Previous Page Table of Contents Next Page


THE ESTIMATION OF JUVENILE SALMON (Salmo salar L.) POPULATIONS IN THE NORTH ESK, SCOTLAND
EVALUATION DE POPULATIONS DE SAUMON JUVENILE (Salmo salar L.) DU NORTH ESK, ECOSSE

by/par

William M. Shearer
Freshwater Fisheries Laboratory
Pitlochry
Scotland/Ecosse

ABSTRACT

The necessity for enumerating or estimating salmon populations at different stages during their life history, in order to effectively manage salmon stocks, is discussed.

Since salmon undertake at least two major migrations during their lifetime, the possibility of enumerating salmon populations directly either in traps or by automatic counters is considered in relation to the cost of their installation in any major salmon river.

Since their present installation cost, except at special sites, prohibits their general use, the possibility of estimating the juvenile salmon production from a major Scottish salmon river, the North Esk, is investigated and two methods based on the tag-recapture principle are described.

The first of these methods estimates the number of smolts which migrate to sea each spring and early summer and the second estimates the total annual recruitment of juvenile salmon from the North Esk to the sea. The limitations, particularly of the first method, are discussed and the evidence is examined for suggesting the absence of particular forms of experimental bias which would invalidate the method.

In the absence of any quantitative estimate of the spawning population, or of the proportion of males to females, the possible sources of error in the estimates are discussed.

RESUME

Il est discuté de la nécessité de dénombrer ou d'évaluer les populations de saumon à différents stades de leur vie, afin d'en exploiter efficacement les stocks.

Le saumon entreprend au moins deux migrations principales au cours de sa vie et, sur ces bases, on considère la possibilité de les populations de saumon directement soit au moyen de pièges ou de compteurs automatiques, eu égard au coût de leur installation sur l'un quelconque des cours d'eau à saumon.

Du fait qu'actuellement, leur coût d'installation, sauf pour quelques emplacements particuliers, empêche leur utilisation généralisée, on enquête sur la possibilité d'évaluer la production de saumon juvénile d'un cours d'eau principal écossais à saumon, le North Esk, et on décrit deux méthodes basées sur le principe de marque-recapture.

La première de ces deux méthodes évalue le nombre de smolts qui émigrent vers la mer chaque printemps et au début de l'été, et la seconde évalue le recrutement total de saumon juvénile du North Esk à la mer. On discute des limites, en particulier, de la première méthode et on en examine une démonstration suggérant l'absence de formes particulières d'erreur systématique expérimentale qui infirmerait la méthode.

On discute des sources possibles d'erreur d'évaluation en l'absence d'estimations quantitatives de la population de reproduction ou de la proportion de mâles et de femelles.

CONTENTS

1. INTRODUCTION

2. MIGRATIONS OF ATLANTIC SALMON

2.1 Juvenile Salmon
2.2 Adult Salmon

3. COUNTING METHODS

3.1 North Esk Trapping System

4. ADULT STOCKS

5. ESTIMATION OF SMOLT MIGRATION

5.1 Sources of Error

6. ESTIMATION OF TOTAL JUVENILE MIGRATION

7. DISCUSSION

REFERENCES

1. INTRODUCTION

The effective management of a salmon (Salmo salar L.) stock, to produce maximum yields for both rod and net fisherman without depleting the spawning stock below an acceptable level, requires more precise information regarding the relationships which exist between stock, recruitment and mortality.

For example, the optimum spawning escapement from the netting area cannot be measured in terms of the number of fish necessary to produce the greatest number of juvenile salmon at the seaward migrating stage (smolts), since this number of individuals is unlikely to be the same as that which would produce the largest angling catch. In addition, the fresh water in which the fish live for a substantial part of their life is also a resource, and may be abstracted from rivers for industrial or domestic use, impounded for the production of power, or sprayed on arable land during periods of drought, when the rivers can least afford to lose it. This alters the natural flow patterns, particularly in the headwaters of rivers, and not infrequently the result is a reduction in the area available for fish production, where the juvenile salmon population is most dense. Thus, optimum production can only be realized when the relationships between numbers of fish, area of substrate and water flow are understood.

Methods of estimating the size of salmon populations at various stages during the life history are therefore required as a first step toward understanding the many relationships which exist between the salmon and its environment, leading eventually to the effective management of the species.

This paper describes two methods which have been used to estimate the numbers of juvenile salmon produced in the River North Esk, in eastern Scotland.

