G. Rasmussen
Inland Fisheries Laboratory
Silkeborg, Denmark
ABSTRACT
During the past century pond-reared trout have been liberated into natural Danish fresh waters in order to compensate for decreasing salmonid production caused by the increasing deterioration of the physical environment. At the beginning of the century such stocking was performed in a very uncritical way without any regard to the biological requirements of the fish and no beneficial effect of these stockings has been demonstrated. From the middle of the thirties, however, stocking of trout fry was systematized with regard to the habitat requirement of this developmental stage. Today, modern stocking schemes operate with fry, half-yearlings, yearlings and two-year old fish (smolts). At present about 56 percent of Danish catchment areas are managed according to such modern schemes, according to which about 1 600 000 fry, 350 000 half-yearlings, 200 000 yearlings and 130 000 two-year old fish are liberated per annum.
The paper discusses the biological and technical principles for these schemes and gives an example of their effects on the trout stocks. The paper finally discusses the actual problems and outlines the work to be done in the future.
RESUME
Au siècle dernier, on a lâché des truites élevées en étang dans les eaux douces naturelles du Danemark pour compenser la diminution de la production de salmonidés, diminution causée par la dégradation de plus en plus marquée de l'environnement. Au début du siècle, ces opérations étaient encore effectuées de façon empirique, sans tenir compte des exigences biologiques du poisson; rien ne permet de dire qu'elles ont été effectivement bénéfiques. C'est vers 1935 qu l'on a commencé à repeupler méthodiquement les eaux avec du frai de truites en tenant dûment compte de l'habitat nécessaire. Aujourd'hui, on déverse chaque année dans les eaux intérieures environ 1 500 000 alevins de moins de six mois, 500 000 alevins de six mois, 300 000 sujets d'un an et 220 000 sujets de deux ans. Environ 56 pour cent des bassins versants du Danemark sont ainsi aménagés.
L'auteur examine ces opérations du point de vue technique et biologique et donne des exemples de leurs effets sur les stocks de truites. Il passe en revue les problèmes qui subsistent et indique brièvement le travail qui reste à accomplir.
Denmark is typically lowland with sandy or gravelly soils.
The streams of western Jutland are characteristically cold in summer (maximum 12–15°C), relatively poor in nutrients (150–300 μS/cm) and have a large run-off in summertime (about 7–15 1/sec/km2). These streams are potentially excellent for spawning and growth of trout (Salmo trutta L.) provided that physical interventions are reduced or can be performed with better environmental understanding.
The streams of eastern Jutland and of the islands rise in morainic soils and in the summer may reach temperatures above the optimum temperature for trout (about 15°C) and even close to lethal temperature (about 25°C). The streams are rich in nutrients (200–600μS/cm) and are today more or less influenced by biologically purified sewage. In critical periods during summer, run-off may reach values as low as 1 l/sec/km2. Furthermore, the greatly increased extraction of ground water especially in Zealand must be blamed for the disappearance of formerly good trout streams, whereas the streams in eastern Jutland and in Funen are still affected by agro-technical interventions and the presence of obstructions to spawning migration. In these streams, therefore, the possibilities for reproduction are reduced, whereas the conditions for growth are excellent.
When summer run-off exceeds a level of ≥ 1–2 1/sec/km2, Danish streams are potentially excellent waters for spawning and growth (Larsen, 1955; 1961; 1972).
During the last century agro-technical interventions such as land drainage, culverting of small streams, straightening, canalization and diversion of larger streams changed the hydraulic conditions in the streams, principally by increasing the annual fluctuations between minimum and maximum run-off and the conveyance of suspended material from the fields to and transport in the streams. The discharge of acid ferruginous water was also increased by drainage and lignite mining. Interventions for agricultural improvements of this kind have been followed by increased applications of fertilizers which have resulted in greater weed growth in the streams. To counteract this, intensive weed cutting and stream maintenance are needed with up to three weed cuttings per season and frequent dredging of bottom and bank material, which greatly reduces the areas for spawning and growth of salmonids.
Since the Middle Ages weirs at mills and industrial plants have effectively prevented spawners from reaching the spawning grounds. Water intakes for irrigation and fish farms have not only obstructed the migration of spawners but also the downstream migration of smolts. Because stocking with trout in streams and coastal waters have been undertaken with varying success throughout this period, it is not possible to evaluate the degree to which original trout populations have naturally been reduced to the present level.
