Creating a Disturbance: Manipulating Slackwaters in a Lowland River

The dynamic nature of habitat patches in rivers is driven primarily by flow regime. Altered hydrology, through river regulation, can limit the size and distribution of slackwater patches; important areas for the development of young fish and for shrimp in lowland rivers. Between late October 2002 and late January 2003, we investigated responses of fish, shrimp and their potential prey to the experimental creation of slackwaters and, conversely, to the experimental creation of flowing patches, by diverting water away from flowing patches and into slackwater patches, respectively. A pre-experimental survey indicated that slackwaters contained many more fish than flowing patches, and fish larvae were flushed out of slackwaters during the construction of flowing patches. Creation of slackwaters resulted in increased abundance of fish and shrimp, with the opposite occurring when slackwaters were changed into flowing patches. Converting slackwaters into flowing patches, and vice versa, altered the species composition of zooplankton and microbenthic assemblages but did not change their densities. Thus, standing crop of potential prey alone could not explain the differences in fish or shrimp abundance found between patch types. We hypothesize that slackwaters primarily act as refuges from current and provide energetic advantages to the young stages of fish and to shrimp. River regulation has the potential to affect the recruitment success of fish and shrimp by affecting the size, arrangement and availability of slackwater patches. Copyright # 2006 John Wiley & Sons, Ltd. key words: fish; shrimp; river regulation; experiment; larvae INTRODUCTION Habitat patches in rivers are formed by interactions among hydrology, geomorphology and structural elements, such as coarse woody debris, boulders or macrophytes (Thorp et al., in press). The dynamic nature of these habitat patches is driven primarily by characteristics of the flow regime (Hill et al., 1991). This dynamism is most obvious during flooding, when ephemeral off-channel habitat patches, such as billabongs (oxbow lakes) and other floodplain wetlands, become inundated and connected to the main channel. Prior to inundation, these habitat patches are terrestrial in nature; and it is only once flooded that they become available to aquatic organisms (Boulton and Lloyd, 1992; Nielsen et al., 2002; Brock et al., 2003). This shifting nature of habitat patches may also occur within the main channel of rivers; thus, a patch that may act as a pool (slow flowing or still) under particular flow conditions may act as a run (fast current, unbroken water) under other flow conditions. The persistence of habitat patches is dependent on temporal characteristics of a flow regime: timing, duration and variation. A patch may begin to exhibit the characteristics of a pool as flow declines into the normally dry period (timing). This pool may persist for several months (duration), although local rainfall events may disrupt its pool-like nature temporarily (variation). Riverine biota must run the gauntlet of the variation in habitat patches due to changes in flow, although in the case of highly mobile organisms, such as adult fish, it should not normally cause undue stress. This is partly because of their typically more generalist requirements (e.g. food, shelter, depth) relative to younger stages, partly because they can move from one patch to another as flow changes, and partly Received 10 February 2005 Revised 13 March 2005 Copyright # 2006 John Wiley & Sons, Ltd. Accepted 14 September 2005 *Correspondence to: Dr Paul Humphries, The Johnstone Centre, School of Environmental and Information Sciences, Charles Sturt University, PO Box 789, Albury, NSW 2640, Australia. E-mail: [email protected] y Present address: The Johnstone Centre, School of Environmental and Information Sciences, Charles Sturt University, PO Box 789, Albury, NSW 2640, Australia. because of the often predictable nature of broad, seasonal hydrological patterns within which biota have evolved (Lytle and Poff, 2004). However, for less mobile organisms or life stages—such as attached eggs and undeveloped larvae of fish—changes to flow, and the resulting effect on critical rearing habitat, may result in significant mortality (Harvey, 1987; Schlosser and Angermeier, 1990; Chapman and Kramer,1991; De Angelis et al., 1997) and may be a significant determinant of recruitment success in any one year. Alterations to natural patterns of river flow are extremely common throughout the world, as part of the widespread regulation of rivers (Ward and Stanford, 1979; Petts, 1984). The natural hydrologies of reaches, sections and sometimes entire rivers have been disrupted by the capture of rainfall and the subsequent release of water from dams, often for hydro-electric or irrigation purposes. This flow alteration, by definition, has resulted in altered type, frequency and extent of particular habitat patches. In the case of the lower River Murray in South Australia