SEDIMENT GENERATION FROM FORESTRY OPERATIONS AND ASSOCIATED EFFECTSON AQUATIC ECOSYSTEMS

Paul G. Anderson

Alliance Pipeline Ltd., 400, 605 – 5 Avenue S.W. Calgary, Alberta, Canada, T2P 3H5Phone (403) 716 0389 email – andersp@alliance-pipeline.com

Anderson, P.G. 1996. Sediment generation from forestry operations and associated effects on aquaticecosystems. Proceedings of the Forest-Fish Conference: Land Management PracticesAffecting Aquatic Ecosystems, May 1-4, 1996, Calgary, Alberta. pp-pp.

Abstract

Timber harvest operations have been shown tohave many effects on adjacent watercoursesand on the aquatic ecosystems they support.This may occur from introductions or loss ofwoody debris, loss of riparian vegetation,accelerated stream bank and bed erosion, thealteration of natural channel form and process,and the reduction of stream habitat diversity.However, the existing literature indicates one ofthe most insidious effects of logging is theelevation of sediment loads and increasedsedimentation within the drainage basin.

Sediment generation from various forestrypractices has been studied extensively in thepast. Forestry practices which generatesuspended sediments include all operations thatdisturb soil surfaces such as site preparations,clear-cutting, log skidding, yarding, slash burns,heavy equipment operation and roadconstruction and maintenance. From thesesources, construction, use and maintenance oflogging roads located in close proximity towatercourses produce by far the highest levelsof suspended sediment generation in streams.

Three aspects of logging road development andmaintenance are known to elevate sedimentloads in watercourses: 1) instream and near-stream construction operations; 2) reduction inretention time and associated increase inerosion in the drainage basin; and 3) mass soilmovements and/or landslides associated withlogging road design and placement.

This literature review examined the effects ofincreased sediment load and sedimentation onaquatic ecosystems emphasizing forestryoperations that generate elevated sedimentloads. The review included the effects ofsediment on fish (behavioral, physiological and

population effects) and the effects ofsedimentation on fish habitats (includingspawning, rearing, food production, summer andoverwintering habitats). A habitat effectsrelationship was presented which related theconcentration and duration of specific sedimentexposure events to the alteration of fish habitats.This relationship allows for post-disturbanceevaluation of the potential effects on fish habitat.

Many advances have been made in the designand construction of logging roads over the pastthree decades. Many of these advances havebeen made in the design of road accesssystems, while other advances have been madein the development of more sophisticatedmitigation techniques. These advances aredescribed and discussed in relation tominimizing sediment generation and subsequentsedimentation in aquatic environments.

Introduction

Forest harvesting practices can elicit a numberof physical changes within a watershed. Thesechanges can set up associated responses in awide range of physical, chemical and biologicalprocesses and can substantially alter aquatichabitats and communities. Forestry relatedactivities can influence: stream hydrologic regimeby reducing the time between peak rainfall andpeak stream discharge and consequentlyincreasing the magnitude of peak seasonal flows;water quality parameters such as temperatureand sediment load; and, stream geometry byincreasing erosive forces, channel migration,width to depth ratio and altering stream form andprocess (Meehan 1991; Salo and Cundy 1987)).

Forestry-related activities are not always harmfulto freshwater fish communities. Long term andintensive study of logging on fish communities ofCarnation Creek, Vancouver Island, British

Columbia, revealed the annual growth of juvenilecoho salmon (Oncorhynchus kisutch) increased inyears immediately following logging which waslikely in response to increases in air and watertemperatures (Hartman and Scrivener 1990;Holtby 1988; Tschaplinski and Hartman 1983).However, forestry activities have been shown tohave adverse effects on resident and migratoryfish communities, often related to an increase inthe delivery of sediments to streams.

Although frequently viewed as synonymous,suspended sediments and sedimentation are twodiscrete processes that affect aquaticcommunities in different ways. Sedimenttransport in watercourses occurs in three forms,which include wash load, suspended, load andbed load. The entry of sediments to watercoursesfrom upland sources is a natural process;however, when the rate entering a watercourseexceeds the capacity of the watercourse totransport or assimilate the sediment, stress mayoccur to the aquatic community and the suitabilityand/or productivity of aquatic habitats may bealtered. Excess sediment in rivers and streamshas been identified as the largest and mostpervasive water pollution problem faced byaquatic systems in North America (Sweeten1995).

This paper provides a collection of informationrelated to biological responses to sedimentsynthesized from the literature and attempts touncover relationships which are present betweenelevated sediment episodes and biologicalresponse. For additional information pertainingto this topic or to obtain additional references,the reader is directed to Meehan 1991, Salo andCundy 1987, Newcombe (1994), Newcombe andJensen (1995), Kerr (1995), Waters (1995) andAnderson et al. (1996).

Sources of Elevated Sediment Loads

Anthropocentric Sources of ElevatedSediments

Erosion within streams is a natural process andis affected by parameters such as stream flow,channel structure and stability, streambedcomposition, and disturbance within thewatershed such as fire, landslide or ice scour.The disturbance of lands accelerates erosionand increases the delivery of sediment to streamsystems. Any activity that is undertaken within a

watershed that disturbs land surfaces has thepotential to increase sediment delivery tostreams. Most of man’s activities will increaseerosion to some extent in forested watersheds.Man’s activities which most frequently increasesediment loads in watercourses includeagriculture, mining, forestry, urban development,and stream channel alteration such as dams andchannelization, and instream constructionassociated with developments such as bridges,roads or pipeline and transmission crossings.This paper discusses the effects of sedimentincrease associated with episodic elevatedsediment events in general, however, types ofmitigation to minimize sediment load increaseswhich are specific to the forestry industry arediscussed in the Mitigation Measures sectionand sediment sources specific to forest harvestoperations are discussed below.

Forestry Sediment Sources

Forestry operations frequently increase sedimentdelivery into streams. Logging operations havebeen shown to increase sediment productionabove natural sedimentation rates (Megahanand Kidd 1972). The activities which commonlyresult in increased sediment delivery includeclear-cutting, skidding, yarding, site preparationfor replanting and road construction, use andmaintenance (Waters 1995).

Among forestry harvest activities, disturbanceassociated with logging road construction andoperation produces the greatest sediment loadincrease (Waters 1995; Furniss et al. 1991;Cederholm et al. 1981). Roads associated with ajammer logging system in the Payette NationalForest, Idaho, increased sediment production anaverage of approximately 750 times over thenatural rate over a six-year period followingconstruction (Megahan and Kidd 1972). Theaverage erosion rate from roads on the jammerunit for 1.35 years preceding logging was 56.2tons/mile2/day and the average rate for 4.8 yearsfollowing logging was 51.0 tons/mile2/day(Megahan and Kidd 1972).

Effects of Elevated Sediment Loads

Sediment Effects on Fish

In response to changes in the environment,ecosystems often undergo changes in

community composition and structure.Organisms respond to environmental change inorder to avoid or minimize effects on fitness. Ifan organism can not compensate for a changein the environment and suffers a reduction infitness, the environmental change is termed astress (Brett 1958; Kohen and Bayne 1989).Therefore, a stress limits either the rate ofresource acquisition or growth and reproductionso that fitness is reduced (Grime 1989).

Stress has been defined as “the sum of all thephysiological responses by which an animal triesto maintain or re-establish a normal metabolismin the face of a physical or chemical force”(Selye 1950). Stress occurs when thehomeostatic or stabilizing processes of the fishor organism are extended beyond thecapabilities of the organism to compensate forthe biotic or abiotic challenges. Anthropogenicinputs of sediments into stream and riverineenvironments can cause stress to aquaticsystems and thereby, directly and indirectlyimpact upon fish behaviour and health.Increased concentrations of suspendedsediments can have direct effects on fishbehaviour, fish physiology and fish populations(Anderson et al. 1996).

Behavioural Effects

Changes in fish behaviour are some of the firsteffects evoked from increasing concentrations ofsuspended sediments. Behavioural changes aregenerally considered benign and transitory. Theyare easily reversed and do not exhibit a long-lasting impact (Newcombe 1994). Typicalresponses include an increased frequency of thecough reflex, avoidance of suspendedsediments, reduction in feeding and temporarydisruption of territoriality. The severity of thebehavioural response is associated with thetiming of disturbance, the level of stress (andassociated energy cost) and the importance ofthe habitat that the fish may be being excludedfrom.

The avoidance of suspended sediment plumesis one of the first reactions. Bisson and Bilby(1982) observed this behaviour evoked injuvenile coho salmon at total suspendedsediment (TSS) concentrations as low as 88mg/L. Similar results were recorded by McLeayet al. (1987) who found that Arctic grayling(Thymallus arcticus) avoided concentrationsgreater than 100 mg/L.

Increased concentrations of suspendedsediment have also been correlated with areduction in feeding. Feeding rate may be afunction of prey visibility. McLeay et al. (1987)states that Arctic grayling exposed to suspendedsediment concentrations greater than 100 mg/Lwere slower to recognize the food and morefrequently missed a food item when theyattempted to eat it. Sigler et al. (1984) believes,however, that a reduced feeding is morecomplex than reduced ability to see prey items,as many fish species (especially benthicfeeders) do not use sight to identify prey items,but still exhibit reduced levels of feeding inresponse to elevated sediment loads.

