New Paradigm for Sizing Riparian ... - Dr. M.Todd WalterNew Paradigm for Sizing Riparian Buffers to Reduce Risks of Polluted Storm Water: Practical Synthesis M. Todd Walter, M.ASCE1;

New Paradigm for Sizing Riparian Buffers to Reduce Risksof Polluted Storm Water: Practical Synthesis

M. Todd Walter, M.ASCE1; Josephine A. Archibald2; Brian Buchanan3; Helen Dahlke4; Zachary M. Easton5;Rebecca D. Marjerison6; Asha N. Sharma7; and Stephen B. Shaw8

Abstract: Riparian buffers are commonly promoted to protect stream water quality. A common conceptual assumption is that buffers“intercept” and treat upland runoff. As a shift in paradigm, it is proposed instead that riparian buffers should be recognized as the partsof the landscape that most frequently generate storm runoff. Thus, water quality can be protected from contaminated storm runoff bydisassociating riparian buffers from potentially polluting activities. This paper reviews and synthesizes some simple engineering ap-proaches that can be used to delineate riparian buffers for rural watersheds based on risk of generating runoff. Although reference is madeto specific future research that may improve the proposed methods for delineating riparian buffers, the approaches described here provideplanners and engineers with a set of currently available scientifically defensible tools. It is recommended that planners and engineers useavailable rainfall and stream discharge data to parameterize the buffer-sizing equations and use variable-width buffers, based on atopographic index, to achieve a realistic representation of runoff generating areas.

DOI: 10.1061/�ASCE�0733-9437�2009�135:2�200�

CE Database subject headings: Stormwater management; Runoff; Hydrology; Water quality.

Introduction

For over two decades, riparian buffers have been aggressivelyencouraged as part of the environmental protection policies of theUnited States �e.g., Calhoun 1988; Welch 1991; Lee et al. 2004�.There is copious published research demonstrating that protectingnear-stream parts of the landscape typically leads to improvedstream water quality �e.g., Lowrance et al. 1997�. In fact, riparianbuffers are promoted to protect an incredibly wide range of water-and habitat-quality attributes. Unfortunately, the data correlatingbuffer size with improved water quality are generally quite scat-tered or inconsistent. The basic underlying research has been re-viewed and re-reviewed many times �e.g., Lammers-Helps and

1Assistant Professor, Dept. of Biological and Environmental Engi-neering, Cornell Univ., Ithaca, NY 14853 �corresponding author�. E-mail:[email protected]

2Graduate Student, Dept. of Biological and Environmental Engineer-ing, Cornell Univ., Ithaca, NY 14853.



5Post-doctoral Researcher, Dept. of Biological and Environmental En-gineering, Cornell Univ., Ithaca, NY 14853.




Note. Discussion open until September 1, 2009. Separate discussionsmust be submitted for individual papers. The manuscript for this paperwas submitted for review and possible publication on March 19, 2008;approved on August 4, 2008. This paper is part of the Journal of Irriga-tion and Drainage Engineering, Vol. 135, No. 2, April 1, 2009. ©ASCE,
ISSN 0733-9437/2009/2-200–209/$25.00.
200 / JOURNAL OF IRRIGATION AND DRAINAGE ENGINEERING © ASCE

Downloaded 19 Mar 2009 to 128.253.239.201. Redistribution subject to

Robinson 1991; USACE 1991; Van Deventer 1992; Lowranceet al. 1997; Wegner 1999; Christensen 2000; Broadmeadow andNisbet 2004; Parkyn 2004; Lee et al. 2004; Hawes and Smith2005; USEPA 2005�, consistently concluding that, in light of sci-entific uncertainty and inconsistent, quantifiable impacts on waterquality, “professional judgment” �Lowrance et al. 1997; Haber-stock et al. 2000� as well as social, economic, and other noneco-logical factors carry heavy weight in establishing riparianvariances and guidelines �e.g., Fischer and Fischenich 2000; Leeet al. 2004�. This is often explicitly noted in most regional/statebuffer guidelines, which are generally available online. As a re-sult, it is not surprising that recommendations for riparian bufferwidths range over two orders of magnitude �e.g., Hawes andSmith 2005�, with typical U.S./Canadian averages in the range of10–30 m �Lee et al. 2004�.

Some of the most scientifically consistent riparian buffer re-search has demonstrated that relatively thin ��20 m� buffers canhave remarkably positive impacts on stream water temperature�e.g., Lynch et al. 1984; Beschta et al. 1987; Kochenderfer andEdwards 1991; Johnson and Jones 2000; Murray et al. 2000;Jackson et al. 2001; Macdonald et al. 2003�, bank stabilization�e.g., Erman et al. 1977; Wipple et al. 1981; Beschta and Platts1986�, and sediment trapping �e.g., Wenger 2003; Broadmeadowand Nisbet 2004; Hawes and Smith 2005�. However, the data areparticularly imprecise with regards to buffer sizes required foreffective reduction in soluble nonpoint source pollutants such asnutrients �e.g., Daniels and Gilliam 1996; Wenger 2003; USEPA2003�, pesticides �e.g., Neary et al. 1993; Hatfield et al. 1995;Arora et al. 1996; USDA-NRCS 2000; Lin et al. 2002�, andpathogens �Doyle et al. 1977; Zhang et al. 2000; Jamieson et al.2002; Ferguson et al. 2003; Tyrrel and Quinton 2003; Unc andGoss 2004; Oliver et al. 2005 Pachepsky et al. 2006�. This isprobably because there is not sufficient understanding of the fate-transport processes in the environment to accurately predict mo-bility and transformations in riparian areas �e.g., Polyakov et al.
2005�; in fact, most of the references cited previously with respect
/ MARCH/APRIL 2009

ASCE license or copyright; see http://pubs.asce.org/copyright

to pathogens are review papers highlighting the complexity ofpathogen transport in the environment, yet buffers are commonlypromoted as effective pathogen traps based on little actual bufferdata �e.g., Barling and Moore 1994; USDA-NRCS 1998, Wegner1999; Parkyn 2004�. Nitrogen may constitute the best understoodnutrient with respect to riparian processes and even in this casethe data are wildly scattered �e.g., see nitrate data in U.S. ArmyCorps of Engineers �USACE� �1991� and USGS �2004�� and thepersistent nitrogen-related problems in the Gulf of Mexico �e.g.,Rabalais et al. 1999; Alexander et al. 2000; Howarth et al. 2000�and Chesapeake Bay �e.g., Staver and Brinsfield 2001; IAN-EcoCheck 2007� testify to the critical knowledge gaps. Despiteenormous research efforts, phosphorus may represent one of themost poorly understood pollutants and studies are at best incon-clusive with regards to the ability of riparian buffers to “filter”phosphorus �e.g., USACE 1991; Lowrance et al. 1997�. Indeed,some management practices designed to reduce phosphorus loadsto streams actually result in enhanced loadings �Dillaha et al.1988, 1989a,b; Gaynor and Findlay 1995; Daverede et al. 2003;Novotny 2003�.

