Det här verket har digitaliserats vid Göteborgs universitetsbibliotek och är fritt att använda. Alla tryckta texter är OCR-tolkade till maskinläsbar text. Det betyder att du kan söka och kopiera texten från dokumentet. Vissa äldre dokument med dåligt tryck kan vara svåra att OCR-tolka korrekt vilket medför att den OCR-tolkade texten kan innehålla fel och därför bör man visuellt jämföra med verkets bilder för att avgöra vad som är riktigt. Th is work has been digitized at Gothenburg University Library and is free to use. All printed texts have been OCR-processed and converted to machine readable text. Th is means that you can search and copy text from the document. Some early printed books are hard to OCR-process correctly and the text may contain errors, so one should always visually compare it with the ima- ges to determine what is correct. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 C M FISHERY BOARD OF SWEDEN INSTITUTE OF FRESHWATER RESEARCH DROTTNINGHOLM Report No 50 LUND 1970 CARL BLOMS BOKTRYCKERI A.-B. FISHERY BOARD OF SWEDEN INSTITUTE OF FRESHWATER RESEARCH DROTTNINGHOLM Report No 50 LUND 1971 CARL BLOMS BOKTRYCKERI A.-B. ■ Contents The winter migrations of char, Salvelinus alpinus L., in the hydroelectric reservoirs Tunhovdfjord and Pålsbufjord, Norway; Per Aass ............................................................. 5 On the ecology of the ide Idus idus (L.) in the River Kävlingeån, south Sweden; Plutarco Cala ..................................................................................................................................... 45 Fishing pressure, growth and recruitment in a small high mountain lake; T. Lindström, Å. Fagerström and K. J. Gustafson ........................................................................................... 100 Spatial distribution of pelagic fishes in Lambarfjärden (Mälaren, Sweden) with par­ ticular reference to interaction between Coregonus albula and Osmerus eperlanus\ T. G. Northcote and H. Rundberg.............................................................................................. 133 The food ecology and the annual food cycle in the Baltic population of fourhorn sculpin, Myoxocephalus quadricornis (L.). Pisces; Lars Westin .................................... 168 Observations on the nest digging of fourhorn sculpin Myoxocephalus quadricornis (L.) ; Lars Westin ........................................................................................................................................ 211 The winter migrations of char, Salvelinus alpinus L., in the hydroelectric reservoirs Tunhovdfjord and Pålsbufjord, Norway By Per Aass Ministry of Agriculture, Vollebekk, Norway. I. Introduction The conversion into reservoirs of Scandinavian arctic or high boreal char lakes has a pronounced effect on the winter movements of the char, Salvelinus alpinus L. The water is stored mainly in spring and summer, while most of the draining and nearly all the lowering of the reservoirs take place in winter. Large-scale winter migrations are unknown in unimpounded lakes, where the winter dispersion of char apparently does not differ much from the autumn distribution. In reservoirs, however, the impoundments at once set off extensive and often far-ranging migrations, the char gathering upstream of the outlets and in sounds if such exist. The phenomenon is widely known, and the fishermen are quick to take advantage of the new situation. These migrations have been briefly described by Aass (1963), Runnstr0m (1964) and Hanell & Holmberg (1968) and in more detail by Aass (1965). The circumstances releasing the migrations and their extent and range are, however, inadequately known. A better understanding of the subject will have theoretical as well as practical consequences, as the altered dispersion of the char has a drastic effect on the fishing. The experi­ ments undertaken were carried out in the lake reservoirs Tunhovdfjord and Pålsbufjord, which have long served investigations concerning the effects of impoundments on different aspects of fishing. II. The lakes The oligotrophic lakes Pålsbufjord and Tunhovdfjord are situated in the upper course of the River Numedalslågen, which drains the north-eastern parts of the Hardangervidda mountain area. Their position appears from Fig. 1. Another big