High concentration of suspended sediments hasalso been associated with the loss of territorialityand interruption of movements of salmonids.Berg and Northcote (1985) found that territorialbehaviour was lost at concentrations exceeding30 NTU. They indicate that territoriality isinfluential in the allocation of food and habitatresources. Disruption to territoriality can occurwhen turbidity limits the visual distance thatindividuals can see, and when the downstreamdrift of fish avoiding increased concentrations ofsuspended sediment disrupts existing territories.

Physiological Effects

Physiological changes can be measured in fishas a response to the increased stress ofsuspended sediments. The typical measuredresponses include impaired growth, histologicalchanges to gill tissue, alterations in bloodchemistry, and an overall decrease in health andresistance to parasitism and disease. Theeffects of sediment exposure on each of thesephysiological effects are discussed below.When compared to sediment exposure eventsthat elicit behavioural responses, longerexposure periods and/or higher concentrationsare generally required before physiologicalresponses are expressed. In this respect,physiological responses are more of a chroniceffect. The effects are usually a gradedresponse to increasing sediment dose. Impactsevoked from lower doses can be transitory, whilethose resulting from higher doses can be morelasting and severe.

GrowthAn impaired growth rate is generally one of themore sensitive physiological responses to an

increase in suspended sediment concentration.Unlike behavioural responses, impaired growthgenerally requires a longer exposure periodbefore effects are manifested. Sigler et al.(1984) found that growth was impaired injuvenile steelhead trout (Oncorhynchus mykiss)and coho salmon exposed to fire clay orbentonite clay at concentrations between 84 and120 mg/L during a 14 to 21 day exposure period.Similar concentrations of 100 mg/L or greaterwere found to significantly impair growth in Arcticgrayling under-yearlings (McLeay et al. 1987),largemouth bass (Micropterus salmoides),bluegill (Lepomis macrochirus), and redearsunfish (Lepomis gibbosus) (Buck 1956).However, growth impairment may be relatedmore to the metabolic demands resulting fromstress caused by increased suspended sedimentthan from a reduction in feeding. The timerequired before growth impairment wasmeasurable ranged from a low of two weeks forjuvenile steelhead trout and coho salmon to ahigh of six weeks for Arctic grayling under-yearlings (Sigler et al. 1984; McLeay et al. 1987).

Blood Chemistry

Alteration in blood chemistry resulting from theincreased stress of suspended sediments havebeen found associated with concentrationsranging between 500 to 1500 mg/L (Redding andSchreck 1982; Servizi and Martens 1987). Thechanges most commonly recorded include anincrease in haematocrite, erythrocyte count,hemoglobin concentration, and elevated bloodsugar levels (hyperglycemia), plus decreases inblood chloride content, and depletion of liverglycogen (Wedemeyer et al. 1990; Servizi andMartens 1987). These increases coincide withthe release of stress hormones (i.e., cortisol andepinephrine) and traumatization of the gill, andpresumably represent a compensatory responseto a decrease in gill function (Newcombe 1994).In addition, Sherk et al. (1973) found thesechanges to be associated with a reduction in theswimming endurance of white perch (Moroneamericana) exposed to 650 mg/L of TSS. Mostof the observed changes resulted after four tofive days of exposure (Newcombe 1994).Exceptions to this, however, were noticed byRedding and Schreck (1982) who found asignificant increase in haematocrit volume withinsteelhead trout after only nine hours of exposureto 500 mg/L of volcanic ash, clay and topsoil.

Gill Trauma

Increased concentration of suspendedsediments are known to physically traumatize gilltissue. The primary mechanisms of action isthrough physical abrasion of tissue and particleadsorption onto the gill. The types of tissuechanges observed include swelling of secondarylamella and hypertrophy (cell swelling) ofepithelial cells (Sherk et al. 1973); hyperplasia(increase in cell number) of gill tissue (Simmons1984); and tissue necrosis (Servizi and Martens1987).

The severity of damage appears to be related tothe dose of exposure, as well as the size andangularity of the particles involved. Greaterdamage is typically observed with larger, moreangular particles (Servizi and Martens 1991).These factors could account for the large rangein responses seen for different exposure rates.For example, concentrations as low as 270 mg/Lare known to cause gill damage in rainbow trout(Oncorhynchus mykiss) (after 13 days ofexposure (Herbert and Merkens 1961) and yetMcLeay et al. (1987) found no gill damage inyoung-of-the-year (YOY) Arctic grayling thatwere exposed to concentrations as high as 1300mg/L; the duration of exposure was, however,only four days.

Secondary effects resulting from an infestationof parasitic protozoans were found in juvenilerainbow trout that were exposed to extremelyhigh concentrations of suspended sediments.The trout were exposed to 4887 mg/L for aperiod of 16 days (Goldes 1983). This author didnoted that the protozoan infection and gillarchitecture was found to be normal 58 daysafter the exposure ceased.

Resistance

Increased concentrations of suspendedsediments have been associated with an overalldecrease in the ability to defend against diseaseand to tolerate chemical toxins. For example,Herbert and Merken (1961) observed rainbowtrout to be more susceptible to infestations of finrot when fish were exposed for 121 days toconcentrations of 270 mg/L of diatomaceousearth. Likewise, Servizi and Martens (1991)found a correlation between the prevalence of aviral kidney infection and an increasedconcentration of suspended sediments in cohosalmon. When concentrations of suspended

sediments exceeded 100 mg/L, the tolerance ofArctic grayling to the toxicant pentachlorophenol(PCP) decreased (McLeay et al. 1987). Thisobservation by McLeay et al. (1987) indicates ageneral decrease in tolerance to increasedenvironmental stressors.

Phagocytosis

A process that may be closely linked to reducedresistance, is phagocytosis. Newcombe andJensen (1995) discuss the process by which fineparticles are enveloped by cells within fish gilland gut tissues and are transported to internalrepository tissues. The main organ of repositoryin fish is the spleen (Newcombe and Jensen1995). It is hypothesized that through thisprocess, particles could reduce resistance toother stressors by impairing fish health. Inaddition, particles could trigger tumour induction,especially in circumstances where contaminantswere adsorbed to particles in suspension(Newcombe and Jensen 1995).

Lethal Effects

Increased concentrations of suspendedsediments and increased sedimentation rateshave the potential to affect fish populations. Theprimary mechanisms of action are throughincreased egg mortality, reduced egg hatch, areduction in the successful emergence of larvae,plus the sediment-induced death of juvenile andadult fish. These mechanisms are discussedbelow.

Egg Mortality

The primary cause of egg death is generallyfrom burial by settled particles. Thin coverings (afew mm) of fine particles are believed to disruptthe normal exchange of gases and metabolicwastes between the egg and water.Sedimentation rates of 0.03 to 0.14 g dry weightsediment/cm2 (i.e., 1-4 mm depth of silt and clay)significantly reduced the survival of lakewhitefish (Coregonus clupeaformis) eggs (Fudgeand Bodaly 1984). The effects upon eggmortality appear to be more closely related to thesedimentation of particles and less related to theconcentration of suspended sediments. Zallen(1931) observed that concentrations of 1000 to3000 mg/L had no effect upon the survival ofmountain whitefish eggs (Prosopium

williamsoni). Campbell (1954), however, found100 percent mortality in rainbow trout eggsexposed to TSS concentrations of 1000 to 2500mg/L. Differences in egg mortality effectsassociated with elevated sediment loads isrelated to the size of the sediment particlesinvolved and rates of sediment deposition.

In addition to the concentration of suspendedsediments and the size of the particles involved,the duration of exposure appears to be key afactor in determining the effects of sediments onegg survival. Slaney et al. (1977) noticed thathatching success for rainbow trout was reducedafter 2 months of exposure to 57 mg/L. Asignificant reduction in the hatching success ofwhite perch and striped bass (Morone saxatilis)was observed in only 7 days after exposure toabout 1000 mg/L TSS (Auld and Schubel 1978).The magnitude of the effect of sedimentexposure may also be influenced by the timing ofsediment exposure with respect tot the stage ofembryo development. The dose of sedimentrequired to induce egg mortality is greatlyinfluenced by the physical characteristics of thestream which, in turn, affect sediment transportcapabilities and the capacity to maintainsediments in suspension or otherwise to result intheir deposition.

Juvenile and Adult Fish Death

Juvenile and adult fish generally appear to bemore resilient to stress from suspendedsediments than other life history stages. Shortterm increases in TSS concentrations between11000 and 55000 mg/L appear to be the point atwhich salmonid mortality significantly increases(Stober et al. 1981; Servizi and Martens 1987;Smith 1940). McLeay et al. (1983) reportedsurvival of Arctic grayling subjected tomoderately high concentrations (1000 mg/L) offine grained materials (mining silt). Lloyd (1987),in a review of existing information, reportedlethal effects to fish at concentrations rangingfrom 500 to 6000 mg/L. Sigler et al. (1984)reported mortality in young of the year cohosalmon and steelhead trout at 500 to 1500 mg/L.Based on the information available on sedimentand acute effects to fish, it is apparent that theseverity of effect caused is a function of manyfactors which, in addition to concentration,duration particle size and life history stage, mayinclude temperature, physical and chemicalcharacteristics of the particles, associatedtoxicants, acclimitization, other stressors and

interactions of these and other factors (Waters1995).

Habitat Effects

Habitat Exclusion and Habitat Alteration

In addition to the direct impacts of suspendedsediments on fish, increases in sediment loadscan also alter fish habitat or the utilization ofhabitats by fish (Scullion and Milner 1979, Lisleand Lewis 1992). High sediment loads can alterfish habitats temporarily by affecting waterquality, making a stream reach unsuitable foruse by fish. This habitat exclusion, if timedinappropriately, could have impacts on fishpopulations if the affected habitat is critical to thepopulation during the period of elevatedsediment load. This principle of habitatexclusion is very important one; however, thisissue is separate from the issue of direct habitatalteration that will be discussed below.

