2 4 1

HERBICIDE%PROTECTING LONG-TERMSUSTAINABILITY AND WATER QUALITY

IN FOREST ECOSYSTEMS

DANIEL G. NEARY

USDA Forest Service,2500 Pine Knoll Drive, Flagstaff, AZ 86001, United States

and JERRY L. MICHAEL

USDA Forest Service,Devall Street, Auburn University, AL 36849, United States

ABSTRACT

World-wide, sediment is the major water quality problem. The use of herbicides forcontrollingcompeting vegetation during stand establishment can be benci ic ial to forestecosystem sustainability and water quality by minimising off-site soil loss, reducing on-site soil and organic matter displacement, and preventing deterioration of soil physicalproperties. Sediment losses from sites where competing vegetation is controlled bymechanical methods can be I to 2 orders of magnitude greater than natoral Iosscs fromundisturbed watersheds. On a watershed basis, vegetation management techniques ingeneral incrcaseannualerosion by<7%. Hcrbicidesdonotincreasenaturalcrosio~~rates.Organic matter and nutrients that are critical to long-term site productivity can beremoved off-site by mechanical vegetat ion-managcmcnt techniques and fix, orredistributed on-site in a manner that rcduccs availability to the next stand.

For several decades, research has been conducted on the fate of forcitry-useherbicides in various watersheds throughout the southern and western l lni tcd Stntcs,Canada, andAustrnlia.‘fhisrcscarch hasevaluatedchemicalssuch as2,4-D,glyphosate,hexazinone,imarapyr,mctsulfuronmethy:,picloram,sulfometuronmethyl,tebuthiuron,and triclopyr. Losses in strcamflow, and leaching to groundwater have been evaluated.F i e l d s t u d y d a t a i n d i c a t e t h a t rcsiduc c o n c e n t r a t i o n s t e n d t o b e low. e x c e p t w h e r e d i r e c ta p p l i c a t i o n s a r c m a d e t o ephemeral channels o r s t r e a m s , a n d d o n o t persist f o r e x t e n d e dperiods of time. Regional environmental impact statements in the United Statesdcmonstrntc that forestly herbicide presence in surface and groundwar is not asignificant risk to water quality or human health. They also clearly indicate thatherbicides can greatly reduce water quality deterioration that is produced by erosion 2ndsedimentation.

Keywords: herbicides; sediment:uatcrquality;environment:~iteproductiviry;f~~rcstry:vrgetstion control.

INTRODUCTIONA critical component of inter-rotation forest management is the manipulation of

successional vegetation to ensure adequate survival and growth of the next forest crop.

242 New Zealand Journal of Forestry Science 26( l/2)

Techniques such as manual removal, mechanical control, prescribed tire, and herbicideapplication have been used to reduce competition from undesired vegetation. Herbicideshave been incorporated into vegetation management programmes on intensively managedforests more frequently in the past two dccadcs (USDA Forest Service 1989a, b, 1990).

In many countries with intensive forestly programmes, considerable controversy hasdeveloped concerning the environmental impacts of herbicides. Human health risks ofcommonly used forestry herbicides andothcr vcgetationmanagement techniques have beenaddrcsscd by scvcral intcnsivc cnvironmcntal impact analyses (USDAForest Service 1989a,b. 1990). On- and off-site impacts on water quality continue to bc the subject ofmuch debateand scientif ic analysis (Norris 19X 1; Nearyetal. 1993). Other important parameters ofwaterqual i ty such as scdimcnts and nutr ients have been mostly ignored in the continuing focus onherbicide residues. Indeed, the major water quality problem in arcas with intcnsivc forestmanagement is sediment, not herbicides (Marion & Ursic 1993; Nary & Hombeck 1994).

Another issue relating to forest harvesting, vegetation managcmcnt, and the choice oftechniques for manipulating forest vegetation to enhance productivity, is long-tamsustainability (Dyck & Skiuncr 1990; Powers et ul. 1990). Kimmins (1994) and Nay el al.(1990) identified some of the key processes affecting long-term site productivity. Theseinclude adequate root system development, sufficient soil moisture availability to maintainnutr ient f lux to t ree root systems, suitable supplicsofplant macro- and micro-nutrients in therhizosphere. fully functioning microbiological processes, and adcquatc hydrologicalfunctioning. Some vegetation management techniques can adversely affect site organicmatter rcscrvcs. nutrient pools, and soil physical properties. Improperly used vegetationmanagcmcnt tcchniqucs can cffcctively displace up to five times the amounts of nutrientsremoved in whole-tree harvesting (Ballard 1978; Morris ef al. 19X3). This situation occurswhen organic matterand topsoil are concentrated into small port ions of inter-rotat ion standstrcatcd to control unwanted vcgctation.

Several hypotheses can be formulated about the role of herbicides in forestry. They areconsidered by some sectors of international environmental interest groups to be agents ofenvironmental degradation. Arguments can bc made that they produce “either positivebcncfit nor advcrsc impact. Another hypothesis is that the proper use of herbicides actuallyhasapositivcrolc inprotectingenvironmentalquality. llerbicideusageduringintcr-rotationvegetation management of forest stands can do this by maintaining the commercialsustainabi l i ty of forest ccosystcms and protecting water quali ty. The objective of this paperis to synthcsisc scicntilic information on both forest ecosystem sustainability and waterquali ty, focusing on theroleofherbicidcsinkecpingsoilrcsourccson-sitewithout degradingwater quality.

METHODS AND MATERIALS

Thispaperisasynthesisofmanypublicationsdealing withherbicideresiduefate, erosionandsedimentat ion, vegetationmanagement, soilphysicalconditions, andnutr icnt d is t r ibut ionsin forest ecosystems. Standard hydrological, soil physics and chemistry, erosion, vcgctationmanagcmcnt, and hcrbicidc rcsiduc methodologies were used in the conduct ofthc researchreviewed in this paper. A detailed discussion of the specific methods and materials can bcfound in the references ci ted in this synthesis .

Nary & MichacCHcrbicides protecting forest ecosystems 243

RESULTS AND DISCUSSION

Sustainability

The concept of sustainabil i ty used in this paper only addresses whether a site can supplysufficient water and nutr ients to support successive rotat ions of commercial forcst s tands.Forest ecosystem sustainabil i ty is a much broader concept that encompasses the entirety ofthe fauna and flora and associated ecological process occurring within forests. A narrowdefinition of sustainability was selected because of the forum at which this paper waspresented.

Trees require adequate supplies of nutrients and water to grow, and roots need a well-structured soi l to devclop large enough systems to support that growth (Nary et ul. 1990).So, the keys to long-term sustainabi l i ty are organic matter , nutr ient supply, soi l hydrologicfunct ion. and soi l physical condit ions (Powers era/. 1990). Detrimental changes in the statllsof any of these site characteristics can cause a decl ine in forest productivi ty.

Inmost intensivelymanagedforests, somcformofsitepreparationispractised toimprovemicrosite condition. control competing vegetation, or reduce logging slash to facilitateplanting (Cmtchfield & Martin 1982). However, it may product adverse effects on sitecharacteristics which control productivity. Intense tires can consume much of the residualorganic matter in slash, litter, and the mineral soil. volatilising nitrogen and leaving nutrient-rich ash susceptible to water or wind transport off-site (DeBano & Conrad 197X; Nary rfa/. 1978). Soils left bare by hot fires increase surface run-off and often develop water-repellent horizons, therebymakingsitescrosion-proneanddrier(DeBano 19Xl).Mcchanicalsite preparation can redistribute organic matter, cffcctively removing from seedlings manytimes more nutrients than whole-tree harvesting (Nary er al. 19X4; Balncaves et al. 199 I).Soils arc often left bare and susceptible to surface run-off and erosion. Additional machinerypasses can increase bulk density in susceptible, mainly tine-textured soils, significantlyreducing both rooting volume and available moisture-holding capacity. Herbicides do notproduce the adverse effects associated wi th scverc tire and mechanical site preparation, andthcrcforc work to minimise impacts on s i te product ivi ty and forest sustainabi l i ty (Nary eful. 1990). Hcrbicideapplications tocontrolcompctingvcgetationdo notdisturb thcnutrient-rich litter layer, do not crcatc additional amounts of bare soil, and do not adversely afCcctwatershed condition. Among other things, soils on recently harwsted sites trcatcd withherbicides have better moisture contents due to the reduction of surface run-off and thetranspiration component of evapotranspiration. These soils are bcttcr able to supply thenutrients necdcd for early growth of the succeeding forcst crop (Carter et al . 19X4: Nary eta/. 1990; Smcthurst ef ai. 1993).

Direct evidence of declines in forest stand sustainability due to inter-rotation sitepreparation and vegetation managmcnt is scarce because intensive inter-rotation fores@/management is relat ively rcccnt and there is a lack of good long-term databases (Powers eta/. 1990). Productivity declines of 20% to 30% due to mechanical site preparation (Ballard1978; Swindel ef al. 19X6; Powers et al. 198X; Fox el al. 19X9; Dyck & Skinner 1990) andfire (Keeves 1966; Squire etal. 19X5) have been documented. Increases in the productivityof pine stands after hcrbicidc USC in the south-eastern United States have been noted byMichael (19X0), Knoweet a/. (19X5)_ Nary er al. (lY90), Bramlctt &al. (1991), and Laueret al. (lYY3). Balncavcs et ul. (1991) reported similar results from New Zealand. A few

244 New Zealand Journal of Forestry Science 26( 112)

aspects ofthe sustainabi l i ty quest ion relat ive to organic matter , soi l physical condit ions, andnutrient supply will be discussed here. Two other papers (Powers ef u[. 1995; Powers&Ferrell 19963 cover these topics in more detail.

@~‘,niC IllOffW’ ‘U,d PlLlt,%‘?ll .SUJ’Jl!”

Powers et ul. (I 990) discussed the importance of organic matter to forest productivitythroughitsfunctionofsupplyingnutrien!s,au~entingcationexchangecapacity,improvingsoil stmchue; andchelaringmetal cations.Afterharvesting,themainlossesoforganicmatterresult from decomposition. erosion, oxidation in tires, residue displacement by mechanicalsite preparation. or a reduction in new organic matter recruitment (litterfall, fine rootturnover, etc.). Since organic matter in the forest tloor and surface horizons of the mineralsoil is the major nutrient reservoir in forest ecosystems, especially for nitrogen and boron(McCall & Powers 1984). additional losses during vegetation management arc a conccm.

