ASSESSING CHEMICAL HAZARDSON GOLF COURSESby C. J. BORGERT; S. M. ROBERTS; R. D. HARBISON; J. L. CISAR,z G. H. SNYDER3UNIVERSITY OF FLORIDAlCenter for Environmental & Human Toxicology, Alachua, FL2Ft. Lauderdale Research & Education Center, Ft. Lauderdale, FL3Everglades Research & Education Center, Belle Glade, FL

FEAR of chemicals in the environmentranks high on the list of anxieties formany Americans. Because this con-

cern extends to nearly every industry andactivity in our society today, it should comeas no surprise that fears have arisen in regardto chemicals used on golf courses, such aspesticides.

Before we can decide how concerned weshould be about chemicals in our environ-ment, and on golf courses in particular, thereare some basic, common-sense rules thatshould be kept in mind. First, any chemicalcan produce toxicity in living organisms.Second, for a toxic effect to be produced, anorganism must actually be exposed to thechemical and the exposure must result in adose sufficient to produce toxicity. In otherwords, the dose makes the poison. Third,chemicals have specific and consistenteffects.

These simple facts are often misunder-stood in our society. Many assume, incor-rectly, that the mere presence of a chemicalconstitutes a health threat. Yet we all knowthis is not true. If the mere presence of achemical in our environment could producehealth effects, then aspirin could relieveheadaches without being swallowed. If thedose did not determine the poison, one glassof wine would be as inebriating as an entirebottle, and it would make no difference ifthe wine was swallowed or poured on one'sfeet! If chemicals did not have specific andpredictable effects, their use as medicineswould be impossible. Too often we fail toapply the common sense of dose andresponse to chemicals with which we areunfamiliar.

Risk assessment. is the application ofthese and other principles of toxicology tohelp us rationally decide our level of concernabout chemicals encountered in the environ-ment. Risk assessments are methods forcomparing levels of chemicals in the en-vironment with doses that produced no ad-

Leather patches, backed with aluminumfoil, were stapled to a shoe sale prior to walking on apesticide-treated turf surface.

MARCH/APRIL 1994 11

Table 1Exposure Point Quantities of Pesticides

*Chlorpyrifos was not detected on cloth. However, because chlorpyrifos wasknown to be present on the turf, we used a value of one-half the detectionlimit of the assay for chlorpyrifos on cloth, which was 0.000025 mg. Thisapproach is consistent with USEPA guidance for "non-detects."

Quantity of Pesticide on Grip (QP g)Diazinon 0.0000975 mgIsazofos 0.0000552 mgChlorpyrifos 0.0000495 mg

Quantity of Pesticide on Pant Knee (QPk)

Diazinon 0.00994 mgIsazofos 0.00198 mgChlorpyrifos* 0.000025 mg*

attached to one knee while kneeling for 10seconds to simulate aligning a putt, b) a 10cm square piece of leather attached to a shoesole following 10 steps on the treated turfsurface, and c) a golf ball putted 36 timesover a distance of 4 meters per putt. Fromthe amount of pesticide retained on leathershoe bottoms, we estimated the amount thatmight be retained on leather golf club gripslaid on the putting green. These data arepresented in Table 1.

Using the levels of pesticides listed inTable 1 for each exposure point, we calcu-lated the dose of each pesticide that a golfermight receive from each of the exposurepathways and summed the doses from allpathways to arrive at the golfer's total dose.The equations used to calculate dermal andoral doses are listed in Table 2. The golfer's

Quantity of Pesticide on Shoe Sole (QPs)Diazinon 0.000650 mgIsazofos 0.000368 mgChlorpyrifos 0.000330 mg

Legend for Table 1(QPk): Quantity of pesticide adsorbed to 10cm square of cotton cloth afterkneeling. We assumed two kneeling contacts with turf per hole.(QPs): Quantity of pesticide retained on leather soles of size 10 shoes after10 steps on turf surface. This exposure assumed that the golfer's handscontacted the entire sole of each shoe during cleaning after each round.(QP g): Quantity of pesticide retained on two leather club grips was estimatedby assuming retention rates equal to the leather shoe sole, andthat 15 square centimeters of each club grip contacts the turf.(QPb): Quantity of pesticide retained on ball following 36 putts of 4 meterseach. We assumed that the golfer licked and swallowed all of the pesticideon the surface of the ball.