Sediment episodes can have a prolonged effecton the suitability of habitats within a streamreach through increased levels of sedimentation.In fact, sedimentation is the single mostimportant effect associated with sediment loadincreases, since sediment loads can alter thegross morphology of streams as well as thecomposition of the stream bed and associatedhabitats.

Changes in Stream Bed Porosity

Larger-sized materials, such as fine to coarsesand are quick to settle onto the stream bed.This material may accumulate on the surface ofthe stream bed or filter down into the inter-gravelspaces. Interstitial spaces can become cloggedby the downward or, to a lesser extent, by thehorizontal movement of sediment (Beschta andJohnson 1979).

Water movement through the stream bedmaterials is important for the benthiccommunities which reside there, and for thedeveloping embryos of fish species who burytheir eggs. Inter-gravel water movement iscontrolled by several hydraulic and physicalproperties of the stream and its bed. Thepermeability of the stream bed is determined bysize composition of the substrate material,viscosity of the water (temperature dependent)and the packing of the substrate material (Stuart

1953; Cooper 1965). A small increase in theproportion of fine material can severely reducethe porosity and permeability of the gravel bed(Lisle and Lewis 1992) and the ability of alevinsto receive adequate oxygen and emerge fromthe gravel.

Changes in Stream Morphology

In addition to altering stream bed composition,elevated sediment loads can also changechannel geometry (Klein 1984). Elevated levelsof sediment deposition can reduce the depth ofpools and produce a net reduction in riffle areas.This accumulation of streambed deposits canreduce available habitat. For example,deposition of sediments in pools and other areasof instream cover can cause a decrease in thefish holding capacity of a stream reach (Bjornn etal. 1977). Smith and Saunders (1965) found thatdecreased brook trout (Salvelinus fontinalis)populations were related to infilling of availablecover. Alexander and Hansen (1992) also notedthan a decrease in sand bedload sediment wasassociated with an increase in rainbow trout andbrown trout (Salmo trutta) populations. Changesin physical morphology of the stream can alsoinhibit the movement of fish or change thedistribution of adult fish (Alabaster and Lloyd1982).

Channels affected by sediment derived fromAnthropogenic disturbance are also moretransitory in nature. Fox (1974) found urbanwatersheds exhibited a 33% monthly change ingeometry as compared to a 5% change in lessdisturbed rural drainages. Sediment materialdeposited within streams can be in constantmotion as bedload transport slowly moves thedeposited materials through the system. Thismaterial in motion can increase bed scour andbank erosion as the sediment increases theerosive force of the water, by creating a “sandblasting” effect.

Sedimentation Effects on Spawning Habitats

River spawning salmonids typically deposit theireggs in gravel beds commonly found in theupper reaches of river systems. For example,brown trout typically bury eggs in interstitialspaces of the substrata to depths of 9 to 12 cm(Scullion and Miller 1979). Alevins remain in theinterstitial spaces until the start of exogenousfeeding. The percolation of water through the

incubation substrate is an essential factor indetermining the survival rate of incubating eggs(Lisle and Lewis 1979).

An increase in percent of fine material in thestream bed can have impacts on egg survivalrates (Shaw and Maga 1943; Cordone andKelley 1961) since it reduces streambedpermeability. Lowered permeability reduces theinterchange between stream flow and watermovement through the redd, resulting in areduction in the supply of dissolved oxygen tothe egg and a hindrance to the removal ofmetabolites. Slaney et al. (1977) reported thatrainbow trout egg survival was significantlyreduced when spawning gravel contained morethan 3% of fines (diameter 0.297 mm). Inaddition, Hall and Lantz (1969) determined thathatching success of coho salmon and cutthroattrout (Onchorhynchus clarki) was reduced by 40to 80% when spawning substrates contain 20 to50% fines (1-3 mm diameter).

Even if intergravel flow is adequate for embryodevelopment, sand that plugs the interstitialareas near the surface of the stream bed canprevent alevins from emerging from the gravel(Koski 1966; Phillips et al. 1975). For example,the emergence success of westslope cutthroattrout was reduced from 76% to 4% when finesediment was added to redds (Weaver andFraley 1993).

Female stream spawning salmonids typicallyclean an area of the stream bed in which theybury their eggs. This nest building activityflushes sediments and increases the stream bedpermeability. With time, sediment conditionswithin the redd gradually return to ambient levels(Wickett 1954; McNeil and Ahnell 1964). Undernormal conditions, this slow increase in sedimentintrusion is not a problem; however, increasedlevels of sediment within a system as a result ofanthropocentric disturbance increase the rateand level of sediment intrusion and reduces theperiod of time in which the redd is clean. Theperiod of time before sediment intrusion into theredd is very important with respect to the survivalof salmonid larvae. Studies by Wickett (1954)suggest that sediment accumulation during earlyembryonic development may result in higher eggmortalities than if deposition occurs after thecirculatory system of developing larvae isfunctional. This may be due to the higherefficiency in oxygen uptake by the embryo or

alevin with a functional circulatory system (Shawand Maga 1942).

Ringler and Hall (1975) documented increasedtemperature and reduced dissolved oxygenlevels of intragravel water in salmon and troutspawning beds because of clearcut loggingpractices. An associated reduction in residentcutthroat trout populations was attributed to thisreduction in spawning habitat suitability.However, the failure to document seriousreductions in coho salmon could be related tosediment clearing and removal by these largerfish during redd construction.

Sedimentation Effects on Fish RearingHabitat

Sediment deposition also affects rearing habitatof juvenile fish since young salmonids frequentlyuse the interstitial spaces in the stream bed forcover. Thus, a reduction in the suitability ofpotential rearing habitat as a result of sedimentintroduction is related to a reduction in the spaceavailable for occupancy (Reiser et al. 1985).When pools and interstitial spaces in gravel fillwith sediment, the total amount of habitatavailable for rearing is reduced (Bjornn et al.1977). Griffith and Smith (1993) found thatnumbers of juvenile rainbow trout and cutthroattrout decreased due to lack of available cover inheavily embedded gravel substrata. Interstitialspace is particularly important during winterbecause juvenile fish live in these areas makingthem especially susceptible to impacts fromincreased sedimentation (Bjornn et al. 1977).Without these inter-gravel refugia, young fishmay abandon the stream or move to lesssuitable areas where survival rates may bereduced.

Sedimentation Effects on Food Supply

Sedimentation can affect fish populations byaltering the available food supply. Increasedconcentrations of suspended sediments andincreased rates of sedimentation can reduce theprimary productivity of the impacted area.Periphyton communities are likely the mostsusceptible to the scouring action of suspendedparticles or burial by sediments. Atconcentrations exceeding 115 mg/L, suspendedsediments can reduce light penetration andprimary productivity (Singleton 1985). Areduction in primary productivity has the potential

to appreciably decrease the food supply ofmacrobenthos that graze on periphyton(Newcombe and Macdonald 1992). Manymacrobenthic organisms are, in turn, used as afood source by fish.

Increased sediment loads in streams can alsohave an effect on zooplankton andmacrobenthos. Sediment release can effect thedensity, diversity and structure of residentinvertebrate communities (Gammon 1970; Lenatet al. 1981). A number of studies havedemonstrated decreases in invertebratedensities and biomass following sedimentationevents (Wagener 1984; Mende 1989).Increases in sediment input may reduce thedensity of invertebrates by directly affectingaspects of their physiology or by altering theirhabitat. Suspended sediments can have anabrasive effect on invertebrates and interferewith the respiratory and feeding activities ofbenthic animals (Tsui and McCart 1981).Increased sediment deposition may also reducethe biomass of invertebrates by filling theinterstitial spaces with sediments and byincreasing invertebrate drift or covering thebenthic community in a blanket of silt (Cordoneand Kelley 1961; Tsui and McCart 1981).Increases in sediment deposition that affect thegrowth, abundance, or species composition ofthe periphytic (attached) algal community willalso have an effect on the macroinvertebrategrazers that feed predominantly on periphyton(Newcombe and MacDonald 1991).

A change in particle-size distribution in thestream bed can alter the habitat and make itunsuitable for certain species of invertebrates.Gammon (1970) noticed that an increase insuspended sediments from 40 to 120 mg/Lresulted in a 25 to 60% decrease in the densityof stream macroinvertebrates. Likewise, Slaneyet al. (1977) found that a 16 hour pulse ofsuspended sediments (2500 to 3000 mg/L) ledto a 75% reduction of invertebrate biomasswithin the most affected areas.

Sedimentation can alter the structure of thebenthic invertebrate community by causing ashift in the proportion from one functional groupto another. For example, streams with clearwater normally contain a high proportion ofinvertebrates in the shredder group; however, ifsediment deposition is substantially increased,shifts to other groups such as grazers (Bode1988) or collector-gatherers may occur

(Wagener 1984). Some studies indicate thatincreased inputs of sediments cause a shifttowards chironomid-dominant benthiccommunities (Rosenberg and Snow 1975;Dance 1978; Lenat et al. 1981).

