to equilibrate soil moisturz. It µ�as hcpothesized that the rz- na-ezn plots (9-il% 2ad a sudace so�t that «"as a se�'erzlv bl�dzs. aad the oriznt tion of [he� b�omn �h erodzd Hzazrsto�.�n serizs classitizd as a clay (0.23 kg ke . sand. 0.36 kg ko ' siic. 0.41 ka i:g '��a�). The local seologp slope ticould form preferen.ixi f!o�v channels for runoff and �cas a frac[urzd karst and depth to bedrock ranged from � to se Rur.oif anater san pl s�iere tak n an �'fodei 60 cm. At the bzginains of this stud}' in 199�, the top S em _ �Q� orjable «zrer samplzr (ISCO. Ine.. Lincotn. �E). To of soil had a pH of 7.1. P iz��el of Si ks p ha '. and K lecel F of 26"I kg K ha Plots «'e:e mowed ro 13 mm ���ich dippin�s eser2et a runoff samptz. ihz potyeih}'lenz samplin� access tu z tur�f thac � ould be ca112 found on a olf fa�nat}iY The sampler���as intzrfaced �nith an ISCO p4odel �0 flo� Each ploi containzd 21 R'zathermatic (Garland. TX) pop- mecer chat «�as programmzd so that after zcery i5.6 L of p onvas an �coaced con e t [e he e bo lha m collecez a r�u�oa mL sa ple of runoff «ater in thz splitiing chamber. Thic and direc�eand sam equipmzna \Vatzr the chute r�� utes a d an� a��era z unoffb olamu for� turE pre- mzasuring P floa'ed into a polyethylene splitcin� chambu (for runoff sub- `i�iz eom tzd int c� o�00-mL(boulzs. The 40 mLsamples samplz cotlzction) znd in[o a parti[ionzd steel tank (Harrison from thz f rst 907 L of runoff �verz composited into the firsc et al., 1993). FunofE �olume and rates �tzrz mzasurzd wing ' an I 0 dates ` from J nz 1994 totOctOoberC199?��runoff was bhe co d�botde. Nhen r�u off �aas L hzne ach botde forced w�ich irri�ation at a rate of 139 mm h to generate contained �SO mL of w'ater. Samplin� «'as stoppzd after the thesz irr �atzd e��ents Pe�e cor.ductz [ appco�imate]}'. e e�z Y L, lhz�n zcond concained �SO mLnRunoff aused Z�ck, dzpzndino on �i'eather and availabiliq� of laboc Durin� by rains[orn:s �cas measurzd and samplzd �cith the equipment to�a[natual`siatz ethreoard alzd99�)bzcauseinsthoe,studesrunoffflo�.'ratescausedby' <10 n�pla Based�on`data from aL (199�)�irrigation In lsfl�rms d de ot and samplers forthis� smdy �ere 1[m¢O duration ��'as set at 2� min for bent�rass plots and 1� min to onz. for ryzgrass plots for most ecznts. The purpose of different L,zacha[e water ���as sampled from four pan-lysimeter cotumesfor tYle[toOluiSSSlOC2 tat2LS847}Jlltt�rwaSElO�SQa ea fabz of each (Hahri on et Abo u I> after Fur the 6 and 24 Szpt. 199� events, both turfs �i�ere irrigatzd for�each plo[ t �Ppakin,, q ai an oun[s from of tl�ie for 2� min each. four sam lers. Approzimaeely 12 h prior to each runoff ecen[. the pl�ots "�a�z,.pamplzs «'zre analyzed for NO,-�. TKN, and phos- .eere irri�ated ��'ich a senes of shocc duracion (2-3 mia) irri�a- phate (orihophosphatz) aceordin� ro the procedures described tion sets at a ratz of li9 mm h"' un[il runoff was �'isua4ly one o four depe d b g n thz antzcedeen soil mo sture co t nt ru off arerz uszd to�cal ulat nutnent t o r s s tion runofa for The purpose of thz pr� zcent irrieaciors H•a� to equilibrate acerage of bo[h mrfs. Total runoff from each piot and thz the soii moisture content for all plo[s so that data comparisons ave.age of the wncentr2cions in the first and szmad flon'- bet�iezn dates could bz made. Lindz et aL (1994) did not paced zunoff samptes w'erz used ro calculate losses. The �ol- equilib[a[e soil moismre bzforz irriga[zd ecents, thus they had ume of «�ater that could be held in one subsurface sampler � � 1 L w�as used ro catculatz nutrient loss in leachate. Sedimznt concentrations of the runoff samples �vere detzo- ro limie comparisons to individual dates becausz major soil (-� moistute differences esisred between dates and betwezn [urfs. mined �racimetrically by measurin� the amount of inorgamc For four oE the runotf z�'znts, ptoes a�zre fertilizzd w�ih a �atzrial tra ed by' filtec papet (1 µm diam. pores) aftzr 19-1.3-li.S (N-P-K) fertitizzr (O.�l. Scott & Sons.:�larysville, Fp . filterine the enure kno«�n samplz colume. Filters «'zre place OH) usin, a broadcas[ spreader at the ratz of 4.9 g N m - � p_, o p m and �7.1 g K m about 3 h afcer prz-e��en[ irngation in a o�en at 42�`C for S h and then wzighed. It «as hypo� z 2nd S h prior to thz runoft event. The f:rtilizer contained sized thai the soil dismrbance caused by �'ertical mo«"�no � 0.6°rb NH,-�'. 15 �%6 u;ea-N coated to procide 73 slou �i'ould tikzh� incrzase szdimznt trxnsport shortly after mo«�n° ' relzase �. P dericzd from monoammoniem phosphate. and and thzn decrzase as the rurf reco�'°-rzd. Treamtents (turf species) «erz arranazd in a random�zzd ' K from Ii Beforz fertilizer applica[ions. rmtoff collzction cum letz bluck dzsign ��'ich [hrze replieacions and blocking ' .ezirs locatzd at t4e boctom of each plot «ere co.�zrzd «'i:h bas d on suriacz siope- In a concurrent runoff study b}' Lindz ' plastic to psecen[ =ranules from enterina an}" part ot the weir. 1996 . i: «as dztermined that the pre-e�'znt irrigation procz- dure �cas zffecticz in equi(ibrating the soit moismre contznt ' On sz�"en othzr datzs, supplzmental maintznance N ti�'as ap- � plizd as a liquid or aranular application of urea (46�-f��) a� {or all plots beforz runoft. Thz}' found tha[ [hz a�'eraee soii � 2.4 or �J g� m'. ' columzvic ��'ater cuntznt of the ploU ius[ prior to irrigatzd ' Abeu: 6 h prior to another four e��znts. all plots w'zre czrti- 0 0' in 199?. Thus. ' cuc once using a R}an >tata�cay ��ereical mo��zr \4ode154�3� 3 I`� s�`z` �o m ' arisons�bzc�' 9 een z''a OS °°� izn[ and sedimznc � (Cushman. Inc. Lincoln. �E) [hat had 20. 1.6-mm-��'ide blades p i blades� pznzt atzd the soii appro�an2tel�d3Smm.aPlots az usins a mzasures analbs seofbariance bacau<z � �eere �'erticut Izngth�rise do«n the stope. Thz majority of the rzpzated measurements ��'erz made on thz samz experimznta ' ,:� J�b� � !�: `_,,.. ;n 8 =a�(�rc hv .+- mo.rino �vas units o�zr time. ,� , factor. O ecperiment �va<_ : �;� �,�, t��c .� �_ ni-,., „F t�mr � eolleC[ad by rakin� and ��ei�hed. A�;}>;����°.�lci�� ti Ii boio�� i j7_jO 1. E�VIRON. QUAL. vOL.?6. SEP'i'EMBER-OCiOBER 1997 Tabie 1. �lean n co ncentrations for 199 irrigated e � tiitrnte-.S P h osp hate R unoEF sample+ Runoff sa mple Date FPl FP_' Leacha[e FPl FP? Leachate 29 June 13 ]uh ' p� ?? .�.u�.4 3 Sept. ?0 Sept.7{ S Oct. se 1.0 : \S 0 03 \S 03 \S 0.1 \S OZ 0.1 0.6 O \5 0 C �J 0.3 �5 0.1 r5 0 \5 0.? U.t OS \j 0 \S 0 !�5 OS \S 03 \S 0 \S 03 0.? 1.67 2.�1 \S 2.71 1YS 4.13 1.90 3.85 0.89 0.p'_ '** "" Significant at the 0.05, 0.01, and 0.001 probability� levels, respectivei�'. i FPl = lst �u�s-paced runotT sample, FP2 =?nd itow-paced mnof5 sample. - Mean comparisons are for adjacent d�tes hithin columns. §\S = not significant; SE = standard error. Q Fectilized 8 h beFore e�'eat at �ate of 4.9-03-1.1 g m' oE N-P-K. arran�zd as a spfit-block in a randomized completz block desi�n accordin� to 5[zzi and Torrie (1950). Usin� the Huynh- Feldt epsilon valuzs calculatzd by� thz repeatzd statzment in SAS's �eneraV linear model proczdure (SAS Inst,1990}, Linde (1996) dztermined that levet� of time were indepzndent for the runofi data. Sincz levzts of time were independen[, [hen [hz Ftasts of thz usua( sp(it block analysis werz valid and a univariate analyss w'a> conductzd that provided information on the Izast square mzans and thz probabilities associated �cith thzir compa;ison. Thz prz-plannzd comparisons for this study included comparing spzci;s within days and comparin� conszcutive days within spzcies for thz variables neasurzd. RESULTS AND DISCUSSION Nutrient Transpod Mean nutrienc concentrations, bv evznt, sample type, and }ear are przsenizd in Tablas 1 ar.d 2. The concentza- �������� Total Iijeldahi-ti R un o EE sample FPI FP'_ Leachste mg L I.li O.T 0.0? 0 0 • \53 \S \S � Z.Ol Lli 0.03 O.L' 0.19 �S \S �S �5 1.76 1.13 0.39 0.13 0.0�4 iER 1y ids '� 4.17 1.95 b.3-! � 90 3.93 �}a #Ri s Ysi x 151 0.83 130 OSl 0.86 a. t:. :<s 2.66 ?.4i 5.78 2.53 ?.30 . war 0.61 OAI 0.0? 0 0.06 0.?? O.tS 033 0.13 0.10 tions reported are those detectzd in thz samples minus the concentration found in the irrigation or rain watec for each date, thus they represent the nutrient contribu- tion from the turf plot alone. No si�nificant turf specizs effects werz found on any date for any sample type; [herefore, values for both species ���ere averaged. Sig�ifi- cant date efiects, ho��ever, were found and are pre- sented as comparisons bzt��'een ad}acent dates in Tables 1 and 2. Nutrient concentrations, dzpth of water applied and depth of runoff (Table 3) wzre used to calculatz nutrient inputs and losses. \lean NO concentrations w'ere alwags found to be lowzr than the 10 m� L drinking watzr standard set by the USEPA. This findin� concurs �rith NO concentrations found in turf runoff studizs done by Linde et al. (1994), Harrison et al. (1993), Gross et al. (1990), and blorton et aL (1933). The hi�hzst mean Table? >Iean nu[rient concentrations for 199: irrigated e Ciittate-N Phosphate ' Runoff sample` RunotF sample k D at e F FP? Lea<hate FP1 FP2 Leachate m� L 1 16 �Sm 0.6 �1 U:? 1.?7 09J L01 ' - " rS$ x <. . 31 Afar^,I 13 L� 0.1 9.96 7.67 d.?d x. ... �5 x: . . 11J�me 03 0.6 _ 0 253 1J9 1.2? P$ s.� \S \S \S � ?83unt 0? 0.? 0 133 1.60 0.99 \S n \S .-.. ... .., 1? July¶ OS 0.� 0.1 1�.39 5.51 1.9? �s �s ».. >k. ... .,. ?6 Jul. Ob OS 2.6 219 I.85 1S6 ' � \S *"" *• \S °` \5 6 Sept. 0.6 0- 03 ].95 7.30 Y3i i i ti5 r5 �5 NS \5 1 23Sept. OA 01 0 1.i7 0.93 I.Ol � SE 0? 0.03 OS 0?3 0.09 039 "° $ignilicant at the 0.05, 0.01, and OAO( pcobl6ilit� IereB, tespecti�el.. - FPl = lit flo�.-paced mnotF sample, FP2 =?nd fluw-paced mnotT sample. �'• ` Dlean comparisons are for adjucent dstes x[�p(n c0lumns. 1 I. § NS = not s�gniFicant; $E = stsndard error. f ¶ Fertilized S h befoce e�ent at rate of 4.9-03-1.1 g m oF N-P-Ii. ���: Tot Kei d�hi N RunofF sample FP1 FP2 Leachate 0.�8 S.J2 ♦S� 0.�7�1 a.na 553 o �, 1�5 0.3? \5 0=9 0.09 0.20 3%3 031 s 0 3.23 �r< 0.18 �S 0.06 \5 0.?? 0.09 0.�33 2.00 0.83 • 0.0# 2.60 0.25 C�S 0.37 \S 0.?7 0.26 G�-1\\�- Table3. «'ater applied and mean total runoff for bentgrass and n�egrass. Li\DE g w'ATSCHKE: RU�OFF FRO�t BE�TGRASS A\D RYEGRASS TI;RFS {t'alec applied Date Bent Rye 1993 29 ]une 13 3u1y 14 ]vl.i ai ���: , >? ]uir ? Aug: 17 Aug.: 2'_ Aug. 3 Sepi. 17 Septi 20 Sept. 8 Oct. 1 No�:9 at �o..� Z$ tiov.T 1995 16 bfay 31 Dtav 13 ]une ?S June 6 Julyi 12 July 26 Julr 6 Sept. 23 Sept. ZO Oct3 5& 58 �; is � 58 93 58 58 3tS 5g 58 26 ia 49 58 SS Sg >$ Z.8 58 93 cg 58 93 3j 2% is � 35 93 35 35 38 35 35 26 is 49 35 35 3> :s �g 3> 93 53 <g 93 ; RainC�ll e+used runofi. }fean total runolf depth Bent Rte 16.1 1>.8 0.1 0 ?.1 15.1 o.s 79.1 173 0.8 za.a ?OS 0 o.a 3.2 15.5 18.6 23.5 18.7 I.1 21.1 13.2 17.9 16.9 13 135 1L7 I.9 0.6 33 1�? 6.6 13A 30.8 0.3 15.0 L? 0.8 ia 0.7 SL7 11.'_ 1?.9 11.0 0.6 123 9.6 2?1 19.7 3.2 1�0 N concentration was 2.6 m� L that �a'as found in the leachate samples on 26 July 199�. Runoff and leachate losses of NO N�cere consis- tendy lower than irri�ation inputs of NO� N(Tables 4 and 5). For both years, mean losses in runoff, which ranged from 0 to 0.02 g m'-, �vere lo«er than mean losses in leachate, which ranaed from 0 to 0.16 � m Linde et al. (1994) reported similar findings; however, they described nutrient losses as nutrient loadin� rates. For all samptz types, phosphate concentrations si�nif- i2sz icantly increased for e��ents conducted 8 h after fertilizer application (Tables 1 and 2). Concentrations «erz sie- nificantiv less bv thz nest evznt. usuallv conducted within about 2�ck. The highest mzan phosphate conczn- tracion found in the flo���-paced runoff samples �cas 1039 ma L for an e�•ent conducted S h after fertilization on 2I Ju]y 199�. Escludino the events which fertilizer was applied S h prior, phosphatz concentrations �cere similar to those found by Linde ec al. (199�). Theg applied the same rate and source of P. monoammonium phosphate, to the same turf plots used in this study and reportzd 6.06 mg L as the hi�hest mean phosphate conczntration in runoff. ���ith most concentrations <3 mo L Linde et al. (1994) found little indication in runoff or leachate samples thac P fertilizzr had been applied approximately 24 h before an irri�ated e��znt. The hi�her phosphatz concentrations detected for events immzdiately follo«in� fertilization compared to findings by Linde et al. (199�) could be attributed to hieher soil moisture contents as a result of pre-event irrigation used in this study. The soil was likely wetter (near saturation), thus a �reater portion of the soluble monoammonium phosphate fertilizer couid move off- site in the runoff. Thz a�erage soil volumetric water content of the plots just prior to irrigated events was 030 g k� in 1994 and 0.40 g k� in 199� (Linde,1996). Soil moisture levels were likely less in the study by Linde et aL (1994) because they did not use pre-event irrieation to equilibrate moismre ]evels. Based on soil test results for 6 Apr. 1994 and 23 Au�. 199�, soil P levels in the top S cm of soii �vere �enerally in the low range. In 199�, P le��els averaeed S� k� P ha ' and ran�ed from 73 to 11Q k� P ha In 199�, Izvels avera�ed 73 k� P ha and ranged from 36 to 91 k� P ha `. Since soil P levels �cere low, then excessive levels of soil P�rere not the cause of the increased P transport from the plots. Runoff and leachacz losses of phosphate-P were often Table 4. 199A Chronoloa� of inean inputs and mnofl and leachate losses otrotal N(TKN + 1�0,-\) and 10 for bentgrass and q�egrass. Irrigation and t inp uts Runoll losses Leachafe los Fert. BenL R}' e. Bent. Rye. Be nt. R)e. inpu<s Date of'.� \ \Orti \ \Orti N 10r\ 1 \Or\ \ \OrV \ \O�-\ 7 June 3.7 29 June 0.31 0.?8 0.19 57u1y 2A 53 Sutv 0.}i 0.3i 0?0 13Iuk"r 0.03 0.0= 0.03 _'1 Juhi 0.05 0 O.Oi Z' Jul�fi O.OI 0.01 0.03 ? Au�. 0.31 0.38 D.?5 b �ug. 2.3 17.4ug.� 0.01 0.01 0.03 22 au� 9.9 0.?7 D.?7 0.76 3 Sept. 0.?S 0.28 0.17 17 Sept.�= 0.07 0 0.07 ?0 Sept 4.9 0.?6 0.26 016 S OcL 0.?7 0.?7 0.16 9 Oct. ?.4 I \o.�.�: 0.01 0 0.01 21 \ov.i� 0 0 0 23 \ov.>: 0.01 0.01 0.01 T Rainfall caused runoR. _ Leachate samples xere not coliected. 0.17 0:20 0.0? 0 0.01 0.23 0.01 0.16 0.17 0 016 0.16 0 0 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0 0 0 0 0.01 0 0.0? 0 0 0 0 0.01 0 0.0? 0 0 0 0 0 0 0.09 0 0 0 0 0.01 0.01 0.05 0 0 0 0 0 0 0.03 0 0 0.01 0.01 0.0? o.ii o o.io o.oi o.�� 0.01 0 0.01 0 0.01 0 0 0 0 O.11 0 0.06 0 0.03 0.01 0.01 0 0 0.0? 0.01 0.13 o �+_i 0.01 0.05 0 0.09 U.02 0.01 0.03 0 0.01 0.09 0.03 0 0.13 o.aa O.OI 0 0.01 1��� J. ENV(RO�. QUAL. VOL.26. SEPTE?��BER-OCCOBER 1997 r O�-���a- 'I Table 5. 1995 Chro n o ioa,v of in ean inp and runuFf and l eacha te losses of totnl ( TF\ + NOz-ti) and ti Or� for bentpctcs and q�egrass. Icrig a[ion and xainfali input R unoE F l oss e s Leachate lozses Fett- BenL Rpe. Bent. Rre. Be R ve• inputi Date oE\ � �Or� � �Oy� � �O,-� � �Or1 � \Or� � 1��r� gm : 16 Diac 0.?9 0.� O.IS O.li 0.01 0.01 0.01 0 0 0 0 0 22 JIa�' 37 3Illa�� 3.9 0.?J 021 01> 013 0.09 34lune 0?0 0.?0 01? 012 0.03 ?3 June 0?6 0?? 0.15 0.13 0 6luh�; 0.01 U.Ol 0.04 0.01 0 12July 4.9 031 018 01S 017 0-09 26 Suh 635 0.?L 0.1> 013 OAS 31 Aug.?-3 O.L' 0.01 O.L' 0.01 1 Sept. 4.9 6 Sept.II 0.17 0.13 0.17 0.13 0.01 2a SepLn 0.20 O.1S 0.?U O.1S 0 ? Oct. 4.9 20 Oct.'s_ 0.01 0 0.0� 0 0 � R�inCall evused mnol£. _ Leachate ssmples eere no[ collected. § Lightning disabied runoff inea�uring equipment. ¶ Both turFs were init�ted at same duration (?5 min). greater than irri�ation inputs (Tab(es 6 and 7). For both years, mean P losszs randed from 0 to 0.06 g m' for runoff and 0 to 0.07 g m for leachate. Since the hi�hest losses were found for events that had P fzrtilizer applied S h baforz runoff, then a portion of the applied P was transported in runoff and leachate. For esample. on 22 Au�. 199=4, an avera�e of S% of the applied P was de- tected in runoff and 7% in leachatz. On 12 July 199i, an averaoz of 17% of the appli�d P was detected in runoff and 20% in leachate. Linde et aL (199�) found mean P losses (repo.ted as ]oadino ratzc) ran�ed from 0 to 0.01 � m for runoff and 0.01 to 0.04 g m for leachate. Mean TKt�'� concentrations followed a simitar pattern as phosphate. Conczntrations sionificantly increased for events conducted 8 h aftzr ferti(ization (Table 1 and 2). blean TK\ concentrations range3 from 0 to 6.84 m� L'' in 199-4 and 0 to 5.�8 me L in 199�. The higher concentrations werz dzrec[ed in the first flow-paced run- Tabie 6. 1994 Chroaoloe,v of inean inputs and runoff and leachate losses of phospha[e-P for bentgrass and q�e�rass. Grisution and rainF�}� Runoff Leacfiate Fert. p�nput losses oP P los5 af P inpu[c Date o(P Bent Rre BenL R�e. Ben[. Rye. a m . 29 June 0.0? 0.01 0.01 0.01 0.01 0.01 13Juh� 0.03 0.0? 0.01 0.01 0.01 OA2 11]ul��� 0 0 0 0 0.0? OA3 ZI Suiv` 0 0 0 0 0 0.03 3? Juii�: 0 0 0 0.01 0.01 0.03 Z:1u�. 0.0'- 0.01 0.01 0.01 OAl 0.0'_ 17,aug.� 0 0 0 0.01 0.01 O.Dl 32 Aug. 0.3 O.OZ 0.41 0.03 0.02 0.0t 0.03 3 Sept. 0.02 0.01 0.01 0.01 0.01 0.01 17 Sept.:_ 0 0 0 U ?0 Sept 03 0.0? 6.01 0.03 0.01 0.03 0.03 S Oct. U.Oi OAI 0 0 0 0.01 1 �us.`_ �) U 0 0 21 Nu�.*- 0 0 D 0 38 \o�.�_ 0 0 0 0 i Rainfall caused mnOtT. `y Leach�te sampies were not coilected. 0.0? 0.07 0-Ol 0.06 0 0.10 0 0.0'_ 0.01 0 0.0.1 0 0.03 0 0 0 0 0 0 0 0 0 0 0 0.01 0.07 0.01 0.09 0.01 0.1? 0 0.01 0.01 0.01 0.06 0.05 0.17 0.16 0.01 0.0? 0.01 0.06 0.02 0.07 0.02 0 0.01 O.�I 0.01 0 0.01 0 0 0 0 off samples for events immediately followin� fertiliza- tion. The TKY ]evels �cere often higher than levels found by Linde e[ aL (1994). They reported that TK?� concentrations ran�ed from 0 to 3.� mg L Like phos- phate, the higher TIiN levels for this study �rere attrib- uted to hi�hzr soil mois[urz contznts as a result of pre- event irrigation prior to fzrtilizer application. The TKN and NO N results w'ere added together for estimates of total I�'. In both yzars, despite relative increaszs in total N losses for events that fertilizer was applied 8 h before runoff, totat A' losses remained consis- tentty lo«•zr than irrigation inputs (Tables 4 and 5). ln 1994, mean total N losses ran�zd from 0 to 0.11 g N m''- for runoff and 0 to 0.21 g N m for leachate. In 199i, mean losses rangzd from 0 to 0.09 g N m for runoff and 0 to 017 � N m'� for'leachate. Linde et al. (199=!) repocted similar numbers; however, total N losses �vere szldom greater than NO losses. In the currznt studv, totaf \ losses were much hi�her than NO r 2able 7. 1995 Chronolo�y of inean inputs and runoRand leachate losses oF phosphare•P For bentgrass and ryegrass. Irri�ation and cainfni( RunoR losses Leachate Fert. p� oF P losse ot P inputs Dafe ofP Ben1. Rye BenA Rye. BenL Rre. gm' 16 �[a� 0 0 OA1 0 OA1 0.01 31 �ta. 03 0 0 O.Oi 0.03 0.0> 0.06 li June 0 0 0.02 0.01 0.01 0.01 283une 0 0 OAl OAS 0.01 O.OI 6 July�- 0 0 0 0 i? Julr 03 0 0 0.06 0.0� 0.05 0.07 26 Julr 0 0 0.01 0.01 0.02 0.0'_ 314ue.>;§ 0 0 1 Sept. 03 6 Sept11 0 0 0.01 0.01 0.02 0.0' ?15rpt.Q 0 0 0.01 0.01 0.01 0.01 ? Oct. 0.? ?�) Oct.+= 0 0 0 0 i Rainfull caused runoff. - Leachate samples aere not colle<ted. � L�ghtning disnbied runofE me:uuring equipment. �� Both turts were irrivated at same duration ('_5 min). � L1SDE 8 WpTSCHKE: RL'�OFEFRO\f BE�TGRA55 A�D RYEGR:ISS TIiRFS losses for e�'ents that had fzrtilizer applied, thus a por- tion of thz applied �. albeit small. «�as transported in runoff and lzachatz in forms other than \O_ N. For esample, on 22 Aug. 199�. an a��erasz of 2% of the applied \ �cas detzcted in runoff and about 3% in lea- chate. Gross et ai. (1990) also found that ` losses �cere areaczst uhen runoff occurrzd soon after fertilization of a tall fescue!Kentucky blue�rass turf. On 11 datzs durine the study period (�fa;� 199a-Nov. 199�}. dztectable amountc of runoff (>0.6 mm h oc- curred due to rainfall. Runoff caused bv rainfall often did not pro��ide complete data sets because runoff did not ahcays occur on all plots. Thzrefore, mzan nutriznt concentrations for rainfall events «�ere based on an avzr- aoz for both turfs and thz number of plots that provided data. Statistical analysis was not conducted. On 31 Au�. 199�, rainfall produced measurable runoff from each ptot. ho«�e�zr. li�hming disabled the runoff ineasurin� dzr'ices. �4ean nutrient conczntrations in runoff from the rain- fall events (Table S) were slightly higher than those found in the flow-paced runoff samples from the irri- �ated e�ents ��ithout fzrtilization because thz volumes of runoff �� ere much less from the rainfall evznts. A'utri- ent losses. ho�cever. were much lower for rainfall events than irrieated evenu (Tables 4, �. 6, and 7). ConcenVa- tions in leachate were similar to those found for irrigated e.ents «•ithoet fertiiization. No indication of fertiliza- tion «�as e�•ident in samplzs for any rainfall evenL Linde et al. (199�) also reported'nigher nucrient concentrations but ]ower losses in runoff from rainfall events. Thz data from the rainfalt events is ntore reprzsentative of what ��ould likelv occur on a eolf course fairwav since data from the irrieated runoff e��ents that wzre preceded b� pre-event irrioation would represent a worst case scenario for runoff and nutrient transport. Five da}s before the irri�ated event on 6 Sept. 199�, plots «e.e fertilized at a rate of 4.9-03--4.1 o m'- (N-P- K). Immedia[ely foilo�rino fzrtilization, 9 mm of water �vas applied to each plot. In addition, plots werz verticut 6 h after the normal pre-event irri�ation procedure. Despite this mana�emznt scznario. nutrient concentra- tions remainzd ]ow for the e��enL Unlike previous fertil- Table 3. �Iean nutrient concentrations in runot'f and leachate samples for evenfs ihat rainfail caused runoff. \inate-\ Phosphate Total Kjeldahl-\ D�te Runot7 Le�ehate RunotF Leachate Runoft Leachate mg L 79Y3 137uh� 7.6 03 a79 135 0.7? 0.'_3 ?1 Julr: 15 1.1 6.06 1.10 0 0 ?? Jul. 0.7 0.6 6.00 ?.91 0.�? O.L 17 Au�.� 11 1.9 ?.7S 1.00 1.7> 0.21 17Sept: 1.9 ?.i0 1.09 1 \ov.:= 1.7 3.33 Q.90 ?1 \o..i_ ?.6 2.60 0.87 ?3 Xoc_ LO 1.60 0.06 199> 6 ]ulc_ ].6 6.61 0.�6 }1 Au�.= 2.0 4.?3 0.96 _'0 Oct.- OS 1.8_' 0.06 ` aIl repiications did not produce a sample. - Leachate sampies uere mt coliected. O�_�\\� lzs� izations conducted 4 h after pre-event irri�ation and S h before an ecent, it �eas not ecident in the water samplzs that fertilization had occurred. On ;olf course faira�ays, off-site movement of nutri- ents may happzn if runoff occurs soon after application of a�ranular fzrtilizer to a nearly saturated soil. To pre�•ent this scena; io, ��arious manaQement practices can be implemznted by the mrf manaser. One practice k'ould be to a��oid applyinQ fzrtiIizer on nearlv saturated soils, especiall�� �vhen rainfall is expzcted shortly after appiication. hno2her practice �vould be to vrater-in the fertilizer �cith li�ht irri�ation shordy afcer fertilization. Othzr practices to prevent nucrient transport in runoff include fotiar appiication of soluble nutrients and the usz of fertilizzrs thai have significant slo«�-release char- acteristics. Sediment Transport Since S3% of the 237 runoff samples analyzed con- tained no measurabie sediment, statistical analyses u�erz not conducted. Also. no consistent difference was evi- dznt benceen turf species. Therefore, sediment concen- trations from both turfs were averaged for each type of runoff sample on zach date (Table 9). Concentrations were usuall}� hiehest for che first flo�c-paced samp]es and lo�vest for the second flow-paced runoff samples. \Vhen sediment was measured, the amount was very small, evzn after certical mowine down the slope and removin� an average of 67 k� of organic material from the bentora=_s plots and an avera�e of 20 kg from the :yeerass plots. The highest observed sediment concentration for all runoff samples ���as 23� mg L detected in the first flow-paced sample of a rye�rass plot that produced 1675 L of ranoff on 29 June 199�. That 123.5 m� plot was irri�ated at 139 mm h for 1� min. To calculate the potential soif ]oss for that plot, sediment concentrations of both flo�v-paced samples were averaoed and equaled Table 9. A1ean sediment concenirations for irrigated erents. ntean sediment concentntion Runoff samplei D FPl FP2 mg L 1991 ?9 June 68.7 0 13 Juk S.8 3-9 ? Aug.' 0 0 2? Au�. 0 0 3 Sept.; 9.6 0.6 ?0 Sept. ?7.9 12.5 S Oct. 0.6 i-d 1995 16 �fav 0 0 31 >Ia� 0 0 li June_ 0 0 28 June ?.0 1.S 12 Juh 0.7 0 26 luh 0 2�� 6 Sept= 1.7 0 ?3 SepL 0.3 0 i FPS = lst ilo��-paced mnott' sample, FP'_ _'_nd ilo��-paced runoll sample. � Yertical mo��ins conducted 6 h before e�ent. 12�� J EYV(ROX QUAC.., VOL?6. SEPlE�iBER-0CTOBER (9Y7 143 ma L The potzntial soil loss from thz plot was calculated to be 19.4 k� ha for thz li min event. This number rzpresents the highzst potential soil loss for [his studv. Based on the acerages of total runoff volume (1390 L). duration (20 min). and concentration in the flo�s-paced sampies (9.b mg L the averaoz pocential soil loss for afl plots in 199-� was 15 ke ha for a 20 min zcznt. In 199�, thz a��erage potentizl soil loss ���as 0.1 kg ha ' for a 20 min ecznt. lisins S°io-sloped plots of tall fescue turf maintained at 8 cm, Gross et aL (1991) rzported the avera�e poten- tial soil loss for a 30 min, 120 mm h intznsity storm �cas �19 kg ha ' for bare soil and ��' ko ha for mature tall fescue turf szeded at 4SS kg ha �.�The much lower amount oi soil loss for the current study compared to Gross et at. (1991) could hzve been duz to differences in turf density and tortuousit}� of ocerland floa�. Gross et aL (1991) reportzd a mean density of �7 tillers dm' for a mamre tall fzscue turf sezded at 4SS kg ha and maintained at 8 cm. Linde et al. (199�) reported mean densities of 2�93 tillers dm for ma[urz creepin� bent- erass and 275 tillzrs dm'' for mature perennial rye�rass maintained at 1.9 cm. In this study, sedimen[ was seldom detzcted in runoff samples produced bq rainfalL Only 6 of thz �6 total runoff samplzs from rainfaff even[s contained detectable sedimenc Of those sis samplzs, thz highest sediment concen[ra[ion was 26 mg L for a bent�rass plot that produced 170 L of runoff in 100 min. Thz soi! loss for that plot �cas 0.36 k� ha''. In another study, Gross et aI. (1990) reported the average soil loss in runoff caused by rainfall #or sodded tall fescuelKenmcky blue�rass p1oG �vith a 5 to 7%a slope ti�as 0.4� ke ha and 1.47 k� ha for consecutive }'ears. SUDT�IARY AND CONCLUSIONS On avzrage, vzry fittle sediment transport, if any, was found in runoff samplzs. Therefore, for the conditions of thz curren[ studq, 4-yr-old creepino bentorass and pzrennial ryeerass turfs wzrz very effective in rzducine sediment transpurt, e�en aftzr vertical mowing do�vn the slope ok each plo:. I[ ma}� be possiole that if vertical V�����L mo�cina �cas more agaressive, sediment transport could increase bzcause a�reatzr amount of vegztation would be remoced and morz arooves cut into the soil. Nutrient transport, particularly phosphate and TKN, sienificantly increased for runoff events that had pre- e��ent irrigation and were conducted 3 h after fertiliza- tion. For a11 other evenG, nutrient transport �vas consis- tently lo«er. As a rzsul[, off-site movzment of nutrient; from golf course fairways may happen if runoff occurs soon aE[er sranular fertilizzr is applizd to a neariy satu- rated soil. Such conditions w•ould essentiafly represent a«orst casz scenario for runoff and nutrient transport and wouid Izss likely happen in `real world' circum- stances. REFERE\CES Btard. J.B. 197i. Turf�rass scizncz and culture. Przntice-HaL(, En�lz- wood Clift>, rJ. Gro>s, C.DI., J.S. Anglz. R.L. HiII, and 1LS. R'zIItrizn. 1991. RunoF[ and szdimznt losszs from [slt fescuz undzr simula[zd rainfalL I. Environ. Qual. ?0:6d3-607. Gross. Cbt., 7.S. Ano1z, and hLS. Neltzrltn. 1990. K�utrient and szdi- mcn[ losses from mrf�rass. J. Emiron. Qual. 19:66i-66S. Harrison, S.A., T.L �Ya[schke, R.O. hlumma, A-R. Jarre[[, and G.W. Hamilton. 1993. Nutriznt and pzsticidz concentra[ions in wa[zr from chemically treated mrf�ca>s. p. 191-207. I�i K.D. Racke and A.R. Les(iz (zd.) Pesticidzs in urban environments: Fatz and si�nifi- cance. ACS $ymp. Ser. 5??. Am. Chem. Soc.. �Vashington, DC. Lindz, D.T. 1996. RunoEf, erosion, and nucriznt [ransporc from crzzp- in� bznc�rass and pzrennia! ry'e�rass turfs. Ph.D. diss. Pznnsylvania State Univ., Univzrsity Park. PA. Linde, D.T., T.L. ��"atschkz, and J-A. Bor�er.1993. Nutriznt transport in rurtoff from two turf�rass spzcizs. p. 489-496. (n AJ. Cochran and bi.F. Farrally (edJ Science and �oif IL Ptoc. of the 19%Norid $cizntific Con�rev Of 6olf. E& F\ Spon, D'+zw Yotk. Linde, D.T., T.L. Wa[schke, A.R. Jarrett, and J.A. Bor�ec 1995. Surface mnoff assessment from creepin� bzntorass and pzrennial ryzgrass mrE A�roa J. 57:176-152. �[or[on. T.G.. A.J. Gold, and W.�I. Sullivan. 1988. Intluznce of over- warering and fertiliza[ion on ni[roozn losses from homz lawns. !. Eaviron. QuaL 17:1?�i-liQ. $AS Ins[itu[z 1990. SAS;STAT uszr's �uide. �'oL 2. Version 6. A[n ed. SAS Inst., Cary, FC. Stzzl, R.G.D., and J.H. Torrie. 1930. Principles and procedures o( s[a[istics: A biomz[rical approach. 2nd ed. hleGraw-Hilt, he«' York. ��'auchopz. R.D.. F.G. Witliams, and L.R. \farti. 1990. Runoff of sulfomz[uron-mzthyl and tyanazine from small plo[s: Effzc[> of formulation and srass co�er. J. Environ. Qual. 19:119-IZ�. Reprinted Cmm IheJouma! � f L•nvnonmrn/a! Quoliry Volume 23, no. I, Jan.-Fcb. I999. Copyright O I999, ASA, CSSA, SSSA 677 South $egoe Rd„ Mad(mq NI53711 USA 0 � �,��� Relationship between Phosphorus Levels in Three Ultisols and Phosphorus Concentrations in Runoff D. H. Pote,� T. C. Daniel, D. J. Nichols, A. N. Shazpley, P. A. Moore, dr., D. M. Miller, and D. R. Edwards ABSTRACT Soi15 that contsi� high P lerel5 Can become a primary 5ource of dissolved reactive P(DRP) in runoff, and thus contribute to acceler- ated eutrophication of surface waters. In a prerious sNdy on Captina soil, sereral soil te5t P(S1'P) methods gave resulB that wece signifi- canUy mrrelaFed fo DRP lerels in rvnoEf, but disrilled H and NH�- oxalate methods gave the best coaelations. Because results might differ on other soils, mnoff studies were conducted on three additional Ultisols to identify the most cnnsistent STP method for predicting runoffDRP lereis, and determine effecfs of site hydrology on correla- tions between STP and runoff DRP rnncentrations. Sucface soil {U- 2 cm depth) of pastuce plots was analyced 6y hlehlich III, Olsen, Morgan, Bray-Kurtz Pl, NH�-oxalate, and dis611ed H methods. Also, P saNration of each soil was deterntined by three diflerent methods. Simulated rain (75 mm h") produced 30 min of runoff Gom each ploL Ali cortelations of STP to (unoff DRP were significant (P G 0.01) regardless oF soil series or STP method, with most STP methods e Fiag digh correlxtions (r > 0.90) on all three soils. For a given levei of H�O-eztractable STP, low runoff vulumes coincided with low DRP concen[rations. Tf�erefore, when each DRP concentra- tion was di�ided by��olume of piot nmutt; wrrelafions to H:O-exfract- able STP had the same (Y < 0.05) regression line (or every soil. This suegests the importance of site hydrology in determining P Ioss in r�noff, and may provide a means of developing a single reiationship for a range of seil series. E uTxoexicnno:r of streams and lakes can be greatly accelzrated by the influs of nutrients in surface runoff from agricultu-al land. Since P has been identified as the nutrient in rur.off that is usually the most limiting to algal growth, control of P Izvels in runoff is often recommended as the best way to minimize the eutrophi- cation of su:face waters (Rohlich and O'Connor, 1980; Litt1e,198S; Breeuwsma and Silva,1992; Sharpiey et ai., 1994). Phosphorus is often perceived to be so immobiie in soil that losses from agricultural land are not usualiy considzred to be agronomically_iraportant, but even small agronomic losses can have serious environmental consequences. In iact, scils that contain high levels of P from escessive fertilization can become a primary source of dissolved reactive P(DRP) in runoff (Edwards et a1., 1993). Other investigators have found direct correlations be- tween soil P levels and P concentrations in runoff. D.H. Po[e, USDA-ARS, Dale Bumpers Small Fazms Res. Centez, 6833 South Stare Hwy. 23, Booneville, AR "R927-9214; T.C. Daniel, D.J. Nichols, and D-M. Miller, Dep. of Agro�omy, S15 Plan[ $<ience, Univ. of Arkansas, Fayetteville, AR 72701; P.A. Moore, Jr., USDA- ARS, 115 Ptant Scier.ce, Fayetteville, AR 72701; A.N. Sharotey, Pas- ture Systems and Watershed Managem�nt Research Lab., USDA- ARS, Curtin Road, University Park, PA 168023702; and D.R. Ed- wards, Biosystems and A�ricultural Engineering Dep., 128 Agricul- tural Engineering Building, Univ. of Ker,mcky, Lexing�on, KY 40546. Received 29 July 1998. *Correspondin� author (dpo�e@a�.gov). Published in 7. Environ. QuaL 25370.175 (1999). Schreiber (1988) sampled soil and runoff from mono- cropped com (Zea mays L.) or cotton (Gossypium Firsu- tum L.) research plots and watersheds in Mississippi, with various cropping pracaces for corn includin� con- ventional tillage, no-till, crop residue removed for sila�e, and crop zesidue left on the soii surface. Results showed that water-extractable soil test P(STP) was significantly correlated to annual discharge-weighted DRP in runoff. Yli-Halla et al. (1995) analyzed soil and runoff from eight cultivated field plots in southwestern Finland and concluded that mean DRP concentration in runoff de- pended on the water-extractable P level in surface soil. However, both of these studies relied on uncontrolled natural rainfall events to produce runoff, and combined a variety of cultivated crops and management practices, while neither study included uncultivated grassland. In a previous study (Pote et al., 199fi), we concrolled the variabitity of field conditions as much as possible by using consistent dimensions, slope, soil, and grass cocer for all plots, and using simulated rainfaL to pro- duce runofi. The study compared results from several soil test P(STP) extraction methods to determine which were most useful for predicting DRP levels in runoff from fescue (Festuca arundinacea Schreb.) plots on a Captina silt loam (fine-silry, siliceous, mesic Typic Ftagi- udult). Extraction of P in soil samples from the surface soil (0-2 cm depth) showed that the Mehlich III (Meh- (ich, 1984}, Bray-Kurt2 Pl (Bray and Kurtz, 1945), and Olsen (Olsen et al., 1954) extraction methods gave soil P levels with very significant correlations to DRP con- centrations in surface runoff. The soil P-saturation method (Pote et al., 1996) also gave zesuiu that corre- lated very well to runoff DRP, but Fe-oxide strigs (Sharpley, 1993), distilled water, and acidified ammo- nium oxaiate (Pote et a1.,1996) were the STP extractants that gave the best correlations to DRP in runoff. Since this study was only conducted on a single soil, we hy- pothesized that the results might be different for other soils of differing physical and chemical properties, even with;n the same soil order. As severai states are attempting to define threshold STP levels above which DRP enrichment of runoff is unacceptable from a water-quality perspective, more information relatin� soil P to runoff P is needed (Shacpley et al., 1996). Such Field data are essential to development of technically-sound STP levels that can be used to guide P management recommendations. Therefore, runoff studies were conducted on three addi- tional Ultisols. The objectives were to determine (i) which STP method maintains the highest correlation to Abbreviations: CV, coefficient of variation; DRP, dissolved reactive P; ICP, induc[iveiy coupted plasma spectrometer, M3, Mehlich III extraction me[hod for soil P; PSI, P sorp[ion index; SD, siandard deviation; STP, soil tes[ P. y �, 170 O l - �� \'�-- POTE EI AL: PHOSPHORUS LEVELS LY THREE ULTISOLS Tabie 1. Soil charaderisticc (mean) and :esu{ts of various soiS test P(STP) methods from plots on three soils. Nella so� Clay covteat 105% Oiganic C content 3.S% pH 59 Oxalate-Fe, mg kg ' 1909 Oxalate•Al, mg kg ' LiO4 Ne{ls wl Ztange Mean SDi Lioker wil li9q 3.6 % 51 3003 1170 Linker soil Range hlean SD _ �p � � i in lYOark swl 7.4% 4.6 % 6.2 104i 16M13 noack :oa Range Mean SD STP method - - � biehlich III 260.42L 294 73 12I 366 226 � 17-?63 109 � Olsen 79-166 175 28 61-1b2 104 31 7-303 44 30 Morgan 23b5 44 IS 30-108 57 27 0.107 35 34 Bray-Kum PS 161-3d2 7A0 62 121-328 ?A7 76 14-156 90 47 Nil..ozalate 691-1L7 9DD L'9 315-707 442 144 210.613 4W 1'A1 DisNled H�O 37-109 74 25 1& -107 50 3U b-3U 36 24 t Standard deriation. DRP concentra[ions in runoff from a variety of soil series within the Ultisol order, (ii) whether STP levels affect DRP concentrations in runoff consistently across soil series and if not, (iii) what effect soil hydrology has on the relationship between STP and runoff DRP. MATERYALS AND bIETHODS a Field Plots Six field plots were constructed during the fall of 1993 on each of three soils in northwest Arkansas: Nella (fine-loamy, siliceous, thermic Typic Pateudult), Linker (fine-loamy, sili- ceous, thermic Typic Hapluduli), and Noark (clayey-skeletat, mixed, mesic Typic Paleudult) (Table 1). All plots were con- structzd on well-established tall fescue pastures with approxi- ma[ely 7% slope and 100% ground cover as measured by the line-transect me[hod (Laflen et al., 1981). These pastures had previously been amended with various combinations of swine manure slurry, commercial fertilizers, and/or manure from grazing cattle. Some plots had received swine manure [he previous year, but no amendments were allowed on the plots for sevetal months preceding this s[udy. Vegeta[ion height was maintained between 0.1 and 02 m throughout the study by mowing. Each plo[ (1.5 x 3 m) was fitted with aluminum borders (extending 5 cm above and 10 cm below [he surface) for runoff isolation, a downslope trough for runoff collection, and a runoff sampling pit, as described by Edwards and Daniel (1993). Fences were constructed around the plocs to prevent catile from contributing P inp�ts or causing other damage during the study. In May 1995, a simula[or described by Edwards et al. (1992) was used to reduce antecedent moismre variability by applying rainfall (75 mm h'') to each ptot until the su[face layer was saturated. This simulator delivers rainfatl at an eait pressure of 41.4 kPa from four VeeJet nozzles` elevated 3.05 m