The Tay Foundation, in partnership with the Tay District Salmon Fisheries Board and Scottish and Southern Energy plc has supported a study into the effects of flow regulation by hydro dams on invertebrates in the River Lyon. Preliminary studies indicated that damming had resulted in a change to the river’s temperature regime and perhaps other factors and that this may have caused a reduction in the abundance of invertebrates, and consequently fish stocks.

The three year study, a PhD study conducted by the Department of Geography and Environment of the University of Aberdeen, was submitted in the summer of 2006.

Dr. David Summers

The importance of the River Lyon

If on a glance at a map of the Tay system, one had to pick out what ought to be the most important tributary for salmon it would surely be the River Lyon. This long tributary feeds practically every beat on the Tay and should surely be considered the enginehouse of the Tay. Indeed, in the past this would seem to have been so. The Lyon was long famed as a foremost salmon river producing large spring fish. The classical commentators, Augustus Grimble and W.L. Calderwood wrote at length on the upper Tay beats, and the Lyon. In 1921 Calderwood remarks of the Lyon: “any one who has seen the ascent of fish at the Gallan Falls, or Falls of Meggernie, has at any rate been able to realise the great number of fish which the river holds”. However, the same does not appear to be true today.

The Lyon Past and Present

A great change was made to the River Lyon in the 1950s when it was harnessed for hydro-electric production.

Prior to that time, the River Lyon rose in Loch Lyon, a modest upland loch, and flowed onwards for some 50 km to it’s confluence with the Tay. In character the Lyon varied repeatedly between gravelly spawning areas, boulder strewn reaches, sluggish holding pools and rocky gorges. The walls of Glen Lyon present a sheer face and most of the tributary burns are only accessible to salmon for a few hundred yards. The most accessible stream was the Meran, joining the Lyon immediately below the Loch.

The Breadalbane Constructional Scheme entailed the construction of a dam at Lubreoch several miles below the Meran confluence, impounding a new much bigger Loch Lyon. No provision was made to allow salmon passage through this dam as the limited habitat left above did not warrant it. In addition another dam, Giorra Dam, was constructed in a glen to the north east creating a new Loch Daimh out of two small lochs, Daimh and Giorra. To increase the catchment areas of these new reservoirs, water was intercepted and transferred from feeder streams in neighbouring catchments. Loch Lyon received water from the Dochart, Lochay and Orchy and Loch Daimh received water from the Lyon itself. Thus the catchment area of the Loch Lyon was increased by some five-fold over the natural loch.

The water stored in Loch Lyon is released into the River Lyon through a turbine at the base of Lubreoch Dam. Two miles downstream the river is arrested by another smaller dam at Stronuich. At this point water from Loch Daimh is also received via a tunnel bored for several miles through the hills. Most of the water collected in the headpond at Stronuich Dam is diverted into another tunnel through the hills to a power station in the neighbouring Lochay catchment which ultimately discharges into Loch Tay. Thus after a circuitous route, water abstracted from the upper Lochay along with water from the Lyon, Dochart and Orchy finds it’s way back again into the lower Lochay.

At Stronuich Dam, the Lyon proper may be said to commence. A “compensation” flow is released from this point which can vary between 30 million gallons per day and 48.6 million according to the time of year. Freshets are also periodically released. However, a Borland fish lift has been incorporated into this dam so that salmon can and do penetrate right up to Lubreoch Dam.

Further downstream from Stronuich a number of other tributaries, the Milton Burn, the Allt Gleann Da-Eig and the Allt a’ Chobhair are also abstracted, the water from these also being transferred into the Loch Tay catchment.

The character of the Lyon

The character of the Lyon varies. Immediately below Lubreoch Dam flows can vary considerably according to the generation regime. However, below Stronuich Dam they are generally very stable, except occasionally when the dam spills after a great deal of rain. With distance from the dams, flashy tributaries do enter the river and gradually the Lyon reassumes the character of a spate river.

A noticeable feature of the Lyon, is that there tends to be profuse growth of filamenteous algae. This is true from the tailrace of Lubreoch Dam, continuing sometimes for over twenty miles to the lower reaches. This is not a recent phenomenon. In 1965 Lyon fishing interests wrote to this Board complaining of this very subject. When the algae is scraped off stones from the bed in the upper half of the Lyon, they are covered in a black precipitate. Tests have shown this precipitate contains the metals aluminium and manganese. The stones in the tributary burns, on the other hand, sparkle.