Benthic fauna possess behavioural andmorphological adaptations which limit them frombeing displaced in a unidirectional flowenvironment (Hynes 1973). Invertebrate drift,however, is a continuous redistributionmechanism that occurs in most streamecosystems. It is an important factor in theregulation of population density (Williams andHynes 1976), in the dispersion of aggregationsof young larvae (Anderson and Lehmkuhl 1967),in the abandonment of unsuitable areas(Williams and Hynes 1976), and in therecolonization of areas after disturbance (Barton1977).

Invertebrate drift may be induced by elevatedsuspended sediment levels (Rosenberg andWeins 1978). Increased rates of downstreamdrift by macrobenthos can be induced byconcentrations as low as 23 mg/L (Rosenbergand Snow 1975). Drifting affords invertebrate taxathat are sensitive to increased sediment loads theopportunity to avoid areas which becomeunsuitable as a result of high suspended sedimentlevels. Conversely, invertebrate drift is consideredto be the most important component of ecosystemrecovery following stream disturbances (Williamsand Hynes 1976; Barton 1977; Young 1986). Thisis especially true in areas of swift-flowing waters(Waters 1964).

Sedimentation Effects on OverwinteringHabitat

The magnitude of impact upon fish resultingfrom increased concentrations of suspendedsediments and levels of sedimentation can varyseasonally. It has been argued that the loweredmetabolic requirements during winter conditionsmay in some ways provide a protective influenceto conditions such as gill trauma and decreasedgill function (C. Newcombe, BCMELP, pers.comm.). However, the ability of the fish tocompensate for the stress of suspendedsediments is influenced by a number of factorsincluding the physiological condition of the fishand its ability to respond to the stress.

Early live stages (i.e., eggs, alevins) of manysalmonids are found in the stream bed during

the winter months. These stages are particularlysensitive to the effects of increasedconcentrations of suspended sediments and thedeposition of fines sediments. The introductionof sediments during the winter, therefore, hasthe potential to appreciably influence these earlylife stages.

Bjornn et al. (1977) found that the number ofjuvenile salmon that a stream can support inwinter was greatly reduced when the inter-cobblespaces were filled with fine sediment. Thedecreased carrying capacity was a function ofboth a loss of substrate cover for juvenile fishand a reduction in food as benthic invertebratecommunities changed. Bjornn et al. (1977)suggested that the summer rearing or winterholding habitat may be more influential to thecarrying capacity of a stream reach than embryosurvival.

During winter fish generally experiencedecreased energy reserves and as such searchfor habitat that allows them to reduce energyexpenditures (Clapp et al. 1990; Nickelson et al.1992). Preferred habitats are speciesdependent; however, for most salmonidspreferred habitats are located in low velocityareas such as pools and behind instream coverwere focal velocities are low (Vondracek andLonganecker 1993; Griffith and Smith 1993;Modde et al. 1991; Heggenes and Saltveit 1990;Cunjak and Power 1986; Tschaplinski andHartman 1983). By remaining in low velocityareas, fish are able to minimize their energyexpenditures and hence reduce the rate ofmetabolic depletion (Cunjak and Power 1986).

Land-use activities that increase the delivery offine materials to streams can significantly affectthe overwintering survival of resident fish. Amechanism of potential impact is a depletion ofcritical energy reserves as a result of increasedphysiological stress, alterations in behaviourand/or exclusion from preferred sites ofoverwintering habitat. This is particularlydeleterious to fish species and life stages thatprefer to overwinter within the interstitial spacesof the stream bed. The net loss in energyreserves will depend on the concentration ofsediment and the duration of impact. Dependentupon existing energy reserves, the fish may beable to tolerate the energy depletion attributableto an increase in the cough reflex and reducedfeeding, but may not be able to tolerate the

energy depletion associated with displacementfrom critical habitats.

Preferred winter habitat areas of low currentvelocity are often predisposed to sedimentation(Cunjak 1996), and lower flows oftenexperienced during winter may result in higherrates of sediment deposition. Bjornn et al. (1977)found, during sediment experiments, that thespring freshet from snow melt was rarelysufficient to transport sediment out of pools;therefore, the damage to these areas isfrequently of a longer-term than sedimentsdeposited in more erosive habitats. Due tonatural factors, the availability of winter habitat isgenerally less than that of summer habitat andmay be more influential in the determination ofthe stream’s natural carrying capacity (Cunjak1996; Mason 1976). A further reduction in theabundance of already limited winter habitat maysignificantly affect the overall fish population of awatercourse (Hartman and Scrivener 1990).Additive to this problem may be a reduction infood supply resulting from benthic drift or buryingof food supplies. Elwood and Waters (1969)observed that increased sedimentation reducedthe population of invertebrates and hence thecapacity of the stream to support brook trout. Areduced food supply, and a greater expenditureof energy in food search and avoidance of higherconcentrations of suspended sediments maysignificantly impact upon the fish’s ability tocompensate for negative physiological changesand the ability to survive the winter.

Sediment Load Biological ResponseRelationships

The Dose/Response Approach

One method that has been developed to addressthe issue of quantifying the adverse effects of TSSon fish is the ranked effects model first putforward by Newcombe and MacDonald (1991).This model compiled information from more than70 studies on the effects of inorganic suspendedsediments on freshwater fish (mainly salmonids)and invertebrates, and ranked the severity ofimpacts from 1 to 14 (rank effects). Linearregression was used to correlate ranked effectswith intensity (concentration x duration) ofincreased suspended sediment load (Newcombeand MacDonald 1991). Since the effect ofelevated TSS levels on fish is a function of boththe concentration of suspended sediment and

the duration of the exposure, Newcombe andMacDonald (1991) developed a Severity Index(SI). This Index provides a standardized relativemeasure of exposure. It is the natural logarithmof the concentration (mg�L-1) multiplied by hoursof exposure (i.e., Ln mg�h�L-1). This SI providesa convenient tool for predicting effects of episodesof elevated suspended sediments of knownconcentration and duration.

The usefulness of the Newcombe and Macdonald(1991) concentration-duration response modelhas been questioned in the past (Gregory et al.1993). The main concerns with the approach arethe highly variable nature of the data used todevelop the severity of effects model that reducesits predictive power, and the concern that themodel is unrealistically simplistic (Gregory et al.1993). A separate concern associated with theNewcombe and Macdonald (1991) severity ofeffects model is the severity index (Ln mg�h�L-1)assumes a unit increase in episode duration (inhours) has a similar effect as a unit increase inconcentration (in mg/L) (Anderson et al. 1996).

In 1994, the ranked effects model was furtherrefined (Newcombe 1994). Using 140 articles onsuspended sediment pollution, Newcombedeveloped a database of nearly 1200 datapointsconcerning the effects of suspended sedimentsand associated effects upon marine andfreshwater biota. With this database, regressionanalysis was used to relate severity of effect tothe dose of TSS for specific fish species orassemblages. This approach was used todescribe the dose/response relationship for theeffects of suspended sediments on salmonidfishes (Newcombe and MacDonald 1991), onjuvenile salmon (Newcombe 1994) and for othercoldwater fishes (Newcombe 1994) using sub-sets of the dataset which are presented incomplete form in Newcombe (1994).

In addition to the dose/response relationshipspresented in for salmonid fishes (Newcombeand MacDonald 1991) and coldwater fishes andunderyearling trout (Newcombe 1994),Newcombe and Jensen (1995) further expandupon dose/response relationships of aquaticresources to sediment exposure. Newcombeand Jensen (1995) presented six sedimentdose/response relationships for specific fishcommunities exposed to elevated sedimentloads. One of the important additions to theanalysis presented in Newcombe and Jensen(1995) was the linkage of grain size of sediment

to the nature of ill-effect associated withsediment exposure. The dose/responserelationships presented in Newcombe andJensen (1995) were organized according to fourvariables: taxonomic group; life stage; naturalhistory and estimated predominant particle sizerange of the sediment episode. Thedose/response relationships were characterizedby three variables: [x], [y], and [z], where:

[x] = duration of exposure expressed asthe natural log of hours;

[y] = concentration of sedimentexpressed as natural log of mg/L;

[z] = severity of ill effect (SE).

Since the work of Newcombe and MacDonald(1991) and Newcombe (1994) used the StressIndex (Ln concentration · duration), thedose/response relationships presented were ofthe form:

Equation 1: z = a + bxwhere:SE = severity of ill effects a = intercept; b = slope of the regression line x = stress index value

The six dose/response relationships presentedin Newcombe and Jensen (1995) do notcharacterize response using the stress index,and are presented in the form:

Equation 2: z = a + b(Ln x) + c(Ln y)

The dose/response relationships (Newcombeand MacDonald 1991; Newcombe 1994;Newcombe and Jensen 1995) provide insightinto the relationship between sediment releaseand adverse effects on a variety of fishcommunities. These relationships provideincreased precision for the prediction ofresponse of a particular fish species orassemblage of species based on a given dose ofsuspended sediment.

The database used to determine theserelationships relied heavily on the physiologicalresponse of fish to increases in sediment load;therefore, the relationships presented may notbe directly applicable to the prediction of physicalalteration to fish habitats due to sediment loadincreases and increased sedimentation nor thelong term impacts from reduced growth or thepossible exclusion from certain habitats.

Developing the Habitat Effects Database

Anderson et al. (1996) attempted to developeffects relationships for fish habitat response toincreased sediment. The first step in developingmore specific dose/response criteria for habitateffects was to search the literature for studiesreporting TSS concentrations and their effectson fish habitat. Several fundamental criteria hadto be satisfied for data from a report to beincluded in the expanded database that wasdeveloped. The report had to give at least: 1) aconcentration of TSS; 2) a duration of exposureto 3) one or more identified species of fish; and4) a description of the effect. The severity ofeffect, and occasionally other data, sometimeshad to be inferred from the qualitativedescriptions. A total of 18 reports, containingsome 53 new documentations of TSS effects,were retrieved in the literature search (Andersonet al. 1996).