above the soii surface by an aluminum scaffold to obiain drop-size distribution and terminal velocity compazab(e to that of natu- ral rainfall. Tarpaulins attached to the aluminum scaffold sur- round the plot to form wind screens. An elec[rie motor drives the shaft to which the nozzles are attached, causing them to oscillate across openings in the simulator body, with the rain- fall intensity dependent upon the frequency of oscillation. ' Names are necessary to repor[ factualty on available data; how- ever, [he USDA nei[her guarantees nor warrants [he s[andard of the produc[, and [he use oE the name by USDA implies no appmval of, : the product to [he zx<lusion of othecs Ihat may also be suitable. � Following the initiai rainfall application, all plots were allowed to drain for 48 h before simulated rain was applied again a[ an intensity of 75 mm h'' to generate 30 min of runoff from each plot. Sampling Methods Runoff was sampled manually at 5-min intervals throughout the runoff event, beginning 2.5 min after initiation of continu- ous-flow runoff. For each discrete runoff sample, the volume and time required to colleM it were recorded and used to calcula[e mean flow rate and total volume of runoEf for the 5-min interval. Using these runoff data, the six discrete runoff samples from each plot were used to construct a flow-weighted composite sample to represent the total runoff from that plot. An aliquot of each composite runoff sample was filtered (0.45- µm pore diame[er) within 2 h of collection and stored in the dark at 4 until analyzed for DRP by the molybdeoum-blue method (Mucphy and Riley,1962). Total DRP mass loss from each plot was calculated as the plot's total runoff voiume multiplied by DEtP concentration in the flow-weighted com- posite runoff sample from thaf plot. . Just prior to applying simulated rainfall to a given pbt, a representative composite soil sample was collected by combin- ing 10 discrete soil cores (2.54 cm diam.) taken randomly from the surface layer (0-2 cm depth) of the plot. All composite soil samples were stored in the dark at 4 until air dried and sieved (2 mm) to remove larger rock particles and most of the plant material. Two complete runofF events were conducted on each plot, separated by a 2-d interval. For each separate runoff event, soil samples were collected jus[ prior to simulated rainfall ap- pHcation. Soil Analyses Each soil sample was analyzed for extractable P by six methods: Morgan (Morgan, 194I), Mehlich III (Mehlich, 1984), Bray-Kurtz Pl (Bray and Kurtz, 1945), Olsen (Olsen et al., 1954), distilled water, and acidified ammoaium oxaLate. The Morgan, Mehlich III, Bray-Kurtz Pl, and Olsen chemical extractan[s were selected because they aze commonty used for STP analysis in soil testing laboratories. These methods were not originally devetoped to predict runoff water quality, but rathez to assess the fertility status of soil for aop produc- tion. Distilled water most closety simulates actual runoff solu- tion, and may thus be the most appropriate for predicting runoff DRP. One „cram of soil was mized with 25 mL of o � - ����. I�Z J. ENVIRON. QUAL., VOL. 28, JANUARY-FEBRUARY 1999 distilled water, shaken end-over-end for 1 h, centrifuged for 5 min a[ 266 m s�(27100 g), Eiltered (0-45 µm), and the supzmatant analyzed for P by the molybdenum-blue method (Murphy and Riley, 1962)_ Acidified ammonium oxalate has been used in severaf pre�,ious studies (van der Zee et al., 1987; van der Zee and van Riemsdijk, 1988; A�folina et al., 1991; Breeuwsma and Silva, 1992; Freese et al., 1992), theoretically [o release into solution potentially dzsorbable P, as it dissolves the compounds (noncrys[alline oxides oi iron and aluminum) controlting P sorption in acid soils (Table 1). In our smdy, ammonium oxalate extractant was made by mixing 02 M oxalic acid with 0.2 M ammonium oxalatz (approximately 535 mL of oxalic acid with 700 mL of ammonium oxalate) until the combined-solution pH was 3.0. A 20-mL aliquot of the ammonium oxalate solution was then mixed with 0.5 g of soil, shaken in the dark for 2 h, cenirifuged for 20 min at 131 m s '(14481g),anddecantedforPanalysis.Oxalate-extractable P, A., and Fe were also used to calculate the P sorption- sa[uration of each soil as described below. Mehlich III, Bray- Kurtz Pl, and acidified ammonium oxalate extracts were ana- lyzed for P by inductively coupled plasma spectrometer (ICP), while n4organ, Olsen, and d'utitied water extracts were ana- ]yzed colorimetrically by the molybdenum-blue me[hod (Mur- phy and Riley, 1962). A single-point P sorption index (PSI) described by Mozaf- fari and Sims (1994) was also determined on each soil. A P sorption solution (containing 300 mg P per liter) was made by dissolving 1.315 g of KH in enough distilied, deionized H to make 1 L of solucion. The PSI was determined by weighing 1.00 g of soil in[o a�0-mL centrifuge [ube, adding 20 mL of 0.0125 M_ CaCi 2H>O, and adding � mL of P sorption solution to make a combined solution containing 0.01 M CaCi and 60 mg P per liter. After two drops of toluene were added and the tubes sealed, the mixture was shaken for 18 h on a reciprocating shaker, centrifuged for 10 min at 266 m s' (27100 g), filtered (0.45-µm), and analyzed for P by induc- tively coupied plasma spectrometer (ICP). The PSI was calcu- lated as X(lo� P where X is P sorbed (mg kg'`) _[(P�)(V) —(P (kg of soil) P is initial P concentration in sorption solution (mg L V is volume of P sorption solution (L) P is final P concentration in solu[ion (m� L Phosphorus Saturation of Soil The P saturation (%) of each soit sample was calculated by two different methods; (i) oxalate-extractable P(mmol kg") divided by thz oxalate-extractable AI and Fe (mmot kg con[ent, and multiplied by 100, and (ii) initial STP con- ten[ (mg kg ') divided by P (mg kg and multiplied by 100. For this second method, the PSI value was used to approximate the maximum amount of P(P.,,, that could be adsorbed by the soil. Mozaffari and Sims (1994) found that P� can be estima[ed by the equation P�� =(PSI + 51.9)/ OS, �iven that P Mqr < 1400 mg kg STP extracta�ts selected to obtain the initial STP conten[ were Mehlich III (M3-PSI method) and distilled H2O (H2O-PSI method). Statistical Methods For each soil, comparisons were made between STP meth- ods by correlating STP results to DRP concentrations in runoff from the ptots, developing a linear regression from [he 12 data points, and calcula[ing [he sample corrzla[ion coefficient (r value) for each. For each soil test method, analysis of covari- ance was used to determine whe[her there were statistica! differences be[ween regression slopes and in[ercepts of ihe three soils. RESULTS AND DISCUSSION Soil Phosphorus For each of these soils, the range, mean, and standard deviation of STP contents are shown in Table 1. Distilled water, Mor�an, and Olsen methods extracted the least amounts of P from soil, while Mehlich III and Bray- Kurtz Pl methods extracted larger amounts. NH,-oxa- late extracted much larger amounts of soil P than did other extractants, suggesting that most of the P in these soils is sorbed or precipitated on amorphous oxides of Fe and Al. Relationship between STP and Runoff DRP For each soil, correlations of STP to runoff DRP were not significantly affected by the time interval (2 d) between the two runoff events. Therefore, the data from both runoff events were combined to �ive a total of 12 data points for each soil. The correlation coefficient (r) and linear regression equation are given in Table 2 for each STP correlation to DRP in runoff. For all soiLs, the STP values obtained by each method were significantly corretated (P < 0.01) to DRP concentrations in plot runoff. Yet, when the extraction methods were com- pared using r values to see how closely the data peints fit the regression line, it was apparent that some STP methods were more closely related to DRP concentra- tions in runoff than other methods (Table 2). For exam- ple, the NH and Olsen methods each gave a weaker correlation r< 0.90) to DRP concentrations in runoff from at least one soil, while all other STP meth- ods gave correlations with r> 0.90 for all three soils (Table 2). However, if previous studies (Pote et al., 1996) are considerzd, the H2O-estractable soil P has shown the most consistently high correlation to DRP concentrations in runoff, even when rainfall intensity, slope, and seasonal conditions varied. Although the usefulness of an STP method for pre- dicting runoff DRP concentrations depends largely on its ability to produce data poinu that closely fit a regres- sion line on any given soil, it would also be very helpful to have an STP method that produces approximately the same regression for all soils (or at least a large group of soils). Such a method would eliminate the need to use soil series as the basis for maximum soii P recom- mzndations, thus saving the time and expense of accu- rately identifying the soil series of each individual site. If the data points from all three soils are combined into a single data set, the P-saturation (oxalate method) might seem to be a good choice for this purpose because it gives a good linear correlation (r = 0.887), and the fit is even better for a second-order regression (r = 0.931 for the curve where y= 03053 — 0.0353x + 0.0014x However, if the data points are separated into regression lines for each soil, differences between some slopes be- come apparent (Fig. l). When the regression-line �raphs of each method were compared visually, the P-satura- POTE EI' AL PHOSPHORUS LEVEIS IN THREE ULTISOLS 0.905 0$69 0.907 ' . 0.913 0.806 0.9'� 0903 0.916 0932 Table 2. Results of soil test P(STP) methods correlated to dissolved reactive P(DRP) in runoH from three Ultisola CortelaRon coefident (r) t S1 P(m kg"') co rrelated t DRP (mg L STP method Nella w� Linker so0 Noark wii Mehtich III Olsea Mocgan Brav-Kurtz PS NH,-Ozalate Disb7led H.O P saNlation (oxalate method) P satuntion @13-PSI metbod) P satucdtion (H_O-PSI metLod) $TP method Mehlich III Oisen Morgan Bnv-Kurtz Pl NI-I�-Ozalate Distilled H P saturation (o:Wate met6od) P saturation (M}pSI method) P saturation (H.O-PSI metho� 0.916 0.864 0.941 0.950 0.914 0.928 0.928 0.928 0921 �\�\\\b- 173 0.932 0935 0.932 0943 0908 0.965 0.933 0.937 0.978 Regression line equa for ST P (mg kg �) <onelated to DRP (mg L NeO soii Linker sol Noark soil y= 0.0036z - OAS y= 0.01135c - 038 y = 0.0016x + 0.00 y= O.00SSz - 0.43 y= 0.0093z - 056 y= O.00d3x - 0.02 y= O.OlSlz - 0.18 y= O.O1LSZ - 025 y= 0.0038x -F 0.04 Y=O.00d3x-0.42 y=0.0041ac-0.4b y=0.0027x-0.02 y= 0.001Sx - 1.03 y= 0.002Lc - 0.63 y= 0.0009x - 019 y = 0.0107x - 0.18 y = 0.0104x - 0.11 y= O.00SSx - 0.03 y= 0.0820x - 203 y = 0.0397x - 0.62 y = 0.0251c - 01A y= O.00S� - 0.08 y= 0.0065x - 0.04 y= 0.0045x + 0.03 y = 0.0?62x + 0.03 y = 0.0215z + 0.06 y= 0.0759x + 0.01 i All correla6on coeffidents were signifitant (a = 0.01). tion (PSI methods), Mehlich III, Bray-Kurtz Pl (Fig. 2), and distilled H (Fig. 3) methods each appeared to have regression lines that were relatively close together with similar slopes for all soils, but statistical analysis showed that none of the methods for correlating STP to DRP in runoff gave the same (P � 0.05) regression line for all three soils. This result was not surprising, given the differences in chemieal and physical properties between soils. The P saturation status of each soil in this study was significantly (P < 0.01) related to DRP concentrations in runoff, regardless of the method used to calcutate P saturation. All three methods gave high correlations to DRP in runoff but none gave the same regression line on all three soils, so their vatue as universal predictors of DRP concentrations in runoff is questiouable. i � E 0 � c a � 0 1.6 �,a 7.2 7.0 0.8 0.6 0.4 0.2 • Nella (r = 0.903) � Linker (r = 0.928) ONoark (r = 0.933) 0.0 ' " ' 0 10 20 30 � 40 � • • • � � • O • M .� Effects of Runoff Volume Because site hydrology of each soil is likely to impact the relationship between soil P and runoff P(Gburek and Sharpley, 1998), the effect of runoff votume on P transport from our plots was evaluated. The average rainfall application required to produce 30 min of con- tinuous runoff is included in Table 3, along with mean runoff volume for each soil. Runoff from the Nella and Linker soils averaged about the same volume and the variability was also similar, while the Noark soil had the lowest amount of runoff and the least variability. The differences in runoff among soils are reflected in the correlations of water-extractable STP to runoff DRP. For esample, when water-extractable STP was correlated to mass losses (loads) of DRP in runoff (Fig. 1.6 7.4 7.2 �� rn 1.0 E 0.8 ' 0.6 c a 0.4 ¢ 0 0.2 0.0 0 700 200 300 400 Bra Kurtz extractabie soil P m k - Soil P Saturation (%) Y � 9 9) Fig. 1. Relationship between P samration (oxalate method) of suhace Fig. 2. Relationship behveen Bray-Kurtz Pl extractabfe P in sudace soil and dissoived reactive P(DRP) in runoff from three soils. soil and dissolved reac[ive P(DRP) im m�oll. . 174 7.6 7.4 1.2 i � 7.0 0 0.8 c i 0.6 � 0.4 0 0.2 a.o 0 20 40 60 J. ENVIRON. QUAL., VOL 28, JANUARY-FEBRUARY 1999 80 700 120 Water eztractable soil P(mg k9 Fg.3. Relationship between water-eztraMabie P in surface soil and dissolved reactive P(DRP) in runoS 4), the Noark correlation was best (r value = 0.963) because mass losses depend on boSh.the P concentration and the volume of runoff (which" was highly consistent for the Noark soil). Runoff volumes were more variable for the other two soils, and therefore mass losses of runoff DRP show a poorer correlation to STP (Fig. 4). The variability of runoff volume is also reflected in the r values for the correlation of water extractable STP to DRP concentrations in runoff (Fig. 3). For example, Nella soil had the most variable runoff volume, and it also had ihe lowest r value, while Noark soil had the least variable runoff volume and the highest r value. Finally, for a given level of water-extractable STP, soils with the lowest mean runoff volume also had the ]owest concentration of DRP in runoff (Fig. 3). For example, Noark soil produced the least amount of run- off, but for any given level of water-extractable STP, it also had the lowest concentration of DRP in the runoff. No previous studies have investigated the relationship between runoff volume and DRP concentration in the runoff; and our observations at first seemed rather counter-intuitive because we expected higher volumes of runoff to generally produce lower DRP concentra- tions due to greater dilution. This unexpected trend may result from the rapid movement of DRP into the soil profile of soIls with low runoff volumes (high infiltration rates), thus taking it away from the primary zone of transfer to surface runoff. In soils with lower infiltration Table 3. Rainfal{ aud runo8 data from simufated rain appliration to 6eld plots on three soiLs. Ne17a soii Linker soii Noark sol Rainfall meant, mm 4$.9 475� 53.6 Ruuo@'mean,mm 21.7 7A.6 133 Runoff CV, % 30.0 29.1 175 i Amount required to produce 30 min of rumR. Eath mean tepresenls 12 runoH events. 400 �0 300 ._°: 0 � 200 i c v a 100 a 0 0 b a\ -\\\�� •Nella (r = 0.724) �Linker (r = 0.847� ONoark (r = 0.963) •• • � � • • � � � � � �• � • . o 20 40 60 80 100 120 Water extracfoble soil P(mg kg Fg.0. Relafionsltip between water-extractable P in surCace soil and dissolved reactive P(DRP) load in runoff. rates, much more of the dissolved P may remain near the soil surface long enough to be lost in runoff water. In an attempt to define these processes, we normal- ized DRP concentration for each plot. When the DRP concentration in runoff from each plot was divided by the depth of runoff from that plot, and related to the water extractable STP level, regression lines for all soils were statistically the same line (P < 0.05) (Fig. 5). Thus, by combining water-extractable STP data with hydro- logic data, it may be possible to make reasonably accu- rate predictions of DRP levels in runoff from a range of soils. Acquiring the necessary hydrologic da[a on runoff volumes from a soil may sometimes be just as difficult as accurately identifying the soil series of each specific site, but this at least provides an a!temate 7.2 E � 1.0 0 � 0.8 � ^ o.s � � E o.4 `o � 0.2 � c a 0.0 � 0 0 . . . • • • • . � b .� . 20 40 60 SO 100 120 Water eztractable soil P(mg kg Fg. 5. Relafio�ship between watervextzactable P in sudace soil and the ratio of dissolved reactire P(DRP) im m�otf to the total amount of runoR. •Nella (r = 0.866) - �Linker (r = 0.847) oNoark (r = 0.919) POTE EC AL: PHOSPHORUS LEVELS I1V THREE ULTISOLS method for predicting DRP concentrations in runoff. For water-quality modelers, it also supplies important information conceming the relationship between vol- ume of runoff and DRP concentration in runoff. Most importantly, i[ shows the strong influence of site hydrol- ogy on processes controlling P loss in surface runoff. CONCLUSIONS The results of this study reinforce previous evidence of a linear relationship between P levels in surface soil (0-Z cm deep) and DRP concentra[ions in runoff from the soil surface, but this study also extends our knowl- edge by showing that such a relationship exists on a variety of Ultisols. On each soil that was tested, a signifi- cant (P < 0.01) linear relationship was apparent, regard- less of the method used to determine STP. Because most STP extractants gave results that were highly cor- related (r > 0.90) to DRP in runoff from all three soils, this study did not clearly identify any particular STP method for maintaining the highest correlation to DRP concentrations in runoff from all soils tested. However, the study did show that several STP extractants may be useful for predicting DRP concentrations in runoff, including extractants such as distilled water that were supported by the results of previous work (Pote et al., 1996) conducted under different rainfall intensity, slope, and seasonal conditions. This study showed that effects of STP levels on DRP concentrations in runoff are not always consistent across soil series, and much of the difference can be attributed to soil hydrology. The fact that total plot runoff was much more variable on some soil series than on others was apparently reflected in correlations be[ween STP and runoff DRP, as soils with the most consistent vol- ume of plot runoff had the best correlations of water- extractable STP to both concentrations and mass losses of DRP in runoff. Also, for any given level of water- extractable STP, soils that produced the lowest volumes of runoff also had the lowest concentrations of DRP in the runoff. When this information was used to normalize the data for DRP concentrations in runoff (divide each DRP concentration by the volume of runoff from that plot), the resulting correlations to water-extractable STP had statisTically the same (P < 0.05) re�ression line for every soil. This implies that knowledge of site hydrology can improve the usefulness of STP data for predicting DRP concentrations in runoff. REFERENCES Bray, R.H., and L.T. Kurtz. 1945. Determination of [otal, organic, and available foans of phosphorus in soils. Soil Sci. 59:39-45. Breeuwsma, A., and S. Silva. 1992. Phosphorus fertilisa[ion and en�i- ronmencal effec[s in The Netherlands and [he Po region (I[aly). Rep. 57. Agric. Res. Dep. The Winand S[aring Cen[re for In[e- (j \-\\\�--- 175 grated Iand, Soil and �'da[er Research, �'lageningen, the Ne[h- edands. Edwards, D.R., and T.C. Danie1.1993. Effects of poultry lit[er applica- tion rate and rainfall intensity on quality of runoff from fescuegrass plots. J. Enviton. QuaL 22361-365. Edwards, D.R, T.G Daniel, J.F. Murdoch, and P.F. Vendrell. 1993. The Moore's Creek BMP effectiveness monitoring pcoject. Paper 93208�- ASAE, St. Joseph, MI. Edwards, D.R., LD. Nocton,T.C. Daniel, J_T. Walker, D.L. Ferguson, and G.A. Dwyer. i992. Perfonnance of a rainfall simularor. Arkan- sas Farm Res. 41:1'ri4. Freese, D., S.E.A.T_M. van der Zee, and W.H. van Riemsdijk. 199"t Compadsons of different modzls for phosphate sorption as a func- [ion of the iron and aluminium orzides of soils. J. Soii Sci.43:'729-738. Gburek, W.J., and A.N. Sharpley.1998- Hydrologic con[rols on phos- phorus loss from upland agricul[ural wa[enheds. 7. Emiron. Qual. 27.267-277. I.aflen. J., M. Amemiya, and E.A. Fiiniz.19S1. Measuring crop residue cover. J. Soil Water Conserv. 6:341-343. Little, C.E. 1988. Rurai dean water. The Okeechobee story- 7. $oil Water Conserv. 43:38C�390. Mehlich, A. 1984. Mehiich 3 soil [es[ extractanC A modifica[ion of Mehlich 2 extrac[ant Commun. Soil Sci. Plant AnaL 15:1409-1416. Molina, E., E. Bomemisza, F. Sancho, and D.L. Kass. 1991. $oil aluminum and iron fractions and their relacionships wiih P immobi- liza[ion and other soil properties in andisols of Costa Rica and Panama. Commun. Soil Sci. Plant AnaL 221459-1476. Morgan, M.F. 1941. Chemical soil diagnosis by [he universal soil testing system. Conn. Agric. Exp. Stn. (New Haven, CT) Bul(. 450. Mozaffari, M., and I.T. Sims. 1994. Phosphorus availability and sorp- [ion in an A[lan[ic coas[al plain wa[ershed domina[ed by animal- based agriculture. Soil Sci. 157(2):97-107. Muiphy, J., and J.R. Riley. 1962. A modified single solu[ion method for [he de[ermination of phosphate in na[ural waters. Anal. Chem. 2731-36. Olsen, S.R., C.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estima- tion of available phosphorus in wils by extraction with sodium bicarbona[e. USDA Circ. 939. U.S. Go�. PrinC Office, Washing- ton, DC. Pote, D.H., T.C. Daniel, A.N. Sharpley, P.A. Moore, Jr., D.R. Ed- wards, and D.J. Nichols.1996. Relating ex[ractable soil phosphorus to phaspho�us losses in runoff. Soil Sci. Soc. Am. J. 60:&55�59. Rohlich, G.A., and D.I. O'Connor. 1980. Phosphorus managemen[ for [he Grea[ Lakes. Fival Rep., Phosphorus Management S[ra[e- gies Task Force, Int. Ioint Commission (IJC). Poilution from Land Use Ac[ivities Reference Group Tech. Rep. Phosphorus Manage. Strategies Task Focce, Windsor, ON. Schreiber, 7.D. 1988. Estimating soluble phesphorus (POrP) �n ag- ricultural runoff. J. Miss. Acad. Sci. 33:1-15. Sharpley, A.N.1993. An innovative approach [o estimate bioavailable phosphorus in agriculiural runoH using iron oxide-impregna[ed paper. 7. Environ. Qual. 22:597fi01. Sharpley, A.N., S.C. Chapra, R. Wedepohl, I.T. Sims, T.C. Daniel, and K.R. Reddy. 1994. Managing agricultural phosphorus for pro- tection oE wrface waters: Issues and op[ions. J. Envimn. Qual. 23: 437�51. Sharpiey, A.N., T.C. Daniel, J.T. Sims, and D.H. Pote.1996. Decertnin- ing environmentally sound soil phosphorus levels. J. Soil Water Conserv. 51(2):160-166. van der Zee, S.E.A.T.bt.. L.G.J. Fokkink, and W.H. van Riemsdijk. 1987. A new tzchnique for assessmen[ of reversibiy adsorbed phos- pha[e. Soii Sci- Soc. Am. 7. 51599�04. van der Zee, S.E.A.T.M., and W.H. van Riemsdijk. 1988. Model for lon�-term phospha[e reac[ion kinetics in soil. 7. Environ. Qual. 1735�1. Yli-Halla, M., H. Hartikainen, P. Ekholm, E. Tur[ola, M. Puustinen, and K. Kallio. 1995. Assessmen[ of soluble phosphorus load in surface runoff by soil analyses. Agda Ecosyst. Environ. 56:53-b2. �A�.��a� �d - �o� a$ a.OQ, Councii File # � �+'� �� Green Sheet # 1 \'j �. � y ORDINA OF SAINT P� Presented Referred To ..� Committee Date � �o 1 An ordinance amending Saint Paul Legislative Code Chapter 377 t� ��the use of fertilizers containing 2 phosphorus 3 THE COUNCIL OF THE CITY OF SA1NT PAUL DOES ORDAIN: Section 1 5 Chapter 377 of the Saint Paul Legislative Code is hereby amended to read as follows: 6 Sec. 377.01. Definitions. For the purposes of this chapter, the terms defined in this secrion have the meanings ascribed to them: 8 Person means any person, firm or corporation engaged in the business of lawn fertilizer or pesticide 9 applications and includes those persons licensed by the State of Minnesota pursuant to Minnesota Statutes, Secrion 10 18A-21 et seq. 11 Pesticide means any substance or mixture of substances intended for prevenring, destroying, repelling or 12 mifigating any pest, and any substance or mixture of substances intended for use as a plant regulator, defoliant or 13 desiccant. It also means any chemical or combination thereof registered as a pesticide with the U.S. Environmental 14 Protection Agency, any agency later assuming registration in the U.S. federal government, the State of Minnesota 15 Agricultural Deparhnent, or any other State of Minnesota government agency. 16 Sec. 377.02. License required; council approval. 17 (a) No person shall engage in the business of lawn fertilizer or lawn pesticide application in Saint Paul without 18 a license issued by the City of Saint Paul. 19 (b) All city programs for pesticide use shall be reviewed and approved by the city council prior to any application 20 upon city property. 21 Sec. 377.03. Fee. 22 The fee required for a license shall also be established by ordinance as specified in section 310.09(b) of the 23 Saint Paul Legislative Code. ; ► �' ;. . 1 Sec. 377.04. Employees licensed by state. O \ � �\\9� All ofiicensee's employees actually engaged in lawn pesticide applications shail be duly licensed by the State of Miunesota and shall be trained and qualified in the proper methods of handling and applications of pesticides. Satisfactory evidence that such employees are licensed by the state shall be maintained on file in the office of the license inspector. 6 Sec. 377.05. Division of health. 7 The � - DirectoroftheOfficeofLicense,Inspections 8 and Environmental Protection or his/her desi¢nee is directed to monitor the health and safety effects ofthe chemical 9 applications to lawns and to advise the ciTy council of any suspected hazards or violations. 10 Sec. 377.06. Class I license. 11 The license granted pursuant to the provisions of this chapter is designated as a Class � R license, subj ect to 12 the procedures applicable to Class � R licenses in Chapter 310. 13 Sec. 377.07. Pesticide applications; posting. 14 15 16 17 18 19 20 All persons who apply pesticides outdoors are required to post or affix warning signs on the street frontage ofthe properry so treated. The warning signs must protrude a minimum of eighteen (18) inches above the top ofthe grass line. The warning signs must be of a material rain-resistant for at least a forty-eight-hour period and must remain in place for at least a forty-eight-hour period or longer if the human re-enhy interval prescribed in the pesticide label specifies a longer hourly or daily interval. The information printed on the sign must be printed in contrasring colors and capitalized letters at least one-half inch or in another format approved by the coxnmissioner of agriculture, and shall provide the following information: 21 (1) 22 23 (2) 24 25 26 27 The name of the company applying the pesticide or, if not a company, the name of the person having done the application. The following language: "This area chemically treated. Keep children orpets offuntil (date of safe entry--at least forry- eight (48) hours after applicafion or longer if specified on pesricide label)" ar a universally accepted symbol and text approved by the commissioner of agriculture specifying a date of safe entry as specified herein. The warning sign may include the name of the pesticide used. 28 The sign shall be posted on the lawn or yard no closer than two (2) feet from the sidewalk ar right-of-way and no 29 further than five (5) feet from the sidewalk or right-of-way. 30 31 Sec. 377.08. Fertilizer Content. No person licensed under this chapter sha11 appl�y lawn fertilizer. Iiquid or 32 p_ranulaz. within the Citv of Saint Paul that is labeled to contain mare than 0% phosphate (P O ,�prohibition 33 shall not appplv to: 1 a. The naturallv occurrin� phosnhorus in unadulterated natural or organic fertilizing O l-1\�Y 2 products such as vard waste compost; 3 b. Use on newlv established or developed turf and lawn azeas durin¢ their first growing 4 season• 5 c. Turf and lawn azeas which soil tests taken according to Universit�of Minnesota 6 ¢uidelines and analyzed in a State of Minuesota certified laboratorv confirm are 7 below phosphorus levels established by the Universitv of Minnesota. In such cases, 8 lawn fertilizer application shall not exceed the Universitv ofMinnesota recommended 9 application rate for phosphorous. 10 11 12 13 14 15 16 17 Lawn fertilizers contaiuingphos hro orus applied pursuant to the above-listed exceptions shall be watered into the soil where the phosphorus can be nnmobilized and enerallv nrotected from loss bv runoff. Fertilizer applied to impervious services, such as sidewalks. drivewavs and streets is to be removed bv sweepins or other means immediately after fertilizer application is completed. Fertilizer is not to be apniied to frozen soil. saturated soil or under conditions ofim ep nding heaw rainfall. The Office ofLicense, Inspections and Environmental Protection shall be notified at least 24 hours prior to the application of anv lawn fertilizer containing_phosphorus that such fertilizer will be used. the amount to be used and the reason for its a�lication. Section 2 18 This ordinance sha11 take effect and be in force thiriy (30) days following its passage, approval and publication. Yeas Na s Absent Benanav i /' Blakey f Bostrom „/ Coleman ✓ Harris ,/ Lantry �/' Reiter �' Adopted by Council: Date - �. S� 1,oa � AdopUon Certified by Council Secretary By: 2 . � --Q- Approved by Mayor: Date By: � Requested by Deparnnent o£ � Form Approved by Ciry Attomey �� By: Approved by Mayor for Submission to Council � By: � 2 7 'A'� : �� °—�.- S a9�J�� V < �,.. , IST BE ON CIXRJCILAGB�IDA BY OCt. Z�F, ZOO1 a „��,m. ,_ _ ,. C 1- ►1\3 o/laio� GREEN SHEET No 113684 � TOTAL # OF SIGNATURE PAGES u��� u��— ❑ arvwnoa�r ❑ rnrcuu _ ❑ waxr,n�fEau¢ESO.� ❑ qmcu�mnri�eero ❑ ❑ (CLJP ALL LOCATIONS FOR SICaNATURE) An ordinance amending Saint Paul Legislative Code Chapter 377 to prohibit the use of fertilizers containing phosphorus. PLANNING COMMISSION qB COMMITTEE CNIL SERVICE CAMMISSION IIy_\ �=,ti]qd�7 ' TOTAL AMOUNT OP TRANSACTION S �� ' FllNOING SOURCE FlNqNCIAL INFORMAl10N (EXPWf� H36 th15 p2fSMl�fi112R! VqIKIM UlMN 8 COIIffdC[ f0f fhi6 d2p3lIIT2M? VES NO Has thia Pe���m e<u been a dty employee9 ' YES NO Does tlds PersaVfim+ V� as4dN nM na�matHP� b'f anY cwrent dty emWoyee't YES MO Is ttiis peBOMfirtn a targeted �endoYt YES � COST/REVENUE BUDGEfED (GRCLE ON� ACTIVIiY NUMBER VES NO ❑ YINNESOTA BOARD OF WATEfl ANC SOIL NESOURCES NORTHERN FiEGION 394 S Lake Ave Room 403 Duluth, MN 55802-2325 PHONE (218)723-2350 FAX (218) 723-4794 �MINNESOTA BOAND OF WATEfl AND SOIL RESOURCES METRO REGION One W Water SL, Suite 200 St. Paul, MN 55107-2039 r� , Water Resources Education Date: November 12, 2001 To: Councilmember 7ay Benanav Councilmember Jerry Blakey Councilmember Dan Bostrom Councilmember Chris Coleman Councilmember Pat Harris Councilmember Kathy Lantry Councilmember Jun Reiter ! �� UNIVERSITY OF MINNESOTA Extension �� O I - ��l'�- � B � ,M,�,.,,,,,,. ' �CC�G�4�L n'YV!' 1 3 2 �Di �4TY ��ERK Mayor Norm Coleman From: Ron Struss, University of Minnesota Extension __'Sf Re: UM comments on proposed lawn fertilizer ordinances and regulations PHONE (651) 215-1950 FAx Attached are comments from a team of University of Minnesota specialists (651) 297-5615 developed to heip inform the City of St. Paul on their proposed lawn fertilizer [{ MiNNE50TA 80ARD OF ordinance and the Minnesota Senate on their November 15, 2001 hearing on W4TER AND SOIL 1aW11 fe1�.111ZEI$. * RESOUXCES SOU7HERN REGION 261 Highway 15 S New Ulm, MN 56073-8915 PHONE (507) 359-6090 If you would like clarification of these comments or fizrther information, please contact Dr. Carl Rosen at 612-625-811A or crosen(�a,soils.umn.edu. �12.i11C }�011. FAX (507) 359-6018 . � .. � << 0 � -I��a' Comments on proposed ordinances and legislation relating to lawn fertilizers University of Minnesota - November 9, 20U1 Carl Rosen, Extension Soil Scientist, Dept. of Soil, Water, and Climate, Univ. of Minnesota Brian Horgan, Extension Turf Specialist, Dept. of Horticulhu�al Science, Univ. of Minnesota Don White, Professor, Dept. of Horticulrisal Science, Univ. of Minuesota Robert Mugaas, Extension Educator, Hennepin County, Univ. of Minnesota Doug Foulk, Extension Bducator, Ramsey County, Bniv. of Minnesota Ron Struss, Extension Educator, Water Resources Center, Univ. of Minnesota Phosphoms is an essential element required by all forms of life. However, high phosphorus inputs have been linked to degradafion of lakes and rivers through promoting excessive algae gowth. The overall intent of the proposed ordinances and legislation is to reduce the amount of phosphorus entering surface waters and improve water quality. Proposed ordinances and legislation will restrict the use of phosphoms containing fertilizer for established lawns unless a need is indicated by a soil test. The rationale for this is supported by two sound premises: 1) Surveys conducted over the past 30 years have shown that 70% to 80% of the lawns in the Twin City Metropolitan Area have soil phosphorus levels in the very high range and would not require additional phosphorus for oprimal huf growth, and, 2) Applicafion of phosphorus to lawns not requiring phosphorus is a waste of a limited resource. An underlying assumprion is that reshicring the use of phosphorus on lawns will reduce the amount of phosphoms entering surface waters. Unfortunately, the scientific evidence to show that such a restricrion will improve water quality is lacldng. In fact, the pioneering studies conducted by the University of Minnesota in the 1970's suggest that in the short term, use of phosphorus fertilizer on lawns has little impact on phosphorus runoff compared to- the amounts of phosphorus in runoff resulting from breakdown of orgaxuc material such as leaves and grass clippings. Clearly, more quanfitanve research is needed to determine the fate of phosphorus in the lawn landscape and how resh-icting phosphoms fertilizer use for lawns will impact water quality. Reseazch proposals have been submitted by a team of turf gass and soil scientists at the University of Minnesota to deteinune the fate of phosphorus applied to lawns and to define management practices that will minimize movement of phosphorus into surface water runoff. Proposed ordinances and legisiarion may also raise expectarions that water quality will dramatically improve once lawns aze not fertilized with phosphorus. The problem is more complicated than simply restricting fertilizer use and will require a much more integrated approach to improve lake q_uality. In addition to reseazch, educational efforts should be implemented to address all pracrices that affect or contribute to phosphorus runoff in urban areas. UM comments on lawn fertilizer ordinances and regulations Page 7 of 2 �� � Y One fmal comment concerns a loophole in exisring and proposed ordinances that allows for application of organic fertilizer containing phosphorus. Most organic fertilizers have a nitrogen-to- phosphorus ratio that is mucfl lower than the common inorganic lawn fertilizers used today. Since the rate of lawn fertilizer applicarion is based on the amount of nitrogen applied, there will likely be more phosphoms applied when an organic fertiIizer is used tban when a more common inorganic lawn fertilizer is used. Since organic fertilizers with a 0% phosphorus label aze available in Mumesota, the reshiction should be for both inorganic and organic fertilizers. In summary, our comments on proposed ordinances and legislation are: • They aze based on a sound premise that regulaz app2icafion of phosphorus is not needed on most Twin City lawns. • Reseazch Yias not yet shown that restricting phosphorus fertilizer use on lawns will improve lake water quality. • They should be considered as one part of an overall phosphoms runoff management program. Lawn fertilizer ordinances or legislation will not solve the water quality problems of Twin City lakes on their own. • Educafian will be needed for successful compliance and reduction of phosphorus in urban nmof£ • The exemption provided for organic fertilizers is neither warranted nor advised. Thank you for the opportunity to comment. If you would like clarificarion of these comments or further information, please contact Carl Rosen at 612-625-8114 or crosen(c�soils.umn.edu. UM comments on lawn fertilizer ordinances and regulations Page 2 of 2 a De�n Vietor, 12:00 PM 11/6/01 -0600, Re: P in turf runoff X-From : dvietor@taexgw.tamu.edu Tue Nov 612:02:48 2001 X-Mailer: Novell GroupWise Intemet Agent 5.5.5.1 Date: Tue, 06 Nov 20�1 12:0037 -0600 From: "Don Vietor" <dvietor@taexgw.tamu.edu> To: Leslie A Everett <evere003@tc.umn.edu> Subject: Re: P in turf runoff Page 1 of Z O 1 — \\\1— I have attached Word97 files of a manuscript for which 1 wili submit revisions to Assoc. Editor of JEQ next week. The title page, text, and tables are in separate files. The manuscript and references represent the latest work we have (that is near publication) for runoff of P fertilizer and manure P from a relatively steep slope of turf. We applied P fertilizer rates that were refatively large, but comparabie to the farge P amounts observed in soi4 sampfes submitted from urban counties in Texas. If I can be of further help, please let me know. We are very interested in the new urban regulations being proposed for the twin cities. ls Ron Struss our best source for informafion related the new regulations? Don Vietor Donald M. Vietor Soil & Crop Sciences Texas A&M University College Station, TX 77843-2474 Tel. (979) 845-5357 FAX (979) 845-0456 email dvietor@tamu.edu »> "LesiieA. Everett" <evere003@tc.umn.edu> 11/05/01 03:41PM »> He44o Don, Now I've got a request for you! The cities of St. Paui and Minneapolis are in the middle of passing or implementing ordinances regarding phosphorus fertilizer use on lawns. The state legislature is also looking at the issue, with a hearing next week. Some people tell me there is no research data out there to support instituting controls on P fertilizer application to turf. My guess is that there must be some, and ihere should be some as weii regarding P fosses from inorganic fertilizer applied to pasture or hayland as an analogous system. Most current research focuses on manure applied to cropland and pastureland, which wouid not represent the lawn situation well. Ftave you got any references along this line? If so, p(ease send me a{ist, as well as an indication of who else 1 should contact. Thanks much, Les Everett � Gaumanu.doc � JEQtable.doc Printed for "Leslie A. Everett" <evere003@tc.umn.edu> 11/7/Ol O�-���� Response of Turf and Quality of Water Runoff to Manure and Fertilizer J.E. Gaudreau RH. White , D. M. �etor , T.L. Provin and C.L. Munster ' Soil & Crop Sciences Depattment and Z Agricultural Engineering Department, Texas A&M University, Coilege Station, Texas 77843-2474 ABBREVIATIONS DP, dissolved phosphorus; NO3 N, nitrate nitrogen; NHa-N, ammonium nitroDen; PP, particulate phosphorus; TKN, total Kjeldahl nitrogen. . < < 1 �\ _\\�.'