Quantity of Pesticide on Golf Ball (QPb)

Diazinon 0.000775 mgIsazofos 0.000241 mgChlorpyrifos 0.000240 mg

For the sake of presenting a reasonablemaximum exposure, we took measurements24 hours following application of diazinon,chlorpyrifos (Dursban 2E), and isazofos(Triumph 4E) at rates of 470, 57, and 229 mgactive ingredient per square meter of turf.The measurements were taken from a pre-liminary study conducted on a Tifgreen ber-mudagrass surface at the Ft. LauderdaleResearch and Education Center, Universityof Florida, and the pesticide analysis wasconducted at the Everglades Research andEducation Center, Belle Glade, University ofFlorida. This research on pesticide dislodge-ability was sponsored by the Florida Turf-grass Association and by the USGA GreenSection.

We measured the amounts of pesticidesretained on a) a 10cm square piece of cotton

verse effects in laboratory animal studies orenvironmental toxicity tests. These method-ologies can be applied to turfgrass systemsto help ensure that chemicals are used inamounts and frequencies that do not poseunacceptable health or environmentalhazards.

An example of the methodology used toassess the concern about chemicals used ongolf courses is illustrated here by consideringthree pesticides applied to a putting green.Our approach is consistent in principle withthe Baseline Human Health Risk Assessmentused by the U.S. Environmental ProtectionAgency (U.S.E.~A.) to evaluate hazardouswaste sites and chemical contamination ofsoil, but is modified and refmed specificallyfor a putting green. Our assessment is pre-liminary in nature and should in no way beconsidered thorough or complete. It is in-tended only to illustrate some principles ofhealth risk assessment as applied to golfcourses and is not a defmitive evaluation.Our goal was to suggest how risk assessmentmethods can be used to address concernsregarding chemicals used on golf courses.We utilized preliminary data to conduct alimited health risk assessment for a puttinggreen and illustrated how such informationcan assist decision-making regarding levelsof chemicals that may warrant concern.

The first step in health risk assessment isto evaluate the ways a person might comeinto contact with chemicals at a particularsite, in this instance a putting green, and totake measurements of chemicals at thosepoints of potential exposure. We evaluatedthe potential exposure to three pesticides ona putting green for a golfer who plays 18holes of golf We considered four pathwaysof exposure for this golfer. We assumed thatthis theoretical golfer would 1)kneel on thegreen to align putts, 2) handle golf club gripsthat have been laid on the green, and 3) con-tact the soles of golf shoes while cleaningthem after the round. These are dermal ex-posure pathways, i.e., those that involveabsorption of chemicals through the skin.

Because skin is an effective protectivebarrier against entry of most chemicals intothe body, very little chemical that contacts theskin is actually absorbed in most instances.The dermal permeability factor for a chemi-cal reflects the fraction of chemical appliedto skin that might actually be absorbed.Although oral exposure pathways wouldprobably be less significant for most golfersthan dermal pathways, as an extreme case weassumed that the golfer would 4) clean hisgolf ball by licking it. This is an ingestionexposure pathway. We assumed that all ofthe pesticide ingested is actually absorbedinto the body because the intestinal tractgenerally is not an effective barrier to theabsorption of organic chemicals.

12 USGA GREEN SECTION RECORD

total doses were then compared with dosesconsidered by the USEPA to be safe for aperson to receive every day for a lifetime (adosage called the "Chronic Reference Dose,"or RID). Chronic Reference Doses take intoaccount that toxicity can accumulate forsome chemicals in some organ systemswhen the chemical is received as frequentlyas every day. Although we calculated singledoses from one round of golf, we comparedthese doses with chronic RIDs, which aresafe doses that can be received daily for alifetime. The comparison was made by cal-

culating a "Hazard Quotient," which is theperson's total dose divided by the RID (seeTable 3). Doses below the RID yield HazardQuotients less than 1, and those greater thanthe RID yield Hazard Quotients greater than1. If the calculated dose is equal to the"safe" dose (RID), then the Hazard Quotientequals 1.

