In Canada, the monitoring of environmental levels of DGS has for the most part been concentrated on fresh water environments. Some data have been collected for the marine environment, but they are very limited. The data which exist will be described in the following sections. In many aquatic and marine environments where DGS has been shown to cause GBT in fish and invertebrates there are no data for comparable Canadian environments. Where appropriate, data from the United States will be used to provide an indication of the DGS levels which could occur in Canada.

5.1 Fresh Water Environments

5.1.1 Rivers and Lakes

Most data on DGS levels in fresh water environments come from British Columbia. Beginning in 1968, the provincial Ministry of Environment conducted an extensive program of monitoring dissolved gas levels in rivers and lakes throughout the province (Clark 1977). A large part of these data were collected from the Columbia River in the southern part of the province. At the time, there were concerns about dissolved gas levels in the Columbia River resulting from the dams which had been built as a part of the Columbia River Treaty with the United States. As pointed out earlier, delta P levels approaching 350 mm Hg had been observed. Table A1 of Appendix A gives a detailed description of the results of the province-wide monitoring program. Table 2 summarizes the data from Table A1 for specific lakes and river systems. Table 3 further summarizes the data in terms of those river systems having dissolved gas levels which, at the time of measurement, were considered harmful to fish.

In Tables 2 and 3 it is apparent that dissolved gas levels in the Columbia River below the Hugh Keenleyside Dam are very high, with delta P levels above 200 mm Hg occurring many times throughout the year. Data from more recent studies by BC Hydro (Maxwell 1985) are shown in Table 4. Comparing these data with those of Table A1, it is apparent that dissolved gas levels in the Columbia River have not changed significantly since they were first measured in 1968. An important feature of the high levels of DGS in the Columbia River in Canada is that there have been only minor recorded occurrences of GBT in fish (Hildebrand 1991). This contrasts sharply with the Columbia, Snake, and Bighorn Rivers in the United States where massive fish kills have been encountered. The differences between these systems will be considered further in Section 8.1.1.3.

In addition to DGS associated with dams on the Columbia River, it is apparent in Table A1 that very high levels of DGS exist in the Kootenay River below the Upper and Lower Bonnington Dams, below the Slocan Dam, and below the Brilliant Dam. At many times, the levels in the Kootenay River are comparable to those in the Columbia River. The same is true below the Waneta Dam on the Pend d'Oreille River. Occasionally high delta P levels exist in the Peace River below the W.A.C. Bennett Dam, in the Stave River below the Ruskin Dam, below the Aberfeldie Dam on the Bull River, and below the Duncan Dam on the Duncan River (Table A1).

It is interesting to note in Table A1 of Appendix A that although dams are clearly a major source of DGS in some British Columbia rivers, there are numerous rivers (without dams) and lakes where delta P levels range from 40 to above 76 mm Hg (sea level TGP% about 105% to 109%). For example, Clark (1977) reported delta P levels well above 76 mm Hg in the Fraser River at Lillooet, above and

Table 2: Range of TGP% Levels in Some British Columbia Rivers
(Adapted from Clark 1977)

TGP% Levels

Table 3: TGP% Levels in British Columbia Waters Considered to be Harmful to Fish
(Adapted from Clark 1977)

TGP% Levels

Table 4: TGP% in the Columbia River Below the Hugh Keenleyside Dam 1983-1984
(Maxwell 1985)

Note: A + indicates that gates or ports are open; however, the number of each is unknown

below Hells Gate, at Yale, at Hunter Creek, and at Agassiz. Clark (1977) also found that both Kootenay Lake and the Kootenay River frequently have delta P levels above 76 mm Hg. It is not clear in many cases why these high levels exist, although solar heating and primary production may be factors. In the case of the Fraser River, the high levels of DGS may be related to the extreme turbulence of the river in the Fraser Canyon where deep plunge pools and large rapids are a common feature. It should be noted that although relatively high levels of DGS appear to be a natural feature of many of the province's water bodies, the majority of the province's natural waters are seldom supersaturated.

