Type 5 Streams and Small Wetlands Literature Review

Return to the: DNR Homepage or Type 5 Stream Literature Review (main DNR site)

How do Type 5 streams interact with upland and aquatic ecological functions to maintain the integrity of western Washington forests?

Statements within this literature review are based on the reported findings from various published sources. Due to differences in sample design, methodology, geographical location, site characteristics, data analysis and interpretation, these statements may or may not agree with the results of other published reports. Despite contradictory results, one cannot say the findings of one author are more or less valid than another, due to the wide range of site characterists. Readers are urged to consider how and where information was collected when interpreting the value of the following conclusions. Headwater stream science is, in many ways, a relatively new field. Early observations and case studies may or may not represent widespread relationships.

Click on the citation following each statement to view the annotated bibliography

Organize statements by:

  • This study found that decidious trees had a "higher invertebrate mass per stem" than conifers. Taxa composition varied considerably between plant species. (Allan et al. 2003).


  • Lepidopteran species were strongly associated with decidious vegetation. (Allan et al. 2003).


  • Diptera species predominated in the hemlock stands. (Allan et al. 2003).


  • Plecoptera species were more abundant on spruce and hemlock. (Allan et al. 2003).


  • Stream heating depends on the following factors: daily average solar insolation, local air temperature, shade by riparian vegetation, air velocity, relative humidity, and groundwater input. Groundwater plays an important role in lowering the temperature in small streams. (Adams and Sullivan 1989).


  • Instream production of algae and invertebrates contribute to terrestrial food webs and feed many organisms such as, grasshoppers, dippers, ducks, and water shrews. (Baxter et al. 2005).


  • The diets of winter wrens (Troglodytes troglodytes) consist mainly of emergent insects. (Baxter et al 2005).


  • This study found that channel widths were greater in the 1990's than in the 1860's. This is due to loss of root strength in harvested riparian areas, which increased bank erosion and sediment loading. (Beechie et al. 2000).


  • Not only do debris flows fill the channel creating a gravel-bed morphology, but they can also scour the channel. This results in a mixed bedrock and boulder-bed morphology. (Benda 1990).


  • Debris flows initiate in bedrock hollows. They transport sediment and organic debris, to first and second order streams. (Benda 1990).


  • Debris flows occur in first and second order channels at approximately 750-1500 year intervals. (Benda and Dunne 1987).


  • Landslides occur in bedrock hollows at 6,000 year intervals. (Benda and Dunne 1987).


  • Small or large storms do not affect the rate of flow through a forested wetland, because groundwater gradients and their corresponding velocities are not different after storm events. Gravitational flow, dictated by gradient, is the main force that moves water in the Pacific Northwest. (Berlin 2000).


  • Harvesting around stream channels can produce unsuitable habitat for hyporheic invertebrates, by increasing sediment that destroys invertebrate habitat, and decreasing dissolved oxygen in the hyporheic zone. (Boulton et al. 1997).


  • After harvesting, diurnal soil temperature variations increased in the harvested areas. Upslope soil temperatures influence groundwater temperatures more than near stream soil temperatures. (Brosofske et al. 1997).


  • Increased wind speeds in narrow buffers are related to decreased humidity levels. (Brosofske et al. 1997).


  • The most common sites for dam-break flows are in steep narrow, low order streams. (Coho and Burges 1991).


  • First- and second-order channels store mass wasting events, and release them through episodic debris flows. (Coho and Burges 1994).


  • Post harvest woody-debris loading, supports high densities of detritivores for many years after harvest. Depletion of these allochthonous materials may not happen until decades after the initial cut. (Cole et al. 2003).


  • Amphibian species richness was greater in the uncut forests than in the harvested forests. (Corn and Bury 1989).


  • Harvesting upstream from the un-cut area did not effect the presence, density, or biomass for any of the amphibian species studied. (Corn and Bury 1989).


  • Dicamptodon tenebrosus (Pacific Giant salamanders) and Rhyacotriton olympicus (Olympic salamanders) were positively correlated with stream gradients in logged stands. However, this was not the case in un-logged stands. This suggests that disruptive effects, possibly from harvesting, can increase sediment yields in low gradient streams. (Corn and Bury 1989).