tributary of Pålsbufjord is the River Skurdalselv. Into Tunhovdfjord runs the River Rodungselv, which originates in the reservoir Lake Rodungen. The natural catchment area of Tunhovdfjord is 1,809 sq. km, and the mean flow 42 m3/sec. (Fossedirektören 1932). In 1968 three BJ 0R N SR U D 6 PER AASS Fig. 1. The lake reservoirs Tunhovdfjord and Pâlsbufjord and their position. The two out­ lines indicate areas at maximum and minimum regulated levels. Place names of interest in connection with char winter migrations. THE WINTER MIGRATIONS OF CHAR 7 Table 1. Tunhovdfjord and Pålsbufjord. Particulars of impoundments. Start of impound­ ment Damming | Lowering in metres Annual amplitude in metres Regulated levels metres above sea level Areas in sq.km Mini­ mum Maxi­ mum Mini­ mum Maxi­ mum Tunhovdfjord 1919 18.00 18.00 718.00 736.00 14.25 25.35 Pålsbufjord 1927 — 9.00 9.00 — — — — 1946 12.50 — 21.50 — ■--- — — 1958 — 12.00 24.50 726.00 750.50 5.25 19.50 small rivers running to the east and west of the lake were transferred to Tunhovdfjord by means of tunnels, the drainage area being thereby increased by 271 sq. km. Originally Tunhovdfjord consisted of one big and two small basins, con­ nected by short rapids. In 1919—20 the level of the lake was raised 18 metres above normal level, and in order to drain the reservoir completely the rapids were canalized, thus forming narrow sounds at low water. Pålsbufjord, situated immediately above, was lowered by 9 metres below normal level in 1927—28. Later the lowering was increased and the lake level also raised by damming. Since 1958 the maximum amplitude has been 24.5 metres. Particulars of the impoundments will be found in Table 1. The reservoirs are filled mostly in spring and summer, and the water levels are relatively stable until the lowering starts in late autumn or early winter. Both res­ ervoirs are emptied by tunnels situated close to their natural outlets and they are at their lowest in late April (Fig. 2). Tunhovdfjord serves as intake TUNHOVDFJORD PÅLSBUFJORD MAX.REG.LEVEL NORMAL & MIN.REG.LEVEL 750 MAX.REG.LEVEL 746 742 -738-NORMAL LEVEL 734 - 730 -726-MIN.REG. LEVEL 722 H 718 S 1 0 1 N 1 D Fig. 2. Water-level fluctuations of Tunhovdfjord and Pålsbufjord 1963. The lowering of Pålsbufjord in June—July is due to draining to facilitate floating. 8 PER AASS reservoir for the Nore power plant. The section of the River Numedalslågen connecting the two lakes is about 400 metres long when Tunhovdfjord is at its highest level, but lengthens to about 1,800 metres when the reservoir reaches low-water level. In winter the upper half of the riverbed is laid dry, the water running instead through the tunnel which empties in the middle of the course. Only when the reservoir in Pålsbufjord has reached its original level can water run through the whole section. Winter comes early, and shallow bays freeze in the middle of November. The main basins are mostly covered with ice from December to the first half of May. Brown trout (Salmo trutta L.) was for a long time the only fish species present in both lakes. About 1915 minnow (Phoxinus phoxinus L.) was in­ troduced by fishermen who used this species as living bait. In summer 1910 char fry was planted in a mountain lake that drained into the River Nume­ dalslågen, about 35 kilometres upstream from Pålsbufjord. Char was first caught in this lake in 1919 (Dahl 1920). No records exist of the first appearance of char in Tunhovdfjord, but probably they entered this lake at about the same time. That is at the start of impoundment of Tunhovd­ fjord. In Pålsbufjord the stock was of considerable size when this lake was impounded. In both lakes char is now the predominant species, constituting about 85 per cent of the total yearly catch in weight. III. Matlierial and methods During the years 1951 and 1954—68, in all 21,073 char were tagged in the two lakes and the connecting river. Of these 10,475 were tagged in winter, 8,714 in autumn and 1,884 in spring and summer. For details of number and tagging places, see Table 2 and Fig. 1. A small home-made celluloid disc tag was used in 1951, and Lea’s hy­ drostatic tag in 1954—55. During the winter taggings in 1956—58, difficulties over the supply of Lea tags led to comparisons with Carlin tags (Carlin 1955) and modifications of this type. Tags with