In order to consider the entire puttinggreen as a unit, we summed the HazardQuotients for all three pesticides to arrive ata "Hazard Index" for the putting green (seeTable 3). This takes into account any

potential for additive toxicity from two ormore chemicals. A Hazard Index less than 1would indicate that the person's dose ofeach pesticide is below its respective "safedose" or RID, and that the additive potentialdoes not exceed a "total safe dose." TheUSEPA considers a Hazard Index less than1 to indicate that there is no increased healthrisk. In other words, a Hazard Index less than1 indicates that all contaminants are presentat concentrations below those that couldcause effects in humans, even if the chemi-cals have additive effects.

Table 2Dermal and oral doses of three pesticides expected from

exposure to putting greens during a round of golf

0.0000172 mg/kg 0.0000125 mg/kg 0.0000298 mg/kg0.0000010 mg/kg 0.0000039 mg/kg 0.0000049 mg/kg0.0000002 mg/kg 0.0000039 mg/kg 0.0000041 mg/kg

mg/kg = milligrams pesticide per kilogram body weight

lDermal Dose == _(Q_P_k_+_Q_P_s_+_Q_P....:::g~)_X_D_PBW (kg)

QPb (mg) X ABS20ral Dose == -------

BW (kg)

Preliminary ConclusionUnder the assumptions of this risk assess-

ment, the exposures evaluated could betripled without exceeding levels consideredsafe for daily lifetime exposure. Because wecompared the doses our theoretical golfermight receive from one round of golf withchronic RIDs, this golfer could receive thesedoses every day of his life without concernfor cumulative toxicity. We caution that this"conclusion" is made as an example onlyand cannot be applied generally becauseconditions and pesticide use can vary widelyfrom site to,site.

Total DoseOral Dose2Dermal DoselPesticideDiazinonIsazofosChlorpyrifos

where:

BW == Body Weight: 62 kg (age-adjusted male body weight)

ABS Oral Absorption Constant: 100% (assumption for organic chemicals)

DP Dermal Permeability: DP Diazinon == 0.10DP Isazofos == 0.025DP Chlorpyrifos == 0.025

Table 3Calculation of Hazard Quotients and "Hazard Index" -

a comparison of the estimated dose with the Reference Dose (RfD),a dose considered safe for lifetime exposure by USEPA

Total Dose USEPAaandPesticide (oral + dermal) OPPbRfDs Hazard Quotientsl

Diazinon 0.0000298 mg/kg 0.0009 mg/kg/d 0.0331Isazofos 0.0000049 mg/kg 0.00002 mg/kg/d 0.2450Chlorpyrifos 0.0000041 mg/kg 0.003 mg/kg/d 0.00137

Hazard Index2 = 0.2795

a United States Environmental Protection Agencyb Office of Pesticide Programs

mg/kg = milligrams chemical per kilogram body weightmg/kg/d = milligrams chemical per kilogram body weight per day

1H d Q . fi P .. d Total Dose of Pesticideazar uotIent or estlcl e == ----------Reference Dose of Pesticide

2Hazard Index for Putting Green == Sum of Hazard Quotients for Pesticides

Interpretation of Resultsand Uncertainty Analysis

The focused risk assessment presentedhere would indicate that under these theo-retical conditions and assumptions, a golfer'sexposures to chlorpyrifos, diazinon, andisazofos on putting greens would be con-sidered acceptable because the Hazard Indexis much less than 1. But how would weinterpret a Hazard Index greater than I?Although a Hazard Index of 1 or less isconsidered safe, it is not accurate to say thata Hazard Index greater than 1 is thereforeunsafe. Because of the large safety factorsoften employed in developing ReferenceDoses (10 to 10,000), doses many timesgreater than the Reference Dose couldpotentially be all right without adverseeffects. A Hazard Index greater than 1indicates that we are less certain of thepotential for adverse health effects fromcontact with the site, but it does notnecessarily indicate that the site is a threatto health. Hazard Quotients and HazardIndices are interpreted similarly, as sum-marized in Table 4.

There are numerous sources of uncertaintyinherent in the risk assessment process. Theextrapolation of toxicity data from laboratoryanimal studies to human exposure scenariosis an inexact science that introduces muchuncertainty into the process. Yet, it is uponthese extrapolations that we must often relyto determine doses that are safe from toxicity,such as Reference Doses. Similarly, studies

MARCH/APRIL 1994 13

Cotton patches, backed with aluminum foil, were pinned over one knee for tests simulating kneelingwhile aligning a putt.