This paper proposes a new paradigm for delineating riparianbuffers that is scientifically defensible and associated with quan-tifiable risks. Buffers are consistently conceptualized as areasthrough which water flows and pollutants are removed �Welsch�1991�, widely adopted by a variety of agencies e.g., USDA For-est Service �1996�, USDA-NRCS/USEPA �1997��. The idea thatbuffers intercept runoff is not unique to fixed-width buffers, butalso underlies many precision or variable-width buffer designs�e.g., Dosskey et al. 2002, 2005; Polyakov et al. 2005�. This viewof buffers probably originated with early soil conservation re-search, which was focused on removing suspended sedimentsfrom overland flows. Indeed, even narrow ��10 m�, well-vegetated buffers, if properly maintained, have been shown to bevery effective at slowing overland flow and allowing sedimentand sediment-borne nonpoint source pollution to settle out ofstorm runoff �e.g., Peterjohn and Correll 1984; Dillaha et al.1988; Magette et al. 1989�. The flow-interception buffer conceptis also often appropriate for denitrificating subsurface nitratefluxes that surface in riparian areas �e.g., Lowranceet al. 1997; USEPA 2005�. However, the writers suspect that thisconcept may not be useful for designing riparian buffers for re-ducing dissolved pollutant loads from storm flows �e.g., Walteret al. 1979�, which may explain why riparian buffers have hadinconsistent benefits with respect to dissolved phosphorus �Shep-pard et al. 2006�. Although, as noted earlier, there is especiallypoor understanding of the relevant processes associated withphosphorus fate-transport in storm runoff, relatively good insightsexist into the processes that generate storm runoff.

In humid, well-vegetated parts of the world, such as the north-eastern United States, hydrologists recognize that storm runoff istypically generated from small parts of a watershed, specifically,those where the soil profile is prone to saturating �e.g., Dunne1970; Dunne and Black 1970a,b; Dunne et al. 1975; Dunne andLeopold 1978� or at least getting wet enough to promote rapidlateral drainage �Lyon et al. 2006a,b�. The extent of these runoffcontributing areas varies throughout the year as the landscapedries and wets and, thus, they are typically referred to as “variablesource areas” or VSAs. This mechanism for generating stormflow is in contrast to so-called Hortonian flow, which is producedwhen and where rainfall intensities exceed soil infiltration capac-ity �e.g., Horton 1933, 1940�; a situation that appears to occuronly rarely in the northeastern United States �Walter et al. 2003�.

The writers propose delineating riparian buffers based on their

JOURNAL OF IRRIGATION A


risk of generating runoff and that they should be protected todisassociate areas prone to generating storm runoff from areaslikely to receive potentially polluting activities. Although claimedto be a new paradigm, the basic concept has been suggested inpeer-reviewed literature for several years �Walter and Walter1999; Gburek et al. 2000; Walter et al. 2000, 2001; Gburek et al.2002; Walter et al. 2005; Agnew et al. 2006; Qui et al. 2007� andeven demonstrated to be surprisingly effective for reducing phos-phorus loads to streams �James et al. 2007; Easton et al. 2008�.This paper is a review that synthesizes previous research to dem-onstrate some simple ways of quantifiably delineating riparianbuffers based on their propensity to generate storm runoff, i.e., todelineate them based on the best science instead of relying soheavily on professional judgments and inconclusive research find-ings �see the reviews cited earlier�. It is recognized that there aresome other similarly unique perspectives on the roles of riparianbuffers including stream–buffer interactions during flooding �Il-hardt et al. 2000; Skally and Sagor 2001� and urban riparian areasas sinks for stream nitrate �Kaushal et al. 2004; Groffman et al.2005�. The proposed method uniquely addresses the role of ripar-ian areas as hotspots of storm runoff generation.

A two-step approach for identifying runoff generating areas:will be presented first, use a simple rainfall-runoff model to de-termine the extent or size of the area generating runoff and, sec-ond, locate or delineate that area in the landscape. For the secondstep, both a set-width buffer and a variable-width buffer based ontopography are considered.

Determining Extent of Contributing Area

There are several good simulation models that can capture VSAhydrology �e.g., TOPMODEL �Beven and Kirkby 1979�,DHSVM �Wigmosta et al. 1994�, SMR �Frankenberger et al.1999�, VSLF �Schniederman et al. 2007�� but these generally re-quire more input data and/or calibration than are typically avail-able for planning and designing best management practices likeriparian buffers. Thus, the writers propose using the U.S. Depart-ment of Agriculture, Natural Resources Conservation Service��NRCS�, formerly the Soil Conservation Service or SCS� CurveNumber �CN� equation for estimating storm flow �USDA-SCS1972; USDA-NRCS 1986�: The CN equation relates storm runoff�Q� to effective rainfall �Pe� and a watershed storage factor �S�,usually referred to as the maximum soil storage—all in units ofdepth:

Q =Pe

2

Pe + S�1�

where S is typically related to tabulated CN values ��mm�: S=25,400 /CN-254� and the CN values are based on qualitativeland use and soil infiltration characteristics. The use of thesetables implicitly links the CN equation with Hortonian runofftheory �Walter and Shaw 2005�. Steenhuis et al. �1995� demon-strated that the concepts underpinning the equation itself are moreindicative of VSA hydrology. The runoff in Eq. �1� is in units ofdepth and can be converted to a runoff volume by multiplying bythe watershed area �Aws�. In accordance with VSA theory, thevolume of runoff can be estimated as the depth of effective rain-fall over the runoff contributing area �Ac�. Noting that the runoffvolume based on Eq. �1� must equal the VSA runoff volume�PeAc�, Gburek et al. �2000, 2002� suggested one way to estimate
the average runoff contributing area �avg�Ac�� during a storm
ND DRAINAGE ENGINEERING © ASCE / MARCH/APRIL 2009 / 201


avg�Ac� = � Pe

Pe + S�Aws �2�

Steenhuis et al. �1995� considered the dynamic nature of the con-tributing area during a storm and developed an equation for themaximum contributing area �Max�Ac�� associated with Pe

max�Ac� = �1 −S2

�Pe + S�2�Aws �3�

Using the methods presented here, the risk of a particular ex-tent of runoff contributing area is associated with the rainfall fre-quency, i.e., the 10-year rainfall would produce the 10-yearcontributing area. This, of course, is not strictly true, but it hasbeen a common assumption in engineering design and will beadopted here. Because the original CN equation had no variablefor time, the writers’ contributing area equations, Eqs. �1� and �2�,similarly lack this detail and it is typical for engineers to simplyuse daily rainfall. However, it is likely that any time period isacceptable as long as it is short enough that storm flow �overlandflow and rapid lateral drainage� dominates the water budget, i.e.,percolation, evapotranspiration, and micropore drainage are notsubstantial components to the water balance.

Determining S is challenging. For gauged watersheds it can beback-calculated from Eq. �1� for a range of different storm sizes�e.g., Steenhuis et al. 1995, Lyon et al. 2004, Shaw and Walter,personal communication, 2008�. For ungauged watersheds, Gbu-rek et al. �2002� suggest simply using the published tables �e.g.,USDA-SCS 1972; USDA-NRCS 1986� to estimate CN based onland uses and soil characteristics. Unfortunately, as mentionedearlier, the tables implicitly assume Hortonian flow is the domi-nant storm runoff process and this paper proposes a way of de-lineating riparian buffers based on VSA hydrology. So there is anuncomfortable discontinuity in using the tabulated CNs in thiscontext. Ideally, one would back-calculate S for nearby water-sheds with similar characteristics to the one for which riparianbuffers are to be developed. This will be explored further in theexample application.