Electrofishing surveys

To gain some insight into the productivity of the Lyon, an electrofishing survey was conducted at various sites along the main stem of the river in August 1999 and the mainstem and selected tributaries in late August / September 2000.

Electrofishing was performed by a generator driven pulsed DC machine. At each site a single pass was made by the electrode and fish caught in a banner net. It is not usual to catch all fish on one pass of the electrode, so the numbers caught must be regarded as minimum estimates of numbers present. While the figures obtained are not precise, they may be considered a relative measure as an identical technique was used at all sites.

Results of electrofishing surveys

The most meaningful measure of the productivity of salmon in a river from electrofishing is the density of large “pre-smolt” parr present. It is considered that parr greater than about 90mm in length (8 grammes in weight) in the autumn are likely to become smolts the following year and smaller fish will not. Densities of fish over this size are presented for 2000 in Figure 2 (data for 1999 were presented in the 1999 Annual Report and show a similar pattern). For comparative purposes data are also presented from sites electrofished in other upland Tay tributaries.

Figure 1. River Lyon Electrofishing Sites, September 2000

Figure 2. Densities of pre-smolt parr (>90mm), autumn 2000

It is noticeable that densities of pre-smolts in the mainstem Lyon are generally the lowest found, especially the upper Lyon. They are even higher in the Lyon’s tributaries, and much higher in the upper tributaries of the River Ericht for example. This difference in pre-smolt numbers is borne out in differences in catches between these rivers. Some beats on the Ericht system catch about as many fish as the entire Lyon in a year (about 200).

One reason for the low numbers of pre-smolts in the Lyon must be their growth rate, which appears to be the poorest in the Tay catchment. This is demonstrated by Figures 3 and 4. These figures represent plots of the average weights of salmon fry and 1 year old parr for different sites across the upper Tay catchment against the density of parr present. The reason the data are presented in this manner is to account for the impact of competition between fish and allow a clearer picture of the true productivity of the river. Fry are especially disadvantaged by the presence of older parr, which being much stronger, force the fry into sub-optimal habitats along the margins of streams where they find less food.

It is noticeable from Fig. 3 that, for any given parr density, fry in the mainstem Lyon grow slowly relative to elsewhere, including Lyon tributaries. Most astonishing is a burn called the Allt Chiorlaich only a mile from Lubreoch Dam where the fish are at high density but still averaged 2.5 grammes. In the adjacent mainstem Lyon fry were present at one tenth the density and yet only weighed 0.7 grammes. Even in the mainstem of the River Tilt, the wildest and most barren looking river in the district, where parr densities are as low as the Lyon fry, growth rates are considerably higher. In the Ericht they are much higher. The same also holds true for one year old parr (Fig 4), except that in the Lyon tributaries the difference is lost. However, because the Lyon tributaries are only accessible to salmon for a short distance, it may be that there is interchange in parr with the mainstem Lyon and so we may not be entirely looking at fish which have lived all their lives in separate environments.

Slow growth means that in the upper half of the Lyon at least, most smolts are three years old, as opposed to two years elsewhere. Thus the juvenile habitat is occupied for longer and the turnover and overall productivity of the population is less.

Figure 3. Density of parr plotted against mean weight of fry, September 2000

Figure 4. Density of parr plotted against mean weight of one year old parr, Sept. 2000

Temperature

One factor which might account for the slow growth in the Lyon is temperature, which does have a profound effect on the growth of juvenile salmon. Below 7 degrees salmon cease to feed, fastest growth occurring at 16 degrees.

However, it has widely been found that in rivers which are regulated by dams, especially those which release water from the bottom of the dam, spring and summer water temperatures can be reduced. This is especially of concern because the most important period for growth in young salmon is the spring. To investigate this possibility, in the early spring of 2000, automatic water temperature loggers taking a reading every hour, were installed at three locations on the River Lyon. One is located at Lubreoch at the outflow from the base of the Dam, another at Kenknock a mile downstream from Stronuich Dam and the third at Invervar, some 15 miles downstream from Stronuich.

