Lake Trout Ecosystems in a Changing Environment - Chapter 5 pot

Climatic warming and reservoirsImpacts of reservoirs use on lake trout reproduction Characteristics of reproduction sites Effects of raised water level Effects of water level fluctuation

Climatic warming and reservoirs

Impacts of reservoirs use on lake trout reproduction

Characteristics of reproduction sites

Effects of raised water level

Effects of water level fluctuations

Management recommendations for lake trout in reservoirs

Integrated management of water and wildlife

Introduction of a strain of deep-spawning lake trout

Creating deep-zone spawning areas

Even though there are many reasons for creating reservoirs (such as drinking water supply,

reason for the existence of reservoirs on the Precambrian Shield Exploitation of Canadian

hydroelectric potential began in the last century and has advanced northward like a wavethat is now cresting in the midnorthern latitudes (Rosenberg et al., 1987) Already, thetotal area of boreal reservoirs in North America is similar to that of Lake Ontario (Rudd

et al., 1993) The environmental impacts that result from the creation of reservoirs are

numerous and include accumulation of sediments, shore erosion, and accumulation ofmercury in fish and other organisms

There are upwards of 100 reservoirs in Québec, 68 of which are known to supportlake trout populations These represent 7.4% of the province's lake trout lakes In Ontario,the areas with the most concerns about reservoirs and water level variations are theAlgonquin and Eastern regions (Lewis et al., 1990) Together they comprise nearly 22% ofOntario’s lake trout lakes

The lake trout Salvelinus namaycush is among the most studied species in North

America with regard to reservoir management Lake trout have great difficulty adapting

to lakes with regulated water levels or to reservoirs used for hydroelectric generationpurposes (Martin, 1955; Wilton, 1985; Gendron and Bélanger, 1993; Benoît et al., 1997).The change in water level, the resulting surface area and type of substrate of the floodedlands, the seasonal drawdown regime employed, and the lotic or lentic origin of thereservoir are some of the many factors affecting lake trout’s adaptation in reservoirs(Machniak, 1975; Evans et al., 1991) Among the foregoing factors, drawdown is oftenpinpointed as the cause of the lake trout's adaptation problems, as it has repercussions

on two aspects of the species’ vital cycle: reproduction success and larvae survival (Martin,1955; Wilton, 1985; Gendron and Bélanger, 1993)

This chapter reviews major modifications made to lake trout ecosystems by the ation of reservoirs on the Precambrian Shield It also discusses impacts of drawdown onthe reproduction of lake trout and suggests management alternatives

cre-Figure 5.1 Reservoir for production of hydroelectricity

Transformations to ecosystems

As a result of reservoir creation and management practices, temporary and permanenttransformations to ecosystems will occur These transformations can be of greater or lessersignificance depending on the reservoir’s intended use (source of drinking water, electricalenergy generation, flow regulation) and its particularities: shape and mean depth, reten-tion time, surface area of flooded lands, density and nature of vegetation flooded, as well

as length of impoundment period (Baxter and Glaude, 1980) Without a doubt the mostspectacular and complex changes, from both physical and biological standpoints, arebrought about by the creation of a reservoir on a river Such a reservoir entails permanenttransformations to ecosystems, as vast tracts of lands are flooded and river stretches areturned into lakes

A lake can also be transformed into a reservoir by regulating its flow In such cases,the transformations to ecosystems are less extensive and, in some instances, only tempo-rary The following section examines major modifications to lake trout ecosystems broughtabout by the creation and management of reservoirs located on the Precambrian Shield

The decomposition of flooded organic matter then follows This phenomenon leads

to consumption of dissolved oxygen, mainly in the deeper strata of the reservoir, lower

oxygen levels and change in the chemical conditions of deep-zone waters are factors likely

to limit the habitat of young lake trout, increase the threat of predation by adults, andultimately reduce recruitment to the population