Delineating the Riparian, Runoff Contributing Area

Riparian buffers are most commonly defined by a fixed set-backdistance or buffer width, although, recently various ways of de-lineating nonuniform buffers have been proposed �e.g., Ilhardt etal. 2000; Dosskey et al. 2002, 2005; Polyakov et al. 2005�. In theproposed method for delineating a riparian buffer, a fixed bufferwidth, w, can easily be estimated by

w = � Ac

2Ls� �4�

where Ls=total length of stream in the watershed. This approachwas suggested by Gburek et al. �2000, 2002�. In this project,fixed-width buffers were created using the ArcGIS �ESRI soft-ware, Redlands, Calif.� buffer tool.

A more realistic pattern of VSAs is captured by using a topo-graphic index �Agnew et al. 2006; Lyon et al. 2006a,b�. There aremany variants on the topographic index concept �e.g., Beven andKirkby 1979; O’Laughlin 1986; Ambroise et al. 1996; Habets andSaulnier 2001; Walter et al. 2002� and they all result in similarpatterns, although the actual values of the indices will vary. Here
only the patterns of relative likelihood of runoff generation are of


concern, so the specific method used is probably not too impor-tant. In this investigation the soil topographic index was calcu-lated as

� = ln� a

KsD tan�� 5�

where �=soil topographic index; a=upslope contributing area perunit contour length �m2�; tan��=topographic slope; Ks=meansaturated hydraulic conductivity of the soil �m day−1�; and D=soil depth �m� �Walter et al. 2002�. The contributing areas weredetermined in ArcView 3.3 �ESRI software� based on a hydro-logically corrected 10 m digital elevation model �DEM� using amultidirectional flow algorithm �Montgomery and Foufoula-Georgiou 1993�. Slopes were calculated from the same DEMusing the method of Horn �1981�. The soil characteristics, Ks andD, were obtained from the SURRGO database �USDA-NaturalResources Conservation Service�. Higher values of � representhigher frequency of generating runoff, i.e., perennial streams willhave the highest � values. For a given Ac, the places most likelyto saturate were determined by determining a threshold � suchthat the area covered by all � values greater than this thresholdwas equal to Ac �Fig. 1�. The combination of Eq. �1� and thetopographic index �Eq. �5�� have been corroborated with fieldobservations of soil saturation degree �e.g., Schneiderman et al.2007; Easton et al. 2008�.

Example Applications

Site Description

To demonstrate the methodologies used for delineating riparianbuffers the 230 km2 Salmon Creek Watershed was considered,which flows into Cayuga Lake north of Ithaca, N.Y. �Fig. 2�. Theland use distribution is typical of the region, about 70% is agri-

Fig. 1. Schematic illustrating how runoff contributing areas weredetermined; �a� a “map” of topographic indices, �, �Eq. �4�� is ana-lyzed to determine �b� the continuous distribution of topographic in-dices. For any fractional contributing area �Ac /Aws� �from Eqs. �2�and �3�� there is a threshold � value that �c� corresponds to theboundary of the runoff contributing area.

cultural land, 28% is mixed forest, and the remaining 2% is a mix

/ MARCH/APRIL 2009


of residential, commercial, and mixed urban land. According tothe USGS National Hydrography Dataset, Ls=330 km. A USGSgauge �#04234018, latitude 42°33�13�, longitude 76°32�08�� isabout 2 km upstream from Cayuga Lake. Daily streamflow datawere collected by the USGS from October 1964 to September1968 and resumed in March 2007. Precipitation is measured at aNortheast Regional Climate Center weather station near the Cor-nell University campus, approximately 10 km south of the streamgauge.

Determining S

Two methods were used to estimate S �Eqs. �1�–�3��: it was back-calculated from measured stream discharge-precipitation data andwas estimated based on published tables of CN �USDA-SCS1972�.

Fig. 2. The Salmon Creek Watershed. �Map projection: UTM meters,Conservation and U.S. Geological Survey �stream network and lakes�NRCS �watershed boundary�; U.S. Census Bureau �roads and countycal Survey �land use�.�

S values were back-calculated using Eq. �1� for 28 well-



defined events, i.e., events in which the precipitation and resultingstorm hydrograph were clearly separated from adjacent events.�Note: it was assumed Pe=measured rainfall.� For each event,storm runoff, Q, was separated from base flow by subtracting thepre-event streamflow from the hydrograph. Using Pe–Q pairs foreach event, Eq. �1� was rearranged to solve for S and a unique Svalue for each event was back-calculated. The back-calculated Svalues varied substantially from event to event. Each S value waspaired with its corresponding pre-event base flow and it wasfound that pre-event base flow correlated well with S �Fig. 3�. Thewriters hypothesize that base flow is a good indicator of anteced-ent wetness �Troch et al. 1993; Shaw and Walter, personal com-munication, 2008�. Other base flow separation methods result insimilar relationships as shown in Fig. 3, although the absolutevalues of S vary with the base flow separation method used �data

18N, datum NAD 1983� �Data sources: NYS Dept. of EnvironmentalDept. of Environmental Conservation and U.S. Dept. of Agriculture-aries�; and U.S. Environmental Protection Agency and U.S. Geologi-

zone; NYSbound

not shown�. To calculate an average or representative S, measured



stream discharge was used to determine the median base flow�0.72 cm� and the corresponding S value was calculated using thebase flow–S correlation �see Fig. 3 caption�; S=23.5 cm. Onecould use a bivariate approach to determine a more precise aver-

Fig. 3. S as a function of base flow, Qb. Symbols are values calcu-lated from measured rainfall and stream discharge. The line is thebest fit power function, S=1.52Qb

−1.04 �R2=0.77�. The relationshipsuggests that the available water storage, S, decreases as base flowincreases, which agrees with earlier work suggesting base flow is agood indicator of landscape moisture conditions, i.e., high-base flowindicates a wet landscape �e.g., Troch et al. 1993�.

Table 1. Determination of Salmon Creek, N.Y. Curve Numbers �CN� U

Land use Hydrologic soil groupa Percent of watersh

Crop/pasture A 0.45

B 51.11

C 15.78

D 2.70

Mixed forest A 1.18

B 15.90

C 8.31

D 2.86

Commercial B 0.07

C 0.04

D 0.01

Other agricultural B 0.26

C 0.03

D 0.01

Mixed urban A 0.07

B 0.53

C 0.16

D 0.01

Gravel pitt/quarry B 0.10

C 0.02

D 0.01

Residential A 0.05

B 0.21

C 0.10

D 0.01

Area weighted averagesaFor soils identified as a combination of hydrologic soil groups, the firstalso, all land uses with no associated soil group ��1% of the area� werebThe writers’ best guesses assumed: crop /pasture=crops, straight row,
agricultural=pasture, good condition; residential=1 /2 acre residential lots �all o


age S �e.g., Shaw and Walter, personal communication, 2008�, butthat is beyond the scope of this paper. Because most rivers are notgauged, practitioners will have to extrapolate back-calculated Svalues from nearby gauged streams. S was calculated for threeadditional rural watersheds located between the Catskill Moun-tains in southeastern New York and near Lake Erie in westernNew York and all were similar to Salmon Creek �S=22.8, 25.5,and 27.8 cm�. In this example it was found that setting Pe to beequal to the measured rainfall worked best. Usually engineersdefine Pe as the total rainfall minus an initial abstraction, Ia,which is typically defined as 20% of S �USDA-SCS 1972�. Thewriters have found that Ia is negligible for most rural watershedsin this region, especially for rainfall events �1 cm �data notshown�. Other researchers have also found Ia to be much smallerthan 0.2S �Jiang 2001; Mello et al. 2003; Woodward et al. 2003;Baltas et al. 2007�.

