ÐÓ°ÉÔ­´´

Anglers had been complaining for years that their catches were declining, but it was not until the Australian bass and golden perch disappeared almost completely from some of southeastern Australia’s rivers that biologists realised something was seriously wrong. It has taken 10 years of research to pinpoint the problem and devise a solution.

Southeastern Australia has 83 species of freshwater fish. Almost half of them must migrate at least once during their life cycle. These include fish popular with anglers, such as the Australian bass and golden perch; barramundi and striped or sea mullet which are fished commercially, and threatened species such as the Macquarie perch and Australian grayling.

In 1913, David Stead, a biologist working for the government of New South Wales, recognised that the migrating fish needed a way to get past the increasing number of dams and weirs that were being built to provide water for the area’s growing population. The following year, a simple 10-metre-long ramp was built at Audley on the Hacking River, approximately 25 kilometres south of Sydney, with the aim of helping fish past the weir, but it soon became obvious that very few fish could struggle up its 1 in 10 slope. This led to an interest in the pool-type fish ladders or ‘fishways’ that had proved highly successful on the salmon rivers of North America and Europe, where they help adult salmon negotiate dams and weirs as they journey upstream towards their spawning grounds.

Between 1925 and the mid-1980s, more than forty of these fishways were built on the rivers of southeastern Australia. And they provide a striking example of the pitfalls of adopting a foreign technology without first determining its suitability for local conditions.

Pool fishways consist of a ‘staircase’ of pools, and fish swimming upstream have to leap up the steps from one to the next. The fish they were designed to cater for – adults of the salmonid family – are larger and faster-swimming than Australia’s migratory fish. Their migratory patterns are different too. Three-quarters of the old-style fishways built in Australia have gradients steeper than 1 in 8, and in most of them the step from one pool to the next is at least 30 centimetres. The velocity of water flowing through such fishways is about 2.4 metres per second, and turbulence reaches more than 190 watts per cubic metre – stormy waters even for a fish.

John Harris of the Fisheries Research Institute in New South Wales has estimated that the habitat available to fish in the coastal rivers of southeastern Australia has halved since the dams and weirs were built. Grayling and bass have disappeared completely from some coastal rivers. Further inland, silver perch have declined by up to 93 per cent over the past 50 years, and it is now listed as a threatened species.

By the 1980s, Australian biologists had begun to realise that it was important for some native migratory fish to be able to move around within the river systems throughout the year, and not just during the breeding season. This was the starting point for my own research group at the Fisheries Research Institute, which the government of New South Wales commissioned in 1985 to find out how to make the fishways more effective.

The waters of southeastern Australia can be divided into the coastal rivers and the Murray-Darling river system (see Map). In the coastal waters, including the coastal reaches of the Murray River, the most common migratory pattern is for the adult fish to live in the rivers, from where they move downstream in winter to spawn in the estuary or the sea. The adults return to their usual habitats upstream in early spring, followed by the juvenile fish in the late spring and summer. Australian bass, barramundi and the striped or sea mullet all follow this ‘catadromous’ migration pattern, which is the exact opposite to the ‘anadromous’ migration pattern of salmon.

But other common migratory fish in coastal rivers follow the ‘amphidromous’ pattern, which is driven not by breeding but by the search for food and the need to avoid predators. Throughout the year, young and adult members of amphidromous species, such as the climbing galaxias, mangrove jack and flat-tai mullet, are moving up and down rivers and to and from the sea.

In the light of this knowledge, it becomes obvious that fishways at coastal weirs need to accommodate different sizes of fish with different swimming abilities. Fishways designed for a 10-kilogram adult salmon that is 700 millimetres long are unlikely to suit a native juvenile fish that may weigh just a few grams and measure 40 millimetres in length. Even the adult catadromous and amphidromous species generally weigh less than 2 kilograms.

In the large, inland Murray-Darling river system, the most common migratory fish are ‘potamodromous’: they move entirely within the freshwater system. Examples include silver and golden perch, which weigh up to 3 kilograms when fully grown, and the huge Murray cod, which can grow to 80 kilograms. Tagging studies on the golden perch, carried out in the 1970s by Fred Reynolds of South Australia’s Department of Fisheries, showed that the fish could travel 1000 kilometres during the spring and summer, averaging 10 kilometres a day. Silver perch migrated about 500 kilometres upstream. Reynolds and other biologists concluded that these extensive upstream movements of perch were, like the movements of salmon, connected to spawning. Perch eggs, however, are semibuoyant and float downstream where they hatch within a day or two, while salmon eggs spend three months on the bed of the river. For this reason, the movement of immature fish was generally agreed to be insignificant, and it seemed logical to assume that the design of fishways for inland rivers could be based on the swimming ability of adult perch.

Given that the native migratory fish are weaker swimmers than salmon, the water speed in fishways had to be reduced. Other biologists have estimated that the maximum swimming speed of fish is between 9 and 21 body lengths per second, but few researchers have studied how fish other than salmonids move and behave in an experimental fishway. The size of pools, steepness of channel and type of opening between pools all influence the speed of water flowing through the fishway, and so the ability of fish to make it to the top.