Fire can have a ma.jor effect on organic matrcr, depending on i ts in tensi ty and durat ion.Organic matter oxidation in tires not only affects nutrient pools, but can &I affect soilmoisture that is critical for ion flux to plant roots. nitrogen fixation, and mycorrhizaldevelopment (Jurgensen r, ui. 1990; Nary ef ul. 1990).

The main effect of mechanical site preparation relat ive to organic matter is displacementin windrows or slash piles, or erosional losses. The impact of site preparation on nitrogenbalances in Coastal Plain and Piedmont s i tes ofthe south-eastern United States is i l lustratedin Table I. Organic matter displacement in windrows can have a major effect on nitrogenbalances. Ballard ( 1978) reported a 16% reduction in the volume of a second-rotat ion Pinu.sradiutu D.Don stand on the central volcanic plateau ofNew Zealand after piling of loggingslash and topsoi l in to windrows. Balneavcs er al. (1991) documented that root raking, pi l ingof slash, and burning in the Nelson district of New Zealand removed greater quantities ofnitrogen and some macronutrients from portions ofP mdiatu sites than harvesting of saw-and chip-log material.

TABLE I-~.Effccts ofsrcm harvest, vegetation management, and fertilirer in the south-castcrn UnitedStates on nitrogen balances (Mgh) after 15 years (from Nary ei al. 1984, 1YXl)*.

Coasral Plain PiedmontSix preparalion Flatwoods Wet Flats Uplands

Chop, Herbicide +O.l23 +0.147 +o.lohChop, Rum +0.063 +0.0x3 +0.045Windrou (careful) 4.222 4058 -0.264Windrow (carcIcss) 4.449 4.674 41.354

* Assumes applicalian al‘O~200 Mg Nib* (itI?% ecosystem rccovc~y~. atmosphciic inpuls (0.005 Mghalycar~, andfixaim ((I.001 Mg~haiymr~: nulrients dispiaccd by uindrowing considcied mavailahlc.

In combinat ion with residual slash chopping, herbicides have the least impact on nitrogenpools after 15 years (Table I). Although herbicides do not directly affect organic matterstatlls. theycanrcsultin somenutrient lossesvia leachingandrun-offbyrcducingtheamountof successional vegetation available to take up nutrients released through mincralisationprocesses. This effecl is visible in increased nitrate-nitrogen losses sometimes measuredafter herbicide applications for vegetation management after harvesting. This topic isdiscussed in mme detai l in the sect ion on water qual i ty .

Neary & Michael-Herbicides prutrcting forest ccosystcms 245

Vegetation management techniques can alter the physical characterist ics of soils and inturn affect both hydrologic function and si te productivity. Fires can leave soils bare of forestfloor material and therefore subject to raindrop impact, run-off, and erosion. In some soilsand vegetation types, hydrophobic layers develop, causing excessive run-off and erosion(DeBano 1981). Mccbanical site preparation can have variable effects on soil physicalconditions (i.e., increasing porosity and infiltrarion talcs in some locations and decreasingthem claewhcre through compaction) depending on soi l texhrcs and moisture, preparationtcchniqucs, the type of equipmenl. and the skills of operators. Vegetation control bymechanical methods usually Ieavcs larger areas of bare soil than harvesting, thus increasingthcamount ofrun-offandcrosion. Compactionarsociatcdwithvehiclc traffic onclayorsilt-textured soi ls , during ei ther harvest ing or site preparation. reduces soil macroporosity. Theresult is reduced rooting volumes and moisture storage capacity. Both affect the ability ofplants to obtain water and nutrients necessary to sustain productivi ty, thus reducing growthin the subsequent rotation. However, on sites with cxtremcly coarse-textured soils theopposite effect can occur. Hcrbicidcs do not increase the amount ofbare soi l . and except for~otnc compaction from ground applicat ion equipment, do not adversely affect soi l physicalproperties.

Large amounts ofthe forest f loor and nutrient-rich surface soil horizons can bc displacedduring mechanical site-preparation. Although this material is not rcmovcd off-site,displacement in to windmws or slash piles can result in net nutr ient losses to 80% or 90% ofthe stand. equivalent to 2 to 5 times that of whole-tree hawesting (Ballard 197X; Morris rtui. 19X3: Tew rf a/. 19X6). In a study of soil and organic matter displacement after bladingand windrowing, Morris of ui. (I 9X3) found that I80 Mg soil and organic matter/ha weredisplaced into the windrows. From 24% (nitrogen) to 64% (phosphorus) of the nutrientreserves remaining on-site were concentrated on to about 5% of the harvested stand’s arcn(Table 2). This type ofnutricnt displacement (loss) is of particular canccm for intensiveplantation-foresttysustainabilityonnutrient-poorsoils. Similarresults have been mcasurcdelscwherc in the southern United States and New Zealand. Early effects on the productivityof the next tree rotation may not be apparent due to compensating mechanisms such as thecontrol of herbaceous weeds, use of new genetic material, improved soil porosity, reducedtranspiration. However, distinct dcclincs of 20% to 30% in the productivity of succccdirtgstands have been documented after intcnsivc mechanical site-preparation involvingwindrowing (Ballard 197X; Swindel rf a/. 1986; Tcw e’1 nl. 1986; Fox rf ui. 19X9; Dyck &Skinner 1990).

‘TABLE Z-Effects of lbawests and windrowing on nutrient pools (MgIlra) in a 40.year-old slash pinef”rest ~from Ncarv el 01. ,984. 19901

Comilonent N P K Cd Me

Ecosystem tut.4 * I.550 0.02x 0.0x0 0.303 0.104.Abovc-ground tree harvest 0.1 IO 0.010 0.036 U.118 0.027stcmon1y harvest 0.066 wtiih 0.022 0.090 0.020Uindnxv 0.373 0.01X 0.027 0. I63 9.041

2 4 6 New Zealand Journal of Forestry Science 26(1/2)

Water QualitySediment

Sedimentat ion, or the erosion and transport of rocks, mincral soil , and organic debris tostreams, has long been the most obvious and important concern in forestry regarding waterquality (Nary & Hombeck 1994). Sediment yields from major river systems range from0.2 Mg/ha/year (Wairau; New Zealand) to 1.2 Mg/ha/year (Columbia; United States), to140.0 Mg/ha/year (Hung Ho; China), and reflect the climate, hydrology, geology, soils,vegetation, physiographic regions, and land-use history of each basin. Natural rates ofsediment yield from smaller, forested watersheds are normally low (CO. 100 Mgihaiyear) butcanvarytre~nendously(uptofiveordersofmagnitude--O’Loughlin&Ziemer 1982). Waterqual i ty in streams emanating from forested watersheds is very important since these streamsare typically used for water supplies throughout the world. In addition, these streams areimportant as habitat and refugia for aquatic biota.

Except during catastrophic mass wasting events, floods, or where bedrock is naturallyhighly erosive ( e.g., Eel River, California; Snake River, Idaho; Waipaoa River, NewZealand), sediment is usually not an important problem in undisturbed forest ecosystems.DebrisavalanchescancauseinajorsedimentproblemsinharvestedforestsofthcPacificRimand other stecplands. These episodic, spectacular events can account for much of thesediment transported off harvested stands, and seriously affect forest resources and valuessuch as water quality, fish habitat, engineering structures, buildings, recreation areas,reservoir capacity; downstream farmland, etc . The loss of soi l s t rength on steep slopes dueto tree root decay 4 to X years af ter cut t ing is usual ly the mechanism predisposing slopes toavalanching (Zicmer 1981). Depending on soil type, geology, climate, and slope, forestharvesting can increase both the erosion rate (factor of four) and frequency of debrisavalanches. but not necessarily the average six (Swanson a al . 19X 1). Road constructionaggravates all debris avalanche hazard factors (erosion rate 120 times that of undisturbedsteepland forests).

However, vegetation management after harvesting generally does notappearto aggravatedebris avalanching or other mass failures except on highly erosive soi ls or unstable geologicformations. In these instances, spot spraying ofherbicidcs rather than broadcast applicationcan reduce mass wasting hazards. Another technique used to reduce erosion after forestharvest ing on highly erosive s tccplands is to oversow wi th grasses orherbaccous species thatcan quickly colonise a sire and stabilise the soils.

Sediment yields from disturbed and undisturbed forest watersheds have been measuredand documented in numerous studies throughout the world (Nary & Hombeck 1994). It isclearly evident that disturbances which create large areas of bare soil, aggravated by highrainfall. unstable geologic formations. erosive soils, and steep terrain, produce the mostsediment yield. Except for some unusual situations with highly erosive, fine-textured soils(Marion & Ursic 1993), erosion losses from harvest-disturbed forested lands usual ly do notapproach those of agriculture (5 to 13 Mdha/year--Larsen et al. 1983). They also do notpersist from the same landscape units. as do sediment losses from agricultural land uscs, ifnormal forest regeneration or n-establishment occurs.

The main impact on water quali ty from inter-rotat ion vegetat ion management is increasedsedimentation (Ncary & Hornbcck 1994). Next to roads and logging skid trails, the ma~jor

Ncay & Michael--Herbicides protecting forest ecosystems 247

source of sediment comes from any ground-disturbing activity. Off-site movement ofsediment frommechanical , burning, and hcrbicidcsitepreparationtechniquesreported in thelitcrarurerangesfrom97 toO.17Mgihalyear.Naturalratesofsedimentlossfromundisturbedforest watersheds are usually 10. I Mgihalycar but in some locations can range up to 0.5 Mg/ha/year. Sediment yields during site preparation are affected by geology, soil, slopes,vegetationandlittercover,andclimate. Theytypicallyareatamaximumduringthcfirstyearafter site preparation, and decline as vegetation recovers on the treated area (up to 4 years).The highest losses have been documented in China (La1 1984). Under intensive high-yieldforest management in the United States , the highest documented losses (14.25 Mg/ha/year)have occurred on si l t - textured soils in the upper coastal plain ofMississippi after cutt ing andbedding. On clay-textured soils in the Piedmont of North Carolina, sediment losses of0.97 Mgihalyear have been reported after mechanical site-preparation (blading andwindrowing) to control compaing vegetation. In New Zealand, maximum sediment yieldsafter clearfelling and site preparation were estimated to be 3.43 Mg/ha/year with skidderlogging and burning with a 20-m riparian buffer, but were much less (0.6 I Mg/‘ha/ycar) withcable logging and burning with no buffer strip (O’Loughlin ef al. 1980).