The following review describes the past and present practices in Denmark for stocking of trout into natural freshwater areas. Furthermore, it recommends intensified biological investigations of existing natural trout populations, investigations of stocking material, together with a re-examination of policies for future stocking.
The history and the biological principles of the Danish trout stocking programme have been described by Otterstr∅m (1938) and Larsen (1972). Apart from the comparatively recent introduction of electrical fishing (Larsen, 1955) the present practice has not changed since the middle thirties.
Up to the middle thirties trout stocking was carried out uncritically with no proper consideration of the suitability of the streams for fry or which year-class was to be preferred for stocking. In those cases where it has been possible to check the effect of such stockings (Poulsen, 1935) it has been demonstrated that generally speaking they had been without value.
Today the following year-classes are being used: fry, half-yearlings, yearlings and two-year olds and, if desired, older fish, the latter usually in connexion with “put-and-take” fishery. Stocking is performed according to the following principles:
The stocking material is distributed as widely as possible in the river system in order to make use of its total productive capacity.
The size of the material stocked must conform to the summer depth of the stream in question.
The numbers of fish liberated must be adjusted according to the estimated carrying capacity of the locality, minus any natural trout population already present.
(i) and (ii) are relatively easily demonstrated on site, whereas (iii) at least implies a quantitative survey of the trout population by means of electro-fishing over a representative section of the stream, if necessary combined with scale sampling. The carrying capacity of the site for an individual year-class of trout is evaluated first according to (ii), then with respect to the different requirements of the various trout sizes (i.e., year-classes) to the physical conditions of the stream. Thus a high quality biotope for fry will typically be a shallow (depth about 10 cm) small stream with gravel or pebbles, constituting a mosaic of permanent territories on the bottom. Water weeds and overhanging vegetation on the banks may also create territories during the spring and summer, but it is essential that territories are also present during wintertime when weeds normally are sparse. A suitable fry biotope with such characteristics may also be found in the main stream provided that the original structure of the bottom with alternations between shallows and pools is retained. A locality evaluated as potentially suitable for liberation of larger fish requires a greater amount of shelter such as undercut banks, overhanging bank vegetation or alder roots. Localities of this kind are typically found in the greater parts of a river system.
Any site is evaluated according to a “trout biotope scale”, which is linear from 0 (i.e., unsuitable because of stagnant water, pollution, total lack of shelter, etc.) up to 5, which typically satisfies a trout population, physically as well as chemically. However, a biotope score of 5 for fry differs in its physical constitution from a level 5 biotope for, for example, two-year old fish.
Following this evaluation, an on-site investigation of the actual need for stocking is undertaken by determining the existing trout population by means of electro-fishing. In practice, the field investigations take place in the autumn, when the young of the year resulting from natural reproduction have reached a size when they may be caught with a reasonable sampling accuracy. As the investigation aims at quantifying the natural reproduction potential of the reach in question, no fry are stocked in the preceding spring. However, releases of older fish are allowed. As half-yearlings are only liberated after electro-fishing it is possible in this way to demonstrate the natural breeding of young of the year (half-yearlings), together with breeding of any older fish which have been liberated in the preceding spring, possibly mixed with naturally-hatched trout.
The need for stocking is then worked out on the basis of knowledge gained from previous stockings together with the experimentally determined size and composition of the population, and a knowledge of the obstacles in the stream system. It should be stressed that a stocking programme covers a whole river system and not only a part of it. The exact quantity to be stocked is calculated on the basis of the area of the stream sections and on the demonstrated need for stocking.
A so-called trout stocking scheme is then worked out in the laboratory, which is based on stream descriptions and the results of the electro-fishings, and also makes provision for stockings in the years to come. The scheme is accompanied by a map (Fig. 4) which shows that the localities where fry are found are typically in the headwaters of the tributaries, where the average width of the streams varies between 0.5 to 1.5 m. The localities for half-yearlings are found downstream of the fry localities indicating that young trout gradually move downstream as they grow. This downstream movement of the fry must also be taken into consideration when evaluating the quantity of half-yearling fish to be liberated. The distribution of the localities for stocking one and two-year old fish are determined in a similar way. A stocking scheme also includes suggestions for the liberation of two-year olds (or older) at localities where it has not been possible to electro-fish.
Stockings with fry and one-year old fish takes place in April with densities of up to 200 and 20 fish/100m2 respectively for the best biotopes, whereas half-yearlings are stocked in October up to a density of 50 fish/100 m2.
