by/par
C.P. Ruggles
Fisheries and Marine Service
Department of the Environment
P.O. Box 550, Halifax, Nova Scotia
Canada
ABSTRACT
A variety of devices used to enumerate populations of anadromous fishes in Eastern Canada are described. Population changes in Saint John River Atlantic salmon (Salmo salar), American shad (Alosa sapidissima), and the alewife (Alosa pseudoharengus) as measured at fish passage facilities at three hydro-electric dams are discussed. Changes in run timing, age composition and age at first maturity of Miramichi River Atlantic salmon as measured by an estuarine sampling trap are discussed in terms of their effects on natural reproduction. The author warns that extreme care must be exercised when entire runs of fish are intercepted at dams or fish fences. Population changes may be a direct effect of handling and sampling. Some units of constant fishing effort, perhaps in the form of modified commercial fishing gear can often supply good indices of population change when sound sampling theory is applied.
RESUME
L'auteur décrit une variété de dispositifs utilisés pour dénombrer les populations de poissons anadromes de l'Est du Canada. Il discute des changements parmi les populations du saumon de l'Atlantique (Salmo salar), de Alosa sapidissima et de Alosa pseudoharengus du Saint John River tels qu'ils sont mesurés aux installations de passage de poissons de trois barrages hydro-électriques. Il discute également des changements de période de migration, composition d'âges et de l'âge de la première maturité du saumon de l'Atlantique de Miramichi River, tels qu'ils sont mesurés au moyen d'une échelle à échantillonnage aux estuaires, en fonction de leurs effets sur la reproduction naturelle. L'auteur fait ressortir qu'un soin extrême doit être pris lorsque des migrations entières de poisson sont interceptées aux barrages ou aux barrières à poisson. Les modifications parmi les populations peuvent être causées directement par la manipulation et l'échantillonnage. Une certaine unité d'effort de pêche constant, peut-être sous forme d'engin de pêche commerciale modifié, fournit souvent de bons indices de changement de population lorsque l'on applique une saine théorie d'échantillonnage.
CONTENTS
1. INTRODUCTION
2. USE OF UPSTREAM FISH PASSAGE FACILITIES
2.1 Changes in Anadromous Fish Populations in the Saint John River
3. USE OF DOWNSTREAM FISH PASSAGE FACILITIES
4. USE OF SAMPLING TRAPS AND FISH WEIRS
4.1 Sampling Traps
4.2 Changes in the Salmon Populations in the Miramichi River
4.3 Fish Weirs
5. DISCUSSION
ACKNOWLEDGEMENTS
REFERENCES
A variety of devices have been used in Eastern Canada to intercept populations of anadromous species. Many of these devices were initially designed to aid fish past man-made and natural obstructions. In some instances, fish weirs and sampling traps have been constructed specifically to obtain biological information on stocks of Atlantic salmon (Salmo salar). In all cases the effectiveness of the catching gear depends on the migratory behaviour of the fish species being studied.
Anadromous species lend themselves to quantitative study because of their mass migrations to and from the sea. These migrations provide an opportunity to enumerate and otherwise sample entire populations as they move past a fixed point in a river or estuary. The ability to estimate the number of juveniles migrating to sea and the number of adults returning to spawn is the cornerstone of fishery research and management for most anadromous species.
The purpose of this paper is to describe some of the more important fish passage facilities used to monitor fish runs in Eastern Canada and to describe specific sampling gear used for assessing population characteristics of Atlantic salmon. Emphasis is placed on large rivers rather than on streams and the collection and interpretation of data are related to the solution of practical fishery management problems. In particular, recent changes in timing, age composition, and age at first maturity of Atlantic salmon, and the resultant effects on natural reproduction are discussed.
The subject material is divided into two broad areas: (i) the collection and interpretation of fish population data supplied from a variety of fish passage facilities; and (ii) the collection and interpretation of data supplied by sampling traps and fish weirs specifically designed for Atlantic salmon enumeration studies.