  • Logged ephemeral streams, were found to have two to three times more fine sediment than in the un-logged streams. Fine sediment infiltration in ephemeral streams is enhanced by logging in steep terrain. (Davies and Nelson 1993).


  • Catchment disturbance due to forestry, and related practices may account for elevated sediment levels downstream. (Davies and Nelson 1993).


  • Materials eroded from the uplands are stored over time in hollows, and then suddenly released as a debris flow, which may affect downstream processes. (Dietrich 1989).


  • Runoff from roads can increase the frequency of landslides from hollows. (Dietrich 1989).


  • The author speculates that there is a positive correlation between litterfall, and wind speed at the treated sites. (Grady 2001).


  • Sediment levels increased in the harvested watersheds due to the following conditions: poorly designed roads, large storm events, surface erosion after burning, and continous mass wasting events in the upper parts of the watersheds. (Grant and Wolff 1991).


  • The amount and type of organic matter that enters the stream channel is dependent upon the composition of the riparain zone. (Gregory et al. 1991).


  • 17% of the uprooted trees delivered sediment to the stream channels. (Grizzel and Wolff 1998).


  • 75% of the in-stream large woody debris was recruited after harvest, and was suspended above the bankfull channel. (Grizzel and Wolff 1998).


  • Smaller pieces of woody debris are more mobile during high stream flows. (Grizzel et al. 2000).


  • Streams with second growth buffers are more prone to windthrow than old-growth stands. This negates the long-term recruitment potential of woody debris. (Grizzel et al. 2000).


  • The woody debris recruited by second growth usually has a longer mean length and smaller mean diameter, than downed wood from old-growth forests. (Grizzel et al. 2000).


  • This study found that instream woody debris recruitment was meeting frequency target levels, but the size of this debris is below target levels. (Grizzel et al. 2000).


  • Although this study reports that 50% of recruited wood originates within 15m of the stream bank, they state that narrow buffers limit woody debris recruitment. (Grizzel et al. 2000).


  • Woody debris recruitment frequency was approximately 14 tree/pieces per 100m for 20-30m buffers, and 8.1 tree/pieces per 100m for buffers less than 20m. (Grizzel et al. 2000).


  • Two factors that affect water temperature are: groundwater inflow, and the stream source. (Hagan 2000).


  • In first-order streams, macroinvertebrate food sources mainly consist of detritus from woody material. "Shredder" populations were most abundant in these streams because of the large allocthonous input of woody debris from uplands. (Haggerty et al. 2002).


  • Invertebrate "collector" populations increased after harvesting. "Shredder" populations also climbed after harvest, probably due to availability of organic materials from instream slash buildup. (Haggerty et al. 2004).


  • When slash is covering the stream channel, algal growth is limited because of low water temperatures and lack of light for photosynthesizing. This inhibits the growth of "scraper" populations. (Haggerty et al. 2004).


  • In the year following harvest, organic sediment was greater in the clearcut streams, but then went back to equilibrium in the following year. In the buffer treatments, organic sediment rose gradually and steadily in the year following harvest. In the first year of post-treatment, inorganic sediments decreased in clearcuts, but went up significantly in buffers and then decreased slightly two years after harvest. (Haggerty et al. 2004).


  • Wetland trees that are associated with seeps are: Alnus rubra (red alder), Fraxinus latifolia (Oregon ash), and Populus balsamifera spp. trichocarpa (black cottonwood). Plants that are associated with seeps are: Hepaticae spp. (leafy liverworts), Lysichiton americanum (skunk cabbage), Oplopanax horridus (devil's club), and Tolmiea menziesii (piggy-back). Seeps were relatively close to the stream, however seep formation increases as side-slope elevation decreases. The local geology also affected the location of the seeps. (Hayes and Quinn 2001).


  • Older stands had a greater diversity of amphibians than younger stands. (Hayes and Quinn 2001).