�- 1 ABSTRACT 2 Manure applications can provide nutrients and other benefits to turfgrass produetion and 3 unused nutrients in manure residues can be exported through sod harvests. Yet, unused nutrients 4 near the soil surface could be transported in surface runoff and be detrimental to water quality. In 5 addition to measurements of bermudagrass (Cynodon dactydon var. Guymon) turf responses, 6 volumes and P and N concentrations of surface runoff were monitored during evaluations of 7 composted manure applications in turfgrass production. Manure rates that supplied 50 and 100 8 kg P ha' at the start of each of two monitoring periods were compared to P fertilizer rates of 25 9 and 50 kg ha' and an unfertilized control. Two applications of [NHa]zSOa (100 kg N ha" were 10 applied with the P fertilizer. Three replications of treatments were estabiished on a Boonviile 11 fine-sandy-loam (fine, smectitic, thermic Ruptic-vertic Albaqual� that was excavated to create 12 an 8.5% slope. Compared to initial soit tests, nitrate concentrations decreased to 2 mg kg 1 and P 13 concentrations increased aRer two manure and fertilizer applications and eight rain events over 14 the two monitoring periods. The fertilizer sources of N and P produced 19% more dry weight 15 and 21% lazger N concentrations in grass clippings than manure sources. Runoff volumes did not 16 differ between manure and fertilizer sources of P, but average volumes recorded for the 17 unfertilized control were 22% greater than either source or rate of P during the second 18 monitoring period. Dissolved P concentration (30 mg L in runoff was 5 times greater for 19 fertilizer than for manure when rain occuned 3 d after P applications at the same rate. Similarly, 20 total dissolved P losses in zunoff above those of the control were 1.4 times greater for fertilizer 21 than for manure when both were applied in two applications at equal P rates (100 ka P ha 1 y') 22 Under the relatively large P rates on a steep slope of turfgrass, P and N losses in runoff during 23 natural rain events were no greater for composted manure than for fertilizer sources of P. � m z p � _ � � �'�- 1 INTRODUCTION 2 Additions of organic amendments, inciuding composted sewage sludge, can reduce soil 3 bulk density and increase water infiitration rate and nutrient holdin� capacity of soil in turf�ass 4 production (An�le, 1994). In addition, the amendments can enhance turfgrass estabiishment and 5 quality compued to fertilizer sources of nutrients. Aithough costs of haulin? and handling 6 organic sources of nutrients are relatively large (Daniel et al., 1998), the high economic values of 7 turfgrass, including sod, can offset those costs. 8 Despite agronomic advantages of sludge and manure applications on turfgrass, nutrient 9 concentrations can increase near the soil surface (Vitosh et al. 1973, Kin�ery et al. 1994, and 10 Lund and Doss, 1980). After mineralization, accnmulations of manure sources of P neaz the soil 11 surface are transportable as both soluble and sediment-bound P in surface runoff (Vitosh et al. 12 1973, Kingery et al. 1994, Romkens et at., 1973). Similar increases of P concentration in surface 13 runoff were observed as rates of P fertilizer on grassland increased (Austin et al., 1996). 14 Large nitrate-N (NOs-I� concentrations in soil can similariy contribute to losses through 15 surface runof�' In addition, inorganic N in fertilizer appiications is soluble in water and readily 16 transported in water flow over and through soil. Linde and Watschke (1997) indicated NO3-N 17 losses in runoff were largest in initial runoff events after fertilizer applications. Runoff losses l8 decline as fertilizer N dissolves and infiltrates with water into soil (Schuman et al. 1973). Unlike 19 fertilizer N, organic N in manure is released slowly through mineralization and nitrification 20 processes. Slow release of the manure N could minimize the portion of N applied on turfgrass 21 that is transported in water, compared to inorganic N sources. 22 Sediment and associated nutrients aze transported with soluble nutrient forms in runoff. 23 In the case of turfgrass, Linde et. al (1995) reported an inverse relationship between plant density J 1 and sediment loss. Similarly, Gross et al. (1991) observed less sediment loss at dense compared 2 to sparse seeding rates of turfgrass. The relatively lazge plant densities of turfgrass could reduce 3 sediment and nutrient loss in runoff compazed to grasslands used for grazing and fora�e 4 production (Romkens et al. 1977). 5 The use and eaport of manure sources of nutrients through turfgrass sod production has 6 been proposed as a practice for reducing P loads on watersheds containing large densities of 7 animal feeding operations (Griffith, 2000). Sod harvest can remove and reduce P concentrations 8 near the soil surface, but potential losses of P and N after surface applications of manure on turf 9 need to be evaluated. The objectives of this study were: l.) Evaluate turf quality and P and N 10 concentrations of turFgrass clippings and soil in response to increasing rates of P and N in dairy 11 manure and inorganic fertilizer, 2.) Compare volumes and P and N concentrations of surface 12 runofF between manure and inorganic fertilizer treatments, and 3.) Relate the rate and source of 13 applied P and N to losses in surface runoff. 01-11 14 . < 4 O 1 _ 1 2 MATERTALS AND METHODS 3 Plot Design and Treatments. 4 5 Common bermudagrass was established on a slope of 8.5%. The ent'ue plot area was 6 hydro-seeded at a rate of 50 kg pure live seed ha 1 during October, 1997. Irrigation maintained 7 soil water content for seedling establishment and turfgrass growth without runoff. Plot 8 dimensions were 4 m x 1.5 m. Sheet metal strips (thickness = 1.9 mm) were inserted 5 cm into 9 soil around the perimeter of each plot to contain runoff. Runoff of each rain event was collected 10 through an H-flume at the base of each plot into an uncovered, 311-L container. 11 Applications of composted dairy manure and inorganic fertilizer comprised five 12 treatments on the slope of bermudagrass. Three replications of the treatments were distributed 13 along the slope in a randomized complete block design during monitoring periods in 1998 and 14 1999. The treatments were: control (no P), 100 and 200 kg P ha 1 y 1 as manure, and 50 and 100 15 kg P ha I y 1 as inorganic fertilizer. Experimental results were analyzed as a split-split plot 16 arrangement of the experimental design. Two monitoring periods (1998 and 1999) were main 17 plots, nutrient sources were sub-plots, and nutrient rates were sub-sub plots within three 18 replications. A single control plot (0 kg P and N ha i y 1 ) was included in each replication. 19 Dairy manure was analyzed before application usin� methods of the Texas A&M Soil, 20 Water and Forage Laboratory (Parkinson and Allen, 1975). Tatal P and N concentrations in 21 composted manure averaged 5.0 and 15.Sg kg i , respectively. The rates of total P, applied as 22 composted dairy manure, were two times those applied as inorganic fertilizer to compensate for 23 the slow release ofP from manure. The inorganic P in fertilizer was assumed completely soluble 24 after prills were applied on the plot surface. The rates of P applied as inorganic fertilizer 25 maintained or increased ea�tractable soil P concentrations above 40 mg kg' and similar to the P � • C 5 p � _\�\lr 1 levels in more than 70% of soil samples submitted from selected urban counties of Texas (T.L. 2 Provin, Personal Communication). 3 Composted dairy manure was applied at the start of monitoring periods in 7une, 1998 and 4 Mazch, 1999. Each application supplied 50% of the total P rate. Similazly, inorganic P was 5 broadcast at rates of 25 and 50 kg P ha I to the respective fertilizer treatmems before runoff 6 monitoring began during each period. In addition to P, inorganic N(100 kg N ha 1 as [NH4] 7 ZSOa) was applied to the two fertilizer treatments. The N rate for each period was split between 8 broadcast applications before runoff monitoring started and applications 61 and 40 days later 9 during the respective monitoring periods in 1998 and 1999. 10 Turl'grass Responses. 11 Plots were clipped 3.8-cm above the soil surface when turf reached a height of 5 to 7.5 12 cm. The first clipping date occurred 17 d after application of both P sources during the first 13 monitoring period. Piant uptake of nutrients was quantified through digestion, and analysis of 14 clipping samples taken during selected mowing dates. Clipping sampies were dried and analyzed 15 for total N and P by the Texas A&M University Soil, Water, and Forage Testing Laboratory 16 (Feagley et al., 1994, McGeehan and Naylor, 1988). 17 Color, density, and quality of turfgrass in plots were rated visually. The monthly ratings, 18 startin� 5 d after initial P applications, were based on a scale of 1 to 9. Brown turf was given a 19 color rating of 1 and dark green turf was rated 9. The density of an open turf canopy with 20 exposed soil was rated 1 and a closed canopy of tillers and leaves was rated 9. Quality ratings 21 integrated consistency, color, density and aesthetics into a single numerical value. Quality, 22 density, and color ratings near 5 represenied an average turfgrass that could be used for a home 23 lawn or sod production 24 Volume and Nutrient Concentration of Runoff. ��F � < � 1 Total runoff volume was determined by multiplying water depth, as a proportion of the 2 maximum, by the container volume. Daily rain amounts were recorded for natural events at an 3 onsite monitoring station. Rain depth for the 24-h period in which measurable runoff occurred 4 was subtracted from the depth of zunoff in containers. After each runoff event, SOOmL was 5 sampled after mixing the volume coilected in containers of each plot_ The samples were frozen 6 immediately to prevent microbial Meakdown of nutrients within the water sample. 7 The particulate fraction of N and P in the SOOmL samples was removed during filtration 8 through 1-µm glass microfiber filter. The 1-µm pore size permitted suction filtering of the 9 sample volume without plugging by organic and clay colloids and total dissolved P(DP) in the 10 fiitrate could be analyzed through Inductively Coupled Plasma optical emission spectroscopy 11 (ICP). In addition, the glass filter disk and particulate fraction were digested to detennine total P 12 and Total Kjeldahl Nitrogen (TKN) (Parkinson and Allen, 1975). Total P in digests of the 13 particulate fraction was analyzed through ICP. The TKN in the digests and the NO3-N and NHa- 14 N of the filtrate were measured in an auto analyzer. The NOs-N was analyzed using cadmium 15 reduction (Dorich and Nelson, 1984) and the NF3a-N was analyzed colorometrically (Dorich and 16 Nelson, 1983, Isaac and Jones, 1970). The NI-7a-N concentrations were measured in runoff of the 17 first three events in 1998 and the initial event in 1999. 18 A tea analysis was completed for three soil samples taken at random across the 3 19 replications of plots on the siope. Each sample comprised 12 to 15 cores, which were 2.5 cm in 20 diameter and 7.5 cm in depth. The soil is described as a U5DA sandy-loam or sandy-ciay-loam 21 containing 56% sand, 24% silt, and 20% clay. The native soil, a Boonville fine-sandy-loam (fine, 22 smectitic, fhermic Ruptic-vertic Albaqual fl, was excavated to construct the 8.5% slope. 23 Each plot was sampled and analyzed prior to the initial N and P applications and after 24 each monitoring period. Ten to 15 soil cores (2.5-cm diameter and depth of 7.5 cm) were �\����r ,� • 4 7 ��_\\\�' i randomly sampled and mixed to provide a plot composite. Bxtractable P and NO3-N of the 2 sampie from each plot were analyzed by the Texas A&M University Soil, Water, and Forage 3 Testing Laboratory. An acidified ammonium acetate - EDTA was used to estimate plant- 4 available P{Hons et al. 1990) and soil nitrate was extracted and analyzed using methods 5 described by Dorich and Nelson (1984). 6 Statistical Analysis. 7 The Statistical Analysis System (SAS, 1988) was used to analyze variation of turf 8 responses, runoff volumes, and P and N concentrations of runoff and soil among monitoring 9 periods, rain events, P sources, and P rates. Numerical ratings of turf and weights and nutrient 10 concentrations of clippings were pooled over the sampling dates of both monitoring periods for i l analysis. The Generalized Lineaz Models Procedure (SAS, 1988) was used to analyze variation 12 of soil nutrients and of volume and DP, NO3-N, and NH quantities for runoff filtrates. 13 Variation of total P and TKN in particulate fractions of runoff was similarly analyzed. When 14 interactions of effects of monitoring periods with P sources and rates were significant (P=0.05), 15 monitoring periods were analyzed separately. Similarly, when interactions between effects of 16 rain events and of P sources and rates were significant (P=0.05), rain events were analyzed 17 separately. The P rates were treated as class variables in the statistical model. � � 4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 � RESULTS AND DISCUSSION Turfgrass Responses. Applications of N totaling 200 kg N ha I with the inorganic P fertilizer contributed to significantly greater ratings (P = 0.05) of turf color, quality, and density than the other treatments during the two monitoring periods (Table 1). The mean color and quality ratings oftreatments fertilized with inorganic P and N were 20% greater than control or manure treatments. The slow release of available N from the composted manure could have limited N availability to turf compared to treatments feztilized with [NH Similar to differences in visual ratings, the treatments fertilized with inorganic P and N yielded 19% a eater dry weights and 21% larger 1V concentrations of clippings than treatments supplied manure sources of P and N(Table 1). The lack of treatment differences in P concentrations of clippings (3.9 mg g') indicated variation of inorganic N suppiy was the principal determinant of larger ratings and yield of turf fertilized with inorganic fertilizer rather than manure. Soil Analysis. Variation of extractable soil P between the two P rates applied as manure or fertilizer was statistically si�nificant (P= 0.05) on sampling dates in March and June, 1999 (Table 2). The soil tests reveal P accumulation neaz the soil surface after applications of dairy manure and inorganic fertilizer. Large amounts of manure residue near the soil surface facilitate removal of manure-P amounts in turf sod that are much greater than P removed in biomass harvests of other grass crops (Crriffitt�, 2000). The slow release of available P from manure was evident in smaller increases of extractable soil P for manure treatments on the two sampling dates in 1999, despite two-fold larger P rates in manure than in fertilizer applications. The relationship between applied �i -1\ lY k �� 9 01-\t�a 1 and ezctractable soil P is consistent with previous reports (Vitosh et al. 1973, Kingery et al. 1994, 2 and Lund and Doss, 1980). 3 In contrast to P, soil NOs-N concentrations of all treatments decreased from beginning to 4 end ofthe study (Table 1). The decrease in soil IvT03-N is consistent with amounts ofN removed 5 in clippings (up to 36 kg ha and estimates of equal or greater amounts of N in grass parts 6 below the cutting height (Schuman et al., 1973). Although not measured in this study, leaching 7 and volatilization losses could have contributed to losses of appiied N and small NOs-N 8 concentrations in soil (7ohnson et al., 1995, Tennan, 1979). 9 Runoff Votume. 10 Runoff volumes differed significantly (P=0.01) among four rain events during each 1 I monitoring period and the first event did not occur unti160 d after the P applications during 1998 12 (Table 2). Asynchrony between rainfall and runoff measurements during 2 days of a pzolonged 13 rain event resulted in a runoff depth greater than the 24-h rain total for event D in 1998. A 14 portion of the rain recorded for event C contributed to runoff measured for D. In addition, the 15 antecedent rainfall of event C saturated the soil and maximized the portion af rain lost as runoff 16 during event D. 17 In addition to event differences, runoff volumes of the unfertilized control were 18 significantly (P=0.05) greater than treatments that received either manure or fertilizer P in 1999. 19 The average volume of the control was 22°lo greater than volumes recorded for either rate or 20 source of P. Relatively large clipping dry weights and density ratings for the two inorganic P 21 rates, which included 100 kg ha 1 of inorganic I�i (Table 1), were consistent with observed 22 differences in runoff volume. Runoff volume was expected to decrease as the plant density 23 ratings of the turf increased (I.inde and Watschke, 1997). . . io O�-ttta- 1 Nutrient Concentrations in Runoff. 2 An interaction between rain events and P rates was significant (P=0.01) for DP in runoff 3 during each monitoring period (1998 and 1999). In contrast to the initial rain event in 1999, DP 4 concentrarions in runoff for the latter three events in 1999 and all four events during 1998 were 5 relatively small (Table 4). Irrigation during the 60-d period between P applications on 6 bermuda�rass turf and the first rain event in 1998 reduced DP concentrations at the soil surface 7 and limited DP concentrations in runoff compared to 1999. The lazge reduction of DP 8 concentrations in runoff a8er the initial event in 1999 was similaz to previous studies of turf and 9 pasture (Edwazds and Daniel, 1994, McLeod and Hegg, 1984, Austin et al., 1996, Linde and 10 Watschke, 1997). 11 The vaziation of runoff concentrations of DP among P rates, including the control, and 12 between P sources was significant (P=0.05) on seven of the rain dates during the two monitoring 13 periods (Table 4). Significant interactions (P=0.001) between P rate and sources revealed greater 14 differences in DP of runoff between P rates of manure than between P rates of fertilizer for five 15 rain events. During seven rain events, vaziation of DP concentrations in runoff corresponded 16 with relative differences in P rate between fertilizer and manure P sources (Table 4). The DP 17 concentrations ofthe 100-kg rate of manure P averaged 2 times greater than the 50-kg rate of 18 fertilizer P for all events in 1998 and the latter three events in 1999. 19 The differences in runoff concentrations of DP between P rates were largest during the 20 initial rain event 3 d after manure and fertilizer applications in 1999 (Tables 3 and 4). In contrast 21 to seven other rain events, differences in mean DP concentrations of runoff of this initial rain 22 between the two P rates of fertilizer were 3 times greater than differences between the two P 23 rates of manure (Table 4). Similarly, differences in DP of runoff between each rate of P fertilizer 24 and the control were 3 times greater than DP differences between respective smaller and larger ! '., ll ��-��,� 1 rates of manure P and the control. The DP concentrations in runoff from the 50-kg rate of 2 fertilizer P were 206% larger than runoff from the 100-kg rate of manure P for first rain in 1999. 3 In a previous comparison between fertilizer and poultry (Gallus gallus domestieus) litter, 4 differences in DP concentration of runoff were greatest between P sources during the first 5 simulated rain event after application on tall fescue (Festuca arunciinacea, Schreber) (Edwards 6 and Daniel, 1994). Dissolved P concentration in runoff from fertilized tall fescue was 2 times 7 greater than runoff concentrarions after the same P rate was applied as poultry litter. The clipping 8 height of tall fescue was 2.4 times taller than that of bermudagrass in the present study. Yet, DP 9 concentrations in the initial runoff after application of comparable P rates were similar between 10 the studies of ta11 fescue and bermudagrass (Table 4). ll Similaz to DP, NO3-N and NH concentrations were largest in the first runoff event 3 d 12 after the manure and fertilizer applications in 1999. In addition, the N source by rate interaction 13 was significant (P=0.01) for NO3 N and NHa-N in runoff of this initial event during 1999. The 14 large NO N concentration in the initial rwioff of the lazger manure rate in 1999 was consistent IS with 5.3 times more total N in the manure than in the initial applicatian of 50 kg N ha as 16 [NHa}ZSOa (Table 5). Previous evaluations of plant uptake of N during the first year after dairy 17 manure application indicated 21% of the I�i in manure was equivalent to N appiied as fertilizer 18 (Klausner et al., 1994). The relatively lazge I�O concentrations in nznoff 3 d after application 19 of the two manure rates during 1999 indicated more than 21% of the total N in composted 20 manure was in nitrate form. Uniike the firsY rain event, the NO concentrations in runoff of 21 fertilized treatments were significantly greater (P=0.05) than manure treatments during rain 22 events B, C, and D of 1999 (Table 5). Larger NO3 N concentrations and losses in runoff from 23 fertillzer compared to manure or organic sources of N have previously been reported (Edwards 24 and Daniel, 1994, McLeod and Hegg, 1984). � 1 12 01-t11� 1 The variation of NO3-N concentrarion in runoff between the total N rates in manure or 2 fertilizer was significant (P=0.05) for five of eight events during both monitoring periods (Table 3 5). The NO concentration in runoff of fertilized piots was 2 to 10 times greater than controls. 4 Concentrations of NOs-N in runoff from the larger manure rate were greater than the control plot 5 for 7 of the 8 runoff events. Austin et ai. (1996) observed comparable increases in NO3-N losses 6 as fertilizer rate was increased. 7 The initia] application of [NHa]zSO4 with P fertilizer in 1999 contributed to 32 mg L� of 8 NHa-N in runoff 3 d later. Similaz NH4-N concentrations were observed in runoff of simulated 9 rainfall shortiy after N feRilizer was applied to tall fescue stands (Edwards and Daniel, 1994). 10 During the monitoring period in 1998, NH concentrations in runoff (3.2 mg L 11 days after 11 the second [NHa]zSOa application were smaller than the initial event in 1999. Irrigation during 12 the 11 d before the rain event couid have dissolved and transported the NHa-N into soii. In 13 contrast to observations after fertilizer appiications, NHa-N concentrations in runoff shortly after 14 composted manure applications in 1998 and 1999 were < 1 mg L (data not shown). Near-zero 15 NHc-N concentrations were observed in simulated runoff 14 d or more after poultry Iitter was 16 applied to tall fescue (Edwazds and Daniel, 1994). 17 Nutrient iosses in runoSf. 18 The potential for removing and exporting lazge amounts of manure P and N through sod 19 is an incentive for lazge manure rates that exceed P and N amounts needed for turf growth 20 (Crriffith, 2004). The volumes and P and N concentrations of runoff on the steep slope of 21 bermudagrass provide estimates of potential P and N losses and environmental impacts of the 22 large manure and fertilizer rates on turf. The DP amounts in runoff differed significantly 23 (P=0.05) between rates and between manure and fertilizer sources during 1999. During eight rain 24 events, the 200 kg of P in two manuze applications contributed 7.1 kg ha 1 more DP to runoff � , li 01-1\\� 1 than the control. A similar loss of DP in runoff was observed after two fertilizer applications 2 totaling 100 kg P ha i . The DP losses during eight rain events following applications of lower 3 manure rates totaling 100 kg P ha' were 3.0 kg ha 1 greater than the control and similar to two 4 fertilizer applications totaling 50 kg P ha 1 . 5 The portion of total P in turf cligpings and runoff attributed to manure (controi amounts 6 were subtracted) was only 2.8 to 3.8% of P applied during both monitoring periods (Table 6). 7 Comparable percentages of P in pouitry litter applications were collected in runoff during four 8 simulated rain events on a 5% slope of perennial grass (Edwards and Daniel, 1994). The small 9 amounts collected in clippings and runoff and extracted from soil (Table 2) indicate most of the 10 P in applied manure remained on or in soil and available for harvest with sod. 11 Similar to DP, NO losses in runoff differed significantly (P=0.05) between rates and 12 between manwe and fertilizer sources during 1999. During eight rain events of both monitoring 13 periods, 3.9 kg ha' more NOs-N was lost in runoff from the higher manure rate (two applications 14 of 267 kg N ha �) than from the control. The NOs-N losses in runoff of the larger manure rate 15 were 2 times greater than the lower manure rate and treatments fertilized with 200 kg N as 16 [NHa]zSOa. 17 Losses of NT3 in runoff soon after N applications revealed an advantage of manure 18 over fertilizer applications on tur£ The largest loss comprised 10.3 kg NHa N ha 1 in runoff 3 d 19 after 50 kg N was applied as [NH in 1999. The total NH4-N losses in runoff during two 20 rain events following N fertilizer applications were 2.9 times greater than total NO3-N losses 21 from the larger manure rate during all eight rain events in 1998 and 1999. The NHa-N losses 22 made up 40 to 42 % of total N amounts in clippings and runoff (Table 6). Similar to the NHa-N 23 loss from fertilizer, DP losses in runoff above those of the control were 2.7 times greater for the 24 50-kg rate of fertilizer-P than for the 100-kg rate of manure-P during the first rain event in 1999. � 14 Ol �111�- 1 An advantage of turf in a system for eacporting manure P and N was evident in negligible 2 losses of particulate forms of P and N after surface application of composted manure. Caiculated 3 total losses of PP and TKN after manure or fertilizer applications during the eight rain events in 4 1998 and 1999 did not differ from the control (Table 7). In addition, amounts of PP and TKN in 5 runoff decreased significanfly (P =0.05} in both years after the first rainfall event (Table 7). 6 Reductions in PP and TKN after the initial runoff event of each monitoring period could be 7 attributed to increases in turfgrass plant density over time (Linde and Watschke, 1997, McLeod 8 and Hegg, 1984). The density ratings and clipping dry weights (Table 1) indicate additions ofN 9 fertilizer with manure P could increase plant density and minimize losses of particulate forms of 10 P from turf. Yet, large runoff losses of fertilizer N compared to manure alone could be 11 problematic (Table 6). 12 CONCLUSIONS 13 The slow release of P and I3 from composted manure can limit turf growth and 14 quality compared to timely applications of soluble fertilizers. Yet, the slow release of manure P 15 and N resulted in smaller DP, NOs-N, and NHa-N concentrations in runoffthan fertilizer P and N 16 during rain events after both were applied. At equal P rates, runoff losses of DP attributed to a 17 recent application of P was 58% less for manure than for fertilizer P. Similariy, runoff losses of 18 DP totaled over eight rain events were A4% less for manure than for fertilizer applied at equal P 19 rates. Applications of N fertilizer with manure could increase turf quality and P and N amounts 20 in clippings, but timing of applications in relation to rain events will be critical to prevent lazge 21 runoff losses of N on steep slopes. 22 The surface application manure on turf optimizes potential removal and export of excess 23 P and N during harvest of the sod layer. One disadvantage of the large manure rates was evident 24 in the relatively large DP concentrations and losses observed in runoff, which could raise ,x , 15 0�-��1Y 1 concentrations of DP and accelerate eutrophication in surface waters (Daniel et al., 1998). The 2 observations of runoff losses on the steep slope did represent a worst-case situarion for turfgrass 3 sod production, but it is clear than manure rates need to be managed on a site-specific basis to 4 prevent edge-of-field losses of P and N in runoff. 5 REFERENCES 6 Angle, 7.S. 1994. Sewage sludge compost for estabiishmern and maintenance of turfgrass. p. 45- 7 52, In Anne R. Leslie (ed), Handbook of integrated pest management for turf and 8 omamentals. Lewis Publishers, Boca Rotan. 9 Austin, N.R., J.B. Prendergast, and M.D. Collins. 1996. Phosphorous losses in irrigation runoff � 10 from fertilized pasture. J. Environmental Quality. 25:63-68. 11 Daniei, T.C., A.N. Sharpley, and 7.L. Leymunyon. 1998. Agricukural phosphorous and 12 eutrophication: A symposium overview. J. Environmental Quality. 27251-257. 13 Dorich, R.A., and D.W. Nelson. 1984. Evaluation of manual cadmium reduction methods for 14 determination of nitrate in potassium chloride extracts of soils. Soil Sci. Soc. Am. J. 15 48:72-75. 16 Dorich, R.A., and D.W. Nelson. 1983. Direct colorometric measurement of ammonium in 17 potassium chloride extracts of soil. Soil Sci. Soc. Am. J. 47:833-836. 18 Edwazds, D.R, and T.C. Daniel. 1994. Quality of runoff from fescuegrass plots treated with ✓ 19 poultry litter and inorganic fertilizer. J. Environ. Qual. 23:579-584. 20 Feagley, SB., and M.S. Valdez, and W.H. Hudnall. 1994. Papermill sludge, phosphorous, 21 potassium, and lime e£fect on clover grown on a mine soil. 7. Environmental Quality. 22 23:759-765. 23 Griffith, E.N. 2000. Export of manure sources of phosphorus and nitrogen throu�h turfgrass sod_ 24 M.S. Thesis. Texas A&M University, College Station, Texas. 43 pages. F r. 16 0�-Il I� 1 Gross, C.M., J.S. Angle, RL. Hill, and M.S. Welterlen. 1991. Runoff and sediment losses from � 2 tall fescue under simulated rainfall. 7. Environmental Quality. 20:604-607. 3 Hons, F_M., L.A_ Lazson-Vollmer, and MA. Locke. 1990. NH 4 phosphorous as a soil test procedure. Soil Sci. 149249-256. 5 Isaac, RA., and J.B. Jones, 7r. 1970. Auto-analysis for the analysis of soil and plant rissue 6 extracts. P. 57-64. In Advances in Automated A.nalysis, Technicon Congr. Proc., 7 Technicon Corp., Tanytown, N.Y. 8 7ohnson, A.F., D.M. Vietor, F.M. Rouquette, 7r., V.A. Haby, and M,L. Wolfe. 1445. Estimating 9 probabilities of nitrogen and phosphorus loss from animal waste application. P. 411-418, 10 In K. Steele (ed), Animal waste and the land-water interface. Lewis Publishers, Boca 11 Raton. 12 Kingery, W.L., C.W. Wood, D.P. DeLaney, J.C. Williams, and G.L. Mullins. 1994. Impact of �/ 13 long-term land application of broiler litter on environmentaliy related soil properties. J. 14 Environmental Quality. 23:139-147. 15 Klausner, S.D., V.R. Kanneganti, and D.R. Bouldin. 1994. An approach for estimating a decay 16 series for organic nitrogen in animal manure. Agron. J. 86:897-903. 17 Linde, D.T., and T.L. Watschke. 1997. Nutrients and sediment in runoff from creeping � 18 bentgrass and perennial ryegrass turfs. J. Environmental Quality. 26:1248-1254. 19 Linde, D.T., T.L. Watschke, A.R. 7arrett, J.A. Borger. 1995. Surface mnoff assessment from ✓ e 1 � �- 20 creeping bentgrass and perennial ryegrass turfs. J. En�ironmental Quality. 87:176-182. 21 Lund, Z.F. and B.D. Doss. 1980. Coastal bermudagrass yield and soil properties as affected by 22 surface-applied dairy manure and its residue. 7. Environmental Quality. 9:157-162. 23 McLeod, R.V. and R.O. Hegg. 1984. Pasture runoffwater quality fromapplication of inorganic . 24 and organic nitrogen sources. 3. Environmental Quality. li:122-126. n, � 1� o1-11t�- 1 Parkinsoq 7.A, and S.E. Allen. 1975. A wet o�cidation procedure for detemunation of nitrogen 2 and mineral nutrients in bioloaical material. Comm. Soii Sci. and Plant Anal. 6:1-11. 3 Romkens, J.M.M., D.W, Nelson, and 7.V. Mannering. 1973. Nitrogen and phosphorous 4 composition of surface runoff as affected by tillage method. J. Environmental Quality. 5 2292-295. 6 Schuman, G.E., R.E.Burwell, R.F. Piest, and R.G. Spomer. 1973. Nitrogen losses in surFace 7 runoff from agricultural watersheds on Missouri Valley Loess. 8 J. Environmental Quality. 2:299-302. 9 SAS Iastitute. 1988. SAS/STAT user's guide: Statistics, version 6.03 ed. SAS Institute, Cary, 10 N.C. 11 Terman, G.L. 1979. Volatilization losses of nitrogen as ammonia from surface-applied fertilizers, 12 organic amendments, and crop residues. Advances in Agronomy 31: 189-222. 13 Vitosh, M.L., J.F. Davis, and B.D. Knezek 1973. Long-term effects of manure, fertilizer, and 14 plow depth on chemical properties of soils and nutrient movement in a monoculture corn 15 system. J. Environmental Quality. 2:296-299. 16 Y 4 6�-t��� y 'O O �L d ' M C �i O a+ O � � � � .-+ � G C. � � L d > O 'Q ... O O c N C 0 L d U c 0 V C d .; � C. � a "a A W A �c R • U � _ Q y b0 G Ri 1 � � i Gi.n� i 0 F C.1 � H b0 G . 4 � ; ; 1 �i C � 00 rl � M 7 � M 7 i .,� b0 � �I .�• oo r '» o M M M M M i � ��C ...r 1(� �O M 7 � � T � � � � i N M � N C vi �o �o Vi vi M 00 �O C1 N �A N N 7 �A 00 O� �O N O 7 N V� 7 N � ^� � � � N N V�] � � i � d < i QI L d d � � '� O y � U W � �� o a � L �%J Vl UI 'L^ w C G .. � �3 x � s � � Y � � z x N � � u i, � � O � � Q O •= U � O � L V h 'K N i � �% .� . V � L K II ` w � 7 � C C V Q � L d � � � � � ^^ i # V � U O L 4 O i a� p "�S a�i V ./+ sC � .,I�.�' h � i ��' � � � �" � x • � c° 3 >, K II � o� a a > d 'O b. � i.+ C � C c� p Q Q U R C. C d °i � 3 �� � d z�. U L ~ i! O � d � y ..' ¢ s, ..�.� Il p * p � ,� II .,^. p � O ic x � � t � v i �' � i V1 Qw O ""i '� Q„ y�, 'C �j� d d y � 6� Li �s, � � 3 ri y d 0 L y O � = U O ... c cC .> +' v v' iA � O y � X p � � � � y � R ..- C 'C 'r �y y R a r R� C dD CC y � �a d r. w n K a• c W i •"� h d �a (in i � •� i Fi �C � 'C � 3 �D � 3 w v�i -.'i'+ y w � � � 7 � +..� > 3 � . ..� K K a� r 'O d > Ti CC � � Q� � � � % R Q� a' E; C. G�i �' � �e w � a� � k � � ° .?'s � � °...' a � a �, = � i L O'3 r�i� 6� � 8� O �y �..� E"� � A �p 'X V1 wc �a+ � v �/. 'K^ K ti �, � 1\ 1 }� � d L � � � � U K o � b0 � C � G C � • y Q � � � .� Y � d � d � "a C .� £ � :.+ � C � O O £. w d � +.+ Q L � r. R: �' N C N � O � .� a+ � � a � �. U O 4 � w' 0. pD s�, r a +.+ a •O � L'.. N CJ U :-. O w £ a � � L p r C � R' i Z i y � Y � _ � � � N �C '�^ � d K s. � � d � F� a .. � H � !'r� !"I .�-i N N � � � � � rl �i 00 F, � N � � i i i � � � � i � � i � � � � 1 + .r L bD � N V�J a, x t. d O � L �. v y r" i O c� w � � � � Y�1 .-i N � � N N �' � � 1�l1 O .-� � L '3 C R' � y � h .ri 'X ri w * x x ,*� x .Y �' y .K ,�. o' x ri' C C y N H � � �i H N y ri {�i Zi d � r. L y e " o `� ;; a Q � ' L �a � '} � '3 U1 CSr V� I� d 7 � � � � `O a C y O L �..� � � � O � O ':, O L G:: C U R �C. �-1 R� O L � � � G O L p da L ,� � R G �„� � c U d .��' �+ p y Q �M E � U V � r, +� � d a �a, � a x � t ,.�+ � � �. F. � a � �- � 1 .� • _l. � � �,,, � � a � c r, � a '-I � � o � � r � ta � r'�.i '� 1 N 7 7 O� � � M �-+ N tV � � � 7 A � ai ' � � � 7 a � O�0 7 r N o, o � ,.r .-� ,� � � � � L � �. � � �.y w U � N M � � a i R � a „ , £ � M N 7 � H � w � y a = � N � N � � � ' +-' � r � �o 0 o A V eV O tV M C � fY. 6 � N in r e� � '3 y � �O N VJ DO M � ; e�+ .�-� �-1 � � � a i .. ,_ � �"' N � �--I V1 d £ e}' �-+ V� N p A i � d i E 'O O N � f H i � r � � w d � � o W �.� � Q� C� U A A � � V1 O N * � Q � � a c � � � on > c :r CC U �i R 'C y d L � w+ ' .j � O �° p O � � C y � � � �� �o d F° — � o � � O C Py % Vl 3.' p 'p'� _ �i ,�j,. � � •� W+ f.' a � U a � y t;, � L U � C c d �,� O p So�„ � C�C i � U ` R G' m 4 y,� v �. C.^ L: w � � p �*.. K L � x e� y �,,, C s�. +°�,' C >' c y x" o '�' > A = �" �^ vs P, W .a n.,,, 0 5 — \� \'+� � ri �i O 'G C'i � � C .� i. 3 W i d 4� 4 � O C i d w R p� 3 � e a; y a C � � i � .� o` G � d U V C R U � P�V � 'C a v O > •� h � �, c A °D C � 'i. Y d � C'' O H � � � � H � � � rl t � O� N t� �O lV N t7 'Q V er M C� 7 .ti �i ri .-i N R � � i�O N M C h .��i � N M 7 � �C m � � ' N � � M V � � N N M M �(. a bC �i' c�'i i c� a a v: ci H i ti f`1 N M h i m � �` ; v: '+ 'r t : ao e�-i � ey e�i ti ey N c, C� � H � x c. c+ rl 1 .; m r � � . Y r �o o v+ a? � �G C h C1 e-1 C�f ao a v v a ti ti tV 1V 'S � N V�i h O ey U O � C � U W � � . t x s .x �k x K X ; � i % • i x � � x • x + + x � * x • �' C � a : � � s x * W � � L = x • x « � x K # i . « � : . � x K � � � Y ,. L � L � ri C1. r d � Y Y �i d i "3 C � � � O a) > r. O v 7 q O y " i L d '-I N p � O � O � � y � R � ri � � �Q O R V� CC O H � e A; i y 0. � d � C � o � � �i d Q Q •� � � •y y C 1 �L .� � � C. � � C � O � V � �' R' + � vs u a�i � .�" � '^- x � y a+ Y� L Q �% E"� a+ +i �n %� � " • .4 b1-V\��- a 0 .� d a b� C . O w .� � � bD G .� � "'a w fJ d d w C. .� L L � w 6. O � O O O i C C O � G i w C d U L' O U Q� v G1 L ti z G R � � V � � ti h .__ � c� � � x a c� .�r x x r � e� o .-7 .+ o .-i .-i v, v� o e� O O O O C a � � I� N M V � ti C N N O ti i m ; ..^' 1 � � b d' � rl j C H 'rl C C a bL � c� � � 7 .y N N ef H � p ti eti C C i � � � N 'eT M M �'1 .ti � C N tV C O � � c� ti x a .ti �O M N 00 �p o v .-i Q .-i '. � M M N t7 C C o C C � i t� S t � � � O C M � z 'Y 'r ey '-1 N L h y L +" ' � ^j = • i � v w � y * � 2 % * * d y K * q X i' y � � G � � � e C.� • � � • s ,* x x x • • � . • � � y � C G a U d U L i L � z � � � � •� � a on '" � d i �+ ° � •� L O '-i �T. p U � d L' L K w ri C O O O i+ � V O � � � w , � r. '3 Li C R. � � N C C bD O 'y L � O � i CC N � ... a+ � ^ � � O O �� � � N 'X � x *� � z x� r ,, 4 � \ �\��}� L I OI N � C �L � � w L � « N N i OI � � � � L N m � O w O C 7 L W 0 N C O U l�0 w d Y � 7 Ci F. A C 'a C R � > 0 N N � C � R N � C . G V C � w t� Q j d � O � V � a � a R C pp R � Z � R C F +" C � � N N Q C R F � x N N J � e C d V d a N y N /0 y � L C t3 � Y R z m N 3 O a. L y Q R O J ' L O1 Q. w Y O H R r L R � � Y X N N O J � e C d i a N 41 R J L Y � d Y Z N N � O N J � � r Y � O m ~ � N � � O �„ R Z R' L +�-� Y � d L 3 O � Q ? tn a0 a0 6i o� N M � r � W � � c0 N I� � � � ti r � � � � � O O O O � O O N c'1 7 0 N � � � N � � N � � � � u'i N N I� N N N rn N � �1 N N 0 0 0 � c N N N O N O y U tL d 3 C N � O � v o � a � �o. U � �" . � R t6 � c 6f � C � O � � C � •L � L W � � C � � � a � N y C O C O � � .