Table 4Interpretation of Hazard Quotient and Hazard Index

Hazard Index (Sum of Hazard Quotients)Hazard Index < 1Hazard Index = 1Hazard Index> 1

References:

accurately reflect the toxic potential ofchemicals.

Risk assessments are only as applicableand reliable as the data upon which they arebased. Without adequate data, our ability toidentify true health and environmentalhazards is reduced and anxiety over chemi-cals increases. In the absence of data that isspecific and complete, risk assessors mustresort to conservative assumptions to en-sure that risk assessments overestimate ratherthan underestimate chemical exposures andtoxicities. Costly errors can result whenevaluations are made on the basis of inap-propriate or poorly documented data. Themore accurate the data we use to conductrisk assessments, the more confident wecan be that our efforts to protect the publicand the environment are appropriate andeffective.

The risk assessment we report here,though limited in scope and preliminary innature, illustrates how the methodology canbe applied to turfgrass on golf courses. Inorder to expand and complete this riskassessment, it is necessary to broaden itsscope and to reduce uncertainties inherent inits assumptions. To do this, we must verifyand expand the database on pesticide fate,transport, dislodgeability and toxicology,and on human behaviors that result in poten-tial exposure. These data are optimized forrisk assessments on golf courses when turf-grass scientists and toxicologists collabora-tively design the gathering, testing, andanalysis of the data. This risk assessmentrepresents our initial efforts to expand theexposure database and refine the risk assess-ment methodology for use in golf coursemanagement.

InterpretationDose is safeDose is safeSafety is less certain

InterpretationSite is safeSite is safeSafety is less certain

Hazard QuotientDose/RID < 1Dose/RID = 1Dose/RID> 1

DoseDose <RIDDose = RIDDose > RID

of chemical absorption are rarely done onhuman subjects and may be the source ofconsiderable uncertainty in estimatingchemical intakes. The dermal permeabilitieswe used are rough estimates based uponpublished studies, but a more thoroughexamination of the literature may yieldinformation that enables us to refine theseestimates.

The assumptions we made regardingexposure events and durations are worst-casescenarios and would apply to very fewpeople. Age-adjusted body weights areaverages and actually fit only a small num-ber of people. Summing the toxicity scores(Hazard Quotients) of various chemicalsmay overestimate potential health risks fromchemicals that target different tissues andorgans. Conversely, the potential for syner-gistic toxicity is not directly considered inthe risk assessment process.

The current means of addressing theseuncertainties are through extreme conserva-tism in all extrapolations and assumptionsand by the use of large "uncertainty factors"that reduce the chemical dose consideredsafe. For example, determination of Refer-ence Doses is typically done by finding thedose at which no effects are produced in ratsor mice and dividing that dose by a "safetyfactor" of 10 to 10,000. These safety factorsare used to account for uncertainty and tobe sure that even the most sensitive humanswould not be adversely affected at theReference Dose. Use of such large safetyfactors may often result in RIDs (safe dailydoses) that are actually far below a dose thatcould produce effects in humans. This ap-proach is prudent because the process ofhealth risk assessment is intended to supportdecision making that is protective of publichealth and the environment rather than to

Agency for Toxic Substances and DiseaseRegistry (1992) Public Health AssessmentGuidance Manual. PB92-147164.

Bronaugh, R L., and Barton, C. N. (1993) Pre-diction of human percutaneous absorption withphysicochemical data. InHealth Risk Assessment:Dermal and Inhalation Exposure and Absorptionof Toxicants, chapter 8, RG.M. Wang, 1. B.Knaak, and H. I.Maibach eds. CRC Press, BocaRaton.

Nolan, R 1., Rick, D. L., Freshour, N. L., andSaunders, J. H. (1984) Chlorpyrifos: pharmaco-kinetics in human volunteers. Toxicology andApplied Pharmacology 73; 8-15.

U.S. Environmental Protection Agency (1989)Risk Assessment Guidance for Superfund,Volume I, Human Health Evaluation Manual (PartA). EPA/540/l-891002.

U.S. Environmental Protection Agency (1992)Dermal Exposure Assessment: Principles andApplications. Interim Report. EPA/600/8-91/011B.

14 USGA GREEN SECTION RECORD