  • Connected seeps, offer a more favorable habitat for amphibians than unconnected seeps; possibly because these seeps offer less risk of desiccation to water-dependent amphibians, such as torrent salamanders (Rhyacotriton spp.). (Hayes and Quinn 2001).


  • The Columbia torrent salamanders appeared in streams, and seeps with greater than 18% gradient, regardless of substrate type. (Hayes and Quinn 2001).


  • Fallen woody debris provides step structures, which aid in the reduction of stream flow, and sediment discharge. The streams also use these step structures to adjust to gradient change. (Heede 1972).


  • Drainage density (mile of channel per square mile of basin area), and valley widths of non-perennial streams can vary significantly as a function of underlying geology. (Henderson and DeWalle 2000).


  • Channel scour widths of non-perennial streams were not found to be significantly different in various geologic strata. (Henderson and DeWalle 2000).


  • This study reveals that there is little difference in the vegetative patterns of buffer strips in comparison to undisturbed riparian areas. (Hibbs and Bower 2001).


  • The water yield increased in the first eight years after harvest, but was less than normal for 18 of the 19 years on record. This study hypothesizes that summer low flows will go back to pre-treatment levels when a coniferous canopy replaces a decidious canopy. This may take 40-60 years after harvest. (Hicks et al. 1991).


  • The major source of wood delievered to these low order streams is caused by disease, windthrow, and mortality. (Jackson and Sturm 2001).


  • Slash in the un-buffered streams trapped fine sediment by inhibiting fluvial transport. This interception increased the percentage of fines from 12 to 44%. The increase of fine sediments in headwater streams can be detrimental to certain populations of amphibian species. (Jackson et al. 2001).


  • Landslide disturbances were typically followed by an increase in alder colonization in uplands (providing there is a seed source). Cable logging and road building are other disturbances that can increase the percent cover of alder. (Johnson and Edwards 2002).


  • Confined channels are less likely to be colonized by alders after landslides because the disturbance zone is more narrow (<25m) and incised. Unconfined channels tend to have more alders because the surface area of the disturbance zone is wider and more inhabitable. (Johnson and Edwards 2002).


  • Dicamptodon tenebrosus salamanders had a greater range of movement in forested streams. Compared to other amphibian species, D. tenebrosus living in clearcut streams stayed closer to the channel. (Johnston and Frid 2002).


  • Over the past 50 years, forest harvest has increased peak flows by as much as 50% in small basins, and 100% in large basins. This may be due to flow changes caused by roads, and/or vegetation removal. (Jones and Grant 1996).


  • This paper discusses the comments that were made in the 1998 paper by Thomas and Megahan which covers the contradictory findings, and decisions concerning sample sizes, significant statistical levels, and data analysis. This study concludes that both Thomas and Megahan AND Jones and Grant show that forest harvest increases peak discharges by 50% in small basins, and 100% in large basins. (Jones and Grant 2001).


  • Found 7-14 times more algal biomass in headwater streams in the logged sites, than the control sites. (Kiffney and Bull 2000).


  • The logged site had a 2-4 times increase of inorganic mass, and fine sediment in the periphyton mat, decreasing grazer abundance. (Kiffney and Bull 2000).


  • Narrower buffers had increases in maximum water temperatures, photosynthetically active radiation, chlorophylla abundance, periphyton ash free dry mass, periphyton inorganic mass, and chironomid abundance. (Kiffney et al. 2003).


  • Average solar flux was greatest in the clearcut (58x) compared to the control. (Kiffney et al. 2003).


  • In comparison to the controls, periphyton inorganic mass was 4x greater in the 30m buffer and 9x greater in the 10m buffer and no buffer streams. (Kiffney et al. 2003).


  • Chironomidae (midges) and Ephemeroptera (mayfly) abundance generally increased as buffer width decreased. (Kiffney et al. 2003).


  • Road building and maintenance, was found to result in larger inputs of inorganic sediments to headwater streams, than timber harvesting itself. (Kreutzweiser et al. 2001).


  • Rhyacotriton cascadae (Cascade torrent), and Plethodon dunni (Dunn's) salamanders were positively correlated with water, and negatively correlated with riparian overstory (especially coniferous), elevation, and wood cover. (Lee 1997).