Stockings with older fish normally take place in June or July to avoid excessive migration to the sea. However, contrary to this principle, some two-year old fish (regarded as smolts) are stocked in the mouth of the rivers in April. These fish are expected to leave the stream almost at once, thus not burdening the stocking site with extra fish to compete for territories. The numbers of fish thus stocked depend on the economic possibilities at this time.
Ideally, the principles regulating a stocking scheme should be revised every fifth year, not only because of changes in the state of pollution, obstructions, streams maintenance, etc., but also because self-reproducing trout populations may have arisen as a result of stocking. Alternatively, satisfactory trout populations may not have been created by means of the stocking system adopted, so that a revised scheme may be necessary.
The Inland Fisheries Act states that in a river system where a stocking scheme has been worked out and approved by the fishery authorities trout may only be stocked in accordance with its recommendations. The numbers may, however, be reduced in the case of economic difficulty but the numbers provided for in the scheme are the maximum allowed and correspond to the carrying capacity of the system.
The percentage of various catchment areas for which stocking schemes have been elaborated according to biologically sound principles are shown in Table 1 and Fig. 1 also indicates schemes which have been revised at least once. In 1982, 1983 and 1984 a further six river systems and one delimited area with several independent small streams will be included in these activities.
Where the establishment of a fish farm results in a drop in the potential trout production in any stream, the fish farm is legally obliged to carry out compensatory stocking. The total number of trout stocked under (i) the stocking schemes now in force, and (ii) the compensatory stockings by fish farms (shown in Table 2).
The recommended numbers of trout stocked according to the schemes may be reduced because of a lack of economic capacity on the part of the interested group of people (usually an angling club or a union of clubs). Furthermore, a number of two-year old fish are liberated directly into coastal waters and in river mouths. Thus, actual numbers fall short of those recommended and on an average the annual trout liberations in 1982 are as shown in Table 3.
Stocking in river systems for which no schemes have been worked out must also be included in the costing. It is, therefore, considered that trout at a value of about 2 000 000 D.Kr. were liberated into Danish streams and coastal areas in 1981/82.
There is no public authority responsible for trout stocking in Denmark. All stocking and the associate management is performed by individual angling clubs or unions of clubs which cover the river systems in question.
The field investigations and the development of the stocking schemes are funded by public money, whereas the clubs themselves buy and liberate the fish at their own expense. By far the greater part of the trout liberated are bought as cultured fish from private, commercial trout farms. However, within the last couple of years a few clubs have started experiments in catching spawners, stripping them, hatching the eggs and rearing the different size categories of the fish themselves. Normally the rearing of pre-fed fry involves no problems but except for one case no clubs have had the possibility of producing older year-classes yet.
Several clubs have permission to catch wild spawners every autumn, strip them and place the eggs with commercial trout farms where they are reared to the appropriate stocking sizes. Rearing fry from wild spawners up to say two-year old fish in this way involves a certain selection, because in trout farms fish are selected for other criteria than in natural stocks (fish develop lack of shyness, they willingly take pelleted food and mortality is reduced because of faster growth). There is, however, no doubt that this practice preserves a certain genetic variety and fitness of the trout populations in the river systems in question. Fig. 2 shows those river systems where this kind of fishery management has been performed over a varying number of years.
There are still some unions or smaller clubs which undertake stocking in ignorance of existing schemes for parts of a river system, or they stock fish into systems where no biological investigations have been undertaken. They use fish casually bought from trout farms not knowing their origin, which often implies that the fish stocked are the cheapest, and often discarded fish, poorly adapted to natural waters. Typically this practice applies to two-year old fish for put-and-take fisheries.
The above explanation of the present stocking activities indicates the need for better coordination of biological knowledge of natural trout strains, the definition and description of “optimal” fish for stocking and the selective breeding associated with the development of such strains.
The stockings which have been undertaken over the past years with varying intensity raise the important question of the possible effect or influence of such stockings on the original trout populations. To what extent does stocking affect biological parameters such as brown trout/sea trout relationship, age and size at smoltification, sex distribution of smolts and of sea trout, age and size at sexual maturity at first spawning, etc.? At present we do not know with certainty whether Denmark possesses pronounced brown and sea trout strains which are reproductively isolated.