and without an inserted wire between the tag and the attachment were tried, and some tags were also strengthened by a steel thread running along the tag. No significant difference in recapture was found between the Lea tag and the modified Carlin tag without an intermediate link. Carlin tags with an inserted wire gave a high first-year recapture, owing to their high tendency to fasten in nets. Reinforcing the tags did not increase the rate of recapture over the years. Different kinds of thread were also compared, viz. monofilament nylon, stainless steel and surgical steel. Nylon attachment gave the highest recapture rate, because of the greater tendency of the tags to fasten in nets. With the rather stiff nylon it was difficult to get the tag as close to the THE WINTER MIGRATIONS OF CHAR 9 Table 2. Numedalslågen watercourse. Tagging of char 1951—68. Season Place tagged Number Pålsbufjord Summer 1951 Astveit 42 Autumn 1955—64 Åsodden 2,913„ 1965—68 Bjornsrud 1,211 Winter 1951, 1954—64 Âsodden 1,179 n 1951, 1958—64 Near tunnel intake 1,083 « 1964 Astveit 84 Tunhovdfjord Autumn 1955—68 Tunhovdvik 3,828 » 1965 Bustrom 237n 1961—64 Transferred from Tunhovdvik to different places in the lakes 525 Winter 1955—65 Near Tunhovd Dam 3,735„ 1957—65 Bustrom 3,142 » 1957, 1963 Tunhovdvik 495 » 1961—62 Stornesvik 400„ 1961 Gravikstrom 107 » 1964 Turrsnippen 250 Numedalslågen Spring—summer 1957—61, 1965—66, 1968 1,842 body as it was with steel thread. From 1959 onwards, only Carlin tags without an inserted wire have been used. They are attached with double 0.3 mm stainless steel wire passed through the back of the fish in front of the dorsal fin. Anaesthetics are used in spring, summer and autumn, but not in winter. The tagging places were selected with the intention of obtaining informa­ tion on the extent and range of the movements and their importance to the fishing. They are rather few in number, because it was considered that series of long duration and populations that were to some extent homoge­ neous would show the general pattern and possible changes better than would many short-term experiments scattered along the lakes. The tagging was performed in the periods when it is possible to catch char in large numbers, mainly autumn and winter. During spawning in October char was caught with gill nets. Fishing depths were about 15—20 metres in Tunhovdfjord and 30 metres in Pålsbufjord. The nets were regularly set in the afternoon and lifted early in the morning, but occasionally the occurrence of a gale meant that the nets had to stand for two or more nights. When the nets were lifted, the fish were removed one by one and placed in water, either in tubs or in the half-filled boat. Fresh water was constantly supplied. The procedure could take up to 2—3 hours, especially in windy or snowy weather, but the fish evidently took little harm. After the fish had been transported to the shore, immature and resting mature specimens and all fish with poor scales or floating belly up 10 PER AASS were discarded. The rest were placed in immersed wooden boxes and kept for control for about a week. During the stay about 1—2 per cent of the fish died. To avoid immediate mass recapture, tagging and releasing took place when most of the commercial fishing had ended. The fish were placed in shallow water and allowed to make for deeper water by themselves. All fish in bad condition were killed. Most of the fish tagged were males. Their well-attached scales prevent scratches and the slender and slimy bodies do not stick firmly to the nets. The females mostly get firmly entangled and stand the pressure of the meshes poorly. The slime secretion of females is less than for males and the scales are more easily lost. Most die on the nets or immediately after lifting; the rest are subject to the attack of fungi in the scars. Of 1,159 male spawners tagged in the two lakes in the years 1961—66 a total of 46 per cent were recaptured, not counting returns from the spawning places shortly after tagging. The corresponding figures for females were 285 and 23 per cent. The inferiority of the females is possibly due rather to reduced tolerance to netting than to different behaviour after release. In the winter fishing, which accounts for the bulk of the total catch, the sex ratio is very nearly 50—50. Both sexes