2. MIGRATIONS OF ATLANTIC SALMON

2.1 Juvenile Salmon

Atlantic salmon are particularly suitable subjects for population study experiments because, being anadromous, they undertake at least two major migrations during their life when they pass through the estuary of a river on their way to or from the sea.

The first of these migrations occurs when the juvenile salmon migrate to sea from the rivers in which they were hatched, as smolts measuring between 10 and 20 cm on average. In Scottish rivers most salmon reach this stage during the spring or early summer of their third or fourth year. However, recent evidence from the North Esk suggests that there are movements of parr (the pre-smolt stage) both to and from the estuary of the river at other times of the year. The peak of the upstream migration occurs in September, while most downstream- migrating parr are trapped in October or November. These parr have a similar length frequency distribution to smolts.

2.2 Adult Salmon

The second major migration is the return of the adults from the sea feeding grounds, which in the larger Scottish salmon rivers occurs at any time from one to three years after entering the sea as a smolt. However, more than 75 percent of the maiden adults which return to fresh water do so during the first half of this period.

With relatively few exceptions the lengths of the adult salmon range from 50 to 100 centimetres.

3. COUNTING METHODS

Atlantic salmon, irrespective of whether they are migrating to the sea as juveniles or as spent fish, or returning from the sea as adults, exhibit a behaviour pattern which enables them to be trapped easily. Thus a barrier across a river, with a box placed at the down- stream end, will catch the descending juvenile salmon, and a similar box placed at the opposite end will catch the ascending adults.

The number of salmon which enter or leave fresh water can be obtained directly by trapping the river as near the mouth as practicable and counting all upstream and downstream migrants. However, such a direct approach is not possible on any but a very small salmon river, mainly because of the prohibitive cost of a trap capable of catching, at all expected river flows, all sizes of migrating fish.

Recently the use of automatic fish counters has increased, but as their range of operation is limited the majority have been incorporated in some other structure such as a hydro-electric dam, when the law requires the building of a fish pass. Elsewhere, their installation on a big river could be as expensive as a full-scale trap, and they would provide less information. Although the numbers of some species of Pacific salmon smolts migrating to sea have been estimated on the basis of counts obtained by photographing a known area of river throughout the run, no completely automatic method of counting juvenile fish has yet been devised. Even if the fish could be counted by automatic methods, sampling by some other means is still necessary in order to determine the age, length, weight and sex ratio.

3.1 North Esk Trapping System

The North Esk is one of the five most productive salmon rivers in Scotland. Its catchment area is approximately 586 km2 and its length is almost 48 km from the source formed by the union of the outlet burn from Loch Lee and the second largest tributary, the Mark, to the point where it drains into the North Sea some three miles north of Montrose. The width of the river at the head of tide during H.W.S.T. is approximately 50 metres.

The average annual rainfall recorded in Glen Esk during the last twelve years has ranged between 881 mm in 1971 and 1 613 mm in 1963. In most years the greatest amount of rain falls either in October or November, with a minimum in July. Thus the smolt migration which normally extends from April to mid-June is usually outside the period of maximum river flows.

The sampling site on the North Esk is situated on Kinnaber Mill Lade, which draws its water from the main river at Morphie Dyke (Figure 1). This dam stretches across the North Esk some 3 km upstream from the sea and 2 km above the head of tide at H.W.S.T. The amount of water drawn off is regulated by a sluice to approximately 164 million 1 per day. During drought conditions this amount is greater than that flowing in the 2-km stretch of the main river between the intake and outlet of the Lade, but is only a small proportion of the total flow when the river is in flood. Approximately 95 percent of the water abstracted at Morphie Dyke eventually returns to the main river.

The trap for all the migrants in the Lade is extremely simple in design and consists of a fence across the Lade and a holding tank into which all the fish diverted by the fence are collected and held until released. A set of screens inside the tank separates the down- stream migrants from those which are migrating upstream.

4. ADULT STOCKS

During the fishing season, which extends from 16 February to 31 August, the stock of salmon returning to the North Esk is harvested in “home Waters” by fixed nets set along the coast, and within the river by shore seines and by rod and line. The proportion of salmon, as distinct from grilse (those adults returning during the second year in the sea), in the river catch has steadily declined during the last 50 years and in 1970 grilse outnumbered salmon by a ratio of 1:3:1. At present, there is no measure of the total numbers of salmon entering the river system but accurate catch figures are available, and the proportion of the total run removed by netting within the river is being investigated at the present time. However, evidence suggests that in most years the spawning stock exceeds 5 000 fish.