Dose/Response Relationships of FreshwaterHabitats to Sediment Releases

Severity of ill-effects rankings on a scale from 0-14 were assigned to each documented effectfollowing the severity-of-ill-effects scalepublished by MacDonald and Newcombe (1993),Newcombe (1994), and Newcombe and Jensen(1995). The nature of the observed habitateffect was assigned one of the following classeffect rankings:

SE = 3 Measured change in habitatpreference;

SE = 7 Moderate habitat degradation -measured by a change in theinvertebrate community;

SE = 10 Moderately severe habitatdegradation - as defined bymeasurable reductions in theproductivity of habitat forextended periods (months) orover a large area (kms);

SE = 12 Severe habitat degradation - asmeasured by long-term (years)alterations in the ability ofexisting habitats to support fishor invertebrates; or ,

SE = 14 Catastrophic or total destructionof habitat in the receivingenvironment.

Emphasis was placed on sediment releaseevents (i.e., effects of sediment pollution eventsrather than chronic erosion and sediment loadproblems). The database excluded anydatapoints for which the extent of habitatmodification could not be ascertained from theprimary manuscripts. The database wasreduced to 35 entries (Anderson et al. 1996) andwas used in the analysis described below todevelop a relationship between sediment doseand habitat effects.

The dose/response approach of Newcombe andMacDonald (1991) and Newcombe (1994)defines dose as the product of TSSconcentration (C in mg/L) and duration ofexposure (T in hours). This definition of dose isstrictly empirical and reflects the observation thatthe product of concentration and duration bearsa closer correlation with ranked effects thanconcentration alone. The inherent assumption isthat brief exposures to high doses of TSS areequivalent in effect to prolonged exposure ofmuch lower doses. Since severity of effect isdetermined based on a linear relationship withdose (Ln C·T) of the form SE = a + b (Ln C·T),the biological receptor response (SE) isassumed to respond to an effective dose inwhich concentration and duration are equally asimportant (i.e., the Effective Dose = CnT; wheren = 1).

However, it has been proven in much of theliterature related to the response of biologicalreceptors to toxic agents, that the relationshipbetween concentration and duration is oftenmore complex. That is, a high concentration fora very short time can cause a higher or lowerresponse than can a low concentration for alonger time (Zelt 1995). In essence, byassuming a linear response (as measured bySE) to dose (as a function of Ln C·T), it isassumed that a unit increase in concentration (inmg/L) is equal to a unit increase in time (inhours). This assumption may or may not be avalid one. In an effort to address the potential fornon-linearity in the relationship betweenconcentration and duration in determining theeffective dose of sediment (i.e., Effective Dose=Cn·T; where n≠1), Newcombe and Jensen(1995) used multiple regression analysis todevelop severity of effect relationships based onconcentration and duration:

Equation 3: z = a + b(Ln x) + c(Ln y)

This approach, in effect, allows for differentfactors (slopes) to be assigned separately to thevariables of concentration and duration.

In order to explore the relationship betweenconcentration and duration in influencing habitatchange, Anderson et al. (1996) also usedmultiple regression analysis to analyze thehabitat effects database. This analysis identifieda relationship between sediment exposure andhabitat effects that can be described by theequation:

Equation 4: z = 0.637 + 0.740Ln(X)+ 0.864Ln(Y);

r2(adj) = 0.627; n=35; p<0.001.

Statistics for the multiple regression relationshippresented are summarized in Table 1. The “T”statistic for each slope in the regression (Table2) is an expression of the importance of eachvariable with respect to the relationship derived.The higher score attributed to duration indicatesits importance in determination of habitat effects.This indicates that concentration and durationaffect the extent of habitat alteration in dissimilarways or in other words, that the effective dose ofsediment is a function of a non-linearrelationship between the two predictive variables(i.e., Effective Dose =Cn·T; where n≠1).

Table 1Statistics for the Multiple Regression Relationship (Equation 4)

Variable Coefficient

Std. Error Std Coef Tolerance T P(2Tailed)

Constant 0.0637 1.293 0.000 0.493 0.625Ln Con. 0.864 0.176 0.520 0.973 4.903 0.000

LnDuration

0.740 0.111 0.706 0.973 6.652 0.000

Discussion

Confounding Factors

The dose/response relationships that have beendeveloped make generalizations about theanticipated level of effects to the aquaticenvironment that may result from elevatedsediment levels. Since these aregeneralizations, the actual effects that arerealized by a sediment release episode may bemore or less severe based on a number ofconfounding factors.

The potential for adverse effects on fish andtheir habitats associated with sediment releaseis a function of increasing particle size(Newcombe 1996). More information relatingdose-response relationships between specificfish guilds or habitat types as a function ofparticle size range is required in order to developa better understanding of this confoundingfactor.

The angularity or mineralogy of suspendedparticles may play a important role in thepotential for physiological or toxicity effects

(Newcombe and Jensen 1995). The angularityof a particle may be of particular importance withrespect to gill abrasion of fish within the receivingenvironment, and may also influence the rate ofinfiltration of particles into the stream bed.Meanwhile, the mineralogy of the particle may beimportant since the particle itself may have somepotential chemical activity at the cellular level(Newcombe and Jensen 1995). In addition, thepotential for contaminants adsorbed to sedimentparticles is also a concern, since contaminatedsediments could have more dramatic effectsthan those which might be caused by theincrease of sediment load alone.

The amount of material that intrudes into thegravel bed has been shown to be highlydependent on the grain-size distribution of thetransported sediment as well as that of thegravel bed. If the suspended sediment load iscomposed of very fine material, the gravel porestend to fill from the bottom to the top of thepavement layer. If the suspended particles arelarger in size, angular or platelet in shape, a filmcan develop within the substrata which will tendto limit the intrusion of additional sediments intothe interstitial spaces of the stream bed (Beschtaand Jackson 1979). Beschta and Jackson(1979) concluded that the finer the suspended

sediment, the greater the potential was to fillinterstitial voids.

The shape of the stream bed substrata may alsoaffect sediment deposition. Under low flowconditions, rounded stream bed substrata tend toaccumulate more sediment than angularsubstrata, whereas, during high flows, thereverse is true (Meehan and Swanston 1977).This may be due to the reduced turbulencelevels at the gravel bed in rounded stream bedsduring low flows, while at higher discharges aflow separation zone can develop behind angularmaterials causing greater sediment deposition(Reiser et al. 1985).

The temperature of the water can have animpact on the severity of the effects caused by asediment release event. The oxygen holdingcapacity of water and the metabolic andrespiratory rates of fish are influenced by watertemperature. Consequently, the effects ofsediment exposure may be greater inseasonably warm waters than in seasonably coldwater (Newcombe and Jensen 1995). However,during winter conditions aquatic organisms maybe especially vulnerable to additional stressors.Since stress has been defined as the sum of allphysiological responses, the severity of effectsthat are caused by sediment release will be afunction of the level of stress at the time of, orbefore, the period of elevated sediment load.

Factors Influencing the Risk of HabitatAlteration

Many attributes may influence the extent ofsediment-induced aquatic habitat alteration. Theprincipal factor which influences the extent ofhabitat alteration is the increase in sediment loadassociated with watershed disturbance. Inaddition, the sensitivity of the exposed habitatsand the length of time habitats are likely to beimpaired are also important factors whichinfluence the level of habitat alteration.

Considerations regarding the sensitivity of thereceiving environment include the susceptibilityto alteration of the habitats within the receivingenvironment, and the timing of the elevatedsediment loads. In addition to overall sensitivityof the watercourse and the sensitivity of thehabitats it supports, a related consideration isthe species and life stages which may bepresent during times of instream activityassociated with forest harvesting or road

construction. Certain life stages are especiallysensitive to increases in sediment load (such asdeveloping eggs and larvae, or overwintering fish– particularly juvenile salmonid in streams); as aresult, the presence of these life stages duringinstream construction would increase thesensitivity of the watercourse to disturbance.

An additional related consideration regardingconstruction timing sensitivity is the flowconditions within the watercourse during periodsof elevated sediment loads. Watercoursedischarge influences the concentration ofsuspended sediments, transport and depositionof materials as well as the extent of habitatpresent and the ability of resident biota to avoidareas of elevated sediment.

The duration of habitat impairment is one of themost critical consideration in relating the extentof habitat alteration to aquatic community impactsince habitat alteration will only affect the aquaticcommunity if the altered habitat would have beenused during the period of impairment.Therefore, the level of concern associated withhabitat alteration increases as the duration ofimpairment increases. The duration ofimpairment is considered to be the length of timebefore deposited sediments are flushed from thewatercourse into a less sensitive area such as alake and are normally viewed as the number oflife history stages which are impacted during theperiod in which the habitat is in an altered state.

As a result, evaluation of the extent of habitatalteration associated with elevated sediment loadsmust consider the level of sediment load increase(concentration and duration), the nature of thehabitats and communities affected and theduration of likely impairment caused. The severityof effects approach which have been developedby Newcombe and others are not easily applied tothe prediction of habitat change and attributingjustifiable numbers to abstract concepts such assystem sensitivity is difficult at best. Resourcemanagers need to apply expert judgement toensure that models and assumptions are notapplied blindly and that model results do notviolate the most important management tool,which is common sense.