However, overall water quality within the reservoir is barely influenced by thesebenthic processes Because the volume of water rich in decomposition by-products nearthe bottom is very small relative to the total volume of the reservoir, the concentrations

of these products throughout the reservoir remain low following spring overturn.All of these changes are temporary and subside as time passes (Baxter andGlaude, 1980) For example, in the Robert-Bourassa and Opinaca reservoirs (Québec) as

a whole (Chartrand et al., 1994), physicochemical variations peaked quickly (1 to 4 years)following impoundment (Figure 5.2) Modifications related to the decomposition offlooded organic matter were nearly over 9 to 10 years after impoundment With respect

to the Caniapiscau reservoir, Québec (Chartrand et al., 1994), the modifications measuredwere of the same magnitude as those measured in other reservoirs; however, the maximafor total phosphorus and silica were reached later, between the 6th and 10th year ofimpoundment (Figure 5.2) In this particular instance, the return to values representative

of natural environments was completed 14 years later It appears that because ment occurred more gradually, over a period of 3 years rather than 6 to 12 months as withthe others, the period required for the return of initial conditions was extended

impound-The modification period is brief largely because only a small portion of the floodedorganic matter composing the forest soil and vegetation decomposes easily and rapidly.Only the leaves of trees and bushes, conifer needles, forest ground cover, and the first fewcentimeters of humus decompose rapidly Most of the other flooded material (tree

branches, trunks and roots, and deep soil humus) proves to be difficult to decompose andremains really intact dozens of years after impoundment.

Mercury is a widespread contaminant in freshwater fish Lake trout frequently havehigh concentrations of mercury because of their position at the top of the food chain andbecause the Boreal lakes that they inhabit often have conditions favorable for mercury

Several studies have demonstrated that impoundment brings about a rapid increase

in fish mercury levels (Schetagne et al., 1997) The extent of the increase in bioavailability

of mercury for aquatic wildlife in reservoirs depends on many factors: the land areaflooded, filling time, water residence time, volume of water, proportion of flooded land

in shallow environment (where biotransfer is at its maximum), water quality, food web

of the flooded environment, fish population dynamics, etc (Jones et al., 1986; Brouard etal., 1990; Doyon et al., 1996)

Research conducted at the La Grande complex (Québec) (Schetagne et al., 1997)showed that depending on the fish species and reservoir considered, maximum mercuryconcentrations were 3 to 7 times higher than those measured in natural environments Innonpiscivorous species, mercury levels stop increasing significantly 4 to 5 years afterimpoundment, and the return to concentrations representative of natural environments iswell under way 10 to 15 years after flooding In piscivorous species, maximum valueswere attained later than for nonpiscivorous species Maximum concentrations for walleye

Figure 5.2 Variation of principal water quality variables before, during, and after impoundment inthe major reservoirs of the La Grande complex (⎯ Robert-Bourassa, … Opinaca, - Caniapiscau)

(From Chartrand et al , 1994, Commission Internationale des Grands Barrages, Dix-huitième

Con-grès des Grands Barrages, Durban, Q.69-R-14, pp 165-190.)

Stizostedion vitreum, northern pike Esox lucius, and lake trout were reached between 9 and

13 years after impoundment, depending on the reservoir and species

A gradual decrease in fish mercury levels is observed once mercury release activities,including decomposition of flooded organic matter, as well as the erosion and resuspen-sion of flooded organic matter along banks exposed to wave action have declined Sub-sequently, this erosion of organic matter helps accelerate the decrease in fish mercurylevels by reducing the area of shallow zones in reservoirs that still have organic matteravailable It is in these shallow zones rich in organic matter where most biotransfer occurs(Shetagne et al., 1997) In the majority of reservoirs, a significant decrease begins to beapparent, for all these species, 14 to 15 years after reservoir creation

The data collected at the La Grande complex, as well as in other reservoirs located inthe Precambrian Shield, show that fish mercury levels in reservoirs return to values similar

to those measured in natural environments after a period that may range from 15 to 25years for nonpiscivorous species and from 20 to 30 years for piscivorous species (Shetagne

et al., 1997)

Water impoundment also affects other aspects of water quality and because thechanges result from major alterations in the reservoir’s shape and management, they arepermanent For example, the creation of a reservoir on a river enables the particles thatwere suspended in the river’s running waters to settle to the bottom more easily Thissedimentation process reduces turbidity and increases light penetration into the water(Chartrand et al., 1994)