5. ESTIMATION OF SMOLT MIGRATION

Every year known numbers of smolts, originally caught in Kinnaber Mill trap, are anaesthetized and tagged with individually numbered tags, and their adipose fin removed before being transported upstream by road and returned to the main river at Logie, which is situated on the right-hand bank approximately 1.5 km above Morphie Dyke. These tagged smolts, in common with the untagged members of the migrating population, have to pass Morphie Dyke in order to reach the sea, and when they reach this point they can either continue their journey in the main river or enter Kinnaber Mill Lade. All the smolts which travel by the latter route are caught in the trap at Kinnaber Mill, where each juvenile salmon is recorded as being either tagged or untagged and any fish which had lost its tag would be immediately recognized by its fin clip as having been tagged. Estimates of the total smolt population have been made by Bailey's (1951) modification of Peterson's method.

5.1 Sources of Error

Experimental error from such causes as loss of tags, tagging-induced mortality, the unrepresentative distribution of tagged fish throughout the population, the possibility that the chances of catching tagged fish as compared with those untagged are unequal, and that the ratio m:n is not constant throughout the recapture period, can completely invalidate this method of estimating population size. (Ratio m:n = proportion in sample tagged.)

During this series of experiments, no fish which had been tagged and subsequently released at Logie was recaptured without its tag. Any evidence of tag-induced mortality was particularly difficult to establish, principally because of the lack of knowledge of the direct effect of tags on the predator-prey relationship, and on the behaviour of the fish. For instance, tagged fish can be visible to a predator over a much wider range of light intensities than untagged fish. Observation suggested that tagged fish which are kept until the effect of the anaesthetic disappeared, as indicated by the adoption of a normal swimming posture, react to water currents in the same manner as untagged individuals. Delayed mortalities due to the actual tagging operation are not thought to be significant, since tagged fish have been kept in a tank without loss for up to five months and the majority of the fish recaptured are caught within 14 days.

There is no evidence to suggest that the tagged fish are not representatively distributed throughout the population, or that tagged and untagged fish are not equally liable to capture. During a series of tests tagged fish were released across the width of the river at Logie, as it was thought that perhaps those released nearer the right-hand bank would be more liable to enter the Lade, as the intake is situated on the right-hand bank. However, non-significant differences in the recapture rates of tagged fish released at the same time across the width of the river indicated that there was no bias related to the position of the release point (Table I). Furthermore, the results indicated that there were considerable changes in the positions of individual fish relative to the banks of the river during their 1.5-km journey to Morphie Dyke. This latest movement would cause a considerable mixing in the population before its arrival at the Dyke.

The recapture rate of the tagged smolts released at Logie appeared to be related to the volume of water flowing in the river following their release, since the higher rates of recapture tended to coincide with periods of low water levels when a greater proportion of the total flow entered the Lade. This is illustrated by comparing the mean daily river height in each of the smolt runs (1971, 1972 and 1973) with the corresponding percentage recapture rate of the tagged smolts released at Logie. The mean heights were 16,22 and 14 inches (40, 55 and 35 cm) compared with 5.3, 2.4 and 8.8 percent recapture rates, respectively. Similarly, the largest number of untagged smolts (14 925) was caught in 1973 and the smallest number in 1972 (6 131). Because of the nature of the recapture method, the presence of a tag did not increase the possibility of the recapture of a tagged smolt compared with an untagged one.

With few exceptions, and most of these occurred either at the beginning or toward the end of the smolt run when few fish were migrating, the ratio of untagged to tagged smolts caught in the trap during each 24-hour period was of the same order. This suggests that the migration rate of tagged and untagged smolts was basically the same, so that the constancy of the ratio m:n was maintained for each series of experiments.

6. ESTIMATION OF TOTAL JUVENILE MIGRATION

Although there has been circumstantial evidence each year since 1954 that a proportion of the juvenile salmon of the North Esk migrated beyond the head of tide at periods of the year other than during the normal smolt migration period, it was not until 1970 that this migrations was monitored. The tagging of the proportion of this migrating population caught in the trap at Kinnaber Mill showed that these fish eventually went to sea and returned as grilse or salmon in the same manner as smolts. The peak of their migration was in October or November, and the number caught in each of the periods June 1971 to March 1972 and June 1972 to March 1973 were approximately 44 and 10 percent of the smolts caught in 1972 and 1973, respectively.

There was also an upstream movement of these parr, particularly during September in each of the years 1971, 1972 and 1973, from below the head of tide into fresh water. The number of fish moving upstream each year was less than that moving in the opposite direction and less than 300 individuals were caught in the trap at Kinnaber Mill in any of those years. Although the average age of the parr caught in the autumn was approximately one year less than that of the smolts which migrated in the spring and early summer, the average length of both groups of fish was virtually the same.