The occurrence of early and late sea trout strains in larger river systems has been claimed, and such strains have been regarded as reproductively (i.e., genetically) separated. The early strain appeared in river mouths in March/April and could be fished in the main streams as early as May. Contrary to this the late strain was composed of small fish not present on the spawning grounds until just before and during spawning. It is not, however, clear whether these fish which have differing migration patterns were also reproductively separated. If so, why does the early strain occur so rarely in today's catches. Because reliable statistics for sea trout, distributed on size classes, sex and areas do not exist, we have only limited possibilities to resolve such questions.
In salmon (Salmo salar L.), late spawners (grilse) are dominated by males with a short sea life (i.e., 1.5 years), whereas early spawners (salmon) are composed of the rest of the males and by far the greater part of the females which have not yet spawned and have a sea life of at least 2.5 years before maturity. It is therefore tempting to claim that even if sea trout normally spawn several times in consecutive years, later and early strains in sea trout arise from differences in age and sex composition. In this way, young mature males and females would arrive late in the streams and -because they are small -spawn in the smaller tributaries, whereas the older mature fish (mainly females), together with a few large specimens which spawned previously, will stay a much longer time in the stream. In these the spawning grounds will be concentrated in the main stream. In this way, a certain genetic mixing might result (i.e., from large, but early spawners). Whether or not there is a genetic difference between the two strains is critical to the evaluation of the trout stocks. Among other factors the age and size at smoltification (the fastest growing young fish are first to become smolts, and the slowest growers smoltify later), however the average size of a smolt increases with its age of smoltification. Thus large smolt will become mature at an earlier age, which means that in any smolt year-class one would expect to find an increasing proportion of early smoltifying fish as the year-class progresses. This implies that young fish in small streams that grow relatively slowly would become smolts at two-three years of age but as relatively large smolts which might mature relatively early. Contrary to this, fast-growing young fish in the main stream may already become smolts at an age of one-two years, but their relatively smaller size requires that they return to the spawning grounds as large but late (i.e., by age) mature fish (“the early strain”).
Under this hypothesis, factors which might have been responsible for the change in the trout population could have been the removal of spawning grounds in the main streams for navigation and draining and an intensified coastal and stream fishery for big sea trout. This may have lead to a reduction or removal of the possible genetically determined “large and early” sea trout strain, and a reduction of the natural sea trout population to a relatively later smoltifying and early maturing, slow-growing population with a small maximum size.
At the same time, these natural populations are exposed to great pressure from stocked farm-reared trout which are selected for fast growth and which are therefore relatively large at smoltification, mature early, have decreased growth and attain a small maximum size. It is not known to what extent these genetic mixtures counteract (the growth of young fish) or enhance (early maturity followed by slow growth) each other.
The achievement of the original objectives requires a better understanding of the biological characteristics of the natural sea trout strains in the river systems, particularly of the spawners to be stripped so that selection can be made of specimens especially fit for breeding. It also requires the recording of changes in numbers, period of immigration, size and age at smoltification besides possible changes in age-composition at maturity as a result of breeding and stocking over the next 20 years. Tagging of two-year old smolts and the wild trout from the stream under study and from other river systems as well as electro-fishing following stocking - will also be necessary to monitor the progress of such populations. Better knowledge of the time and size of stocking of all size groups is also needed. For example, it is normally recommended that fry should be pre-fed for three weeks before stocking. In the case of pond-reared fry this means that stocking normally takes place in March/April, whereas naturally hatched fry have not been detected in the streams until a month later.
It is, however, doubtful whether there are any great differences between strains of trout in the various river systems within a geographical region as limited as Denmark. There might originally have been great genetic differences in such parameters as size and age at smoltification and also the size of immigrants in brooks and small streams with direct outflow in the sea, which cannot be entered until October/November when precipitation is highest compared to the larger streams and rivers (i.e., Gudenå, Skjern å, Skive å), into which immigration is practicable at all times of the year. Possibly, the focus should be on developing a strain of trout which is adapted to all larger river systems and disregard the idea of individual strains for individual streams. This would essentially facilitate the production process in fish culture stations where the usual procedure of frequent screenings of equal size groups could be maintained. In this way, costs would also be reduced.
The River Ribe å has a catchment area of about 680 km2. The main stream rises in eastern Jutland about 41 m above sea level in the region close to the main stationary line of the last glaciation, and proceeds from there along a more or less straight line westward. After a course of about 71 km it discharges into the Wadden Sea 10 km west of the city of Ribe. Besides numerous smaller tributaries, the main stream receives two major tributaries, from the right the River Tved å and from the left the River Gels å, both themselves with larger and small tributaries.