Plankton

A rise in nutrients, particularly phosphorus, which generally limits phytoplankton duction on the Precambrian Shield, can cause an increase in phytoplankton quantity.Phytoplankton abundance had increased by five times in Lake Minnewanka, Alberta 3years after its level had been raised (Cuerrier, 1954) and had returned to initial levels 8years later The rise in nutrients noted in all reservoirs at the La Grande complex (Char-

pro-trand et al., 1994), particularly phosphorus, resulted in a threefold increase in chlorophyll

a concentration (Figure 5.3) At the Robert-Bourassa and Opinaca reservoirs, maximumconcentration was reached 3 to 5 years after impoundment Once easily decomposed

organic matter was depleted, nutrient levels dropped to initial levels and chlorophyll a

concentration returned to values similar to those prevailing before impoundment Thereturn to initial values occurred 9 and 10 years, respectively, after impoundment of thosereservoirs At the Caniapiscau reservoir, maximum values were attained some 10 yearsafter impoundment, and the return to initial values was nearly completed after 14 years.Zooplankton abundance and biomass in reservoirs are influenced by water enrichmentand availability of organic matter, produced by flooded vegetation and forest soils, as well

as increased retention time At the Robert-Bourassa reservoir (Chartrand, 1994), ton density and biomass reached maximum values in the fourth summer after impound-ment and dropped off slowly afterwards (Figure 5.3) The maximum values were attained

zooplank-a yezooplank-ar zooplank-after chlorophyll zooplank-a (phytoplzooplank-ankton biomzooplank-ass) (Figure 5.3) zooplank-and phosphorus

the reservoir creations, most notably the flooding of forest and the formation of huge lakesfrom rivers.

After the initial decrease in benthos abundance and change in their compositionfollowing the water level rise in lake Minnewanka (Cuerrier, 1954), benthos increased in

Figure 5.3 Chlorophyll a variation in the major reservoirs of the la Grande complex (—

Robert-Bourassa, … Opinaca, - Caniapiscau) and evolution of zooplankton and benthic organisms biomass

at Robert-Bourassa reservoir following impoundment (From Chartrand et al , 1994, Commission

Internationale des Grands Barrages, Dix-huitième Congrès des Grands Barrages, Durban,

Q.69-R-14, pp 165-190.)

Chlorophyll a

0,6 1,0 1,4 1,8 2,2 2,6 3,0 3,4 3,8

Biomass of Benthic Organism (Annual Means)

0,00 0,15 0,30 0,45 0,60 0,75 0,90 1,05 1,20

abundance over the next 10 years, showing adaptation to new conditions However, thelittoral zone, which is subjected to drawdown 5 months per year, is not productive.Significant variations in water level hinder the growth of vegetation in the riparian zoneand consequently limit the abundance of benthic fauna.

In the reservoirs of the La Grande complex (Chartrand et al., 1994) benthos diversitydecreased, mainly in the first years While species with poor mobility or adapted torunning water became scarce, those with greater mobility or requiring little dissolvedoxygen rapidly took over the new aquatic habitats The great number of anchoring pointsoffered by the flooded vegetation increased the surface area of feeding grounds, leading

as a result to measurements of greater benthos densities and biomass than in natural lakes(Figure 5.3)

Analysis of the stomach contents of reservoir fish and monitoring of fish populationsrevealed that benthos diversity and quantity were sufficient to sustain major increases ingrowth rate and condition factors of fish feeding on benthos These fish included lake

whitefish (Coregonus clupeaformis), and their predators, such as the northern pike (Esox

lucius).