Sediment and vegetation management-%uthern United StatesIn the southern United States, natural erosion rates from forested watersheds are usually

low at ~0. I I Mgihaiyear) but can range up to 0.22 Mglhaiyear (Maxwell &Nary 1991).However, the distorbances that accompany forest harvesting and site preparation, especiallyroad construction, can cause sediment yields to increase. In some physiographic regions withhighly erosive soils, sediment yields after cutting and site preparation for vegetationmanagement have increased temporarily by as much as 278.fold up to the 9 to I4 Mg/harange (Riekerk er a/. 1989) (Tables 3 and 4).

A comprehensive analysis of sediment production from forests of the southern UnitedStates was conducted by Marion & Ursic (1993). They examined data sets from 37

TABLE )-Effect of forest harvesting and vegetation management on scdimcnt yield, United States.

Location Scdimcntyield

(Mg/ha)

Hubbard BrookNew Hampshire, U S A

hloonshine CreekGeorgia, 1JSA

c~ie”,ion ForcstS. Carolina. USA

Picd,nonfN. Carolina, USA

Gulf Coasthlississippi, USA

Bradford Forts,Florida, USA

Ouachita MountainsArkansas. USA

uncut 0.042Clcarcut 3.650

lAcor 0.067Hcrbicidc, cut 0.170

uncut fl.020Cut’bum 0.151Uncut 0.035Cutibladc 9.730

Uncut 0.620Cdbcd 14.250Uncut 0.003CutGndrow 0.036

uncut 0.071Cut/herbicide 0.251

Hombeck ei RI. (1987)

Neary~tui.(I9XG)

Van Lear er oi. (1985)

Douglass 8: Godrin (I 980)

Beasley (1979)

Rickerk (I 983)

Bcasley er al. (19X6)

2 4 X New Zealand Journal of Forestry Science 26( 112)

TABLE &Effect of forest harvesting and site preparation on sediment yield, Europe, South America,Asia. Australia. New Zealand.

Location Trcatmcnt Scdimcntyield

(MfVhaivcar)

Rcfcrcnce

WalesUnited Kingdom

OxapampaPer”

Hong KongChina

KoolauHawaii , USA

Tawhai ForestNew Zealand

Undisturbed 0.037Drainage 0.090UnCUt 0.121Cut/pasture 0 . 5 4 2

Uncut 2 . 0 0 0Partial cut 6 7 . 0 0 0Clearcut 9 7 . 0 0 0

Uncut 0 . 5 3 6CUtlAg. 2 . 0 9 0

Uncut 0 . 4 2 9Cut, bum* 0 . 6 1 ICut , bum-i 3 . 4 3 2

Francis & Taylor (I 989)

Plamondon ef al. (1991)

Lal(l984)

Do,y~ta/.(l981)

O’Loughlin a al. (I 980)

watershedsranging inareafrom 1 to 2266 haandrepresenting I X9 yearsofrecords (Table 5).Sediment data were transformed to concentrat ions (g/m’) to el iminate variat ions caused byhigh rainfall variability in the region (1000 to 2000 mm). Natural background rates ofindividual watershed average annual sediment concentrations ranged from IX to 106 g/m3(potential range of 0.18 to 2.12 Mgihaliear), and for undisturbed watersheds as a groupaveraged 62 g/m3. Marion & Ursic (1993) concluded that post-harvest vegetation controlwith herbicides did not clevatc scdimcnt losses above natural rates of erosion (Table 6).Burning created a sediment loss problem only in Coastal Uplands. The main source ofsediment from vegetation control techniques in the region originated from soil-disturbingmechanical methods on previously eroded soils or s teep terain. This was part icularly true inthe Piedmont and Coastal Uplands where avcragc annual sediment concentrations frommechanical site-preparation wcrc 17. to 43.fold greater than natural backgroundco”centrations.

TABLE >Summary of average annual sediment concentrations (g/m’) in streamflow from 37walersheds throughout the southern United States (adapted from Marion & Ursic 1993).

LOChOIl Natural Har”eSt Site preparationBurning or Mechanicalherbicides

Interior highlands I6 2 2 5 3 9 1Piedmont 3 7 2 1 2 3 5 1611Coastal Uolands 9 8 IO9 1357 1725Coastal L;wlands II <IO <IO IX

As part of a regional vegetation managcmcnt environmental impact analysis, 27representative watersheds in different National Forests of the southern United States,covering all physiographic regions, were evaluated to determine the effect of vegetationmanagcmcnt on sediment yields (Maxwell & Nary 1991). Within each physiographic

TABLE GCumulativc IO-year sediment yields (Mg) from typical watersheds in the Coastal Plain and Piedmont, southern United States (from USDA Forest 3Service 1989a). &

E:c

SO”hX Brushy 2 Payne Cottonwood Hager Red Buck Nine T W O Indian 2 PdttUSO” zPlOQ Mile Barrel Y

0NT&Z21 1 2 987 I 432 I 542 944 2 55x 736 399 417 9 PO7 I 161 3.3USFS

RoadsHUYWVeg. mgt

PrivateRoadsForestCropsPaStWe

m96 34 13 33 79 15 8 6 4 264 396 $

244 19 209 2s 38 4 4 4 266 32 y1375 24 9 6 X 5x 9 7 8 494 59 R

e385 II 15 83 22 5 I 2 2012 131 z

3 1 7 5 46 25 88 442 6 3 2 8 169 343 2I 333 0 71 47 0 53 0 0 25 354 I 065

61 I 0 659 139 I28 1 7 0 0 7 676 161Total increase 6219 134 1001 486 767 109 23 22 48 223 2 187

Percentage increase 4x 9 65 51 30 1 5 6 5 487 188Veg. mgt percentage increase 3 2 1 7 2 I 2 2 5 5

250 New Zealand Journal of Forestry Science 26(1/2)

region, watersheds representing a range of areas (1781 to 22 096 ha), land types (10 in theCoastal Plain and Piedmont alone), and ownerships (United States Federal Government,State, and private, ranging from 56% to 99% Federal) were analysed. Modelling of sedimentyields over a IO-year period indicated that the cumulative effect qf all land managementactivities(forestry,agriculture,grazing,roadmaintenance,etc.)withinthestudiedwatershedswould bcanelevationofnahral scdimentyields(0.022 to0.134Mgihalyear) by5% to4X7%(Table 6). In forest watersheds with mixed land uses, agriculture usually results in thegreatest increase in sediment yield (487%). Forest harvesting has the potential to increasesediment production l--13% above natural rates.

Current low-intensity, post-harvest vegetation management operations required onUnited States National Forest lands (moderate fire; light mechanical, herbicides, orcombination treatments) can increase sediment loss by another <I% to 7% (Maxwell &Neary 1991). Use of high-impact mechanical vegetation control methods could increasesediment loss on port ions of watershed units by one or two orders of magnitude (Table 7).By comparison, roads (usually the largest and most constant source of sediment) on bothNational Forest and private lands, account for sediment yield increases from 2% to 156% ofthenaturalerosionratc. So, Maxwell&Neary(199I)concludcdthat theimpactofvcgetationmanagement techniques on erosion and sedimentation of water resources isherbicides<fire<mechanical. They also concluded that sediment losses during inter-rotat ionvegetat ion management could be sharply reduced by using hcrbicidcs and moderate burninginstead of mechanical methods and heavy burning.

TABLE 7PEstimates ofsedirncnt loss(Mg/ha/year)by landscape type and vegetation managementtreatment in intensively managedforeslsofthe southern United States(from Maxwell&Nary 1991).

~,Landtype Vegetation management cmsion rate-First year only

Moderate Severe

NZLtUFd Bum or Chop Burn HeavyerOS,on herbicides pile rake/bed disk

A. Coastal PlainRolling uplands 0.045 0.040 0.061 0.303 I.211Upper hills 0.024 0.090 0.134 0.672 2.691Loessuplands 0.134 0.133 0.296 0.999 3.994Flatwoods 0.022 0.010 0.015 0.074 0.296Sand ridges 0.022 0.017 0.025 0.123 0.492

B. PiedmontPiedmont 0.044 0.165 0.247 1.237 4.949

Sediment and vegetation manugement-Western United StatesIn thcforestsoftheu,estemUnited States, tireand herbicides have traditionallybecnused

as post-harvesting vegetation management tools because of the frequency of steep anddissected terrain. Mechanical methods such as chopping, and chaining were once usedextensively for vegetation management on low-gradient terrain. Results of these practicesand of wildfires are summarised in Table X. These natural disturbances aggravate erosionjustaboutanywherein thewestemUnitedStates. Onsomcgeologicallyunstablcterrainwitherosive soils , prescribed t ire can dramatically increase sediment yield, but not to the extent

Nary SC Michael-Herbicides protecting forest ecosystems 251

TABLE 8-Estimates of sediment loss (Mg/ha/year) from vegetat ion management in the wcstemUnited States.

Location Treatment SedimcntyieldControl Treated

Refcrencc

I. vegcrarion managementMontana ClearcutTexas Control bumCalifornia Control burnCalifornia Control burnArizona HerbicideArmma Herbicide

2. WildfireArizona WildfireCalifornia WildfireWashineton Wildfire

<IlOO 0.168<O.OOl 0.028<U.UUl <fl.OUI

0.210 7.3400.019 0.0022.565 2.049

DeBylc & Packer (1972)Wells Ff al. (I 979)Wells (lY79)Dcbano & Conrad (I 976)lngebo & Hibbert (1974)Renardelal.(l991)

2.200 50.500 Hibbar (I Yfi5)5.530 55.300 Kramme; (I 9hO)0.011 2.353 mvey (1980)

that wildfires do (Debano & Conrad 1976). On the whole. l ight control bums and herbicidesdo not accelerate erosion. Fast regrowth of sediment-trapping grasses and herbaccous plantsis usually responsible for this phenomenon (Ingebo & Hibben 1974).

Unlikeorganiccl~emicalsandplantnutrientsoriginatingfr~~mfircorchcn~ical vegetat ioncontrol techniques, physical sediment added to stream systems does not degrade, andbecomes part ofnormal fluvial sediment transportand storageproccsses. Theresidencetimcof th is sediment in fluvial geomorphic systems can range from months to hundreds of years(l&de et al. 1988). The residence time of chemicals is much shorter and their persistencein storage sinks is related to the intrinsic rate of degradation of each chemical.