Several different types of device are used to enable fish to migrate upstream past dams, waterfalls, and rapids in Eastern Canada. In the Saint John River, New Brunswick, three types of fish passage facility provide an excellent opportunity to gather biological information on long-term trends in populations of native anadromous fishes. The anadromous species for which data has been gathered, include the American shad (Alosa sapidissima), the alewife (Alosa pseudoharengus), and the Atlantic salmon.
The Saint John River, 676 km long and draining an area of 55 000 km2 is one of the larger North American rivers flowing into the Atlantic Ocean (Figure 1). The river has been extensively developed for hydro-electric power and hence several fish passage problems have been created by the construction of dams on the main stem of the river and on important tributaries.
A dam was constructed in 1953 on the Tobique River, a major Atlantic salmon-producing tributary (Figure 1). The 21-m high dam was provided with fish passage facilities designed primarily to provide upstream access to adult Atlantic salmon. The reinforced concrete structure consists of a collection gallery and a pool and weir fish ladder (Figure 2). A fish trap located immediately above the exit of the fish ladder provides an opportunity for counting and sampling all fish that migrate above the dam. The collection gallery is located along the downstream face of the powerhouse, with an arm at approximately 90° extending down along the left shoreline of the tailrace. This arm leads directly into the lowermost pool of the fish ladder. The fish ladder consists of 73 pools separated by weirs providing a head differential of approximately 31 cm between each successive pool. The uppermost six pools are joined by underwater orifices, each fitted with an adjustable gate to regulate flow rates in the fish ladder.
From 1953 to 1967 a counting trap was fished at the fish ladder exit to count and otherwise sample the runs of ascending salmon. The trap, a temporary wooden structure, was installed at the beginning of each migration season and removed at the end. Fish were released from one to several times daily depending on run intensity. To check the trap, the V-shaped entrance was first closed by sliding a movable barrier into a slot at the apex of the “V”. Hinged trap doors on the timbered deck were then opened to permit observation inside. Next, the floating floor was raised close to the water surface to facilitate counting and tag detection. When it was necessary to check tag numbers or examine fish closely, they were removed by dip net and afterwards returned to the trap to ensure complete recovery. Once these operations were completed, the exit gate (a slatted wooden panel) was lifted by hand and the fish swam out into the forebay to continue their upstream migration.
The second dam where counts of Saint John River anadromous species were obtained was at the Beechwood hydro-electric development, located on the main stem of the river, approximately 40 km downstream from the Tobique River (Figure 3). The 17-m high dam, constructed in 1957, is provided with a mechanical hoist for lifting fish over the dam. Fish enter a collection gallery located over the draft tubes of the powerhouse and are led by attraction flows to a hopper located at the end of the powerhouse (Figure 4). The hopper or “skip” rests in a pool and a V-shaped entrance gate leads in to its side. It is 3.7 m deep at its deepest point, tapering to zero depth along its upstream side. When the skip is raised, fish and a relatively small quantity of water are lifted by a solid steel box floor until the skip reaches the top of the dam. At this point it tips forward dumping its contents of fish and water into the forebay. Counts of fish are made by observation in the hopper. Ease in counting is facilitated by division of the floor of the hopper into four smaller parallel compartments.
The third site where anadromous runs of fish are monitored on the Saint John River is at the Mactaquac Dam (Figure 5). The 34-m high dam was constructed in 1967 on the main stem of the river 130 km below the Beechwood Dam. At Mactaquac, all upstream migrating fish are trapped and transported by tank truck to various destinations in the river system. The fish passage facilities include a collection gallery, cantilevered from the powerhouse wall and supplied with a constant flow of water so as to provide an attraction to upstream migrating species. Entrance to the gallery is by way of six submerged gates spaced along its length (Figure 6). Salmon and other upstream migrating species move against the current in the collection gallery to a holding pool where a mechanical crowder forces them into a primary sorting facility. The sorting facility, comprised of two interconnected brail pools, is designed to exploit the jumping behaviour of Atlantic salmon to isolate them from the large numbers of other migrating species. From the brail pools, fish are lifted by hoppers and released into tank trucks for further distribution.