� O. y c6 V _ � C� O Q Z U 'Q m w a O � N C V � N N O � ."' C � N U C E O_ U • Q •O - O � �p_ Q N N � � N N N +- 'C p_ O O . N � � C 3 p N N O O - ' w � C C 3 3 � _ � C C � � � O O � . Q a Q (� � Z a �J « m . r-�... 0�-\�\s- Table 7. Total P and total Kjeldahl nitrogen (TKN) of particulate fraction of runoff for four rainfall events during monitoripg periods in each 1998 and 1999. Runoff was collected from Yreatments comprising two rates of either composted dairy manure or inorganic fertilizer before runoff monitoring began. 1998 1999 Event Total P TKN Total P TKN ----------- �g Plot 1 ---- ---------- mg plo£I ----- � 57 B 14 C 28 D 41 # MSDo.os 15 265 46 73 152 43 62 34 37 24 16 349 165 180 124 89 f Miniumum significant difference within columns using Tukey's Studentized Range, P=0.05. `' — _ Nutrients and Sediment in Runoff from Creeping Bentgrass and Perennial R` Turfs Douelas T. Linde" and Thomas L. �Vatschke ABSTRACT Althoueh scientist> fia}�e found 4+tde tcanspoct of nutrienu to date in runofF (rum turF_rasses. more research is needed on a x�ider range oEsuil conditions and mans�ement scenar�os. Thii studp ��as designed to as>e>s nutrient and sediment tr�nsport from creepin� bentarass (4grostis pa(usiris HudsJ and perennial rpegrrss (Lolium perersne LJ turG and to asse>s the influence that .�ertical mo..ing had on sediment tmnsport Sloped p�ots of ben[�rass and n e�mss. maintained simitnr to a golf f�irr�a�, ��cre ircisated to forte mnoff Cor the venera- tion oC mnoPF and leachatc �+�ater samples. About 1? h before each runoff e.�ent, iaigation •�as used ro equilibza�e soil moisture for ail plo[s. Foc four erents, pbts ���ere tre�red xith (crtilizer ai a rate of }.9 g N m'. 0.3 g P m and 4.1 g iC m'' about 4 h aFter pre-erent �r��gstion and S h before runotF. For another Four e�cnts, plots »�ere verticut 6 h bePure runoll- «�ter ssmples »ere anskzed fur NO tot�l I:je�dahl-� (T6ti), ph�sphate, and sediment. p]ean \0-�: wn- centrations rarelr ecceeAed 1 mg L Phosphete and TF1 concentra• tions and losses sieniticnnd} incre�sed �+hen runoll �+�s Forced 8 h after fertilization. On a�ersee fur these c��ents, ll % oC �pp��ed P and 2 applied ti»as delectvd in runull and IS % applfed P and 3 applied A' ��'as detected in Ie�chate. Fur all other e�ents nutrient concentrations and los>es e'ere consistentl} lo��er. ��ertical mowine h�d little afFect on sediment transporL 5ediment transport Crom both turfs a�eraged O.S k� ha On golF Eairwu��s, oll-site mo�'ement oF nutrients mar happen if rvnoff occurs soun aEter granular fertilizer is applied tu a neady saturated soil. I � THE RESEAFCH CO d3i2, S :izntisti hace found that nutrient transporc in runoft from turfgrass is small (blortor. et aL. 19SS; Gross et al.. 1990: Harrison et ai., 1993: Linde et al.. 199-'.)• Ho«'e�'er. additional research is needzd on a l��ider ran�e of soil conditions and man- a�ement scenarios beEore �ny generalities can bz made about surtace transporc of nutrients from golf courses. Harrison et al. (1993) found that conczntra[ions of NO ti and phosphatz in runoff and leachatz from turf maincained like a home lawn rarely excreded � and 2 mQ L respectively. Using crzzpme bent�rass and perennial ry'e�rass turG maintainzd Iike a QoIE fainv'ay. Linde et aL (1994) rzported that concentrations of \p—\, phosphate, and TK� in runotif and leachate rarely exceeded 7, �, and 2 mo L respecti�'aly. In fact. nutrient concentrations and losses (presentzd as loading rates) usually reflected those Eound in the «'ater used for irri�ation. For rcnoff ecents wichin ?=� h follo«�ing ierti(ization. the runoft contaiaed an avzrage of 1% oE the applied � and 3?% of thz apptied P for both turfs. The leachate contained an averagz of � 2% of applizd iv and 1�3°io of appliz:l P for both turfi. Linde et al. (199-i) concluded that for similar condi[ions on a eolf D.T. Lindz. Dep. oC A?-�onomc and Emiron. Scizn�'.. Dtlawara �'af- Icc Cufk?e. l0U E. Buder Ave.. Doy(zsto��a PA 1S90t: and T.L. \\�atichk:. Dep. o[ Ao•000my, Prnnsyl�a�ia S:ata Univ.. LL6 ASI Bld,_ (:ni��ervtc Park. PA 1650?. Receivcd I1 Oct. 1996. *Cor.z- sponding au[hor. � Publi>hzd in J. Enciron Qual. 26:12�1S-t?i (1997). 0 � -�\\d— fair«'ay�, it «ould be reasonable to assume that little off- site tran�port of nutrients from the fair�ca�' ��ould occur as a result of fertilization. Linde ec aL (199<} had ro limit s[atistical anat}'sis to individual dates bzcause major soil moisture difYerences esisted bzn�zzn e�ents and be- t«•ezn turf species. In the Simixed studies on erosion from turfgrasszs, scientists have found that turfgrasses �rzatic rzducz ero- sion compared to bare soil (�Vauchop� et a1..1990; Gross et a1.,1990,1991). Usin� a mised stand of bermuda�rass (Cynodon dacrylon L. Pers.) and bahiaarasi (Paspalum notnnun Flug�e vac. suare Parodi). «'auchope et al. (1990) found that thz a��era�z soil loss for simutated rainfalts a[ 69 mm h was 28 k¢ ha ` for bare p(ots and 3 k� ha ' for grassed plots. Using slopzd plots of tall fescue (Fesu�ca arundinacea SchrebJ. Gross et al. (1991) reportzd the a� erage soil loss for a 30-min, 120-mm h intensity scorm was 519 k� ha for bare soil and 54 kg ha`` for mature tall fzscue seeded at 4SS ka ha"'. Gross et al. (1991) concluded that even lo«' dznsity turf stands could si�nificantly reduce erosion and a«'zll-maintainzd stand should not bz a si�nificant sourcz of sediment. The studias conducted by `h'auchopz et al. (1990) and Gross et al. (1990,1991) used turfs maintained at heights >S cm. No published studies wzre found that included inEocmation concernin� soil loss from turfs maintained similar to a �olf fair« ay (about 13 cm heieht). In addi- tion, no studies were found that pro�ided inEormation on the influence that vertica! mowing for thatch manaee- ment had on soil loss from turfgras�. Since �'ertical mo«'- ine for thatcn management typically resul[s in the physi- cal removal of oraanic matcer and shatlo«' grooves in the suil. it is possiblz that soi( loss mac increasz. Crezping bentgrass and perennial rcearass are t«•o turf�rasses commonly used for �olf couriz fairways in the tamperate ciimate regions of thz li.S. Perennia! ry'egrass is a medium-tezturzd, bunch-ttipz species that does not form a dzfinite thatch layzr. ��'hen closel}' mowed, perennial rye�rass forms a turf «'«h a shoot densicy' bzt��'een 100 to 200 shoots dm (Beard, 1973). Crzepin� bent�rass is a fine-tzxtured, sroloniferous spe- cies that forms a definite thatch la}�zr. �Vhen closely mowed, creeping bentgrass forms a turf �cith a shoot densin >200 shoots dm '- (Bzard, 19i_). 7'he objec[ives of this rzsearch «'zre (i) to assess the transport of KO N, TKN, phosphace, and szdiment from the turfs and (ii) ro determine the influencz that vertical mo«�ino for thatch mana�ement had on sedi- ment transport from the turfs. DiETHODS AND DIATERI�LS Sic established tucf runoEE ptots ustd b}' Lir.de et al. (199-' anS 199�) werz used for this smdy. In 1991, tF.ree piots (each 6� m«•idz by 19 m lona) «'erz establishzd ro'Pznneaslz' crzzpin� bentgrass and thrzz to a perennial r}'e�rass blznd izas n � _���� LI�DE & w��.TSCHKE: RC�OFF FRO>t SE�"IGRASS A�D RY'EGRASS Tl'RFS 1249 ('Cita:ion II'. 'Commar.dzr'. 'Ome�a II'). Hacrison et al. .-eTticat mo�cing. Flots rzce:� thz usual pre-even: irri�ation (1993} characczrized thz sue as having a•�ariablz slopz be- to equilibrate soil moisturz. It µ�as hcpothesized that the rz- na-ezn plots (9-il% 2ad a sudace so�t that «"as a se�'erzlv bl�dzs. aad the oriznt tion of [he� b�omn �h erodzd Hzazrsto�.�n serizs classitizd as a clay (0.23 kg ke . sand. 0.36 kg ko ' siic. 0.41 ka i:g '��a�). The local seologp slope ticould form preferen.ixi f!o�v channels for runoff and �cas a frac[urzd karst and depth to bedrock ranged from � to se Rur.oif anater san pl s�iere tak n an �'fodei 60 cm. At the bzginains of this stud}' in 199�, the top S em _ �Q� orjable «zrer samplzr (ISCO. Ine.. Lincotn. �E). To of soil had a pH of 7.1. P iz��el of Si ks p ha '. and K lecel F of 26"I kg K ha Plots «'e:e mowed ro 13 mm ���ich dippin�s eser2et a runoff samptz. ihz potyeih}'lenz samplin� access tu z tur�f thac � ould be ca112 found on a olf fa�nat}iY The sampler���as intzrfaced �nith an ISCO p4odel �0 flo� Each ploi containzd 21 R'zathermatic (Garland. TX) pop- mecer chat «�as programmzd so that after zcery i5.6 L of p onvas an �coaced con e t [e he e bo lha m collecez a r�u�oa mL sa ple of runoff «ater in thz splitiing chamber. Thic and direc�eand sam equipmzna \Vatzr the chute r�� utes a d an� a��era z unoffb olamu for� turE pre- mzasuring P floa'ed into a polyethylene splitcin� chambu (for runoff sub- `i�iz eom tzd int c� o�00-mL(boulzs. The 40 mLsamples samplz cotlzction) znd in[o a parti[ionzd steel tank (Harrison from thz f rst 907 L of runoff �verz composited into the firsc et al., 1993). FunofE �olume and rates �tzrz mzasurzd wing ' an I 0 dates ` from J nz 1994 totOctOoberC199?��runoff was bhe co d�botde. Nhen r�u off �aas L hzne ach botde forced w�ich irri�ation at a rate of 139 mm h to generate contained �SO mL of w'ater. Samplin� «'as stoppzd after the thesz irr �atzd e��ents Pe�e cor.ductz [ appco�imate]}'. e e�z Y L, lhz�n zcond concained �SO mLnRunoff aused Z�ck, dzpzndino on �i'eather and availabiliq� of laboc Durin� by rains[orn:s �cas measurzd and samplzd �cith the equipment to�a[natual`siatz ethreoard alzd99�)bzcauseinsthoe,studesrunoffflo�.'ratescausedby' <10 n�pla Based�on`data from aL (199�)�irrigation In lsfl�rms d de ot and samplers forthis� smdy �ere 1[m¢O duration ��'as set at 2� min for bent�rass plots and 1� min to onz. for ryzgrass plots for most ecznts. The purpose of different L,zacha[e water ���as sampled from four pan-lysimeter cotumesfor tYle[toOluiSSSlOC2 tat2LS847}Jlltt�rwaSElO�SQa ea fabz of each (Hahri on et Abo u I> after Fur the 6 and 24 Szpt. 199� events, both turfs �i�ere irrigatzd for�each plo[ t �Ppakin,, q ai an oun[s from of tl�ie for 2� min each. four sam lers. Approzimaeely 12 h prior to each runoff ecen[. the pl�ots "�a�z,.pamplzs «'zre analyzed for NO,-�. TKN, and phos- .eere irri�ated ��'ich a senes of shocc duracion (2-3 mia) irri�a- phate (orihophosphatz) aceordin� ro the procedures described tion sets at a ratz of li9 mm h"' un[il runoff was �'isua4ly one o four depe d b g n thz antzcedeen soil mo sture co t nt ru off arerz uszd to�cal ulat nutnent t o r s s tion runofa for The purpose of thz pr� zcent irrieaciors H•a� to equilibrate acerage of bo[h mrfs. Total runoff from each piot and thz the soii moisture content for all plo[s so that data comparisons ave.age of the wncentr2cions in the first and szmad flon'- bet�iezn dates could bz made. Lindz et aL (1994) did not paced zunoff samptes w'erz used ro calculate losses. The �ol- equilib[a[e soil moismre bzforz irriga[zd ecents, thus they had ume of «�ater that could be held in one subsurface sampler � � 1 L w�as used ro catculatz nutrient loss in leachate. Sedimznt concentrations of the runoff samples �vere detzo- ro limie comparisons to individual dates becausz major soil (-� moistute differences esisred between dates and betwezn [urfs. mined �racimetrically by measurin� the amount of inorgamc For four oE the runotf z�'znts, ptoes a�zre fertilizzd w�ih a �atzrial tra ed by' filtec papet (1 µm diam. pores) aftzr 19-1.3-li.S (N-P-K) fertitizzr (O.�l. Scott & Sons.:�larysville, Fp . filterine the enure kno«�n samplz colume. Filters «'zre place OH) usin, a broadcas[ spreader at the ratz of 4.9 g N m - � p_, o p m and �7.1 g K m about 3 h afcer prz-e��en[ irngation in a o�en at 42�`C for S h and then wzighed. It «as hypo� z 2nd S h prior to thz runoft event. The f:rtilizer contained sized thai the soil dismrbance caused by �'ertical mo«"�no � 0.6°rb NH,-�'. 15 �%6 u;ea-N coated to procide 73 slou �i'ould tikzh� incrzase szdimznt trxnsport shortly after mo«�n° ' relzase �. P dericzd from monoammoniem phosphate. and and thzn decrzase as the rurf reco�'°-rzd. Treamtents (turf species) «erz arranazd in a random�zzd ' K from Ii Beforz fertilizer applica[ions. rmtoff collzction cum letz bluck dzsign ��'ich [hrze replieacions and blocking ' .ezirs locatzd at t4e boctom of each plot «ere co.�zrzd «'i:h bas d on suriacz siope- In a concurrent runoff study b}' Lindz ' plastic to psecen[ =ranules from enterina an}" part ot the weir. 1996 . i: «as dztermined that the pre-e�'znt irrigation procz- dure �cas zffecticz in equi(ibrating the soit moismre contznt ' On sz�"en othzr datzs, supplzmental maintznance N ti�'as ap- � plizd as a liquid or aranular application of urea (46�-f��) a� {or all plots beforz runoft. Thz}' found tha[ [hz a�'eraee soii � 2.4 or �J g� m'. ' columzvic ��'ater cuntznt of the ploU ius[ prior to irrigatzd ' Abeu: 6 h prior to another four e��znts. all plots w'zre czrti- 0 0' in 199?. Thus. ' cuc once using a R}an >tata�cay ��ereical mo��zr \4ode154�3� 3 I`� s�`z` �o m ' arisons�bzc�' 9 een z''a OS °°� izn[ and sedimznc � (Cushman. Inc. Lincoln. �E) [hat had 20. 1.6-mm-��'ide blades p i blades� pznzt atzd the soii appro�an2tel�d3Smm.aPlots az usins a mzasures analbs seofbariance bacau<z � �eere �'erticut Izngth�rise do«n the stope. Thz majority of the rzpzated measurements ��'erz made on thz samz experimznta ' ,:� J�b� � !�: `_,,.. ;n 8 =a�(�rc hv .+- mo.rino �vas units o�zr time. ,� , factor. O ecperiment �va<_ : �;� �,�, t��c .� �_ ni-,., „F t�mr � eolleC[ad by rakin� and ��ei�hed. A�;}>;����°.�lci�� ti Ii boio�� i j7_jO 1. E�VIRON. QUAL. vOL.?6. SEP'i'EMBER-OCiOBER 1997 Tabie 1. �lean n co ncentrations for 199 irrigated e � tiitrnte-.S P h osp hate R unoEF sample+ Runoff sa mple Date FPl FP_' Leacha[e FPl FP? Leachate 29 June 13 ]uh ' p� ?? .�.u�.4 3 Sept. ?0 Sept.7{ S Oct. se 1.0 : \S 0 03 \S 03 \S 0.1 \S OZ 0.1 0.6 O \5 0 C �J 0.3 �5 0.1 r5 0 \5 0.? U.t OS \j 0 \S 0 !�5 OS \S 03 \S 0 \S 03 0.? 1.67 2.�1 \S 2.71 1YS 4.13 1.90 3.85 0.89 0.p'_ '** "" Significant at the 0.05, 0.01, and 0.001 probability� levels, respectivei�'. i FPl = lst �u�s-paced runotT sample, FP2 =?nd itow-paced mnof5 sample. - Mean comparisons are for adjacent d�tes hithin columns. §\S = not significant; SE = standard error. Q Fectilized 8 h beFore e�'eat at �ate of 4.9-03-1.1 g m' oE N-P-K. arran�zd as a spfit-block in a randomized completz block desi�n accordin� to 5[zzi and Torrie (1950). Usin� the Huynh- Feldt epsilon valuzs calculatzd by� thz repeatzd statzment in SAS's �eneraV linear model proczdure (SAS Inst,1990}, Linde (1996) dztermined that levet� of time were indepzndent for the runofi data. Sincz levzts of time were independen[, [hen [hz Ftasts of thz usua( sp(it block analysis werz valid and a univariate analyss w'a> conductzd that provided information on the Izast square mzans and thz probabilities associated �cith thzir compa;ison. Thz prz-plannzd comparisons for this study included comparing spzci;s within days and comparin� conszcutive days within spzcies for thz variables neasurzd. RESULTS AND DISCUSSION Nutrient Transpod Mean nutrienc concentrations, bv evznt, sample type, and }ear are przsenizd in Tablas 1 ar.d 2. The concentza- �������� Total Iijeldahi-ti R un o EE sample FPI FP'_ Leachste mg L I.li O.T 0.0? 0 0 • \53 \S \S � Z.Ol Lli 0.03 O.L' 0.19 �S \S �S �5 1.76 1.13 0.39 0.13 0.0�4 iER 1y ids '� 4.17 1.95 b.3-! � 90 3.93 �}a #Ri s Ysi x 151 0.83 130 OSl 0.86 a. t:. :<s 2.66 ?.4i 5.78 2.53 ?.30 . war 0.61 OAI 0.0? 0 0.06 0.?? O.tS 033 0.13 0.10 tions reported are those detectzd in thz samples minus the concentration found in the irrigation or rain watec for each date, thus they represent the nutrient contribu- tion from the turf plot alone. No si�nificant turf specizs effects werz found on any date for any sample type; [herefore, values for both species ���ere averaged. Sig�ifi- cant date efiects, ho��ever, were found and are pre- sented as comparisons bzt��'een ad}acent dates in Tables 1 and 2. Nutrient concentrations, dzpth of water applied and depth of runoff (Table 3) wzre used to calculatz nutrient inputs and losses. \lean NO concentrations w'ere alwags found to be lowzr than the 10 m� L drinking watzr standard set by the USEPA. This findin� concurs �rith NO concentrations found in turf runoff studizs done by Linde et al. (1994), Harrison et al. (1993), Gross et al. (1990), and blorton et aL (1933). The hi�hzst mean Table? >Iean nu[rient concentrations for 199: irrigated e Ciittate-N Phosphate ' Runoff sample` RunotF sample k D at e F FP? Lea<hate FP1 FP2 Leachate m� L 1 16 �Sm 0.6 �1 U:? 1.?7 09J L01 ' - " rS$ x <. . 31 Afar^,I 13 L� 0.1 9.96 7.67 d.?d x. ... �5 x: . . 11J�me 03 0.6 _ 0 253 1J9 1.2? P$ s.� \S \S \S � ?83unt 0? 0.? 0 133 1.60 0.99 \S n \S .-.. ... .., 1? July¶ OS 0.� 0.1 1�.39 5.51 1.9? �s �s ».. >k. ... .,. ?6 Jul. Ob OS 2.6 219 I.85 1S6 ' � \S *"" *• \S °` \5 6 Sept. 0.6 0- 03 ].95 7.30 Y3i i i ti5 r5 �5 NS \5 1 23Sept. OA 01 0 1.i7 0.93 I.Ol � SE 0? 0.03 OS 0?3 0.09 039 "° $ignilicant at the 0.05, 0.01, and OAO( pcobl6ilit� IereB, tespecti�el.. - FPl = lit flo�.-paced mnotF sample, FP2 =?nd fluw-paced mnotT sample. �'• ` Dlean comparisons are for adjucent dstes x[�p(n c0lumns. 1 I. § NS = not s�gniFicant; $E = stsndard error. f ¶ Fertilized S h befoce e�ent at rate of 4.9-03-1.1 g m oF N-P-Ii. ���: Tot Kei d�hi N RunofF sample FP1 FP2 Leachate 0.�8 S.J2 ♦S� 0.�7�1 a.na 553 o �, 1�5 0.3? \5 0=9 0.09 0.20 3%3 031 s 0 3.23 �r< 0.18 �S 0.06 \5 0.?? 0.09 0.�33 2.00 0.83 • 0.0# 2.60 0.25 C�S 0.37 \S 0.?7 0.26 G�-1\\�- Table3. «'ater applied and mean total runoff for bentgrass and n�egrass. Li\DE g w'ATSCHKE: RU�OFF FRO�t BE�TGRASS A\D RYEGRASS TI;RFS {t'alec applied Date Bent Rye 1993 29 ]une 13 3u1y 14 ]vl.i ai ���: , >? ]uir ? Aug: 17 Aug.: 2'_ Aug. 3 Sepi. 17 Septi 20 Sept. 8 Oct. 1 No�:9 at �o..� Z$ tiov.T 1995 16 bfay 31 Dtav 13 ]une ?S June 6 Julyi 12 July 26 Julr 6 Sept. 23 Sept. ZO Oct3 5& 58 �; is � 58 93 58 58 3tS 5g 58 26 ia 49 58 SS Sg >$ Z.8 58 93 cg 58 93 3j 2% is � 35 93 35 35 38 35 35 26 is 49 35 35 3> :s �g 3> 93 53 <g 93 ; RainC�ll e+used runofi. }fean total runolf depth Bent Rte 16.1 1>.8 0.1 0 ?.1 15.1 o.s 79.1 173 0.8 za.a ?OS 0 o.a 3.2 15.5 18.6 23.5 18.7 I.1 21.1 13.2 17.9 16.9 13 135 1L7 I.9 0.6 33 1�? 6.6 13A 30.8 0.3 15.0 L? 0.8 ia 0.7 SL7 11.'_ 1?.9 11.0 0.6 123 9.6 2?1 19.7 3.2 1�0 N concentration was 2.6 m� L that �a'as found in the leachate samples on 26 July 199�. Runoff and leachate losses of NO N�cere consis- tendy lower than irri�ation inputs of NO� N(Tables 4 and 5). For both years, mean losses in runoff, which ranged from 0 to 0.02 g m'-, �vere lo«er than mean losses in leachate, which ranaed from 0 to 0.16 � m Linde et al. (1994) reported similar findings; however, they described nutrient losses as nutrient loadin� rates. For all samptz types, phosphate concentrations si�nif- i2sz icantly increased for e��ents conducted 8 h after fertilizer application (Tables 1 and 2). Concentrations «erz sie- nificantiv less bv thz nest evznt. usuallv conducted within about 2�ck. The highest mzan phosphate conczn- tracion found in the flo���-paced runoff samples �cas 1039 ma L for an e�•ent conducted S h after fertilization on 2I Ju]y 199�. Escludino the events which fertilizer was applied S h prior, phosphatz concentrations �cere similar to those found by Linde ec al. (199�). Theg applied the same rate and source of P. monoammonium phosphate, to the same turf plots used in this study and reportzd 6.06 mg L as the hi�hest mean phosphate conczntration in runoff. ���ith most concentrations <3 mo L Linde et al. (1994) found little indication in runoff or leachate samples thac P fertilizzr had been applied approximately 24 h before an irri�ated e��znt. The hi�her phosphatz concentrations detected for events immzdiately follo«in� fertilization compared to findings by Linde et al. (199�) could be attributed to hieher soil moisture contents as a result of pre-event irrigation used in this study. The soil was likely wetter (near saturation), thus a �reater portion of the soluble monoammonium phosphate fertilizer couid move off- site in the runoff. Thz a�erage soil volumetric water content of the plots just prior to irrigated events was 030 g k� in 1994 and 0.40 g k� in 199� (Linde,1996). Soil moisture levels were likely less in the study by Linde et aL (1994) because they did not use pre-event irrieation to equilibrate moismre ]evels. Based on soil test results for 6 Apr. 1994 and 23 Au�. 199�, soil P levels in the top S cm of soii �vere �enerally in the low range. In 199�, P le��els averaeed S� k� P ha ' and ran�ed from 73 to 11Q k� P ha In 199�, Izvels avera�ed 73 k� P ha and ranged from 36 to 91 k� P ha `. Since soil P levels �cere low, then excessive levels of soil P�rere not the cause of the increased P transport from the plots. Runoff and leachacz losses of phosphate-P were often Table 4. 199A Chronoloa� of inean inputs and mnofl and leachate losses otrotal N(TKN + 1�0,-\) and 10 for bentgrass and q�egrass. Irrigation and t inp uts Runoll losses Leachafe los Fert. BenL R}' e. Bent. Rye. Be nt. R)e. inpu<s Date of'.� \ \Orti \ \Orti N 10r\ 1 \Or\ \ \OrV \ \O�-\ 7 June 3.7 29 June 0.31 0.?8 0.19 57u1y 2A 53 Sutv 0.}i 0.3i 0?0 13Iuk"r 0.03 0.0= 0.03 _'1 Juhi 0.05 0 O.Oi Z' Jul�fi O.OI 0.01 0.03 ? Au�. 0.31 0.38 D.?5 b �ug. 2.3 17.4ug.� 0.01 0.01 0.03 22 au� 9.9 0.?7 D.?7 0.76 3 Sept. 0.?S 0.28 0.17 17 Sept.�= 0.07 0 0.07 ?0 Sept 4.9 0.?6 0.26 016 S OcL 0.?7 0.?7 0.16 9 Oct. ?.4 I \o.�.�: 0.01 0 0.01 21 \ov.i� 0 0 0 23 \ov.>: 0.01 0.01 0.01 T Rainfall caused runoR. _ Leachate samples xere not coliected. 0.17 0:20 0.0? 0 0.01 0.23 0.01 0.16 0.17 0 016 0.16 0 0 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0 0 0 0 0.01 0 0.0? 0 0 0 0 0.01 0 0.0? 0 0 0 0 0 0 0.09 0 0 0 0 0.01 0.01 0.05 0 0 0 0 0 0 0.03 0 0 0.01 0.01 0.0? o.ii o o.io o.oi o.�� 0.01 0 0.01 0 0.01 0 0 0 0 O.11 0 0.06 0 0.03 0.01 0.01 0 0 0.0? 0.01 0.13 o �+_i 0.01 0.05 0 0.09 U.02 0.01 0.03 0 0.01 0.09 0.03 0 0.13 o.aa O.OI 0 0.01 1��� J. ENV(RO�. QUAL. VOL.26. SEPTE?��BER-OCCOBER 1997 r O�-���a- 'I Table 5. 1995 Chro n o ioa,v of in ean inp and runuFf and l eacha te losses of totnl ( TF\ + NOz-ti) and ti Or� for bentpctcs and q�egrass. Icrig a[ion and xainfali input R unoE F l oss e s Leachate lozses Fett- BenL Rpe. Bent. Rre. Be R ve• inputi Date oE\ � �Or� � �Oy� � �O,-� � �Or1 � \Or� � 1��r� gm : 16 Diac 0.?9 0.� O.IS O.li 0.01 0.01 0.01 0 0 0 0 0 22 JIa�' 37 3Illa�� 3.9 0.?J 021 01> 013 0.09 34lune 0?0 0.?0 01? 012 0.03 ?3 June 0?6 0?? 0.15 0.13 0 6luh�; 0.01 U.Ol 0.04 0.01 0 12July 4.9 031 018 01S 017 0-09 26 Suh 635 0.?L 0.1> 013 OAS 31 Aug.?-3 O.L' 0.01 O.L' 0.01 1 Sept. 4.9 6 Sept.II 0.17 0.13 0.17 0.13 0.01 2a SepLn 0.20 O.1S 0.?U O.1S 0 ? Oct. 4.9 20 Oct.'s_ 0.01 0 0.0� 0 0 � R�inCall evused mnol£. _ Leachate ssmples eere no[ collected. § Lightning disabied runoff inea�uring equipment. ¶ Both turFs were init�ted at same duration (?5 min). greater than irri�ation inputs (Tab(es 6 and 7). For both years, mean P losszs randed from 0 to 0.06 g m' for runoff and 0 to 0.07 g m for leachate. Since the hi�hest losses were found for events that had P fzrtilizer applied S h baforz runoff, then a portion of the applied P was transported in runoff and leachate. For esample. on 22 Au�. 199=4, an avera�e of S% of the applied P was de- tected in runoff and 7% in leachatz. On 12 July 199i, an averaoz of 17% of the appli�d P was detected in runoff and 20% in leachate. Linde et aL (199�) found mean P losses (repo.ted as ]oadino ratzc) ran�ed from 0 to 0.01 � m for runoff and 0.01 to 0.04 g m for leachate. Mean TKt�'� concentrations followed a simitar pattern as phosphate. Conczntrations sionificantly increased for events conducted 8 h aftzr ferti(ization (Table 1 and 2). blean TK\ concentrations range3 from 0 to 6.84 m� L'' in 199-4 and 0 to 5.�8 me L in 199�. The higher concentrations werz dzrec[ed in the first flow-paced run- Tabie 6. 1994 Chroaoloe,v of inean inputs and runoff and leachate losses of phospha[e-P for bentgrass and q�e�rass. Grisution and rainF�}� Runoff Leacfiate Fert. p�nput losses oP P los5 af P inpu[c Date o(P Bent Rre BenL R�e. Ben[. Rye. a m . 29 June 0.0? 0.01 0.01 0.01 0.01 0.01 13Juh� 0.03 0.0? 0.01 0.01 0.01 OA2 11]ul��� 0 0 0 0 0.0? OA3 ZI Suiv` 0 0 0 0 0 0.03 3? Juii�: 0 0 0 0.01 0.01 0.03 Z:1u�. 0.0'- 0.01 0.01 0.01 OAl 0.0'_ 17,aug.� 0 0 0 0.01 0.01 O.Dl 32 Aug. 0.3 O.OZ 0.41 0.03 0.02 0.0t 0.03 3 Sept. 0.02 0.01 0.01 0.01 0.01 0.01 17 Sept.:_ 0 0 0 U ?0 Sept 03 0.0? 6.01 0.03 0.01 0.03 0.03 S Oct. U.Oi OAI 0 0 0 0.01 1 �us.`_ �) U 0 0 21 Nu�.*- 0 0 D 0 38 \o�.�_ 0 0 0 0 i Rainfall caused mnOtT. `y Leach�te sampies were not coilected. 0.0? 0.07 0-Ol 0.06 0 0.10 0 0.0'_ 0.01 0 0.0.1 0 0.03 0 0 0 0 0 0 0 0 0 0 0 0.01 0.07 0.01 0.09 0.01 0.1? 0 0.01 0.01 0.01 0.06 0.05 0.17 0.16 0.01 0.0? 0.01 0.06 0.02 0.07 0.02 0 0.01 O.�I 0.01 0 0.01 0 0 0 0 off samples for events immediately followin� fertiliza- tion. The TKY ]evels �cere often higher than levels found by Linde e[ aL (1994). They reported that TK?� concentrations ran�ed from 0 to 3.� mg L Like phos- phate, the higher TIiN levels for this study �rere attrib- uted to hi�hzr soil mois[urz contznts as a result of pre- event irrigation prior to fzrtilizer application. The TKN and NO N results w'ere added together for estimates of total I�'. In both yzars, despite relative increaszs in total N losses for events that fertilizer was applied 8 h before runoff, totat A' losses remained consis- tentty lo«•zr than irrigation inputs (Tables 4 and 5). ln 1994, mean total N losses ran�zd from 0 to 0.11 g N m''- for runoff and 0 to 0.21 g N m for leachate. In 199i, mean losses rangzd from 0 to 0.09 g N m for runoff and 0 to 017 � N m'� for'leachate. Linde et al. (199=!) repocted similar numbers; however, total N losses �vere szldom greater than NO losses. In the currznt studv, totaf \ losses were much hi�her than NO r 2able 7. 1995 Chronolo�y of inean inputs and runoRand leachate losses oF phosphare•P For bentgrass and ryegrass. Irri�ation and cainfni( RunoR losses Leachate Fert. p� oF P losse ot P inputs Dafe ofP Ben1. Rye BenA Rye. BenL Rre. gm' 16 �[a� 0 0 OA1 0 OA1 0.01 31 �ta. 03 0 0 O.Oi 0.03 0.0> 0.06 li June 0 0 0.02 0.01 0.01 0.01 283une 0 0 OAl OAS 0.01 O.OI 6 July�- 0 0 0 0 i? Julr 03 0 0 0.06 0.0� 0.05 0.07 26 Julr 0 0 0.01 0.01 0.02 0.0'_ 314ue.>;§ 0 0 1 Sept. 03 6 Sept11 0 0 0.01 0.01 0.02 0.0' ?15rpt.Q 0 0 0.01 0.01 0.01 0.01 ? Oct. 0.? ?�) Oct.+= 0 0 0 0 i Rainfull caused runoff. - Leachate samples aere not colle<ted. � L�ghtning disnbied runofE me:uuring equipment. �� Both turts were irrivated at same duration ('_5 min). � L1SDE 8 WpTSCHKE: RL'�OFEFRO\f BE�TGRA55 A�D RYEGR:ISS TIiRFS losses for e�'ents that had fzrtilizer applied, thus a por- tion of thz applied �. albeit small. «�as transported in runoff and lzachatz in forms other than \O_ N. For esample, on 22 Aug. 199�. an a��erasz of 2% of the applied \ �cas detzcted in runoff and about 3% in lea- chate. Gross et ai. (1990) also found that ` losses �cere areaczst uhen runoff occurrzd soon after fertilization of a tall fescue!Kentucky blue�rass turf. On 11 datzs durine the study period (�fa;� 199a-Nov. 199�}. dztectable amountc of runoff (>0.6 mm h oc- curred due to rainfall. Runoff caused bv rainfall often did not pro��ide complete data sets because runoff did not ahcays occur on all plots. Thzrefore, mzan nutriznt concentrations for rainfall events «�ere based on an avzr- aoz for both turfs and thz number of plots that provided data. Statistical analysis was not conducted. On 31 Au�. 199�, rainfall produced measurable runoff from each ptot. ho«�e�zr. li�hming disabled the runoff ineasurin� dzr'ices. �4ean nutrient conczntrations in runoff from the rain- fall events (Table S) were slightly higher than those found in the flow-paced runoff samples from the irri- �ated e�ents ��ithout fzrtilization because thz volumes of runoff �� ere much less from the rainfall evznts. A'utri- ent losses. ho�cever. were much lower for rainfall events than irrieated evenu (Tables 4, �. 6, and 7). ConcenVa- tions in leachate were similar to those found for irrigated e.ents «•ithoet fertiiization. No indication of fertiliza- tion «�as e�•ident in samplzs for any rainfall evenL Linde et al. (199�) also reported'nigher nucrient concentrations but ]ower losses in runoff from rainfall events. Thz data from the rainfalt events is ntore reprzsentative of what ��ould likelv occur on a eolf course fairwav since data from the irrieated runoff e��ents that wzre preceded b� pre-event irrioation would represent a worst case scenario for runoff and nutrient transport. Five da}s before the irri�ated event on 6 Sept. 199�, plots «e.e fertilized at a rate of 4.9-03--4.1 o m'- (N-P- K). Immedia[ely foilo�rino fzrtilization, 9 mm of water �vas applied to each plot. In addition, plots werz verticut 6 h after the normal pre-event irri�ation procedure. Despite this mana�emznt scznario. nutrient concentra- tions remainzd ]ow for the e��enL Unlike previous fertil- Table 3. �Iean nutrient concentrations in runot'f and leachate samples for evenfs ihat rainfail caused runoff. \inate-\ Phosphate Total Kjeldahl-\ D�te Runot7 Le�ehate RunotF Leachate Runoft Leachate mg L 79Y3 137uh� 7.6 03 a79 135 0.7? 0.'_3 ?1 Julr: 15 1.1 6.06 1.10 0 0 ?? Jul. 0.7 0.6 6.00 ?.91 0.�? O.L 17 Au�.� 11 1.9 ?.7S 1.00 1.7> 0.21 17Sept: 1.9 ?.i0 1.09 1 \ov.:= 1.7 3.33 Q.90 ?1 \o..i_ ?.6 2.60 0.87 ?3 Xoc_ LO 1.60 0.06 199> 6 ]ulc_ ].6 6.61 0.�6 }1 Au�.= 2.0 4.?3 0.96 _'0 Oct.- OS 1.8_' 0.06 ` aIl repiications did not produce a sample. - Leachate sampies uere mt coliected. O�_�\\� lzs� izations conducted 4 h after pre-event irri�ation and S h before an ecent, it �eas not ecident in the water samplzs that fertilization had occurred. On ;olf course faira�ays, off-site movement of nutri- ents may happzn if runoff occurs soon after application of a�ranular fzrtilizer to a nearly saturated soil. To pre�•ent this scena; io, ��arious manaQement practices can be implemznted by the mrf manaser. One practice k'ould be to a��oid applyinQ fzrtiIizer on nearlv saturated soils, especiall�� �vhen rainfall is expzcted shortly after appiication. hno2her practice �vould be to vrater-in the fertilizer �cith li�ht irri�ation shordy afcer fertilization. Othzr practices to prevent nucrient transport in runoff include fotiar appiication of soluble nutrients and the usz of fertilizzrs thai have significant slo«�-release char- acteristics. Sediment Transport Since S3% of the 237 runoff samples analyzed con- tained no measurabie sediment, statistical analyses u�erz not conducted. Also. no consistent difference was evi- dznt benceen turf species. Therefore, sediment concen- trations from both turfs were averaged for each type of runoff sample on zach date (Table 9). Concentrations were usuall}� hiehest for che first flo�c-paced samp]es and lo�vest for the second flow-paced runoff samples. \Vhen sediment was measured, the amount was very small, evzn after certical mowine down the slope and removin� an average of 67 k� of organic material from the bentora=_s plots and an avera�e of 20 kg from the :yeerass plots. The highest observed sediment concentration for all runoff samples ���as 23� mg L detected in the first flow-paced sample of a rye�rass plot that produced 1675 L of ranoff on 29 June 199�. That 123.5 m� plot was irri�ated at 139 mm h for 1� min. To calculate the potential soif ]oss for that plot, sediment concentrations of both flo�v-paced samples were averaoed and equaled Table 9. A1ean sediment concenirations for irrigated erents. ntean sediment concentntion Runoff samplei D FPl FP2 mg L 1991 ?9 June 68.7 0 13 Juk S.8 3-9 ? Aug.' 0 0 2? Au�. 0 0 3 Sept.; 9.6 0.6 ?0 Sept. ?7.9 12.5 S Oct. 0.6 i-d 1995 16 �fav 0 0 31 >Ia� 0 0 li June_ 0 0 28 June ?.0 1.S 12 Juh 0.7 0 26 luh 0 2�� 6 Sept= 1.7 0 ?3 SepL 0.3 0 i FPS = lst ilo��-paced mnott' sample, FP'_ _'_nd ilo��-paced runoll sample. � Yertical mo��ins conducted 6 h before e�ent. 12�� J EYV(ROX QUAC.., VOL?6. SEPlE�iBER-0CTOBER (9Y7 143 ma L The potzntial soil loss from thz plot was calculated to be 19.4 k� ha for thz li min event. This number rzpresents the highzst potential soil loss for [his studv. Based on the acerages of total runoff volume (1390 L). duration (20 min). and concentration in the flo�s-paced sampies (9.b mg L the averaoz pocential soil loss for afl plots in 199-� was 15 ke ha for a 20 min zcznt. In 199�, thz a��erage potentizl soil loss ���as 0.1 kg ha ' for a 20 min ecznt. lisins S°io-sloped plots of tall fescue turf maintained at 8 cm, Gross et aL (1991) rzported the avera�e poten- tial soil loss for a 30 min, 120 mm h intznsity storm �cas �19 kg ha ' for bare soil and ��' ko ha for mature tall fescue turf szeded at 4SS kg ha �.�The much lower amount oi soil loss for the current study compared to Gross et at. (1991) could hzve been duz to differences in turf density and tortuousit}� of ocerland floa�. Gross et aL (1991) reportzd a mean density of �7 tillers dm' for a mamre tall fzscue turf sezded at 4SS kg ha and maintained at 8 cm. Linde et al. (199�) reported mean densities of 2�93 tillers dm for ma[urz creepin� bent- erass and 275 tillzrs dm'' for mature perennial rye�rass maintained at 1.9 cm. In this study, sedimen[ was seldom detzcted in runoff samples produced bq rainfalL Only 6 of thz �6 total runoff samplzs from rainfaff even[s contained detectable sedimenc Of those sis samplzs, thz highest sediment concen[ra[ion was 26 mg L for a bent�rass plot that produced 170 L of runoff in 100 min. Thz soi! loss for that plot �cas 0.36 k� ha''. In another study, Gross et aI. (1990) reported the average soil loss in runoff caused by rainfall #or sodded tall fescuelKenmcky blue�rass p1oG �vith a 5 to 7%a slope ti�as 0.4� ke ha and 1.47 k� ha for consecutive }'ears. SUDT�IARY AND CONCLUSIONS On avzrage, vzry fittle sediment transport, if any, was found in runoff samplzs. Therefore, for the conditions of thz curren[ studq, 4-yr-old creepino bentorass and pzrennial ryeerass turfs wzrz very effective in rzducine sediment transpurt, e�en aftzr vertical mowing do�vn the slope ok each plo:. I[ ma}� be possiole that if vertical V�����L mo�cina �cas more agaressive, sediment transport could increase bzcause a�reatzr amount of vegztation would be remoced and morz arooves cut into the soil. Nutrient transport, particularly phosphate and TKN, sienificantly increased for runoff events that had pre- e��ent irrigation and were conducted 3 h after fertiliza- tion. For a11 other evenG, nutrient transport �vas consis- tently lo«er. As a rzsul[, off-site movzment of nutrient; from golf course fairways may happen if runoff occurs soon aE[er sranular fertilizzr is applizd to a neariy satu- rated soil. Such conditions w•ould essentiafly represent a«orst casz scenario for runoff and nutrient transport and wouid Izss likely happen in `real world' circum- stances. REFERE\CES Btard. J.B. 197i. Turf�rass scizncz and culture. Przntice-HaL(, En�lz- wood Clift>, rJ. Gro>s, C.DI., J.S. Anglz. R.L. HiII, and 1LS. R'zIItrizn. 1991. RunoF[ and szdimznt losszs from [slt fescuz undzr simula[zd rainfalL I. Environ. Qual. ?0:6d3-607. Gross. Cbt., 7.S. Ano1z, and hLS. Neltzrltn. 1990. K�utrient and szdi- mcn[ losses from mrf�rass. J. Emiron. Qual. 19:66i-66S. Harrison, S.A., T.L �Ya[schke, R.O. hlumma, A-R. Jarre[[, and G.W. Hamilton. 1993. Nutriznt and pzsticidz concentra[ions in wa[zr from chemically treated mrf�ca>s. p. 191-207. I�i K.D. Racke and A.R. Les(iz (zd.) Pesticidzs in urban environments: Fatz and si�nifi- cance. ACS $ymp. Ser. 5??. Am. Chem. Soc.. �Vashington, DC. Lindz, D.T. 1996. RunoEf, erosion, and nucriznt [ransporc from crzzp- in� bznc�rass and pzrennia! ry'e�rass turfs. Ph.D. diss. Pznnsylvania State Univ., Univzrsity Park. PA. Linde, D.T., T.L. ��"atschkz, and J-A. Bor�er.1993. Nutriznt transport in rurtoff from two turf�rass spzcizs. p. 489-496. (n AJ. Cochran and bi.F. Farrally (edJ Science and �oif IL Ptoc. of the 19%Norid $cizntific Con�rev Of 6olf. E& F\ Spon, D'+zw Yotk. Linde, D.T., T.L. Wa[schke, A.R. Jarrett, and J.A. Bor�ec 1995. Surface mnoff assessment from creepin� bzntorass and pzrennial ryzgrass mrE A�roa J. 57:176-152. �[or[on. T.G.. A.J. Gold, and W.�I. Sullivan. 1988. Intluznce of over- warering and fertiliza[ion on ni[roozn losses from homz lawns. !. Eaviron. QuaL 17:1?�i-liQ. $AS Ins[itu[z 1990. SAS;STAT uszr's �uide. �'oL 2. Version 6. A[n ed. SAS Inst., Cary, FC. Stzzl, R.G.D., and J.H. Torrie. 1930. Principles and procedures o( s[a[istics: A biomz[rical approach. 2nd ed. hleGraw-Hilt, he«' York. ��'auchopz. R.D.. F.G. Witliams, and L.R. \farti. 1990. Runoff of sulfomz[uron-mzthyl and tyanazine from small plo[s: Effzc[> of formulation and srass co�er. J. Environ. Qual. 19:119-IZ�. Reprinted Cmm IheJouma! � f L•nvnonmrn/a! Quoliry Volume 23, no. I, Jan.