have served to illustrate the pattern of movements, but in estimating the extent of the migrations only males have been used. Non-spawning char are also very difficult to catch or keep alive when caught on nets. The scales tend to get lost, the bodies stiffen and most of the fish die before the fungi have made any visible impact. This category was never used in tagging experiments in autumn. The char tagged in winter were all caught from the ice with spoons baited with prawn, worm or artificial flies. Both sexes were tagged, as in winter they may be difficult to separate by external characters and their bodies are equally slimy and unhurt by capture. The fish were tagged immediately, without being placed in water, which in the low temperatures soon becomes undercooled and is plainly detrimental to the char. They were rushed to the tagging table, a portable wooden box warmed by a primus stove. Sometimes a tent was used for the comfort of the tagger, but the time taken to run to a permanent tagging centre might easily be too long. The climatic conditions were often unfavourable, and the technical difficulties could be considerable. Tagging was performed in temperatures mostly be­ tween minus 10—20°C, occasionally as low as minus 34°C, and sometimes gales tore the tent. After tagging, the char was put through a hole in the ice. Apparently the fish took no harm from the cold, even if they became glaciated during the short stay in the air. Of 4,258 char tagged in Tunhovd- fjord during the winters 1961—65 the number recaptured was 1,971 or 46.3 per cent. The corresponding figures for males tagged in the autumn of the same years were 830 and 398 or 48.0 per cent. In shallow water tagged char could be observed through holes in the ice. They behaved like untagged THE WINTER MIGRATIONS OF CHAR 11 ones, and very soon became interested in the bait again. The shortest time recorded between tagging and recapture was a quarter of an hour. In the river char was caught in spring and summer with baited spoons and kept in running water for control before tagging. In this case, too, both sexes were used. When the tagging started, the short-term positive effects of the impound­ ments (Dahl 1926 & 1932) had expired (Huitfeldt-Kaas 1935, Aass 1957) and the lakes were crowded with the dwarfed char characteristic of many reservoirs. Preliminary experiments showed that fish shorter than 18 cm gave significantly lower returns than did bigger fish and were mostly caught in nets although there was also extensive fishing with hooks. The char tagged in Pålsbufjord were mostly of the size 22—28 cm and in Tunhovd- fjord 20—25 cm. The age has varied considerably according to fluctuations in year-class strength. The tags carried an inscription requesting the fishermen to report recap­ ture, a reward being promised. From the start of the experiment the fishing public were also informed by posters, letters and personal communications. During the season controllers were stationed at the main fishing places and arrangements were also made for the handing in of tags at local shops. Twice a year, after the end of the summer and winter seasons, chalet owners were visited in search of tags and of information. Other persons known to have fished in the lakes were contacted by post. In spite of all these efforts, some tags were not handed in, but so far as we know their recording would not have altered the general picture of the migrations given in what follows. Their main trend is illustrated by a few characteristic years, and the mass of data in evidence of the description is not included. Exceptions from the general pattern are, however, dealt with. To supplement the tagging experiments with fishing statistics, the controllers recorded the catches daily or three times a week. The lakes and river are rather heavily exploited; fishing efforts, measured as daily visits, amount to about 12,000 a year. The winter fishing with spoon attracts most fishermen, but char are fished the year round, nets and fishing otter also being important gear. There is an active search for fish and concentrations of char are soon discovered. Their movements help to enable an impression of the migrations to be formed. IV. Winter migrations in Lake Tunhovdf jord Early winter movements Char for the autumn taggings have chiefly