Since part of the fishery conservation measures required after completion of the Mactaquac hydro-electric development included a large salmon hatchery, further s orting of Atlantic salmon is required to obtain approiate broodstock. This secondary sorting facility is located at the hatchery site, 2 km below the dam. Fish captured at the Mactaquac fish collection facilities are transported by tank truck and released into the dumping pool of the secondary sorting facility (Figure 7). A mechanically operated vertical rising brail encourages fish to enter an inspection area where individual fish are retained momentarily for observation. From here, individual fish are sorted into interconnected holding pools for upriver distribution or retention for broodstock.
Despite extensive hydro-electric development and impaired water quality due to gross industrial pollution, Atlantic salmon still return in relatively large numbers to the Saint John River. Salmon counts collected from the various fish passage facilities at the hydro- electric dams, combined with commercial and sport fishing statistics, enable mean annual estimates of stock distribution to be made (Table I).
The recent decline in salmon abundance is believed to be due to several causes. Elson and Kerswill (1964) concluded that persistent forest spraying with DDT over two salmon generations seriously reduced salmon production in the most important salmon-producing tributary, the Tobique River. Following construction of Beechwood Dam in 1957, salmon angling below the dam improved and salmon abundance in the regions above Beechwood was generally high (Dominy, 1973). It was not until the completion of the Mactaquac Dam in 1968 that environmental changes in the river seriously jeopardized salmon reproduction.
Atlantic salmon management in North America has not developed the degree of sophistication that is evidenced in Pacific salmon management. In particular, the stock-recruit relationships for salmon developed by Ricker (1954) have not been applied in any systematic way to the management of Atlantic salmon. Only recently has the fishery been regulated to provide for specific spawning escapements. Counts of adult salmon collected at fish passage facilities on the Saint John River have formed an important body of data upon which recent salmon management decisions have been made. Estimates of spawning escapement provided by counts at major dams on the river have been particularly useful in forecasting the magnitude of returning runs five and six years hence. Because of the adverse effects of hydro-electric development and water pollution, a drastic reduction in the commercial fishery was required to rebuild an adequate spawning escapement.
By marking hatchery smolts and subsequent recovery at the Mactaquac fish collection facilities, the relative contribution of hatchery-produced Atlantic salmon is known. Data collected in this way reveal that the contribution of hatchery-reared smolts has risen from 1.8 percent of the total run in 1970 to 35 percent in 1973. Numbers of returning hatchery adult fish have risen from 94 in 1970 to 2 250 in 1973. More important from a management point of view has been the opportunity to separate stocks of hatchery-reared Saint John River Atlantic salmon from naturally produced salmon. Thus, selective breeding can be utilized as a hatchery management tool.
American shad have declined dramatically following construction of the Tobique, Beechwood, and Mactaquac dams. Early records indicate that a shad population existed in the Tobique River prior to dam construction. Their presence at the Tobique Dam fishway was noted for a few years but a rapid decline in abundance followed completion of the dam. At the Beechwood fish collection facility shad numbers declined from a high of 1 487 in 1960, four years after construction of the dam, to zero in 1967. At the Mactaquac Dam, shad runs declined from 38 838 in 1968 to 7 363 in 1973.
Whereas shad and salmon populations have declined in abundance since the construction of hydro-electric dams on the Saint John River, two other anadromous species, the alewife and the closely related blueback herring (Alosa aestivalis), have increased in abundance. Annual combined estimates of these species at the Mactaquac fish collection facilities are shown in Table II. Although the exact reasons for the increase in alewife and blueback herring have not been studied, both these anadromous species require relatively still water for spawning and reach their greatest abundance in river systems containing a large amount of lake area. Another significant point may be their ability to negotiate fishways more successfully than shad. Stevenson (1899) observed that: “Shad cannot or will not run up through fishways that are readily used by alewives”. It is interesting to note that the proliferation of alewives and blueback herring has caused congestion problems in the fish transport facility at the Mactaquac Dam.