Figure 5. River Lyon temperature, spring / summer 2000

Figure 6. River Lyon temperature, spring 2001

The findings for the spring and summer of 2000 and spring of 2001 are shown in Figures 5 & 6. These show that during the spring, the temperature of the water leaving the dam is generally considerably colder than would naturally be the case. In both years there was warm weather in early May which succeeded in raising temperatures at Invervar into the teens, while the water leaving the dam did not exceed 8 degrees. Other spot recordings taken in warm afternoons in May 2000 suggest that temperature suppression was indeed still evident at the Kenknock site, but that by Meggernie (10 miles from Lubreoch) the effect had attenuated. Reduced temperatures were also found in the neighbouring River Lochay downstream of the power station which receives water from Stronuich Dam.

It is also interesting to note that during the spring and early summer, the water temperature at Lubreoch remains stable for periods and increases in a stepwise manner. The stable periods tend to coincide with warm periods, as evidenced by the temperatures at Invervar. The jumps occur when ambient temperatures would appear to reduce. This indicates that thermal stratification occurs in Loch Lyon when it is warm and calm (i.e. a layer of warm water, a “thermocline”, floats over colder denser water in the bottom of the loch). However, cooler weather or wind causes the water in the loch to mix, causing bottom temperatures to rise.

Presumably temperature must have an impact on the growth of juvenile salmon, and the insects on which they feed (see below), on the upper Lyon. However, it is the case that there is no really noticeable increase in growth of juvenile salmon by Meggernie, so temperature cannot be the only factor at work.

Invertebrates

In order to account for the slow growth of juvenile salmon it is necessary to investigate the insects in the river on which the fish feed.

Full analysis of aquatic insect populations is an involved process. Owing to their life cycles, many insect larvae are present or are only of a catchable size for a short part of the year and this differs according to species. Thus sampling really needs to be repeated throughout the year. Also, because different species prefer different types of physical habitat, these also must be sampled.

So far it has not been possible to rigorously sample the Lyon invertebrates. However, in June-August 2001, some samples were obtained, and notwithstanding the incompleteness of the data, they do give an indication of something unusual.

Larvae were sampled by the “kick sampling” method. A stout fine meshed net is held in the current and the stones and gravel immediately upstream are disturbed by kicking and shuffling the feet. By trying to restrict each sample to a standard area and time (2 minutes in this case) some crude measure of abundance may be obtained. It should be stressed, however, the strength of current, looseness of stones, swimming ability of larvae etc all conspire to vary the sampling efficiency between sites.

Owing to time constraints larvae were generally not identified to species level, but rather to broader groups. For an initial scoping study this is adequate.

The following groups or species of larvae were identified:

Ephemeroptera (the “mayflies”)

Baetis

These are known to anglers as the “olives”. In subsamples identified to species level, Baetis rhodani (the “large dark olive”) is most common, Baetis muticus (the “iron blue”) also being found. These agile larvae are good swimmers and eat algae off the surface of stones plus rotting detritus.

Relative to most of our invertebrates Baetis have short life cycles and are “multivoline” – i.e. they produce more than one generation per year, unlike most insects in our streams. This means that instead of only being abundant during a short season, Baetis larvae can be abundant for most of the year. Because of this Baetis rhodani is regarded as being one of the most successful freshwater invertebrate in Britain.

In a number of studies baetids have been found to be a most important food item for juvenile salmon, especially fry. This is due to two factors. They are available throughout the year, and also as they swim on the top of stones they are easy prey.

Ephemerella

These will almost certainly be Ephemerella ignita the “blue winged olive” or “sherry spinner” to trout anglers. This larvae lives on detritus (decomposing organic matter – algae, leaves etc) and crawls on the bed as opposed to swimming. It also has a distinct seasonal abundance (summer only). It has less inherent value as food for juvenile salmon than Baetis.

Caenis

These are the smallest of the mayflies and live amongst the sediment and debris on the bed. The also eat detritus. All examples caught are probably Caenis rivulorum.

Heptagenia

This group of mayflies eats algae off stones or detritus. They have a characteristic flattened streamlined shape to allow them to cling to stones in fast currents. They are not particularly good swimmers like Baetis and mainly crawl. They include flies like the March Brown (Rithrogena semicolorata) and a large bright yellow fly (Heptagenia sulphurea) that is sometimes called the “Mayfly” in Scotland or the Yellow May Dun on chalkstreams. Heptagenia are associated with rough streams especially.