Using the published tables of CN �USDA-SCS 1972�, it wasestimated that S=8.7 cm �CN=74�. However, because of thequalitative nature of the tables, this value is somewhat subjectiveand could, reasonably, vary substantially, S=6.6–14.5 cm �CN=79 and 64, respectively� �Table 1�. Note, use of the tables ap-pears to presume Ia=0.2S.

Delineating Riparian Buffers

To illustrate the various proposed approaches to sizing riparianbuffers, runoff contributing areas, Ac, were calculated for rain

blished, Tabulated Values �USDA-SCS, 1972�—�S=25,400 /CN−254�

a Best-guess CNb Highest tabulated CN Lowest tabulated CN

67 72 39

78 81 61

85 88 74

89 91 80

30 45 30

55 66 55

70 77 70

77 83 77

92 92 92

94 94 94

95 95 95

61 68 61

73 79 73

80 86 80

81 81 81

88 88 88

91 91 91

83 83 83

85 85 85

89 89 89

91 91 91

61 61 61

75 75 75

83 83 83

87 87 87

74.4 79.3 63.6

ted was assumed, e.g., soil hydrologic group C/D was assumed to be C;ed soil group B.

ondition; mixed forest=woods �farm woodlots�, good condition; other

sing Pu

ed are

type lisassign

good c
ther land uses have only a single CN for each soil group�.
/ MARCH/APRIL 2009


events ranging from 0.03 to 12.9 cm, which corresponds to theobserved range of events during 1971–2001. Fig. 4�a� shows thecumulative distribution of daily rainfalls and Figs. 4�b and c�show the corresponding fixed width of riparian buffers �Eq. �4��that would correlate to average �Eq. �2�� and maximum �Eq. �3��runoff contributing areas, respectively. Figs. 4�b and c� also showhow the buffer widths differ when a backcalculated S �solid lines�versus tabulated S �dashed lines� are used. As a reference point,only 3% of rain days experience rain depths �2.5 cm �1 in.�

Fig. 4. �a� Cumulative distribution of rainfall events �1971–2000�and the fraction of events that generate runoff contributing areaslarger than a fixed-width buffer based on �b� the average contributingarea �Eq. �2�� and �c� the maximum contributing area �Eq. �3��.Dashed-lines are based on tabulated CN values, the thin dashed linescorrespond to the writers’ estimated uncertainty in determining CNfrom the tables �S=3 and 12 cm�. Solid lines are based on backcal-culated S values; the heavy lines use the S value backcalculated forSalmon Creek and the thin lines correspond to the highest and lowestbackcalculated S values �22.8 and 27.8 cm� for similarly rural water-sheds. �Note: the lines corresponding to S=22.8 cm cannot be distin-guished from the heavy lines.� The arrows illustrate that �a� 3% of thedaily rainfall “events” are �2.5 cm; �b� average; and �c� maximumrunoff contributing areas produced by a 2.5 cm rainfall, which willonly be exceeded by 3% of the events, correspond to the areas en-compassed by 30 or 60 m riparian buffer, respectively, for thiswatershed.

�arrow in Fig. 4�a��. Thus, it was anticipated that only 3% of the



rain days would experience runoff contributing areas larger thanthat generated by a 2.5 cm rainfall; this is indicated with arrowsin Figs. 4�b and c�. Using the backcalculated S �solid lines inFigs. 4�b and c��, the Ac generated by a 2.5 cm storm correspondsto the area encompassed by 30 m buffer based on avg�Ac� �Eq.�2�, Fig. 4�b�� and 60 m buffer based on the max�Ac� �Eq. �3�,Fig. 4�c��. It is interesting, although perhaps not overly notewor-thy, that, for any given level of risk �frequency larger�, the bufferwidth of Fig. 4�b� is approximately half that in Fig. 4�c� based onthe backcalculated S values.

Clearly, whether one uses a backcalculated S or one based onpublished tables makes an enormous difference in buffer size,especially for small, frequent events ��1 cm� �Figs. 4�b andc��, i.e., those exceeded by �15% of the events �Fig. 4�a��. Forlarger events ��1.4 cm�, those which are only exceeded by10% of rainfall events �Fig. 4�a��, the buffer widths based onthe backcalculated S are within the range of buffer widths due tothe writers’ uncertainty in interpreting the published CN tables�Figs. 4�b and c��. However, buffer widths based on tabulatedvalues suggest no runoff until �1.7 cm �0.6 in.� of rain, which isan artifact of the initial abstraction and unrealistic as storm runoffis observed in the hydrograph for some daily rainfalls �1.7 cm�data not shown�. By contrast, the buffer widths based on thebackcalculated S values suggest continuum of growing runoffcontributing areas with increasing storm size, which is more con-sistent with VSA-hydrologic theory.

It is noteworthy that the variability in buffer widths due todifferences in backcalculated S values was small, especially com-pared to the uncertainty introduced in using the tabulated CNs�Fig. 4�. Even though relatively consistent S values were foundacross four regional watersheds, these values will vary somewhatwith land use. For example, S for a relatively urbanized watershednear Rochester, N.Y. was found to be 8.7 cm, about a third of thevalues found for the rural, agricultural watersheds. This is ex-pected because impervious areas generally reduce the water hold-ing capacity of the landscape; for a wholly paved watershed, S0 cm.

As previously noted by Agnew et al. �2006�, fixed-width buff-ers only approximately correlate with actual runoff contributingareas. To illustrate this, the writers created maps of both fixed-width and topographic-index delineated riparian buffers for the3% risk level �2.5 cm �1 in.� rain� using the maximum runoffcontributing area approach �Eq. �3�� Fig. 5�. Interestingly, inmany places the fixed-width buffer covers parts of the landscapethat are not likely to generate runoff �Fig. 5�b��. There are severalheadwater areas that appear to be especially prone to generatingrunoff even though they are not necessarily adjacent to the peren-nial stream �Fig. 5�c��. The writers speculate that these headwatersource areas may be largely responsible for generating runoff thatis often observed to “short-circuit” fixed-width buffers. One pos-sible way to help reconcile the differences between fixed-width-and topographic-index-based buffers might be to develop a hy-drologically based way of determining the effective stream net-work, which includes ephemeral and intermittent streams that areoften not comprehensively included in regularly mapped hydrog-raphy and may include human modifications to the natural drain-age system �e.g., ditches�; some such methods have beenproposed �Tarboton 1997; Duke et al. 2003, 2006; Seibert andMcGlynn 2007� but need to be tested in the context of this ripar-ian buffer work and may benefit from improved topographical
data, such as that generated by LiDAR.