Sediment losses resulting from inter-rotat ion vegetat ion management affect both on- andoff-siteenvironmentalquality. Mcchanicalsitepreperation, whichproducesthelargestmassof sediment loss; can result in nitrogen and phosphorus losses 20 to 30 timcs the normalannual rate ofundisturbed forest watersheds (Nary rf a/. 1984). While these Iosscs arc lowcomparedtoagriculture-relatednutrientlosses(Larscnetal. 19X3), theydopresentaconcemfor long-term forest management. For example, some forests in the southern United Statesnow under intensive forcstmanagemcnt were highly eroded during abusive agriculture in thelate nineteenth and early twentieth centuries. Because of loss ofnutricnt-rich A horizons,these forests remain sensitive to potential productivity decline unless augmented withfertilisers or vegetation control.

Since herbicide applications do not disturb the forest floor and slash material from theprevious stand, herbicides work to protect water quality and maintain site productivity byretaining nutrient-r ich organic matter and soil surface horizons on-si te. Sediments retainedon-si te do not contr ibute to addi t ional nutr ient loadings or physical deter iorat ion of aquat icecosystems and water resources.

Environmentalfare: A large number of herbicides are used for vegetation managementin forest ecosystems throughout the world, but a dozen account for the majori ty ofthe usage,

252 New Zealand Journal of Foresby Science 26(1/2)

in both frequency and total amounts applied. These herbicides are 2,4-D, 2,4-DP, dicamba,fosamine, glyphosate, hexazinonc, imazapyr, metsulfuron methyl, pi&ram, sulfomcturonmethyl , tcbutbiuron, and t r ic lopyr . This discussion wil l focus on those herbicides which arein wide-spread use for inter-rotation vegetation managment in forest s tands.

Norris (19x1). USDA Forest Service (19X9a), Michael & Nary (1993), Nary el al.(1993), Rashin & Grabcr (1993), and Nary & Hombeck (1994) discussed herbicide fatestudies in North American forest ecosystems. They l is ted numerous s tudies that examinedsampling matrices such as water, soil , and vegetation, and measured peak concentrations insome detail.

Maximum obscrvcd concentrations of hexazinone, imazapyr, sulfometuron methyl,picloram, triclopyr, and 2,4-D measured in strcamflow in a large number of s tudies in NorthAmerica are summarised in Tables 9-l 3. There arc several common features of these data.Firstly, measured peak concentrations were of short duration. Secondly, the highestconcentrations (>I30 mg/mi) occurred whcrc buffer strips were not used or streams wereaccidentally overflown during a herbicide application.

Instantaneous and 24.hour average water quality standards have been recommended bytoxicologists or set by either Canadian or the United States regulatory agencies based onhuman or plant toxicology concerns. A standard process has not been developed for sett ingwater quality standards for herbicides, so some disagreements exist. The most commonlyused instantaneous water quali ty s tandards in the United States are: glyphosate 700 mg/mi.hexazinone 200 mg/m), imazapyr IO 000 mg/&, pi&ram 500 mg/m3, triclopyr ester30 mg/m3, and 2,4-D ester 70 mg/mi. Standards for Canada are currently lower, being 190,280, and 100 mg/m3 for picloram, glyphosate, and 2,4-D, respectively. Except for thoseinstances where buffer strips were not used or streams were overflown, water qualitystandards have not been exceeded by forestry chemical vegetation management operations.

Ofthcnewersilvicultural herbicides (~20 years old), hexazinone has the largest databaseon residues in strcamflow or standing water. There are three instances reported in the

TABLE %Maximum observed hcnarinone residues in streamflow or surface water from treated sitesin North America.

Location Rate Buffer Concentration Reference(kg/ha) (m.g/mJ)

Quebec, Canada 3.6 YCS 15 Legris (1987)Quebec, Canada 3.6 YCS 5 Legris (I 988)Quebec, Canada 3.6 N O 820 Legris (1988)Georgia, USA 1.7 N O 442 Neary da/. (1983)Geoqja, U S A 1.6 Yes 6 Michael &Nary (1993)Georgia, U S A 1.6 Yes 9 Michael & Nary (I 993)Tennessee, U S A 1.7 N O 0 Neary (1983)Arkansas, USA 2.0 Yes 20 Bouchard ef al. (1955)West Virginia, USA I .4 Yes 9 Lavy el ul. (1989)Alabama, USA 2.9 YCS 37 Michael & Nary (1993)Alabama, USA 2.9 Yes 24 Michael &Neary (1993)Alabama, USA 2.9 Yes 23 Michael & Neary (1993)Alabama, USA 2.9 Yes Michael &Neary (1993)Alabama, USA 0.8 N O Miller & Bate (1980)

Nary & Michael-Herbicides protecting Sorcst ecosystems 253

literature where the water quality standard (200 mg/m3) was exceeded during operational

use. The two highest concentrations measured, 2400 and X20 m&m3 (Table 9). occurred

when herbicide pellets were placed directly into adry channel, and when applicationaircraft

overflew surface water. In the third (442 mgim)), hexarinone pellets were distributed

uniformly across small watersheds containing many ephemeral, first-order channels. In all

other instances, hcxarinone residues did nor exceed 37 mg/m3.

lmarapyrandsulfometuronmethylshowasimilarpatremtohexarinone. withthehighest

concentration (imazapyr 680 mg/m’) associated with an aerial application on areas having

no buffer strip (Table IO). A concentration of 130 mg imarapyr/m3, well below the 10 000

m&water quality standard, was measured in Alabama. even with a buffer strip in use,

because of surface run-off. Sulfometuron methyl, which hydrolyses readily in acidic water,

has not been detected above 44 mg/m3 in streamflow.

Clyphosate has been used frequently in forest ecosystems because of its low mobility. It

is readily immobiliscd by organic matter in the forest floor. Most studies (Table I I) have

measured peak glyphosate concentrations in streamflow at or below 10 mg/m’ (more than

an order of magnitude below the 700 mg/m’ water quality standard). As seen with other

herbicide data, the highcst glyphosatc peak concentration (270 mg/m3) occurred whcrc a

buffer strip was no+ used as a Best Management Practice.

TABLE I(tManimumobserved imarapyrandsulfometuronmcthylrcsidues in streamfloworsurfaccwater from treated sites.

I . ImazapyrAlabama. L;SA 2.2 NO 6 8 0Alabama, USA 2.2 Yes 130Washington, USA 0.1 YCS IWashington, USA 0.1 Yes I

2. Sulfometurcn methylR4ississippi. 1JSA 0.4 Yes 2 3Mississippi, USA 0.4 Yes 4 4Florida, lJSA 0.4 Yes 5Florida. LISA 0.4 YCS 7

Michael &Nary (1993)Michael & Nuary (1993)Rashin & Grabcr (1993)Rashin & Grnber (lY93)

Michael & Nary ( 1993)Michael 81 Nary (1993)Neary & Michael (1989)Neaw & Michael (1989)

TABLE 1 I-Maximum”bserved~lyphosatercsiduesinstreamflorxorsurfacewatcrfr”mtreatedsites.

LOCatIOn Rate(ke/hai

Buffer C”ncentrati”n Reference(me/m?

Quebec, CanadaQuebec, CanadaQuebec, CanadaQuebec, CanadaWashingtonWashingtonWashingtonOregon

I.5I.?I.51.51.31.71.3I.3

YesYesYe4YesYtXYesYesNO

5 Legris ff a/. (I 9X5)5 Legris (lY87)

IO Legris (19%)0 Legris &Couture (1989)2 Rashin & Graber (1993)R Rashin & Graber (19Y3)4 Rashin & Graber (1993)

2 7 0 Newton et ni. (1984)

254 New Zealand Journal of Forestry Science 26( l /2)

Maximum measured concentrations ofpicloram, triclopyr, and 2,4-D are l isted in Tables12 and 13. The pattcm in the data is the same as obserwd in the other herbicides, namely thathigh concentrat ions (X&620 mgim)) arc associated with a lack of buffer strips. Otherwise,peak concentrations ofthese three herbicides did not exceed 40 mg/mj. The only exceptionis the picloram concentration of370 mg/m’ reported by Davis & Ingebo (1973). That studyinvolved a very high application rate (10.4 kg/ha) of a pa&tent herbicide in a desertenvironrnentwhichhasalowhcrbicidcrcsiduedegradationrateasaresultofthcaridclimatc.Evenunder these condit ions. the human health water quali ty standard forpicloram (500 mgim3 in the United States) was not exceeded. Some agricultural crops can be affected bypi&ram levels ~10% of that s tandard.

Where buffer strips are used or other mitigatory techniques are employed, forcstlyherbicidcsgcnerallydonotposeathreat towaterquality.Peakconccntrationsareusuallylow(<lo0 mg/m’) and do not persist for long periods of time (6 months) (Michael & Nealy1993).

I ’ABLE I2-Maximum observed picloram and hiclopyr residues in streamflow or surface water fromtreated sites.

Location RateWha)

Buffer Concentration R e f e r e n c e(mglm’)

PicloramGeorgia, USAGmxgia, USAGeorgia. USAKentucky, USAKentucky. USA‘Tennessee, CSAAlabama, USAN.Cnrolina, USAArizona. USAArizona, USA

Triclonvr&rida, USA 2 . 0 YesW. Virginia. lJSA I l . 2 N OBrit. Columbia, Canada 0 . 9 NOOntario. Canada 3 . 9 N O

0 . 3 Yes0 . 3 Yes0 . 3 Yes1 . 3 Yes

0 . 3 YCS0 . 6 Yes5 . 6 N O5 . 0 Yes

1 0 . 4 Yes2 . x N O

0 Michael & Nary (1993)0 Michael&Navy (IYY3)6 Michael & Nary (1993)

21 Michael & Nary (1993)IO Michael & Nary (1993)

4 Michael & Neary (1993)4 4 2 Michael CI a/. (IYXY)

IO Nary ii/ nl. (1985)370 Davis & lngebo (1973)320 Johnscn ( 1980)

2 Nary & Michael (IYXY)8 0 McKel lar el ol. (1982)

6 2 0 Wan (19X7)3 5 0 Thomnsan E, ii/. I 199) /

.TABLE 13-Maximum obscrvcd 2,4-D residues in streamflow or surface water from treated sites.

Location Buffer Concentration R e f e r e n c e(mg/d)

Washington, USA 2 . 1Oregon, USA 2 . 2Oregon, USA 4 . 6Oregon, USA 6 . 7Pacific NW, USA* w

YesN OYesYesYIN

2 Rashin & Graber (1993)1 3 2 Norris (1967)

2 2 Norris et al. (I 982)IO Norris et a/. (1982)

4 0 USDA Forest Service (1984)

* 1 3 3 s e p a r a t e s p r a y i n g s I I7 with no detected residues, I3 with 2.4~” < 5 mgim’, and 2 with 2.4-D of&10 mg/1

+ V”,ri”“s rates w e r e u s e d .

