Influence of flow on the biota

Field studies that compare the abundances of macroinvertebrates to hydraulic variables measured at the typical sampling scale of 0.1m2 or less provide strong support for the influence of the flow environment on benthic organisms. In the Ardeche River, a tributary of the River Rhone in southern France, Merigoux and Doledec (2004) collected invertebrates from microhabitats and estimated hydraulic conditions at sampling locations using FST hemispheres. Some dependence of distribution on hydraulic parameters was found in nearly 70% of the taxa analyzed (54 in spring, 31 in autumn), whereas others (the mayfly Epeorus in spring, the stonefly Leuctra in autumn) appeared indifferent to the range of hydraulic conditions. The ordination of the spring assemblage against shear stress (Figure 5.5) illustrates the influence of hydraulic variables; of these, shear stress and Fr were better predictors than substrate size. In riffle microhabitats (0.07 m2) of the Kangaroo River of southeastern Australia, the majority of the macroinvertebrate community was associated with riffle areas of lowest near-bed turbulence (Brooks et al. 2005). Macroinvertebrate abundance and number of taxa were negatively related to roughness Re, shear velocity, velocity, and Fr. In particular, some mayflies of the families Leptophlebiidae and Baetidae, and the water penny Psephenidae, were associated with

Flow 1 Flow 2 Flow 3

Flow 1 Flow 2 Flow 3

0 200 400 0 400 800 0 400 800 1200

FIGURE 5.4 Average longitudinal velocity over cobbles with (open triangles) and without (closed triangles) the moss Frissidens rigidulus in a laboratory flume at three flow levels (1 is lowest, 3 is highest). The vertical axis is distance above the streambed. The existence of the internal boundary layer (IBL) and influence of the moss are clearly evident. (Reproduced from Nikora et al. 1998.)

0 200 400 0 400 800 0 400 800 1200

Average longitudinal velocity (mm s~1)

FIGURE 5.4 Average longitudinal velocity over cobbles with (open triangles) and without (closed triangles) the moss Frissidens rigidulus in a laboratory flume at three flow levels (1 is lowest, 3 is highest). The vertical axis is distance above the streambed. The existence of the internal boundary layer (IBL) and influence of the moss are clearly evident. (Reproduced from Nikora et al. 1998.)

FIGURE 5.5 Ordination of the fauna collected from the Ardeche River, France, during spring sampling versus a hydraulic axis constructed from hydraulic parameters including shear stress estimated using the FST-hemisphere method, Froude number, and depth and substrate measures. The bottom axis (a) denotes the hydraulic axis. (b) Taxa are positioned according to their locations along the axis, and the area of each circle is proportional to taxon abundance. Horizontal lines represent the standard deviation of the hydraulic score. (Reproduced from Merigoux and Doledec 2004.)

FIGURE 5.5 Ordination of the fauna collected from the Ardeche River, France, during spring sampling versus a hydraulic axis constructed from hydraulic parameters including shear stress estimated using the FST-hemisphere method, Froude number, and depth and substrate measures. The bottom axis (a) denotes the hydraulic axis. (b) Taxa are positioned according to their locations along the axis, and the area of each circle is proportional to taxon abundance. Horizontal lines represent the standard deviation of the hydraulic score. (Reproduced from Merigoux and Doledec 2004.)

FIGURE 5.6 Relationship between roughness Reynolds number and (a) number of invertebrate taxa and (b) macroinvertebrate abundance in sampled areas of 0.07 m2 within three riffles in the Kangaroo River, New South Wales, Australia. Dotted lines indicate 95% confidence intervals. (Reproduced from Brooks et al. 2005.)

FIGURE 5.6 Relationship between roughness Reynolds number and (a) number of invertebrate taxa and (b) macroinvertebrate abundance in sampled areas of 0.07 m2 within three riffles in the Kangaroo River, New South Wales, Australia. Dotted lines indicate 95% confidence intervals. (Reproduced from Brooks et al. 2005.)

low Re (Figure 5.6). Although each of these studies makes a strong argument for the utility of characterizing the flow microenvironment using both simple and complex hydraulic variables, at this time it is not obvious which variables might be more important, nor their mode of action.