Plecoptera (“Stoneflies”)

There are many different species of stoneflies in our rivers. However, apart from one species Perla bipunctata (the large stonefly), a large carnivorous larvae of very distinct appearance, the stoneflies were not differentiated further. Stoneflies are associated with water of very high quality. Some, like Perla are carnivorous and others like Leuctra, which were probably the main group encountered, eat detritus. Stoneflies can be an important food for salmon.

Tricoptera (“caddisflies)

Hydropsyche

This family of caddis flies does not build a case to live in and is of a grub-like appearance. Hydropsyche feed on fine detrital material or plankton which is filtered from the current in fine silk nets which they spin on stones for that purpose. They live very much in or under the substrate.

Ryacophila and Polycentropus

These also are caseless caddis larvae, but are predators of other insect larvae. Again they crawl amongst and under the stones and so may be of lesser value to salmon.

Cased caddis

These include many species of which have not been identified further. Cased caddis exibit a range of feeding behaviours. Some only eat detritus but some do scrape algae too.

Diptera (Flat winged flies)

Chironomids (midges)

There are many different species of chironomids and these have not been differentiated. Likewise their habitats also vary. Chironomids are small larvae and often prolific and so have widely been found to be a most important salmon food. They are often associated with silt deposits into which they burrow.

Simulium (blackflies)

Simulium larvae attach themselves by one end to stones or plants by sticky threads. They sway in the current and filter detritus from the water. They can be prolific being multivoltine, and can be an important food for salmon. Being attached in exposed places, they are picked off by juvenile salmon.

Gammarus (freshwater shrimps)

These again are detritus eating, living on dead leaves and other matter accumulated on the river bed.

Results of invertebrate survey

The numbers of different types of invertebrates caught at different locations are shown on page 34.

The general pattern is that in the main stem of the Lyon detritivores such as Ephemerella and Hydropsyche caddis flies are most abundant. It should be noted, however, that at Pubil and Stronuich, Ephemerella were not observed on 13 June. However, they were present at Pubil on 3 July and also in an unrecorded sample at Stronuich on 16 July, many of which were small. They were very abundant at Cashlie, 500m below Pubil, on 5 August. Ephemerella is known to have a long egg incubation followed by rapid larval growth in summer (Brittain and Saltveit, 1989). The delayed hatching of Ephemerella in the upper Lyon may be related to the lower water temperature experienced in spring.

In the tributaries of the Lyon, Ephemerella and Hydropsyche were scarce, the fauna generally being dominated by Baetis. This difference was not only relative, but absolute. In the main stem Baetis were generally scarce. The only exception was the site at Milton Eonan which was also unusual in respect of Heptagenia and especially Simulium which were scarcely found elsewhere in the upper Lyon. However, the current at this site was much stronger than at any of the others and so there might have been local habitat factors at work.

In fact looking more widely at samples from other upland rivers in the district, it can be seen that Baetis tend to be plentiful. The more “upland” the stream the greater the dominance by Baetis appears to be. Ephemerella become more numerous as streams get larger and presumably richer in detritus. It is notable that the Ephemerella production in the Tay at Kenmore is vast.

The stonefly Perla bipunctata was not found at all in the main stem of the Lyon but was found in all the Lyon tributaries. Looking across the sites, this species seems to occur in the type of unenriched upland streams where the Baetis are most dominant.

A notable stream is the Allt Chiorlaich, where in 2000 fry grew by far the fastest (Table 1, p. 34). Owing to drought this stream was actually dry in May 2001. On sampling in early July after it had started flowing again few invertebrates were found. However, on another sample on 5 August, it was found to have a very high number of Baetis present.

Discussion

A number of studies have shown that Baetis, if present, are a very important component of the diet of young salmon, especially fry (Maitland 1965, Egglishaw 1967), as can Simulium (Egglishaw 1967). It is tempting to speculate that the apparent lack of these groups in the mainstem of the Lyon might be a major element in the salmon’s problems.

The lack of these invertebrates may in some way be related to the profuse growths of filamenteous algae. Algal growth can smother every stone in the river for over 20 miles on occasions (personal observations). However, in the Lyon tributaries the stones “sparkle” and have only ever a fine coating of algae. Presumably in the tributaries, the invertebrates find such algae as there is palatable and probably succeed in grazing it down. Perhaps in the main stem they find the algae unpalatable. This algae has been tested and found not to contain toxins. However, it is apparently the case that filamenteous algae is rarely eaten when alive (Allan 1995). Simulium have also been found to do badly where algae is profuse as they cannot obtain anchoring sites (Boon 1988).

