FISHES

J.D. McPhail

PRESENT STATE OF THE FISH FAUNA - NATURAL IMPACTS

With the exception of commercially exploited species like Pacific salmon, and recreational species like rainbow trout, the fish fauna of most of the Canadian Montane Cordilleran Ecozone is still in a natural state. A natural state, however, does not imply stasis and geographic distributions, especially of temperature sensitive species, are dynamic and expand or contract as climate changes. The chiselmouth, Acrocheilus alutaceus, is a warmwater fish with a distribution in the MCE that suggests a temperature sensitive species. The northern edge of its range consists of a few widely scattered, isolated populations and these remnant populations imply a wider, once continuous, distribution. A plausible explanation for these scattered northern populations is range expansion during the postglacial warm "hypsithermal" period followed by range contraction as the climate cooled. Coldwater species display similar patterns of scattered, isolated populations at the contracting southern margins of their ranges. An example is the bull trout, Salvelinus confluentus. The southern margin of this species' distribution is characterized by scattered, isolated populations. Presumably, the bull trout's range expanded south during the last glaciation and is now contracting to the north. Interestingly, in the MCE there is a curious lacuna in the distribution of the bull trout --- although widely distributed in the Columbia system, this species is absent from the Kettle, Okanagan and Similkameen drainages. Summer temperatures in these waters are among the warmest in the MCE and the absence of bull trout in these rivers probably reflects the species' preference for cool waters.

Although temperature clearly influences the distribution of fishes, the impacts of temperature are not necessarily direct. For example, there is evidence that temperature influences the balance of competitive interactions between some species (e.g., bull trout and rainbow trout; Parkinson and Haas 1996), and it is this competitive balance that determines the sites where these species can coexist and sites where one of the species replaces the other.

The expansion of ranges through natural climate changes sometimes also brings species into contact that have had no previous history of coexistence, and this can result in extensive hybridization. An example in the MCE involves the slimy sculpin and the shorthead sculpin in the Flathead River (Hughes and Peden 1984). This river is the only place in the extensive geographic ranges of these species where they co- occur. In the mainstem Flathead River, immediately north of the U.S. border, and in the lower reaches of some tributaries (e.g., Howell and Sage creeks), the shorthead sculpin has replaced the slimy sculpin. In the upper parts of the Flathead, and in the upper reaches of these same tributary streams, only the slimy sculpin is present. This distribution pattern suggests that contact between these species is relatively recent (i.e., late postglacial) and that the interaction between the species is still in a state of flux. Again, the interaction appears to be mediated through temperature --- the shorthead sculpin dominating in warmer areas and the slimy sculpin in cooler areas. Interestingly, there are contact zones in both tributaries and the main river, and evidence of hybridization at each of the contact sites.

Natural landscape changes can also bring species into contact and generate episodes of hybridization. The MCE is a mountain dominated region and, although connections between drainage systems are rare, they most often occur through stream piracy. These minor drainage changes usually involve headwaters and often occur at considerable altitude in mountain passes. The fauna of such streams is limited and, consequently, only a few species disperse by this means. An example in the MCE is the movement of kokanee (landlocked sockeye salmon) across two drainage divides: 1) from the McGregor River (a Fraser tributary) into the upper Parsnip River (a Peace tributary), and 2) from the Sustat system (a Skeena tributary) into the Thutade system (another Peach tributary). Since the kokanee is a lacustrine species, these transfers probably involved small lakes. Another species, the Dolly Varden (Salvelinus malma) also appears to have used the Sustat-Thutade connection. Interestingly, neither the kokanee nor the Dolly Varden have expanded their ranges in these new drainages, and this suggests that the connections were relatively recent. In the case of the Dolly Varden, there is also hybridization (with the bull trout) in the area of recent contact (Baxter et al 1996).

ANTHROPOGENIC IMPACTS

Although most of the waters in the Canadian Montane Cordilleran Ecozone are relatively unsullied, human activities are beginning to alter the natural environment at an alarming and accelerating pace. The major agents of change are the resource extraction industries (e.g., forestry, mining, and fisheries), agriculture, transportation (railways and roads), impoundments for hydroelectric generation and water storage, the introduction of exotic species, and runaway urbanization.