A conceptual model (Biggs et al. 2005) outlines the mechanisms by which current and its associated forces will commonly have offsetting influences on the biota of running waters. Current exerts a drag force on individual organisms and, due to episodic fluctuations and substrate dislodgement, can be a powerful disturbance force as well. However, current benefits the biota through mass-transfer processes, transporting dissolved nutrients and gases to plants and food resources to animals. Biggs et al. (2005) further contend that large-scale temporal events predominantly affect lotic ecosystems through physical drag processes (drag-disturbance), whereas small-scale flow variations affect ecosystems through mass-transfer processes (including invertebrate and fish food-uptake). Drag-disturbance and mass-transfer related processes are viewed as the opposite ends of a continuum of the influence of flow variability, with moderate temporal-scale flow variability affecting ecosystems through both processes. Thus the location where organisms are found at steady state is a reflection of their ability to hold position and acquire resources, and, if a recent disturbance has been sufficiently extreme, their ability to survive and recolonize. Traits associated with avoiding or resisting the forces of flow, acquiring resources under particular flow conditions, and with movement and dispersal should therefore be expected to match certain taxa to particular flow environments.

The growth rates and distribution of periphy-ton illustrate the beneficial effects of small-scale fluctuations in flow (see also Section 6.1.1). Mixing of nutrients in the water column and reduction of the thickness of the viscous sublayer of the boundary layer surrounding microorganisms are expected to enhance rates of molecular diffusion across external membranes and thus the biological uptake of inorganic nutrients (Larned et al. 2004). Indeed, velocity variation over small spatial scales generates gradients in the biomass and growth form of periphyton patches, in which the biomass of dense, mucilaginous diatom mats increases with increasing near-bed velocities while the biomass of loosely aggregated filamentous green algal mats declines (Biggs and Hickey 1994, Biggs et al. 1998). In accord with the expectation that filter-feeding invertebrates will occupy regions of high near-bed current and collector-gatherers locations where particles are deposited, Merigoux and Doledec (2004) found that filter feeders displayed a positive relationship with shear stress and collector-gatherers a negative relationship. By holding low-velocity positions behind current obstructions, stream-dwelling salmonids optimize the tradeoff between the energy supply from drifting invertebrates and the energy cost of swimming (Fausch 1984). Position choice in drift-feeding Arctic grayling (Thymallus arcti-cus) was estimated well by a model in which net energy intake depended on capture rate, which was a function of visual reaction distance, depth, and velocity; and on swimming cost, which depended on velocity (Hughes and Dill 1990). Because a fish must intercept prey entering its field of view before the prey is swept downstream, velocity increases the encounter rate but decreases the proportion captured.

Microhabitat preferences can be strongly influenced by competitive dominance hierarchies and predation risk, especially in young individuals, and so habitat structure and overhead cover provide visual isolation from competitors and predators. Offered a choice among artificial structures placed in a natural stream, young steelhead trout (Oncorhynchus mykiss) were most influenced by overhead cover and favored a combination of cover, visual isolation, and velocity refuge, whereas young coho salmon (O. kisutch) showed strongest preference for velocity refuges and thus appeared to emphasize net energy gain (Fausch 1993).