  • The young forests had a lower species richness, and a greater affect on the total amphibian density, and biomass. (Lee 1997).


  • Rhyacotriton cascadae (Cascade torrent salamander)) were more abundant in units with cobble substrates, and less abundant on sites with fine sediment, gravel, boulder, and bedrock. (Lee 1997).


  • Plethodon dunni (Dunn's) salamanders were the dominant species in the second growth forests, while Rhyacotriton cascadae (Cascade torrent) were more dominant in the old-growth forests. (Lee 1997).


  • Generally, roads contribute far greater sediment yields to the stream, than does vegetation removal. Better road construction and maintenance techniques can mitigate this. (MacDonald and Ritland 1989).


  • This publication reports that when timber harvesting coincided with road building in western Washington, annual sediment rates increased 10-30 times, in small watersheds. The sediment yields returned to background levels after 3 years. (MacDonald and Ritland 1989).


  • Vegetation removal impacts soil moisture by reducing or eliminating evapotranspiration, and alters infiltration rates by compaction of soils from equipment. (MacDonald and Ritland 1989).


  • The results from this study show that stream temperatures went from an average range of 1.0-1.3°C before treatment, to 2.0-3.0° C after treatment. The effect was especially pronounced in the low retention areas. (MacDonald et al. 2003).


  • Higher retention of riparian trees is correlated to better thermal protection of small streams. (MacDonald et al. 2003).


  • The volume of wood found in the stream was positively correlated with the time since the last debis flow. (May 2002).


  • The main mechanisms for woody debris recruitment are: slope instability and windthrow. (May and Gresswell 2003).


  • Small streams in steep locations are more prone to mass wasting, which increases the potential to recruit wood from further upslope. (May and Gresswell 2003).


  • 11% of the total coarse woody debris originated within 1m of the channel, and 70% originated within 20m, regardless of stream order. However, stands with taller trees (old growth conifers) contributed coarse woody debris from greater distances than stands with shorter trees. (McDade et al. 1990).


  • Macroinvertebrate diversity seemed to be the same in both the debris torrent impacted streams, and the non-impacted streams. (McHenry 1991).


  • The macroinvertebrate populations were 75% greater in the non-impacted debris torrent streams, than in the impacted streams. (McHenry 1991).


  • Road drainages can affect erosion processes, and the length of the channel network. At a study site in Oregon, the road drainage was discharged into the head of a hollow causing landslides, and accelerating sediment transport. (Montgomery 1994).


  • Road drainages from ridgetops can cause landslides, and can integrate the channel and the road network together. (Montgomery 1994).


  • Channels begin at the first point downslope from a drainage divide, where there is enough gradient to form a recharge area that can support a channel. (Montgomery and Dietrich 1988).


  • Channels initiate in recharge and high gradient areas. (Montgomery and Dietrich 1988).


  • Channel heads on steep slopes form by subsurface flow from unstable colluvial fill. (Montgomery and Dietrich 1989).


  • Channel heads are controlled by hillslope processes. Channel heads on steep gradients are caused by landslides. On gentler slopes, channel heads form from overland flow, and seepage erosion. (Montgomery and Dietrich 1989).


  • Due to the steep topography of the Pacific Northwest, timber harvesting increases the frequency of shallow rapid landslides. (Montgomery et al. 2000).


  • Woody debris is more important in small, steep streams than was previously thought. (Moore and Richardson 2003).


  • After timber harvest, high gradient streams have less fine sediment deposits, and more coarse sediment accumulation than low gradient streams. (Murphy and Hall 1981).


  • Clear-cutting increases productivity (increasing invertebrate biomass, density, and species richness) within streams. However, this increase declines 10-20 years after the understory is re-established. (Murphy and Hall 1981).


  • This study demonstrated that streams in disturbed watersheds had a greater range of water temperatures, than those in undisturbed watersheds. The treatment streams had water temperature fluctations between 1.2-15.4°C, while the control stream ranged from 5.5--12.1°C. Although the treated streams had higher maximum temperatures, they were still within the limits for water quality standards. (Murray et al. 2000).