When the parr migrating downstream in the autumn are transported upstreams and released at Logie, only a proportion of them return downstream and are caught in the trap that autumn, as the remainder remain above the trap and descend with the smolts migrating the following spring and early summer, before resuming their downstream journey. Therefore, it is not possible to estimate the abundance of this group of fish in the same manner as the smolts.

Between 9 000 and 15 000 juvenile salmon migrating to the sea from the North Esk are tagged each smolt year, and a proportion of them survive to return either as grilse or salmon. The seine nets (net and coble method) fishing in the river catch a known number of tagged and untagged fish, while the remainder escape to migrate above the netting area.

The assumption that the proportion of tagged individuals out of the original number tagged which return in any year equals the proportion of the total population returning is quite critical. However, the data from all years have been aggregated and Bailey's estimate has been used to estimate the total juvenile recruitment.

Before Bailey's estimate can be applied to the figures it is necessary to know, in addition to the number of juvenile salmon tagged in a given year, the number of tagged and untagged fish caught in the North Esk net and coble catch. The netsmen supply catch figures and details of all the tagged fish which they catch, and give permission for the total catch in the river on two days each week to be scaled, measured and weighed, so that each fish can be aged. On the basis of the age composition obtained, the total catch can be divided into groups according to smolt year. Analysis of the samples of scales removed from recaptures is also necessary, as any fish which has spawned previously must be omitted from the final total of recaptures, since such fish escaped capture during their initial spawning migration.

The same assumptions are implicit in this use of Bailey's method, as are described in section 5.1, but in this instance the validity of the assumptions cannot at present be tested, principally because the fish remain virtually unobserved from the time they leave the river until they are recaptured. The fishermen have reported no difference in the likelihood of untagged and tagged fish being captured due to tags becoming entangled in the meshes of the net. Fish which have lost their tag, but are recognized by the netsmen as having been tagged by the absence of their adipose fin, are included in their returns. Although there is no direct evidence at present of the existence of a differential mortality between tagged and untagged fish during their stay in the sea, this possibility cannot be excluded when considering the results.

7. DISCUSSION

The estimate from recaptures in the netting catch of the number of juvenile salmon produced annually in the North Esk, and which survive to migrate to sea, is greater than the number which migrate as smolts, estimated by the mark-and-recapture method from downstream migrants (Table II). However, on the basis of the limited results at present available, and in the absence of any measure of the spawning stock, the difference between those two estimates is greater than that which could reasonably have been expected.

The estimate of the smolt migration does not include an unknown number of parr which also migrate to sea and would contribute to the adult return. The estimate of the production of juvenile salmon provided by the net recaptures would be too high if the mortality of tagged fish in the sea is significantly higher than that of untagged fish.

A further error in the calculation from net recaptures may be due to the higher proportion of grilse which enter the weir when the nets are not in use, as compared with salmon. These fish, being smaller, would produce fewer eggs, and the estimate of smolts produced from them may in turn be too high. There is also evidence that the ratio of males to females among the grilse is higher than for salmon. All of the above factors will be examined in future studies.

TABLE I
Recapture of tagged smolts from different release sites
 Site number
 1234
Total released625625625625
(between 30 April and 17 May)    
Total recapture in trap 32 34 32 30

Note:Site 1 - right bank
 Site 4 - left bank

TABLE II
Smolt
year
Estimated
annual smolt
production
95% confidence limitsEstimated annual
juvenile salmon
production
95% confidence limits
1961  
824 600
636 000–1 013 200
1962  
969 000
695 000–1 248 200
1963  
725 500
587 300– 863 700
1964
316 900
262 400–371 400
929 800
766 800–1 092 800
1965
212 000
176 300–247 700
1 617 000
1 223 000–2 011 000
1966
199 700
163 400–236 000
1 175 400
856 200–1 494 600
1967
146 700
124 700–168 700
1 344 500
909 700–1 779 300
1968
191 800
139 800–243 800
1 450 000
1 016 000–1 884 000
1969a 
1 070 000
448 000–1 692 000
1970b 
912 000
690 000–1 134 000
1971
160 000
134 400–185 600
  
1972
256 000
196 400–315 600
  
1973
172 200
153 200–191 000
  

a Trap not functioning
b New trap not completed

Figure 1

Figure 1 Trapping arrangements on the North Esk


Previous Page Top of Page Next Page