Little effort has been expended in Eastern Canada to provide for the passing of fish downstream past obstructions. Ducharme (1972) describes an application of louver deflectors for guiding Atlantic salmon smolts from a power canal at a hydro-electric site at Nova Scotia. A portion of the smolt run was enumerated by means of this device.
Louvers consist of a series of vertical slats, much like those of a Venetian blind, placed at an angle of from 10–20° to the flow. Each individual slat is set at right angles to the direction of flow and spaced from 5–10 cm apart. The bulk of the water passes between the slats, and downstream migrating fish are diverted along the louver line to a bypass at the downstream end. Ducharme (1972) concludes that louvers are a practical means of guiding Atlantic salmon smolts when the fish must be extracted from a narrow, fast-flowing body of water with velocities greater than 0.6 m/s.
The Nova Scotian louver system (Figure 8) consisted of two louver lines placed in a V-arrangement at a 12° angle from the power canal wall. The louver slats were placed 5 cm apart and every sixth slat was extended to form a flow-straightening vane. The apex of the “V” consisted of a bypass 45 cm in width and extending in depth to 3.6 m. The bypass discharged its flow of approximately 1.4 m3/s via an underground conduit to a fish counting facility, which utilized a modified Wolf trap to further separate fish from excess water for individual counting and inspection.
After a four-year development period, this louver installation successfully guided about 80 percent of the downstream migrants within the river system. Estimates of egg to smolt survival, age composition, timing, size, sex composition and smolt to adult survival were obtained.
Another possibility for capturing downstream migrating salmon smolts exists in the gate wells of powerhouse intakes. The gate wells are normally used for inserting a gate to seal off the flow of water to the powerhouse during dewatering procedures. The provision of an exit from each gate well offers an attractive solution to providing fish an alternate route downstream past the turbines. Although this solution has not been tried in Eastern Canada, fish counts have been obtained from the gate wells on the Beechwood hydro-electric dam. Unfortunately, the downstream migration period occurs when large and varying amounts of water are being spilled over the crest of the dam and it has been impossible to estimate the proportion of the total run counted. This site has been used, however, to annually sample for growth and age composition.
The need to monitor changes in Atlantic salmon abundance in undammed rivers of Eastern Canada has resulted in a variety of fish sampling traps and weirs being utilized for the capture of migrating adults and young as they enter or leave specific rivers. Since sampling traps capture only a portion of the run, care must be taken in extrapolating sampling results to the entire population. Fish weirs or fences as they are known in Canada, on the other hand, are usually constructed so as to block off the entire river, enabling biologists to obtain exact counts of adult salmon moving upstream to spawn. Most weirs are designed so as to enable the downstream migrants to be counted as well.
In some of the more important salmon rivers, fish are captured for enumeration and tagging in a trap similar to that used by commercial fishermen. The size of the trap is largely dependent on river width, depth, and volume of flow. For the capture of adult salmon, the trap (Figure 9) is constructed of polypropylene net material with 9-cm mesh in the pound section, and 15-cm mesh in the leader section. The leader guides migrating fish toward the pound while allowing a considerable portion of debris to pass through. The mesh in the pound section retains all adult salmon and the fish are normally in excellent condition. By varying the size of the mesh, the trap can be converted to capture seaward migrating salmon smolts. In this case, the leader is composed of 6-cm mesh, while the pound section utilizes 3-cm mesh. Data gathered by means of these modified commercial trap nets play an important role in documenting year to year changes in timing, composition and relative abundance of individual river populations of Atlantic salmon.
These traps have proven extremely versatile. By modifying their size, the gear can be adapted to a variety of river conditions. The fact that fish are captured alive, in good condition, enables mark recapture estimates to be made on total population size. The size of the pound section of the trap has been found to be critical in terms of determining sampling mortality. High mortalities, due apparently to overcrowding in the trap, have been overcome by increasing the size of the pound. In the relatively still waters of estuaries the traps can be suspended from floats. Where the current is swifter, the traps are supported by pickets driven into the stream bottom and supported by guy wires.