-Fcb. I999. Copyright O I999, ASA, CSSA, SSSA 677 South $egoe Rd„ Mad(mq NI53711 USA 0 � �,��� Relationship between Phosphorus Levels in Three Ultisols and Phosphorus Concentrations in Runoff D. H. Pote,� T. C. Daniel, D. J. Nichols, A. N. Shazpley, P. A. Moore, dr., D. M. Miller, and D. R. Edwards ABSTRACT Soi15 that contsi� high P lerel5 Can become a primary 5ource of dissolved reactive P(DRP) in runoff, and thus contribute to acceler- ated eutrophication of surface waters. In a prerious sNdy on Captina soil, sereral soil te5t P(S1'P) methods gave resulB that wece signifi- canUy mrrelaFed fo DRP lerels in rvnoEf, but disrilled H and NH�- oxalate methods gave the best coaelations. Because results might differ on other soils, mnoff studies were conducted on three additional Ultisols to identify the most cnnsistent STP method for predicting runoffDRP lereis, and determine effecfs of site hydrology on correla- tions between STP and runoff DRP rnncentrations. Sucface soil {U- 2 cm depth) of pastuce plots was analyced 6y hlehlich III, Olsen, Morgan, Bray-Kurtz Pl, NH�-oxalate, and dis611ed H methods. Also, P saNration of each soil was deterntined by three diflerent methods. Simulated rain (75 mm h") produced 30 min of runoff Gom each ploL Ali cortelations of STP to (unoff DRP were significant (P G 0.01) regardless oF soil series or STP method, with most STP methods e Fiag digh correlxtions (r > 0.90) on all three soils. For a given levei of H�O-eztractable STP, low runoff vulumes coincided with low DRP concen[rations. Tf�erefore, when each DRP concentra- tion was di�ided by��olume of piot nmutt; wrrelafions to H:O-exfract- able STP had the same (Y < 0.05) regression line (or every soil. This suegests the importance of site hydrology in determining P Ioss in r�noff, and may provide a means of developing a single reiationship for a range of seil series. E uTxoexicnno:r of streams and lakes can be greatly accelzrated by the influs of nutrients in surface runoff from agricultu-al land. Since P has been identified as the nutrient in rur.off that is usually the most limiting to algal growth, control of P Izvels in runoff is often recommended as the best way to minimize the eutrophi- cation of su:face waters (Rohlich and O'Connor, 1980; Litt1e,198S; Breeuwsma and Silva,1992; Sharpiey et ai., 1994). Phosphorus is often perceived to be so immobiie in soil that losses from agricultural land are not usualiy considzred to be agronomically_iraportant, but even small agronomic losses can have serious environmental consequences. In iact, scils that contain high levels of P from escessive fertilization can become a primary source of dissolved reactive P(DRP) in runoff (Edwards et a1., 1993). Other investigators have found direct correlations be- tween soil P levels and P concentrations in runoff. D.H. Po[e, USDA-ARS, Dale Bumpers Small Fazms Res. Centez, 6833 South Stare Hwy. 23, Booneville, AR "R927-9214; T.C. Daniel, D.J. Nichols, and D-M. Miller, Dep. of Agro�omy, S15 Plan[ $<ience, Univ. of Arkansas, Fayetteville, AR 72701; P.A. Moore, Jr., USDA- ARS, 115 Ptant Scier.ce, Fayetteville, AR 72701; A.N. Sharotey, Pas- ture Systems and Watershed Managem�nt Research Lab., USDA- ARS, Curtin Road, University Park, PA 168023702; and D.R. Ed- wards, Biosystems and A�ricultural Engineering Dep., 128 Agricul- tural Engineering Building, Univ. of Ker,mcky, Lexing�on, KY 40546. Received 29 July 1998. *Correspondin� author (dpo�e@a�.gov). Published in 7. Environ. QuaL 25370.175 (1999). Schreiber (1988) sampled soil and runoff from mono- cropped com (Zea mays L.) or cotton (Gossypium Firsu- tum L.) research plots and watersheds in Mississippi, with various cropping pracaces for corn includin� con- ventional tillage, no-till, crop residue removed for sila�e, and crop zesidue left on the soii surface. Results showed that water-extractable soil test P(STP) was significantly correlated to annual discharge-weighted DRP in runoff. Yli-Halla et al. (1995) analyzed soil and runoff from eight cultivated field plots in southwestern Finland and concluded that mean DRP concentration in runoff de- pended on the water-extractable P level in surface soil. However, both of these studies relied on uncontrolled natural rainfall events to produce runoff, and combined a variety of cultivated crops and management practices, while neither study included uncultivated grassland. In a previous study (Pote et al., 199fi), we concrolled the variabitity of field conditions as much as possible by using consistent dimensions, slope, soil, and grass cocer for all plots, and using simulated rainfaL to pro- duce runofi. The study compared results from several soil test P(STP) extraction methods to determine which were most useful for predicting DRP levels in runoff from fescue (Festuca arundinacea Schreb.) plots on a Captina silt loam (fine-silry, siliceous, mesic Typic Ftagi- udult). Extraction of P in soil samples from the surface soil (0-2 cm depth) showed that the Mehlich III (Meh- (ich, 1984}, Bray-Kurt2 Pl (Bray and Kurtz, 1945), and Olsen (Olsen et al., 1954) extraction methods gave soil P levels with very significant correlations to DRP con- centrations in surface runoff. The soil P-saturation method (Pote et al., 1996) also gave zesuiu that corre- lated very well to runoff DRP, but Fe-oxide strigs (Sharpley, 1993), distilled water, and acidified ammo- nium oxaiate (Pote et a1.,1996) were the STP extractants that gave the best correlations to DRP in runoff. Since this study was only conducted on a single soil, we hy- pothesized that the results might be different for other soils of differing physical and chemical properties, even with;n the same soil order. As severai states are attempting to define threshold STP levels above which DRP enrichment of runoff is unacceptable from a water-quality perspective, more information relatin� soil P to runoff P is needed (Shacpley et al., 1996). Such Field data are essential to development of technically-sound STP levels that can be used to guide P management recommendations. Therefore, runoff studies were conducted on three addi- tional Ultisols. The objectives were to determine (i) which STP method maintains the highest correlation to Abbreviations: CV, coefficient of variation; DRP, dissolved reactive P; ICP, induc[iveiy coupted plasma spectrometer, M3, Mehlich III extraction me[hod for soil P; PSI, P sorp[ion index; SD, siandard deviation; STP, soil tes[ P. y �, 170 O l - �� \'�-- POTE EI AL: PHOSPHORUS LEVELS LY THREE ULTISOLS Tabie 1. Soil charaderisticc (mean) and :esu{ts of various soiS test P(STP) methods from plots on three soils. Nella so� Clay covteat 105% Oiganic C content 3.S% pH 59 Oxalate-Fe, mg kg ' 1909 Oxalate•Al, mg kg ' LiO4 Ne{ls wl Ztange Mean SDi Lioker wil li9q 3.6 % 51 3003 1170 Linker soil Range hlean SD _ �p � � i in lYOark swl 7.4% 4.6 % 6.2 104i 16M13 noack :oa Range Mean SD STP method - - � biehlich III 260.42L 294 73 12I 366 226 � 17-?63 109 � Olsen 79-166 175 28 61-1b2 104 31 7-303 44 30 Morgan 23b5 44 IS 30-108 57 27 0.107 35 34 Bray-Kum PS 161-3d2 7A0 62 121-328 ?A7 76 14-156 90 47 Nil..ozalate 691-1L7 9DD L'9 315-707 442 144 210.613 4W 1'A1 DisNled H�O 37-109 74 25 1& -107 50 3U b-3U 36 24 t Standard deriation. DRP concentra[ions in runoff from a variety of soil series within the Ultisol order, (ii) whether STP levels affect DRP concentrations in runoff consistently across soil series and if not, (iii) what effect soil hydrology has on the relationship between STP and runoff DRP. MATERYALS AND bIETHODS a Field Plots Six field plots were constructed during the fall of 1993 on each of three soils in northwest Arkansas: Nella (fine-loamy, siliceous, thermic Typic Pateudult), Linker (fine-loamy, sili- ceous, thermic Typic Hapluduli), and Noark (clayey-skeletat, mixed, mesic Typic Paleudult) (Table 1). All plots were con- structzd on well-established tall fescue pastures with approxi- ma[ely 7% slope and 100% ground cover as measured by the line-transect me[hod (Laflen et al., 1981). These pastures had previously been amended with various combinations of swine manure slurry, commercial fertilizers, and/or manure from grazing cattle. Some plots had received swine manure [he previous year, but no amendments were allowed on the plots for sevetal months preceding this s[udy. Vegeta[ion height was maintained between 0.1 and 02 m throughout the study by mowing. Each plo[ (1.5 x 3 m) was fitted with aluminum borders (extending 5 cm above and 10 cm below [he surface) for runoff isolation, a downslope trough for runoff collection, and a runoff sampling pit, as described by Edwards and Daniel (1993). Fences were constructed around the plocs to prevent catile from contributing P inp�ts or causing other damage during the study. In May 1995, a simula[or described by Edwards et al. (1992) was used to reduce antecedent moismre variability by applying rainfall (75 mm h'') to each ptot until the su[face layer was saturated. This simulator delivers rainfatl at an eait pressure of 41.4 kPa from four VeeJet nozzles` elevated 3.05 m above the soii surface by an aluminum scaffold to obiain drop-size distribution and terminal velocity compazab(e to that of natu- ral rainfall. Tarpaulins attached to the aluminum scaffold sur- round the plot to form wind screens. An elec[rie motor drives the shaft to which the nozzles are attached, causing them to oscillate across openings in the simulator body, with the rain- fall intensity dependent upon the frequency of oscillation. ' Names are necessary to repor[ factualty on available data; how- ever, [he USDA nei[her guarantees nor warrants [he s[andard of the produc[, and [he use oE the name by USDA implies no appmval of, : the product to [he zx<lusion of othecs Ihat may also be suitable. � Following the initiai rainfall application, all plots were allowed to drain for 48 h before simulated rain was applied again a[ an intensity of 75 mm h'' to generate 30 min of runoff from each plot. Sampling Methods Runoff was sampled manually at 5-min intervals throughout the runoff event, beginning 2.5 min after initiation of continu- ous-flow runoff. For each discrete runoff sample, the volume and time required to colleM it were recorded and used to calcula[e mean flow rate and total volume of runoEf for the 5-min interval. Using these runoff data, the six discrete runoff samples from each plot were used to construct a flow-weighted composite sample to represent the total runoff from that plot. An aliquot of each composite runoff sample was filtered (0.45- µm pore diame[er) within 2 h of collection and stored in the dark at 4 until analyzed for DRP by the molybdeoum-blue method (Mucphy and Riley,1962). Total DRP mass loss from each plot was calculated as the plot's total runoff voiume multiplied by DEtP concentration in the flow-weighted com- posite runoff sample from thaf plot. . Just prior to applying simulated rainfall to a given pbt, a representative composite soil sample was collected by combin- ing 10 discrete soil cores (2.54 cm diam.) taken randomly from the surface layer (0-2 cm depth) of the plot. All composite soil samples were stored in the dark at 4 until air dried and sieved (2 mm) to remove larger rock particles and most of the plant material. Two complete runofF events were conducted on each plot, separated by a 2-d interval. For each separate runoff event, soil samples were collected jus[ prior to simulated rainfall ap- pHcation. Soil Analyses Each soil sample was analyzed for extractable P by six methods: Morgan (Morgan, 194I), Mehlich III (Mehlich, 1984), Bray-Kurtz Pl (Bray and Kurtz, 1945), Olsen (Olsen et al., 1954), distilled water, and acidified ammoaium oxaLate. The Morgan, Mehlich III, Bray-Kurtz Pl, and Olsen chemical extractan[s were selected because they aze commonty used for STP analysis in soil testing laboratories. These methods were not originally devetoped to predict runoff water quality, but rathez to assess the fertility status of soil for aop produc- tion. Distilled water most closety simulates actual runoff solu- tion, and may thus be the most appropriate for predicting runoff DRP. One „cram of soil was mized with 25 mL of o � - ����. I�Z J. ENVIRON. QUAL., VOL. 28, JANUARY-FEBRUARY 1999 distilled water, shaken end-over-end for 1 h, centrifuged for 5 min a[ 266 m s�(27100 g), Eiltered (0-45 µm), and the supzmatant analyzed for P by the molybdenum-blue method (Murphy and Riley, 1962)_ Acidified ammonium oxalate has been used in severaf pre�,ious studies (van der Zee et al., 1987; van der Zee and van Riemsdijk, 1988; A�folina et al., 1991; Breeuwsma and Silva, 1992; Freese et al., 1992), theoretically [o release into solution potentially dzsorbable P, as it dissolves the compounds (noncrys[alline oxides oi iron and aluminum) controlting P sorption in acid soils (Table 1). In our smdy, ammonium oxalate extractant was made by mixing 02 M oxalic acid with 0.2 M ammonium oxalatz (approximately 535 mL of oxalic acid with 700 mL of ammonium oxalate) until the combined-solution pH was 3.0. A 20-mL aliquot of the ammonium oxalate solution was then mixed with 0.5 g of soil, shaken in the dark for 2 h, cenirifuged for 20 min at 131 m s '(14481g),anddecantedforPanalysis.Oxalate-extractable P, A., and Fe were also used to calculate the P sorption- sa[uration of each soil as described below. Mehlich III, Bray- Kurtz Pl, and acidified ammonium oxalate extracts were ana- lyzed for P by inductively coupled plasma spectrometer (ICP), while n4organ, Olsen, and d'utitied water extracts were ana- ]yzed colorimetrically by the molybdenum-blue me[hod (Mur- phy and Riley, 1962). A single-point P sorption index (PSI) described by Mozaf- fari and Sims (1994) was also determined on each soil. A P sorption solution (containing 300 mg P per liter) was made by dissolving 1.315 g of KH in enough distilied, deionized H to make 1 L of solucion. The PSI was determined by weighing 1.00 g of soil in[o a�0-mL centrifuge [ube, adding 20 mL of 0.0125 M_ CaCi 2H>O, and adding � mL of P sorption solution to make a combined solution containing 0.01 M CaCi and 60 mg P per liter. After two drops of toluene were added and the tubes sealed, the mixture was shaken for 18 h on a reciprocating shaker, centrifuged for 10 min at 266 m s' (27100 g), filtered (0.45-µm), and analyzed for P by induc- tively coupied plasma spectrometer (ICP). The PSI was calcu- lated as X(lo� P where X is P sorbed (mg kg'`) _[(P�)(V) —(P (kg of soil) P is initial P concentration in sorption solution (mg L V is volume of P sorption solution (L) P is final P concentration in solu[ion (m� L Phosphorus Saturation of Soil The P saturation (%) of each soit sample was calculated by two different methods; (i) oxalate-extractable P(mmol kg") divided by thz oxalate-extractable AI and Fe (mmot kg con[ent, and multiplied by 100, and (ii) initial STP con- ten[ (mg kg ') divided by P (mg kg and multiplied by 100. For this second method, the PSI value was used to approximate the maximum amount of P(P.,,, that could be adsorbed by the soil. Mozaffari and Sims (1994) found that P� can be estima[ed by the equation P�� =(PSI + 51.9)/ OS, �iven that P Mqr < 1400 mg kg STP extracta�ts selected to obtain the initial STP conten[ were Mehlich III (M3-PSI method) and distilled H2O (H2O-PSI method). Statistical Methods For each soil, comparisons were made between STP meth- ods by correlating STP results to DRP concentrations in runoff from the ptots, developing a linear regression from [he 12 data points, and calcula[ing [he sample corrzla[ion coefficient (r value) for each. For each soil test method, analysis of covari- ance was used to determine whe[her there were statistica! differences be[ween regression slopes and in[ercepts of ihe three soils. RESULTS AND DISCUSSION Soil Phosphorus For each of these soils, the range, mean, and standard deviation of STP contents are shown in Table 1. Distilled water, Mor�an, and Olsen methods extracted the least amounts of P from soil, while Mehlich III and Bray- Kurtz Pl methods extracted larger amounts. NH,-oxa- late extracted much larger amounts of soil P than did other extractants, suggesting that most of the P in these soils is sorbed or precipitated on amorphous oxides of Fe and Al. Relationship between STP and Runoff DRP For each soil, correlations of STP to runoff DRP were not significantly affected by the time interval (2 d) between the two runoff events. Therefore, the data from both runoff events were combined to �ive a total of 12 data points for each soil. The correlation coefficient (r) and linear regression equation are given in Table 2 for each STP correlation to DRP in runoff. For all soiLs, the STP values obtained by each method were significantly corretated (P < 0.01) to DRP concentrations in plot runoff. Yet, when the extraction methods were com- pared using r values to see how closely the data peints fit the regression line, it was apparent that some STP methods were more closely related to DRP concentra- tions in runoff than other methods (Table 2). For exam- ple, the NH and Olsen methods each gave a weaker correlation r< 0.90) to DRP concentrations in runoff from at least one soil, while all other STP meth- ods gave correlations with r> 0.90 for all three soils (Table 2). However, if previous studies (Pote et al., 1996) are considerzd, the H2O-estractable soil P has shown the most consistently high correlation to DRP concentrations in runoff, even when rainfall intensity, slope, and seasonal conditions varied. Although the usefulness of an STP method for pre- dicting runoff DRP concentrations depends largely on its ability to produce data poinu that closely fit a regres- sion line on any given soil, it would also be very helpful to have an STP method that produces approximately the same regression for all soils (or at least a large group of soils). Such a method would eliminate the need to use soil series as the basis for maximum soii P recom- mzndations, thus saving the time and expense of accu- rately identifying the soil series of each individual site. If the data points from all three soils are combined into a single data set, the P-saturation (oxalate method) might seem to be a good choice for this purpose because it gives a good linear correlation (r = 0.887), and the fit is even better for a second-order regression (r = 0.931 for the curve where y= 03053 — 0.0353x + 0.0014x However, if the data points are separated into regression lines for each soil, differences between some slopes be- come apparent (Fig. l). When the regression-line �raphs of each method were compared visually, the P-satura- POTE EI' AL PHOSPHORUS LEVEIS IN THREE ULTISOLS 0.905 0$69 0.907 ' . 0.913 0.806 0.9'� 0903 0.916 0932 Table 2. Results of soil test P(STP) methods correlated to dissolved reactive P(DRP) in runoH from three Ultisola CortelaRon coefident (r) t S1 P(m kg"') co rrelated t DRP (mg L STP method Nella w� Linker so0 Noark wii Mehtich III Olsea Mocgan Brav-Kurtz PS NH,-Ozalate Disb7led H.O P saNlation (oxalate method) P satuntion @13-PSI metbod) P satucdtion (H_O-PSI metLod) $TP method Mehlich III Oisen Morgan Bnv-Kurtz Pl NI-I�-Ozalate Distilled H P saturation (o:Wate met6od) P saturation (M}pSI method) P saturation (H.O-PSI metho� 0.916 0.864 0.941 0.950 0.914 0.928 0.928 0.928 0921 �\�\\\b- 173 0.932 0935 0.932 0943 0908 0.965 0.933 0.937 0.978 Regression line equa for ST P (mg kg �) <onelated to DRP (mg L NeO soii Linker sol Noark soil y= 0.0036z - OAS y= 0.01135c - 038 y = 0.0016x + 0.00 y= O.00SSz - 0.43 y= 0.0093z - 056 y= O.00d3x - 0.02 y= O.OlSlz - 0.18 y= O.O1LSZ - 025 y= 0.0038x -F 0.04 Y=O.00d3x-0.42 y=0.0041ac-0.4b y=0.0027x-0.02 y= 0.001Sx - 1.03 y= 0.002Lc - 0.63 y= 0.0009x - 019 y = 0.0107x - 0.18 y = 0.0104x - 0.11 y= O.00SSx - 0.03 y= 0.0820x - 203 y = 0.0397x - 0.62 y = 0.0251c - 01A y= O.00S� - 0.08 y= 0.0065x - 0.04 y= 0.0045x + 0.03 y = 0.0?62x + 0.03 y = 0.0215z + 0.06 y= 0.0759x + 0.01 i All correla6on coeffidents were signifitant (a = 0.01). tion (PSI methods), Mehlich III, Bray-Kurtz Pl (Fig. 2), and distilled H (Fig. 3) methods each appeared to have regression lines that were relatively close together with similar slopes for all soils, but statistical analysis showed that none of the methods for correlating STP to DRP in runoff gave the same (P � 0.05) regression line for all three soils. This result was not surprising, given the differences in chemieal and physical properties between soils. The P saturation status of each soil in this study was significantly (P < 0.01) related to DRP concentrations in runoff, regardless of the method used to calcutate P saturation. All three methods gave high correlations to DRP in runoff but none gave the same regression line on all three soils, so their vatue as universal predictors of DRP concentrations in runoff is questiouable. i � E 0 � c a � 0 1.6 �,a 7.2 7.0 0.8 0.6 0.4 0.2 • Nella (r = 0.903) � Linker (r = 0.928) ONoark (r = 0.933) 0.0 ' " ' 0 10 20 30 � 40 � • • • � � • O • M .� Effects of Runoff Volume Because site hydrology of each soil is likely to impact the relationship between soil P and runoff P(Gburek and Sharpley, 1998), the effect of runoff votume on P transport from our plots was evaluated. The average rainfall application required to produce 30 min of con- tinuous runoff is included in Table 3, along with mean runoff volume for each soil. Runoff from the Nella and Linker soils averaged about the same volume and the variability was also similar, while the Noark soil had the lowest amount of runoff and the least variability. The differences in runoff among soils are reflected in the correlations of water-extractable STP to runoff DRP. For esample, when water-extractable STP was correlated to mass losses (loads) of DRP in runoff (Fig. 1.6 7.4 7.2 �� rn 1.0 E 0.8 ' 0.6 c a 0.4 ¢ 0 0.2 0.0 0 700 200 300 400 Bra Kurtz extractabie soil P m k - Soil P Saturation (%) Y � 9 9) Fig. 1. Relationship between P samration (oxalate method) of suhace Fig. 2. Relationship behveen Bray-Kurtz Pl extractabfe P in sudace soil and dissoived reactive P(DRP) in runoff from three soils. soil and dissolved reac[ive P(DRP) im m�oll. . 174 7.6 7.4 1.2 i � 7.0 0 0.8 c i 0.6 � 0.4 0 0.2 a.o 0 20 40 60 J. ENVIRON. QUAL., VOL 28, JANUARY-FEBRUARY 1999 80 700 120 Water eztractable soil P(mg k9 Fg.3. Relationship between water-eztraMabie P in surface soil and dissolved reactive P(DRP) in runoS 4), the Noark correlation was best (r value = 0.963) because mass losses depend on boSh.the P concentration and the volume of runoff (which" was highly consistent for the Noark soil). Runoff volumes were more variable for the other two soils, and therefore mass losses of runoff DRP show a poorer correlation to STP (Fig. 4). The variability of runoff volume is also reflected in the r values for the correlation of water extractable STP to DRP concentrations in runoff (Fig. 3). For example, Nella soil had the most variable runoff volume, and it also had ihe lowest r value, while Noark soil had the least variable runoff volume and the highest r value. Finally, for a given level of water-extractable STP, soils with the lowest mean runoff volume also had the ]owest concentration of DRP in runoff (Fig. 3). For example, Noark soil produced the least amount of run- off, but for any given level of water-extractable STP, it also had the lowest concentration of DRP in the runoff. No previous studies have investigated the relationship between runoff volume and DRP concentration in the runoff; and our observations at first seemed rather counter-intuitive because we expected higher volumes of runoff to generally produce lower DRP concentra- tions due to greater dilution. This unexpected trend may result from the rapid movement of DRP into the soil profile of soIls with low runoff volumes (high infiltration rates), thus taking it away from the primary zone of transfer to surface runoff. In soils with lower infiltration Table 3. Rainfal{ aud runo8 data from simufated rain appliration to 6eld plots on three soiLs. Ne17a soii Linker soii Noark sol Rainfall meant, mm 4$.9 475� 53.6 Ruuo@'mean,mm 21.7 7A.6 133 Runoff CV, % 30.0 29.1 175 i Amount required to produce 30 min of rumR. Eath mean tepresenls 12 runoH events. 400 �0 300 ._°: 0 � 200 i c v a 100 a 0 0 b a\ -\\\�� •Nella (r = 0.724) �Linker (r = 0.847� ONoark (r = 0.963) •• • � � • • � � � � � �• � • . o 20 40 60 80 100 120 Water extracfoble soil P(mg kg Fg.0. Relafionsltip between water-extractable P in surCace soil and dissolved reactive P(DRP) load in runoff. rates, much more of the dissolved P may remain near the soil surface long enough to be lost in runoff water. In an attempt to define these processes, we normal- ized DRP concentration for each plot. When the DRP concentration in runoff from each plot was divided by the depth of runoff from that plot, and related to the water extractable STP level, regression lines for all soils were statistically the same line (P < 0.05) (Fig. 5). Thus, by combining water-extractable STP data with hydro- logic data, it may be possible to make reasonably accu- rate predictions of DRP levels in runoff from a range of soils. Acquiring the necessary hydrologic da[a on runoff volumes from a soil may sometimes be just as difficult as accurately identifying the soil series of each specific site, but this at least provides an a!temate 7.2 E � 1.0 0 � 0.8 � ^ o.s � � E o.4 `o � 0.2 � c a 0.0 � 0 0 . . . • • • • . � b .� . 20 40 60 SO 100 120 Water eztractable soil P(mg kg Fg. 5. Relafio�ship between watervextzactable P in sudace soil and the ratio of dissolved reactire P(DRP) im m�otf to the total amount of runoR. •Nella (r = 0.866) - �Linker (r = 0.847) oNoark (r = 0.919) POTE EC AL: PHOSPHORUS LEVELS I1V THREE ULTISOLS method for predicting DRP concentrations in runoff. For water-quality modelers, it also supplies important information conceming the relationship between vol- ume of runoff and DRP concentration in runoff. Most importantly, i[ shows the strong influence of site hydrol- ogy on processes controlling P loss in surface runoff. CONCLUSIONS The results of this study reinforce previous evidence of a linear relationship between P levels in surface soil (0-Z cm deep) and DRP concentra[ions in runoff from the soil surface, but this study also extends our knowl- edge by showing that such a relationship exists on a variety of Ultisols. On each soil that was tested, a signifi- cant (P < 0.01) linear relationship was apparent, regard- less of the method used to determine STP. Because most STP extractants gave results that were highly cor- related (r > 0.90) to DRP in runoff from all three soils, this study did not clearly identify any particular STP method for maintaining the highest correlation to DRP concentrations in runoff from all soils tested. However, the study did show that several STP extractants may be useful for predicting DRP concentrations in runoff, including extractants such as distilled water that were supported by the results of previous work (Pote et al., 1996) conducted under different rainfall intensity, slope, and seasonal conditions. This study showed that effects of STP levels on DRP concentrations in runoff are not always consistent across soil series, and much of the difference can be attributed to soil hydrology. The fact that total plot runoff was much more variable on some soil series than on others was apparently reflected in correlations be[ween STP and runoff DRP, as soils with the most consistent vol- ume of plot runoff had the best correlations of water- extractable STP to both concentrations and mass losses of DRP in runoff. Also, for any given level of water- extractable STP, soils that produced the lowest volumes of runoff also had the lowest concentrations of DRP in the runoff. When this information was used to normalize the data for DRP concentrations in runoff (divide each DRP concentration by the volume of runoff from that plot), the resulting correlations to water-extractable STP had statisTically the same (P < 0.05) re�ression line for every soil. This implies that knowledge of site hydrology can improve the usefulness of STP data for predicting DRP concentrations in runoff. REFERENCES Bray, R.H., and L.T. Kurtz. 1945. Determination of [otal, organic, and available foans of phosphorus in soils. Soil Sci. 59:39-45. Breeuwsma, A., and S. Silva. 1992. Phosphorus fertilisa[ion and en�i- ronmencal effec[s in The Netherlands and [he Po region (I[aly). Rep. 57. Agric. Res. Dep. The Winand S[aring Cen[re for In[e- (j \-\\\�--- 175 grated Iand, Soil and �'da[er Research, �'lageningen, the Ne[h- edands. Edwards, D.R., and T.C. Danie1.1993. Effects of poultry lit[er applica- tion rate and rainfall intensity on quality of runoff from fescuegrass plots. J. Enviton. QuaL 22361-365. Edwards, D.R, T.G Daniel, J.F. Murdoch, and P.F. Vendrell. 1993. The Moore's Creek BMP effectiveness monitoring pcoject. Paper 93208�- ASAE, St. Joseph, MI. Edwards, D.R., LD. Nocton,T.C. Daniel, J_T. Walker, D.L. Ferguson, and G.A. Dwyer. i992. Perfonnance of a rainfall simularor. Arkan- sas Farm Res. 41:1'ri4. Freese, D., S.E.A.T_M. van der Zee, and W.H. van Riemsdijk. 199"t Compadsons of different modzls for phosphate sorption as a func- [ion of the iron and aluminium orzides of soils. J. Soii Sci.43:'729-738. Gburek, W.J., and A.N. Sharpley.1998- Hydrologic con[rols on phos- phorus loss from upland agricul[ural wa[enheds. 7. Emiron. Qual. 27.267-277. I.aflen. J., M. Amemiya, and E.A. Fiiniz.19S1. Measuring crop residue cover. J. Soil Water Conserv. 6:341-343. Little, C.E. 1988. Rurai dean water. The Okeechobee story- 7. $oil Water Conserv. 43:38C�390. Mehlich, A. 1984. Mehiich 3 soil [es[ extractanC A modifica[ion of Mehlich 2 extrac[ant Commun. Soil Sci. Plant AnaL 15:1409-1416. Molina, E., E. Bomemisza, F. Sancho, and D.L. Kass. 1991. $oil aluminum and iron fractions and their relacionships wiih P immobi- liza[ion and other soil properties in andisols of Costa Rica and Panama. Commun. Soil Sci. Plant AnaL 221459-1476. Morgan, M.F. 1941. Chemical soil diagnosis by [he universal soil testing system. Conn. Agric. Exp. Stn. (New Haven, CT) Bul(. 450. Mozaffari, M., and I.T. Sims. 1994. Phosphorus availability and sorp- [ion in an A[lan[ic coas[al plain wa[ershed domina[ed by animal- based agriculture. Soil Sci. 157(2):97-107. Muiphy, J., and J.R. Riley. 1962. A modified single solu[ion method for [he de[ermination of phosphate in na[ural waters. Anal. Chem. 2731-36. Olsen, S.R., C.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estima- tion of available phosphorus in wils by extraction with sodium bicarbona[e. USDA Circ. 939. U.S. Go�. PrinC Office, Washing- ton, DC. Pote, D.H., T.C. Daniel, A.N. Sharpley, P.A. Moore, Jr., D.R. Ed- wards, and D.J. Nichols.1996. Relating ex[ractable soil phosphorus to phaspho�us losses in runoff. Soil Sci. Soc. Am. J. 60:&55�59. Rohlich, G.A., and D.I. O'Connor. 1980. Phosphorus managemen[ for [he Grea[ Lakes. Fival Rep., Phosphorus Management S[ra[e- gies Task Force, Int. Ioint Commission (IJC). Poilution from Land Use Ac[ivities Reference Group Tech. Rep. Phosphorus Manage. Strategies Task Focce, Windsor, ON. Schreiber, 7.D. 1988. Estimating soluble phesphorus (POrP) �n ag- ricultural runoff. J. Miss. Acad. Sci. 33:1-15. Sharpley, A.N.1993. An innovative approach [o estimate bioavailable phosphorus in agriculiural runoH using iron oxide-impregna[ed paper. 7. Environ. Qual. 22:597fi01. Sharpley, A.N., S.C. Chapra, R. Wedepohl, I.T. Sims, T.C. Daniel, and K.R. Reddy. 1994. Managing agricultural phosphorus for pro- tection oE wrface waters: Issues and op[ions. J. Envimn. Qual. 23: 437�51. Sharpiey, A.N., T.C. Daniel, J.T. Sims, and D.H. Pote.1996. Decertnin- ing environmentally sound soil phosphorus levels. J. Soil Water Conserv. 51(2):160-166. van der Zee, S.E.A.T.bt.. L.G.J. Fokkink, and W.H. van Riemsdijk. 1987. A new tzchnique for assessmen[ of reversibiy adsorbed phos- pha[e. Soii Sci- Soc. Am. 7. 51599�04. van der Zee, S.E.A.T.M., and W.H. van Riemsdijk. 1988. Model for lon�-term phospha[e reac[ion kinetics in soil. 7. Environ. Qual. 1735�1. Yli-Halla, M., H. Hartikainen, P. Ekholm, E. Tur[ola, M. Puustinen, and K. Kallio. 1995. Assessmen[ of soluble phosphorus load in surface runoff by soil analyses. Agda Ecosyst. Environ. 56:53-b2. �A�.��a� �d - �o� a$ a.OQ, Councii File # � �+'� �� Green Sheet # 1 \'j �. � y ORDINA OF SAINT P� Presented Referred To ..� Committee Date � �o 1 An ordinance amending Saint Paul Legislative Code Chapter 377 t� ��the use of fertilizers containing 2 phosphorus 3 THE COUNCIL OF THE CITY OF SA1NT PAUL DOES ORDAIN: Section 1 5 Chapter 377 of the Saint Paul Legislative Code is hereby amended to read as follows: 6 Sec. 377.01. Definitions. For the purposes of this chapter, the terms defined in this secrion have the meanings ascribed to them: 8 Person means any person, firm or corporation engaged in the business of lawn fertilizer or pesticide 9 applications and includes those persons licensed by the State of Minnesota pursuant to Minnesota Statutes, Secrion 10 18A-21 et seq. 11 Pesticide means any substance or mixture of substances intended for prevenring, destroying, repelling or 12 mifigating any pest, and any substance or mixture of substances intended for use as a plant regulator, defoliant or 13 desiccant. It also means any chemical or combination thereof registered as a pesticide with the U.S. Environmental 14 Protection Agency, any agency later assuming registration in the U.S. federal government, the State of Minnesota 15 Agricultural Deparhnent, or any other State of Minnesota government agency. 16 Sec. 377.02. License required; council approval. 17 (a) No person shall engage in the business of lawn fertilizer or lawn pesticide application in Saint Paul without 18 a license issued by the City of Saint Paul. 19 (b) All city programs for pesticide use shall be reviewed and approved by the city council prior to any application 20 upon city property. 21 Sec. 377.03. Fee. 22 The fee required for a license shall also be established by ordinance as specified in section 310.09(b) of the 23 Saint Paul Legislative Code. ; ► �' ;. . 1 Sec. 377.04. Employees licensed by state. O \ � �\\9� All ofiicensee's employees actually engaged in lawn pesticide applications shail be duly licensed by the State of Miunesota and shall be trained and qualified in the proper methods of handling and applications of pesticides. Satisfactory evidence that such employees are licensed by the state shall be maintained on file in the office of the license inspector. 6 Sec. 377.05. Division of health. 7 The � - DirectoroftheOfficeofLicense,Inspections 8 and Environmental Protection or his/her desi¢nee is directed to monitor the health and safety effects ofthe chemical 9 applications to lawns and to advise the ciTy council of any suspected hazards or violations. 10 Sec. 377.06. Class I license. 11 The license granted pursuant to the provisions of this chapter is designated as a Class � R license, subj ect to 12 the procedures applicable to Class � R licenses in Chapter 310. 13 Sec. 377.07. Pesticide applications; posting. 14 15 16 17 18 19 20 All persons who apply pesticides outdoors are required to post or affix warning signs on the street frontage ofthe properry so treated. The warning signs must protrude a minimum of eighteen (18) inches above the top ofthe grass line. The warning signs must be of a material rain-resistant for at least a forty-eight-hour period and must remain in place for at least a forty-eight-hour period or longer if the human re-enhy interval prescribed in the pesticide label specifies a longer hourly or daily interval. The information printed on the sign must be printed in contrasring colors and capitalized letters at least one-half inch or in another format approved by the coxnmissioner of agriculture, and shall provide the following information: 21 (1) 22 23 (2) 24 25 26 27 The name of the company applying the pesticide or, if not a company, the name of the person having done the application. The following language: "This area chemically treated. Keep children orpets offuntil (date of safe entry--at least forry- eight (48) hours after applicafion or longer if specified on pesricide label)" ar a universally accepted symbol and text approved by the commissioner of agriculture specifying a date of safe entry as specified herein. The warning sign may include the name of the pesticide used. 28 The sign shall be posted on the lawn or yard no closer than two (2) feet from the sidewalk ar right-of-way and no 29 further than five (5) feet from the sidewalk or right-of-way. 30 31 Sec. 377.08. Fertilizer Content. No person licensed under this chapter sha11 appl�y lawn fertilizer. Iiquid or 32 p_ranulaz. within the Citv of Saint Paul that is labeled to contain mare than 0% phosphate (P O ,�prohibition 33 shall not appplv to: 1 a. The naturallv occurrin� phosnhorus in unadulterated natural or organic fertilizing O l-1\�Y 2 products such as vard waste compost; 3 b. Use on newlv established or developed turf and lawn azeas durin¢ their first growing 4 season• 5 c. Turf and lawn azeas which soil tests taken according to Universit�of Minnesota 6 ¢uidelines and analyzed in a State of Minuesota certified laboratorv confirm are 7 below phosphorus levels established by the Universitv of Minnesota. In such cases, 8 lawn fertilizer application shall not exceed the Universitv ofMinnesota recommended 9 application rate for phosphorous. 10 11 12 13 14 15 16 17 Lawn fertilizers contaiuingphos hro orus applied pursuant to the above-listed exceptions shall be watered into the soil where the phosphorus can be nnmobilized and enerallv nrotected from loss bv runoff. Fertilizer applied to impervious services, such as sidewalks. drivewavs and streets is to be removed bv sweepins or other means immediately after fertilizer application is completed. Fertilizer is not to be apniied to frozen soil. saturated soil or under conditions ofim ep nding heaw rainfall. The Office ofLicense, Inspections and Environmental Protection shall be notified at least 24 hours prior to the application of anv lawn fertilizer containing_phosphorus that such fertilizer will be used. the amount to be used and the reason for its a�lication. Section 2 18 This ordinance sha11 take effect and be in force thiriy (30) days following its passage, approval and publication. Yeas Na s Absent Benanav i /' Blakey f Bostrom „/ Coleman ✓ Harris ,/ Lantry �/' Reiter �' Adopted by Council: Date - �. S� 1,oa � AdopUon Certified by Council Secretary By: 2 . � --Q- Approved by Mayor: Date By: � Requested by Deparnnent o£ � Form Approved by Ciry Attomey �� By: Approved by Mayor for Submission to Council � By: � 2 7 'A'� : �� °—�.- S a9�J�� V < �,.. , IST BE ON CIXRJCILAGB�IDA BY OCt. Z�F, ZOO1 a „��,m. ,_ _ ,. C 1- ►1\3 o/laio� GREEN SHEET No 113684 � TOTAL # OF SIGNATURE PAGES u��� u��— ❑ arvwnoa�r ❑ rnrcuu _ ❑ waxr,n�fEau¢ESO.