FORESTRY

The impacts of forestry on aquatic environments are well documented (Koski 1992; Murphy 1995) and in British Columbia (i.e., most of the MCE) almost half the provincial forest has been logged (Slaney and Martin 1997). The best timber is at lower elevations and here the cut is estimated to be much greater than 50 percent. Indeed, some of the clear-cuts in the MCE are among the largest on earth (e.g., the Bowron clear-cut). Such massive removals of ground cover and, until recently, the practice of logging stream-banks, dramatically alter aquatic habitats. Logging changes the amount of solar radiation reaching streams, water temperatures, forest canopy, streambank vegetation, streambank stability, suspended solids, fine woody debris, large woody debris, channel morphology, substrate sediments, streambed stability, nutrient supply, and stream flows (Slaney and Martin 1997). Most of these impacts are negative, although sometimes fish growth rates increase after logging; however, any such gains are off-set by increased mortality rates (Murphy et al 1986). Some effects of logging are transient and measured in decades, but other effects are measured in centuries. For example, in small streams large woody debris provides essential winter and summer habitat for juvenile salmonids. When streams are logged to the banks no new large woody debris enters the system, and the existing debris is slowly removed by floods and decay. The result is a loss of essential habitat that may persist for well over a century before windfalls from new growth in the riparian zone reach a size sufficient to again contribute large debris to the stream (Koski 1992). Besides the direct effects of removing forest cover, logging has indirect effects. For example, road building associated with logging increases access to previously inaccessible areas, and increased access is a major threat to species susceptible to angling (e.g., bull trout and Arctic grayling). Road building also increases erosion which increases siltation, the frequency of landslides, and in some cases the widening of flood plains (Hogan and Ward 1997). In the MCE in the past, log drives and the splash dams built to facilitate log drives, have contributed to the loss of major salmon stocks in the upper Fraser system (Roos 1991). Thus the cumulative effects of logging on aquatic environments are unequivocally negative and, although logging has not caused the extinction of any fish species in the MCE, it has contributed to a loss of biodiversity at the population level (Roos 1991).

FISHERIES

Commercial fisheries, and especially attempts to manage and "improve" these fisheries, have had major impacts on the diversity of Pacific salmon stocks and trout populations in the MCE. Traditionally, the management of Pacific salmon has concentrated on a relatively small number of populations that support major fisheries and ignored the myriad of minor populations that make up the bulk of the genetic diversity in these fishes (Hyatt 1996). With the advent of the Salmonid Enhancement Program, this benign neglect of small populations became more serious. This program was designed to "enhance" salmonid populations through the application of various "techno-fixes" such as modern hatcheries, artificial spawning channels, lake fertilizations, and barrier removals. Although some of these projects have been successful (i.e., produced increases in some runs), more is not necessarily better and there is increasing concern about the impacts of enhanced stocks on wild fish (Hilborn and Winton 1993). Artificially inflated stocks do not exist in a vacuum --- their migration routes and run-times overlap those of smaller populations and other valuable species (e.g., steelhead), and gear aimed at enhanced stocks does not discriminate between the target stock and other fishes. The result is severe over-fishing on small wild salmon stocks and on other species. Over time, the incremental effects of harvesting enhanced stocks will almost certainly reduce fish diversity in the west-flowing rivers of the MCE.