Mitigation Measures

Although most disturbances within a watershedinevitably increase the amount of erosion, thedelivery of sediments to streams resulting from

disturbances can be largely circumvented byproper design and planning. As discussedpreviously, logging road construction andoperation can dramatically increase sedimentloads in streams. The amount of disturbancecaused by road construction and maintenancedepends upon its design standard, gradient, totaldistance of road and intensity of use. Forexample, Megahan and Kidd (1972) indicatedthat proper siting and construction of roads couldeliminate much of the sediment loadingassociated with mass erosion in steep terrain.

The density of logging road distribution can be amajor factor in determining the associatedincrease in sediment loads in streams (Waters1995). Cederholm et al. (1981) documented thegreatest accumulation of fine sediments instreambeds associated with road areas thatexceeded 2.5% of the total basin area. Inaddition, the length of logging roads alsoinfluences sediment delivery to streams.Cederholm et al. (1981) calculated total roadlengths of 2.5 km of road per km2 of watershedbasin produced sediment more than four timesnatural rates. As a result, sediment mitigationmeasures have concentrated on the minimizingsediment associated with logging roads.

The logging system used is a critical factor in thedetermination of road density. For example, theuse of jammer logging systems in the past hasrequired a dense network of roads since theselogging systems require a maximum roadspacing of approximately 150 m. In areas ofsteep terrain, this approach to logging maydisturb soil on 25 percent of the total loggingarea. High lead logging is a method thatreduces the level of disturbance since a reducedroad network is required to support this method.In steep topography areas in Idaho, high leadlogging has reduced disturbance to less than10% of the logging area. On the same types ofslopes, jammer logging roads spaced 60-120 mapart disturbed 25-30% of the total area (Rice et

al. 1974). Skyline logging systems permit aneven wider road spacing of 500 m or more,depending on local conditions and topography.This reduces the area disturbed by 75% andmay provide a greater buffer area for sedimentfiltration between the road and the streamchannel (Megahan and Kidd 1972). Skyline,balloon and helicopter systems have beendeveloped to permit the logging of steeptopography with a minimum amount of roaddisturbances (Rice et al. 1974). Binkley (1965)estimated that skyline yarding would save from2.2 to 2.9 kilometres of road over that requiredfor high lead in a 1600 ha drainage area to belogged.

Control measures to minimize sediment deliveryto streams associated with accelerated erosionwithin watersheds disturbed by forest harvestoperations are of paramount importance inminimizing elevated sediment loads in streams.Since the best mitigation measure to minimizessediment loads in streams is to minimize theamount of erosion in the watershed, designfeatures for access planning can beimplemented which will minimize erosionassociated with logging roads. Waters (1995)summarizes design features for the reduction oferosion from logging roads. This summary tableis reproduced as Table 2.

Disturbance within watersheds inevitablyincreases sediment loads within adjacentwatercourses. These increases in sedimentload can have substantial effect on fish and ontheir habitat. Through the proper planning anddesign of forest harvest systems, the level ofsediments delivered to streams can beminimized allowing for the existence of bothforestry and fish.

Table 2Logging Road Mitigation Measures

(Taken from Water 1995)

Design Feature Method Purposeroad placement avoid streams and steep slopes reduce erosion & mass soil

movementroad length few, short roads reduce total area of exposed roadbedroad width narrow as practicable reduce area of disturbanceroad grade 5-15%, not flat, minimum 3% for drainage avoid rapid run-offroad surface gravel, crushed rock roadbed reduce roadbed erosioncut slopes vertical or near vertical cut reduce excavation and erosion of

slopefill slopes avoid road drainage and woody debris in

fillstabilize fill slopes

road drainage outslope drainage on shallow slopes,inside drainage on steep grades

disperse drainage

inside drainage ditch inside road carry run-off along roadcross-drainageculvert

underground pipe or log construction drain inside ditches or waterways

water bar low hump, 30 o angle downslope disperse run-off from roadbedbroadbased dip wide drainage dip or bench disperse run-off from roadbedstream crossing minimize number, use appropriate

methodreduce direct aquatic impact

vegetation planting seed grass, plant trees reduce exposed surfacedaylighting cut canopy to permit sunlight penetration promote drying of roadbedabandonment close access, remove crossings, install

dips and water barsavoid subsequent use andmaintenance

Acknowledgements

The habitat database and habitat effectsrelationship presented in this paper were firstdeveloped and reported in Anderson et al. 1996.For more information on the database and thedevelopment of the relationship, the reader isdirected to that manuscript. The author wouldlike to acknowledge the work of Waters (1995),and the on-going work of C. Newcombe. Theauthor would like to thank the reviewers of thispaper for their time and insightful comments andto staff of Golder Associates for their assistance.

References

Alabaster, J.S. and R. Lloyd. 1982. Finelydivided solids. Pages 1-20 In J.S.Alabaster and R. Lloyd editors. Waterquality criteria for freshwater fish, 2ndedition. Butterworth, London.

Alexander, G.R. and E.A. Hansen. 1988. Declineand recovery of a brook trout streamfollowing an experimental addition ofsand sediment. Michigan Department ofNatural Resources Fisheries Division.Fisheries Research Report No. 1943.

Anderson, N.H. and D.M. Lehmkuhl. 1967.Catastrophic drift of insects in awoodland stream. Ecology 49: 198-206.

Anderson, P.G., B.R. Taylor and G.C. Balch.1996. Quantifying the effects ofsediment release on fish and theirhabitats. Canadian Manuscript Reportof Fisheries and Aquatic Sciences No.2346: 110 p. + 3 Append.

Auld, A.H. and J.R. Schubel. 1974. Effects ofsuspended sediment on fish eggs andlarvae. Special Report 40, Reference74-12, Chesapeake Bay Institute, JohnHopkins University, Baltimore, Maryland.

Auld, A.H. and J.R. Schubel. 1978. Effects ofsuspended sediment on fish eggs andlarvae. A laboratory assessment.Estuarine and Coastal Marine Science.6:153-164.

Barton, B.A. 1977. Short-term effects ofhighway construction on the limnology ofa small stream in southern Ontario.Freshwater Biology, 7: 99-108.

Berg, L. and T.G. Northcote. 1985. Changes interritorial, gill-flaring, and feedingbehaviour in juvenile coho salmon(Oncorhynchus kisutch) following short-term pulses of suspended sediment.Canadian Journal of Fisheries andAquatic Sciences. 42:1410-7

Beschta, R.L. and W.L. Jackson. 1979. Theintrusion of fine sediments into a stablegravel bed. Journal Fisheries ResearchBoard of Canada. 36(2): 204-210.

Bethlahmy, N. And W.J. Kidd. 1966. Contollingsoil movement from steep road fills.USDA Forest. Service Research NoteINT-45. Intermountain Forest and RangeExperimental Station, Ogden, Utah. 4 p.

Binkley, V.W. 1965. Economics and design of aradio-controlled skyline yarding system.USDA Forest. Service Research PaperPNW-25. Pacific NW Forest and RangeExperimental Station, Portaland,Oregon. 30 p.

Bisson, P.A. and R.E. Bilby. 1982. Avoidance ofsuspended sediment by juvenile cohosalmon. North American JournalFisheries Management, 4:371-374.

Bjornn, T.C., M.A. Brusven, M.P. Molnau, J.H.Milligan, R.A. Klamt, E. Chacho and C.Schaye. 1977. Transport of graniticsediment in streams and its effects oninsects and fish. Office of WaterResearch and Technology, U.S.Department of the Interior. Washington,D.C. 20240. Research TechnicalCompletion Report, Project B-036-IDA.43 p.

Bode, R.W. 1988. Methods for rapid biologicalassessment of streams. StreamBiomonitoring Unit, Bureau of Monitoringand Assessment. Division of Water,NYS Department of EnvironmentalConservation.

Brett, J.R. 1958. Implications and assessmentsof environmental stress. In: P.A. Larkin(editor). Investigation of Fish-PowerProblems. pp. 69-83. Vancouver: H.R.MacMillan Lectures in Fisheries,University of British Columbia.

Buck, D.H. 1956. Effects of turbidity on fish andfishing. Transactions of the 1956 NorthAmerican Wildlife Conference, 21:249-261.

Campbell, H.J. 1954. The effects of siltationfrom gold dredging on the survival ofrainbow trout and eyed eggs in PowderRiver. Fisheries Bulletin, Oregon StateGame Commission, Portland, Oregon. 3pp.

Cederholm, C.J., L.M. Reid, and E.O. Salo.1981. Cumulative effects of loggingroad sediment on salmonid populationsin the Clearwater River, JeffersonCounty, Washington. Pages 38-74 inWashington Water Research Council.1981. Proceedings from the conferenceon salmon-spawning gravel: arenewable resource in the PacificNorthwest. Washington StateUniversity, Washington Water ResearchCentre, Report 39, Pullman.

Clapp, D.F., R.D. Clark, Jr and J.S. Diana. 1990.Range, activity, and habitat of large,free-ranging brown trout in a Michigan(USA) stream. Transactions of theAmerican Fisheries Society. 119:1022-1034.

Cooper, A.C. 1965. The effect of transportedstream sediments on the survival ofsockeye and pink salmon eggs andalevin. Int. Pac. Salmon Fish. Comm.Bull. 18 (71 pp)

Cordone, A.J. and D.W. Kelley. 1961. Theinfluences of inorganic sediment on theaquatic life of streams. California Fishand Game 47: 189-228.