Conclusions and Recommendations

The new paradigm presented for delineating riparian buffers spe-cifically addresses the problem of stream water contamination bypolluted storm runoff. It should be emphasized that buffers pro-vide a variety of environmental benefits beyond this problem and,thus, the methods presented here are meant as additional tools forassessing, planning, and designing riparian buffers. For example,if a planner is interested in trapping sediments, for which thetransport processes are different from solutes, this should prob-ably be considered independently of the context of the “new para-digm” presented here. Although, many previous riparian bufferliterature reviews have emphasized the need to carefully considerthe riparian hydrology, usually with respect to flow paths that maybypass a riparian buffer, the methods put forth here uniquely con-sider the buffer itself to be the source of storm flow. Water qualityis protected by keeping potentially polluting activities out of theseareas �e.g., James et al. 2007; Easton et al. 2008�

With regards to the suite of methods synthesized here, it isrecommended that planners try to determine S from availablestream discharge and rainfall data rather than use published tables�USDA-SCS 1972; USDA-NRCS 1986�. This will require locat-ing nearby gauged watersheds that are as similar as possible to theone for which riparian buffers are being designed. If the tables areused to determine S, practitioners need to recognize that there will

Fig. 5. Maps of fixed-width �light areas� and variable-width �dark areaCreek, N.Y. watershed. Insets �b� and �b� show examples of parts of thbuffers and vice versa, respectively.

be substantial uncertainty in the resulting buffer size. The results



presented here suggest that the tabulated values will give similarresults to those based on rainfall-discharge data for large rainfallevents. If one were to design a riparian buffer using the 1-year24-h rainfall, it is likely that buffer size would be the same fortabulated and backcalculated S values. For example, the 1-year24-h rainfall for Salmon Creek is 5.7 cm �Hershfield 1961� andfor which buffer widths determined with either tabulated or back-calculated S values are essentially the same, although large �Figs.4�b and c��.

It is also recommended that, when possible, preference begiven to variable-width buffers, based on a topographic index,over fixed-width buffers. Although the likelihood of an area gen-erating saturation-excess runoff generally decreases with increas-ing distance from a stream, the patterns are only looselycorrelated to distance from stream �Agnew et al. 2006�. Topo-graphic indices appear to capture much more of the realistically,irregular patterns of runoff generating areas �e.g., Agnew et al.2006; Lyon et al. 2006a, b; Schneiderman et al. 2007�. However,fixed-width buffers are still likely to be effective �e.g., Eastonet al. 2008� although they will not capture some runoff-proneareas and will likely encompass some areas that have little like-lihood of generating runoff.

The immediate research needs that will add substantially to thewriters’ confidence in the proposed methods for delineating ripar-ian buffers are: �1� developing a way to estimate S for ungauged

rian buffers associated with a 2.5 cm �1 in.� rain event in the Salmonrshed where fixed-width buffer area is greater than the variable-width

s� ripae wate

watersheds; �2� formalizing a way to incorporate antecedent wet-

/ MARCH/APRIL 2009


ness into quantifying the risk a runoff contributing area extent�Ac�; and �3� assessing reliable ways of delineating a watershed’sdrainage network beyond that shown in published hydrography.

Acknowledgments

The writers would like to explicitly recognize Dr. Tammo Steen-huis and Dr. Michael Walter �Cornell University�, Dr. Jane Fran-kenberger �Purdue University�, Drs. Jan Boll and Erin Brooks�University of Idaho�, and Dr. Bil Gburek �retired, USDA-ARS/Penn State University� for their early, significant, and direct con-tributions to this summary work. The writers would also like tothank the three reviewers of this paper for their valuable feedbackand suggestions.

References

Agnew, L. J., Lyon, S., Gérard-Marchant, P., Collins, V. B., Lembo, A. J.,Steenhuis, T. S., and Walter, M. T. �2006�. “Identifying hydrologicallysensitive areas: Bridging science and application.” J. Environ. Man-age., 78�1�, 64–76.

Alexander, R. B., Smith, R. A., and Schwarz, G. E. �2000�. “Effect ofstream channel size on the delivery of nitrogen to the Gulf ofMexico.” Nature (London), 403�17�, 758–761.

Ambroise, B., Beven, K., and Freer, J. �1996�. “Toward a generalizationof the TOPMODEL concepts: Topographic indices of hydrologicalsimilarity.” Water Resour. Res., 32�7�, 2135–2145.

Arora, K., Mickelson, S. K., Baker, J. L., Tierney, D. P., and Peters, C. J.�1996�. “Herbicide retention by vegetative buffer strips from runoffunder natural rainfall.” Trans. ASAE, 39�6�, 2155–2162.

Baltas, E. A., Dervos, N. A., and Mimikou, M. A. �2007�. “Technicalnote: Determination of the SCS initial abstraction ratio in an experi-mental watershed in Greece.” Hydrology Earth Syst. Sci., 11�6�,1825–1829.

Barling, R. D., and Moore, I. D. �1994�. “Role of buffer strips in man-agement of waterway pollution: A review.” Environ. Manage. (N.Y.),18�4�, 543–558.

Beschta, R. L., Bilby, R. E., Brown, G. W., Holtby, L. B., and Hofstra, T.D.�1987�. “Stream temperature and aquatic habitat: fisheries and for-estry interactions.” Streamside management: Forestry and fishery in-teractions, E. O.Salo and T. W.Cundy, eds., Inst. of For. Res., Univ. ofWashington, Seattle, Conrtib. 57, 191–232.

Beschta, R. L., and Platts, W. S. �1986�. “Morphological features of smallstreams: Significance and function.” Water Resour. Bull., 22�3�, 370–379.

Beven, K. J., and Kirkby, M. J. �1979�. “A physically-based, variablecontributing area model of basin hydrology.” Hydrol. Sci. Bull.,24�1�, 43–69.

Broadmeadow, S., and Nisbet, T. R. �2004�. “The effects of riparian for-est management on the freshwater environmentl: A literature review.”Hydrology Earth Syst. Sci., 8�3�, 286–305.

Calhoun, J. M. �1988�. “Riparian management practices of the Depart-ment of Natural Resources.” Streamside management: Riparian wild-life and forestry interactions, contribution No. 59, K. Raedeke, ed.,Institute of Forest Resources, Univ. of Washington, Seattle, 207–211.

Christensen, D. �2000�. “Protection of riparian ecosystems: A review ofbest available science, Jefferson County Health Division.” �http://www.co.jefferson.wa.us� �Feb. 5, 2008�.

Daniels, R. B., and Gilliam, J. W. �1996�. “Sediment and chemical loadreduction by grass and riparian filters.” Soil Science Society of Am. J.,60�1�, 246–251.

Daverede, I. C., Kravchenko, A. N., Hoeft, R. G., Nafziger, E. D., Bul-lock, D. G., Warren, J. J., and Gonzini, L. C. �2003�. “Phosphorus
runoff: Effect of tillage and soil phosphorus levels.” J. Environ. Qual.,


32�4�, 1436–1444.Dillaha, T. A., Reneau, R. B., Mostaghimi, S., and Lee, D. �1989a�.

“Vegetative filter strips for agricultural nonpoint source pollution con-trol.” Trans. ASAE, 32�2�, 513–519.

Dillaha, T. A., Sherrard, J. H., and Lee, D. �1989b�. “Long-term effec-tiveness of vegetative filter strips.” Water Environ. Soc., 1�3�, 419–421.

Dillaha, T. A., Sherrard, J. H., Lee, D., Mostaghimi, S., and Shanholtz, V.O. �1988�. “Evaluation of vegetative filter strips as a best managementpractice for feedlots.” J. Water Pollut. Control Fed., 60�7�, 1231–1238.

Dosskey, M. G., Eisenhauer, D. E., and Helmers, M. J. �2005�. “Estab-lishing conservation buffers using precision information.” J. SoilWater Conservat., 60�6�, 349–354.