Commercial fisheries are not the only fisheries with negative impacts on biodiversity. Recreational fisheries and, again, attempts to manage and "improve" these fisheries also reduce biodiversity. Extensive stocking programs have spread domesticated strains of rainbow trout throughout the MCE. Often these domesticated trout were stocked into "barren" lakes, but they were also stocked into lakes with wild populations. Stocking domesticated trout into lakes containing wild trout has the potential to reduce the diversity of wild stocks through genetic swamping (Philipp et al 1993). In the case of "barren lakes", stocking trout probably has altered the invertebrate communities in ways at which we can now only guess. All this was done to "improve" fishing in an area that was already world renowned for the quality and quantity of its recreational fisheries. Another management technique for improving recreational fisheries is the introduction of exotic species. An example in the MCE is the brook trout (Salvelinus fontinalis). Brook trout were once widely stocked into waters thought to be marginal for rainbow trout while a related native species, bull trout, was despised by both anglers and fisheries managers. Now, bull trout are the fish species of most concern in the MCE, and there is evidence that brook trout not only out-compete bull trout in some areas but also hybridize with bull trout and eventually replace them (Markle 1992). Thus, ironically, a species once enthusiastically introduced throughout the MCE is now viewed as a major threat to a newly canonized native species. There is an obvious moral to this story but it seems to be lost on many recreational fisheries managers. Probably because occasionally an exotic species creates a new and popular recreational fishery (e.g., the introduction of walleye, Stizostedion vitreum, into the Columbia system). Such successes, however, are bought at a price and walleye are now estimated to account for up to one third of the annual loss to predation of salmon smolts in the Columbia system (McMahon and Bennett 1996). Translocations (the introduction of species within their natural range to localities where they did not originally occur) can also cause a loss of diversity. For example, lake trout (Salvelinus namaycush) were introduced into a number of lakes in the Canadian Cordilleran National Parks. Originally, some of these lakes contained bull trout, but this native species is now gone from the lakes where lake trout were introduced (Donald and Alger 1993). Perhaps, however, the most insidious recreational fisheries management technique of the past was lake "rehabilitation". This is a euphemism for the use of rotenone or toxaphene to poison an entire lake. In BC alone in access of 100 lakes in the MCE were "rehabilitated" to remove the native fishes and replace them with a monoculture of domestic trout. Sometimes, the managers knew what was in the lake before it was "rehabilitated" and occasionally a sample of the dead fish was collected from the poisoned lake. In one case (Dragon Lake near Quesnel), belatedly it was discovered that the lake contained a sympatric pair of lake whitefish. Similar species pairs occur at a few widely scattered sites across northern North America and are used extensively in research on the ecology of rapid speciation (Lindsey 1981, Bernatchez and Dodson 1990). The Dragon Lake whitefish were the only example of this phenomenon known in the MCE (Lindsey 1981).

MINING

Relative to the total area of the MCE only a small amount of land has been disturbed by mining. The earliest mining activities in the region were placer and shaft mines. The impacts of small-scale placer mines were mainly local and transitory, although some large-scale placer mines (e.g., hydraulic mining in the Quesnel area of British Columbia) produced long-term impacts on the landscape. Recent mining activity in the ecozone is primarily large-scale surface mining. Like their predecessors, the impacts of surface mines tend to be local and, if properly reclaimed, transitory. They have, however, the potential to produce long-term "off-site" impacts that negatively effect biodiversity. Large-scale surface mining usually entails exploration and site preparation, as well as mining and processing ore or coal (Stearnes and Gasper 1996). In their earliest stages exploration impacts usually are minor, but once site preparation begins the impacts can be important. In the MCE, site preparation often involves extensive road building which leads to many of the problems associated with forestry roads (e.g., increased access, siltation, and erosion) and, in remote areas, the construction of entire new towns. These new towns result in all of the usual impacts of urbanization on aquatic ecosystems --- pollution, stream diversions, channelization for flood control, erosion, and gravel removal. The actual mining can involve extensive stream diversion, huge open pits, and even mountain removal. If the mine is a heavy metal mine, the rock is usually crushed on site. This produces an enormous problem in waste rock management that often entails large piles of tailings and the construction of settling ponds. The purpose of the settling ponds is to prevent toxic fines from reaching streams, but eventually retaining dikes fail and when this happens large amounts of toxic material are released into adjacent streams. Even large ore mines are transitory, and when such mines close the problems of reclaiming heavy-metal contaminated sites is difficult and usually result in a long-term contaminated seepage problems. problems. If, as in the Rocky Mountains, the surface mines are coal mines, the waste management problems are compounded by acid mine drainage. In summary, although the past the effects of mining in the MCE have been relatively minor and local, modern surface mines have the potential for impacts that can transcend local areas and cause widespread reductions in aquatic diversity.