been caught at the spawning places all round Tunhovdvik at the upper end of the lake. When released the fish have returned quickly to the original site of capture. After spawning 12 PER AASS is finished in the second half of October, the concentrations of char at the spawning places slowly disperse. The first recaptures of spent fish indicate that the movements immediately following spawning are rather short, though a few stray fish are caught far away. The char move in great numbers into the upper littoral from the spawning areas along the shore or on shallows out in the lake. Sheltered coves with muddy bottom are preferred. When the bays and coves freeze up in the middle of November, a greater number of mature fish are found in shallow water than at any other time of the year. To illustrate the accumulation, the fishery in Bergevik, a little bay with an area of about 4—5 hectares and a maximum depth of about 2 metres, may be mentioned. The catch here, made in few weeks, may amount to 3,000 char. The proportion of spent fish heading for the littoral cannot be determined, as the ice conditions restrict fishing. The concentration of old fish in the upper few metres is of short duration. Most withdraw to deeper water from the middle of December. This very often coincides with a decline in the penetrability of the ice by light, caused either by a cover of snow or by growing thickness. Simultaneously with the withdrawal from shallow water, char begin to accumulate just above the tunnel intake near the Tunhovd Dam. Ice fishing on a small scale starts here in the second half of December. The catch gradually increases, generally attaining its maximum during the last days of February. Thereafter the yield decreases rather rapidly, but in most years fishing continues until the end of April (Table 3). Owing to changes in ice conditions the fishing area shifts a few hundred metres during the winter, but it never measures more than about 100X150 metres. It is evident that char must be moving into this and the other main winter fishing places, all situated in the southern part of the lake. Most probably the fish that appear first come mainly from localities nearby, but only one tagging experiment directly showing this has been carried out. In this region spawning places are small and scattered and the char are difficult to come by in autumn. Spawners tagged in 1965 at Bustrom, close to the fishing places, were recaptured earlier in winter than were char tagged in more Table 3. Tunhovdfjord. Estimated numbers of char caught at Tunhovd Dam in winters 1960/61—1966/67. Year December January February March April May Total 1960/61 ..................... 500 3,600 10,700 7,000 1,450 23,250 1961/62 ..................... 750 4,750 6.900 4,300 1,550 100 18,350 1962/63 ..................... 100 2,900 4,500 4,900 1,600 — 14,000 1963/64 ..................... 900 2,050 1,300 500 200 — 4.950 1964/65 ..................... 800 1,900 5,000 4,300 500 — 12.500 1965/66 ..................... 1,400 7,500 7,300 10,250 900 50 27,400 1966/67 ..................... 1,600 5,100 10,200 6,400 1,000 — 24,300 THE WINTER MIGRATIONS OF CHAR 13 Fig. 3. Recaptures in winter 1966 of char tagged as spawners at Bustrom October 1965. On all maps arrows indicate tagging place and a dot denotes one recapture. The recap­ tures are more concentrated than the figures indicate. distant parts. The rate of winter recapture was high, hut all the 1966 recap­ tures but one were made in the same part of the lake, indicating that the southern spawning populations move regularly in winter though not very far (Fig. 3). Through winter tagging it may also be possible to trace the region where the mature fish start their migration. The homing at their spawning areas is rather strong in the case of Tunhovdfjord char. Fish tagged near the Tunhovd Dam in January are recaptured as spawners mostly in the lower third of the lake, an indirect proof that the majority of the early entries belong to local populations. Downstream migration The growth in the catch at the Tunhovd Dam during midwinter is accompanied by an increased recapture of fish tagged as spawners at Tunhovdvik at the opposite end of the lake. From the upper part of Table 4, it will be seen that the recaptures are made mainly in February and March, the two months which account for the greater part of the total catch. Apparently char belonging to the spawning populations of the northern and largest basin of the lake constitute an increasing — possibly a predomi­ nant — part of the fish concentration above the dam as the winter proceeds. 