When using such sampling gear, fishing must extend over the entire run duration; and the gear must be designed to sample accurately the various sizes of fish encountered. If the gear takes a selective rather than a representative sample of the run, estimates of the composition of the run can be subject to large error, as in the case of estimating the proportion of grilse and older salmon.
The use of a sampling trap located in the Miramichi River estuary, New Brunswick has enabled recent declines in Atlantic salmon, originating from this important and productive river, to be analysed. In particular, the relationship between grilse (fish returning to spawn after one year at sea), large salmon (fish returning to spawn after two or more years at sea), and natural reproduction was examined by Ruggles and Turner (1973).
The sampling trap was installed soon after ice cleared the river (about mid-May) and remained in operation until it ceased to catch fish in late October or early November. Fishing was conducted on a daily basis with one or two lifts per day, depending on tides and weather conditions. Very few days were lost each year because of weather or net repair requirements. All salmon were counted and released unless otherwise retained for tagging or further analyses. Data are available over a period of 19 years.
Figure 10 depicts the yearly catch of both large salmon and grilse, together with the estimated potential egg deposition represented by these fish. The potential egg deposition was calculated by determining the sex ratios in the catch and the average fecundity of various sized females. Recent declines in large salmon have had a serious effect on the potential egg deposition in the river. Potential egg deposition has shown a downward trend over the past decade except for some modest improvement indicated for the last two years. This improvement is due to the complete closure of the local commercial fishery which was capable of taking up to 80 percent of the large female salmon entering the estuary. The commercial fishery is composed of a gillnet fleet operating a few kilometres off the river mouth and a trapnet fishery operating in the estuary. Since 86 percent of the large salmon are female and their relatively large size provides a greater fecundity than the smaller female grilse, potential egg deposition depends heavily on large female salmon. Large runs of grilse during the period 1963 to 1967, therefore, did not reverse the decline in potential egg deposition.
From 1954 to 1962 the catch of grilse and large salmon, although fluctuating from year to year, was composed of about equal percentages of large salmon and grilse (Figure 11). Since 1963, however, the catch has been composed of an average of 87 percent grilse. This rather striking change in what would appear to be the normal components of grilse and salmon is discussed in detail in the paper by Ruggles and Turner (1973).
Data gathered by the sampling trap also provide information on seasonal timing of salmon and grilse entering the estuary. Grilse enter the Miramichi estuary about the end of May or early in June (Figure 12). A first peak of abundance occurs in the first week of July and a second, larger peak occurs during September. By the first week of November, the catch in the sampling trap has fallen to less than one grilse every two days. Grilse are not captured in any significant numbers by the driftnet fishery, but some exploitation occurs in the trapnet fishery. It is interesting to note that the peak catch in the early portion of the grilse run coincides with the two-week closure of the trapnet fishery. The catch of large salmon at the sampling trap shows a similar pattern. Large salmon arrive earlier and the first peak of abundance is undeveloped; however, the fall peak of abundance is more abrupt. By mid-November the run of large salmon has also ended.
No one knows when North American Indians first began to capture anadromous fish by means of brush weirs or fences constructed across rivers and streams. The French explorer, Nicholas Denys in 1680 described an Indian fish fence common to the rivers of Northern New Brunswick (Ganong, 1908). Biologists in Eastern Canada have used fish weirs to study the populations of Atlantic salmon in dozens of rivers and streams throughout the region. One of the earliest descriptions of such a structure was provided by White (1939). The weirs, or fences as they are called in North America, are installed in a river so as to form a complete barrier to fish movement.
Blair (1957) describes a counting fence made of netting that was successfully employed to count ascending adult salmon and descending salmon smolts in a number of Newfoundland rivers. The net described was X-shaped, with the upper arms extending upstream to each bank and thus leading smolts moving downstream to a central smolt trap. The lower arms of the “X” extended downstream and led adult salmon migrating upstream into a central adult trap. The traps and barrier netting could be easily dismantled and transported to other rivers. A modification of this fence has been described by Murray (1968).