� ❑ qmcu�mnri�eero ❑ ❑ (CLJP ALL LOCATIONS FOR SICaNATURE) An ordinance amending Saint Paul Legislative Code Chapter 377 to prohibit the use of fertilizers containing phosphorus. PLANNING COMMISSION qB COMMITTEE CNIL SERVICE CAMMISSION IIy_\ �=,ti]qd�7 ' TOTAL AMOUNT OP TRANSACTION S �� ' FllNOING SOURCE FlNqNCIAL INFORMAl10N (EXPWf� H36 th15 p2fSMl�fi112R! VqIKIM UlMN 8 COIIffdC[ f0f fhi6 d2p3lIIT2M? VES NO Has thia Pe���m e<u been a dty employee9 ' YES NO Does tlds PersaVfim+ V� as4dN nM na�matHP� b'f anY cwrent dty emWoyee't YES MO Is ttiis peBOMfirtn a targeted �endoYt YES � COST/REVENUE BUDGEfED (GRCLE ON� ACTIVIiY NUMBER VES NO ❑ YINNESOTA BOARD OF WATEfl ANC SOIL NESOURCES NORTHERN FiEGION 394 S Lake Ave Room 403 Duluth, MN 55802-2325 PHONE (218)723-2350 FAX (218) 723-4794 �MINNESOTA BOAND OF WATEfl AND SOIL RESOURCES METRO REGION One W Water SL, Suite 200 St. Paul, MN 55107-2039 r� , Water Resources Education Date: November 12, 2001 To: Councilmember 7ay Benanav Councilmember Jerry Blakey Councilmember Dan Bostrom Councilmember Chris Coleman Councilmember Pat Harris Councilmember Kathy Lantry Councilmember Jun Reiter ! �� UNIVERSITY OF MINNESOTA Extension �� O I - ��l'�- � B � ,M,�,.,,,,,,. ' �CC�G�4�L n'YV!' 1 3 2 �Di �4TY ��ERK Mayor Norm Coleman From: Ron Struss, University of Minnesota Extension __'Sf Re: UM comments on proposed lawn fertilizer ordinances and regulations PHONE (651) 215-1950 FAx Attached are comments from a team of University of Minnesota specialists (651) 297-5615 developed to heip inform the City of St. Paul on their proposed lawn fertilizer [{ MiNNE50TA 80ARD OF ordinance and the Minnesota Senate on their November 15, 2001 hearing on W4TER AND SOIL 1aW11 fe1�.111ZEI$. * RESOUXCES SOU7HERN REGION 261 Highway 15 S New Ulm, MN 56073-8915 PHONE (507) 359-6090 If you would like clarification of these comments or fizrther information, please contact Dr. Carl Rosen at 612-625-811A or crosen(�a,soils.umn.edu. �12.i11C }�011. FAX (507) 359-6018 . � .. � << 0 � -I��a' Comments on proposed ordinances and legislation relating to lawn fertilizers University of Minnesota - November 9, 20U1 Carl Rosen, Extension Soil Scientist, Dept. of Soil, Water, and Climate, Univ. of Minnesota Brian Horgan, Extension Turf Specialist, Dept. of Horticulhu�al Science, Univ. of Minnesota Don White, Professor, Dept. of Horticulrisal Science, Univ. of Minuesota Robert Mugaas, Extension Educator, Hennepin County, Univ. of Minnesota Doug Foulk, Extension Bducator, Ramsey County, Bniv. of Minnesota Ron Struss, Extension Educator, Water Resources Center, Univ. of Minnesota Phosphoms is an essential element required by all forms of life. However, high phosphorus inputs have been linked to degradafion of lakes and rivers through promoting excessive algae gowth. The overall intent of the proposed ordinances and legislation is to reduce the amount of phosphorus entering surface waters and improve water quality. Proposed ordinances and legislation will restrict the use of phosphoms containing fertilizer for established lawns unless a need is indicated by a soil test. The rationale for this is supported by two sound premises: 1) Surveys conducted over the past 30 years have shown that 70% to 80% of the lawns in the Twin City Metropolitan Area have soil phosphorus levels in the very high range and would not require additional phosphorus for oprimal huf growth, and, 2) Applicafion of phosphorus to lawns not requiring phosphorus is a waste of a limited resource. An underlying assumprion is that reshicring the use of phosphorus on lawns will reduce the amount of phosphoms entering surface waters. Unfortunately, the scientific evidence to show that such a restricrion will improve water quality is lacldng. In fact, the pioneering studies conducted by the University of Minnesota in the 1970's suggest that in the short term, use of phosphorus fertilizer on lawns has little impact on phosphorus runoff compared to- the amounts of phosphorus in runoff resulting from breakdown of orgaxuc material such as leaves and grass clippings. Clearly, more quanfitanve research is needed to determine the fate of phosphorus in the lawn landscape and how resh-icting phosphoms fertilizer use for lawns will impact water quality. Reseazch proposals have been submitted by a team of turf gass and soil scientists at the University of Minnesota to deteinune the fate of phosphorus applied to lawns and to define management practices that will minimize movement of phosphorus into surface water runoff. Proposed ordinances and legisiarion may also raise expectarions that water quality will dramatically improve once lawns aze not fertilized with phosphorus. The problem is more complicated than simply restricting fertilizer use and will require a much more integrated approach to improve lake q_uality. In addition to reseazch, educational efforts should be implemented to address all pracrices that affect or contribute to phosphorus runoff in urban areas. UM comments on lawn fertilizer ordinances and regulations Page 7 of 2 �� � Y One fmal comment concerns a loophole in exisring and proposed ordinances that allows for application of organic fertilizer containing phosphorus. Most organic fertilizers have a nitrogen-to- phosphorus ratio that is mucfl lower than the common inorganic lawn fertilizers used today. Since the rate of lawn fertilizer applicarion is based on the amount of nitrogen applied, there will likely be more phosphoms applied when an organic fertiIizer is used tban when a more common inorganic lawn fertilizer is used. Since organic fertilizers with a 0% phosphorus label aze available in Mumesota, the reshiction should be for both inorganic and organic fertilizers. In summary, our comments on proposed ordinances and legislation are: • They aze based on a sound premise that regulaz app2icafion of phosphorus is not needed on most Twin City lawns. • Reseazch Yias not yet shown that restricting phosphorus fertilizer use on lawns will improve lake water quality. • They should be considered as one part of an overall phosphoms runoff management program. Lawn fertilizer ordinances or legislation will not solve the water quality problems of Twin City lakes on their own. • Educafian will be needed for successful compliance and reduction of phosphorus in urban nmof£ • The exemption provided for organic fertilizers is neither warranted nor advised. Thank you for the opportunity to comment. If you would like clarificarion of these comments or further information, please contact Carl Rosen at 612-625-8114 or crosen(c�soils.umn.edu. UM comments on lawn fertilizer ordinances and regulations Page 2 of 2 a De�n Vietor, 12:00 PM 11/6/01 -0600, Re: P in turf runoff X-From : dvietor@taexgw.tamu.edu Tue Nov 612:02:48 2001 X-Mailer: Novell GroupWise Intemet Agent 5.5.5.1 Date: Tue, 06 Nov 20�1 12:0037 -0600 From: "Don Vietor" <dvietor@taexgw.tamu.edu> To: Leslie A Everett <evere003@tc.umn.edu> Subject: Re: P in turf runoff Page 1 of Z O 1 — \\\1— I have attached Word97 files of a manuscript for which 1 wili submit revisions to Assoc. Editor of JEQ next week. The title page, text, and tables are in separate files. The manuscript and references represent the latest work we have (that is near publication) for runoff of P fertilizer and manure P from a relatively steep slope of turf. We applied P fertilizer rates that were refatively large, but comparabie to the farge P amounts observed in soi4 sampfes submitted from urban counties in Texas. If I can be of further help, please let me know. We are very interested in the new urban regulations being proposed for the twin cities. ls Ron Struss our best source for informafion related the new regulations? Don Vietor Donald M. Vietor Soil & Crop Sciences Texas A&M University College Station, TX 77843-2474 Tel. (979) 845-5357 FAX (979) 845-0456 email dvietor@tamu.edu »> "LesiieA. Everett" <evere003@tc.umn.edu> 11/05/01 03:41PM »> He44o Don, Now I've got a request for you! The cities of St. Paui and Minneapolis are in the middle of passing or implementing ordinances regarding phosphorus fertilizer use on lawns. The state legislature is also looking at the issue, with a hearing next week. Some people tell me there is no research data out there to support instituting controls on P fertilizer application to turf. My guess is that there must be some, and ihere should be some as weii regarding P fosses from inorganic fertilizer applied to pasture or hayland as an analogous system. Most current research focuses on manure applied to cropland and pastureland, which wouid not represent the lawn situation well. Ftave you got any references along this line? If so, p(ease send me a{ist, as well as an indication of who else 1 should contact. Thanks much, Les Everett � Gaumanu.doc � JEQtable.doc Printed for "Leslie A. Everett" <evere003@tc.umn.edu> 11/7/Ol O�-���� Response of Turf and Quality of Water Runoff to Manure and Fertilizer J.E. Gaudreau RH. White , D. M. �etor , T.L. Provin and C.L. Munster ' Soil & Crop Sciences Depattment and Z Agricultural Engineering Department, Texas A&M University, Coilege Station, Texas 77843-2474 ABBREVIATIONS DP, dissolved phosphorus; NO3 N, nitrate nitrogen; NHa-N, ammonium nitroDen; PP, particulate phosphorus; TKN, total Kjeldahl nitrogen. . < < 1 �\ _\\�.'�- 1 ABSTRACT 2 Manure applications can provide nutrients and other benefits to turfgrass produetion and 3 unused nutrients in manure residues can be exported through sod harvests. Yet, unused nutrients 4 near the soil surface could be transported in surface runoff and be detrimental to water quality. In 5 addition to measurements of bermudagrass (Cynodon dactydon var. Guymon) turf responses, 6 volumes and P and N concentrations of surface runoff were monitored during evaluations of 7 composted manure applications in turfgrass production. Manure rates that supplied 50 and 100 8 kg P ha' at the start of each of two monitoring periods were compared to P fertilizer rates of 25 9 and 50 kg ha' and an unfertilized control. Two applications of [NHa]zSOa (100 kg N ha" were 10 applied with the P fertilizer. Three replications of treatments were estabiished on a Boonviile 11 fine-sandy-loam (fine, smectitic, thermic Ruptic-vertic Albaqual� that was excavated to create 12 an 8.5% slope. Compared to initial soit tests, nitrate concentrations decreased to 2 mg kg 1 and P 13 concentrations increased aRer two manure and fertilizer applications and eight rain events over 14 the two monitoring periods. The fertilizer sources of N and P produced 19% more dry weight 15 and 21% lazger N concentrations in grass clippings than manure sources. Runoff volumes did not 16 differ between manure and fertilizer sources of P, but average volumes recorded for the 17 unfertilized control were 22% greater than either source or rate of P during the second 18 monitoring period. Dissolved P concentration (30 mg L in runoff was 5 times greater for 19 fertilizer than for manure when rain occuned 3 d after P applications at the same rate. Similarly, 20 total dissolved P losses in zunoff above those of the control were 1.4 times greater for fertilizer 21 than for manure when both were applied in two applications at equal P rates (100 ka P ha 1 y') 22 Under the relatively large P rates on a steep slope of turfgrass, P and N losses in runoff during 23 natural rain events were no greater for composted manure than for fertilizer sources of P. � m z p � _ � � �'�- 1 INTRODUCTION 2 Additions of organic amendments, inciuding composted sewage sludge, can reduce soil 3 bulk density and increase water infiitration rate and nutrient holdin� capacity of soil in turf�ass 4 production (An�le, 1994). In addition, the amendments can enhance turfgrass estabiishment and 5 quality compued to fertilizer sources of nutrients. Aithough costs of haulin? and handling 6 organic sources of nutrients are relatively large (Daniel et al., 1998), the high economic values of 7 turfgrass, including sod, can offset those costs. 8 Despite agronomic advantages of sludge and manure applications on turfgrass, nutrient 9 concentrations can increase near the soil surface (Vitosh et al. 1973, Kin�ery et al. 1994, and 10 Lund and Doss, 1980). After mineralization, accnmulations of manure sources of P neaz the soil 11 surface are transportable as both soluble and sediment-bound P in surface runoff (Vitosh et al. 12 1973, Kingery et al. 1994, Romkens et at., 1973). Similar increases of P concentration in surface 13 runoff were observed as rates of P fertilizer on grassland increased (Austin et al., 1996). 14 Large nitrate-N (NOs-I� concentrations in soil can similariy contribute to losses through 15 surface runof�' In addition, inorganic N in fertilizer appiications is soluble in water and readily 16 transported in water flow over and through soil. Linde and Watschke (1997) indicated NO3-N 17 losses in runoff were largest in initial runoff events after fertilizer applications. Runoff losses l8 decline as fertilizer N dissolves and infiltrates with water into soil (Schuman et al. 1973). Unlike 19 fertilizer N, organic N in manure is released slowly through mineralization and nitrification 20 processes. Slow release of the manure N could minimize the portion of N applied on turfgrass 21 that is transported in water, compared to inorganic N sources. 22 Sediment and associated nutrients aze transported with soluble nutrient forms in runoff. 23 In the case of turfgrass, Linde et. al (1995) reported an inverse relationship between plant density J 1 and sediment loss. Similarly, Gross et al. (1991) observed less sediment loss at dense compared 2 to sparse seeding rates of turfgrass. The relatively lazge plant densities of turfgrass could reduce 3 sediment and nutrient loss in runoff compazed to grasslands used for grazing and fora�e 4 production (Romkens et al. 1977). 5 The use and eaport of manure sources of nutrients through turfgrass sod production has 6 been proposed as a practice for reducing P loads on watersheds containing large densities of 7 animal feeding operations (Griffith, 2000). Sod harvest can remove and reduce P concentrations 8 near the soil surface, but potential losses of P and N after surface applications of manure on turf 9 need to be evaluated. The objectives of this study were: l.) Evaluate turf quality and P and N 10 concentrations of turFgrass clippings and soil in response to increasing rates of P and N in dairy 11 manure and inorganic fertilizer, 2.) Compare volumes and P and N concentrations of surface 12 runofF between manure and inorganic fertilizer treatments, and 3.) Relate the rate and source of 13 applied P and N to losses in surface runoff. 01-11 14 . < 4 O 1 _ 1 2 MATERTALS AND METHODS 3 Plot Design and Treatments. 4 5 Common bermudagrass was established on a slope of 8.5%. The ent'ue plot area was 6 hydro-seeded at a rate of 50 kg pure live seed ha 1 during October, 1997. Irrigation maintained 7 soil water content for seedling establishment and turfgrass growth without runoff. Plot 8 dimensions were 4 m x 1.5 m. Sheet metal strips (thickness = 1.9 mm) were inserted 5 cm into 9 soil around the perimeter of each plot to contain runoff. Runoff of each rain event was collected 10 through an H-flume at the base of each plot into an uncovered, 311-L container. 11 Applications of composted dairy manure and inorganic fertilizer comprised five 12 treatments on the slope of bermudagrass. Three replications of the treatments were distributed 13 along the slope in a randomized complete block design during monitoring periods in 1998 and 14 1999. The treatments were: control (no P), 100 and 200 kg P ha 1 y 1 as manure, and 50 and 100 15 kg P ha I y 1 as inorganic fertilizer. Experimental results were analyzed as a split-split plot 16 arrangement of the experimental design. Two monitoring periods (1998 and 1999) were main 17 plots, nutrient sources were sub-plots, and nutrient rates were sub-sub plots within three 18 replications. A single control plot (0 kg P and N ha i y 1 ) was included in each replication. 19 Dairy manure was analyzed before application usin� methods of the Texas A&M Soil, 20 Water and Forage Laboratory (Parkinson and Allen, 1975). Tatal P and N concentrations in 21 composted manure averaged 5.0 and 15.Sg kg i , respectively. The rates of total P, applied as 22 composted dairy manure, were two times those applied as inorganic fertilizer to compensate for 23 the slow release ofP from manure. The inorganic P in fertilizer was assumed completely soluble 24 after prills were applied on the plot surface. The rates of P applied as inorganic fertilizer 25 maintained or increased ea�tractable soil P concentrations above 40 mg kg' and similar to the P � • C 5 p � _\�\lr 1 levels in more than 70% of soil samples submitted from selected urban counties of Texas (T.L. 2 Provin, Personal Communication). 3 Composted dairy manure was applied at the start of monitoring periods in 7une, 1998 and 4 Mazch, 1999. Each application supplied 50% of the total P rate. Similazly, inorganic P was 5 broadcast at rates of 25 and 50 kg P ha I to the respective fertilizer treatmems before runoff 6 monitoring began during each period. In addition to P, inorganic N(100 kg N ha 1 as [NH4] 7 ZSOa) was applied to the two fertilizer treatments. The N rate for each period was split between 8 broadcast applications before runoff monitoring started and applications 61 and 40 days later 9 during the respective monitoring periods in 1998 and 1999. 10 Turl'grass Responses. 11 Plots were clipped 3.8-cm above the soil surface when turf reached a height of 5 to 7.5 12 cm. The first clipping date occurred 17 d after application of both P sources during the first 13 monitoring period. Piant uptake of nutrients was quantified through digestion, and analysis of 14 clipping samples taken during selected mowing dates. Clipping sampies were dried and analyzed 15 for total N and P by the Texas A&M University Soil, Water, and Forage Testing Laboratory 16 (Feagley et al., 1994, McGeehan and Naylor, 1988). 17 Color, density, and quality of turfgrass in plots were rated visually. The monthly ratings, 18 startin� 5 d after initial P applications, were based on a scale of 1 to 9. Brown turf was given a 19 color rating of 1 and dark green turf was rated 9. The density of an open turf canopy with 20 exposed soil was rated 1 and a closed canopy of tillers and leaves was rated 9. Quality ratings 21 integrated consistency, color, density and aesthetics into a single numerical value. Quality, 22 density, and color ratings near 5 represenied an average turfgrass that could be used for a home 23 lawn or sod production 24 Volume and Nutrient Concentration of Runoff. ��F � < � 1 Total runoff volume was determined by multiplying water depth, as a proportion of the 2 maximum, by the container volume. Daily rain amounts were recorded for natural events at an 3 onsite monitoring station. Rain depth for the 24-h period in which measurable runoff occurred 4 was subtracted from the depth of zunoff in containers. After each runoff event, SOOmL was 5 sampled after mixing the volume coilected in containers of each plot_ The samples were frozen 6 immediately to prevent microbial Meakdown of nutrients within the water sample. 7 The particulate fraction of N and P in the SOOmL samples was removed during filtration 8 through 1-µm glass microfiber filter. The 1-µm pore size permitted suction filtering of the 9 sample volume without plugging by organic and clay colloids and total dissolved P(DP) in the 10 fiitrate could be analyzed through Inductively Coupled Plasma optical emission spectroscopy 11 (ICP). In addition, the glass filter disk and particulate fraction were digested to detennine total P 12 and Total Kjeldahl Nitrogen (TKN) (Parkinson and Allen, 1975). Total P in digests of the 13 particulate fraction was analyzed through ICP. The TKN in the digests and the NO3-N and NHa- 14 N of the filtrate were measured in an auto analyzer. The NOs-N was analyzed using cadmium 15 reduction (Dorich and Nelson, 1984) and the NF3a-N was analyzed colorometrically (Dorich and 16 Nelson, 1983, Isaac and Jones, 1970). The NI-7a-N concentrations were measured in runoff of the 17 first three events in 1998 and the initial event in 1999. 18 A tea analysis was completed for three soil samples taken at random across the 3 19 replications of plots on the siope. Each sample comprised 12 to 15 cores, which were 2.5 cm in 20 diameter and 7.5 cm in depth. The soil is described as a U5DA sandy-loam or sandy-ciay-loam 21 containing 56% sand, 24% silt, and 20% clay. The native soil, a Boonville fine-sandy-loam (fine, 22 smectitic, fhermic Ruptic-vertic Albaqual fl, was excavated to construct the 8.5% slope. 23 Each plot was sampled and analyzed prior to the initial N and P applications and after 24 each monitoring period. Ten to 15 soil cores (2.5-cm diameter and depth of 7.5 cm) were �\����r ,� • 4 7 ��_\\\�' i randomly sampled and mixed to provide a plot composite. Bxtractable P and NO3-N of the 2 sampie from each plot were analyzed by the Texas A&M University Soil, Water, and Forage 3 Testing Laboratory. An acidified ammonium acetate - EDTA was used to estimate plant- 4 available P{Hons et al. 1990) and soil nitrate was extracted and analyzed using methods 5 described by Dorich and Nelson (1984). 6 Statistical Analysis. 7 The Statistical Analysis System (SAS, 1988) was used to analyze variation of turf 8 responses, runoff volumes, and P and N concentrations of runoff and soil among monitoring 9 periods, rain events, P sources, and P rates. Numerical ratings of turf and weights and nutrient 10 concentrations of clippings were pooled over the sampling dates of both monitoring periods for i l analysis. The Generalized Lineaz Models Procedure (SAS, 1988) was used to analyze variation 12 of soil nutrients and of volume and DP, NO3-N, and NH quantities for runoff filtrates. 13 Variation of total P and TKN in particulate fractions of runoff was similarly analyzed. When 14 interactions of effects of monitoring periods with P sources and rates were significant (P=0.05), 15 monitoring periods were analyzed separately. Similarly, when interactions between effects of 16 rain events and of P sources and rates were significant (P=0.05), rain events were analyzed 17 separately. The P rates were treated as class variables in the statistical model. � � 4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 � RESULTS AND DISCUSSION Turfgrass Responses. Applications of N totaling 200 kg N ha I with the inorganic P fertilizer contributed to significantly greater ratings (P = 0.05) of turf color, quality, and density than the other treatments during the two monitoring periods (Table 1). The mean color and quality ratings oftreatments fertilized with inorganic P and N were 20% greater than control or manure treatments. The slow release of available N from the composted manure could have limited N availability to turf compared to treatments feztilized with [NH Similar to differences in visual ratings, the treatments fertilized with inorganic P and N yielded 19% a eater dry weights and 21% larger 1V concentrations of clippings than treatments supplied manure sources of P and N(Table 1). The lack of treatment differences in P concentrations of clippings (3.9 mg g') indicated variation of inorganic N suppiy was the principal determinant of larger ratings and yield of turf fertilized with inorganic fertilizer rather than manure. Soil Analysis. Variation of extractable soil P between the two P rates applied as manure or fertilizer was statistically si�nificant (P= 0.05) on sampling dates in March and June, 1999 (Table 2). The soil tests reveal P accumulation neaz the soil surface after applications of dairy manure and inorganic fertilizer. Large amounts of manure residue near the soil surface facilitate removal of manure-P amounts in turf sod that are much greater than P removed in biomass harvests of other grass crops (Crriffitt�, 2000). The slow release of available P from manure was evident in smaller increases of extractable soil P for manure treatments on the two sampling dates in 1999, despite two-fold larger P rates in manure than in fertilizer applications. The relationship between applied �i -1\ lY k �� 9 01-\t�a 1 and ezctractable soil P is consistent with previous reports (Vitosh et al. 1973, Kingery et al. 1994, 2 and Lund and Doss, 1980). 3 In contrast to P, soil NOs-N concentrations of all treatments decreased from beginning to 4 end ofthe study (Table 1). The decrease in soil IvT03-N is consistent with amounts ofN removed 5 in clippings (up to 36 kg ha and estimates of equal or greater amounts of N in grass parts 6 below the cutting height (Schuman et al., 1973). Although not measured in this study, leaching 7 and volatilization losses could have contributed to losses of appiied N and small NOs-N 8 concentrations in soil (7ohnson et al., 1995, Tennan, 1979). 9 Runoff Votume. 10 Runoff volumes differed significantly (P=0.01) among four rain events during each 1 I monitoring period and the first event did not occur unti160 d after the P applications during 1998 12 (Table 2). Asynchrony between rainfall and runoff measurements during 2 days of a pzolonged 13 rain event resulted in a runoff depth greater than the 24-h rain total for event D in 1998. A 14 portion of the rain recorded for event C contributed to runoff measured for D. In addition, the 15 antecedent rainfall of event C saturated the soil and maximized the portion af rain lost as runoff 16 during event D. 17 In addition to event differences, runoff volumes of the unfertilized control were 18 significantly (P=0.05) greater than treatments that received either manure or fertilizer P in 1999. 19 The average volume of the control was 22°lo greater than volumes recorded for either rate or 20 source of P. Relatively large clipping dry weights and density ratings for the two inorganic P 21 rates, which included 100 kg ha 1 of inorganic I�i (Table 1), were consistent with observed 22 differences in runoff volume. Runoff volume was expected to decrease as the plant density 23 ratings of the turf increased (I.inde and Watschke, 1997). . . io O�-ttta- 1 Nutrient Concentrations in Runoff. 2 An interaction between rain events and P rates was significant (P=0.01) for DP in runoff 3 during each monitoring period (1998 and 1999). In contrast to the initial rain event in 1999, DP 4 concentrarions in runoff for the latter three events in 1999 and all four events during 1998 were 5 relatively small (Table 4). Irrigation during the 60-d period between P applications on 6 bermuda�rass turf and the first rain event in 1998 reduced DP concentrations at the soil surface 7 and limited DP concentrations in runoff compared to 1999. The lazge reduction of DP 8 concentrations in runoff a8er the initial event in 1999 was similaz to previous studies of turf and 9 pasture (Edwazds and Daniel, 1994, McLeod and Hegg, 1984, Austin et al., 1996, Linde and 10 Watschke, 1997). 11 The vaziation of runoff concentrations of DP among P rates, including the control, and 12 between P sources was significant (P=0.05) on seven of the rain dates during the two monitoring 13 periods (Table 4). Significant interactions (P=0.001) between P rate and sources revealed greater 14 differences in DP of runoff between P rates of manure than between P rates of fertilizer for five 15 rain events. During seven rain events, vaziation of DP concentrations in runoff corresponded 16 with relative differences in P rate between fertilizer and manure P sources (Table 4). The DP 17 concentrations ofthe 100-kg rate of manure P averaged 2 times greater than the 50-kg rate of 18 fertilizer P for all events in 1998 and the latter three events in 1999. 19 The differences in runoff concentrations of DP between P rates were largest during the 20 initial rain event 3 d after manure and fertilizer applications in 1999 (Tables 3 and 4). In contrast 21 to seven other rain events, differences in mean DP concentrations of runoff of this initial rain 22 between the two P rates of fertilizer were 3 times greater than differences between the two P 23 rates of manure (Table 4). Similarly, differences in DP of runoff between each rate of P fertilizer 24 and the control were 3 times greater than DP differences between respective smaller and larger ! '., ll ��-��,� 1 rates of manure P and the control. The DP concentrations in runoff from the 50-kg rate of 2 fertilizer P were 206% larger than runoff from the 100-kg rate of manure P for first rain in 1999. 3 In a previous comparison between fertilizer and poultry (Gallus gallus domestieus) litter, 4 differences in DP concentration of runoff were greatest between P sources during the first 5 simulated rain event after application on tall fescue (Festuca arunciinacea, Schreber) (Edwards 6 and Daniel, 1994). Dissolved P concentration in runoff from fertilized tall fescue was 2 times 7 greater than runoff concentrarions after the same P rate was applied as poultry litter. The clipping 8 height of tall fescue was 2.4 times taller than that of bermudagrass in the present study. Yet, DP 9 concentrations in the initial runoff after application of comparable P rates were similar between 10 the studies of ta11 fescue and bermudagrass (Table 4). ll Similaz to DP, NO3-N and NH concentrations were largest in the first runoff event 3 d 12 after the manure and fertilizer applications in 1999. In addition, the N source by rate interaction 13 was significant (P=0.01) for NO3 N and NHa-N in runoff of this initial event during 1999. The 14 large NO N concentration in the initial rwioff of the lazger manure rate in 1999 was consistent IS with 5.3 times more total N in the manure than in the initial applicatian of 50 kg N ha as 16 [NHa}ZSOa (Table 5). Previous evaluations of plant uptake of N during the first year after dairy 17 manure application indicated 21% of the I�i in manure was equivalent to N appiied as fertilizer 18 (Klausner et al., 1994). The relatively lazge I�O concentrations in nznoff 3 d after application 19 of the two manure rates during 1999 indicated more than 21% of the total N in composted 20 manure was in nitrate form. Uniike the firsY rain event, the NO concentrations in runoff of 21 fertilized treatments were significantly greater (P=0.05) than manure treatments during rain 22 events B, C, and D of 1999 (Table 5). Larger NO3 N concentrations and losses in runoff from 23 fertillzer compared to manure or organic sources of N have previously been reported (Edwards 24 and Daniel, 1994, McLeod and Hegg, 1984). � 1 12 01-t11� 1 The variation of NO3-N concentrarion in runoff between the total N rates in manure or 2 fertilizer was significant (P=0.05) for five of eight events during both monitoring periods (Table 3 5). The NO concentration in runoff of fertilized piots was 2 to 10 times greater than controls. 4 Concentrations of NOs-N in runoff from the larger manure rate were greater than the control plot 5 for 7 of the 8 runoff events. Austin et ai. (1996) observed comparable increases in NO3-N losses 6 as fertilizer rate was increased. 7 The initia] application of [NHa]zSO4 with P fertilizer in 1999 contributed to 32 mg L� of 8 NHa-N in runoff 3 d later. Similaz NH4-N concentrations were observed in runoff of simulated 9 rainfall shortiy after N feRilizer was applied to tall fescue stands (Edwards and Daniel, 1994). 10 During the monitoring period in 1998, NH concentrations in runoff (3.2 mg L 11 days after 11 the second [NHa]zSOa application were smaller than the initial event in 1999. Irrigation during 12 the 11 d before the rain event couid have dissolved and transported the NHa-N into soii. In 13 contrast to observations after fertilizer appiications, NHa-N concentrations in runoff shortly after 14 composted manure applications in 1998 and 1999 were < 1 mg L (data not shown). Near-zero 15 NHc-N concentrations were observed in simulated runoff 14 d or more after poultry Iitter was 16 applied to tall fescue (Edwazds and Daniel, 1994). 17 Nutrient iosses in runoSf. 18 The potential for removing and exporting lazge amounts of manure P and N through sod 19 is an incentive for lazge manure rates that exceed P and N amounts needed for turf growth 20 (Crriffith, 2004). The volumes and P and N concentrations of runoff on the steep slope of 21 bermudagrass provide estimates of potential P and N losses and environmental impacts of the 22 large manure and fertilizer rates on turf. The DP amounts in runoff differed significantly 23 (P=0.05) between rates and between manure and fertilizer sources during 1999. During eight rain 24 events, the 200 kg of P in two manuze applications contributed 7.1 kg ha 1 more DP to runoff � , li 01-1\\� 1 than the control. A similar loss of DP in runoff was observed after two fertilizer applications 2 totaling 100 kg P ha i . The DP losses during eight rain events following applications of lower 3 manure rates totaling 100 kg P ha' were 3.0 kg ha 1 greater than the control and similar to two 4 fertilizer applications totaling 50 kg P ha 1 . 5 The portion of total P in turf cligpings and runoff attributed to manure (controi amounts 6 were subtracted) was only 2.8 to 3.8% of P applied during both monitoring periods (Table 6). 7 Comparable percentages of P in pouitry litter applications were collected in runoff during four 8 simulated rain events on a 5% slope of perennial grass (Edwards and Daniel, 1994). The small 9 amounts collected in clippings and runoff and extracted from soil (Table 2) indicate most of the 10 P in applied manure remained on or in soil and available for harvest with sod. 11 Similar to DP, NO losses in runoff differed significantly (P=0.05) between rates and 12 between manwe and fertilizer sources during 1999. During eight rain events of both monitoring 13 periods, 3.9 kg ha' more NOs-N was lost in runoff from the higher manure rate (two applications 14 of 267 kg N ha �) than from the control. The NOs-N losses in runoff of the larger manure rate 15 were 2 times greater than the lower manure rate and treatments fertilized with 200 kg N as 16 [NHa]zSOa. 17 Losses of NT3 in runoff soon after N applications revealed an advantage of manure 18 over fertilizer applications on tur£ The largest loss comprised 10.3 kg NHa N ha 1 in runoff 3 d 19 after 50 kg N was applied as [NH in 1999. The total NH4-N losses in runoff during two 20 rain events following N fertilizer applications were 2.9 times greater than total NO3-N losses 21 from the larger manure rate during all eight rain events in 1998 and 1999. The NHa-N losses 22 made up 40 to 42 % of total N amounts in clippings and runoff (Table 6). Similar to the NHa-N 23 loss from fertilizer, DP losses in runoff above those of the control were 2.7 times greater for the 24 50-kg rate of fertilizer-P than for the 100-kg rate of manure-P during the first rain event in 1999. � 14 Ol �111�- 1 An advantage of turf in a system for eacporting manure P and N was evident in negligible 2 losses of particulate forms of P and N after surface application of composted manure. Caiculated 3 total losses of PP and TKN after manure or fertilizer applications during the eight rain events in 4 1998 and 1999 did not differ from the control (Table 7). In addition, amounts of PP and TKN in 5 runoff decreased significanfly (P =0.05} in both years after the first rainfall event (Table 7). 6 Reductions in PP and TKN after the initial runoff event of each monitoring period could be 7 attributed to increases in turfgrass plant density over time (Linde and Watschke, 1997, McLeod 8 and Hegg, 1984). The density ratings and clipping dry weights (Table 1) indicate additions ofN 9 fertilizer with manure P could increase plant density and minimize losses of particulate forms of 10 P from turf. Yet, large runoff losses of fertilizer N compared to manure alone could be 11 problematic (Table 6). 12 CONCLUSIONS 13 The slow release of P and I3 from composted manure can limit turf growth and 14 quality compared to timely applications of soluble fertilizers. Yet, the slow release of manure P 15 and N resulted in smaller DP, NOs-N, and NHa-N concentrations in runoffthan fertilizer P and N 16 during rain events after both were applied. At equal P rates, runoff losses of DP attributed to a 17 recent application of P was 58% less for manure than for fertilizer P. Similariy, runoff losses of 18 DP totaled over eight rain events were A4% less for manure than for fertilizer applied at equal P 19 rates. Applications of N fertilizer with manure could increase turf quality and P and N amounts 20 in clippings, but timing of applications in relation to rain events will be critical to prevent lazge 21 runoff losses of N on steep slopes. 22 The surface application manure on turf optimizes potential removal and export of excess 23 P and N during harvest of the sod layer. One disadvantage of the large manure rates was evident 24 in the relatively large DP concentrations and losses observed in runoff, which could raise ,x , 15 0�-��1Y 1 concentrations of DP and accelerate eutrophication in surface waters (Daniel et al., 1998). The 2 observations of runoff losses on the steep slope did represent a worst-case situarion for turfgrass 3 sod production, but it is clear than manure rates need to be managed on a site-specific basis to 4 prevent edge-of-field losses of P and N in runoff. 5 REFERENCES 6 Angle, 7.S. 1994. Sewage sludge compost for estabiishmern and maintenance of turfgrass. p. 45- 7 52, In Anne R. Leslie (ed), Handbook of integrated pest management for turf and 8 omamentals. Lewis Publishers, Boca Rotan. 9 Austin, N.R., J.B. Prendergast, and M.D. Collins. 1996. Phosphorous losses in irrigation runoff � 10 from fertilized pasture. J. Environmental Quality. 25:63-68. 11 Daniei, T.C., A.N. Sharpley, and 7.L. Leymunyon. 1998. Agricukural phosphorous and 12 eutrophication: A symposium overview. J. Environmental Quality. 27251-257. 13 Dorich, R.A., and D.W. Nelson. 1984. Evaluation of manual cadmium reduction methods for 14 determination of nitrate in potassium chloride extracts of soils. Soil Sci. Soc. Am. J. 15 48:72-75. 16 Dorich, R.A., and D.W. Nelson. 1983. Direct colorometric measurement of ammonium in 17 potassium chloride extracts of soil. Soil Sci. Soc. Am. J. 47:833-836. 18 Edwazds, D.R, and T.C. Daniel. 1994. Quality of runoff from fescuegrass plots treated with ✓ 19 poultry litter and inorganic fertilizer. J. Environ. Qual. 23:579-584. 20 Feagley, SB., and M.S. Valdez, and W.H. Hudnall. 1994. Papermill sludge, phosphorous, 21 potassium, and lime e£fect on clover grown on a mine soil. 7. Environmental Quality. 22 23:759-765. 23 Griffith, E.N. 2000. Export of manure sources of phosphorus and nitrogen throu�h turfgrass sod_ 24 M.S. Thesis. Texas A&M University, College Station, Texas. 43 pages. F r. 16 0�-Il I� 1 Gross, C.M., J.S. Angle, RL. Hill, and M.S. Welterlen. 1991. Runoff and sediment losses from � 2 tall fescue under simulated rainfall. 7. Environmental Quality. 20:604-607. 3 Hons, F_M., L.A_ Lazson-Vollmer, and MA. Locke. 1990. NH 4 phosphorous as a soil test procedure. Soil Sci. 149249-256. 5 Isaac, RA., and J.B. Jones, 7r. 1970. Auto-analysis for the analysis of soil and plant rissue 6 extracts. P. 57-64. In Advances in Automated A.nalysis, Technicon Congr. Proc., 7 Technicon Corp., Tanytown, N.Y. 8 7ohnson, A.F., D.M. Vietor, F.M. Rouquette, 7r., V.A. Haby, and M,L. Wolfe. 