Cunjak, R.A. 1996. Winter habitat of selectedstream fishes and potential impacts fromland-use activity. Canadian Journal ofFisheries and Aquatic Sciences.: 53(Supplement 1).

Cunjak, R.A. and G. Power. 1986. Seasonalchanges in the physiology of brook trout,Salvelinus fontinalis (Mitchell), in a sub-Arctic river system. Journal of FishBiology 29:279-288.

Dance, K.W. 1978. Macroinvertebrates diversityin two streams as affected byagricultural land use. Paper presentedat the annual meeting of the NorthAmerican Benthological Society,Winnipeg, Manitoba.

Eaglin, G.S. and W.A. Hubert. 1993. Effects oflogging and roads on substrate and troutin streams of the Medicine Bow NationalForest, Wyoming. North AmericanJournal of Fisheries Management. 13:844-846..

Elwood, J.W. and T.F. Waters. 1969. Effects offloods on food consumption andproduction rates of a stream brook troutpopulation. Transactions of theAmerican Fisheries Society. 99: 253-267.

Fox, H.L. 1974. Effects of urbanization on thePatuxent River, with special emphasison sediment transport, storage andmigration. Ph.D. J. Hopkins Univ.Baltimore., MD.

Fudge R.J.P., and R.A. Bodaly. 1984.Postimpoundment winter sedimentationand survival of lake whitefish(Coregonus clupeaformis) eggs insouthern Indian Lake, Manitoba.Canadian Journal of Fisheries andAquatic Sciences. 41:701-5.

Furniss, M.J., T.D. Roelofs, and C.S. Yee. 1991.Road construction and maintenance.Pages 297-323 In Meehan, W.R. 1991.Influences of forest and rangelandmanagement on salmonid fishes andtheir habitats. American FisheriesSociety Special Publication 19.

Gammon, J.R. 1970. The effect of inorganicsediment on stream biota. U.S. E.P.A.Water Quality Office, Pollution ControlResearch Service. 141 pp.

Goldes, S.A. 1983. Histological andultrastructural effects of the inert claykaolin on the gills of rainbow trout(Salmo gairdneri Richardson). M.Sc.Thesis, University of Guelph, Guelph,Ontario, 103 p. (cf Newcombe 1994a).

Gregory, R.S., J.A. Servizi, and D.W. Martens.1993. Comment: Utility of the stressindex for predicting suspended sedimenteffects. North American J. FisheriesManagement. 13(4): 868-873.

Griffith, J.S. and R.W. Smith. 1993. Use ofwinter concealment cover by juvenilecutthroat and brown trout in the southfork of the Snake River, Idaho. NorthAmerican Journal of FisheriesManagement 13:823-830.

Grime, J.P. 1989. The stress debate: symptomof impending synthesis? BiologicalJournal of the Linnean Society, 37: 3-17.

Hall, J.D. and R.L. Lantz. 1969. Effects oflogging on the habitat of coho salmonand cutthroat trout in coastal streams.In T.G. Northcote (ed) Symposium onsalmon and trout in streams. Institute ofFisheries, University of British Columbia.pp. 355-375.

Hartman, G.F., and J.C. Scrivener. 1990.Impacts of forest practices on a coastalstream ecosystem, Carnation Creek,British Columbia. Canadian Bulletin ofFisheries and Aquatic Sciences 223.

Haupt, H.F. 1959. Road and slopecharacteristics affecting sedimentmovement from logging roads. Journalof Forestry. 57: 329-332.

Heggenes, J., T.G. Northcote, and P. Armin.1991. Seasonal habitat selection andpreferences by cutthroat trout(Oncorhynchus clarki) in a small, coastalstream. Canadian Journal of Fisheriesand Aquatic Sciences. 48:1364-1370.

Heggenes, J. and S. J. Saltveit. 1990. Seasonaland spatial microhabitat selection andsegregation in young Atlantic salmon,Salmo salar L., and brown trout, Salmotrutta L. , in a Norwegian river. Journal ofFish Biology 36:707-720.

Herbert, D.W.M. and J.C. Merkens. 1961. Theeffect of suspended mineral solids onthe survival of trout. International Journalof Air and Water Pollution. 5(1):46-55.

Hillman, T.W., J.S. Griffith. and W.S. Platts.1987. Summer and winter habitatselection by by juvenile chinook salmonin a highly sedimented Idaho stream.Transactions of the American FisheriesSociety. 116: 185-195.

Holtby, L.B. 1988. Effects of logging on streamtemperatures in Carnation Creek, BritishColumbia and associated impacts on thecoho salmon. Canadian Journal ofFisheries and Aquatic Sciences. 45:502-515.

Hynes, H.B.N. 1973. The effects of sediment onthe biota in running water. Proc. 9thCan. Hydrol. Symp. Fluvial Processesand Sedimentation. Univ. of Alberta.Inland Waters Directorate, Departmentof the Environment. pp 652-663.

Kerr, S.J. 1995. Silt, turbidity and suspendedsediments in the aquatic environment:An annotated bibliography and literaturereview. Southern Region Science andTechnology Transfer Unit. TechnicalReport TR-008.

Klein, R.D. 1984. Effects of sediment pollutionupon the aquatic environment. Dated 30November 1984. MD. UnpublishedManuscript.

Koehn, R.K. and B.L. Bayne. 1989. Towards aphysiological and geneticalunderstanding of the energetics of thestress response. Biological Journal ofthe Linnean Society, 37: 157-171.

Koski. 1966. The survival of coho salmon fromegg deposition to emergence in threeOregon coastal streams. M.Sc. Thesis.Oregon State University.

Lamberti, G.A., S.V. Gregory., L.R. Ashkenas.,R.C. Wildman. and K.M.S. Moore. 1991.Stream ecosystem recovery following acatastrophic debris flow. CanadianJournal of Fisheries and AquaticSciences. 48: 196-208.

Lenat, D.R., D.L. Penrose, and K.W. Eagleson.1981. Variable effects of sedimentaddition on stream benthos.Hydrobiologia 79: 187-194.

Lisle, T.E. and J. Lewis. 1992. Effects ofsediment transport on survival ofsalmonid embryos in a natural stream: Asimulation approach. Canadian Journalof Fisheries and Aquatic Sciences. 49:2337-2344.

Lloyd, D.S. 1987. Turbidity as a water qualitystandard for salmonid habitats in Alaska.North American Journal of FisheriesManagement. 7: 34-65.

Macdonald, D.D. and C.P. Newcombe. 1993.Utility of the stress index for predictingsuspended sediment effects: Responseto comment. North American J.Fisheries Management. 13(4): 873-876.

Mason, J.C. 1974. A further appraisal of theresponse to supplemental feeding ofjuvenile coho (Onchorynchus kisutch) inan experimental stream. Dept. of theEnvironment, Fisheries and MarineService, Research and DevelopmentDirectorate. Nanaimo, B.C. 470.

Mason, J.C. 1976. Response of underyearlingcoho salmon to supplemental feeding ina natural stream. J. Wildl. Manag. 40:775-788.

McLeay, D.J. 1977. Development of a bloodsugar bioassay for rapidly measuringstressful levels of pulpmill effluent tosalmonid fish. Journal of the FisheriesResearch Board of Canada, 34:477-485.(cf Noggle 1978).

McLeay, D.J., I.K. Birtwell, G.F. Hartman, andG.L. Ennis. 1987. Responses of Arcticgrayling (Thymallus arcticus) to acuteand prolonged exposure to Yukon placermining sediment. Canadian Journal ofFisheries and Aquatic Sciences. 44:658-673.

McNeil, W.J. and W.H. Ahnell. 1964. Success ofpink salmon spawning relative to size ofspawning bed materials. U.S. Fish andWildlife, Fisheries Bulletin 63(3):575-588.

Meehan, W.R., editor. 1991. Influences ofForest and rangeland management onsalmonid fishes and their habitats.American Fisheries Society SpecialPublication 19. 751 p.

Meehan, W.R. and D.G. Swanston. 1977.Effects of gravel morphology on finesediment accumulation and survival ofincubating salmon eggs. USDA ForestService Research Paper, PNW-220,Portland Oregon.

Megahan, W.F. and W.J. Kidd. 1972. Effects oflogging roads on sediment productionrates in the Idaho Batholith. U.S. ForestService Research Paper INT-123.

Mende J. 1989. River and streaminvestigations: Portneuf Riverassessment.. Idaho Dept of Fish andGame and Idaho State University.Poccatello, Idaho. F-711-R-12 Sub.3,J4. 72p.

Modde, T., D.R. Young, and D.L. Archer. 1991.Evaluation of factors influencingpopulation characteristics and habitatutilization of trout in the Flaming Borgetailwater. 1987-1989. Division of WildlifeResources, Utah Department of NaturalResources. Publication No. 91-10. 192p. ABST

Newcombe, C.P., and J.O.T. Jensen. 1995.Channel sediment and fisheries: asynthesis for quantitative assessment ofrisk and impact. Habitat ProtectionBranch. British Columbia Ministry ofEnvironment, Lands and Parks, Victoria,British Columbia. Canada. 51 p.

Newcombe, C.P. 1994. Suspended sediments inaquatic ecosystems: ill effects as afunction of concentration and duration ofexposure. Habitat Protection Branch,British Columbia Ministry ofEnvironment, Lands and Parks. Victoria,British Columbia, Canada, 298 p.