Dosskey, M. G., Helmers, M. J., Eisenhauer, D. E., Franti, T. G., andHoagland, K. D. �2002�. “Assessment of con-centrated flow throughriparian buffers.” J. Soil Water Conservat., 57�6�, 336–343.

Doyle, R. C., Stanton, G. C., and Wolf, D. C. �1977�. “Effectiveness offorest and grass buffer strips in improving the water quality of manurepolluted runoff.” ASAE Paper No. 77-2501, ASAE, St. Joseph, Mich.

Duke, G. D., Kienzle, S. W., Johnson, D. L., and Byrne, J. M. �2003�.“Improving overland flow routing by incorporating ancillary road datainto digital elevation models.” J. Spatial Hydrol., 3�2�, 1–27.

Duke, G. D., Kienzle, S. W., Johnson, D. L., and Byrne, J. M. �2006�.“Incorporating ancillary data to refine anthropogenically modifiedoverland flow paths more options.” Hydrolog. Process., 20�8�, 1827–1843.

Dunne, T. �1970�. “Runoff production in humid areas.” ARS-41-160, U.S.Department of Agriculture Publication, Washington, D.C.

Dunne, T., and Black, R. D. �1970a�. “An experimental investigation ofrunoff production in permeable soils.” Water Resour. Res., 6�2�, 478–490.

Dunne, T., and Black, R. D. �1970b�. “Partial area contributions to stormrunoff in a small New England watershed.” Water Resour. Res., 6�5�,1296–1311.

Dunne, T., and Leopold, L. �1978�. Water in environmental planning,Freeman, New York.

Dunne, T., Moore, T. R., and Taylor, C. H. �1975�. “Recognition andprediction of runoff-producing zones in humid regions.” Hydrol. Sci.Bull., 20�3�, 305–327.

Easton, Z. M., Walter, M. T., and Steenhuis, T. S. �2008�. “Combinedmonitoring and modeling indicates the most effective agricultural bestmanagement practices.” J. Environ. Qual., 37�5�, 1798–1809.

Erman, D. C., Newbold, J. D., and Roby, K. B. �1977�. “Evaluation ofstreamside buffer strips for protecting aquatic organisms.” CaliforniaWater Resources Center, Univ. of California-Davis, Davis, Calif.

Ferguson, C., de Roda Husman, A. M., Altavilla, N., Deere, D., andAshbolt, N. �2003�. “Fate and transport of surface water pathogens inwatersheds.” Crit. Rev. Environ. Sci. Technol., 33�3�, 299–361.

Fischer, R. A., and Fischenich, J. C. �2000�. “Design recommendationsfor riparian corridors and vegetated buffer strips.” EMRRP TechnicalNotes Collection (ERDC TN-EMRRP-SR-24), U.S. Army EngineerResearch and Development Center, Vicksburg, Miss.

Frankenberger, J. R., Brooks, E. S., Walter, M. T., Walter, M. F., andSteenhuis, T. S. �1999�. “A GIS-based variable source area model.”Hydrolog. Process., 13�6�, 804–822.

Gaynor, J. D., and Findlay, W. I. �1995�. “Soil and phosphorus loss fromconservation and conventional tillage in corn production.” J. Environ.Qual., 24�4�, 734–741.

Gburek, W. J., Drungil, C. C., Srinivasan, M. S., Needelman, B. A., andWoodward, D. E. �2002�. “Variable-source-area controls on phos-phorus transport: Bridging the gap between research and design.” J.Soil Water Conservat., 57�6�, 534–543.

Gburek, W. J., Sharpley, A. N., Heathwaite, L., and Folmar, G. J. �2000�.“Phosphorus management at the watershed scale: A modification ofthe phosphorus index.” J. Environ. Qual., 29�1�, 130–144.

Groffman, P. M., Dorsey, A. M., and Mayer, P. M. �2005�. “N processingwithin geomorphic structures in urban streams.” J. North Am. Benthol.
Soc., 24�3�, 613–625.


Haberstock, A. E., Nichols, H. G., DesMeules, M. P., Wright, J., Chris-tensen, J. M., and Hudnut, D. H. �2000�. “Method to identify effectiveriparian buffer widths for Atlantic salmon habitat protection.” J. Am.Water Resour. Assoc., 36�6�, 1271–1286.

Habets, F., and Saulnier, G. M. �2001�. “Subgrid runoff parameteriza-tion.” Phys. Chem. Earth, Part B, 26�5–6�, 455–459.

Hatfield, J. L., Mickelson, S. K., Baker, J. L., Arora, K., Tierney, D. P.,and Peter, C. J. �1995�. “Buffer strips: Landscape modification toreduce off-site herbicide movement.” Clean water, clean environment,21st Century: Team agriculture, working to protect water resources,Vol. 1. American Society of Agricultural Engineers, St. Joseph, Mich.

Hawes, E., and Smith, M. �2005�. “Riparian buffer zones: Functionsand recommended widths.” Rep. to the Eightmile River Wild and Sce-nic Study Committee, �http://www.eightmileriver.org/resources/digital_library/appendicies/09c3_Riparian%20Buffer%20Science_YALE.pdf� �Feb. 28, 2008�.

Hershfield, D. M. �1961�. “Rainfall frequency atlas for the United Statesfor durations from 30 minutes to 24 hours and return periods from1 to 100 years.” Technical Paper No. 40, U.S. Weather Bureau, Wash-ington, D.C.

Horn, B. K. P. �1981�. “Hillshading and the reflectance map.” Proc. IEEE,69�1�, 14–47.

Horton, R. E. �1933�. “The role of infiltration in the hydrologic cycle.”Trans., Am. Geophys. Union, 14, 446–460.

Horton, R. E. �1940�. “An approach toward a physical interpretation ofinfiltration capacity.” Soil Sci. Soc. Am. Proc., 5, 399–417.

Howarth, R. D., et al. �2000�. “Nutrient pollution of coastal rivers, bays,and seas.” Issues in Ecol., 7, 1–15.

IAN-EcoCheck. �2007�. “2006 Chesapeake Bay report card.” �http://www.eco-check.org/pdfs/2006_chesapeake_report_card.pdf� �alsovia: www.ian.umces.edu or www.eco-check.org� �Jan. 15, 2009�.

Ilhardt, B. L., Verry, E. S., and Palik, B. J. �2000�. “Defining riparianareas.” Riparian management in forests of the continental easternUnited States, E. S. Verry, J. W. Hornbeck, and C. A. Dolloff, eds.,Lewis, New York.

Jackson, C. R., Strum, C. A., and Ward, J. M. �2001�. “Timber harvestimpacts on small headwater stream channels in the coast ranges ofWashington.” J. Am. Water Resour. Assoc., 37�6�, 1533–1549

James, E., Kleinman, P., Veith, T., Stedman, R., and Sharpley, A. �2007�.“Phosphorus contributions from pastured dairy cattle to streams of theCannonsville Watershed, New York.” J. Soil Water Conservat., 62�1�,40–47.

Jamieson, R. C., Gordon, R. J., Sharples, K. E., Stratton, G. W., andMadani, A. �2002�. “Movement and persistence of fecal bacteria inagricultural soils and subsurface drainage water: A review.” Can. Bio-syst. Eng., 44, 1.1–1.9.

Jiang, R. �2001�. “Investigation of runoff curve number initial abstractionratio.” MS thesis, Watershed Management, Univ. of Arizona, Tucson,Ariz.