  • Red alder (Alnus rubra) may contribute to higher instream nitrate levels. (Murray et al. 2000).


  • In coastal areas, instream sodium and chlorine levels are often elevated because of oceanic aerosol deposits. Canopy cover is a factor for determing how much sodium and chlorine will enter the system. (Murray et al. 2000).


  • Events such as windthrow, landslides, debris flows, and timber harvest deliever coarse woody debris (CWD) into first and second order streams. However, CWD transport depends mainly on debris flows. (Nakamura and Swanson 1993).


  • The solid pieces of coarse woody debris tend to be suspended above the stream. (Nakamura and Swanson 1993).


  • Large woody debris (logs and rootwads) is necessary to control scouring of stream beds, and sediment deposition. (Naiman et al. 2000).


  • Large woody debris also acts as a site for vegetation establishment. (Naiman et al. 2000).


  • In natural ecosystems, large woody debris inputs usually occur infrequently, resulting from catastrophic events, such as windstorms, floods, fires, and landslides. (Naiman et al. 2000).


  • Stream discharge can be modified when evapotranspiration is reduced due to harvesting activities. (Naiman et al. 2000).


  • Streams with riparian zones, influence air temperatures up to 60m away from the channel; either by direct cooling or by supplying water for evapotranspiration by vegetation. (Naiman et al. 2000).


  • Alders were most common nearest the stream bank. Conifers were more common at increasing distances from the stream. Alders had the same frequency in both areas, but conifers dominated in the uplands. (Nierenberg and Hibbs 2000)


  • The researchers found that 70% of the downed woody debris was in late stages of decay (between decay classes 3-5). They note that low recruitment of large woody debris may become a habitat issue in the near future. (Olson et al. 2000).


  • Light levels were very low on these study streams, which is a limiting factor in the development of the understory. (Olson et al. 2000).


  • This study found that the following 5 bird species were unique to the upland habitat: Cedar waxwing (Bombycilla cedrorum), dark-eyed junco (Junco hyemalis), Townsend's warbler (Dendroica townsendi x D. occidentalis), band-tailed pigeon (Columba fasciata), and hermit thrush (Catharus guttatus). (Pearson and Manuwal 2001).


  • The following four species were more common in the riparian zone, than in the uplands: American robin (Turdus migratorius), black-throated gray warbler (Dendroica nigrescens), Pacific slope flycatcher (Empidonax difficilis), and winter wren (Troglodytes troglodytes). These species are associated with decidious trees and berry producing shrubs. (Pearson and Manuwal 2001).


  • Canopy type greatly affects the "quantity and biomass of macroinvertebrates exported from headwaters to downstream habitats." (Piccolo and Wipfli 2002).


  • Red alder (Alnus rubra) can link aquatic and terrestrial systems, in headwaters and uplands, to downstream habitats. (Piccolo and Wipfli 2002).


  • Alder canopies export more macroinvertebrates downstream, than clearcuts do. (Piccolo and Wipfli 2002).


  • The disturbed stream had 50% more sediment in the interstitial spaces of the pools and riffles, than was found in the control stream. (Pond 2000).


  • The disturbed stream had a lower invertebrate richness, density, and diversity, than in the control stream. (Pond 2000).


  • Stream size, persistence, and canopy cover are better predictors of invertebrate abundance than measurements of organic detritus, and algal biomass. (Price et al. 2003).


  • The American dipper (Cinclus mexicanus) was more abundant in cutover, young sites. It was concluded that this species prefers wider streams, steeper gradients, and higher elevations. (Raphael et al. 2002).


  • In first- and second-order streams, 80% of the coarse woody debris (CWD) was suspended above the channel or beside the channel. However, in larger streams less than 40% of the CWD was found in this same position. (Robison and Beschta 1990).


  • Root wad presence doubled the frequency of pool formation in small streams. (Rosenfeld and Huato 2003).


  • In small streams, short and long pieces of woody debris function equally. (Rosenfeld and Huato 2003).