A few of the tributaries are more or less polluted from farms and small villages, but the most important deterioration of the river as regards trout fisheries arises from early re-alignments and channelizations and, more recently, with a heavy-handed river management policy and weed cutting.
Besides trout the following fish occur in the river: rainbow trout (Salmo gairdneri Rich.), pike (Esox lucius L.), roach (Rutilus rutilus (L.)), dace (Leuciscus leuciscus (L.)), minnow (Phoxinus phoxinus (L.)), tench (Tinca tinca (L.)), gudgeon (Gobio gobio (L.)), eel (Anguilla anguilla (L.)), burbot (Lota lota (L.)), perch (Perca fluviatilis (L.)), ruffe (Gymnocephalus cernua (L.)), stickleback (Gasterosteus aculeatus (L.)), ten-spined stickleback (Pungitius pungitius (L.)) and also lampreys (Cyclostomata). It is known that salmon previously occurred in the river. Bream (Abramis brama (L.)) and pikeperch (Stizostedion lucioperca (L.)) also occur in adjacent lakes.
An old stocking scheme for trout fry dates back to 1951 which was more or less adhered to. In 1972 systematic electro-fishings and biotope evaluations were undertaken and from the spring of 1974 regular trout stockings were started according to a biologically sound stocking scheme.
Altogether 176 sites were inspected. Out of these 65 were estimated to be unsuitable for trout. Of the remaining 111 sites suitable for trout 33 were not electro-fished either because of difficult access or because the river was too large. The remaining 78 suitable trout sites were electro-fished, 32 of which were devoid of trout. On the remaining 46 sites trout densities between 0.3 and 245 per 100 m2 were found. The total need for restocking was calculated at 89 000 fry distributed at 52 sites, 31 000 half-yearlings at 25 sites, 8 700 yearlings at 4 sites, and 16 600 two-year olds at 7 sites, besides an unspecified number mainly depending on economic capacity, though not exceeding 14 500 specimens.
This scheme was revised in 1981. This time 194 sites were inspected (Fig. 3). A few new sites were included but generally the sites were similar to those sampled in 1972. Out of the 194 sites 62 were estimated to be unsuitable for trout, of the remaining 132 trout sites electro-fishing was not performed at 27 either because of difficult access or because of the width of the river. Of 105 suitable trout sites where electrofishing was performed, only 15 were found to be devoid of trout. There was, therefore, a distinct improvement as compared to the previous investigation. At the remaining 90 sites suitable for trout densities of between 0.8 and 207 specimens per 100 m2 were found.
The total stocking need today (Fig. 4) is calculated to comprise 47 300 fry at 36 localities, 20 900 half-yearlings at 16 localities, 13 300 yearlings at 14 localities and 16 400 2-year olds at 9 localities. Besides this an unspecified number (not exceeding 14 500) of two-year olds distributed in the main river at five localities. In conclusion, it can be stated that an immediate semi-quantitative estimation of the results of the electro-fishings shows an allover improvement of the trout population on the localities examined for the whole river system.
For a more reliable interpretation, localities examined in 1972 and 1981 were grouped in comparable and incomparable localities.
In 1972 a total of 82 localities of a total area of 9 946 m2 were electro-fished, corresponding to 4 096 m of stream. In 1981 a total of 113 localities of a total area of 11 263 m2 were electro-fished, corresponding to 5 140 m of stream.
Sixty-three localities are immediately comparable. At these localities electro-fishing was performed in over 3 088 m of stream, corresponding to an area of 6 372 m2, whereas in 1981 at the same localities electro-fishing took place in over 2 890 m of stream corresponding to an area of 6 613 m2. Even if both surveys were undertaken during autumn, there was great difference in individual run-offs. The summer of 1981 was very rainy, stream widths and depths were greater than in 1972. Thus the broad, shallow sea trout river Gels å was 18 percent wider on average and the other localities 14 percent wider in 1981 than in 1972. This meant increased difficulties when fishing in the River Gels å (average width in 1981 - 5.21 m) and at the same time it was possible for the fry to spread over a relatively larger area. Another significant factor is the year-class strength. During the investigation supplementary scale sampling was undertaken, and this showed that in the localities examined in the Gels å the 1980 year-class had been exceptionally small, whereas the 1981 year-class was of the same strength as the 1972 year-class. This situation was not detected in other parts of the Ribe å system. Because of this discrepancy comparisons between the two years were made including and excluding the Gels å.