Beginning in the late 1970's, the federal Department of Fisheries and Oceans, as part of the Salmonid Enhancement Program, began measuring dissolved gas levels in streams and lakes which were being used or considered for use as salmon hatchery water sources (MacKinlay 1984, Miller et al. 1987). Ground water sources were also being developed for hatcheries and these were also examined for DGS. Table B1 of Appendix B summarizes some of the results of these surveys. As with the measurement program of the provincial government (Clark 1977), it was found that DGS is a natural feature of many provincial waters, with delta P levels commonly of 50 to 80 mm Hg. Also evident in Table B1 are the elevated levels of DGS associated with wells which are used as hatchery water sources. Delta P levels of 30 to 50 mm Hg are a characteristic of many well sources with the Chahalis C1 well having delta P levels approaching 70 mm Hg.

High levels of DGS have been reported in other parts of Canada as well. As described earlier, GBT was identified as the cause of large mortalities in Atlantic salmon and eels below the hydroelectric generating dam on the Mactaquac River of New Brunswick (MacDonald and Hyatt 1973, Penney 1987). Subsequent studies found that dissolved gas levels varied greatly depending on power generation levels. Although delta P levels were not reported directly, dissolved oxygen levels ranged from slightly above saturation to slightly below saturation. Dissolved nitrogen levels were found to range from saturation values to 127% of saturation values.

In Manitoba, it was found that during the spring, high levels of DGS (resulting from solute freeze out) in frozen shallow lakes produced significant levels of mortality in rainbow trout which had been stocked through the ice (Lark et al. 1979, Mathias and Barica 1985). Total dissolved gas levels were reported as being 1.7 times the atmospheric saturation value.

5.1.2 Water Falls

Surprisingly, there is very little in the literature which describes the production of DGS by water falls or the signs of GBT in fish exposed to this form of DGS. In British Columbia, field research related to the Kemano completion project on the Nechako River has provided an example of this effect. Alderdice and Jensen (1985a) and Rowland and Jensen (1988) have shown that Cheslatta Falls, at the head of the Nechako River, produces significant increases in river DGS. Depending on flow-rate, delta P levels below the falls range from about 13 to 112 mm Hg. Some of the delta P is the result of background levels which exist above the falls. However, up to about 50 mm Hg is added by the falls (Alderdice and Jensen 1985a).

5.1.3 Solar Heating

In Canada, all of the reported occurrences of DGS resulting from solar heating have taken place in British Columbia. These incidents have involved both lakes and rivers in the province. At various times of the year, the Puntledge River on Vancouver Island is known to be supersaturated with dissolved gases as a result of rapid heating of the lake from which the river flows (Wright and McLean 1985, Table B1 of Appendix B). Heating of the river also contributes to the elevated dissolved gas levels. This has led to low, but lethal levels of DGS in a salmon hatchery on the river.

Harvey and Smith (1961) reported that Cultus Lake, a source of water for a provincial trout hatchery, was supersaturated with dissolved gases. The high levels of DGS were identified as the source of GBT outbreaks in the hatchery. At a depth of ten metres, dissolved oxygen was at 126% and nitrogen was at 116% of their respective air saturation values.

Lower Arrow Lake, which forms a reservoir behind the Hugh Keenleyside Dam, exhibits elevated levels of DGS during the spring, summer, and fall months (Table A1, Appendix A). Part of the DGS observed in the lake may be related to solar heating; however, detailed studies have not been conducted to confirm this. DGS has also been observed in several other large lakes in British Columbia, (Table A1 of Appendix A). Although the source of the DGS was not identified, solar heating was likely involved, since dams and water falls were not a contributing factor. In some cases, photosynthesis from primary production may have also contributed to the elevated delta P levels.