Fish abundance

It is known that fish populations are often numerous in the first years of a reservoir'sexistence (Ellis, 1941, in Baxter and Glaude, 1980) In some cases, reservoir creation canalso help increase the fisheries resources of a region (Baxter and Glaude, 1980) The rapidincrease in abundance of certain fish species often observed in new reservoirs may haveoccurred for a number of reasons Among these could be increased reproduction ratebrought about by secure spawning grounds and protection of fry afforded by floodedvegetation Increased food availability is another

At the time the La Grande complex reservoirs were impounded (DesLandes et al.,1995), overall abundance for all species dropped significantly but then rose over the next

3 years before decreasing slightly up to the 10th year The drop noted in the first yearcould have been related to rising water levels and the dilution effect this entails Thesubsequent increase in abundance varied in range and duration depending on the species.While abundance of species such as the northern pike and lake whitefish greatly increased,

the abundance of longnose sucker Catostomus catostomus, white sucker Catostomus

com-mersoni, walleye, and cisco Coregonus artedi, declined sharply following the initial increase.

The northern pike is a species that generally does well in reservoirs In the first years

of a reservoir’s creation, its abundance and growth usually increase (Machniak, 1975).When the La Grande 2 (DesLandes et al., 1995) and Caniapiscau (Belzile et al., 2000)reservoirs were created, the relative abundance of northern pike rose sharply and remainedsteady up to 10 and 17 years, respectively, following impoundment

Although few data are available on lake trout response to the trophic pulse followingreservoir creation, the ecosystem changes do not seem favorable to lake trout (DesLandes

et al., 1995)

Lake trout populations are generally low in Québec’s reservoirs despite the fact thatreservoirs present abiotic and biotic factors considered very good for the species (Lacasseand Gilbert, 1992; Gendron and Bélanger, 1993) Québec’s reservoirs containing lake troutpopulations tend to have much larger surface area than most natural lake trout lakes inthe region (Table 5.1) This pattern is also reflected to a lesser extent in the mean andmaximum depth The physicochemical properties of many of the reservoirs on the Pre-cambrian Shield provide good life-sustaining conditions for lake trout populations: near-neutral pH, high concentration of dissolved oxygen, and cold temperatures (Table 5.1)(Gendron and Bélanger, 1993; Benoît et al., 1997) However, adverse effects on fish popu-

lations spawning in shallow waters can occur if reservoir drawdown exposes spawninggrounds during egg incubation or larva development

In Québec, generally speaking, the most abundant lake trout populations are found

in deep (maximum depth >30 m), small area (<1,500 ha) reservoirs with an annual down below 1.6 m (except for the Mitis reservoir: 3.0 m) Conversely, populations are lessabundant in large upstream reservoirs (>25,000 ha) with strong annual and interannualdrawdowns (7.8 m), where impoundment resulted in a sharp rise in water level (Gendronand Bélanger, 1993) However, even though the habitat appears suited to the species anddrawdowns are relatively low, in reservoirs created from rivers, lake trout populationsare scarce

draw-More recent and specific studies of five Québec reservoirs harboring lake trout

pop-ulations (Benoît et al., 1997; Doyon, 1997) revealed that the poppop-ulations in four of the five

reservoirs had been decimated and exhibited significant recruitment problems, most likely

in response to drawdown effects

Climatic warming and reservoirs

in this chapter, these gases are the major end products of the microbial decomposition of

(Duchemin et al., 1995; Kelly et al., 1997) have demonstrated that reservoirs are sources

of these gases to the atmosphere However, the net effect of reservoir creation relative toother electric generation options (more specifically: gas, oil, coal), in terms of greenhousegas production, is controversial

Moreover, predictions of a warmer and drier climate resulting from greenhouse gasaccumulation might shift the balance between evaporation and precipitation, which in

turn will lead to overall declines in both river flows and lake levels (Magnuson et al.,

1997) Under this scenario, reductions in runoff will negatively impact hydroelectric powergeneration, thus creating a demand for new dams Building reservoirs in PrecambrianShield would flood more wetlands and terrestrial soils, thus further contributing to cli-matic warming by increasing greenhouse gas fluxes to the atmosphere

Finally, potential impacts of climate warming on reservoirs are reductions in nutrientloading and recycling for many lakes on the Precambrian Shield (Schindler and Gunn,