Nary & MichaeCHcrbicides protecting forest ecosystems 255

EL@’ strQ~s: Zones of undistorbed vegetation alongside rip&m areas and other surfacewaters. arc frequently employed as “Best Management Practices” to reduce the impact ofherbicides on aquatic ecosystems. The efficacy of buffer strips in mitigating pesticidetransport into wetlands or riparian zones is quite varied due to the many factors which canaffect pesticide transport (Comerford et al. 1992). Except for the work of Rashin & Grabcr(1993), none of the environmental fate studies summariscd in Tables l&13 was designedto investigate the effects and functions of differing buffer strip sizes. Where buffer stripswere used, other criteria determined the buffer strip size or orientation.

Herbicide chemistry, application rate, distribution method, buffer size. and wcatherconditions are very important in determining how well buffer strips work (Comcrford ef al.1992). In all studies listed in Tables I@-13 where ranking streamflow concentrations werehigh (>I30 mgjm’), no buffer strips were used or the buffer was violated during hcrbicidcapplication. Generally speaking, buffer strips of I5 m or larger are effective in minimisingpesticide residue contamination ofstreamtlow (Nary et al. 1993). The use ofbuffcr stripscan keep hcrbicidc residue concentrations within water quality standards. They are notabsolute and one as large as 140 m did not keep residues out of a perennial stream in NorthCarolina(iiearyrtui. 1985). However, themcasuredpeakconcentrationwas SO times lowerthan the water quali ty standard.

Gvoundwntw: Hcrbicidccontaminationofgroundwaterhasbecomcapriorityenvironmcntalissue in the past few years because of growing incidents of agricultural herbicide residuesbeing detected in well samples. In most rural arcas, rcsidenccs are dependent upongroundwaterforawatcrsupply. Also, significantareasofNorthArnericautilisegroundwaterfor major municipal water sources. A major contamination of an aquifer system would notbe easily reversed because of long residence times of wntcr in aquifer systems. Thus it isimportant to address the issue of potential groundwater pollution from operational use offorestly herbicides.

In general tans. forestry use of herbicides poses a low pollution risk to groundwaterbecause of i ts UYC pattern For instance, herbicide use in forestry is only 10% of agriculturalusage and l ikely to occur only once or twice in rotat ions of 25 to 75 years. Applicatmn ratesare generally low (~2 kg/ha) and animal toxicities arc low. Some of the silvicultumlhcrbicidcs can affect non-target plants at low concentrations (<20 mglm? and could affectthe quality of water for irrigation purposes. Within large watersheds where extensivegroundwater recharge occurs, intensive use of silvilcultural herbicides would affect <S% ofthe area in any one year. The greatcst potential hazard to groundwater comes from storedconccntratcs, not operational application of diluted mixtures.

Regional, confined, groundwater aquifers are not likely to be affected by silviculturalherbicides(Neary 19X5). Surface,uncontinedaquifersinthe immediatevicinityofherbicideapplication zones have the most potential for contamination. It is these aquifers which arcdirectly exposed to leaching of residues from the root zone. Several examples are given.

In Georgia, United States, hexarinone was applied at a rate of I .68 kg/ha to four small(<I hs)first-orderwaIcrsheds(Nearyrta/. 1983). Hcxarinoneconcentrationsingroundwatcrentering perennial stream channels as baseflow were very low (~24 mg/m3). and did notpersist for more than 30 days. Bouchard et al. (1985) reported a very different si tuation withhexarinone applied to an I I .5-ha watershed in Arkansas at 2.0 kg/ha. Hexazinone residues

256 NW Zealand Journal of Forestry Science 26( l/2)

( 14 mg/m3) were consistently measured in groundwater entering perennial stream channelsfor over a year after application. In South Carolina, application of hexarinone at 2.8 kg/hadid not produce any groundwater contamination in sandy soils where the water table rangedfrom 2 to 14 m below surface (Bush el al. 1990). On a Florida site with similar soils anda lower application rate (1.7 kg/ha), hexazinone was detected in groundwater (17 to35 mg/m3), but not until a year later.

Sulfomehmm methyl was applied at a rate of 0.4 kg/ha to 4.ha watersheds in Florida(Neaty&Michaell989). Residuesofthisherbicidedidnotpenetratetoshallowgroundwater(~1 m deep). A structurally similar herbicide, metsulfuron methyl, applied to a similar sitein Florida also did not leach into shallow (<l m) groundwater (Michael & Ncary 1991).

Triclopyr was applied to small watersheds (4 ha) in Florida in both the amine (2.0 kg/ha)and ester (I .6 kg/ha) formulations by ground sprayer. Monitoring of both streamflow andsurface groundwatcr (<l m deep) for 5 months following application did not detect anyresidues of triclopyr (Bush et al. 1988). Application of picloram (5.0 kg/ha) to steepwatersheds of the Appalachian Mountains produced ephemeral groundwater contamination(Nary et al. 1985). A 140-m buffer strip between the application area and a first-orderperennial stream reduced picloram concentrations in baseflow down to sporadic peaks of<10m~m3duringa l7-monthmonitoringperiod.Intensivesampllngofasp~ngimmediatclybelow the picloram-treated area measured only trace concentrations.

Theonlyknowngroundwatercontaminationincidentsofanyimportance(contaminationof bedrock aquifers, persistence >6 months, concentrations in excess of the water qualitystandard,etc.)in thesouthcmUnitedStates, wheresignificantamountsofforcstryherbicidesare used, involved (I) use of extremely high rates, or (2) spills of concentrates. Because ofthe high concentrat ions in these instances, herbicide residues were detected in groundwater4 to 5 years after the contamination. These si tuations arc defini tely not typical of operationaluse of forestry herbicides. Proper handling precautions during herbicide transport , storage,mixing-loading, and clean-up are extremely important for preventing groundwatercontaminat ion.

Water qua&-Nutrients

Any disturbance to a forest ecosystem (tire, insects, windthrow, harvesting) can alter theequilibrium in biogeochemical cycling, and ultimately produce changes in surface andgroundwater quality. Nitrogen is the element most sensitive to biogeochemical changessince i t can be volat i l isedby t i re and mineral isedby decomposi t ion into highly mobile forms(nitrate-nitrogen; NO,.N). Vitousek & Mellilo (1979) examined the patterns and processesof ni trate-ni trogen losses from disturbed forest ecosystems throughout the world. The onlyinstances where nitrate-nitrogen levels exceeded the 10 mg/m3 water quality standardinvolved additions of herbicides or use of other techniques to inhibit the rcgmwth ofvegetat ion.

A representative range of nitrate-nitrogen peak concentrations in streamflow is given inTables l&16. As indicated byVitousek&Mellilo (1979), all the studies listedin these tableswi th streamflow concentrations >5.3 mg/m) involved herbicides. In the studies where peakconcentrations exceeded the water quality standard, either repeated applications were usedor rates of application were high. Most operational applications of forestry herbicides

Nary & MichaeLHerbicides prutecting forest ecosystems 251

TABLE 14-Effectofvegetationmanagement”nmaximumnitrate-nitrogenconcennationsinstreamfl”w(“astern United States).

Location

Hubbard BrookNew Hampshire, USA

Hubbard BrookNew Hampshire, USA

Fernow ForcstW. Virginia, USA

Coweera LabN. Carolina, USA

Coweeta LabN. Carolina, USA

Moonshine CreekGeorgia, USA

FOlCSfQPe

Northcmhardwoods

Northernhardwoods

Mixedhardwoods

Convert tograss

Mixedhardwoods

Hardwoods

Treatment Maximum ReferenceNO,-N

bwl)

Cd

Cutherbicide

Cut

Cuthcrbicidc

Cut

Herbic ide

6.1 Hornbeck ez al. (1987)

17.x* Pierce el al. (1970)

I .4 Aubenin & Patric (I 974)

0.7 Swank (19%)

0.2 Swank (I’)@)

5.3 Nary el ul. (1986)

TABLE IS-Effectofvegetationmanagement”nmaximumnitrate-nitrogenc”ncentrationsinstreamflow(western United States).

L”C&l@”

Andrew ForestOregon, USA

Alsea BasinOregon, USA

Three BarArizona, USA

Three BarArizona, USA

Forest

type

Douglas-fir

Douglas-fir

Chaparmlgrass

Chaparralgrass

Treatment Maximum RcferenccNO,-N

__-Cut

burnCIA

Herbic ide

Herbic idebum

b&0.6

2.1

15.3

I x.4

Fredricksen et al. (1975)

Brown et al. (1973)

Davis (I 984)

DdViS(1Y87)

involve relatively low rates and are not likely t” be repeated in successive years, Total rnzw

losses of nutrients from watersheds in streamflow are not usually large relative to other

processes (Nay & Hombeck 1994). Therefore, the impact of additional nitrogen lossesfrom herbicide use is minimal.

CONCLUSIONSInseveraldccadesofresearchonthefateandenvironmentaleffcctsofherbicidesonforest

watersheds, suffkient progress has been made t” support several regional environmentalimpact statements (USDA Forest Service 1989a, b, 1990). Additional research will benecessary in the next decade to examine the environmental fate ofnew pesticides as well asdetermine indirect effects and cumulative effects of forestIy herbicide use.

Numerous research and monitoring studies have documented low concentrations andshort persistence of forestly herbicides in surface waters. In the southern United States,

2 5 8 New Zealand Journal of Forestly Science 26(1/2)

TABLE I~Effectofvegetationmanagementonmaximum nitrate-nitrogcnconcentrationsinstreamflow(Canada, Eurooc, New Zealand)

Location ForesttYPc

Narrows Mtn HardwoodsN.B., Canada & conifers

Haney WcstemB.C., Canada hemlock

Okanagan SprKe-flrB.C., Canada

TOt.%laSen Spruce,NOWZiY alda

Tawhai Forcst Beech-New Zealand podocarp

Tawhai Forest Beech-New Zealand podocarp~. .~~.~~~f Post-hurl,

Cutbum

Cut

Maximum ReferenceNO,-N(wd

1.6 Krause (1982)

0.5 Feller & Kimmons (1984)

0.4 Herhcrington (I 976)

9.1 ogner (I 993)

0.2 Ncary el a/. (1978)

0.4* Neary n a/. ( 1978)

applications ofhexazinonc. imazapyr, metsulfumn methyl, picloram, sulfometuron methyl,and triclopyr at rates of 0.3 to 5.6 kg/ha produced peak strcam concentrations <I30 mg/nSwhen buffer strips were maintained (Michael & Nary 1993; Nay a al. 1993). Aerialapplicat ions to ent ire watersheds in both the United States and Canada have resulted in peakstreamflow concentrations in the 442480 mgim) range where buffer strips were not usedor maintained. Higher concentrations (up to 2400 mgi&) have been reported in shortsectionsofstreamsafteraccidcntaloverflights.These~esofpeekstreamflowconccntrationsdo not persist and rapidly attenuate. Although water quality standards do not exist for allforestryherbicidcsorthestandardsarrunderdebate,monitoringexperienceclcarlyindic~atesthat the rates and use patterns of these chemicals do not pose any problem for surface waterquality For instance, the suggested water quality standard for hcxazinonc has only beenexceeded for a short time where ephemeral or pcrcnnial channels were treated. Whereforestly herbicides have been detected in streamflow, the residues usually dissipate withina few months, and persist mainly in low concentrations (~44 mg/&).