1445. Estimating 9 probabilities of nitrogen and phosphorus loss from animal waste application. P. 411-418, 10 In K. Steele (ed), Animal waste and the land-water interface. Lewis Publishers, Boca 11 Raton. 12 Kingery, W.L., C.W. Wood, D.P. DeLaney, J.C. Williams, and G.L. Mullins. 1994. Impact of �/ 13 long-term land application of broiler litter on environmentaliy related soil properties. J. 14 Environmental Quality. 23:139-147. 15 Klausner, S.D., V.R. Kanneganti, and D.R. Bouldin. 1994. An approach for estimating a decay 16 series for organic nitrogen in animal manure. Agron. J. 86:897-903. 17 Linde, D.T., and T.L. Watschke. 1997. Nutrients and sediment in runoff from creeping � 18 bentgrass and perennial ryegrass turfs. J. Environmental Quality. 26:1248-1254. 19 Linde, D.T., T.L. Watschke, A.R. 7arrett, J.A. Borger. 1995. Surface mnoff assessment from ✓ e 1 � �- 20 creeping bentgrass and perennial ryegrass turfs. J. En�ironmental Quality. 87:176-182. 21 Lund, Z.F. and B.D. Doss. 1980. Coastal bermudagrass yield and soil properties as affected by 22 surface-applied dairy manure and its residue. 7. Environmental Quality. 9:157-162. 23 McLeod, R.V. and R.O. Hegg. 1984. Pasture runoffwater quality fromapplication of inorganic . 24 and organic nitrogen sources. 3. Environmental Quality. li:122-126. n, � 1� o1-11t�- 1 Parkinsoq 7.A, and S.E. Allen. 1975. A wet o�cidation procedure for detemunation of nitrogen 2 and mineral nutrients in bioloaical material. Comm. Soii Sci. and Plant Anal. 6:1-11. 3 Romkens, J.M.M., D.W, Nelson, and 7.V. Mannering. 1973. Nitrogen and phosphorous 4 composition of surface runoff as affected by tillage method. J. Environmental Quality. 5 2292-295. 6 Schuman, G.E., R.E.Burwell, R.F. Piest, and R.G. Spomer. 1973. Nitrogen losses in surFace 7 runoff from agricultural watersheds on Missouri Valley Loess. 8 J. Environmental Quality. 2:299-302. 9 SAS Iastitute. 1988. SAS/STAT user's guide: Statistics, version 6.03 ed. SAS Institute, Cary, 10 N.C. 11 Terman, G.L. 1979. Volatilization losses of nitrogen as ammonia from surface-applied fertilizers, 12 organic amendments, and crop residues. Advances in Agronomy 31: 189-222. 13 Vitosh, M.L., J.F. Davis, and B.D. Knezek 1973. Long-term effects of manure, fertilizer, and 14 plow depth on chemical properties of soils and nutrient movement in a monoculture corn 15 system. J. Environmental Quality. 2:296-299. 16 Y 4 6�-t��� y 'O O �L d ' M C �i O a+ O � � � � .-+ � G C. � � L d > O 'Q ... O O c N C 0 L d U c 0 V C d .; � C. � a "a A W A �c R • U � _ Q y b0 G Ri 1 � � i Gi.n� i 0 F C.1 � H b0 G . 4 � ; ; 1 �i C � 00 rl � M 7 � M 7 i .,� b0 � �I .�• oo r '» o M M M M M i � ��C ...r 1(� �O M 7 � � T � � � � i N M � N C vi �o �o Vi vi M 00 �O C1 N �A N N 7 �A 00 O� �O N O 7 N V� 7 N � ^� � � � N N V�] � � i � d < i QI L d d � � '� O y � U W � �� o a � L �%J Vl UI 'L^ w C G .. � �3 x � s � � Y � � z x N � � u i, � � O � � Q O •= U � O � L V h 'K N i � �% .� . V � L K II ` w � 7 � C C V Q � L d � � � � � ^^ i # V � U O L 4 O i a� p "�S a�i V ./+ sC � .,I�.�' h � i ��' � � � �" � x • � c° 3 >, K II � o� a a > d 'O b. � i.+ C � C c� p Q Q U R C. C d °i � 3 �� � d z�. U L ~ i! O � d � y ..' ¢ s, ..�.� Il p * p � ,� II .,^. p � O ic x � � t � v i �' � i V1 Qw O ""i '� Q„ y�, 'C �j� d d y � 6� Li �s, � � 3 ri y d 0 L y O � = U O ... c cC .> +' v v' iA � O y � X p � � � � y � R ..- C 'C 'r �y y R a r R� C dD CC y � �a d r. w n K a• c W i •"� h d �a (in i � •� i Fi �C � 'C � 3 �D � 3 w v�i -.'i'+ y w � � � 7 � +..� > 3 � . ..� K K a� r 'O d > Ti CC � � Q� � � � % R Q� a' E; C. G�i �' � �e w � a� � k � � ° .?'s � � °...' a � a �, = � i L O'3 r�i� 6� � 8� O �y �..� E"� � A �p 'X V1 wc �a+ � v �/. 'K^ K ti �, � 1\ 1 }� � d L � � � � U K o � b0 � C � G C � • y Q � � � .� Y � d � d � "a C .� £ � :.+ � C � O O £. w d � +.+ Q L � r. R: �' N C N � O � .� a+ � � a � �. U O 4 � w' 0. pD s�, r a +.+ a •O � L'.. N CJ U :-. O w £ a � � L p r C � R' i Z i y � Y � _ � � � N �C '�^ � d K s. � � d � F� a .. � H � !'r� !"I .�-i N N � � � � � rl �i 00 F, � N � � i i i � � � � i � � i � � � � 1 + .r L bD � N V�J a, x t. d O � L �. v y r" i O c� w � � � � Y�1 .-i N � � N N �' � � 1�l1 O .-� � L '3 C R' � y � h .ri 'X ri w * x x ,*� x .Y �' y .K ,�. o' x ri' C C y N H � � �i H N y ri {�i Zi d � r. L y e " o `� ;; a Q � ' L �a � '} � '3 U1 CSr V� I� d 7 � � � � `O a C y O L �..� � � � O � O ':, O L G:: C U R �C. �-1 R� O L � � � G O L p da L ,� � R G �„� � c U d .��' �+ p y Q �M E � U V � r, +� � d a �a, � a x � t ,.�+ � � �. F. � a � �- � 1 .� • _l. � � �,,, � � a � c r, � a '-I � � o � � r � ta � r'�.i '� 1 N 7 7 O� � � M �-+ N tV � � � 7 A � ai ' � � � 7 a � O�0 7 r N o, o � ,.r .-� ,� � � � � L � �. � � �.y w U � N M � � a i R � a „ , £ � M N 7 � H � w � y a = � N � N � � � ' +-' � r � �o 0 o A V eV O tV M C � fY. 6 � N in r e� � '3 y � �O N VJ DO M � ; e�+ .�-� �-1 � � � a i .. ,_ � �"' N � �--I V1 d £ e}' �-+ V� N p A i � d i E 'O O N � f H i � r � � w d � � o W �.� � Q� C� U A A � � V1 O N * � Q � � a c � � � on > c :r CC U �i R 'C y d L � w+ ' .j � O �° p O � � C y � � � �� �o d F° — � o � � O C Py % Vl 3.' p 'p'� _ �i ,�j,. � � •� W+ f.' a � U a � y t;, � L U � C c d �,� O p So�„ � C�C i � U ` R G' m 4 y,� v �. C.^ L: w � � p �*.. K L � x e� y �,,, C s�. +°�,' C >' c y x" o '�' > A = �" �^ vs P, W .a n.,,, 0 5 — \� \'+� � ri �i O 'G C'i � � C .� i. 3 W i d 4� 4 � O C i d w R p� 3 � e a; y a C � � i � .� o` G � d U V C R U � P�V � 'C a v O > •� h � �, c A °D C � 'i. Y d � C'' O H � � � � H � � � rl t � O� N t� �O lV N t7 'Q V er M C� 7 .ti �i ri .-i N R � � i�O N M C h .��i � N M 7 � �C m � � ' N � � M V � � N N M M �(. a bC �i' c�'i i c� a a v: ci H i ti f`1 N M h i m � �` ; v: '+ 'r t : ao e�-i � ey e�i ti ey N c, C� � H � x c. c+ rl 1 .; m r � � . Y r �o o v+ a? � �G C h C1 e-1 C�f ao a v v a ti ti tV 1V 'S � N V�i h O ey U O � C � U W � � . t x s .x �k x K X ; � i % • i x � � x • x + + x � * x • �' C � a : � � s x * W � � L = x • x « � x K # i . « � : . � x K � � � Y ,. L � L � ri C1. r d � Y Y �i d i "3 C � � � O a) > r. O v 7 q O y " i L d '-I N p � O � O � � y � R � ri � � �Q O R V� CC O H � e A; i y 0. � d � C � o � � �i d Q Q •� � � •y y C 1 �L .� � � C. � � C � O � V � �' R' + � vs u a�i � .�" � '^- x � y a+ Y� L Q �% E"� a+ +i �n %� � " • .4 b1-V\��- a 0 .� d a b� C . O w .� � � bD G .� � "'a w fJ d d w C. .� L L � w 6. O � O O O i C C O � G i w C d U L' O U Q� v G1 L ti z G R � � V � � ti h .__ � c� � � x a c� .�r x x r � e� o .-7 .+ o .-i .-i v, v� o e� O O O O C a � � I� N M V � ti C N N O ti i m ; ..^' 1 � � b d' � rl j C H 'rl C C a bL � c� � � 7 .y N N ef H � p ti eti C C i � � � N 'eT M M �'1 .ti � C N tV C O � � c� ti x a .ti �O M N 00 �p o v .-i Q .-i '. � M M N t7 C C o C C � i t� S t � � � O C M � z 'Y 'r ey '-1 N L h y L +" ' � ^j = • i � v w � y * � 2 % * * d y K * q X i' y � � G � � � e C.� • � � • s ,* x x x • • � . • � � y � C G a U d U L i L � z � � � � •� � a on '" � d i �+ ° � •� L O '-i �T. p U � d L' L K w ri C O O O i+ � V O � � � w , � r. '3 Li C R. � � N C C bD O 'y L � O � i CC N � ... a+ � ^ � � O O �� � � N 'X � x *� � z x� r ,, 4 � \ �\��}� L I OI N � C �L � � w L � « N N i OI � � � � L N m � O w O C 7 L W 0 N C O U l�0 w d Y � 7 Ci F. A C 'a C R � > 0 N N � C � R N � C . G V C � w t� Q j d � O � V � a � a R C pp R � Z � R C F +" C � � N N Q C R F � x N N J � e C d V d a N y N /0 y � L C t3 � Y R z m N 3 O a. L y Q R O J ' L O1 Q. w Y O H R r L R � � Y X N N O J � e C d i a N 41 R J L Y � d Y Z N N � O N J � � r Y � O m ~ � N � � O �„ R Z R' L +�-� Y � d L 3 O � Q ? tn a0 a0 6i o� N M � r � W � � c0 N I� � � � ti r � � � � � O O O O � O O N c'1 7 0 N � � � N � � N � � � � u'i N N I� N N N rn N � �1 N N 0 0 0 � c N N N O N O y U tL d 3 C N � O � v o � a � �o. U � �" . � R t6 � c 6f � C � O � � C � •L � L W � � C � � � a � N y C O C O � � .� O. y c6 V _ � C� O Q Z U 'Q m w a O � N C V � N N O � ."' C � N U C E O_ U • Q •O - O � �p_ Q N N � � N N N +- 'C p_ O O . N � � C 3 p N N O O - ' w � C C 3 3 � _ � C C � � � O O � . Q a Q (� � Z a �J « m . r-�... 0�-\�\s- Table 7. Total P and total Kjeldahl nitrogen (TKN) of particulate fraction of runoff for four rainfall events during monitoripg periods in each 1998 and 1999. Runoff was collected from Yreatments comprising two rates of either composted dairy manure or inorganic fertilizer before runoff monitoring began. 1998 1999 Event Total P TKN Total P TKN ----------- �g Plot 1 ---- ---------- mg plo£I ----- � 57 B 14 C 28 D 41 # MSDo.os 15 265 46 73 152 43 62 34 37 24 16 349 165 180 124 89 f Miniumum significant difference within columns using Tukey's Studentized Range, P=0.05. `' — _ Nutrients and Sediment in Runoff from Creeping Bentgrass and Perennial R` Turfs Douelas T. Linde" and Thomas L. �Vatschke ABSTRACT Althoueh scientist> fia}�e found 4+tde tcanspoct of nutrienu to date in runofF (rum turF_rasses. more research is needed on a x�ider range oEsuil conditions and mans�ement scenar�os. Thii studp ��as designed to as>e>s nutrient and sediment tr�nsport from creepin� bentarass (4grostis pa(usiris HudsJ and perennial rpegrrss (Lolium perersne LJ turG and to asse>s the influence that .�ertical mo..ing had on sediment tmnsport Sloped p�ots of ben[�rass and n e�mss. maintained simitnr to a golf f�irr�a�, ��cre ircisated to forte mnoff Cor the venera- tion oC mnoPF and leachatc �+�ater samples. About 1? h before each runoff e.�ent, iaigation •�as used ro equilibza�e soil moisture for ail plo[s. Foc four erents, pbts ���ere tre�red xith (crtilizer ai a rate of }.9 g N m'. 0.3 g P m and 4.1 g iC m'' about 4 h aFter pre-erent �r��gstion and S h before runotF. For another Four e�cnts, plots »�ere verticut 6 h bePure runoll- «�ter ssmples »ere anskzed fur NO tot�l I:je�dahl-� (T6ti), ph�sphate, and sediment. p]ean \0-�: wn- centrations rarelr ecceeAed 1 mg L Phosphete and TF1 concentra• tions and losses sieniticnnd} incre�sed �+hen runoll �+�s Forced 8 h after fertilization. On a�ersee fur these c��ents, ll % oC �pp��ed P and 2 applied ti»as delectvd in runull and IS % applfed P and 3 applied A' ��'as detected in Ie�chate. Fur all other e�ents nutrient concentrations and los>es e'ere consistentl} lo��er. ��ertical mowine h�d little afFect on sediment transporL 5ediment transport Crom both turfs a�eraged O.S k� ha On golF Eairwu��s, oll-site mo�'ement oF nutrients mar happen if rvnoff occurs soun aEter granular fertilizer is applied tu a neady saturated soil. I � THE RESEAFCH CO d3i2, S :izntisti hace found that nutrient transporc in runoft from turfgrass is small (blortor. et aL. 19SS; Gross et al.. 1990: Harrison et ai., 1993: Linde et al.. 199-'.)• Ho«'e�'er. additional research is needzd on a l��ider ran�e of soil conditions and man- a�ement scenarios beEore �ny generalities can bz made about surtace transporc of nutrients from golf courses. Harrison et al. (1993) found that conczntra[ions of NO ti and phosphatz in runoff and leachatz from turf maincained like a home lawn rarely excreded � and 2 mQ L respectively. Using crzzpme bent�rass and perennial ry'e�rass turG maintainzd Iike a QoIE fainv'ay. Linde et aL (1994) rzported that concentrations of \p—\, phosphate, and TK� in runotif and leachate rarely exceeded 7, �, and 2 mo L respecti�'aly. In fact. nutrient concentrations and losses (presentzd as loading rates) usually reflected those Eound in the «'ater used for irri�ation. For rcnoff ecents wichin ?=� h follo«�ing ierti(ization. the runoft contaiaed an avzrage of 1% oE the applied � and 3?% of thz apptied P for both turfs. The leachate contained an averagz of � 2% of applizd iv and 1�3°io of appliz:l P for both turfi. Linde et al. (199-i) concluded that for similar condi[ions on a eolf D.T. Lindz. Dep. oC A?-�onomc and Emiron. Scizn�'.. Dtlawara �'af- Icc Cufk?e. l0U E. Buder Ave.. Doy(zsto��a PA 1S90t: and T.L. \\�atichk:. Dep. o[ Ao•000my, Prnnsyl�a�ia S:ata Univ.. LL6 ASI Bld,_ (:ni��ervtc Park. PA 1650?. Receivcd I1 Oct. 1996. *Cor.z- sponding au[hor. � Publi>hzd in J. Enciron Qual. 26:12�1S-t?i (1997). 0 � -�\\d— fair«'ay�, it «ould be reasonable to assume that little off- site tran�port of nutrients from the fair�ca�' ��ould occur as a result of fertilization. Linde ec aL (199<} had ro limit s[atistical anat}'sis to individual dates bzcause major soil moisture difYerences esisted bzn�zzn e�ents and be- t«•ezn turf species. In the Simixed studies on erosion from turfgrasszs, scientists have found that turfgrasses �rzatic rzducz ero- sion compared to bare soil (�Vauchop� et a1..1990; Gross et a1.,1990,1991). Usin� a mised stand of bermuda�rass (Cynodon dacrylon L. Pers.) and bahiaarasi (Paspalum notnnun Flug�e vac. suare Parodi). «'auchope et al. (1990) found that thz a��era�z soil loss for simutated rainfalts a[ 69 mm h was 28 k¢ ha ` for bare p(ots and 3 k� ha ' for grassed plots. Using slopzd plots of tall fescue (Fesu�ca arundinacea SchrebJ. Gross et al. (1991) reportzd the a� erage soil loss for a 30-min, 120-mm h intensity scorm was 519 k� ha for bare soil and 54 kg ha`` for mature tall fzscue seeded at 4SS ka ha"'. Gross et al. (1991) concluded that even lo«' dznsity turf stands could si�nificantly reduce erosion and a«'zll-maintainzd stand should not bz a si�nificant sourcz of sediment. The studias conducted by `h'auchopz et al. (1990) and Gross et al. (1990,1991) used turfs maintained at heights >S cm. No published studies wzre found that included inEocmation concernin� soil loss from turfs maintained similar to a �olf fair« ay (about 13 cm heieht). In addi- tion, no studies were found that pro�ided inEormation on the influence that vertica! mowing for thatch manaee- ment had on soil loss from turfgras�. Since �'ertical mo«'- ine for thatcn management typically resul[s in the physi- cal removal of oraanic matcer and shatlo«' grooves in the suil. it is possiblz that soi( loss mac increasz. Crezping bentgrass and perennial rcearass are t«•o turf�rasses commonly used for �olf couriz fairways in the tamperate ciimate regions of thz li.S. Perennia! ry'egrass is a medium-tezturzd, bunch-ttipz species that does not form a dzfinite thatch layzr. ��'hen closel}' mowed, perennial rye�rass forms a turf «'«h a shoot densicy' bzt��'een 100 to 200 shoots dm (Beard, 1973). Crzepin� bent�rass is a fine-tzxtured, sroloniferous spe- cies that forms a definite thatch la}�zr. �Vhen closely mowed, creeping bentgrass forms a turf �cith a shoot densin >200 shoots dm '- (Bzard, 19i_). 7'he objec[ives of this rzsearch «'zre (i) to assess the transport of KO N, TKN, phosphace, and szdiment from the turfs and (ii) ro determine the influencz that vertical mo«�ino for thatch mana�ement had on sedi- ment transport from the turfs. DiETHODS AND DIATERI�LS Sic established tucf runoEE ptots ustd b}' Lir.de et al. (199-' anS 199�) werz used for this smdy. In 1991, tF.ree piots (each 6� m«•idz by 19 m lona) «'erz establishzd ro'Pznneaslz' crzzpin� bentgrass and thrzz to a perennial r}'e�rass blznd izas n � _���� LI�DE & w��.TSCHKE: RC�OFF FRO>t SE�"IGRASS A�D RY'EGRASS Tl'RFS 1249 ('Cita:ion II'. 'Commar.dzr'. 'Ome�a II'). Hacrison et al. .-eTticat mo�cing. Flots rzce:� thz usual pre-even: irri�ation (1993} characczrized thz sue as having a•�ariablz slopz be- to equilibrate soil moisturz. It µ�as hcpothesized that the rz- na-ezn plots (9-il% 2ad a sudace so�t that «"as a se�'erzlv bl�dzs. aad the oriznt tion of [he� b�omn �h erodzd Hzazrsto�.�n serizs classitizd as a clay (0.23 kg ke . sand. 0.36 kg ko ' siic. 0.41 ka i:g '��a�). The local seologp slope ticould form preferen.ixi f!o�v channels for runoff and �cas a frac[urzd karst and depth to bedrock ranged from � to se Rur.oif anater san pl s�iere tak n an �'fodei 60 cm. At the bzginains of this stud}' in 199�, the top S em _ �Q� orjable «zrer samplzr (ISCO. Ine.. Lincotn. �E). To of soil had a pH of 7.1. P iz��el of Si ks p ha '. and K lecel F of 26"I kg K ha Plots «'e:e mowed ro 13 mm ���ich dippin�s eser2et a runoff samptz. ihz potyeih}'lenz samplin� access tu z tur�f thac � ould be ca112 found on a olf fa�nat}iY The sampler���as intzrfaced �nith an ISCO p4odel �0 flo� Each ploi containzd 21 R'zathermatic (Garland. TX) pop- mecer chat «�as programmzd so that after zcery i5.6 L of p onvas an �coaced con e t [e he e bo lha m collecez a r�u�oa mL sa ple of runoff «ater in thz splitiing chamber. Thic and direc�eand sam equipmzna \Vatzr the chute r�� utes a d an� a��era z unoffb olamu for� turE pre- mzasuring P floa'ed into a polyethylene splitcin� chambu (for runoff sub- `i�iz eom tzd int c� o�00-mL(boulzs. The 40 mLsamples samplz cotlzction) znd in[o a parti[ionzd steel tank (Harrison from thz f rst 907 L of runoff �verz composited into the firsc et al., 1993). FunofE �olume and rates �tzrz mzasurzd wing ' an I 0 dates ` from J nz 1994 totOctOoberC199?��runoff was bhe co d�botde. Nhen r�u off �aas L hzne ach botde forced w�ich irri�ation at a rate of 139 mm h to generate contained �SO mL of w'ater. Samplin� «'as stoppzd after the thesz irr �atzd e��ents Pe�e cor.ductz [ appco�imate]}'. e e�z Y L, lhz�n zcond concained �SO mLnRunoff aused Z�ck, dzpzndino on �i'eather and availabiliq� of laboc Durin� by rains[orn:s �cas measurzd and samplzd �cith the equipment to�a[natual`siatz ethreoard alzd99�)bzcauseinsthoe,studesrunoffflo�.'ratescausedby' <10 n�pla Based�on`data from aL (199�)�irrigation In lsfl�rms d de ot and samplers forthis� smdy �ere 1[m¢O duration ��'as set at 2� min for bent�rass plots and 1� min to onz. for ryzgrass plots for most ecznts. The purpose of different L,zacha[e water ���as sampled from four pan-lysimeter cotumesfor tYle[toOluiSSSlOC2 tat2LS847}Jlltt�rwaSElO�SQa ea fabz of each (Hahri on et Abo u I> after Fur the 6 and 24 Szpt. 199� events, both turfs �i�ere irrigatzd for�each plo[ t �Ppakin,, q ai an oun[s from of tl�ie for 2� min each. four sam lers. Approzimaeely 12 h prior to each runoff ecen[. the pl�ots "�a�z,.pamplzs «'zre analyzed for NO,-�. TKN, and phos- .eere irri�ated ��'ich a senes of shocc duracion (2-3 mia) irri�a- phate (orihophosphatz) aceordin� ro the procedures described tion sets at a ratz of li9 mm h"' un[il runoff was �'isua4ly one o four depe d b g n thz antzcedeen soil mo sture co t nt ru off arerz uszd to�cal ulat nutnent t o r s s tion runofa for The purpose of thz pr� zcent irrieaciors H•a� to equilibrate acerage of bo[h mrfs. Total runoff from each piot and thz the soii moisture content for all plo[s so that data comparisons ave.age of the wncentr2cions in the first and szmad flon'- bet�iezn dates could bz made. Lindz et aL (1994) did not paced zunoff samptes w'erz used ro calculate losses. The �ol- equilib[a[e soil moismre bzforz irriga[zd ecents, thus they had ume of «�ater that could be held in one subsurface sampler � � 1 L w�as used ro catculatz nutrient loss in leachate. Sedimznt concentrations of the runoff samples �vere detzo- ro limie comparisons to individual dates becausz major soil (-� moistute differences esisred between dates and betwezn [urfs. mined �racimetrically by measurin� the amount of inorgamc For four oE the runotf z�'znts, ptoes a�zre fertilizzd w�ih a �atzrial tra ed by' filtec papet (1 µm diam. pores) aftzr 19-1.3-li.S (N-P-K) fertitizzr (O.�l. Scott & Sons.:�larysville, Fp . filterine the enure kno«�n samplz colume. Filters «'zre place OH) usin, a broadcas[ spreader at the ratz of 4.9 g N m - � p_, o p m and �7.1 g K m about 3 h afcer prz-e��en[ irngation in a o�en at 42�`C for S h and then wzighed. It «as hypo� z 2nd S h prior to thz runoft event. The f:rtilizer contained sized thai the soil dismrbance caused by �'ertical mo«"�no � 0.6°rb NH,-�'. 15 �%6 u;ea-N coated to procide 73 slou �i'ould tikzh� incrzase szdimznt trxnsport shortly after mo«�n° ' relzase �. P dericzd from monoammoniem phosphate. and and thzn decrzase as the rurf reco�'°-rzd. Treamtents (turf species) «erz arranazd in a random�zzd ' K from Ii Beforz fertilizer applica[ions. rmtoff collzction cum letz bluck dzsign ��'ich [hrze replieacions and blocking ' .ezirs locatzd at t4e boctom of each plot «ere co.�zrzd «'i:h bas d on suriacz siope- In a concurrent runoff study b}' Lindz ' plastic to psecen[ =ranules from enterina an}" part ot the weir. 1996 . i: «as dztermined that the pre-e�'znt irrigation procz- dure �cas zffecticz in equi(ibrating the soit moismre contznt ' On sz�"en othzr datzs, supplzmental maintznance N ti�'as ap- � plizd as a liquid or aranular application of urea (46�-f��) a� {or all plots beforz runoft. Thz}' found tha[ [hz a�'eraee soii � 2.4 or �J g� m'. ' columzvic ��'ater cuntznt of the ploU ius[ prior to irrigatzd ' Abeu: 6 h prior to another four e��znts. all plots w'zre czrti- 0 0' in 199?. Thus. ' cuc once using a R}an >tata�cay ��ereical mo��zr \4ode154�3� 3 I`� s�`z` �o m ' arisons�bzc�' 9 een z''a OS °°� izn[ and sedimznc � (Cushman. Inc. Lincoln. �E) [hat had 20. 1.6-mm-��'ide blades p i blades� pznzt atzd the soii appro�an2tel�d3Smm.aPlots az usins a mzasures analbs seofbariance bacau<z � �eere �'erticut Izngth�rise do«n the stope. Thz majority of the rzpzated measurements ��'erz made on thz samz experimznta ' ,:� J�b� � !�: `_,,.. ;n 8 =a�(�rc hv .+- mo.rino �vas units o�zr time. ,� , factor. O ecperiment �va<_ : �;� �,�, t��c .� �_ ni-,., „F t�mr � eolleC[ad by rakin� and ��ei�hed. A�;}>;����°.�lci�� ti Ii boio�� i j7_jO 1. E�VIRON. QUAL. vOL.?6. SEP'i'EMBER-OCiOBER 1997 Tabie 1. �lean n co ncentrations for 199 irrigated e � tiitrnte-.S P h osp hate R unoEF sample+ Runoff sa mple Date FPl FP_' Leacha[e FPl FP? Leachate 29 June 13 ]uh ' p� ?? .�.u�.4 3 Sept. ?0 Sept.7{ S Oct. se 1.0 : \S 0 03 \S 03 \S 0.1 \S OZ 0.1 0.6 O \5 0 C �J 0.3 �5 0.1 r5 0 \5 0.? U.t OS \j 0 \S 0 !�5 OS \S 03 \S 0 \S 03 0.? 1.67 2.�1 \S 2.71 1YS 4.13 1.90 3.85 0.89 0.p'_ '** "" Significant at the 0.05, 0.01, and 0.001 probability� levels, respectivei�'. i FPl = lst �u�s-paced runotT sample, FP2 =?nd itow-paced mnof5 sample. - Mean comparisons are for adjacent d�tes hithin columns. §\S = not significant; SE = standard error. Q Fectilized 8 h beFore e�'eat at �ate of 4.9-03-1.1 g m' oE N-P-K. arran�zd as a spfit-block in a randomized completz block desi�n accordin� to 5[zzi and Torrie (1950). Usin� the Huynh- Feldt epsilon valuzs calculatzd by� thz repeatzd statzment in SAS's �eneraV linear model proczdure (SAS Inst,1990}, Linde (1996) dztermined that levet� of time were indepzndent for the runofi data. Sincz levzts of time were independen[, [hen [hz Ftasts of thz usua( sp(it block analysis werz valid and a univariate analyss w'a> conductzd that provided information on the Izast square mzans and thz probabilities associated �cith thzir compa;ison. Thz prz-plannzd comparisons for this study included comparing spzci;s within days and comparin� conszcutive days within spzcies for thz variables neasurzd. RESULTS AND DISCUSSION Nutrient Transpod Mean nutrienc concentrations, bv evznt, sample type, and }ear are przsenizd in Tablas 1 ar.d 2. The concentza- �������� Total Iijeldahi-ti R un o EE sample FPI FP'_ Leachste mg L I.li O.T 0.0? 0 0 • \53 \S \S � Z.Ol Lli 0.03 O.L' 0.19 �S \S �S �5 1.76 1.13 0.39 0.13 0.0�4 iER 1y ids '� 4.17 1.95 b.3-! � 90 3.93 �}a #Ri s Ysi x 151 0.83 130 OSl 0.86 a. t:. :<s 2.66 ?.4i 5.78 2.53 ?.30 . war 0.61 OAI 0.0? 0 0.06 0.?? O.tS 033 0.13 0.10 tions reported are those detectzd in thz samples minus the concentration found in the irrigation or rain watec for each date, thus they represent the nutrient contribu- tion from the turf plot alone. No si�nificant turf specizs effects werz found on any date for any sample type; [herefore, values for both species ���ere averaged. Sig�ifi- cant date efiects, ho��ever, were found and are pre- sented as comparisons bzt��'een ad}acent dates in Tables 1 and 2. Nutrient concentrations, dzpth of water applied and depth of runoff (Table 3) wzre used to calculatz nutrient inputs and losses. \lean NO concentrations w'ere alwags found to be lowzr than the 10 m� L drinking watzr standard set by the USEPA. This findin� concurs �rith NO concentrations found in turf runoff studizs done by Linde et al. (1994), Harrison et al. (1993), Gross et al. (1990), and blorton et aL (1933). The hi�hzst mean Table? >Iean nu[rient concentrations for 199: irrigated e Ciittate-N Phosphate ' Runoff sample` RunotF sample k D at e F FP? Lea<hate FP1 FP2 Leachate m� L 1 16 �Sm 0.6 �1 U:? 1.?7 09J L01 ' - " rS$ x <. . 31 Afar^,I 13 L� 0.1 9.96 7.67 d.?d x. ... �5 x: . . 11J�me 03 0.6 _ 0 253 1J9 1.2? P$ s.� \S \S \S � ?83unt 0? 0.? 0 133 1.60 0.99 \S n \S .-.. ... .., 1? July¶ OS 0.� 0.1 1�.39 5.51 1.9? �s �s ».. >k. ... .,. ?6 Jul. Ob OS 2.6 219 I.85 1S6 ' � \S *"" *• \S °` \5 6 Sept. 0.6 0- 03 ].95 7.30 Y3i i i ti5 r5 �5 NS \5 1 23Sept. OA 01 0 1.i7 0.93 I.Ol � SE 0? 0.03 OS 0?3 0.09 039 "° $ignilicant at the 0.05, 0.01, and OAO( pcobl6ilit� IereB, tespecti�el.. - FPl = lit flo�.-paced mnotF sample, FP2 =?nd fluw-paced mnotT sample. �'• ` Dlean comparisons are for adjucent dstes x[�p(n c0lumns. 1 I. § NS = not s�gniFicant; $E = stsndard error. f ¶ Fertilized S h befoce e�ent at rate of 4.9-03-1.1 g m oF N-P-Ii. ���: Tot Kei d�hi N RunofF sample FP1 FP2 Leachate 0.�8 S.J2 ♦S� 0.�7�1 a.na 553 o �, 1�5 0.3? \5 0=9 0.09 0.20 3%3 031 s 0 3.23 �r< 0.18 �S 0.06 \5 0.?? 0.09 0.�33 2.00 0.83 • 0.0# 2.60 0.25 C�S 0.37 \S 0.?7 0.26 G�-1\\�- Table3. «'ater applied and mean total runoff for bentgrass and n�egrass. Li\DE g w'ATSCHKE: RU�OFF FRO�t BE�TGRASS A\D RYEGRASS TI;RFS {t'alec applied Date Bent Rye 1993 29 ]une 13 3u1y 14 ]vl.i ai ���: , >? ]uir ? Aug: 17 Aug.: 2'_ Aug. 3 Sepi. 17 Septi 20 Sept. 8 Oct. 1 No�:9 at �o..� Z$ tiov.T 1995 16 bfay 31 Dtav 13 ]une ?S June 6 Julyi 12 July 26 Julr 6 Sept. 23 Sept. ZO Oct3 5& 58 �; is � 58 93 58 58 3tS 5g 58 26 ia 49 58 SS Sg >$ Z.8 58 93 cg 58 93 3j 2% is � 35 93 35 35 38 35 35 26 is 49 35 35 3> :s �g 3> 93 53 <g 93 ; RainC�ll e+used runofi. }fean total runolf depth Bent Rte 16.1 1>.8 0.1 0 ?.1 15.1 o.s 79.1 173 0.8 za.a ?OS 0 o.a 3.2 15.5 18.6 23.5 18.7 I.1 21.1 13.2 17.9 16.9 13 135 1L7 I.9 0.6 33 1�? 6.6 13A 30.8 0.3 15.0 L? 0.8 ia 0.7 SL7 11.'_ 1?.9 11.0 0.6 123 9.6 2?1 19.7 3.2 1�0 N concentration was 2.6 m� L that �a'as found in the leachate samples on 26 July 199�. Runoff and leachate losses of NO N�cere consis- tendy lower than irri�ation inputs of NO� N(Tables 4 and 5). For both years, mean losses in runoff, which ranged from 0 to 0.02 g m'-, �vere lo«er than mean losses in leachate, which ranaed from 0 to 0.16 � m Linde et al. (1994) reported similar findings; however, they described nutrient losses as nutrient loadin� rates. For all samptz types, phosphate concentrations si�nif- i2sz icantly increased for e��ents conducted 8 h after fertilizer application (Tables 1 and 2). Concentrations «erz sie- nificantiv less bv thz nest evznt. usuallv conducted within about 2�ck. The highest mzan phosphate conczn- tracion found in the flo���-paced runoff samples �cas 1039 ma L for an e�•ent conducted S h after fertilization on 2I Ju]y 199�. Escludino the events which fertilizer was applied S h prior, phosphatz concentrations �cere similar to those found by Linde ec al. (199�). Theg applied the same rate and source of P. monoammonium phosphate, to the same turf plots used in this study and reportzd 6.06 mg L as the hi�hest mean phosphate conczntration in runoff. ���ith most concentrations <3 mo L Linde et al. (1994) found little indication in runoff or leachate samples thac P fertilizzr had been applied approximately 24 h before an irri�ated e��znt. The hi�her phosphatz concentrations detected for events immzdiately follo«in� fertilization compared to findings by Linde et al. (199�) could be attributed to hieher soil moisture contents as a result of pre-event irrigation used in this study. The soil was likely wetter (near saturation), thus a �reater portion of the soluble monoammonium phosphate fertilizer couid move off- site in the runoff. Thz a�erage soil volumetric water content of the plots just prior to irrigated events was 030 g k� in 1994 and 0.40 g k� in 199� (Linde,1996). Soil moisture levels were likely less in the study by Linde et aL (1994) because they did not use pre-event irrieation to equilibrate moismre ]evels. Based on soil test results for 6 Apr. 1994 and 23 Au�. 199�, soil P levels in the top S cm of soii �vere �enerally in the low range. In 199�, P le��els averaeed S� k� P ha ' and ran�ed from 73 to 11Q k� P ha In 199�, Izvels avera�ed 73 k� P ha and ranged from 36 to 91 k� P ha `. Since soil P levels �cere low, then excessive levels of soil P�rere not the cause of the increased P transport from the plots. Runoff and leachacz losses of phosphate-P were often Table 4. 199A Chronoloa� of inean inputs and mnofl and leachate losses otrotal N(TKN + 1�0,-\) and 10 for bentgrass and q�egrass. Irrigation and t inp uts Runoll losses Leachafe los Fert. BenL R}' e. Bent. Rye. Be nt. R)e. inpu<s Date of'.� \ \Orti \ \Orti N 10r\ 1 \Or\ \ \OrV \ \O�-\ 7 June 3.7 29 June 0.31 0.?8 0.19 57u1y 2A 53 Sutv 0.}i 0.3i 0?0 13Iuk"r 0.03 0.0= 0.03 _'1 Juhi 0.05 0 O.Oi Z' Jul�fi O.OI 0.01 0.03 ? Au�. 0.31 0.38 D.?5 b �ug. 2.3 17.4ug.� 0.01 0.01 0.03 22 au� 9.9 0.?7 D.?7 0.76 3 Sept. 0.?S 0.28 0.17 17 Sept.�= 0.07 0 0.07 ?0 Sept 4.9 0.?6 0.26 016 S OcL 0.?7 0.?7 0.16 9 Oct. ?.4 I \o.�.�: 0.01 0 0.01 21 \ov.i� 0 0 0 23 \ov.>: 0.01 0.01 0.01 T Rainfall caused runoR. _ Leachate samples xere not coliected. 0.17 0:20 0.0? 0 0.01 0.23 0.01 0.16 0.17 0 016 0.16 0 0 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0 0 0 0 0.01 0 0.0? 0 0 0 0 0.01 0 0.0? 0 0 0 0 0 0 0.09 0 0 0 0 0.01 0.01 0.05 0 0 0 0 0 0 0.03 0 0 0.01 0.01 0.0? o.ii o o.io o.oi o.�� 0.01 0 0.01 0 0.01 0 0 0 0 O.11 0 0.06 0 0.03 0.01 0.01 0 0 0.0? 0.01 0.13 o �+_i 0.01 0.05 0 0.09 U.02 0.01 0.03 0 0.01 0.09 0.03 0 0.13 o.aa O.OI 0 0.01 1��� J. ENV(RO�. QUAL. VOL.26. SEPTE?��BER-OCCOBER 1997 r O�-���a- 'I Table 5. 1995 Chro n o ioa,v of in ean inp and runuFf and l eacha te losses of totnl ( TF\ + NOz-ti) and ti Or� for bentpctcs and q�egrass. Icrig a[ion and xainfali input R unoE F l oss e s Leachate lozses Fett- BenL Rpe. Bent. Rre. Be R ve• inputi Date oE\ � �Or� � �Oy� � �O,-� � �Or1 � \Or� � 1��r� gm : 16 Diac 0.?9 0.� O.IS O.li 0.01 0.01 0.01 0 0 0 0 0 22 JIa�' 37 3Illa�� 3.9 0.?J 021 01> 013 0.09 34lune 0?0 0.?0 01? 012 0.03 ?3 June 0?6 0?? 0.15 0.13 0 6luh�; 0.01 U.Ol 0.04 0.01 0 12July 4.9 031 018 01S 017 0-09 26 Suh 635 0.?L 0.1> 013 OAS 31 Aug.?-3 O.L' 0.01 O.L' 0.01 1 Sept. 4.9 6 Sept.II 0.17 0.13 0.17 0.13 0.01 2a SepLn 0.20 O.1S 0.?U O.1S 0 ? Oct. 4.9 20 Oct.'s_ 0.01 0 0.0� 0 0 � R�inCall evused mnol£. _ Leachate ssmples eere no[ collected. § Lightning disabied runoff inea�uring equipment. ¶ Both turFs were init�ted at same duration (?5 min). greater than irri�ation inputs (Tab(es 6 and 7). For both years, mean P losszs randed from 0 to 0.06 g m' for runoff and 0 to 0.07 g m for leachate. Since the hi�hest losses were found for events that had P fzrtilizer applied S h baforz runoff, then a portion of the applied P was transported in runoff and leachate. For esample. on 22 Au�. 199=4, an avera�e of S% of the applied P was de- tected in runoff and 7% in leachatz. On 12 July 199i, an averaoz of 17% of the appli�d P was detected in runoff and 20% in leachate. Linde et aL (199�) found mean P losses (repo.ted as ]oadino ratzc) ran�ed from 0 to 0.01 � m for runoff and 0.01 to 0.04 g m for leachate. Mean TKt�'� concentrations followed a simitar pattern as phosphate. Conczntrations sionificantly increased for events conducted 8 h aftzr ferti(ization (Table 1 and 2). blean TK\ concentrations range3 from 0 to 6.84 m� L'' in 199-4 and 0 to 5.�8 me L in 199�. The higher concentrations werz dzrec[ed in the first flow-paced run- Tabie 6. 1994 Chroaoloe,v of inean inputs and runoff and leachate losses of phospha[e-P for bentgrass and q�e�rass. Grisution and rainF�}� Runoff Leacfiate Fert. p�nput losses oP P los5 af P inpu[c Date o(P Bent Rre BenL R�e. Ben[. Rye. a m . 29 June 0.0? 0.01 0.01 0.01 0.01 0.01 13Juh� 0.03 0.0? 0.01 0.01 0.01 OA2 11]ul��� 0 0 0 0 0.0? OA3 ZI Suiv` 0 0 0 0 0 0.03 3? Juii�: 0 0 0 0.01 0.01 0.03 Z:1u�. 0.0'- 0.01 0.01 0.01 OAl 0.0'_ 17,aug.� 0 0 0 0.01 0.01 O.Dl 32 Aug. 0.3 O.OZ 0.41 0.03 0.02 0.0t 0.03 3 Sept. 0.02 0.01 0.01 0.01 0.01 0.01 17 Sept.:_ 0 0 0 U ?0 Sept 03 0.0? 6.01 0.03 0.01 0.03 0.03 S Oct. U.Oi OAI 0 0 0 0.01 1 �us.`_ �) U 0 0 21 Nu�.*- 0 0 D 0 38 \o�.�_ 0 0 0 0 i Rainfall caused mnOtT. `y Leach�te sampies were not coilected. 0.0? 0.07 0-Ol 0.06 0 0.10 0 0.0'_ 0.01 0 0.0.1 0 0.03 0 0 0 0 0 0 0 0 0 0 0 0.01 0.07 0.01 0.09 0.01 0.1? 0 0.01 0.01 0.01 0.06 0.05 0.17 0.16 0.01 0.0? 0.01 0.06 0.02 0.07 0.02 0 0.01 O.�I 0.01 0 0.01 0 0 0 0 off samples for events immediately followin� fertiliza- tion. The TKY ]evels �cere often higher than levels found by Linde e[ aL (1994). They reported that TK?� concentrations ran�ed from 0 to 3.� mg L Like phos- phate, the higher TIiN levels for this study �rere attrib- uted to hi�hzr soil mois[urz contznts as a result of pre- event irrigation prior to fzrtilizer application. The TKN and NO N results w'ere added together for estimates of total I�'. In both yzars, despite relative increaszs in total N losses for events that fertilizer was applied 8 h before runoff, totat A' losses remained consis- tentty lo«•zr than irrigation inputs (Tables 4 and 5). ln 1994, mean total N losses ran�zd from 0 to 0.11 g N m''- for runoff and 0 to 0.21 g N m for leachate. In 199i, mean losses rangzd from 0 to 0.09 g N m for runoff and 0 to 017 � N m'� for'leachate. Linde et al. (199=!) repocted similar numbers; however, total N losses �vere szldom greater than NO losses. In the currznt studv, totaf \ losses were much hi�her than NO r 2able 7. 1995 Chronolo�y of inean inputs and runoRand leachate losses oF phosphare•P For bentgrass and ryegrass. Irri�ation and cainfni( RunoR losses Leachate Fert. p� oF P losse ot P inputs Dafe ofP Ben1. Rye BenA Rye. BenL Rre. gm' 16 �[a� 0 0 OA1 0 OA1 0.01 31 �ta. 03 0 0 O.Oi 0.03 0.0> 0.06 li June 0 0 0.02 0.01 0.01 0.01 283une 0 0 OAl OAS 0.01 O.OI 6 July�- 0 0 0 0 i? Julr 03 0 0 0.06 0.0� 0.05 0.07 26 Julr 0 0 0.01 0.01 0.02 0.0'_ 314ue.>;§ 0 0 1 Sept. 03 6 Sept11 0 0 0.01 0.01 0.02 0.0' ?15rpt.Q 0 0 0.01 0.01 0.01 0.01 ? Oct. 0.? ?�) Oct.+= 0 0 0 0 i Rainfull caused runoff. - Leachate samples aere not colle<ted. � L�ghtning disnbied runofE me:uuring equipment. �� Both turts were irrivated at same duration ('_5 min). � L1SDE 8 WpTSCHKE: RL'�OFEFRO\f BE�TGRA55 A�D RYEGR:ISS TIiRFS losses for e�'ents that had fzrtilizer applied, thus a por- tion of thz applied �. albeit small. «�as transported in runoff and lzachatz in forms other than \O_ N. For esample, on 22 Aug. 199�. an a��erasz of 2% of the applied \ �cas detzcted in runoff and about 3% in lea- chate. Gross et ai. (1990) also found that ` losses �cere areaczst uhen runoff occurrzd soon after fertilization of a tall fescue!Kentucky blue�rass turf. On 11 datzs durine the study period (�fa;� 199a-Nov. 199�}. dztectable amountc of runoff (>0.6 mm h oc- curred due to rainfall. Runoff caused bv rainfall often did not pro��ide complete data sets because runoff did not ahcays occur on all plots. Thzrefore, mzan nutriznt concentrations for rainfall events «�ere based on an avzr- aoz for both turfs and thz number of plots that provided data. Statistical analysis was not conducted. On 31 Au�. 199�, rainfall produced measurable runoff from each ptot. ho«�e�zr. li�hming disabled the runoff ineasurin� dzr'ices. �4ean nutrient conczntrations in runoff from the rain- fall events (Table S) were slightly higher than those found in the flow-paced runoff samples from the irri- �ated e�ents ��ithout fzrtilization because thz volumes of runoff �� ere much less from the rainfall evznts. A'utri- ent losses. ho�cever. were much lower for rainfall events than irrieated evenu (Tables 4, �. 6, and 7). ConcenVa- tions in leachate were similar to those found for irrigated e.ents «•ithoet fertiiization. No indication of fertiliza- tion «�as e�•ident in samplzs for any rainfall evenL Linde et al. (199�) also reported'nigher nucrient concentrations but ]ower losses in runoff from rainfall events. Thz data from the rainfalt events is ntore reprzsentative of what ��ould likelv occur on a eolf course fairwav since data from the irrieated runoff e��ents that wzre preceded b� pre-event irrioation would represent a worst case scenario for runoff and nutrient transport. Five da}s before the irri�ated event on 6 Sept. 199�, plots «e.e fertilized at a rate of 4.9-03--4.1 o m'- (N-P- K). Immedia[ely foilo�rino fzrtilization, 9 mm of water �vas applied to each plot. In addition, plots werz verticut 6 h after the normal pre-event irri�ation procedure. Despite this mana�emznt scznario. nutrient concentra- tions remainzd ]ow for the e��enL Unlike previous fertil- Table 3. �Iean nutrient concentrations in runot'f and leachate samples for evenfs ihat rainfail caused runoff. \inate-\ Phosphate Total Kjeldahl-\ D�te Runot7 Le�ehate RunotF Leachate Runoft Leachate mg L 79Y3 137uh� 7.6 03 a79 135 0.7? 0.'_3 ?1 Julr: 15 1.1 6.06 1.10 0 0 ?? Jul. 0.7 0.6 6.00 ?.91 0.�? O.L 17 Au�.� 11 1.9 ?.7S 1.00 1.7> 0.21 17Sept: 1.9 ?.i0 1.09 1 \ov.:= 1.7 3.33 Q.90 ?1 \o..i_ ?.6 2.60 0.87 ?3 Xoc_ LO 1.60 0.06 199> 6 ]ulc_ ].6 6.61 0.�6 }1 Au�.= 2.0 4.?3 0.96 _'0 Oct.- OS 1.8_' 0.06 ` aIl repiications did not produce a sample. - Leachate sampies uere mt coliected. O�_�\\� lzs� izations conducted 4 h after pre-event irri�ation and S h before an ecent, it �eas not ecident in the water samplzs that fertilization had occurred. On ;olf course faira�ays, off-site movement of nutri- ents may happzn if runoff occurs soon after application of a�ranular fzrtilizer to a nearly saturated soil. To pre�•ent this scena; io, ��arious manaQement practices can be implemznted by the mrf manaser. One practice k'ould be to a��oid applyinQ fzrtiIizer on nearlv saturated soils, especiall�� �vhen rainfall is expzcted shortly after appiication. hno2her practice �vould be to vrater-in the fertilizer �cith li�ht irri�ation shordy afcer fertilization. Othzr practices to prevent nucrient transport in runoff include fotiar appiication of soluble nutrients and the usz of fertilizzrs thai have significant slo«�-release char- acteristics. Sediment Transport Since S3% of the 237 runoff samples analyzed con- tained no measurabie sediment, statistical analyses u�erz not conducted. Also. no consistent difference was evi- dznt benceen turf species. Therefore, sediment concen- trations from both turfs were averaged for each type of runoff sample on zach date (Table 9). Concentrations were usuall}� hiehest for che first flo�c-paced samp]es and lo�vest for the second flow-paced runoff samples. \Vhen sediment was measured, the amount was very small, evzn after certical mowine down the slope and removin� an average of 67 k� of organic material from the bentora=_s plots and an avera�e of 20 kg from the :yeerass plots. The highest observed sediment concentration for all runoff samples ���as 23� mg L detected in the first flow-paced sample of a rye�rass plot that produced 1675 L of ranoff on 29 June 199�. That 123.5 m� plot was irri�ated at 139 mm h for 1� min. To calculate the potential soif ]oss for that plot, sediment concentrations of both flo�v-paced samples were averaoed and equaled Table 9. A1ean sediment concenirations for irrigated erents. ntean sediment concentntion Runoff samplei D FPl FP2 mg L 1991 ?9 June 68.7 0 13 Juk S.8 3-9 ? Aug.' 0 0 2? Au�. 