AGRICULTURE

The nature of the soils and landscape in most of the MCE mitigates against large-scale agricultural activities; however, two exceptions are livestock grazing and fruit growing (including viniculture). These activities can have negative impacts on aquatic ecosystems. The primary effect of livestock grazing on aquatic ecosystems is through damage to riparian zones (Armour et al 1991). Livestock grazing changes the extent and kinds of riparian vegetation, breaks down banks, and causes erosion, channel widening, and siltation (Wohl and Carline 1996). These changes increase summer stream temperatures, reduce cover, degrade water quality, and change both flow regimes and stream morphology. Although livestock grazing is one of the principal factors contributing to the decline of native trout in southwestern North America (Armour et al 1991), its impacts in the MCE are minimal. This is because livestock densities on the Interior Plateau of British Columbia and the eastern foothills of the Rocky Mountains are relatively low and in the summer dispersed over large areas of summer range. Typically, although bottom lands in these areas are used extensively for hay production, they are not heavily grazed. Commercial fruit growing (including vineyards) in the MCE is confined to the warmer, arid parts of the Columbia Basin (e.g., the Okanagan and Similkameen regions). The main impact of fruit growing on aquatic ecosystems is the demand for irrigation water. The construction of storage dams on lake outlets and the diversion of water for irrigation started early in the 20th century and continues to this day. Koshinsky and Andres (1972) and Northcote (1996) document the history and extent of such diversions in the Okanagan system. The primary impacts of irrigation are reduced summer flows and increased summer temperatures in rivers and streams. Both of these changes are detrimental to native species, and often tip the competitive balance within fish assemblages towards exotic species (e.g., carp, Cyprinus carpio, and pumpkinseed, Lepomis gibbosus). Thus, in the southern Okanagan Valley, many of the low gradient streams and small lakes are now dominated by exotic species --- a clear signal that the integrity of these environments is breaking down (Karr 1981).

TRANSPORTATION

In mountain dominated landscapes, major transportation corridors are concentrated along valley floors and, inevitably, railways and roads share the valley floors with rivers. Since fish diversity is highest in valley-bottom waters, the potential impact of transportation corridors on biodiversity is disproportionate to the land area occupied by roads and railways. Historically, the single greatest ecological disaster to befall the Fraser River was the 1913 Hells Gate landslide in the Fraser Canyon. This slide occurred during railroad construction and either decimated, or destroyed, many of the upper Fraser salmon populations (Roos 1991). Indeed, one of the major effects of transportation corridors in narrow valleys is their impact on slope stability and, consequently, on the probability of landslides. In addition, to prevent wash-outs, rivers alongside transportation corridors are often constrained or deflected to reduce erosion at specific sites. This alters the natural development of the river course and can cause major downstream changes in river morphology. Generally, the system becomes unstable --- old side-channels (important fish rearing areas) are abandoned, new side-channels are established, bed loads shift and fill pools, new pools are scoured, spawning gravels are transported to new sites, and large woody debris is moved downstream. The cumulative effects of such instability on fish diversity are negative and may take centuries to re-stabilize.

URBANIZATION

With the exception of a few foci of urbanization (e.g., the south Okanagan, Kamloops, and Prince George) human population densities in the MCE are low. Consequently, relative to the entire ecozone, the impacts of urbanization on aquatic environments are generally small and local. However, the scattered urban centers in the MCE are growing rapidly and, inevitably, this growth produces negative impacts on aquatic environments. Typically, in urban areas water quality (in both surface and ground waters) diminishes, sedimentation increases, run-off patterns change because of increases in impervious surfaces (e.g., roads and parking lots), water temperatures rise, and waste management (domestic and industrial) becomes a problem (Hall and Schreier 1996). Further, as population increases developers, abetted by compliant local authorities, build housing or industrial parks on inappropriate sites (e.g., swamps and floodplains). Inevitably, this leads to drainage and flood control measures (ditching, diking, and channelizing flowing waters) which destroy fish habitat and reduce diversity. Nowhere in the MCE are the negative effects of urbanization on aquatic ecosystems more obvious than in the south Okanagan (Northcote 1996). Here, aquatic environments have been altered to the point that in the summer many streams no longer contain sufficient water to support fish.