Forestry herbicides have been detected in shallow, surficial groundwater (unconfinedaquifer of soil, colluvium, or saprolite) only from broadcast applications and then only inabout halfthe srudies that monitored for them. lnnone ofthese si tuat ions were the herbicideresidue concentrations ofany toxicological significance. No cases exist ofa bedrock aquiferbeingcontaminatedonlocalisedorlandscapescalesbyoperationaluseofforestryherbicides.Transport and storage of concentrated herbicide products are the only activit ies with any riskfor localised contamination of major aquifers.

From both the water quality and sustainability perspectives, herbicides have a realadvantage for stand establishment and inter-rotation vegetation management. By keepingsoil on s i te and not in s treams, long-term forest sustainabil i ty is protected and water qual i tyis not adversely affected. Considerable research and monitoring studies have shown thatoperational use of forestry herbicides for inter-rotation vegetation management does notcreate a significant risk to water quality as far as herbicide residues arc concerned.

Nary & Michael-Herbicides protecting forest ecosystems 259

However? when the scientific evidence of risks and benefits is carefully analysed,

herbicides actually have a positive role in protecting environmental quality. They do this by

maintaining the sustainability of forest ecosystems and protecting water quality.

REFERENCESAUBERTIN, GM.; PATRIC, J.H. 1974: Water quality aftcrclcarcutting a small watershed in West

Virginia. Journal ~/Em~imnmen~a/ Qua/i+ 3: 243-9.

BALLARD; R. I Y7R: Effect of slash and soil removal on the productivity of second rotation radiatapine on a pumice soil. Neu, Zeoland.Ioun~al qfForesty Science II: 24X-58.

BALNEAVES, J.M.: SKINNER, M.F.; LOWE, A:I. 1991: Improving the rc-establishmentofradiatapincon imp~~verishrdsoilsinNelson,NcwLcaland. Pp.137-StIinDyck, W.J.; Mees,C.A.(Ed.)“Long-term Field Trials to Assess Environmental Impacts of Harvesting”. Proceedings, IEA/BET61A6 Workshop, Florida, USA.February 1990. IEA/BET6iA6ReportNo.i. ,VewZrulirndMinimL~ “fF”,rsrr): FRI Bu,le,in ,N”. 161.

BEASLEY, R.S. 1979: Intensive site prcparatiun and sediment loss on steep watersheds in the GulfCoastal Plain. Soii Science Sot@ <$,fAmerico Jvurnn14.7: 4 I2~-7.

BEASLEY, R.S.; GRANII.LO, A.B.; ZILLMER, V. ,986: Sediment losses from forest management:mechanical vs. chemical site preparation after cutting. ./oumul q/Environmentai Qua1it.b IS:413-h.

BOUCHARD, D.C.; LAVY, J.L.; LAWSON; E.R. 1985: Mobility and persistence of henazinone ma forested watershed. Journal o/‘Envi,anmentai Quo1il.v 14: 229-33.

BRAMLETT, D.L.; JONES, E.P.: WADE; D.C. 1991: Herbicide and bum site preparation in theGeorgia Piedmont. Pp.13X-46 in Coleman, S.S.; Nary, D.G. (Ed.) Proceedings of the SixthBiennial Southern Silviculturc Research Conference, Memphis, TN, 30 October- I November1990. Sou~hra.~trm Foresr Experiment Srarion General Technical Repon SE-70.

BROWN. G.W.; GAHLER, A.R.: MARSTON, R.B. 1973: Nutrient lossesaftcrclcarcut logging andslash burning in the Oregon Coast IRange. Wulrr Re.wurcr.~ Researc;h Y: 145&3.

BLSH, P.B.; NEARY. D.C.: TAYLOR, J.W. 19X8: Effect ofhiclopyr amine and ester fomlulationson groundwatcr and surface runoff water quality in the Coastal Plain. Pmceedin~s of IkeSourhtwn Wwd Science Socirty 3% 262-70.

BUSH, P.B.; MICHAEL. J.L.; NEARY, D.G.; MILLER, K.V. 1990: Effect of hexazinonc “ngroundwaterquality inthcCuastal Plain. P,vcredin~~ofrkeSoufhrrn W~edScienceSocie~43:I X4-‘)4.

CARTER, (;.A.; MILLER, J.H.; DAVIS, D.E.; PATTERSON, R.M. 1984: Effect of vegetativecompetition on the moistwc and nutrient status of lobtoltv pint. Canadian Journal ofFore.sfR~&rck 14: I -Y.

COMERFORD, N.B.; NEARY, D.G.; MANSELL, R.S. IYYZ: The effcctiveneessofbuffer strips foramel!oratineoffsitctrans~ortofsediment,nutrients,andpesticidesfrom silvicultural operatmns.,Norionul C&nci/oftke &per lndust,yfi,r AirandSrream hnprovemen~. N~MJ York, Technicrdilul lrr in No. 63 1.

CRUTCHFIELD, D.M.;MARTIN,J.P. 19Y2:Sitepreparatio~oastalPlain.Pp.4P-57inProceedingsof Loblolly Pine Ecosystem (East Region) Symposium, North Carolina State University andUSDA Forest Service Southeastern Forest Experiment Station.

DAVIS, EA. 1984: Conversion of Arizona chaparral increases water yield and nitrate loss. WaterRrxwws Ruenrch 23: 2 I >24.

-1987: Chaparral conversion to increase streamflow in Arizona: sequential treatments extendduration of nitrate loss to stream water. Forest Science 33: 8%I 0 3 .

DAVIS,E.A.;INGEBO,P.A. 1973: Piclorammovementfromachaparralwatershed. WalerResourcesReenrch 9: 1304-I 3.

DeBANO, L.F. 1981: Waterrepellentwils: astate-of-the-art. USDA ForesrService, Pac@cSouthwestFwesr and Range Experimem Starion General Technical Report PSW-46.

2hfl New Zealand Journal of Forestry Science 26( I/2)

DeBANO, L.F.; CONRAD, C.E. 1976: Nutrients lost in debris and runoff water from a burnedchaparral watershed. Pp.3.13 ro 3-27 in Proceedings of the 3rd Interagency SedimentationConference, Denver, CO, March.

-I 97X: The effect of fire on nutrients in B chaparral ecosystem. Ecolo,q 59: 48%97.DEBYLE, N.V.; PACKER, P.E. 1972: Plant nutrient and soil losses in overland flow from burned

forestclcarcuts. Pp,29~~307inProceedings,NationalSympociumon”WatershedsinTransition”,Fort Collins, Colorado, I !&22 June. American Water Resources Association, Urbana, Illinois.

DOl’Y, R.D.; WOOD, H.B.; MERRIAM, R.A. 1981: Suspended sediment production from forestedwatersheds on Oahu, Hawaii. Wu~er,erKrsouwr~ Bullerin 17: 399-405.

DOUGLASS. J.E.; GODWIN, R.C. 1980: Runoffand soil erosion from site preparation practices. Pp.51-73 in “U.S. Forestly and Water Quality: What Course in the X0’s’? Richmond, Virginia,Water Pollution Control Federation. Washington, D.C.

DYCK, W.J.; SKINNER, M.F. 1990: Potential for productivity decline in New Zealand rndinta pintfores& Pp.3 I %32 in “Sustained Productivity of Forest Sods”. Proceedings of the 7th NorthAmerican Forest Soils Conference, Vancouver, British Columbia, July 19x8. University ofBritish Columbia, Vancouver. 524 p.

FF.LLER, M.C.: KIMMONS, J.P. 1984: Effects of clearcutting and slash burning on streamwaterchemistry and watershed notricnf budgets in southwestern British Columbia. Waler ResourwsRrsenrdz Xl 2WO.

FOX,T.II.; MORRIS, L.A.; MAIMONE, R.A. 1989: Windrowing reduces growth in B loblolly pineplantation in the North Carolina Piedmont. Pp. 133-Y in Miller, J.H. (Camp.) ProceedingsoftheFifthBiennialSouthemSilvicultureResearchConfcrcnce,Memphis,TN, I-3Novembcr 19%.USDA Forest Servicr. Sourhem Forerr Experbnml Station, rVew Or/ram, Genwal TedznimlRep,-, x-74.

FRANCIS, 1,s.; TAYLOR, J.A. 1989: The effect of forestry drainage operations on upland sedimentyields: a study of two peat-covered catchments. Earlh Surfre Processes und Lon@+ms /4:7.%Yi.

FREDRICKSEN,R.L.;MOORE,D.G.;NORRIS,L.A. 1975:Theimpactoftimberharvest,feltili~atlon,and herbicide treatments on stwrnwatcr quality in western Oregon and Washington. Pp.283314 i,z Bcmier, B.; Winget, C.H. (Ed,) “Forest Soils and Fores1 Land Management”. Lava1University Press, Quebec, Canada. 675 p.

HEEDE, B.H.; HARVEY, M.D.; LAIRD, J.R. 198X: Sediment delivery linkages in a chaparralwatershed following a wildfirc. Environmenral Management IZ: 349-58.

HELVEY. J.D. 1980: Effects of a north central Washington wildfire on runoff and sedimentproduction. Warer Rrsourcrr &&in 16: 627-34.

HETHERINGTON, E.D. 1976: “Dennis Creek: A Look at Water Quality Following Logging in theOkanagan Basin”. Environment Canada Forest Service. 2X p.

HIBBERT,A.R. 1985: Storm runoffandsedirnentproductionafter wildiireinchapanal. Pp.3142 in“Hydrology and Water Resources in Arizona and the Southwest”, Proceedings of the 1985Meeting of the Arizona-Nevada Academy of Science, Las Vegas, NV, 21 April.