A possible cause of the profuse growth of algae is the fact that the Lyon is regulated. Indeed, as referred to earlier, proprietors on the Lyon complained about the amount of “slime” in the river even in 1965 which they then believed was a consequence of the water originating in reservoirs.

It is the case that reservoirs can have profound effects on the ecology of rivers downstream, especially those from which water drains from the bottom of the dam. In a review based on American experience as early as 1963, Neel included among possible impacts, profuse algal growth, temperature and chemistry changes and changes to the flow regime and sediment movement. Regarding chemistry, the discharge of iron, manganese and sulphides were specifically mentioned.

Boon (1988) reviewed studies of the downstream impacts of reservoirs in the U.K. He remarked on the paucity of studies relative to North America, citing a number of studies in northern England and Wales, but surprisingly for the number of dams, none at all in Scotland.

The biological effects of dams varies greatly according to local factors. In some instances Baetis have increased, in others decreased (Brittain and Saltveit 1989). For example, in the River Elan, Wales, which was dammed a century ago to supply water to the Midlands, mayfly species have been severely reduced including Baetis. In this river the stones are coated in a curious deposit containing iron and manganese, which was considered responsible. As referred to earlier, the stones in the upper Lyon are in fact coated in a black precipitate which testing has shown contains aluminium and manganese. The significance of this is as yet unknown.

Downstream of dams there can often be an increase in invertebrates which live on detritus. Increases in Ephemerella are commonly reported (Brittain and Saltveit 1989) as have increases in Hydropsyche (Boon 1988). Brittain and Saltveit quote a German example where Baetis and Heptagenia decreased and Ephemerella increased, similar to the Lyon. In general the stone clinging Heptagenia appear to be adversely affected by profuse algal growths below dams (Brittain and Saltveit) as are Simulium, even though they as filter feeders might be expected to do well (Boon 1988).

The life cycles of invertebrates have also been found to be modified by changes in temperature regime. Decreased spring and summer temperatures can delay hatching and reduce growth, and in some cases can result in the elimination of particular species (Boon 1988). A temperature impact on Ephemerella, is noticeable in the upper Lyon.

The absence of Perla bipunctata, as here, has been reported elsewhere (Boon 1988).

Conclusions

On the basis of numerous studies performed elsewhere, it appears possible that the differences in invertebrate fauna between the mainstem River Lyon and it’s tributaries are some consequence of the damming of the headwaters and subsequent regulation of the river. While cause and effect have not been truly established it is tempting to speculate that the differences in invertebrate fauna have resulted in a combination of reduced growth rates and abundance of juvenile salmon, effecting a loss in salmon production from this system.

Exactly, what the cause is will require further study to determine how invertebrate populations and algal abundance changes seasonally along the river and it’s tributaries. This then needs to be related to factors like temperature, hydrology and water chemistry. When the mechanisms are clearly understood, then perhaps some ameliorative action might be identified.

References

Allan J. D. (1995) Stream Ecology. Chapman & Hall. London.

Boon P.J. (1988) The impact of river regulation on the invertebrate communities in the U.K. Regulated Rivers: Research and Management, 2, 389 – 409.

Brittain J.E. and Saltveit S.J. (1989) A review of the effect of river regulation on Mayflies (Ephemeroptera). Regulated Rivers: Research and Management, 3, 191-204.

Egglishaw H. J. (1967) The Food, Growth and Population Structure of Salmon and Trout in two Streams in the Scottish Highlands. Freshwater and Salmon Fisheries Research 38. HMSO, Edinburgh.

Maitland P.S. (1965) The feeding relationships of salmon, trout, minnows, stone loach and three-spined sticklebacks in the River Endrick, Scotland. Journal of Animal Ecology, 34, 109-133.

Neel J.K. (1963) Impact of Reservoirs. In Limnology in North America (ed. D.G. Frey), pp. 575-593. University of Wisconsin Press, Madison, Wisconsin.

Acknowledgements

Thanks are due to the Tay Foundation for purchasing the temperature loggers used in this study and the FRS Freshwater Fisheries Laboratory for arranging the chemical testing of algae and sediment samples. I also wish to thank Jane Keay of FFL for invaluable assistance in the identification of invertebrates.