Effects of Roads on Aquatic Habitat

The effects of roads on aquatic habitat are believed to be widespread and profound. Mechanisms for these effects include: (1) movement of fine sediment, (2) changes in streamflow, (3) changes in water temperature caused by loss of shade cover or conversion of groundwater to surface water, (4) migration barriers, (5) vectors of disease, (6) introduction of exotic fishes, (7) changes in channel configuration from encroachment, and (8) increased fishing pressure. At the landscape scale, correlative evidence suggests that roads are likely to influence the frequency, timing, and magnitude of disturbance, which are likely to influence the structure of aquatic communities.

Effects of Fine Sediment

Increases in fine sediment in stream gravel has been linked to decreased fry emergence, decreased juvenile densities, loss of winter carrying capacity, and increased predation of fishes. Increased fine sediment can reduce benthic organism populations and algal production. Survival of incubating salmonids from embryos to emergent fry has been negatively related to the proportion of fine sediment in spawning gravels (Chapman 1988, Everest and others 1987, Scrivener and Brownlee 1989, Weaver and Fraley 1993, Young and others 1991). Increased fine sediment in stream gravel can reduce intragravel water exchange, thereby reducing oxygen concentrations, increasing metabolic-waste concentrations, and restricting movements of alevins (Bjornn and Reiser 1991, Coble 1961, Cordone and Kelley 1960). Survival of embryos relates to dissolved oxygen and apparent velocity of intragravel water, and to gravel permeability and gravel size (Chapman 1988, Everest and others 1987). Consequently, juvenile salmonid densities decline as fine sediment concentrations increase in rearing areas (Alexander and Hansen 1986, Bjornn and others 1977, Chapman and McLeod 1987, Everest and others 1987, Shepard and others 1984).

Increases in fine sediment also can reduce winter carrying capacity of streams by destroying concealment cover (Bjornn and others 1977, Chapman and McLeod 1987, Thurow 1997) increasing the likelihood of predation (Chapman and McLeod 1987). Pools function as resting habitats for migrating adults, rearing habitats for juveniles (Bjornn and Reiser 1991), and refuges from natural disturbances (Sedell and others 1990). Pools that lose volume from sediment (Jackson and Beschta 1984, Lisle 1982) support fewer fish (Bjornn and others 1977), and fish that reside in them may suffer higher mortality (Alexander and Hansen 1986). Similarly, populations of tailed frogs can be severely reduced or eliminated by increased sedimentation (Corn and Bury 1989, Welsh 1990), presumably because of their dependence on unembedded interstitial areas in the stream substrate where they hide and overwinter (Brown 1990, Daugherty and Sheldon 1982). Increased sediment reduces populations of benthic organisms by reducing interstitial spaces and flow and by reducing algal production, the primary food source of many invertebrates (Chutter 1969, Hynes 1970).

See also: Geomorphic Effects.

Effects of Barriers to Migration

Improper culvert placement at roadstream crossings can reduce or eliminate fish passage (Belford and Gould 1989), and road crossings are a common migration barrier to fish (Clancy and Reichmuth 1990, Evans and Johnston 1980, Furniss and others 1991). In a large river basin in Washington State, 13 percent of the historical coho habitat was lost as a result of improper culvert barriers (Beechie and others 1994). Roads built adjacent to stream channels pose additional effects.

Effects of Temperature Change

Changes in temperature and light regimes resulting from removing the riparian canopy can have both positive and negative effects on fish populations. Sometimes increased food availability can mitigate negative effects of increased summer water temperatures (Bisson and others 1988). Beschta and others (1987) and Hicks and others (1991) document negative effects, including elevation of stream temperatures beyond the range of preferred rearing, inhibition of upstream migrations, increased disease susceptibility, reduced metabolic efficiency, and shifts in species assemblages.

Effects of Altered Streamflow

The size, timing, duration, and frequency of streamflow changes also strongly influence salmonid  reproductive success and overwintering survival (McFadden 1969). For example, high flows after spawning can wash out eggs or displace fry (Latta 1962, Mortensen 1977, Shetter 1961). The effect of roads on peak flows is relatively modest, and the issues of changing stability and predictability because of roads may be of little importance to aquatic habitat suitability.

See also: Hydrologic Effects

Effects of Stream Crossings

Road-stream crossings have effects on stream invertebrates. Hawkins and others (in press) found that the aquatic invertebrate species assemblages (observed versus expected, based on reference sites) were related to the number of stream crossings above a site. Total taxa richness of aquatic insect larvae including mayflies, (Ephmeroptera), stoneflies, (Plecoptera), and caddisflies (Trichoptera) were negatively related to the number of stream crossings. Another study (Newbold and others 1980) found significant differences between macroinvertebrate assemblages above and below road-stream crossings.

Effects of Road Density

Several studies at broad scales document aquatic habitat or fish density changes associated with road density or indices of road density. Eaglin and Hubert (1993) show a positive correlation with numbers of culverts and stream crossings and amount of fine sediment in stream channels, and a negative correlation with fish density and numbers of culverts in the Medicine Bow National Forest. Macroinvertebrate diversity negatively correlates with an index of road density (McGurk and Fong 1995). Increasing road densities are associated with decreased likelihood of spawning and rearing of nonanadromous salmonids in the upper Columbia River Basin, and populations are negatively correlated with road density (Lee and others 1997).