Detailed design criteria for a variety of adult and juvenile salmon fences are provided by Clay (1961). In Eastern Canada, fish fences have evolved to a fairly uniform structure, similar in detail to Figure 13. The fence superstructure is usually secured to a permanent ballast-filled timber crib extending completely across the river and set flush with the natural river bed. Occasionally, a permanent base is provided by constructing a reinforced concrete slab or by driving timber or steel piles. The superstructure, consisting usually of timber or steel members, is secured to the counting fence base so as to support movable screen panels. The alignment, and consequently the length of the fence, is determined by calculating the net open area requirements through the screens to prevent significant head loss. Fish traps are secured to the counting fence superstructure and to the base, and stability is provided by structural members. Frequently the fish traps are provided with a vertically rising false floor to facilitate fish counting. Best results have been achieved by setting the fence diagonally across the stream so as to maximize the guiding effect of the fence on upstream and downstream migrating fish. A portable steel fence, similar in design but easily transported by helicopter, has recently been developed in Newfoundland. This steel fence utilizes closely spaced electrical conduit pipes placed in holes in horizontal I-beam stringers. Durability is better than with older type fences, although operational failure is higher than with permanent structures secured to a weighted base.
One of the most successful fences in terms of data collection was installed on a tributary to the Miramichi River, New Brunswick. The tributary, the Northwest Miramichi River, has been fenced since 1950. Kerswill (1971) described the fence and presented an analysis of data, in part gathered by means of the fence, on the comparative utilization and escapement of wild salmon originating from that river. Saunders (1967) was able to determine the breeding isolation of early and late running grilse and salmon by comparing relative spawning distributions above the fence. Elson (1957) used the fence for estimating the recruitment needs for maintenance of Atlantic salmon. Saunders and Sprague (1967) used the counting fence on the Northwest Miramichi River to show that large numbers of ascending adult Atlantic salmon returned downstream after encountering copper-zinc pollution from a bese metal mine upstream of the fence. The fence provided a unique opportunity to document avoidance reactions of salmon to pollution in the fish's natural environment. In his analysis of the effects of economic growth and industrial development on the ecology of the native Atlantic salmon population, Elson (1974) summarizes the population dynamics data gathered at the fence over a 20-year period.
Most fishery biologists are painfully aware of the limitations of conventional catch statistics in the analysis of long- and short-term changes in fish populations, hence the apparent insatiable appetite for data supplied by fish passes, traps and weirs. The fact that anadromous species conveniently “present themselves” in space and time at two important phases in their life history has been a boon to those who wish to quantitatively study the population dynamics of these species.
The passion with which some of these studies has been pursued could explain, in some instances, the downward trends in abundance that have often been revealed. Whereas biologists usually abhor dams at which data can be gathered on fish populations, they seem to have a strong affinity for fish fences. Many fishery biologists can discuss at great length the adverse effects of dams on anadromous fish populations, but too few seem to acknowledge that similar effects may result from poorly designed and operated fish fences.
All counting fences, to my knowledge, delay adult upstream spawning migrations. The extent of this delay may determine the reproductive success of a given year class. In some instances delays will cause a redistribution of spawners in favour of reaches below the fence. Delays may also change the patterns of angler exploitation with resultant changes in spawning escapement and the creation of some vexing public relations problems.
Extreme care must be exercised when entire runs of fish are intercepted at either dams or fences. Population changes may be a direct effect of handling and sampling. Operating regimes should be developed to adequately handle the maximum number of fish encountered. More attention should be given to sampling theory so as to minimize the number of specimens physically handled, and therefore lessen the stress on the entire population. Care must be exercised to minimize stresses that may be selective in terms of species, individual races, size, sex, or other less readily recognized characteristics.
With appropriate operating safeguards, fish passes and fences are excellent tools to document changes in fish populations. In the original design of a fish pass or fence, adequate provision should be made for the biological sampling and enumeration of the fish runs of interest. Some of the latest developments in photography and electronics offer promise for counting large numbers of mixed species. These techniques, however, are still in the developmental stage. No routine application of these methods, to my knowledge, exists in Eastern Canada.