Table 5.1 Characteristics of Québec's Natural Lakes and Reservoirs having

a Lake Trout Population

Parameters Natural lakes ReservoirsSurface area (ha) 766 ± 7332 28421 ± 73187

N = 906 N = 67Mean depth (m) 14.1 ± 8.6 22.2 ± 14.5

N = 171 N = 35Maximum depth (m) 36.7 ± 26.1 64.4 ± 54.6

N = 359 N = 45Secchi disk transparency (m) 5.2 ± 1.9 5.7 ± 2.9

N = 283 N = 14Conductivity (µS/cm at 25°C) 46.8 ± 56.7 33.0 ± 31.8

Chapter 8, this volume) The change of thermal regime would also cause shrinkage ofsummer habitats for cold-water fish species such as lake trout.

Impacts of reservoirs use on lake trout reproduction

The following

first examines the impacts on lake trout reproduction by raising the water level duringreservoir impoundment Second, it reviews possible effects of various water level man-agement practices on the quality of reproduction sites and on egg and fry mortality

Characteristics of reproduction sites

The lake trout is a fish that reproduces almost exclusively in lakes, although reproductiveactivities have been documented in the rivers of Ontario and Québec (Loftus, 1958; Vincentand De Serres, 1963; Séguin and Roussell, 1970) The spawning period varies with thelatitude For the Precambrian Shield, this usually means that it takes place in October.Incubation extends over a period of 4 to 5 months, depending on water temperature(Martin and Olver, 1980) The great majority of eggs hatch around March 1, but hatchingmay occur as early as the end of January or as late as the beginning of April (Chabot andArchambault, 1981; Pariseau, 1981) Lake trout embryos can move extensively within andabove the substrate immediately after hatching (Baird and Krueger, 2000) After their yolksac is fully resorbed, approximately 2 months after hatching, it appears that the youngfish immediately migrate to deep waters (Martin and Olver, 1980)

To date, it has been impossible to determine whether lake trout return to their place to reproduce However, it is clear that specific sites are used by spawning stock,sometimes year after year (Gunn, 1995) Lake trout spawning grounds in Québec’s Mau-ricie region are generally located near the shore at a depth of less than 2 m The substrate

birth-is composed mainly (>90%) of cobbles and boulders (40 to 500 mm) without sand or silt,and with numerous and deep interstices The sites are subject to strong wave action, have

a relatively steep slope (>20%), and are located near a deep zone (>30 m) (Benoît et al.,

1999) These characteristics are similar to those of many lake trout spawning grounds

observed in Ontario (MacLean et al., 1990).

For the lake trout, spawning in very deep zones appears more of an exception thanthe rule because in Ontario lakes (excluding the Great Lakes), 98% of the spawning

grounds are less than 4.5 m deep The average overall depth is 1.4 m (MacLean et al.,

1990) The reason for this is that the occurrence of coarse substrate decrease with depth(Chabot and Archambault, 1981) As a result, most of the substrate at depths of greaterthan 3 to 5 m generally consists of fine particles (sand and silt) However, some spawninggrounds at depths below 5 m have been reported (Machniak, 1975)

Spawning depth seems to result from a compromise between the forces needed tokeep the substrate clear of fine particles and the forces that can cause egg disturbance ormortality A significant relationship between spawning depth and lake size was obtained

Effects of raised water level

Few studies have focused on the lake trout’s reproductive behavior following reservoircreation, but raising the water level appears to have less of an impact than lowering thelevel in winter

At the Minnewanka reservoir, Cuerrier (1954) observed spawning as usual at tional sites that had maintained a suitable spawning substrate despite the raising of thewater level by 23 m However, spawning took place over a wider vertical range thanbefore construction of the dam because the lake trout were able to use new rocky areasdown to a depth of about 9.5 m On the contrary, when the water level in Bark Lake,Ontario was raised 11 m in the late 1930s, the lake trout stopped using traditional sites(Wilton, 1985) Inventories carried out from 1966 to 1972 revealed that the lake troutspawned at depths of less than 3 m during that period.