Johnson, S. L., and Jones, J. A. �2000�. “Stream temperature responses toforest harvest and debris flows in western Cascades, Oregon.” Can. J.Aquat. Sci., 57 �Suppl. 2�, 30–39.

Kaushal, S. S., Groffman, P. M., and Findlay, S. E. �2004�. “Ecosystemrestoration and denitrification in an urban stream.” Conf. Information:89th Annual Meeting of the Ecological Society of America—Lessonsof Lewis and Clark: Ecological Exploration of Inhabited Landscapes,01–06, Ecological Society of America Annual Meeting Abstracts, 89.

Kochenderfer, J. N., and Edwards, P. J. �1991�. “Effectiveness of threestreamside management practices in the central Appalachians.” Proc.,6th Biennial Southern Silviculture Research Conf., Vol. 2, Memphis,Tenn., S. S. Coleman D. G. Neary, comps., eds., Gen. Tech. Rep.SE-70, U.S. Dept. of Agriculture, Forest Service, Southeastern ForestExperiment Station, Ashville, N.C., 1:688–700.

Lammers-Helps, H., and Robinson, D. �1991�. “Literature review andsummary pertaining to buffer strips.” Final Rep., research subprogramof the National Soil and Water Conservation Program.

Lee, P., Smyth, C., and Boutin, S. �2004�. “Quantitative review of ripar-
ian buffer width guidelines from Canada and the United States.” J.


Environ. Manage., 70�2�, 165–180.Lin, C. Y., Chou, W. C., and Lin, W. T. �2002�. “Modeling the width and

placement of riparian vegetated buffer strips: A case study on theChi-Jia-Wang Stream, Taiwan.” J. Environ. Manage., 66�3�, 269–280.

Lowrance, R., et al. �1997�. “Water quality functions of riparian forestbuffers in Chesapeake Bay Watersheds.” Environ. Manage. (N.Y.),21�5�, 687–712.

Lynch, J. E., Corbett, E., and Mussallem, K. �1985�. “Best managementpractices for controlling nonpoint source pollution of forested water-sheds.” J. Soil and Water Conservation, 40�1�, 164–167

Lyon, S. W., Gérard-Marcant, P., Walter, M. T., and Steenhuis, T. S.�2004�. “Using a topographic index to distribute variable source arearunoff predicted with the SCS-curve number equation.” Hydrolog.Process., 18�15�, 2757–2771.

Lyon, S. W., Lembo, A. J., Walter, M. T., and Steenhuis, T. S. �2006a�.“Defining probability of saturation with indicator kriging on hard andsoft data.” Adv. Water Resour., 29�2�, 181–193.

Lyon, S. W., Seibert, J., Lembo, A. J., Walter, M. T., and Steenhuis, T. S.�2006b�. “Geostatistical investigation into the temporal evolution ofspatial structure in a shallow water table.” Hydrology Earth Syst. Sci.,10, 113–125.

Macdonald, J. S., Macisaac, E. A., and Herunter, H. E. �2003�. “Theeffect of variable-retention riparian buffer zones on water tempera-tures in small headwater streams in sub-boreal forest ecosystems ofBritish columbia.” Can. J. For. Res., 33�8�, 1371–1382.

Magette, W. L., Brinsfield, R. B., Palmer, R. E., and Wood, J. D. �1989�.“Nutrient and sediment removal by vegetated filterstrips.” Trans.ASAE, 32, 663–667.

Mello, C. R. de, Lima, J. M. de, Silva, A. M. da, and Lopes, D. �2003�.“Initial abstraction of small watersheds of ephemeral flood.” RevistaBrasileira de Engenharia Agricola e Ambiental, 7�3�, 494–500.

Montgomery, D. R., and Foufoula-Georgiou, E. �1993�. “Channel net-work source representation using digital elevation models.” Water Re-sour. Res., 29�12�, 3925–3934.

Murray, G. L. D., Edmonds, R. L., and Marra, J. L. �2000�. “Influence ofpartial harvesting on stream temperatures in forests on the westernOlympic Peninsula, Washington.” Northwest Sci., 74�2�, 151–164

Neary, D. G., Bush, P. B., and Michael, J. L. �1993�. “Fate, dissipationand environmental effects of pesticides in southern forests: A reviewof a decade of research progress.” Envir. Toxicol. Chem., 12�3�, 411–428.

Novotny, V. �2003�. Water quality: Diffuse pollution and watershed man-agement, 2nd Ed., Wiley, New York.

Oliver, D. M., Clegg, C. D., Haygarth, P. M., and Heathwaite, A. M.�2005�. “Assessing the potential for pathogen transfer from grasslandsoils to surface waters.” Adv. Agron., 85, 125–180.

O’Louglilin, E. M. �1986�. “Prediction of surface saturation zones innatural catchments by topographic analysis.” Water Resour. Res.,22�5�, 794–804.

Pachepsky, Y. A., Sadeghi, A. M., Bradford, S. A., Shelton, D. R., Guber,A. K., and Dao, T. �2006�. “Transport and fate of manure-bornepathogens: Modeling perspective.” Agric. Water Manage., 86�1–2�,81–92.

Parkyn, S. �2004�. “Review of riparian buffer zone effectiveness.” Tech-nical Paper No. 2004/05, Ministry of Agriculture and Forestry, Well-ington, N.Z.

Peterjohn, W. T., and Correll, D. L. �1984�. “Nutrient dynamics in anagricultural watershed: Observation on the role of a riparian forest.”Ecology, 65�5�, 1466–1475.

Polyakov, V., Fares, A., and Ryder, M. H. �2005�. “Precision riparianbuffers for the control of nonpoint source pollutant loading into sur-face water: A review,” Environ. Impact. Asses. Rev., 13, 129–144.

Qiu, Z., Walter, M. T., and Hall, C. �2007�. “Managing variable sourcepollution in agricultural watersheds.” J. Soil Water Conservat., 62�3�,115–122.

Rabalais, N. N., Turner, R. E., Justic, D., Dortch, Q., and Wiseman, W. J.�1999�. “Characterization of hypoxia: Topic 1.” Rep. for the Inte-grated Assessment on Hypoxia in the Gulf of Mexico, Decision Analy-
sis Series No. 15, NOAA Coastal Ocean Office, Silver Spring, Md.
/ MARCH/APRIL 2009


Schneiderman, E. M., et al. �2007�. “Incorporating variable source areahydrology into curve number based watershed loading functions.” Hy-drobiol. Bull., 21�25�, 3420–3430.

Seibert, J., and McGlynn, B. L. �2007�. “A new triangular multiple flowdirection algorithm for computing upslope areas from gridded digitalelevation models.” Water Resour. Res., 43, W04501.

Sheppard, S. C., Sheppard, M. I., Long, J., Sanipelli, B., and Tait, J.�2006�. “Runoff phosphorus retention in vegetated field margins onflat landscapes.” Can. J. Soil Sci., 85�5�, 871–884.

Skally, C., and Sagor, E. �2001�. “Comparing riparian management zonesto riparian areas in a pilot study area: A mid-term follow-up action inresponse to the riparian/seasonal pond peer reviews from the Minne-sota Forest Resources Council.” Rep. to the Minnesota Forest Re-sources Council, St. Paul, Minn., �http://www.frc.state.mn.us/Info/MFRCdocs/rr1001.pdf� �Jan. 15, 2009�.