On occasion, fish counts at a fishway may indicate fewer pollution effects on migratory fish than are suggested by other data. The Exploits River in Newfoundland, for example, receives waste from a base metal mining operation, from a pulp and paper mill, and from several towns. Routine water quality sampling showed periodic oxygen levels as low as 60 percent saturation and many heavy metal concentrations exceeding that known to impede salmon migrations. Fishway counts at a dam 1 km above tidewater, however, indicated no apparent effects from the pollution until river discharges fell to very low levels. Further analysis of the data revealed that low dissolved oxygen levels were localized in the estuary where the fish could avoid them, and that the pulp mill waste was reducing the heavy metal toxicity by chelation on organic matter.
The use of local commercial fishing gear often offers an inexpensive but effective method for obtaining useful information on changes in population size and character. Many of the disadvantages, in terms of interfering with the natural migrations of relatively large runs of wild fish, can be overcome by sampling with a constant fishing effort. The catch method will then determine how representative the sample is of the entire population. Several units of gear can be utilized to obtain estimates of variance and conventional sampling theory can determine the sample size for any chosen degree of accuracy. The modified commercial salmon trap used by Ruggles and Turner (1973) is an example of an unsophisticated sampling device capable of assessing a variety of population changes in Atlantic salmon.
It is a pleasure for me to acknowledge assistance received from a number of Resource Development Branch staff. In particular, I should like to thank K.E.H. Smith, R.E. Cutting, G.E. Turner, D.C. Riley and T.L. Marshall for firsthand information on a variety of fish- ways, traps and weirs. P.F. Elson, St. Andrews Biological Station, kindly provided unpublished data from the northwest Miramichi River salmon fence. The figures were done by N. Whynot and were based largely on drawings provided by Resource Development Branch engineering staff.
| Time period | Commercial fishery | Sport fishery | Spawning escapement | Total |
| 1954–58 | 11 800 | 2 300 | 6 700 | 20 800 |
| 1959–63 | 13 000 | 2 100 | 4 600 | 19 700 |
| 1964–68 | 19 000 | 1 700 | 5 300 | 26 000 |
| 1969–73 | 3 200 | 1 700 | 8 800 | 13 700 |
| Year | Numbers |
| 1968 | 22 000 |
| 1969 | 106 000 |
| 1970 | 85 500 |
| 1971 | 394 000 |
| 1972 | 1 737 000 |
| 1973 | 1 400 000 |
Figure 1 The Saint John River, New Brunswick, showing the location of Mactaquac, Beechwood and Tobique dams.
Figure 2 Tobique River dam with pool-and-weir fishway, showing the fish counting trap at the fishway exit.
Figure 3 Beechwood dam showing the location of the fish collection gallery and fish hoist
Figure 4 Beechwood fish collection facility, showing the fish collection gallery and the skip hoist used to pass upstream migrating fish over the dam.
Figure 5 Mactaquac dam showing location of spillway, powerhouse, fish collection gallery and fish handling facilities.
Figure 6 Details of the Mactaquac fish collection and primary sorting facility.
Figure 7 Secondary sorting facility located at the Mactaquac Fish Culture Station 2 km below the Mactaquac dam.
Figure 8 Louver deflectors used for guiding Atlantic salmon smolts from a power canal at a hydroelectric site in Nova Scotia.
Figure 9 The adult sampling trap used to monitor Atlantic salmon runs in the Miramichi River estuary, New Brunswick.
Figure 10 Atlantic salmon catch at Miramichi adult sampling trap from 1954 through 1973 and potential egg deposition represented by these fish.
Figure 11 Percentage composition of large salmon and grilse in the Miramichi sampling trap catch, 1954–1973.
Figure 12 Seasonal timing and relative abundance of grilse and large salmon entering the Miramichi estuary, based on average daily catch per 2-week period at the sampling trap from 1954 through 1970. The timing of the local commercial fishing season for driftnets and trapnets is also depicted (Ruggles and Turner, 1973).
Figure 13 A hypothetical Atlantic salmon counting fence similar to several used in Eastern Canada for enumerating upstream and downstream migrating fishes.