tradi-Whether it is in terms of substrate or depth, the characteristics of reproduction sitesfound in reservoirs (Lacasse and Gilbert, 1992; Bélanger and Gendron, 1993; Martin, 1955)are similar to those of sites in natural environments (Dumont et al., 1982) When the waterlevel is raised, fine particles eventually cover traditional reproduction sites (P.G Sly,

personal communication in Evans et al., 1991), rendering them less attractive to breeders

because of reduced substrate permeability and cleanliness However, wave action can clearsubstrate, presenting characteristics favorable to reproduction and hence creating newsites in shallow areas McAughey and Gunn (1995) demonstrated that the species wasclearly capable of seeking alternative spawning sites when traditional ones weredestroyed Hence, the lake trout will readily abandon traditional reproduction sites andselect new, more favorable ones in the vicinity

Effects of water level fluctuations

Management of artificial bodies of water designed for hydroelectric production purposesusually requires that the reservoir be filled when water inputs are high and emptiedgradually during periods of heavy energy demand On the Precambrian Shield, threemajor steps are involved in dam management The first step, characterized by a ratherstable but high water level, covers the summer season (May to August included) Thesecond step, known as drainage, sometimes begins after a slight fall high water stage andlasts until spring thaw This step makes it possible to supply power plants throughout thewinter when energy demand is heavy and water inputs are low It rests nearly entirely

on water loads accumulated during the previous spring thaw or drawn from groundwatertables Winter drawdowns (October to April) are typical of this step In the third and last

shows typical water level fluctuations in a reservoir used for hydroelectric productionpurposes A similar management pattern is observed in reservoirs used to regulate springflooding

Figure 5.4 Relation between lake size (km2, log 10 scale) and lake trout spawning depth (m) (From

Fitzsimons, J.D., 1994, Canadian Technical Report of Fisheries Aquatic Sciences, No 1962.)

0,0 2,0 4,0

A survey of various Ministry of Natural Resources districts of Ontario (Lewis et al.,1990) determined the pattern and importance of drawdowns in 85 lake trout lakes A largenumber (50) of these lakes had drawdowns occurring after lake trout spawning Draw-downs occurred before lake spawning on 39 lakes and during spawning on 14 lakes (somelakes had drawdowns occurring at more than one time) The modal drawdown depth was0.5 to 1.0 m for all three periods (before, during, and after spawning), with drawdowndepths varying from 0.5 m to more than 9.5 m In Québec, the modal drawdown depthwas 0.1 to 0.5 m, with depths varying from 0.1 to 11.2 m (Gendron and Bélanger, 1993).The impacts on lake trout of fluctuations in water level are many and have beenextensively documented (Machniak, 1975; Evans et al., 1991; Gendron and Bélanger, 1993).Because lake trout spawn in fall, generally in shallow littoral zones, and the fry leave thespawning grounds towards late spring, the species is vulnerable to winter drawdowns

Water management effects on lake trout populations seem to vary according to voir management parameters, especially the extent of winter drawdowns Some reservoirswhere winter drawdowns are rather extensive nonetheless manage to support healthylake trout populations For example, the Kempt (Benoît et al., 1997) and Mitis (Gendronand Bélanger, 1992) reservoirs in Québec are characterized by significant winter draw-downs (2.0 and 3.0 m) but harbor relatively abundant and stable lake trout populations

reser-A large proportion of spawning grounds used by lake trout in those lakes are deep enough

to be unimpacted by the effect of drawdown

Fluctuations in water level create shoreline erosion and stir up significant quantities

of sediment Spatial distribution of sediment particles is strongly influenced by downs (Baxter and Glaude, 1980), and, depending on shoreline instability and seasonalvariations in water use, a decline in the quality of spawning grounds can occur as a result.Sand and silt deposits on these grounds can seriously disrupt lake trout reproduction, asthey often result in highly reduced interstitial oxygen concentrations

draw-Among the factors likely to modify the lake trout’s reproductive behavior is thedisappearance of reproduction sites as a result of changes in their chemistry and physicalattributes caused by the adverse effects of water level fluctuations The species may have

Figure 5.5 Typical water level fluctuations of reservoir in the Precambrian Shield used for electric production purposes; critical period for lake trout reproduction is indicated

Larvae Emerging Period