0 0 3 Sept.; 9.6 0.6 ?0 Sept. ?7.9 12.5 S Oct. 0.6 i-d 1995 16 �fav 0 0 31 >Ia� 0 0 li June_ 0 0 28 June ?.0 1.S 12 Juh 0.7 0 26 luh 0 2�� 6 Sept= 1.7 0 ?3 SepL 0.3 0 i FPS = lst ilo��-paced mnott' sample, FP'_ _'_nd ilo��-paced runoll sample. � Yertical mo��ins conducted 6 h before e�ent. 12�� J EYV(ROX QUAC.., VOL?6. SEPlE�iBER-0CTOBER (9Y7 143 ma L The potzntial soil loss from thz plot was calculated to be 19.4 k� ha for thz li min event. This number rzpresents the highzst potential soil loss for [his studv. Based on the acerages of total runoff volume (1390 L). duration (20 min). and concentration in the flo�s-paced sampies (9.b mg L the averaoz pocential soil loss for afl plots in 199-� was 15 ke ha for a 20 min zcznt. In 199�, thz a��erage potentizl soil loss ���as 0.1 kg ha ' for a 20 min ecznt. lisins S°io-sloped plots of tall fescue turf maintained at 8 cm, Gross et aL (1991) rzported the avera�e poten- tial soil loss for a 30 min, 120 mm h intznsity storm �cas �19 kg ha ' for bare soil and ��' ko ha for mature tall fescue turf szeded at 4SS kg ha �.�The much lower amount oi soil loss for the current study compared to Gross et at. (1991) could hzve been duz to differences in turf density and tortuousit}� of ocerland floa�. Gross et aL (1991) reportzd a mean density of �7 tillers dm' for a mamre tall fzscue turf sezded at 4SS kg ha and maintained at 8 cm. Linde et al. (199�) reported mean densities of 2�93 tillers dm for ma[urz creepin� bent- erass and 275 tillzrs dm'' for mature perennial rye�rass maintained at 1.9 cm. In this study, sedimen[ was seldom detzcted in runoff samples produced bq rainfalL Only 6 of thz �6 total runoff samplzs from rainfaff even[s contained detectable sedimenc Of those sis samplzs, thz highest sediment concen[ra[ion was 26 mg L for a bent�rass plot that produced 170 L of runoff in 100 min. Thz soi! loss for that plot �cas 0.36 k� ha''. In another study, Gross et aI. (1990) reported the average soil loss in runoff caused by rainfall #or sodded tall fescuelKenmcky blue�rass p1oG �vith a 5 to 7%a slope ti�as 0.4� ke ha and 1.47 k� ha for consecutive }'ears. SUDT�IARY AND CONCLUSIONS On avzrage, vzry fittle sediment transport, if any, was found in runoff samplzs. Therefore, for the conditions of thz curren[ studq, 4-yr-old creepino bentorass and pzrennial ryeerass turfs wzrz very effective in rzducine sediment transpurt, e�en aftzr vertical mowing do�vn the slope ok each plo:. I[ ma}� be possiole that if vertical V�����L mo�cina �cas more agaressive, sediment transport could increase bzcause a�reatzr amount of vegztation would be remoced and morz arooves cut into the soil. Nutrient transport, particularly phosphate and TKN, sienificantly increased for runoff events that had pre- e��ent irrigation and were conducted 3 h after fertiliza- tion. For a11 other evenG, nutrient transport �vas consis- tently lo«er. As a rzsul[, off-site movzment of nutrient; from golf course fairways may happen if runoff occurs soon aE[er sranular fertilizzr is applizd to a neariy satu- rated soil. Such conditions w•ould essentiafly represent a«orst casz scenario for runoff and nutrient transport and wouid Izss likely happen in `real world' circum- stances. REFERE\CES Btard. J.B. 197i. Turf�rass scizncz and culture. Przntice-HaL(, En�lz- wood Clift>, rJ. Gro>s, C.DI., J.S. Anglz. R.L. HiII, and 1LS. R'zIItrizn. 1991. RunoF[ and szdimznt losszs from [slt fescuz undzr simula[zd rainfalL I. Environ. Qual. ?0:6d3-607. Gross. Cbt., 7.S. Ano1z, and hLS. Neltzrltn. 1990. K�utrient and szdi- mcn[ losses from mrf�rass. J. Emiron. Qual. 19:66i-66S. Harrison, S.A., T.L �Ya[schke, R.O. hlumma, A-R. Jarre[[, and G.W. Hamilton. 1993. Nutriznt and pzsticidz concentra[ions in wa[zr from chemically treated mrf�ca>s. p. 191-207. I�i K.D. Racke and A.R. Les(iz (zd.) Pesticidzs in urban environments: Fatz and si�nifi- cance. ACS $ymp. Ser. 5??. Am. Chem. Soc.. �Vashington, DC. Lindz, D.T. 1996. RunoEf, erosion, and nucriznt [ransporc from crzzp- in� bznc�rass and pzrennia! ry'e�rass turfs. Ph.D. diss. Pznnsylvania State Univ., Univzrsity Park. PA. Linde, D.T., T.L. ��"atschkz, and J-A. Bor�er.1993. Nutriznt transport in rurtoff from two turf�rass spzcizs. p. 489-496. (n AJ. Cochran and bi.F. Farrally (edJ Science and �oif IL Ptoc. of the 19%Norid $cizntific Con�rev Of 6olf. E& F\ Spon, D'+zw Yotk. Linde, D.T., T.L. Wa[schke, A.R. Jarrett, and J.A. Bor�ec 1995. Surface mnoff assessment from creepin� bzntorass and pzrennial ryzgrass mrE A�roa J. 57:176-152. �[or[on. T.G.. A.J. Gold, and W.�I. Sullivan. 1988. Intluznce of over- warering and fertiliza[ion on ni[roozn losses from homz lawns. !. Eaviron. QuaL 17:1?�i-liQ. $AS Ins[itu[z 1990. SAS;STAT uszr's �uide. �'oL 2. Version 6. A[n ed. SAS Inst., Cary, FC. Stzzl, R.G.D., and J.H. Torrie. 1930. Principles and procedures o( s[a[istics: A biomz[rical approach. 2nd ed. hleGraw-Hilt, he«' York. ��'auchopz. R.D.. F.G. Witliams, and L.R. \farti. 1990. Runoff of sulfomz[uron-mzthyl and tyanazine from small plo[s: Effzc[> of formulation and srass co�er. J. Environ. Qual. 19:119-IZ�. Reprinted Cmm IheJouma! � f L•nvnonmrn/a! Quoliry Volume 23, no. I, Jan.-Fcb. I999. Copyright O I999, ASA, CSSA, SSSA 677 South $egoe Rd„ Mad(mq NI53711 USA 0 � �,��� Relationship between Phosphorus Levels in Three Ultisols and Phosphorus Concentrations in Runoff D. H. Pote,� T. C. Daniel, D. J. Nichols, A. N. Shazpley, P. A. Moore, dr., D. M. Miller, and D. R. Edwards ABSTRACT Soi15 that contsi� high P lerel5 Can become a primary 5ource of dissolved reactive P(DRP) in runoff, and thus contribute to acceler- ated eutrophication of surface waters. In a prerious sNdy on Captina soil, sereral soil te5t P(S1'P) methods gave resulB that wece signifi- canUy mrrelaFed fo DRP lerels in rvnoEf, but disrilled H and NH�- oxalate methods gave the best coaelations. Because results might differ on other soils, mnoff studies were conducted on three additional Ultisols to identify the most cnnsistent STP method for predicting runoffDRP lereis, and determine effecfs of site hydrology on correla- tions between STP and runoff DRP rnncentrations. Sucface soil {U- 2 cm depth) of pastuce plots was analyced 6y hlehlich III, Olsen, Morgan, Bray-Kurtz Pl, NH�-oxalate, and dis611ed H methods. Also, P saNration of each soil was deterntined by three diflerent methods. Simulated rain (75 mm h") produced 30 min of runoff Gom each ploL Ali cortelations of STP to (unoff DRP were significant (P G 0.01) regardless oF soil series or STP method, with most STP methods e Fiag digh correlxtions (r > 0.90) on all three soils. For a given levei of H�O-eztractable STP, low runoff vulumes coincided with low DRP concen[rations. Tf�erefore, when each DRP concentra- tion was di�ided by��olume of piot nmutt; wrrelafions to H:O-exfract- able STP had the same (Y < 0.05) regression line (or every soil. This suegests the importance of site hydrology in determining P Ioss in r�noff, and may provide a means of developing a single reiationship for a range of seil series. E uTxoexicnno:r of streams and lakes can be greatly accelzrated by the influs of nutrients in surface runoff from agricultu-al land. Since P has been identified as the nutrient in rur.off that is usually the most limiting to algal growth, control of P Izvels in runoff is often recommended as the best way to minimize the eutrophi- cation of su:face waters (Rohlich and O'Connor, 1980; Litt1e,198S; Breeuwsma and Silva,1992; Sharpiey et ai., 1994). Phosphorus is often perceived to be so immobiie in soil that losses from agricultural land are not usualiy considzred to be agronomically_iraportant, but even small agronomic losses can have serious environmental consequences. In iact, scils that contain high levels of P from escessive fertilization can become a primary source of dissolved reactive P(DRP) in runoff (Edwards et a1., 1993). Other investigators have found direct correlations be- tween soil P levels and P concentrations in runoff. D.H. Po[e, USDA-ARS, Dale Bumpers Small Fazms Res. Centez, 6833 South Stare Hwy. 23, Booneville, AR "R927-9214; T.C. Daniel, D.J. Nichols, and D-M. Miller, Dep. of Agro�omy, S15 Plan[ $<ience, Univ. of Arkansas, Fayetteville, AR 72701; P.A. Moore, Jr., USDA- ARS, 115 Ptant Scier.ce, Fayetteville, AR 72701; A.N. Sharotey, Pas- ture Systems and Watershed Managem�nt Research Lab., USDA- ARS, Curtin Road, University Park, PA 168023702; and D.R. Ed- wards, Biosystems and A�ricultural Engineering Dep., 128 Agricul- tural Engineering Building, Univ. of Ker,mcky, Lexing�on, KY 40546. Received 29 July 1998. *Correspondin� author (dpo�e@a�.gov). Published in 7. Environ. QuaL 25370.175 (1999). Schreiber (1988) sampled soil and runoff from mono- cropped com (Zea mays L.) or cotton (Gossypium Firsu- tum L.) research plots and watersheds in Mississippi, with various cropping pracaces for corn includin� con- ventional tillage, no-till, crop residue removed for sila�e, and crop zesidue left on the soii surface. Results showed that water-extractable soil test P(STP) was significantly correlated to annual discharge-weighted DRP in runoff. Yli-Halla et al. (1995) analyzed soil and runoff from eight cultivated field plots in southwestern Finland and concluded that mean DRP concentration in runoff de- pended on the water-extractable P level in surface soil. However, both of these studies relied on uncontrolled natural rainfall events to produce runoff, and combined a variety of cultivated crops and management practices, while neither study included uncultivated grassland. In a previous study (Pote et al., 199fi), we concrolled the variabitity of field conditions as much as possible by using consistent dimensions, slope, soil, and grass cocer for all plots, and using simulated rainfaL to pro- duce runofi. The study compared results from several soil test P(STP) extraction methods to determine which were most useful for predicting DRP levels in runoff from fescue (Festuca arundinacea Schreb.) plots on a Captina silt loam (fine-silry, siliceous, mesic Typic Ftagi- udult). Extraction of P in soil samples from the surface soil (0-2 cm depth) showed that the Mehlich III (Meh- (ich, 1984}, Bray-Kurt2 Pl (Bray and Kurtz, 1945), and Olsen (Olsen et al., 1954) extraction methods gave soil P levels with very significant correlations to DRP con- centrations in surface runoff. The soil P-saturation method (Pote et al., 1996) also gave zesuiu that corre- lated very well to runoff DRP, but Fe-oxide strigs (Sharpley, 1993), distilled water, and acidified ammo- nium oxaiate (Pote et a1.,1996) were the STP extractants that gave the best correlations to DRP in runoff. Since this study was only conducted on a single soil, we hy- pothesized that the results might be different for other soils of differing physical and chemical properties, even with;n the same soil order. As severai states are attempting to define threshold STP levels above which DRP enrichment of runoff is unacceptable from a water-quality perspective, more information relatin� soil P to runoff P is needed (Shacpley et al., 1996). Such Field data are essential to development of technically-sound STP levels that can be used to guide P management recommendations. Therefore, runoff studies were conducted on three addi- tional Ultisols. The objectives were to determine (i) which STP method maintains the highest correlation to Abbreviations: CV, coefficient of variation; DRP, dissolved reactive P; ICP, induc[iveiy coupted plasma spectrometer, M3, Mehlich III extraction me[hod for soil P; PSI, P sorp[ion index; SD, siandard deviation; STP, soil tes[ P. y �, 170 O l - �� \'�-- POTE EI AL: PHOSPHORUS LEVELS LY THREE ULTISOLS Tabie 1. Soil charaderisticc (mean) and :esu{ts of various soiS test P(STP) methods from plots on three soils. Nella so� Clay covteat 105% Oiganic C content 3.S% pH 59 Oxalate-Fe, mg kg ' 1909 Oxalate•Al, mg kg ' LiO4 Ne{ls wl Ztange Mean SDi Lioker wil li9q 3.6 % 51 3003 1170 Linker soil Range hlean SD _ �p � � i in lYOark swl 7.4% 4.6 % 6.2 104i 16M13 noack :oa Range Mean SD STP method - - � biehlich III 260.42L 294 73 12I 366 226 � 17-?63 109 � Olsen 79-166 175 28 61-1b2 104 31 7-303 44 30 Morgan 23b5 44 IS 30-108 57 27 0.107 35 34 Bray-Kum PS 161-3d2 7A0 62 121-328 ?A7 76 14-156 90 47 Nil..ozalate 691-1L7 9DD L'9 315-707 442 144 210.613 4W 1'A1 DisNled H�O 37-109 74 25 1& -107 50 3U b-3U 36 24 t Standard deriation. DRP concentra[ions in runoff from a variety of soil series within the Ultisol order, (ii) whether STP levels affect DRP concentrations in runoff consistently across soil series and if not, (iii) what effect soil hydrology has on the relationship between STP and runoff DRP. MATERYALS AND bIETHODS a Field Plots Six field plots were constructed during the fall of 1993 on each of three soils in northwest Arkansas: Nella (fine-loamy, siliceous, thermic Typic Pateudult), Linker (fine-loamy, sili- ceous, thermic Typic Hapluduli), and Noark (clayey-skeletat, mixed, mesic Typic Paleudult) (Table 1). All plots were con- structzd on well-established tall fescue pastures with approxi- ma[ely 7% slope and 100% ground cover as measured by the line-transect me[hod (Laflen et al., 1981). These pastures had previously been amended with various combinations of swine manure slurry, commercial fertilizers, and/or manure from grazing cattle. Some plots had received swine manure [he previous year, but no amendments were allowed on the plots for sevetal months preceding this s[udy. Vegeta[ion height was maintained between 0.1 and 02 m throughout the study by mowing. Each plo[ (1.5 x 3 m) was fitted with aluminum borders (extending 5 cm above and 10 cm below [he surface) for runoff isolation, a downslope trough for runoff collection, and a runoff sampling pit, as described by Edwards and Daniel (1993). Fences were constructed around the plocs to prevent catile from contributing P inp�ts or causing other damage during the study. In May 1995, a simula[or described by Edwards et al. (1992) was used to reduce antecedent moismre variability by applying rainfall (75 mm h'') to each ptot until the su[face layer was saturated. This simulator delivers rainfatl at an eait pressure of 41.4 kPa from four VeeJet nozzles` elevated 3.05 m above the soii surface by an aluminum scaffold to obiain drop-size distribution and terminal velocity compazab(e to that of natu- ral rainfall. Tarpaulins attached to the aluminum scaffold sur- round the plot to form wind screens. An elec[rie motor drives the shaft to which the nozzles are attached, causing them to oscillate across openings in the simulator body, with the rain- fall intensity dependent upon the frequency of oscillation. ' Names are necessary to repor[ factualty on available data; how- ever, [he USDA nei[her guarantees nor warrants [he s[andard of the produc[, and [he use oE the name by USDA implies no appmval of, : the product to [he zx<lusion of othecs Ihat may also be suitable. � Following the initiai rainfall application, all plots were allowed to drain for 48 h before simulated rain was applied again a[ an intensity of 75 mm h'' to generate 30 min of runoff from each plot. Sampling Methods Runoff was sampled manually at 5-min intervals throughout the runoff event, beginning 2.5 min after initiation of continu- ous-flow runoff. For each discrete runoff sample, the volume and time required to colleM it were recorded and used to calcula[e mean flow rate and total volume of runoEf for the 5-min interval. Using these runoff data, the six discrete runoff samples from each plot were used to construct a flow-weighted composite sample to represent the total runoff from that plot. An aliquot of each composite runoff sample was filtered (0.45- µm pore diame[er) within 2 h of collection and stored in the dark at 4 until analyzed for DRP by the molybdeoum-blue method (Mucphy and Riley,1962). Total DRP mass loss from each plot was calculated as the plot's total runoff voiume multiplied by DEtP concentration in the flow-weighted com- posite runoff sample from thaf plot. . Just prior to applying simulated rainfall to a given pbt, a representative composite soil sample was collected by combin- ing 10 discrete soil cores (2.54 cm diam.) taken randomly from the surface layer (0-2 cm depth) of the plot. All composite soil samples were stored in the dark at 4 until air dried and sieved (2 mm) to remove larger rock particles and most of the plant material. Two complete runofF events were conducted on each plot, separated by a 2-d interval. For each separate runoff event, soil samples were collected jus[ prior to simulated rainfall ap- pHcation. Soil Analyses Each soil sample was analyzed for extractable P by six methods: Morgan (Morgan, 194I), Mehlich III (Mehlich, 1984), Bray-Kurtz Pl (Bray and Kurtz, 1945), Olsen (Olsen et al., 1954), distilled water, and acidified ammoaium oxaLate. The Morgan, Mehlich III, Bray-Kurtz Pl, and Olsen chemical extractan[s were selected because they aze commonty used for STP analysis in soil testing laboratories. These methods were not originally devetoped to predict runoff water quality, but rathez to assess the fertility status of soil for aop produc- tion. Distilled water most closety simulates actual runoff solu- tion, and may thus be the most appropriate for predicting runoff DRP. One „cram of soil was mized with 25 mL of o � - ����. I�Z J. ENVIRON. QUAL., VOL. 28, JANUARY-FEBRUARY 1999 distilled water, shaken end-over-end for 1 h, centrifuged for 5 min a[ 266 m s�(27100 g), Eiltered (0-45 µm), and the supzmatant analyzed for P by the molybdenum-blue method (Murphy and Riley, 1962)_ Acidified ammonium oxalate has been used in severaf pre�,ious studies (van der Zee et al., 1987; van der Zee and van Riemsdijk, 1988; A�folina et al., 1991; Breeuwsma and Silva, 1992; Freese et al., 1992), theoretically [o release into solution potentially dzsorbable P, as it dissolves the compounds (noncrys[alline oxides oi iron and aluminum) controlting P sorption in acid soils (Table 1). In our smdy, ammonium oxalate extractant was made by mixing 02 M oxalic acid with 0.2 M ammonium oxalatz (approximately 535 mL of oxalic acid with 700 mL of ammonium oxalate) until the combined-solution pH was 3.0. A 20-mL aliquot of the ammonium oxalate solution was then mixed with 0.5 g of soil, shaken in the dark for 2 h, cenirifuged for 20 min at 131 m s '(14481g),anddecantedforPanalysis.Oxalate-extractable P, A., and Fe were also used to calculate the P sorption- sa[uration of each soil as described below. Mehlich III, Bray- Kurtz Pl, and acidified ammonium oxalate extracts were ana- lyzed for P by inductively coupled plasma spectrometer (ICP), while n4organ, Olsen, and d'utitied water extracts were ana- ]yzed colorimetrically by the molybdenum-blue me[hod (Mur- phy and Riley, 1962). A single-point P sorption index (PSI) described by Mozaf- fari and Sims (1994) was also determined on each soil. A P sorption solution (containing 300 mg P per liter) was made by dissolving 1.315 g of KH in enough distilied, deionized H to make 1 L of solucion. The PSI was determined by weighing 1.00 g of soil in[o a�0-mL centrifuge [ube, adding 20 mL of 0.0125 M_ CaCi 2H>O, and adding � mL of P sorption solution to make a combined solution containing 0.01 M CaCi and 60 mg P per liter. After two drops of toluene were added and the tubes sealed, the mixture was shaken for 18 h on a reciprocating shaker, centrifuged for 10 min at 266 m s' (27100 g), filtered (0.45-µm), and analyzed for P by induc- tively coupied plasma spectrometer (ICP). The PSI was calcu- lated as X(lo� P where X is P sorbed (mg kg'`) _[(P�)(V) —(P (kg of soil) P is initial P concentration in sorption solution (mg L V is volume of P sorption solution (L) P is final P concentration in solu[ion (m� L Phosphorus Saturation of Soil The P saturation (%) of each soit sample was calculated by two different methods; (i) oxalate-extractable P(mmol kg") divided by thz oxalate-extractable AI and Fe (mmot kg con[ent, and multiplied by 100, and (ii) initial STP con- ten[ (mg kg ') divided by P (mg kg and multiplied by 100. For this second method, the PSI value was used to approximate the maximum amount of P(P.,,, that could be adsorbed by the soil. Mozaffari and Sims (1994) found that P� can be estima[ed by the equation P�� =(PSI + 51.9)/ OS, �iven that P Mqr < 1400 mg kg STP extracta�ts selected to obtain the initial STP conten[ were Mehlich III (M3-PSI method) and distilled H2O (H2O-PSI method). Statistical Methods For each soil, comparisons were made between STP meth- ods by correlating STP results to DRP concentrations in runoff from the ptots, developing a linear regression from [he 12 data points, and calcula[ing [he sample corrzla[ion coefficient (r value) for each. For each soil test method, analysis of covari- ance was used to determine whe[her there were statistica! differences be[ween regression slopes and in[ercepts of ihe three soils. RESULTS AND DISCUSSION Soil Phosphorus For each of these soils, the range, mean, and standard deviation of STP contents are shown in Table 1. Distilled water, Mor�an, and Olsen methods extracted the least amounts of P from soil, while Mehlich III and Bray- Kurtz Pl methods extracted larger amounts. NH,-oxa- late extracted much larger amounts of soil P than did other extractants, suggesting that most of the P in these soils is sorbed or precipitated on amorphous oxides of Fe and Al. Relationship between STP and Runoff DRP For each soil, correlations of STP to runoff DRP were not significantly affected by the time interval (2 d) between the two runoff events. Therefore, the data from both runoff events were combined to �ive a total of 12 data points for each soil. The correlation coefficient (r) and linear regression equation are given in Table 2 for each STP correlation to DRP in runoff. For all soiLs, the STP values obtained by each method were significantly corretated (P < 0.01) to DRP concentrations in plot runoff. Yet, when the extraction methods were com- pared using r values to see how closely the data peints fit the regression line, it was apparent that some STP methods were more closely related to DRP concentra- tions in runoff than other methods (Table 2). For exam- ple, the NH and Olsen methods each gave a weaker correlation r< 0.90) to DRP concentrations in runoff from at least one soil, while all other STP meth- ods gave correlations with r> 0.90 for all three soils (Table 2). However, if previous studies (Pote et al., 1996) are considerzd, the H2O-estractable soil P has shown the most consistently high correlation to DRP concentrations in runoff, even when rainfall intensity, slope, and seasonal conditions varied. Although the usefulness of an STP method for pre- dicting runoff DRP concentrations depends largely on its ability to produce data poinu that closely fit a regres- sion line on any given soil, it would also be very helpful to have an STP method that produces approximately the same regression for all soils (or at least a large group of soils). Such a method would eliminate the need to use soil series as the basis for maximum soii P recom- mzndations, thus saving the time and expense of accu- rately identifying the soil series of each individual site. If the data points from all three soils are combined into a single data set, the P-saturation (oxalate method) might seem to be a good choice for this purpose because it gives a good linear correlation (r = 0.887), and the fit is even better for a second-order regression (r = 0.931 for the curve where y= 03053 — 0.0353x + 0.0014x However, if the data points are separated into regression lines for each soil, differences between some slopes be- come apparent (Fig. l). When the regression-line �raphs of each method were compared visually, the P-satura- POTE EI' AL PHOSPHORUS LEVEIS IN THREE ULTISOLS 0.905 0$69 0.907 ' . 0.913 0.806 0.9'� 0903 0.916 0932 Table 2. Results of soil test P(STP) methods correlated to dissolved reactive P(DRP) in runoH from three Ultisola CortelaRon coefident (r) t S1 P(m kg"') co rrelated t DRP (mg L STP method Nella w� Linker so0 Noark wii Mehtich III Olsea Mocgan Brav-Kurtz PS NH,-Ozalate Disb7led H.O P saNlation (oxalate method) P satuntion @13-PSI metbod) P satucdtion (H_O-PSI metLod) $TP method Mehlich III Oisen Morgan Bnv-Kurtz Pl NI-I�-Ozalate Distilled H P saturation (o:Wate met6od) P saturation (M}pSI method) P saturation (H.O-PSI metho� 0.916 0.864 0.941 0.950 0.914 0.928 0.928 0.928 0921 �\�\\\b- 173 0.932 0935 0.932 0943 0908 0.965 0.933 0.937 0.978 Regression line equa for ST P (mg kg �) <onelated to DRP (mg L NeO soii Linker sol Noark soil y= 0.0036z - OAS y= 0.01135c - 038 y = 0.0016x + 0.00 y= O.00SSz - 0.43 y= 0.0093z - 056 y= O.00d3x - 0.02 y= O.OlSlz - 0.18 y= O.O1LSZ - 025 y= 0.0038x -F 0.04 Y=O.00d3x-0.42 y=0.0041ac-0.4b y=0.0027x-0.02 y= 0.001Sx - 1.03 y= 0.002Lc - 0.63 y= 0.0009x - 019 y = 0.0107x - 0.18 y = 0.0104x - 0.11 y= O.00SSx - 0.03 y= 0.0820x - 203 y = 0.0397x - 0.62 y = 0.0251c - 01A y= O.00S� - 0.08 y= 0.0065x - 0.04 y= 0.0045x + 0.03 y = 0.0?62x + 0.03 y = 0.0215z + 0.06 y= 0.0759x + 0.01 i All correla6on coeffidents were signifitant (a = 0.01). tion (PSI methods), Mehlich III, Bray-Kurtz Pl (Fig. 2), and distilled H (Fig. 3) methods each appeared to have regression lines that were relatively close together with similar slopes for all soils, but statistical analysis showed that none of the methods for correlating STP to DRP in runoff gave the same (P � 0.05) regression line for all three soils. This result was not surprising, given the differences in chemieal and physical properties between soils. The P saturation status of each soil in this study was significantly (P < 0.01) related to DRP concentrations in runoff, regardless of the method used to calcutate P saturation. All three methods gave high correlations to DRP in runoff but none gave the same regression line on all three soils, so their vatue as universal predictors of DRP concentrations in runoff is questiouable. i � E 0 � c a � 0 1.6 �,a 7.2 7.0 0.8 0.6 0.4 0.2 • Nella (r = 0.903) � Linker (r = 0.928) ONoark (r = 0.933) 0.0 ' " ' 0 10 20 30 � 40 � • • • � � • O • M .� Effects of Runoff Volume Because site hydrology of each soil is likely to impact the relationship between soil P and runoff P(Gburek and Sharpley, 1998), the effect of runoff votume on P transport from our plots was evaluated. The average rainfall application required to produce 30 min of con- tinuous runoff is included in Table 3, along with mean runoff volume for each soil. Runoff from the Nella and Linker soils averaged about the same volume and the variability was also similar, while the Noark soil had the lowest amount of runoff and the least variability. The differences in runoff among soils are reflected in the correlations of water-extractable STP to runoff DRP. For esample, when water-extractable STP was correlated to mass losses (loads) of DRP in runoff (Fig. 1.6 7.4 7.2 �� rn 1.0 E 0.8 ' 0.6 c a 0.4 ¢ 0 0.2 0.0 0 700 200 300 400 Bra Kurtz extractabie soil P m k - Soil P Saturation (%) Y � 9 9) Fig. 1. Relationship between P samration (oxalate method) of suhace Fig. 2. Relationship behveen Bray-Kurtz Pl extractabfe P in sudace soil and dissoived reactive P(DRP) in runoff from three soils. soil and dissolved reac[ive P(DRP) im m�oll. . 174 7.6 7.4 1.2 i � 7.0 0 0.8 c i 0.6 � 0.4 0 0.2 a.o 0 20 40 60 J. ENVIRON. QUAL., VOL 28, JANUARY-FEBRUARY 1999 80 700 120 Water eztractable soil P(mg k9 Fg.3. Relationship between water-eztraMabie P in surface soil and dissolved reactive P(DRP) in runoS 4), the Noark correlation was best (r value = 0.963) because mass losses depend on boSh.the P concentration and the volume of runoff (which" was highly consistent for the Noark soil). Runoff volumes were more variable for the other two soils, and therefore mass losses of runoff DRP show a poorer correlation to STP (Fig. 4). The variability of runoff volume is also reflected in the r values for the correlation of water extractable STP to DRP concentrations in runoff (Fig. 3). For example, Nella soil had the most variable runoff volume, and it also had ihe lowest r value, while Noark soil had the least variable runoff volume and the highest r value. Finally, for a given level of water-extractable STP, soils with the lowest mean runoff volume also had the ]owest concentration of DRP in runoff (Fig. 3). For example, Noark soil produced the least amount of run- off, but for any given level of water-extractable STP, it also had the lowest concentration of DRP in the runoff. No previous studies have investigated the relationship between runoff volume and DRP concentration in the runoff; and our observations at first seemed rather counter-intuitive because we expected higher volumes of runoff to generally produce lower DRP concentra- tions due to greater dilution. This unexpected trend may result from the rapid movement of DRP into the soil profile of soIls with low runoff volumes (high infiltration rates), thus taking it away from the primary zone of transfer to surface runoff. In soils with lower infiltration Table 3. Rainfal{ aud runo8 data from simufated rain appliration to 6eld plots on three soiLs. Ne17a soii Linker soii Noark sol Rainfall meant, mm 4$.9 475� 53.6 Ruuo@'mean,mm 21.7 7A.6 133 Runoff CV, % 30.0 29.1 175 i Amount required to produce 30 min of rumR. Eath mean tepresenls 12 runoH events. 400 �0 300 ._°: 0 � 200 i c v a 100 a 0 0 b a\ -\\\�� •Nella (r = 0.724) �Linker (r = 0.847� ONoark (r = 0.963) •• • � � • • � � � � � �• � • . o 20 40 60 80 100 120 Water extracfoble soil P(mg kg Fg.0. Relafionsltip between water-extractable P in surCace soil and dissolved reactive P(DRP) load in runoff. rates, much more of the dissolved P may remain near the soil surface long enough to be lost in runoff water. In an attempt to define these processes, we normal- ized DRP concentration for each plot. When the DRP concentration in runoff from each plot was divided by the depth of runoff from that plot, and related to the water extractable STP level, regression lines for all soils were statistically the same line (P < 0.05) (Fig. 5). Thus, by combining water-extractable STP data with hydro- logic data, it may be possible to make reasonably accu- rate predictions of DRP levels in runoff from a range of soils. Acquiring the necessary hydrologic da[a on runoff volumes from a soil may sometimes be just as difficult as accurately identifying the soil series of each specific site, but this at least provides an a!temate 7.2 E � 1.0 0 � 0.8 � ^ o.s � � E o.4 `o � 0.2 � c a 0.0 � 0 0 . . . • • • • . � b .� . 20 40 60 SO 100 120 Water eztractable soil P(mg kg Fg. 5. Relafio�ship between watervextzactable P in sudace soil and the ratio of dissolved reactire P(DRP) im m�otf to the total amount of runoR. •Nella (r = 0.866) - �Linker (r = 0.847) oNoark (r = 0.919) POTE EC AL: PHOSPHORUS LEVELS I1V THREE ULTISOLS method for predicting DRP concentrations in runoff. For water-quality modelers, it also supplies important information conceming the relationship between vol- ume of runoff and DRP concentration in runoff. Most importantly, i[ shows the strong influence of site hydrol- ogy on processes controlling P loss in surface runoff. CONCLUSIONS The results of this study reinforce previous evidence of a linear relationship between P levels in surface soil (0-Z cm deep) and DRP concentra[ions in runoff from the soil surface, but this study also extends our knowl- edge by showing that such a relationship exists on a variety of Ultisols. On each soil that was tested, a signifi- cant (P < 0.01) linear relationship was apparent, regard- less of the method used to determine STP. Because most STP extractants gave results that were highly cor- related (r > 0.90) to DRP in runoff from all three soils, this study did not clearly identify any particular STP method for maintaining the highest correlation to DRP concentrations in runoff from all soils tested. However, the study did show that several STP extractants may be useful for predicting DRP concentrations in runoff, including extractants such as distilled water that were supported by the results of previous work (Pote et al., 1996) conducted under different rainfall intensity, slope, and seasonal conditions. This study showed that effects of STP levels on DRP concentrations in runoff are not always consistent across soil series, and much of the difference can be attributed to soil hydrology. The fact that total plot runoff was much more variable on some soil series than on others was apparently reflected in correlations be[ween STP and runoff DRP, as soils with the most consistent vol- ume of plot runoff had the best correlations of water- extractable STP to both concentrations and mass losses of DRP in runoff. Also, for any given level of water- extractable STP, soils that produced the lowest volumes of runoff also had the lowest concentrations of DRP in the runoff. When this information was used to normalize the data for DRP concentrations in runoff (divide each DRP concentration by the volume of runoff from that plot), the resulting correlations to water-extractable STP had statisTically the same (P < 0.05) re�ression line for every soil. This implies that knowledge of site hydrology can improve the usefulness of STP data for predicting DRP concentrations in runoff. REFERENCES Bray, R.H., and L.T. Kurtz. 1945. Determination of [otal, organic, and available foans of phosphorus in soils. Soil Sci. 59:39-45. Breeuwsma, A., and S. Silva. 1992. Phosphorus fertilisa[ion and en�i- ronmencal effec[s in The Netherlands and [he Po region (I[aly). Rep. 57. Agric. Res. Dep. The Winand S[aring Cen[re for In[e- (j \-\\\�--- 175 grated Iand, Soil and �'da[er Research, �'lageningen, the Ne[h- edands. Edwards, D.R., and T.C. Danie1.1993. Effects of poultry lit[er applica- tion rate and rainfall intensity on quality of runoff from fescuegrass plots. J. Enviton. QuaL 22361-365. Edwards, D.R, T.G Daniel, J.F. Murdoch, and P.F. Vendrell. 1993. The Moore's Creek BMP effectiveness monitoring pcoject. Paper 93208�- ASAE, St. Joseph, MI. Edwards, D.R., LD. Nocton,T.C. Daniel, J_T. Walker, D.L. Ferguson, and G.A. Dwyer. i992. Perfonnance of a rainfall simularor. Arkan- sas Farm Res. 41:1'ri4. Freese, D., S.E.A.T_M. van der Zee, and W.H. van Riemsdijk. 199"t Compadsons of different modzls for phosphate sorption as a func- [ion of the iron and aluminium orzides of soils. J. Soii Sci.43:'729-738. Gburek, W.J., and A.N. Sharpley.1998- Hydrologic con[rols on phos- phorus loss from upland agricul[ural wa[enheds. 7. Emiron. Qual. 27.267-277. I.aflen. J., M. Amemiya, and E.A. Fiiniz.19S1. Measuring crop residue cover. J. Soil Water Conserv. 6:341-343. Little, C.E. 1988. Rurai dean water. The Okeechobee story- 7. $oil Water Conserv. 43:38C�390. Mehlich, A. 1984. Mehiich 3 soil [es[ extractanC A modifica[ion of Mehlich 2 extrac[ant Commun. Soil Sci. Plant AnaL 15:1409-1416. Molina, E., E. Bomemisza, F. Sancho, and D.L. Kass. 1991. $oil aluminum and iron fractions and their relacionships wiih P immobi- liza[ion and other soil properties in andisols of Costa Rica and Panama. Commun. Soil Sci. Plant AnaL 221459-1476. Morgan, M.F. 1941. Chemical soil diagnosis by [he universal soil testing system. Conn. Agric. Exp. Stn. (New Haven, CT) Bul(. 450. Mozaffari, M., and I.T. Sims. 1994. Phosphorus availability and sorp- [ion in an A[lan[ic coas[al plain wa[ershed domina[ed by animal- based agriculture. Soil Sci. 157(2):97-107. Muiphy, J., and J.R. Riley. 1962. A modified single solu[ion method for [he de[ermination of phosphate in na[ural waters. Anal. Chem. 2731-36. Olsen, S.R., C.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estima- tion of available phosphorus in wils by extraction with sodium bicarbona[e. USDA Circ. 939. U.S. Go�. PrinC Office, Washing- ton, DC. Pote, D.H., T.C. Daniel, A.N. Sharpley, P.A. Moore, Jr., D.R. Ed- wards, and D.J. Nichols.1996. Relating ex[ractable soil phosphorus to phaspho�us losses in runoff. Soil Sci. Soc. Am. J. 60:&55�59. Rohlich, G.A., and D.I. O'Connor. 1980. Phosphorus managemen[ for [he Grea[ Lakes. Fival Rep., Phosphorus Management S[ra[e- gies Task Force, Int. Ioint Commission (IJC). Poilution from Land Use Ac[ivities Reference Group Tech. Rep. Phosphorus Manage. Strategies Task Focce, Windsor, ON. Schreiber, 7.D. 1988. Estimating soluble phesphorus (POrP) �n ag- ricultural runoff. J. Miss. Acad. Sci. 33:1-15. Sharpley, A.N.1993. An innovative approach [o estimate bioavailable phosphorus in agriculiural runoH using iron oxide-impregna[ed paper. 7. Environ. Qual. 22:597fi01. Sharpley, A.N., S.C. Chapra, R. Wedepohl, I.T. Sims, T.C. Daniel, and K.R. Reddy. 1994. Managing agricultural phosphorus for pro- tection oE wrface waters: Issues and op[ions. J. Envimn. Qual. 23: 437�51. Sharpiey, A.N., T.C. Daniel, J.T. Sims, and D.H. Pote.1996. Decertnin- ing environmentally sound soil phosphorus levels. J. Soil Water Conserv. 51(2):160-166. van der Zee, S.E.A.T.bt.. L.G.J. Fokkink, and W.H. van Riemsdijk. 1987. A new tzchnique for assessmen[ of reversibiy adsorbed phos- pha[e. Soii Sci- Soc. Am. 7. 51599�04. van der Zee, S.E.A.T.M., and W.H. van Riemsdijk. 1988. Model for lon�-term phospha[e reac[ion kinetics in soil. 7. Environ. Qual. 1735�1. Yli-Halla, M., H. Hartikainen, P. Ekholm, E. Tur[ola, M. Puustinen, and K. Kallio. 1995. Assessmen[ of soluble phosphorus load in surface runoff by soil analyses. Agda Ecosyst. Environ. 56:53-b2.