We decided to study four species of migratory fish as representative of the smallest fish likely to migrate upstream in the two river systems. From the coastal rivers we chose two sizes of juvenile Australian bass (40 and 64 millimetres) and 43-millimetre juvenile barramundi. From the Murray-Darling river system we chose 250-millimetre adult silver perch and 440-millimetre adult golden perch.

Our experimental fishway was based on a vertical-slot design that has been used successfully in North America for about 50 years. It consists of a channel divided into a series of pools by baffles. Water passes down from pool to pool through a vertical slot in the baffle. Fish move up the fishway by swimming through the slot into the next pool. One of the main advantages of the vertical-slot fishway is that the slot can be angled to minimise turbulence. In addition, fish can pass through the fishway whether the level of the river is high or low – an important consideration in areas such as southeast Australia, where river levels fluctuate widely with rainfall.

GOING AGAINST THE FLOW

In our experiments we counted the proportion of each type of fish that could ascend our fishway at various water velocities. As we had suspected, the maximum velocity they could swim against was significantly less than the 2.4 metres per second of existing pool-type fishways. For example, in the fishway juvenile barramundi could not negotiate water flowing faster than 0.66 metres a second. The 64-millimetre juvenile Australian bass could cope with water speeds up to 1.4 metres a second, and its younger and smaller 40-millimetre counterpart could not negotiate water flowing faster than 1.02 metres a second. The adult golden and silver perch could pass through slots when water was flowing at 1.8 metres a second or less.

Since the late 1980s, the Sydney Water Board, the Department of Water Resources in New South Wales and the Murray-Darling Basin Commission have built six vertical-slot fishways. There are four on the coastal reaches of the Hawkesbury and Nepean rivers and two on the Murray-Darling river system. The channel in these fishways extends downstream from the top of the weir and doubles back so that its entrance is close to the face of the weir where fish congregate.

In a fishway on the upper reaches of the Hawkesbury River at Penrith, which is nearest the estuarine spawning grounds, water flow is designed to be 1.25 metres a second. Those built farther upstream at Camden, Sharpes and Cobbitty, on the lower reaches of the Nepean River, are designed for a water flow of 1.4 metres a second because the juvenile fish will have grown bigger and stronger by the time they reach them. On the Murray-Darling fishways, water velocity is 1.8 metres a second, suitable for the passage of adult perch. The gradients are no steeper than 1 in 15, and turbulence is kept between 40 and 100 watts per cubic metre – little more than half, at most, of the turbulence in salmonid pool-type fishways.

The new fishways have cost between A $250 000 ( £120 000) and A $500 000 to build – between 3 and 5 per cent of the total cost of a new weir, depending on their size and the accessibility of the site. This is two to three times the cost of old-style pool fishways. Are the new fishways worth the extra?

When the Torrumbarry fishway, built at a 6-metre weir on the Murray River, started operating in February 1991 we kept a small-mesh cage over the exit of the fishway to trap fish as they reached the top. Every day, staff at the weir counted the fish before releasing them. Over the 16-month monitoring period, more than 12 000 native fish were counted.

This told us that fish could negotiate the fishway, but not whether all the fish congregating at the weir could find the entrance and make their way to the top. Nor did we know if the fishway was big enough for all the fish that wanted to pass. So, in a jointly-funded research project with the Murray-Darling Basin Commission, we examined the distribution of fish above and below the weir. We put nets across the river in three spots – one above and two below the weir – once a month from January 1990 until June 1992, and compared the numbers of fish caught in each net. During 1990, we caught large numbers of fish in the river below the weir during the spring and summer, and almost none above. But after the fishway opened in February 1991, we found that the number of fish above the weir evened out, with no build-up of fish below queueing to get past.

We then sampled other parts of the Murray-Darling system with traps and nets. One of these sampling locations was 500 kilometres downstream of Torrumbarry, at Euston, where there is an old-style pool fishway, reckoned to be one of the most effective of its type. By reducing the amount of water flowing into the top of the Euston fishway we temporarily decreased the velocity and turbulence at the base so that all the fish attempting to migrate upstream could enter the bottom pool. We collected up to 3000 fish each day in the bottom pool, but only 30 at the top pool, confirming the inadequacy of those designs.

EASY PASSAGE

At the same time, we monitored two of the new coastal fishways: at Penrith and Camden. We focused on the known seasons for migration, and in particular the summer, when juveniles are on the move. We found that even the smallest fish could pass through the fishways, and we collected large numbers of many different species, including Australian smelt and gudgeons, at the top of the fishways.

Having established that the vertical-slot fishway was effective for native fish, we wanted to see how we could improve on its design, in particular to make it cheaper. From field experiments on the swimming ability of perch in the Torrumbarry fishway, it became clear that we had underestimated the speed at which wild fish could swim through the baffles.

Our first thought was that we might be able to cut costs in future fishways by increasing the water speed, and therefore making them shorter and steeper. But we had to think again when we discovered that the fish were taking up to 3 hours to ascend through all 39 pools of the Torrumbarry fishway, even though they can swim much faster than this over a short distance. Clearly they were having to rest on the way up, which suggested that it would not be wise to make their passage through the fishway any more arduous.