The results of the comparison are shown in Table 1. As the material is based on a reasonably large number of sites it is clearly demonstrated that stocking has had an exceedingly good effect on the total trout population in the whole Ribe å system.
Larsen, K., 1955 Fish population analyses in some small Danish trout streams by means of D.C. electro-fishing; with special reference to the population of trout (Salmo trutta L.). Medd.Danm.Fiskeri Havunders.(Ny Ser.), 1(10):1–69
Larsen, K., 1961 Fish populations in small Danish streams. Verh.Int.Ver.Theor.Angew.Limnol., 14:769–72
Larsen, K., 1972 New trends in planting trout in lowland streams. The results of some controlled Danish liberations. Aquaculture, 1:137–71
Otterstrøm, C.V., 1938 On methodical liberations of salmon and trout fry in watercourses with special reference to the Gudenaa area. Rep.Dan.Biol.Stn., 42:3–32
Poulsen, E.M., 1935 Recent investigations into the stock of salmon and sea trout in the River Gudenaa. Rep.Dan.Biol.Stn., 40:9–37
Table 1 Stream areas subject to trout stocking schemes in Denmark (up to 1981)
| Area (km2) | % with stocking scheme | |
| Jutland (1984) | 29 643 | 59 |
| Funen | 2 984 | 85 |
| Zealand (w.islands) | 9 310 | 26 |
| Bornholm | 588 | 100 |
| Total area | 42 555 |
Table 2 Recommended numbers of trout stocked into Danish waters (1981)
| Stocking schemes | Compensatory stocking | |
| Fry | 1 600 000 | 27 000 |
| Half-yearlings | 437 000 | 83 000 |
| Yearlings | 356 000 | 99 000 |
| Two-year olds | 273 000 | 15 500 |
Table 3 Actual numbers of trout stocked into Danish waters (1982)
| Number | Value (D.kr.) | |
| Freshwater: | ||
Fry | 1 600 000 | 152 000 |
Half-yearlings | 350 000 | 273 000 |
Yearlings | 200 000 | 240 000 |
Two-year olds | 130 000 | 533 000 |
| River mouths/coastal waters | 75 000 | 308 000 |
| Compensatory stocking from fish culture, all categories | 250 000 | |
| Total | 1 756 000 | |
Table 4 River Ribe å - comparable localities in 1972 and 1981
| 1972 | 1981 | Difference % | |
| No. of sites: 63 (10 of these in Gels å) | |||
| Metres electrofished - total | 3 086.0 | 2 890.0 | |
| Metres electrofished - Gels å | 550.0 | 440.0 | |
| No. of trout caught - total | 2 012.0 | 2 273.0 | |
| No. of trout caught - Gels å | 1 139.0 | 831.0 | |
| No. of trout/100 running m - total | 65.2 | 78.7 | 21.5 |
| No. of trout/100 running m - Gels å | 207.1 | 188.9 | -8.8 |
| No. of trout/100 running m - outside Gels å | 34.4 | 58.9 | 71.2 |
| No. of trout fry caught - total | 1 236.0 | 1 312.0 | |
| No. of trout fry caught - Gels å | 782.0 | 721.0 | |
| No. of trout/100 running m - total | 40.1 | 45.4 | 13.2 |
| No. of trout fry/100 running m - Gels å | 142.2 | 163.9 | 15.3 |
| No. of trout fry/100 running m - outside Gels å | 17.9 | 24.1 | 34.6 |
| No. of older trout/100 running m - total | 25.1 | 33.3 | 32.7 |
| No. of older trout/100 running m - Gels å | 64.9 | 25.0 | -61.5 |
| No. of older trout/100 running m - outside Gels å | 16.5 | 34.7 | 110.2 |
 | Schemes based upon only one field survey and not revised since. |
 | Schemes revised at least once. |
 | Schemes planned for autumn 1982–1984. |
Fig. 1 Geographical distribution of stocking schemes based on catchment areas
 | Pond reared fish of more or less known origin. |
 | Schemes previously fulfilled with fish from the systems' own strains, but now being fulfilled with pond reared fish of more or less known origin. |
| Schemes throughout the years fulfilled with fish from own strains |
| Schemes previously fulfilled with fish bought at random, but where now a production of trout of own strains has started. |
| The Funen - scheme includes a total of 60 streams systems, fulfilled with fish from the river Odense å. |
Fig. 2 The origin of stocking material
Fig. 3 Test fishing map of the Ribe River system
Fig. 4 Liberation map of the Ribe River system