Numerous fish kills have occurred in British Columbia lakes with no apparent cause having been identified (Shepherd 1993 - personal communication). These have happened at times when solar heating, perhaps combined with dissolved oxygen supersaturation caused by primary production, could have caused high levels of DGS (Shepherd 1991). However, at the time of the fish kills, water dissolved gas levels were not monitored. As described earlier, a characteristic of DGS resulting from solar heating and primary production is the widely varying levels of dissolved gas tension between daylight hours and darkness (White et al. 1991). Thus, the delta P levels which can cause fish mortalities may be easily missed if measurements are done on a periodic spot-check basis only.

An example of a massive fresh water fish kill caused by GBT which was directly associated with solar heating and primary production occurred in 1940 in Lake Waubesa, Wisconsin (Woodbury 1941). Although the full extent of the mortalities was not determined, several thousand fish were involved in a limited observation. Affected fish had clear evidence of sub-dermal emphysema and massive bubble formation in gill filaments with significant damage to the gill structure. Most fish species of the lake were involved; however, only the adult fish of each species were affected. Dissolved oxygen levels were recorded at over 300% of equilibrium with the source identified as heating and photosynthesis involving the algae Chlamydomonas. Although delta P levels were not measured, the assumption that dissolved nitrogen levels were at atmospheric saturation values would lead to an estimated delta P of over 300 mm Hg. With the source of the DGS involving solar heating as a component, it is likely that nitrogen was supersaturated as well and that delta P levels were much greater than 300 mm Hg. Delta P levels of 300 mm Hg can kill fish in less than three hours (Section 6.1.2). Woodbury (1941) also reported that mortalities of this nature were common each spring in the lake, but not at the levels observed in 1940.

5.1.4 Industrial or Power Generation Cooling Water Effluents

Although there have been no instances of GBT in Canadian waters as a result of cooling water discharges, there have been several cases documented in the United States. At Cape Cod Bay, Massachusetts in 1973 an estimated 43 000 Atlantic menhaden (Brevoortia tyrannus) died from GBT in the heated discharge canal of the Pilgrim Nuclear Station Unit 1 of the Boston Edison Company installation (Marcello and Fairbanks 1976, Bridges and Anderson 1984). Although dissolved gas levels were not reported, there was clear evidence of GBT in the fish examined.

The Duke Power Company Marshall Steam Station (coal fired) on Lake Norman in North Carolina was identified as causing major losses of fish from GBT produced by cooling water discharges (DeMont and Miller 1971, Adair and Hains 1974, Miller 1974). DeMont and Miller (1971) reported that thirteen species of fish in Lake Norman exhibited severe signs of GBT during the winter of 1970-1971. Although delta P levels were not reported, dissolved nitrogen was up to 144% of atmospheric values.

5.2 Marine Environments

There have been only a few measurements of DGS in Canadian marine environments. Again, these come from British Columbia and are summarized in Tables A1 and B1. The measurements, from both the BC Ministry of Environment (Clark 1977) and the federal Department of Fisheries and Oceans (MacKinlay 1984, Miller et al. 1987), indicated elevated dissolved gas tensions, but delta P values were not greater than 76 mm Hg. Since these data involved so few measurements, it is unlikely that they represent Canadian marine conditions as a whole.

Very high levels of DGS have been recorded in other marine environments. In January, 1959, approximately 300 seatrout (Cynoscion nebulosus) were found dead in Galveston Bay, Texas (Renfro 1963). Other affected fish included largescale menhaden (Brevoortia patronus), bay anchovies (Anchoa mitchilli), Atlantic croakers (Micropogon undulatus), eels (Myrophis punctatus), and longnose gar (Lepisosteus osseus). GBT was easily identified as the cause of death in these animals. Extensive blistering of the skin and bubble formation in the cardiovascular system was common in most animals. Delta P levels in the bay were measured at 200 to 250 mm Hg. Although fish kills have been reported for British Columbia marine environments (Taylor 1993, Taylor and Haigh 1993), there are no reports in which GBT has been investigated as a possible cause.