14 PER AASS Table 4. Tunhovdfjord. Winter recaptures 1960—67 at Tunhovd Dam, Bustrom and Turrsnippen of char tagged as spawners at Tunhovdvik. Year Jan. Feb. March April May Total 1960 ............... Tunhovd Dam 2 5 3 1 11 1961 ............... 2 16 11 3 — 32 1962 ............... 1 11 11 3 — 26 1963 ............... 4 3 5 2 — 14 1964 ............... 1 2 — — — 3 1965 ............... — 2 3 — — 5 1966 ............... 1 1 7 1 — 10 1967 ............... 3 7 5 2 — 17 12 45 47 14 1 118 1960 ............... 4 2 Bustrom 22 16 1 45 1961 ............... 1 6 18 23 1 49 1962 ............... ■--- 4 17 17 — 38 1963 ............... — 1 43 8 — 52 1964 ............. 10 14 5 4 — 33 1965 ............... — 8 13 2 — 23 1966 ............... — 3 10 7 --- - 20 1967 ............... — 9 15 4 4 32 15 47 143 81 6 292 1960 ............. Turrsnippen — 2 2 1961 ............. — — — 1 — 1 1962 ............... — 1 — — — 1 1963 ............. — — — 1 — 1 1964 ............. — 5 13 4 — 22 1965 ............. — — 1 — — 1 1966 ............. 1 — 2 — — 3 1967 ............. 1 — 1 — — 2 2 6 19 6 — 33 The facts suggest that a migration from north to south takes place regularly every winter. The decreasing yield from the beginning of March is due to hydrological conditions and not to a reduction in fishing efforts. The fisher­ men are rather conservative and do not move to another fishing place until the downward trend is well established. The place accounting for the greatest winter catch is situated at Bustrom, about 6 kilometres above the Tunhovd Dam. Fishing starts as soon as the lake is covered, but during the first months the yield is inconsiderable. Normally a sharp increase in the catch occurs in late February and the daily maximum is obtained in the second half of March. April may also provide good fishing, but by the time the ice breaks in May the yield has for some time been insignificant (Table 5). The interdependence of the fishing at the Tunhovd Dam and at Bustrom is striking. When the yield starts to decline at the Tunhovd Dam, it improves THE WINTER MIGRATIONS OF CHAR 15 Table 5. Tunhovdfjord. Estimated numbers of char caught at Bustrom in winters 1960/61—1966/67. Year Dec. Jan. Feb. March April May 1960/61 ........... ___ 600 2,500 7,700 13,600 9,100 300 1961/62 ........... 550 1,050 3,200 9,000 8,800 500 1962/63 ........... ___ 150 600 2,150 10,300 3,100 300 1963/64 ........... 500 4,750 7,850 4,100 1,900 100 1964/65 ........... 500 1,500 7,500 14,400 3,200 1,000 1965/66 ........... 100 1,800 10,300 19,800 6,000 300 1966/67 ........... 600 2,000 15,000 15,300 8,300 2,300 Total 33,800 23.100 16,600 18.300 28.100 38.300 43,500 \ Fig. 4. Upper part. Number of char caught during winter 1961/62 at Tunhovd Dam and Bustrom. Lower part. Number of char caught during winter 1963/64 at Tunhovd Dam, Bustrom and Turrsnippen. The catches of two days are combined. Tunhovd Dam, ------ Bustrom,------ Turrsnippen. ------ 16 PER AASS Fig. 5. Winter recaptures 1962 (left) and 1963 (right) of char tagged as spawners at Tunhovdvik, October 1961 and 1962. greatly at Bustrom (upper part of Fig. 4). The large number of recaptures of fish tagged in the north makes it evident that the fishery at Bustrom, too, is largely maintained by char migrating southwards, and consequently the fish caught at the two places must in part belong to the same spawning populations. Fig. 5 shows the winter recaptures in 1962 and 1963 of spawners tagged at Tunhovdvik. In contrast to the Bustrom tagging, Fig. 3, the recap­ ture at the tagging place is insignificant. During the fishing season the situation at the Tunhovd Dam and Bustrom is characterized by a water velocity that is above the average for the res­ ervoir. The currents formed at these places attract the char and the resul­ ting concentrations form the basis of the rich fishing which develops. At high water early in winter the char can pass unimpeded from north to south. There are no currents strong enough to stop the fish, except at the intake near the dam. Some fish may interrupt their downward movement at Bustrom, but both the fishing and the number of recaptures indicate that normally the char pass on. Tagging shows that many char that stop at Bustrom in January and most of February are only halting on their way THE WINTER MIGRATIONS OF CHAR 17 Fig. 6. Winter recaptures of char tagged at Bustrom in January and March 1961 (upper and lower left) and 1965 (upper and lower right). Only returns from the tagging winters are marked off. to the dam (upper part of Fig. 6). The proportion of fish permanently remaining at Bustrom depends on the height of the water, and this is why the tagging experiments may give slightly different results in different years, e.g. 1961 and 1965. When the reservoir is lowered during the winter a current is formed at Bustrom, which is one of the old rapids transformed by canalization into a narrow strait at low water. In most years the current is observable from 2 18 PER AASS Fig. 7. Winter recaptures of char tagged at Tunhovd Dam in March 1962 (left) and 1963 (right). Returns the same winters. the end of February. Tagging experiments at Bustrom in March show the char to be strongly attached to the current (lower part of Fig. 6). Only a few stray fish move with the water to the dam, and char returning to the main basin late in winter have never been recorded. The fish standing above the Tunhovd Dam show the same preference for the current, although a few stray char have been recaught later in the winter in currents higher up in the lake (Fig. 7). Another of the former rapids, Gravikstrom, is situated between Bustrom and the Tunhovd Dam. The current here is less concentrated, and the tendency of the char to stop is not as great as at the places mentioned above. A small fishery develops in middle of winter, just when the decrease sets in at the Tunhovd Dam, and dwindles when the fishing at Bustrom reaches its peak. The yearly catch may amount to 3,000—3,500 char. Returns from the Tunhovdvik taggings prove the fish to be partly of northern origin. A tagging experiment carried out in March 1960 showed a winter movement towards the dam. Recaptures from outside the tagging area were all made near the tunnel intake (Fig. 8). THE WINTER MIGRATIONS OF CHAR 19 Fig. 8. Winter recaptures of char tagged at Gravikstrom in March 1960. Returns the same winter. Water level and char concentration The tendency of the char to make a permanent stop at Bustrom in the latter half of the winter must be the reason why the catch decreases drasti­ cally in the lake further south. Supply from the northern basin is severely reduced and the local populations are small and have been exploited through the whole season. Thanks to the close connection between water level and fishing output at Bustrom, it is possible to predict when the change will occur. Usually the abrupt increase in catch starts when the reservoir is down 4—5 metres, but in some years the reaction is delayed until the lowering has reached 7—8 metres. The best fishing takes place when the reservoir is 8—10 metres below highest regulated level. This pattern is reflected in the recapture of northern spawners. Apart from a few stray individuals, tagged fish enter the catches when the reservoir is lowered 4 metres and most recaptures are made when the water level is down by 8—10 metres. Fig. 9 shows the relation between water level and recaptures at Bustrom in the winters 1956—68. The increase in both catch and recaptures is entirely dependent on the height of the water level, not on the time. This is best illustrated by the events in 1964. The reservoir was drained exceptionally early and rapidly to enable the Tunhovd Dam to be rebuilt, and the water level normal for 20 PER AASS 735 731 Ml N.REG.LEVELMAX.REG. LEVEL RES.DOWN 0 4 8 12 16 METRES Fig. 9. Relation between water level and number of returns made at Bustrom during the winters of 1956—68. The char were tagged as spawners at Tunhovdvik. March—April was reached in the first half of February (Fig. 10). This pul forward the catch at Bustrom of char tagged in the north by more than a month, the majority being caught in January and February against March and April in an ordinary year (middle of Tabled). At the Tunhovd Dam only three northern spawners were recaptured, and the proportion between recaptures at the two places was 1: 10, compared with an average of 1:3. