Estimation of preferred water flow parameters for four species of Simulium (Diptera: Simuliidae) in small clear streams in South Africa

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/258029710 Estimation of preferred water flow parameters for four species of Simulium (Diptera: Simuliidae) in small clear streams in... Data in African Journal of Aquatic Science · December 2006 DOI: 10.2989/16085910609503895 CITATIONS READS 2 53 4 authors, including: Ferdinand C. De Moor Sharon A Birkholz 69 PUBLICATIONS 660 CITATIONS 7 PUBLICATIONS 23 CITATIONS Albany Museum and Rhodes University SEE PROFILE Cranfield University SEE PROFILE Robert Wiliam Palmer Nepid Consultants 26 PUBLICATIONS 272 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Flight behaviour in Trichoptera View project All content following this page was uploaded by Ferdinand C. De Moor on 19 December 2016. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, letting you access and read them immediately. Copyright © NISC Pty Ltd African Journal of Aquatic Science 2006, 31(2): 261–269 Printed in South Africa — All rights reserved AFRICAN JOURNAL OF AQUATIC SCIENCE EISSN 1727–9364 Estimation of preferred water flow parameters for four species of Simulium (Diptera: Simuliidae) in small clear streams in South Africa Nicholas A Rivers-Moore1*, Ferdinand C de Moor2, Sharon A Birkholz1 and Robert W Palmer3 1 Institute for Water Research, Rhodes University, PO Box 94, Grahamstown 6140, South Africa; current address: KZN Wildlife, PO Box 13053, Cascades 3202, South Africa 2 Albany Museum (Department of Freshwater Invertebrates, Makana Biodiversity Centre) and Rhodes University, PO Box 94, Grahamstown 6140, South Africa 3 Nepid Consultants, PO Box 4349, White River 1240, South Africa * Corresponding author, e-mail: riversmn@kznwildlife.com Received 19 April 2006, accepted 14 July 2006 Blackfly larvae typically occur in fast-flowing riffle sections of rivers, with different blackfly species showing preferences for different hydraulic conditions. Very little quantitative data exist on hydraulic conditions linked to the blackfly species occurring in South African streams. Stones-in-current biotopes (i.e. fast riffle flows over cobbles) were sampled from four sites in three small clear streams in the Eastern and Western Cape provinces of South Africa. Mean water column velocities at each sampled stone were measured using a mini current meter, while flow velocities closer to the boundary layer where blackfly larvae occurred were estimated using indirect techniques (standard hemispheres and aerating tablets). Standard hemispheres were also used to calculate more complex hydraulic parameters such as Froude and Reynolds numbers. Four species of Simuliid were sampled in sufficient numbers to show trends in flow velocity preferences. Simulium impukane and S. rutherfoordi both occurred at their highest densities at velocities of 0.3m s–1, while S. merops preferred velocities of 0.7m s–1. Simulium nigritarse SL attained the highest densities of all the blackfly species sampled, and its relative abundances were greatest at velocities of 0.8–0.9m s–1. Within the streams surveyed, all blackfly species occurred in subcritical-turbulent flows — based on a classification using Froude and Reynolds numbers — although two of the species were also found in high densities in supercritical flows where these existed at the sites. Local hydraulics within the stones-in-current biotope are complex, but in the absence of fine-scale equipment for measuring micro-velocities, standard hemispheres are a useful, cost-effective technique for the initial quantification of hydraulic parameters in small, clear streams. Such an approach facilitates further understanding of links between hydraulics and aquatic invertebrates in South African streams. Keywords: aerating tablets, blackfly, flow velocity, Froude number, hydraulics, Reynolds number, standard hemispheres Introduction Hydraulic properties of stream flow are valuable descriptors of the physical environmental template of lotic organisms (Statzner and Higler 1986), whose zonation patterns typically relate to downstream changes in hydraulic properties (Statzner and Higler 1986, Statzner et al. 1988, Statzner and Borchardt 1994). In general, abiotic processes are considered one of the major determinants of river communities (Hildrew and Giller 1994). The stone/water column interface is of particular importance to aquatic invertebrates, as it is a gradient region where flow velocity changes from zero at the stone surface to that experienced in the water column (Vogel 1981). Within this gradient region, or boundary layer, various factors including drag, the extent of laminar versus turbulent flow, and viscosity, affect aquatic invertebrates by affecting, inter alia, access to food and mobility (Gordon et al. 1992). Boundary layer thickness has considerable implications for filter-feeding aquatic macroinvertebrates, such as blackfly larvae (Diptera: Simuliidae), which depend on the accessibility of micro-particles of food in this layer. Blackflies are ubiquitous and occur in most riverine systems worldwide (Snaddon and Davies 2000), their larvae usually being found on hard substrata in fast flowing conditions (de Moor 1982). Carlsson (1967) found that the distribution of blackfly larvae depended largely on flow velocity, although depth, type of substratum, light and water temperature also contributed to habitat preference. For example, de Moor (1982) illustrated the link between blackfly densities, discharge, mean water column velocity and colonisable area in the Vaal River. His study showed that an increase in discharge provides more habitat, higher velocities, and better dispersion capacity, while falling water levels and unpredictable flows reduce suitable habitat. He also showed that the abundance of S. chutteri increased with increasing flow velocities (0.18–1.32m s–1), with 0.56m s–1 considered as a suitable lower threshold for colonisation. A small tributary of the Vaal River with slower velocities (0.3–0.4m s –1 ) had a lower prevalence of S. chutteri. Conversely, S. adersi showed a preference for flow velocities 262 of 0.3–0.6m s–1 (de Moor 1982). Similarly, Palmer (1997) observed that low flows in the Orange River (<21m3 s–1) were ideal for S. adersi, S. damnosum and S. mcmahoni, flows of around 200 m3 s–1 were suboptimal for S. chutteri, while discharges in excess of 300 m3 s–1 were ideal for forming large populations of S. chutteri. These discharges relate directly to flow velocities and stream depths, where the greater the discharge the higher the flow velocity and the greater the amount of available habitat for S. chutteri. Palmer and Craig (2000) showed that a relationship existed between labral fan structures of different species of blackfly larvae and habitat parameters (flow velocity). This relationship was useful in predicting distribution patterns of different species of blackfly larvae relative to flow velocities. For ecological purposes, flow velocities near the bottom of a channel, or at the water surface, are of greater significance than averages across a river channel cross-section or over a reach. Variations of flow velocity with time may also be critical to aquatic insects (Gordon et al. 1992) and, at smaller scales, velocity governs the hydraulic conditions that mediate biotic interactions (Hilldrew and Giller 1994). Additional hydraulic parameters of interest to understanding distribution patterns of aquatic freshwater invertebrates are the dimensionless Reynolds (Re) and Froude numbers. Re numbers, central to biological fluid mechanics (Vogel 1981), describe the ratio of inertial to viscous forces and assist in characterising the microhabitat of lotic organisms. Large Re numbers indicate turbulent conditions, while small Re numbers indicate laminar flows (Gordon et al. 1992). The Froude number is the ratio of inertial to gravitational forces and is useful in describing ‘bulk flow characteristics such as surface waves’ (Gordon et al. 1992: 276). For example, under conditions of high turbulence (i.e. higher Re numbers), Simulium chutteri feeds more efficiently than S. nigritarse SL (Barber-James 2004), due to differences in their labral fan structures and in how efficiently food particles are intercepted under different flow conditions (Palmer and Craig 2000). Estimating flow velocity at the surface of an in-stream stone, as well as associated hydraulic, parameters — including viscosity and laminar versus turbulent flows — are critical to understanding the abiotic templates that structure the zonation patterns of simuliids. One challenge is to use appropriate sampling techniques to collect invertebrate data and relate these to hydraulic parameters. The aims of this study were to evaluate the use of standard hemispheres for characterising the hydraulic habitat of Simulium spp. occurring in small clear streams, and to estimate the preferred flow velocities of these species. Study area Data were collected from two small clear streams in the Eastern Cape province, and one stream in the Western Cape province, South Africa. Two of the Eastern Cape sites were on the Bloukrans River, a tributary of the Kowie River, in the Belmont Valley 6.5km east of Grahamstown. Site 1 (BV1) (33°19’25.2”S; 26°36’00.8” E; 480m amsl) was downstream of a small bridge, while Site 2 (BV2) (33°19’21.4”S; 26°36’49.7” E; 470m amsl) was 1.5km further downstream. Rivers-Moore, de Moor, Birkholz and Palmer These sites were chosen because the downstream site represented a less polluted replicate of the upstream site. The third site in the Eastern Cape, Site 3, was located on the Palmiet River, a tributary of the Kariega River, at Howison’s Poort (33°22’05”S; 26°28’30”E; 400m amsl) 8km west of Grahamstown. Site 4 (33°58’30.6”S; 23°31’17.9”E; 20m amsl), in the Western Cape province, was located on the Salt River in the De Vasselot Nature Reserve. Methods Collection of invertebrates and hydraulic data Aquatic invertebrate communities were sampled between 2 and 12 August 2005, from 6–10 stones at each site from the stones-in-current biotope (i.e. fast riffle flows over cobbles). Stones that could be lifted with one hand were removed from riffle habitat, and their positions marked using polystyrene floats attached to lead sinkers. Each stone was visually assessed (after lifting) to determine where the highest concentration of simuliid larvae was to be found, relative to flow velocity. These densities were rated according to Palmer’s (1994) 10-point semi-logarithmic scoring system, where a score of 1 represents no blackfly larvae on a stone, and a score of 10 represents in excess of 500 000 larvae m–2. The lifted stones were held over a 250 µm mesh hand net, before being scrubbed over a bucket (Chutter 1968, de Moor 1982). Bucket contents were sieved through a 250 µm mesh hand net into collecting jars containing 70% ethanol. Aquatic invertebrates were identified to family and, where possible, to species level before being counted. Discharges (m3 s–1) at each site were calculated using standard hydrological techniques, as detailed below (Gordon et al. 1999). Discharges at Sites 3 and 4 were calculated using a Scientific Instruments (1980) mini current meter. In this method, the number of revolutions of the current meter at 0.4 of the water depth (measured from the river bed) were counted over one minute, at 0.25m intervals of the stream width at a site where the channel was constrained to account for all the flow. These counts were converted to velocities using calibration curves (Scientific Instruments 1980), and discharge was calculated using flow velocities and channel surface areas. Discharges at Sites 1 and 2, which were downstream of a bridge where all the flow flowed through eight pipes, were estimated using Equations 1a–c, which is an appropriate discharge estimation in partially-filled pipes: Q = vA (1a) A = 0.5r 2 (θ − Sinθ ) (1b) ⎛r−d ⎞ θ = 2 cos −1 ⎜ ⎟ ⎝ r ⎠ (1c) where Q is discharge (m3 s–1), v is average flow velocity (m s–1), A is cross-sectional area of pipe (m2), θ is in radians, d is depth of water in pipe (m), and r is radius of pipe (m). The following water quality parameters were also African Journal of Aquatic Science 2006, 31(2): 261–269 recorded: total dissolved salts (TDS) (WTW LF95 conductivity/TDS meter), pH (Cyberscan 20 pH meter, or Lovibond Comparator), conductivity (WTW LF95 conductivity/TDS meter) and water temperature (glass alcohol thermometer). Turbidity was not measured, since all streams had clear water. Mean water column velocities at each selected stone were determined, using the mini current meter, by recording the number of revolutions recorded by the current meter over one minute and converting these to flow velocities. Flow velocities, as close to the stone surface as possible, were also measured using two indirect methods. In the first, standard hemispheres were placed on a levelled plexiglass plane at each site which had previously been marked with a float showing where a stone had been removed. In this approach, the maximum density of the last hemisphere to move within the current was used to calculate velocity, using the relationship between density and flow velocity defined by Statzner and Müller (1989). Additional hydraulic parameters considered to be of greatest significance to the larval simuliids in this study, namely Froude and Reynolds numbers, were calculated using standard hemispheres and associated relationships (Statzner and Müller 1989). In the second method, after the scrubbed stones had been replaced, flow velocities close to the stone surface were estimated, based on mass loss of Livoxine ® aerating tablets, using an approach described by Madsen and Warncke (1983). A calibration curve was established between the mass loss of aerating tablets over a fixed time period (5min) and flow velocity measured using a mini current meter. Aerating tablets were held in place between two dowels (Ø = 3mm X 300mm long), attached at 90° to a wooden pole (Ø = 20mm X 1m) which was able to slide up or down a second pole. Using this system, the aerating tablets could be positioned in-stream at any chosen locality and depth, thus allowing the estimation of mean water column velocities, or velocities at the surface of in-stream stones. Flow patterns were also visualised using dye (red food colourant mixed with milk) (Clayton and Massey 1967, Vogel 1981), which provided an indirect measure of how flow velocities changed as water passed around the stones. This mixture provides a rough visual estimate of flow patterns, since the dye remains as a dark stain around the instream object for varying amounts of time, depending on flow velocity — less than 1sec for fast (1m s–1) flows, and more than 5sec for slow flows (<1m s–1), particularly around rocks covered with filamentous algae. Substratum size classes and substratum roughness were also calculated, using the methods of Statzner et al. (1988). In this method, stone sizes were visually assessed by size class, and ranked from most dominant to third most dominant size class, in order to estimate a weighted contribution to substratum roughness. Data analyses Larval densities (number of larvae m–2 of stone) were calculated by using an averaged value of 42% colonisable area on stones. Colonisable area was defined as the amount of stone surface exposed to flows, and excluded surfaces embedded in the river bed or colonised by algae. The proportion of colonisable area on stones-in-current biotope 263 was estimated from a pooled subsample of stones removed from high-velocity biotopes from each of the four sites. Areas on stones actually colonised by larvae were marked with a permanent marker. Each rock was wrapped in aluminium foil in such a way as to avoid foil overlaps, and from this pieces were cut out to approximate the colonised areas. The percentage of colonisable surface area actually used by simuliid larvae was estimated using a linear relationship between foil mass and surface area. Mean (± standard deviation) larval densities, as well as their coefficients of variation (%), provided estimates of the uniformity of distribution within the stones-in-current biotopes for the Simulium species collected. Owing to the non-normal distribution of blackfly larvae, relative abundances of blackfly larvae were log(x + 1) transformed prior to further analyses in testing for correlations against flow velocity data. Evaluation of the different techniques to measure flow velocities was achieved through correlations using simple linear regressions. Estimated microhabitat velocities (i.e. velocities calculated using the standard hemispheres and the aerating tablets) were regressed against measured average water column velocities to determine the relationship between mean flow velocities in the water column and velocities closer to the boundary layer. Also, flow velocity estimates using standard hemispheres were regressed against those estimated using the aerating tablets, to determine how similarly each of these techniques estimated velocities close to the boundary layer. In considering the hydraulic habitat of blackfly larvae, flow velocities estimated using the standard hemispheres were used in the graphs of blackfly larval abundances. This allowed for the calculation of Reynolds and Froude numbers associated with larval abundances based on a single measurement using the standard hemispheres. The preferred flows of individual simuliid species were classified according to Davis and Barmuta (1989), by relating maximum relative abundances to corresponding Reynolds numbers and Froude numbers. Results All river sites were less than 6m wide and 0.42m deep, with similar slopes (0.03‰), substratum roughness (2.78), and water temperatures (Table 1). Of the three river sites sampled, the Palmiet was the most acidic and oligotrophic system (Table 1). Both sites on the Bloukrans River were typical of highly eutrophic systems, with visible evidence of algal filamentous blooms (Kleynhans et al. 2005). Aquatic invertebrate communities in the Salt and Palmiet Rivers showed the greatest species diversity (Table 2). The Bloukrans River sites had few species present in relatively high abundances, and only one species of blackfly (S. nigritarse SL) which had relatively high abundances of both larvae and pupae (Table 2). The Palmiet River had the highest number of blackfly species, supporting three species typical of small streams, but all were present in low numbers, with only one species (S. medusaeformae) being found at more than one site. Coefficients of variation of larval densities of blackfly varied from 40% (S. merops) to 270% for S. impukane (Table 3). The distribution of larval S. nigritarse SL showed 264 Rivers-Moore, de Moor, Birkholz and Palmer Table 1: Spot values for physical and water quality variables at four sites in three river systems in the Eastern and Western Cape provinces in August 2005, and the number of stones sampled per site Parameters\Rivers Bloukrans (Site BV1) Water temperature (°C) TDS (mg l–1) Conductivity (µS cm–1) pH Discharge (m3 s–1) Width (m) Substratum particle size (m) No. of stones sampled 10.5 874 1 020 7.6 0.073 3.5 0.15 6 Bloukrans (Site BV2) Salt Palmiet 11.0 1 220 1 420 7.6 0.103 3.1 0.20 10 9.8 299 350 6.3 0.006 6.0 0.20 8 11.5 128 149 5.7 0.006 1.0 0.08 10 22 Orientation 20 FREQUENCY/DENSITY CLASS 18 Density class Class 4 = 4–9 larvae 4mm–2 Class 5 = 10–22 larvae 4mm–2 16 14 12 10 8 6 4 2 Top Bottom Front (upstream) Back (downstream) ORIENTATION Left Right Figure 1: Orientation of blackfly larvae on stone surfaces, and median class values of highest larval density (larvae m–2) on substratum particles. Larval density classes are on a 1–10 scale after Palmer (1994). Black bars represent absolute frequency of orientation of blackfly larvae on stones uneven clumping on substrata at the two sites on the Bloukrans River, which was probably caused by variations in algal coverage on stones at these two sites; the site with higher algal coverage showed more uniform distribution of blackfly larvae. Of the 32 stones sampled, 21 (65.6%) had their highest densities of blackfly larvae on their upper surfaces, with the next most common orientation (12.5%) being the ‘front’ of the stone, facing into the current (Figure 1). In two instances, where stones were shaped in such a way as to have water flowing underneath them, the highest density of larvae was underneath the stone. Median larval densities corresponded to Palmer’s (1994) Class 4 (3 750 larvae m –2). In all observations, the blackfly larvae orientated themselves in a uniform fashion, attaching themselves to the stones with their longitudinal axes (from posterior proleg to cephalic fans) in an upstream/downstream orientation along water-flow streamlines, as illustrated in Figure 2. The relationship between percentage mass loss of oxygen tablets and flow velocity was significant (n = 13; r2 = 0.81; P < 0.05) (Figure 3). However, this relationship became more variable at higher velocities. Correlations between average water column velocities measured using a mini current meter, and velocities close to each sampled stone, measured indirectly using standard hemispheres and aerating tablets, were weak but significant (n = 32; r2 = 0.51 and 0.53 respectively; P < 0.05) (Figure 4). Both indirect techniques generally overestimated flow velocities, with the standard hemispheres yielding higher flow velocity estimates than did the aerating tablets. Velocities estimated using aerating tablets were closer to the measured average water column velocities than were those esti- African Journal of Aquatic Science 2006, 31(2): 261–269 265 Table 2: Species collected at four sites, and class values for relative abundances. Average number of individuals per stone is shown by class (X = 1; XX = 2–10; XXX = 11–100; XXXX = >100). (L) = larva; (P) = pupa; (A) = adults; * = all larvae or nymphs Taxa Higher Taxa Family/species Ephemeroptera* Baetidae Baetis harrisoni (Barnard) Afroptilum sudafricanum (Lestage) Pseudocloeon vinosum (Barnard) Demoreptus capensis (Barnard) Small nymphs Leptophlebiidae Castanophlebia calida (Barnard) Adenophlebia peringueyella (Lestage) Caenidae gen. sp. indet Teloganodidae Lestagella penicillata (Barnard) Simuliidae Simulium impukane (de Meillon) (L) S .impukane (P) S. merops (de Meillon) (L) S. merops (P) S. nigritarse (Coquillett) (L) S. nigritarse (P) S. medusaeforme Pomeroy (L) S. medusaeforme (P) S. rutherfoordi (de Meillon) (L) S. rutherfoordi (P) Chironomidae gen. sp. indet. (L) Athericidae gen. sp. indet. (L) Hydropsychidae Cheumatopsyche afra (Mosely) C. maculata (Mosely) Philopotamidae Dolophilodes sp. Leptoceridae Oecetis sp. Aeshnidae gen. sp. indet. Elmidae gen. sp. indet. (L) Elmidae gen. sp. indet. (A) Helodidae gen. sp. indet. (L) Notonemouridae gen. sp. indet. Corydalidae gen. sp. indet. (L) Oligochaeta gen. sp. indet. Planaria sp. Ancylidae Burnupia sp. Physidae gen. sp. indet. Planorbidae gen. sp. indet. Fam. gen. sp. indet. Hydrachnellae Fam. gen. sp. indet. Diptera Trichoptera* Odonata* Coleoptera Plecoptera* Megaloptera* Annelida Turbellaria Mollusca Collembola Hydracarina mated using the standard hemispheres. Correlations between flow velocities estimated using standard hemispheres and aerating tables were also poor, but significant (n = 32; r2 = 0.28; P < 0.05) (Figure 5). In addition, correlations between water depths where each stone was sampled, and those estimated using standard hemispheres at the same point, were not significant (n = 34; r2 = 0.0003; P > 0.05). Of the five species of Simulium sampled, S. medusaeforme was excluded from analyses, since only three data points were obtained for it. The densities of the four remaining blackfly species peaked at different flow velocities (Figure Rivers sampled Bloukrans BV1 Bloukrans BV2 Salt Palmiet XXX XXX XXX X X X XX X X X XX X XX XX XX X XXX XX XXXX XX XXXX XXX XXX X X XX XXXX XXX X XX XX X X XX XX X XXX XXX X X X X XX X XX X X X X X X X X X XXX X XX X X X X Table 3: Mean densities (m–2), standard deviations and coefficients of variation (%) of five species of blackfly larvae at all sites combined Species Mean SD CV (%) S. merops S. medusaeforme S. impukane S. rutherfoordi S. nigritarse (BV1) S. nigritarse (BV2) 904 398 1 858 349 2 574 5 310 361 866 5 022 391 1 238 6 417 40 218 270 112 48 121 266 Rivers-Moore, de Moor, Birkholz and Palmer Figure 2: Blackfly (Simulium medusaeforme) larvae orientating according to the direction and strength of flow along the surface of a rounded boulder in a small clear stream in the Eastern Cape (photograph by FC de Moor) FLOW VELOCITY (m s–1) 0.8 0.7 0.6 changed the local hydraulic conditions by slowing down flow velocities and increasing the size of the boundary layer, thus making the hydraulic habitat unsuitable for S. nigritarse, possibly by excluding suspended particles from the water column. All four species favoured hydraulic habitats with Froude numbers of less than one, and Reynolds numbers of less than 125 000 (Figure 7). y = 0.012x – 0.0706 R2 = 0.8129 0.5 0.4 0.3 Discussion 0.2 0.1 0 0 10 20 30 40 50 MASS LOSS OF TABLET (%) 60 70 Figure 3: Relationship between percentage mass loss of aerating tablets and flow velocity 6). Simulium impukane and S. rutherfoordi favoured microhabitats with similar, low velocities (~ ~ 0.3m s–1), while S. merops showed preferences for faster flows (0.7m s–1). At the Palmiet site, populations of S. impukane had colonised leaves caught on top of stones-in-current. Simulium nigritarse SL favoured the highest velocities (0.8–0.9m s–1), and also occurred at the highest larval densities. Based on dye movement patterns, it was observed that thick algal mats Characterisation of hydraulic habitat While flow velocities may be the ‘fundamental abiotic factor controlling ecological processes and patterns in streams’ (Hart and Finelli 1999: 375), the complexity of the hydraulic environment at the boundary layer, particularly in rapids and riffle biotopes, presents difficulties in their measurement (Davis and Barmuta 1989). To overcome these challenges may ultimately require instruments with fine spatial resolution to measure basic hydraulic parameters (Carling 1992, Hoover and Ackerman 2004). This complexity, with associated gradients of velocities over small distances, may be one reason why we found no significant correlations between average water column velocities measured using a mini current meter, and those estimated closer to the boundary layer using indirect techniques. Similarly, Hart et al. (1996) found no correlation between velocities meas- 1.2 AT 1 to 1 Linear (SH) Linear (AT) y = 1.2851x + 0.1471 r2= 0.5127 1.0 0.8 y = 0.6708x + 0.173 r2 = 0.5255 0.6 0.4 0.2 0.2 0.4 0.6 0.8 1.0 1.2 MEASURED FLOW VELOCITY (m s–1) S. nigritarse 1.5 1.0 0.5 0.2 1.4 Figure 6: Relative abundances of four Simulium species versus flow velocities estimated using standard hemispheres Reynolds FROUDE NUMBER FLOW VELOCITY (m s–1) :aerating tablets 0.4 0.6 0.8 1.0 1.2 FLOW VELOCITY (m s–1) Froude 150 000 2.5 1.2 100 000 2.0 1.0 A 1.5 0.8 50000 1.0 0.6 0.5 0.4 0.2 S. rutherfoordi 2.0 3.0 1.4 S. merops 2.5 1.4 Figure 4: Correlation between average water column velocity measured using a mini current meter, and flow velocities calculated using indirect techniques (aerating tablets (AT) and standard hemispheres (SH)). Dashed line represents 1:1 correlation S. impukane 3.0 y = 0.2926x + 0.2384 r2 = 0.2786 0.2 0.4 0.6 0.8 1.0 1.2 1.4 FLOW VELOCITY (m s–1): standard hemispheres Figure 5: Correlation between flow velocity estimated using aerating tablets and that estimated using standard hemispheres ured 2mm above a stone and those measured 10mm above it. Where larvae are found without the need for removal from their stones, the use of hot-wire anemometers — which are small enough to measure flow velocities in a 0.25mm space — would be one approach to resolve these uncertainties. However, such technology is not always readily available or cheap (Davis and Barmuta 1989). In the absence of such technology, Davis and Barmuta (1989) proposed a classification of stream flows useful to ecological studies, where complex hydraulic parameters are calculated from relatively simple measures, including measurement of velocity profiles, depth, slope and substratum roughness. This is a similar approach to that of Statzner et al. (1989), who described a useful approach for calculating hydraulic parameters, also based on simple hydrological measurements. This approach was useful in categorising hydraulic requirements of larvae of the pest blackfly, S. chutteri, in the Great Fish River (Rivers-Moore et al. in press), where velocity was the major hydraulic determinant of larval blackfly distribution. The lack of instruments with fine spatial resolution was therefore not a limiting factor in describing the hydraulic environment of the blackfly larvae sampled in the present study. The average depth of the 0.2 0.4 0.6 0.8 1.0 1.2 FLOW VELOCITY (m s–1) REYNOLDS NUMBER CALCULATED FLOW VELOCITY (m s–1) SH 1.4 267 Log (Relative Abundance + 1) African Journal of Aquatic Science 2006, 31(2): 261–269 1.4 Figure 7: Relationship between flow velocity, and the biologically significant hydraulic parameters Reynolds and Froude numbers, estimated using standard hemispheres. Note that the threshold Froude number separating subcritical-turbulent from supercriticalturbulent flows is indicated by point A, as defined by Davis and Barmuta (1989) sites in this study (<0.42m) made it difficult to measure flow velocities within the water column but, nevertheless, the small size of the streams provided an opportunity to use and evaluate standard hemispheres and aerating tablets to measure and describe the hydraulic habitat of blackfly larvae in small, clear streams. Cost-effective techniques, such as aerating tablets and dye, offer limited usefulness in characterising local hydraulic conditions. The mass loss of aerating tablets at different flow velocities is a useful principle, although the method would need to be refined — primarily through refining the apparatus which holds the tablets. The 3mm Ø dowels which held the aerating tablets could be replaced with thinner (<0.5mm) sprung steel callipers to allow for placement of the tablets closer to the stone surfaces. For visualising flow patterns around stones-incurrent biotopes, coloured dyes mixed with milk are of limited value in stream systems with discharges in the order of 0.01m3 s–1 (10 l s–1). They become of increasingly less value at velocities exceeding 0.26m s –1 and at discharges in streams, such as the Bloukrans River, which have average daily flows above 0.1m3 s–1 (100 l s–1). This is because the 268 dyes are dispersed so rapidly that it is difficult to visualise streamlines and areas of laminar versus turbulent flow. Standard hemispheres are useful as an initial technique for quantifying local hydraulic conditions, which can then be related to aquatic invertebrate distribution patterns, particularly in the absence of more expensive micro-velocity meters. Such hemispheres are effective integrators of local hydraulics around a point (Statzner and Müller 1989), and provide a cost-effective approach for initial research in this area for South African rivers. Standard hemispheres are a simple technique for use in the field, even though their one important disadvantage is that it was sometimes not possible to position a hemisphere in situ where blackfly larvae had been collected. We assumed that a hemisphere would approximate the hydraulic conditions of the stone for which it was a ‘place-holder’, in spite of differences in height between the stone and hemisphere translating into differences in local hydraulic conditions, as shown by the use of coloured dye. However, it is also important that, since the hydraulic parameters calculated from the densities of the standard hemispheres are all highly correlated, multi-colinearity in multivariate ordinations would be an issue. One simple measure (such as velocity) is therefore more pragmatic to use as a ‘place-holder’ in such analyses, and from the statistical relationships between standard hemisphere densities and key hydraulic characteristics (including Reynolds and Froude numbers), which can be used to estimate the local hydraulic habitat of target aquatic invertebrates. The usefulness of standard hemispheres lies in describing the hydraulic environment of those aquatic invertebrates of interest, since the hemispheres integrate sitespecific hydraulic effects through their design and in situ placement. Preliminary hydraulic preferences of blackfly larvae sampled Froude numbers of less than 1 characterise subcritical flow, whereas supercritical flow occurs at Froude numbers exceeding 1 (Vogel 1981, Gordon et al. 1992). All blackfly species in this study preferred subcritical flow regions of the river. Using the classification system of Davis and Barmuta (1989), highest densities of all five species of simuliids occurred within subcritical-turbulent flows (i.e. Froude numbers <1 and Re numbers >2000), although there are indications from these data that S. nigritarse and S. merops tolerate supercriticalturbulent flows. According to Chow (1982), these two flow types are the most common ones in streams and rivers. Since hydraulic preferences for S. nigritarse in this study approach supercritical conditions, it is not surprising that this species is also encountered in more turbulent, larger river systems (relative to South African river systems; mean daily discharge >5m3 s–1) (see, for example, studies by O’Keeffe and de Moor 1988, Palmer and O’Keeffe 1990). This is in contrast to S. chutteri, which is unlikely to occur at densities large enough to result in pest outbreaks at flow velocities of less than 1m s–1 (Rivers-Moore et al. in prep.) and which typically occurs in supercritical-turbulent flows in large, turbid, regulated rivers (de Moor 1994). Thus, certainly within southern Africa, hydraulic properties of streams determine longitudinal macrodistribution patterns of larval blackfly species, which is Rivers-Moore, de Moor, Birkholz and Palmer reflected in feeding efficiency and labral fan patterns (Palmer and Craig 2000). This research also showed that those blackfly species that prefer relatively low flow velocities occurred in lower numbers, while higher velocities and more turbulent conditions sustained higher larval densities. This may be related to food availability and to the ability of different blackfly species to filter feed at different efficiencies under different hydraulic conditions (Harrod 1964, Barber-James 2004). Further research is recommended to obtain a better understanding of the links between turbulence, larval densities and feeding efficiency. Morphology and surface roughness of individual stones are important determinants of local hydraulic conditions. At these fine scales, small changes in hydraulic conditions caused by minor changes in discharges, or even the presence of entrapped leaves (as with S. impukane), may be significant enough to have large impacts on habitat suitability and larval densities of blackfly larvae. We were unable to measure these in this study, although we believe this needs further research, using more sensitive techniques. Furthermore, at these small scales, changes in water temperatures are likely to impact on water viscosity, and consequently on the size of the boundary layer, which is likely to impact on feeding efficiencies. Acknowledgements — We thank the National Research Foundation of South Africa for financial assistance. Geoff and Gill McIlleron and Julie Carlisle of the Nature’s Valley Trust are thanked for hospitality and assistance with fieldwork in the Salt River. Helen Barber-James of the Albany Museum, Grahamstown, is thanked for identification of Ephemeroptera. We also thank the anonymous referees for their constructive criticism and comments. References BARBER-JAMES HM (2004) A preliminary investigation into the influence of turbulence on larval feeding in two species of blackfly, Simulium chutteri Lewis and Simulium nigritarse Coquilett (Diptera: Simuliidae), from the Great Fish River, South Africa. Annals of the Eastern Cape Museums 3: 16–24 CARLING PA (1992) The nature of the fluid boundary layer and the selection of parameters for benthic ecology. Freshwater Biology 28: 273–284 C ARLSSON G (1967) Environmental factors influencing blackfly populations. Bulletin of the World Health Organisation 37: 