DAMS & DIVERSIONS

Throughout the MCE major rivers have been impounded to store water and generate electricity and, in at least one case, a major river (the South Nechako) has been impounded and diverted into another drainage system. In addition, hundreds of smaller streams have been dammed to provide irrigation water. Again, the effects of most of these impoundments are local and the majority of rivers in the MCE still flow free. With over 50 major dams on its mainstem and major tributaries, the Columbia River has the dubious honour of being the most heavily dammed drainage system in North America. These dams are directly responsible for the extinction of a number of salmon and steelhead stocks in southcentral MCE (Fulton 1970; Scholz et al 1985). Fortunately, the mainstem Fraser and most of the east flowing rivers in the MCE have escaped the Columbia's fate. An exception, however, is the Peace River. This magnificent river is dammed where it breaks out of the Rocky Mountains and the impoundment (Williston Reservoir) is the largest man-made feature in the MCE.

The direct effects of major dams on fish are obvious --- dams impede or totally prevent fish passage, they alter natural flow regimes (to which the life histories of local populations often are precisely adapted), change water quality (temperature, turbidity, and dissolved gas pressures), and in many cases convert fluvial environments into lacustrine environments. Unlike small irrigation dams, the effects of major dams are not local and the impacts can be complex and extend for hundreds of kilometers downstream (Hartman 1996).

The indirect effects of dams on fishes also can be complex. For example, a feature of north temperate freshwater fish is their propensity to hybridize in disturbed environments (Hubbs 1955). This is because reproductive isolation in closely related fish species typically is maintained through differences in spawning times or habitats. Impoundments disrupt these factors and, throughout the MCE, interspecific hybridization is strongly associated with impoundments (Nelson 1965, 1974).

INTRODUCTIONS

The total fish fauna of the Canadian Montane Cordillera Ecozone is small (59 species) and it is remarkable that in such a relatively undisturbed area almost a quarter (16) of the species are exotic (i.e., species introduced from outside the MCE). Exotic species were introduced into the waters of the MCE for a variety of reasons, but the commonest reason was to "improve" recreational fishing opportunities. Thus, warmwater species like largemouth and smallmouth bass (Micropterus salmoides and M. dolomieu) were introduced into the Okanagan and Kootenay regions to provide angling opportunities in waters deemed too warm to sustain trout. Other recreational species like brown trout, Salmo trutta, and brook trout were introduced into waters thought to be marginal for rainbow trout, and also to provide "different" angling opportunities. Still other species like carp and lake whitefish were introduced, or transposed in the case of whitefish, with the intention of creating commercial enterprises. The reasons for the introduction of other species (e.g., tench, Tinca tinca; black bullhead, Ameiurus melas; pumpkinseed; black crappie, Pomoxis nigromaculatus, and yellow perch Perca flavescens) are less clear. Probably some dispersed north from introductions in adjacent states (e.g., walleye) and others (e.g., goldfish, Carassius auratus; western mosquitofish, Gambusia affinis; sailfin molly, Poecilia latippina, and African jewelfish, Hemichromis bimaculatus) are aquarium releases. With one or two exceptions (see the fisheries section) the impacts of exotics on native species in the MCE has been minimal and, generally, native species are competitively superior (i.e., better adapted to local conditions) to exotic species in undisturbed habitats. Thus, the dominance of introduced species in waters like the Okanagan River reflects the extent of the environmental damage to this system. However, not all exotic species are relatively harmless and the introduction of a "shrimp", Mysis relicta, into large lakes in the Okanagan and Kootenay regions has had a major --- and continuing --- negative impact on the fish assemblages in these lakes (Northcote 1991).