for example, normally free-flowing sections of river have been turned into virtual lakes by a series of 11 locks (Walker, 1986, 1992). In other cases, capture of runoff in dams has reduced flooding frequency and floodplain connectivity (Ward and Stanford, 1979; Petts, 1984). Less obvious changes to habitat patches, however, may occur within the main channel of rivers, especially when flows are only moderately altered. For example, artificially enhanced flows in rivers during typically low-flow periods—summer in southern Australia—increases current speed and depth within slackwater patches (areas with little or no discernible current) (Thorp and Casper, 2003). This can result in the loss of a substantial proportion of this type of habitat (Bowen et al., 2003), sometimes for short periods— such as during hydro-peaking—and sometimes for months at a time, during irrigation releases. Because slackwaters are important nursery habitats for some species of fishes (Humphries et al., 1999, 2002; Humphries and Lake, 2000; Dieterman and Galat, 2004; King, 2004b) and for shrimp (Richardson et al., 2004), there is the potential for significant harm to recruitment should their extent be diminished. In this study, we investigated the effects on fish, shrimp and measures of potential food (microfauna, benthic chlorophyll-a and organic content) of changes to slackwater patches within a lowland river in southern Australia, through two hydraulic manipulations. In the first manipulation, we diverted flow through natural slackwaters, increasing the current speed considerably (and thus simulating enhanced flows due to releases from dams). In the second manipulation, we diverted flow away from naturally flowing patches, thus stopping the current and effectively ‘creating’ slackwaters. We hypothesized that increasing current speed through natural slackwaters would reduce the abundance of fish, shrimp and microinvertebrates, and reduce the benthic chlorophyll-a and organic content relative to controls. We hypothesized that stopping the current in naturally flowing habitats would have the opposite effects. STUDY LOCATION The study was conducted in the Broken River, Northeastern Victoria, Australia. This river arises at about 1000m a.s.l., just north of the Great Dividing Range and has a total length of approximately 180 km. The mean annual discharge of the Broken River is about 230 10 m, with highest flows between June and September, when water temperatures are coldest ( 7 C) and lowest flows between December and April, when water temperatures can reach > 30 C. The Broken River is moderately regulated, with some storage of winter/spring flows in two impoundments (Lakes Nillahcootie and Mokoan) and later release of this over summer. Two substantial weirs (Casey and Gowangardie weirs), downstream of the major impoundments, are impediments to movement of fish upstream, but have little effect on the hydrology of the river. Much of the catchment has been cleared for agriculture. The experiments were carried out at Morago (145 57’ 20’’ E, 36 31’ 18’’ S–145 57’ 28’’ E, 36 31’ 31’’ S), which typically does not experience enhanced flows over the spring/summer period (as releases from Lake Nillahcootie are routed through Lake Mokoan and re-enter the Broken River downstream of our study reach) and so discharge is relatively stable during the time the experiment was conducted. The study reach was approximately 1 km in length, its width ranged from 20 to 50m, it consisted of alternating pools (slow current speed, between 1 and 2m deep) and runs (moderate current speed, between 0.2 and 1m deep), with frequent piles of large woody debris throughout and less frequent patches of aquatic macrophytes (predominantly Phragmites australis and Vallisneria gigantea). 526 P. HUMPHRIES ET AL. Copyright # 2006 John Wiley & Sons, Ltd. River Res. Applic. 22: 525–542 (2006) Eight species of fish have been commonly caught in this region of the Broken River (Humphries et al., 2002) and include Australian smelt (Retropinna semoni), carp gudgeons (Hypseleotris spp., see below), Murray cod (Maccullochella peelii), crimson-spotted rainbowfish (Melanotaenia fluviatilis), mountain galaxias (Galaxias olidus), European perch (Perca fluviatilis), common carp (Cyprinus carpio) and eastern gambusia (Gambusia holbrooki). The adults of all of these species can be found in both flowing and non-flowing habitats in the Broken River (Humphries unpublished data), although Murray cod and mountain galaxias only rarely in the latter. None of the species could be termed obligate rheophils or obligate limnophils as most exhibit ontogenetic changes in habitat use (see King, 2004b). Three species of shrimp—the atyids Paratya australiensis and Caridina mccullochi and the palaemonidMacrobrachium australiense—are common in this region of the Broken River (Richardson et al., 2004). Larvae of all three shrimp species can be found in slackwater areas, as do juvenile and adult Paratya and Caridina: adultMacrobrachium more commonly occur in flowing channel habitats (Richardson et al., 2004).