We also discovered that the picture is complicated by the behaviour of some species. Whereas many fish in the Murray River move upstream night and day, bony herring, silver perch and Australian smelt only do so during daylight hours. If these fish do not complete their ascent before nightfall, they will drop back down the fishway. In the light of these two findings, we decided to stick with the water velocity we had originally specified for the fishway, despite having underestimated the swimming speed of the native fish.

Even more unexpected than the reluctance of some species to move upstream at night was the discovery of juvenile silver and golden perch in the cages at the top of the Torrumbarry fishway. These immature fish, which were between one-third and two-thirds the size of the adults, accounted for between 65 and 95 per cent of the fish collected at the top of the fishway. Likewise, we found that 99 per cent of the 3000 fish a day we collected at the bottom of the Euston fishway were juveniles. Why juveniles were moving up the fishway is not clear, but it put paid to our assumption that upstream journeys by immature fish are negligible in the Murray-Darling river system.

FUTURE FISHWAYS

At least two more vertical-slot fishways are to be built, one at Liverpool Weir on the Georges River and one at Audley Weir on the Hacking River. The tidal weir at Audley has severely depleted migratory fish, and Australian bass have disappeared completely from the river. Both fishways are being funded by the Department of Public Works and the Environmental Trust, which is financed from levies on companies discharging polluted water into the Sydney sewerage system.

Meanwhile, we are continuing research on the swimming abilities of native fish and looking at the role of movement of immature fish and the environmental cues that stimulate migration (see ‘Factors for fish breeding’). Another area we are tackling is to find ways to ease the passage up and down other rivers for migratory fish. In some of the steepest, traditional fishways, for example, we are experimenting with a prefabricated Denil fishway that rests within the channel of the old fishway. A Denil fishway, named after the Belgian engineer who developed the principle in 1908, is a channel divided by closely spaced U-shaped baffles. These baffles should help reduce the speed of the water in the old, steep fishways.

For weirs not already equipped with fishways, navigation locks could be used to help the fish on their way. At one lock, we tried leaving the gates ajar for the 16 hours of daylight during the main migration season, instead of the 10 minutes they are normally opened for boats to enter the lock. The result was a fortyfold increase in the number of fish passing through. As from this spring, the Murray-Darling Basin Commission plans to leave the gates ajar at 13 locks on the Murray River, making 1300 kilometres of river available during the migration season.

It is now clear that lack of knowledge about migratory fish led to the ineffective design of the older fishways. But Australia is not alone in adopting the pool-type fishway without for non-salmonid fish. Countries as far afield as the former USSR, Latin America, China and Nigeria have built fishways which have proved very poor at passing fish. In future, fishways should only be built after biologists have studied the migrating patterns, swimming ability and behaviour of the native fish. The importance of this was highlighted during our studies of potamodromous species: we found that one of them, the bony herring, would not pass through tunnels. Yet fishways beside dams are often built in tunnels through the abutments.

Our methods for determining the best design for fishways have aroused interest in many parts of the world, including France, Austria, South Africa and the US. But the greatest interest is coming from the Philippines, French Polynesia, Brazil and Bangladesh: in developing countries freshwater fish are an important food resource and non-salmonid species dominate. The consequences can be disastrous if the construction of dams, weirs and other barriers leads to a decline in their numbers. For example, fish is one of the major foods in Bangladesh, but new flood-prevention levees are restricting the movement of fish, which is preventing them completing their life cycles.

With the health of rivers and their ecosystems becoming an increasingly important issue around the world, improvements to the design of fishways have come none too soon.

Martin Mallen-Cooper is a biologist at the Fisheries Research Institute in Cronulla, New South Wales.

* * *

Factors for fish breeding

In efforts to increase the number of fish in a river, fishways are only half the battle. Even if fishways get fish past weirs, the water must be managed to present fish with the right cues for stimulating migration and breeding.

Many fish time their migration by changes in the level of the river or in the flow of water. The winter migration by catadromous species towards their estuary spawning grounds is stimulated by faster flows. Similarly, their return in early spring and the juveniles’ move upstream from the estuary in late spring and summer could be triggered by changes in water flow.

For most potamodromous species, the springtime rise in the water levels stimulates the migration upstream. Some species, such as the Australian smelt and the silver perch, move upstream during spring and summer, or try to, during small rises in the water level. This is particularly true of the juveniles.

Dams can also interfere with two important triggers for breeding: flooding and temperature. Because they even out river levels, dams can affect fish which spawn only during a flood. Some fish whose breeding patterns are attuned to temperature, have become locally extinct below dams that release their cold water into the river during the summer, so that it can be extracted further downstream for irrigation.

In the past, river management schemes have made irrigation and flood control their priorities. From our research at Torrumbarry we can now recommend what flows and water levels are needed for the migration of fish. The Department of Water Resources of New South Wales is implementing a management plan in one river system that will maintain flows during these critical periods, instead of pumping it off into storage for use in irrigation later in the season.