  • Rhyacotriton kezeri (Columbia Torrent) salamanders are found more often on north facing aspects because the habitat is cooler and moister. (Russell et al. 2004).


  • On average, species richness was not significantly different in riparian or upland habitats. (Sabo et al. 2005).


  • In natural settings, hillslope processes of soil movement are more often redistributed on the slope, rather than delivered to the stream channel. (Swanson and Fredriksen 1982).


  • In low order channels, instream large woody debris functions as a long term storage site for sediment deposits. (Swanson and Fredriksen 1982).


  • Significant increases in sediment inputs can occur in headwater channels following logging disturbances. (Swanson et al. 1987).


  • An accelerated rate of debris flows often occur after road building and clearcutting. Clearcuts can increase landslides by two to four times as much as forested areas. Roads can increase slide erosion by several hundred times compared to intact forested zones. (Swanson et al. 1987).


  • Slope failures have the potential to move more than 10,000 cubic meters of debris at a speed of 10 m per second. If and when this happens, channel morphology and riparian vegetation are greatly altered. (Swanson et al. 1987).


  • Conducted a re-analysis of the methods, and data from Jones and Grant (1996), and found that there was no effect on peak flows in one large basin, and the results were statistically inconclusive in the other two. They concluded that forest roads have no effect on peak flows in the small watershed treatments. (Thomas and Megahan 1998).


  • Thomas and Megahan concur with Jones and Grant, that timber harvest practices can increase peak flows by 100% for small events, but do not agree that large storm events effect peak flows in the same manner. They reiterate that Jones and Grant's data on large basins remains inconclusive, and provides no evidence that forest roads increase peak flows. (Thomas and Megahan 2001).


  • This study found that when leaf litter is excluded from the stream for an extended period of time, predator and prey populations decrease, due to the disturbance in the ecological food web. (Wallace et al. 1997).


  • Plant diversity and abundance is greater in the riparian area immediatly adjacent to the stream bank, than it is in the uplands. (Waters et al. 2001).


  • Plant associations, as well as vertebrate abundance, is related to stream size. (Waters et al. 2001).


  • Disturbed streams export more particulate organic matter, especially during storm events. (Webster et al. 1990).


  • In Maine, headwater streams were found to have similar plant communities as adjacent upland areas. (Whitman and Hagan 2000).


  • This research found that herbaceous species can survive and quickly recolonize in clearcut areas. (Whitman and Hagan 2000).


  • After harvest, wetland plant species can temporarily occur in clearcuts because the water table may rise. This is referred to as a "watering up" effect. This short term colonization happens frequently in skidder ruts. (Whitman and Hagan 2000).


  • Larval Ascaphus truei (Tailed frogs) were found only in streams in basalt lithologies and at elevations greater than 300m. (Wilkins and Peterson 2000).


  • Plethodon vandykei (Van Dyke salamander) occupied areas where the streams traversed through basalt lithologies on north facing slopes. (Wilkins and Peterson 2000).


  • Plethodon dunni (Dunn's salamanders) inhabited areas of high gradients and steep sideslopes. (Wilkins and Peterson 2000).


  • Larval Ascaphus truei (tailed frogs) were found only in basalt streams at elevations above 300 m. (Wilkins and Peterson 2000).


  • Streams flowing through basalt lithology had twice the amount of Dicamptodon tenebrosus (giant salamanders) than those in marine sediment. (Wilkins and Peterson 2000).


  • The occurrence of headwater amphibians in second growth forests depends on landform characteristics and basin lithology. (Wilkins and Peterson 2000).


  • The population of Rhyacotriton spp. (Torrent salamanders) increased as channel gradient increased and basin area decreased. (Wilkins and Peterson 2000).


  • This study documented that more detritus and terrestrial invertebrates were exported from alder dominated headwater streams, than from conifer dominated streams. (Wipfli and Musslewhite 2004).


  • Of all the invertebrates sampled, 1/4 were of terrestrial origin. (Wipfli and Musslewhite 2004).



Home Literature Search Focus Questions Research Overview What are Type 5 waters?