Current acts as a negative environmental factor when floods cause dislodgment due to shear forces acting directly on the individual or indirectly through erosion of substrate particles, and when organisms must expend energy to maintain position. Organism distributions are influenced by disturbance events ranging from those of high frequency and low magnitude that result from turbulence and modest rainfall, to low frequency and high magnitude floods. The severity of the impact of any such event is further determined by the presence of flow refuges and the ability of organisms to seek out refuges and recolonize following disturbance. A survey of the snail Potamopyrgus antipo-darum from 48 streams across New Zealand showed that local densities were inversely related to flood frequency, and populations were more abundant in low-gradient streams of foothill regions that tended to have fewer, less severe floods (Holomuzki and Biggs 1999). In flume studies, snails and the mayfly Deleatidium moved into low-velocity crevices on all substrates as current velocities increased, evidently because they were able to "feel" the increase in skin friction, and caddis larvae Pycnocentrodes unreeled their silken drag-lines to reach more sheltered locations (Holomuzki and Biggs 2000). There is much still to be learned about the possible ability of stream organisms to sense and respond behaviorally to changes in their flow environment, but some clearly do (Lytle and Poff 2004). Others, however, may simply accumulate and persist in areas that are sheltered most or all of the time. Because mussels occur within the surface substrate of the stream-bed and are long-lived, one might expect their distribution to be related to hydraulic variables. Strayer (1999b) indeed found that mussel beds occurred in flow-protected locations of two rivers in New York.

Locations that act as refuges from high-flow episodes can be identified at multiple scales, and their utility is a function of many factors. The size of the organism, its position within the water column or the streambed, how the hydraulic environment changes as flows increase to flood stage, and any behaviors such as seeking shelter or modifying body shape and position in response to increasing hydraulic forces will influence the effectiveness of flow refuges. At a large spatial scale, tributaries in a river network and side channels in a river may maintain favorable environmental conditions, such as alkaline tributaries that resist episodic pulses of acidification (Section 4.4.3) or locations in the streambed that are below the depth of substrate scour during a spate. At an intermediate scale, floodplains, stream margins, depositional areas, and debris dams may serve as refuges (Lancaster and Hildrew 1993, Palmer et al. 1995, Francouer et al. 1998). At the finest scale, heterogeneity of the substrate including crevices and surface roughness may provide refuge for small organisms (Dudley and D'Antonio 1991, Bergey 2005).

Although there is ample evidence that some areas experience comparatively less hydrologic disturbance and maintain greater invertebrate abundances following a spate, it can be difficult to establish that specific habitats truly act as refuges. Colonization of artificial refuges (wire cages) placed in streams in the United Kingdom found differential accumulation of organisms in cages with finer mesh (and therefore reduced flow) compared to coarser mesh cages, and this occurred differentially at high flow (Winterbottom et al. 1997). A similar study reported accumulations of certain taxa in cages protected by baffles versus exposed cages following high-flow events in an upland Scottish stream, whereas no differences were seen at low flows (Lancaster 2000). However, whether colonization is by active or passive means is not known. In a sand-bed river in Virginia, the hyporheic zone within the streambed failed to act as a refuge because floods scoured to bedrock, a depth of roughly 20 cm; however, invertebrates associated with woody debris dams were more resistant to displacement than those in the channel (Palmer et al. 1996).

A true refuge effect requires that organisms are able to move actively or passively among habitat patches in response to a disturbance, have a greater likelihood of persistence due to reduced exposure to extreme currents and hydraulic forces, and have sufficient numbers and mobility to subsequently recolonize highly disturbed habitats (Lancaster and Belyea 1997). The flume studies of Holomuzki and Biggs (2000) provide evidence of such movements in response to increasing flows, and Lancaster

(1999) reported that the dytiscid beetle Oreo-dytes sanmarkii and the mayfly Ephemerella ignita accumulated in low flow areas of a laboratory flume by combination of crawling and drifting behaviors in response to a simulated spate. Invertebrates declined on unstable stones but exhibited no change on stable stones in a gravel-bed stream of the South Island of New Zealand, suggesting that organisms on unstable stones detected vibrations that stimulated their departure (Matthaei et al. 2000).

In comparison with invertebrates, stream fishes can more readily shift location in response to high-flow events and seek flow refuges associated with physical habitat complexity. Side-channels, wood, and large roughness elements may be especially important in high-gradient streams, and access to floodplains may be important in low-gradient streams. Schwartz and Herricks (2005) used prepositioned electrofish-ing devices to show that the fish assemblage of small, low-gradient Illinois streams occupied different habitats depending on flood stage. At near-bankfull flows fish were associated with vegetated point bars and concave-bank benches, at half-bankfull conditions fish abundance and biomass were greatest in low-velocity eddies, and at base flow the main channel habitat of pools, riffles, and glides contained higher numbers and greater biomass than did lateral habitat units.