Newcombe, C.P. 1996. Channel sedimentpollution: A provisional fisheries fieldguide for assessment of risk and impact.Habitat Protection Branch. Ministry ofEnvironment, Lands and Parks, Victoria,British Columbia, Canada. 59 p.

Newcombe, C.P. and D.D. MacDonald. 1991.Effects of suspended sediments onaquatic ecosystems. North AmericanJournal of Fisheries Management 11:72-82.

Nickelson, T.E., J.D. Rodgers, S.L. Johnson,and M.F. Solazzi. 1992. Seasonalchanges in habitat use by juvenile cohosalmon (Oncorhynchus kisutch) inOregon coastal streams. CanadianJournal of Fisheries and AquaticSciences. 49:783-789.

Phillips, R.W., R.L. Lantz, E.W. Claire and J.R.Moring. 1975. Some effects on gravelmixtures on emergence of coho salmonand steelhead trout fry. Transactions ofthe American Fisheries Society. 3:461-6.

Redding, J.M. and C.B. Schreck. 1980. Chronicturbidity and stress in juvenile cohosalmon and steelhead trout. ReportPNW-1705-16, Pacific Northwest Forestand Range Experiment Station, U.S.Forest Service, Corvallis, Oregon.

Reiser, D.W., M.P. Ramey and T.R. Lambert.1985. Review of flushing flowrequirements in regulated streams. SanRamon, California. Pacific Gas andElectric Company, Dept. of EngineeringResearch. Z19-5-120-84. pp.1-97 + App.

Rice, R.M., J.S. Rothacher and W.F. Megahan.1974. Consequences of timberharvesting: An appraisal. NationalSymposium on Watersheds inTransition. pp. 321- 329.

Ringler, N.H. and J.D. Hall. 1975. Verticaldistribution of seidment and organicdebrs in coho salmon redds in threesmall Oregon Streams. CanadianJournal of Fisheries and AquaticSciences. 45:742-747.

Rosenberg, D.M. and A.P. Wiens. 1978. Effectsof sediment addition on macrobenthicinvertebrates in a northern CanadianRiver. Water Resources 12, 753-763.

Rosenberg, D.M. and N.B. Snow. 1975.Ecological studies of aquatic organismsin the Mackenzie and Porcupine riverdrainages in relation to sedimentation.Canadian Journal of FisheriesManagment. Serv. Res. Dev. Tech. Rep.547.

Salo, E.O., and T.W. Cundy, editors. 1987.Streamside management: forestry andfishery interactions. University ofWashington, Institute of ForestResources Contribution 57, Seattle,Washington.

Saunders, J.W., and M.W. Smith. 1965.Changes in a stream population of troutassociated with increased silt. Journal ofFisheries Research Board of Canada.22(2):396-404.

Scullion, J., and N. Miller. 1979. Review ofeffects of natural and artificial flowpatterns and sediment dynamics onbenthic invertebrates and salmonids inupland rivers. The University of Wales.Institute of Science and Technology.Cardiff. Deptartment of Applied Biology.

Selye, H. 1950. Stress and the generaladaptation syndrome. British MedicalJournal 1950(1):1383-1392.

Servini , J.A. and D.W. Martens. 1987. Someeffects of suspended Fraser RiverSediments on sockeye salmon(Onchorynchus nerka). Can. Spec.Publ. Fish. Aquat. Sci. 96: 254-264.

Servizi, J.A. and D.W. Martens. 1992. Sub-lethal responses of coho salmon tosuspended sediments. Canadian J.Fisheries and Aquatic Sciences, 49(7):1389-1395.

Servizi, J.A. and D.W. Martens. 1991. Effect oftemperature, season, and fish size onacute lethality of suspended sedimentsto coho salmon (Oncorhynchus kisutch).Canadian Journal of Fisheries andAquatic Sciences. 48:493-497.

Shaw, P.A. and J.A. Maga. 1943. The effect ofmining silt on yield of fry from salmonspawning beds. California Fish andGame. 29(1):29-41.

Sherk, J.A, O'Connor JM, Neumann DA. 1973.Effects of suspended and depositedsediments on estuarine environments.Estuarine Research. 2:541-58. In: L.E.Cronin (ed). Estuarine Research.Academic Press, NewYork.

Sigler, J.W., T.C. Bjornn and F.H. Everest.1984. Effects of chronic turbidity ondensity and growth of steelheads andcoho salmon. Transactions of theAmerican Fisheries Society. 113:142-50.

Simmons, R.C. 1984. Effects of placer miningsedimentation on Arctic grayling ofinterior Alaska. University of Alaska,Fairbanks: 1-75 pp.

Singleton, H.J. 1985. Water quality criteria forparticulate matter. British ColumbiaMinistry of Environment. ResourceQuality Section, Water ManagementBranch. Victoria, B.C.

Slaney, P.A., T.G. Halsey and H.A. Smith. 1977.Some effects of forest harvesting onsalmonid rearing habitat in two streamsin the central interior of British Columbia.Fisheries Management Rept. #71,Ministry of Recreation and Conservation.

Slaney, P.A., T.G. Halsey and A.F. Tautz. 1977.Effects of forest harvesting practices onspawning habitat of stream salmonids inthe Centennial Creek watershed, BritishColumbia. Fisheries ManagementReport. #73.

Smith, M.W. and J.W. Saunders. 1958.Movements of brook trout (Salvelinusfontinalis) between and within fresh andsalt water. Journal of the FisheriesResearch Board of Canada, 15:1403-1449.

Smith, O.R. 1940. Placer mining silt and itsrelation to salmon and trout on thePacific coast. Transactions AmericanFisheries Society, 69: 225-230.

Stober, Q.J., B.D. Ross, C.L. Melby, P.A. Dinnel,T.H. Jagielo, and E.O. Salo. 1981.Effects of suspended volcanic sedimenton coho and chinook salmon in theToule and Cowlitz Rivers. TechnicalCompletion Report. FRI-UW-8124,Fisheries Research Institute, Universityof Washington, Seattle, Washington.147 pp.

Stuart, T.A. 1953. Water current throughpermeable gravels and their significanceto spawning salmonids. Nature. 172:407-408.

Sweeten, J. 1995. Sediment working groupformed to tackle top water pollutant.Fisheries.

Tebo. L.B. 1955. Effects of siltation, resultingfrom improper logging, on the bottomfauna of a small trout stream in thesouthern Appalachians. ProgressiveFish Culturalist April 1955: 64-70.

Tschaplinski, P.J. and G.F. Hartman. 1983.Winter distribution of juvenile cohosalmon before and after logging inCarnation Creek, British Columbia, andsome implications for overwintersurvival. Canadian Journal of Fisheriesand Aquatic Sciences 40: 452-461.

Tsui, P.T.P. and P.J. McCart. 1981. Effects ofstream-crossing by a pipeline on thebenthic macroinvertebrate communitiesof a small mountain stream.Hydrobiologia, 79: 271-276.

Vondracek, B. and D.R. Longanecker. 1993.Habitat selection by rainbow troutOncorhynchus mykiss in a Californiastream: Implications for the instreamflow incremental methodology. Ecologyof Freshwater Fish 2: 173-186.

Wagener, S.M. 1984. Effects of gold placermining on stream macroinvertebrates ofinterior Alaska. University of Alaska,Fairbanks; 1-100 pp.

Wagener, S.M. and J.D. LaPerriere. 1985.Effects of placer miner on theinvertebrate communities of interiorAlaska streams. Freshwater InvertebrateBiology, 4: 208-214.

Waters, T.F. 1964. Interpretation of invertebratedrift of stream invertebrates. Ecology,43: 316-320.

Waters, T.F. 1995. Sediment in streams:sources, biological effects and control.American Fisheries Society Monograph7. Bethesda, Maryland.

Weaver, T.M, Fraley J.J. 1993. A method tomeasure emergence success ofwestslope cutthroat trout fry from varyingsubstrate compositions in a naturalstream channel. North American Journalof Fisheries Management. 13:817-22.

Wedemeyer, G.A. , B.A. Barton, and D.J.McLeay. 1990. Stress and acclimation.pages 451-489. in C.B. Schreck andP.B. Moyle, editors. Methods for fishbiology. American Fisheries Society,Bethesda, Maryland.

Wickett, W.P. 1959. Effects of siltation onsuccess of fish spawning. p16-22 In 5thSymposium on Water PollutionResearch, Pacific Northwest Siltation -It’s Source and Effects on the AquaticEnvironment, U.S. Public HealthService, Portland, Oregon.

Wickett, W.P. 1954. The oxygen supply tosalmon eggs in spawning beds. Journalof the Fisheries Research BoardCanada. 11(6): 933-953.

Williams, D.D. and H.B.N. Hynes. 1976. Therecolonization mechanisms of streambenthos. Oikos 27: 265-272.

Young, R.J. 1986. The effect of winter oilpipeline construction on the benthicmacroinvertebrate community structureof Hodgson Creek, NWT. M.Sc. thesis,Univ. of Guelph, Ontario. December1986. 76 p.

Zallen, M. 1931. Effects of pipeline constructionon juveniles and incubating eggs ofmountain whitefish (Prosopiumwilliamsoni Girard) in the Moyie River,British Columbia. [Unpublished]

Zelt, B.W. Concentration fluctuations in healthrisk assessment. Presented at 1995CPANS Annual General Meeting andTechnical Conference, June 6-7.Edmonton Convention Centre.Edmonton, Alberta, Canada.