139–150 CARLSSON G (1968) Benthonic fauna in African watercourses, with special reference to blackfly populations. Research Report No. 3. Scandinavian Institute of African Studies, Uppsala C HOW V EN TE (1982) Open-channel Hydraulics. McGraw-Hill, Tokyo CHUTTER FM (1968) On the ecology of the fauna of stones in the current in a South African river supporting a very large Simulium (Diptera) population. Journal of Applied Ecology 5: 531–561 CLAYTON BR and MASSEY BS (1967) Flow visualization in water: a review of techniques. Journal of Scientific Instruments 44: 2–11 DAVIS JA and BARMUTA LA (1989) An ecologically useful classification of mean and near-bed flows in streams and rivers. Freshwater Biology 21: 271–282 DE MOOR FC (1982) A Community of Simulium Species in the Vaal River near Warrenton. PhD thesis, Faculty of Science, University of the Witwatersrand, Johannesburg, South Africa African Journal of Aquatic Science 2006, 31(2): 261–269 MOOR FC (1994) Aspects of the life history of Simulium chutteri and S. bovis (Diptera: Simuliidae) in relation to changing environmental conditions in South African rivers. Verhandlung der Internationale Vereinigung für Theoretische und Angewandte Limnologie 25: 1817–1821 G ORDON ND, M C M AHON TA and F INLAYSON BL (1992) Stream Hydrology: an Introduction for Ecologists. John Wiley & Sons, Chichester HARROD JJ (1964) Effect of current speed on the cephalic fans on the larva of Simulium ornatum var. nitidifrons Edwards (Diptera: Simuliidae). Hydrobiologia 26: 8–12 HART DD and FINELLI CM (1999) Physical-biological coupling in streams: the pervasive effects of flow on benthic organisms. Annual Review of Ecology and Systematics 30: 363–395 HART DD, CLARK BD and JASENTULIYANA A (1996) Fine-scale field measurement of benthic flow environments inhabited by stream invertebrates. Limnology and Oceanography 41: 297–308 HILDREW AG and GILLER PS (1994) Patchiness, species interactions and disturbance in the stream benthos. In: Giller PS, Hildrew AG, and Raffaelli DG (eds) Aquatic Ecology: Scale, Pattern and Process. The 34th Symposium of the British Ecological Society, with the American Society of Limnology and Oceanography, University College, Cork, 1992. Blackwell Scientific Publications, Oxford. pp 21–62 H OOVER TM and A CKERMAN JD (2004) Near-bed hydrodynamic measurements above boulders in shallow torrential streams: implications for stream biota. Journal of Environmental Engineering and Science 3: 365–378 KLEYNHANS CJ, LOUW MD, THIRION C, ROSSOUW NJ and ROWNTREE K (2005) River Eco-classification: Manual for Eco-status Determination (Version 1). WRC Report No. KV 168/05. Water Research Commission and Department of Water Affairs and Forestry, Pretoria, South Africa MADSEN TV and WARNCKE E (1983) Velocities of currents around and within submerged aquatic vegetation. Archiv für Hydrobiologie 97: 389–394 PALMER RW (1994) A rapid method of estimating the abundance of immature blackflies (Diptera: Simuliidae). Onderstepoort Journal of Veterinary Research 61: 117–126 DE View publication stats 269 PALMER RW (1997) Principles of integrated control of blackflies (Diptera: Simuliidae) in South Africa. WRC Report No. 650/1/97. Water Research Commission, Pretoria, South Africa PALMER RW and CRAIG DA (2000) An ecological classification of primary labral fans of filter-feeding black fly (Diptera: Simuliidae) larvae. Canadian Journal of Zoology 78: 199–218 PALMER RW and O’KEEFFE JH (1990) Downstream effects of a small impoundment on a turbid river. Archiv für Hydrobiologie 119: 457–473 RIVERS-MOORE NA, DE MOOR FC, MORRIS C and O’KEEFFE J (in press) Effect of flow variability modification and hydraulics on invertebrate communities in the Great Fish River (Eastern Cape Province, South Africa), with particular reference to critical hydraulic thresholds limiting larval densities of Simulium chutteri Lewis (Diptera: Simuliidae). River Research and Applications SCIENTIFIC INSTRUMENTS (1980) Scientific Instruments, Inc. 518 West Cherry Street, Milwaukee] SNADDON CD and DAVIES BR (2000) An assessment of the ecological effects of inter-basin transfer schemes (IBTs) in dryland environments. WRC Report No. 665/1/00. Water Research Commission, Pretoria, South Africa STATZNER B and BORCHARDT D (1994) Longitudinal patterns and process along streams: modelling ecological responses to physical gradients. In: Giller PS, Hildrew AG and Raffaelli DG (eds) Aquatic Ecology: Scale, Pattern and Process. The 34 th Symposium of the British Ecological Society, with the American Society of Limnology and Oceanography, University College, Cork, 1992. Blackwell Scientific Publications, Oxford. pp 113–140 STATZNER B and HIGLER B (1986) Stream hydraulics as a major determinant of benthic invertebrate zonation patterns. Freshwater Biology 16: 127–139 STATZNER B and MÜLLER R (1989) Standard hemispheres as indicators of flow characteristics in lotic benthos research. Freshwater Biology 21: 445–459 STATZNER B, GORE JA and RESH VH (1988) Hydraulic stream ecology: observed patterns and potential applications. Journal of the North American Benthological Society 7: 307–359 VOGEL S (1981) Life in Moving Fluids: the Physical Biology of Flow. Willard Grant Press, Boston, Massachusetts