HORNBECK, J.W.; MARTIN, C.W.; PIERCE, R.S.; BORMANN, F. H.; LIKENS, G.E.; EATON,J.S. 19X7: The northern hardwood forest ecosystem: IO years of recovery from clcarcuttmg.USDA Fore.~~Servi~e, Nonhenstem Fwesr ExperimenlSra~ion, Broomall, Research Paper NE-596. 30 p.

INGEBO, P.A.; HIBBERT, A.R. 1974: Runoff and erosion after brush suppression on the naturaldrainage watersheds in central Arizona. USDA Foral Servvice Research Nole RM-275. 6 p.

JOHNSEN, T.N. Jr 1980: Picloram in water and soil from a semiaid pinyon-juniper watershed..Im,,nal of Environmenrul Qua/r& 9: 601-95.

IURGENSEN, M.F.; HARVEY, A.E.; GRAHAM, RX.; LARSEN, M.J.; TONN, J.R.; PAGE-DUMBROSE, D.S. 1990: Soil organic matter, timber harvesting, and forest productivity in theinland northwest. Pp. 392415 in Gcsscl, S.A. (Ed.) Proceedings of rhe Seventh North

Neary & Michael-Herbicides protecting forcst ecosystems 261

Amzrica~~ForcstSoilsConfcreoce,Vancouver, B.C.,July 1988. UniversityofBritisllColumbia,Vanc”u”cr.

KEEVES, A. 1966: Some evidence of loss of productivity with successive rotations of Pinu radiarain the south-east of Australia. Austrnl ian Forestq~ 30: 51-63.

KIMMINS, J.P. 1994: Forest manipulation impacts on forest ecosystem sustainability. Pp.23-31 inMahcndrappa, M.K.; Simpson,C.M.; Smith,C.T. (Ed.) Proceedingsofthe IEAIBE Workshop,Fredrickton, N.B., l&22 May 1993. Canadiun ForertService- Mari~imer Region, /n/urmarionReport M-.r-IYIE. 9 I p.

KNOWE, S.A.; NELSON, L.R.; GJERSTAD, D.H.; ZUTTER, B.R.: CLOVER, G.R.; MINOGUE,P.J.; DUKES, J.H. Jr 1985: Four-year growth anddevelopmentofplantedloblollypine on siteswith compcririon control. Sourirern .Journnl of.4ppl;ed Foresrg, 9: I 1-I 4.

KRAMMES, J.S. 1960: Erosion from mountain side slopes after fire in southern California. USDAForesr S&vice. Poci/~c Sourhw~sl Fores1 and Range E,y;pe~immt Starion, Re~krley, RmenrchZ’orePSW-,7/. 8p.

KRAUSE, H.H. 1982: Nitrate formation and movement before and after clear-cutting ofa monitoredwatershed in central New Brunswick, Canada. CnnrrdianJournolo/Foresr Research 12: 922-30.

LAL, R. I!IXJ: Soil erosiw from tropical arable lands sod itscor~hol.Ad~unce.~inAgrorron~y27: IX>248.

LARSEN, V.E.; PIERCE, F.J.; DOWDY, R.H. 1983: The threat of soil erosion to long-term cropproduction. Science 2iy: 45845.

LAUER, D.K.; CLOVER, G.R.; GJERSTAD, D.H. 1993: Comparison of duration and method ofherbaccous weed control on loblolly pine response through midrotation. Canadian Jourml g/Fores, Re.rearch 23: 21 1625.

LAVY, T.L.; MATTICE, J.D.; KOCHENDERFER, J.N. 1989: Henazinone persistence and mobilityin a steep for&cd watershed. ./ournal ofEnvironmental Qual iy IX: 507-14.

LEGRIS, J. 1987: Concentrations residuelies dc glyphosate dam I’cau de surface cn milieu forestiersuiteida pulverisations terr+res, 1985. Gouvcmement dcQuebec, MinistCrede I’&ergiectda Ressourccs, Service des Etudes Environmentales, Progress Report. 34 p.

-1988: Projets de suivi environmental des pulverisations de phytocides cn milieu fo%ticrsur lesterrcrpubliqucsQuebecojsrsen 1988. GouvernementdeQucbcc, Ministhedel’EnergicetdesRcssourccs, Service dcs Etudes Environmentales, Progress Report. I6 p.

LEGRIS, J.; COUTURE, G. 1989: Residus de glyphosate dam I’cau et les sediments suite a desul+ations terxstres en milieu forcstier, en 1986. Gowernement de Quebec. Minisrty de/‘Ew~ir rr des Ressource.~. Service da Etudes Enviwnmenra~rs, Puhlica~ion 3322. 26 p.

LECRIS, J.; COUTRE, G.; LABERGE, L.; MAMARBACHI, G. 1985: Concentrationsresiduelles deglyphosate dans I’cau surface en milieu fore+ suite 8. des pulverisations terrestres, 1985.Gouve~~emmr de Quebec, Mininrs/Pre de l ’E,xy~~e ez de.7 Ressources, Service da EruderEnvironmenlale.~, Publ icnt ion 3 3 1 5 . 3 5 p.

MARION, D.A.; URSIC, S.J. 1993: SedimentproductioninforestsoftheCoastal Plain, Piedmont, andInterior Highlands. Pp. I%228 in Proceedings of a Technical Workshop on Sediments, TerrencInstitute, Washington, D.C. I41 p.

MAXWELL, J.R.; NEARY, D.G. 1991: Vegetation management effects on sediment yields. Pp.1 2-55 to 12-63 inSho”-Shou,T.;Yung-Huang,K. (Ed.)Proceedingsofthe5th Federal lnteragcncySedimentation Conference, Volume 2, 18-2 I March, Las Vegas, NV. Federal Energy RegulatoryCommission, Washington, D.C.

McCOLL, J.G.; POWERS, R.F. 1984: Consequences of forest managment on soil-tree relationships.Pp.37W12 NIBowen, G.D.; Nambiar, E.K.S. (Ed.)“NutritionofPlantation Forests”. AcademicPress, NY.

McKELLAR, R.L.; SCHUBERT, O.E.; BYRD, B.C.; STEVENS, L.P.; NORTON, E.J. 1982: Aerialapplication of GARLON 3a to a West Virginia watershed. Down To Earth 38(2): l%l9.

262 New Zealand Journal of Forestry Science 26(1/2)

MICHAEL, J.L. IYXO: Long-term impact of aerial application of 2,4,5-T to longlcaf pine (Pinuspuluslris). Weed Science 28(&i: 255-75.

MICHAEL, J.L.: NEAR?‘, D.G. 1991: Mefsulfuron in surface groundwafer of a norih Floridatlawoods. Proreedings of the Southern Weed Science Sociery 44: 244.

~.-I 993: Herhicidc dissipation studies in southern forest ecosystems. Ewironmenmi To~xicolofi,rind ChmistgJ 12: 405-i 0.

MICHAEL. J.L.; NEARY, D.G.; WELLS, M.J.M. 1989: Picloram movement in soil solution andstreamihv frum a Coastal Plain forest. .,“urno, of Environrnen/ol Q&r/~ 18: X%95.

MIL.ER, J.H.; BACE, A.C. 1980: Streamwatcr contamination after aerial applicarion ofpel let iredherbicide. l/S/IA For-at Swvi~~. Sourhem Forest Expwimenr Station. Re.wnrch No@ SO-25.5.

4 P-MORRIS, L.A.: PRITCHETT, W.L.; SWINDEL, B.F. 1983: Displacement ofnutricnts in windrows

d u r i n g s i t e preparation o f a f l a t w o o d s f o r e s t . S o i l S c i e n c e Swie! l ’ ofAmwica ./ournnl4 7 : 59 I-4 .

NEARY, D.G. 19X3: Monitoring herhicidc residues in springflow after an operational application ofhexazinonc. Southern .Joumal @App,i,iird Fo,rso~ 7: 21 l-23.

-I 985: Fate of pesticides in Florida’s forests: An ovcwiew of potential impacts on water quality.Soii and Crop Science Socir/y of F/on/la Proceedings 44: I l t23.

NEARY, D.G.; HORNBE<:K. J.W. 1994: Chapter 4: Impacts ofharvesting and associated practicesonoff-siteenvironmenralquality. Pp.Xl-I IXin Dyck, W.J.; Cole, D.W.;Comerford,N.B.(Ed.)“Impacts of Forest Harvesting on Long-Term Site Productivity”. Chapman & Hall, London

NEARY, D.G.; MICHAEL, J.L. 1989: Effect of sulfomcturon methyl on ground water and streamquality in coalal plain forest watersheds. Water Resources Bul ir l in 25: 617-23.

NEARY, D.G.; BUSH, P.B.; UOUGLASS, J.E. 1983: Offsitemovcmcnl ofhenazinone in stormflowand haseflow from forest watersheds. WeedScience il: 543,.5 I.

NEARY,D.G.; BtiSH,P.B.;GRANT.M.A. 1986: Watcrqualityofcphemeralforcsrstreamsaftersitepreparation with the herbicide hexarinone. Forat Ecoiog~j and Mmagemmr 4: 2.?-40.

NEARY, D.G.; BUSH, P.B.; MICHAEL, J.L. 1993: Fate, dissipation. and environmental effects ofpesticides in southcm foreifs: B review of a decade of progress. Envimnmenral Torimio~ andChemis,ry i: 41 1-25.

NEARY, D.G.; MORRIS, L.A.: SWINDEL, B.F. 1984: Site preparation and nutrient management insouthern pine forests. Pp.12144 in Stone, E.L. (Ed.) “Forest Soils and Treatment Impacts”,Proceedings ofthe 6th North American Forest SoilsConference, University ofTcnnessee, June1983.454 p.

NEARY, D.C.; BUSH, P.B.: UOUCLASS, J.E.; TODD, R.L 19X5: Picloram mowment in anAppalachian hardwood forest watershed. . /oumal ~/Environmental Qualir~~ 14: 58>92

NEARY, I).G.; PEARCE, A.J.: O’LOUGHLIN, CL.; ROWE, L.K. 1978: Managcmcnt impacts onnutrient fluxes in beech-podocarp hardwood forests. ,Qw ZealandJournul o/Ecology I: 1%26.