The size and longitudinal spacing of roughness elements along the streambed influences the complexity of flow in the near-bed environment, as shown hypothetically in Figure 5.7. When channel depth is shallow relative to substrate roughness, such as in riffles and broken water, flow will be very complex. At depths greater than three times the height of roughness elements, Davis and Barmuta (1989) recognize three additional categories based on longitudinal spacing. When substrate elements are separated by sufficient distance, the wake behind each element dissipates before the next element is encountered. This is called isolated roughness flow.

Surface Roughness Spacing
(b) " A A *l* A >!

FIGURE 5.7 Conceptualization of three types of flow occurring over a rough surface, depending upon differences in relative roughness and longitudinal spacing between roughness elements. (a) Isolated roughness flow, (b) wake interference flow, (c)quasi-smoothflow. (Reproduced from Davis and Barmuta 1989, after Chow 1981.)

FIGURE 5.7 Conceptualization of three types of flow occurring over a rough surface, depending upon differences in relative roughness and longitudinal spacing between roughness elements. (a) Isolated roughness flow, (b) wake interference flow, (c)quasi-smoothflow. (Reproduced from Davis and Barmuta 1989, after Chow 1981.)

When spacing between roughness elements is less, their wakes interfere with one another, producing high local velocities and turbulence, termed wake interference flow. Lastly, skimming flow describes the circumstance when roughness elements are very closely spaced, which allows flow to skim across the tops of elements and produces a relatively smooth flow environment and slow eddies in the intervening spaces.

Bed surface roughness is not only due to stones of various sizes, but also wood and vegetation. Flow measured in and around a common lotic macrophyte, Ranunculus penicillatus, showed that velocities dropped to a low and constant value within 5 cm into the plant bed, forcing most of the flow over and around it. A dead-water zone formed immediately downstream, and then a region of high turbulence (Green 2005). Quinn et al. (1996) added artificial roughness elements (half-pipes of various diameters attached to a plywood base) to a section of a New Zealand gravel-bed river of uniform current and depth. Near-bed velocities declined and turbulence increased with increasing roughness, as expected. All invertebrates and especially filter feeders declined with increasing upstream roughness, whereas the periphyton apparently was affected by both microhabitat conditions and the differential response of grazers. Velocities at the front and in the wake of selected boulders in a western Australian stream, measured with a field-portable ADV, were greatly reduced at 5 mm and 2.5 cm above the bed relative to velocities at channel mid-depth (Bouckaert and Davis 1998). Turbulence and shear stress were greater in the wake than the front area, although not significantly so due to high variability, and both total abundance and species richness were greater in the wake locations (Figure 5.8), suggesting that turbulence resulted in more favorable conditions.

Although a great deal of effort has gone into the characterization of near-bed flow environments, progress has been modest. In a review of studies attempting to relate organism distributions to hydraulic variables, Jowett (2003) concluded that complex hydraulic variables had not yet been shown to be superior to simple measures of velocity, depth, and substrate roughness. Although the many studies reviewed above establish convincingly that organism abundances and ecosystem processes vary with the flow environment, the mechanism of action often is uncertain. Even in the most carefully controlled experimental studies it is difficult to ascertain whether organisms are responding to mean velocity or to the intensity of turbulence, meaning the variance and extremes of velocity. To identify the flow conditions likely to cause an organism to be eroded is difficult because

FIGURE 5.8 Mean invertebrate abundance and species richness (±1SE, n = 10) in front (dotted bars) and wake (shaded bars) of boulders in a western Australian stream. * = significant at 0.05, ** = significant at 0.01. (Reproduced from Bouckaert and Davis 1998.)

dislodgement may result from infrequent, highvelocity turbulence rather than average flow fields, and organisms may modify their posture and location as forces change. Nonetheless, it is apparent that current, interacting with substrate, profoundly influences life in running waters, and we turn next to an examination of these issues from the perspective of substrate.

0 0

Post a comment