Staver, K. W., and Brinsfield, R. B. �2001�. “Agriculture and water qual-ity on the Maryland eastern shore: Where do we go from here?”BioScience, 51�1�, 859–868.

Steenhuis, T. S., Winchell, M., Rossing, J., Zollweg, J. A., and Walter, M.F. �1995�. “SCS runoff equation revisited for variable-source runoffareas.” J. Irrig. Drain. Eng., 121�3�, 234–238.

Tarboton, D. G. �1997�. “A new method for the determination of flowdirections and upslope areas in grid digital elevation models.” WaterResour. Res., 33�2�, 309–319.

Troch, P. A., De Troch, F. P., and Brutsaert, W. �1993�. “Effective watertable depth to describe initial conditions prior to storm rainfall inhumid regions.” Water Resour. Res., 29�2�, 427–434.

Tyrrel, S. F., and Quinton, J. N. �2003�. “Overland flow transport ofpathogens from agricultural land receiving faecal wastes.” J. Appl.Microbiol., 94�s1�, 87–93.

Unc, A., and Goss, M. J. �2004�. “Transport of bacteria from manure andprotection of water resources.” Appl. Soil Ecol., 25�1�, 1–18.

U.S. Army Corps of Engineers �USACE�. �1991�. Buffer strips for ripar-ian zone management, Waltham, Mass.

USDA Forest Service. �1996�. “Riparian forest buffers: Function and de-sign for protection and enhancement of water resources.” PublicationNo. NA-PR-07-91, Washington, D.C.

USDA-Natural Resources Conservation Service �USDA-NRCS�. �1986�.“Technical Release 55: Urban hydrology for small watersheds.” 210-VI-TR-55, USDA-NRCS Engineering Div., Washington, D.C., NTIS�Natonal Technical Information Service�, NTIS # PB87101580, �ftp://ftp.wcc.nrcs.usda.gov/downloads/hydrology�hydraulics/tr55/tr55.pdf�.

USDA-Natural Resources Conservation Service �USDA-NRCS�. �1998�.“Analysis of conservation buffer effectiveness for Toledo Harborproject.” �http://www.glc.org/dredging/soil/toledo/Appx4.pdf� �Jan.15, 2009�.

USDA-Natural Resources Conservation Service �USDA-NRCS�. �2000�.“Conservation buffers to reduce pesticide losses.” �http://www.wsi.nrcs.usda.gov/products/W2Q/pest/docs/newconbuf.pdf� �Jan. 15,2009�.

USDA-Natural Resources Conservation Service �USDA-NRCS�/USEPA.�1997�. Chesapeake Bay riparian forest buffer handbook: A guide forestablishing and maintaining riparian forest buffers, Washington,D.C.

USDA Soil Conservation Service �USDA-SCS�. �1972�. National engi-neering handbook, Pt. 630 Hydrology, Sec. 4, Chap. 10, Washington,D.C.

USEPA. �2005�. “Riparian buffer width, vegetative cover, and nitrogenremoval effectiveness: A review of current science and regulations.”EPA/600/R-05/118, Office of Research and Development, Washington,D.C.

U.S. Geological Survey �USGS�. �2004�. “A science strategy to support



management decisions related to hypoxia in the Northern Gulf ofMexico and excess nutrients in the Mississippi River Basin.” Preparedby the Monitoring, Modeling, and Research Workgroup, MississippiRiver/Gulf of Mexico Watershed Nutrient Task Force, GeologicalSurvey Circular 1270, Washington, D.C.

Van Deventer, J. S. �1992�. “A bibliography of riparian research andmanagement: Fish, wildlife, vegetation, and hydrologic responses tolivestock grazing and other land use activities.” Idaho Riparian Coop-erative, Univ. of Idaho, Moscow, Id.

Walter, M. F., Steenhuis, T. S., and Haith, D. A. �1979�. “Nonpoint sourcepollution control by soil and water conservation practices.” Trans.ASAE, 22, 834–840.

Walter, M. T., Brooks, E. S., Walter, M. F., Steenhuis, T. S., Scott, C. A.,and Boll, J. �2001�. “Evaluation of soluble phosphorus transport frommanure-applied fields under various spreading strategies.” J. SoilWater Conservat., 56�4�, 329–336.

Walter, M. T., Gérard-Merchant, P., Steenhuis, T. S., and Walter, M. F.�2005�. “Closure: A simple estimation of the prevalence of Hortonianflow in New York City’s watersheds.” J. Hydrol. Eng., 10�2�, 169–170.

Walter, M. T., Mehta, V. K., Marrone, A. M., Boll, J., Gerard-Marchant,P., Steenhuis, T. S., and Walter, M. F. �2003�. “Simple estimation ofprevalence of Hortonian flow in New York City watersheds.” J. Hy-drol. Eng., 8�4�, 214–218.

Walter, M. T., and Shaw, S. B. �2005�. “Discussion of Curve numberhydrology in water quality modeling: uses, abuses, and future direc-tions by Garen and Moore.” J. Am. Water Resour. Assoc., 41�6�,1491–1492

Walter, M. T., Steenhuis, T. S., Mehta, V. K., Thongs, D., Zion, M., andSchneiderman, E. �2002�. “Refined conceptualization of TOPMODELfor shallow subsurface flows.” Hydrolog. Process., 16�10�, 2041–2046.

Walter, M. T., and Walter, M. F. �1999�. “The New York City WatershedAgricultural Program �WAP�: A model for comprehensive planningfor water quality and agricultural economic viability.” Water Resour.Impact, 1�5�, 5–8.

Walter, M. T., Walter, M. F., Brooks, E. S., Steenhuis, T. S., Boll, J., andWeiler, K. R. �2000�. “Hydrologically sensitive areas: Variable sourcearea hydrology implications for water quality risk assessment.” J. SoilWater Conservat., 55�3�, :277–284.

Wegner, S. �1999�. “A review of the scientific literature on riparian bufferwidth, extent and vegetation.” Office of Public Service & Outreach,Institute of Ecology, Univ. of Georgia, Athens, Ga.

Wigmosta, M. S., Vail, L. W., and Lettenmaier, D. P. 1994. “A distributedhydrology-vegetation model for complex terrain.” Water Resour. Res.,30�6�, 1665–1679.

Wipple, W., Jr., DiLouie, J. M., and Pytalr, T., Jr. �1981�. “Erosion po-tential of streams in urbanizing areas.” Water Resour. Bull., 17�1�,36–45.

Woodward, D. E., Hawkins, R. H., Jiang, R., Hjelmfelt, A. T., VanMullem, J. A., and Quan, Q. D. �2003�. World Water and Environ-mental Resources Congress, Abstracts of Papers, American Society ofCivil Engineers, Reston, Va.

Welch, D. J. �1991�. “Riparian forest buffers—Functional and design pro-tection and enhancement of water resources.” USDA Forest ServicePublication NA-PR-07-91.

Zhang, Q., Mankin, K. R., and Erickson, L. E. �2000�. “Effectiveness ofvegetative filter strips in reducing nutrients and fecal microorganismsfrom feedlot runoff.” ASAE Annual International Meeting, AmericanSociety of Agricultural Engineers, St Joseph, Mich.



Documents

New Paradigm for Sizing Riparian ... - Dr. M.Todd WalterNew Paradigm for Sizing Riparian Buffers to Reduce Risks of Polluted Storm Water: Practical Synthesis M. Todd Walter, M.ASCE1;