NEARY, D.G.; JOKELA, E.J.; COMERFORD, N.B.; COLBERT, S.R.; COOKSEY, T.E. 1990:Understanding competition for soil nutrients-the key to site productivity on southcastemCoastal Plain spodosols. Pp.432-50 in Gessel, S.A. (Ed.) “Sustained Productivity of ForestSoils”. Proceedings ofthe 7th North American Forest Soils Conference, Vancouver, B.C., July1988. University of British Columbia, Vancouver. 524 p.

NEWTON, M. : HOWARD, K.M. ; KELPAS, B .R. : UANHAUS, R. ; LOTTMAN, CM.;DUBELMAN, S.1984: FateofglyphosateinanOrcgon Forestecosystem.Jou~nulo/AgricuNuraland Food Chemisr~ 37: 1144-51.

NORRIS, L.A. 1967: Chemical brush control and herbicide residues in the forest environment.Pp.l0.%23 in “Herbicides and Vegetation Management”, Oregon State University Press,Corvallis, OR.

-19X1: The movemenf, persistence, and fate of phenoxy herbicides and TCDD in the forest.Residue Reviews RO: f&135.

Nary & Michael-Herbicides protecting forest ecosystems 2 6 3

NORRIS, L.A.; MONTGOMERY, M.L. 1975: Dicamba residues in sfreams after forest spraying.Bullerin o/Em~ironmenlal Conlaminnl ion and Toxicolo~ 13: 14.

NORRIS,L.A.;MONTGOMERY,M.L.; WARREN, L.E.;MOSHER,W.D. 1982: Brushcontrolwithherbicides on hill pasture sites in southern Oregon. Journal o/Ran~e Manogemenr 35: 75-80.

OtiNER, G. 1993: The effect ofglyphosate application on brook water qualiry in a nutrient-rich forest.hleddrlel.wJka Skogti,r.sk 4613): I-14.

O‘LOUGHLIN, C.L.: ZIEMER, R.R. 1982: The importance ofroot strength and dctcrioration ratesupon edaphic stability in stecpland forests. Pp.7&X in Waring, R.H. (Ed.) Proceedings ofIIJFRO Workshop P.I.O7-00 Ecology of Subalpine Zones, 2-3 August, Corvallis, Oregon.

O’LOUGf~ILlN, C.L.; ROWE, L.K.; PEARCE, A.J. 1980: Sediment yield and water quality responsesto clearfelling of cvergrccn mixed forests in westcm New Zealand. Pp.285-92 in Proceedingsofthe Helsinki Symposium, “The Influence ofMan on the Hydrological Regime With SpecialRcfcrence to Representative and Experimental Basins”, June, Intemotionui As.svciotion q/&lm/r~~icu/ Science. Genthrugge, Belgium, Puhi iar ion 130.

PIERCE, KS.; HORNBECK, J.W.; LIKENS,G.E.: BORMANN, F.H. 1970: Effect ofelimination ofvegetation on stream water quality and quantity. lnlewznalionnl Assoeintion of Hydmlologicai~S’hwr~ Puhlrcarion 96: 3,1-2X.

PLAMONDON, A.P.; RUIZ, R.A.; MORALES, C.F.; GONZALEZ, M.C. 1991: Influence ofprotection forests on soil and water conservation (Onapampa, Peru). Forest Ecolof iy and.knagemenr 3ti: 227-38.

POWERS, R.F.; FERRELL, G.T. 1996: Moisture, nutrient, and insect constrain& on plantationgrowlh: The “Garden of Eden” study. New 7.ealandJoumal o/Foresrry Srirnce 26/l/2): I22-44.

POWERS, R.F.; TIARKS, A.E.; BUFORD, M.A. 1995: Influence of site disturbance on fundamentalproductivity and soil processes in pine plantations of the LISThe “LTSP” Expcrimcnt. Pp.68-70 in Gaskin, R.E.; Zabkiewicz, J.A. (Comp.)“Popular Summaries from Second InternationalConference on Forest Vegetation Management”, 20-24 March. ~N&i!ealandForessr ResearchInsrilure. FRI Bullcti,r No. 192.

POWERS, R.F.: WEBSTER, S.R.; COCHRAN, P.H. 1988: Estimating response ofponderosa pintforests to fertilization. Pp.21%25 in Schmitt, W.C. (Comp.) “Future Forests in the MountainWest: A Stand Culture Symposium”, Missouln, MT, 29 September- 3 October 1986. US/IAFomw Service, /ntermounrai,z Research Smt~on, Ogden, General Techniuzl Repwf /NT-M?.

POWERS, R.F.; ALBAN, D.H.; MILLEfI, R.E.; TIARKS, A.E.; WELLS, CG; AVERS, P.E.;CLINE, R.G.: F!TZGERALD,R.O.; LOFTUS, N.S. 1990:SustainingsiteproductiJityinNorlhAmerican forests: prohlcmsandprospects. Pp.4!&79inGessel,S.A.(Ed.)“Sustaincd Productivityof Forest Soils”, Proceedings ofthe 7th North American Forest Soils Conference, Vancouver,B.C., July 1988, University of British Columbia, Vancouver. 524 p.

RASHIN, E.; GRABER, C. 1993: Effectiveness ofhcstmanagment practices for aerial application offorest pesticides. Washington Stare Deparlmenr of Ecolo~)~ Reporl TFW-WQI-93-001. X3 p.

RIEKCRK, H. 19X3: Impacts ofsilviculture on flatwoods runoff, water quality, and nutlicnt budgets.Wafer Rrsouvces Buiirrin 19: 7S9.

RIEKERK, H.: NEARY, D.G.; SWANK, W.T. 19X9: Magnitude of upland silvicultural nonpointsourcepollutioninfheSouth. Pp.X-IRi~Hook,D.D.:Lca,R.(Ed.)ProceedingsofaSyrnposium“The Forcsted Wetlands ofthe Southern United States”, 12.-I4 July 1988, Orlando, FL. US/IAFores S’rni~e, Sorrfhen.rrwn Foresf E.rperfmm! .SlaUnn, .4.rhev~ile, Generai Technical ReportSE-50 16X p.

RENARD, K.G.; LOPEZ, F.A.; SIMANTON, J.R. 1991: Brushcontrol and sediment yield. Pp.12-38to 12-45 in Shou-Shou, T.; Yung-Huang, K. (Ed.) Proceedings ofthe Fifth Federal InteragencyScdimentationConfercnce, Volumc2, I %2l March, LasVegas,NV. Federal Energy RegulatoryCommission, Washington, DC.

SMETHURST, P.J.: COMERFORD, N.B.; NEARY, D.G. 1993: Weed effects on early K and Pnutrition and growth of slash pine on a spodosol. Forest Ecolo~ and Manngemenr tie: 15-26.

264 New Zealand Journal of Forestry Science 26( 112)

SQUIRE, R.O.; FARRELL, P.W.; FLINN, D.W.; AEBERLI, B.C. 1985: Productivity of first andsecondrotationstandsofradiarapineonsandysoils. II. Heightalldvolumegrowthat~veyears.Ausrvalian Foresty 4X: 127-37.

SWANK, W.T. 1988: Chapter 25: Stream chemistry responses to disturbances. Pp.33%57 in Swank,W.T.; Crossley, D.A. (Ed.)“Forest Hydrology and Ecology at Coweeta”. Springer-Verlag, NewYork. 469 p.

SWANSON, F.J.; SWANSO, M.M.; WOOD&C. 1981: Analysisofdcbris-avalancheerosiollin steepforested lands: an example from Mapleton, Oregon. Pp.67-75 in Davies, T.R.H.; Pearce, A.J.(Ed.) ProceedingsoftheChristchurch Symposium,“ErosionandSedimentTransportin PacificRim Steeplands”. lnlernalional Associarion ofHydro/o~irniScience. Publication 132. 654 p.

SWINDEL, B.F.; CONPE, L.F.; SMITH, J.E. 1986: Windrowing affects early growth of slash pine.Sou,hern Jouvnal gfApp,red Foresy 10: 814.

TEW, D.T.; MORRIS, L.A.; ALLEN, H.L.; WELLS, C.G. 19X6: Estimates of nutrient removal,displacement, and loss resulting from harvest and site preparation of a Pinus taeda plantationin the Piedmont of North Carolina. Forest Ecolo~ and Management 15: 257-61.

THOMPSON, D.G.; STAZNIK, B.; FONTAINE, D.D.; MACKAY, T.; OLIVER. G.R.; TROTH, J.199 I: Fate oftriclopyr ester (RELEASE) in a boreal forest stream. Environmrntul Toxicologyand Chemrsry IO: 61%32.

USDA FOREST SERVICE 1984: “Pesticide Background Statements: Volume I. Herbicides”. ti.S.Government Pnnf ing Office, Washin@on, D.C., Agriculture Handbook 633.

-1989a: Final environmental impact statement, vegetation management in the Piedmont andCoastal Plain. Sourhrrn Region Management Bulletin R&MB-23. 1248 p.

-1989b: Final environmental impact statement, vegetation management in the AppalachianMountains. Sourhem Re@on Management Buile,irr RR-MB-38. 163X p.

-1990: Final cnvjronmental impact statement, vegetation management in the OrarklOuachitaMountains. Soulhem Re@on Mnna~ement Builelin RB-MB-45. 1787 p.

VAN LEAR; DOUGLASS, J.E.; FOX, S.K.; AIJGSPURGER, M.K. 1985: Sediment and nutrientexpost in runoff from burned and harvested pine watersheds in the South Carolina Piedmont.Joumni of Envimnmenfal Quul iQ 14: 169-14.

VITOUSEK, P.M.; MELILLO, J.M. 1979: Nitrate Iosscs from disturbed forests: patterns andprocesses. Fores1 Science 25: 60>19.

WAN, M.T. 1987: The persistence of triclopyr and its pyridinol metabolite in a coutal BritishColumbiastream. De~~arrmenl~~~h~rhe~nvironrnen~, Conse~votiunandProtect;on, Envimonmen~aiProreckm, Pac~$c and Yukon Region. Regional Program Repor No. 86-24. 21 p.

WELLS, C.G.; CAMPBELL, R.E.; DeBANO, L.F. 1979: Effects of fire on soil: a statc-of-the-knowledge review. USDA Forest Service, WmhinRfon, General TechnicalReport WO-7.34 p.

ZIEMER, R.R. 1981: Rootsand thestabilityofforested slopes. Pp.34=1 in Davies, T.R.H.; Pearce,A.J. (Ed.) Proceedings of the Christchurch Symposium, “Erosion and Sediment Transport mPacific Rim Steeplands”. lnterna~ional Associarion ofHydmlogica1 Science, Publication 132.654 p.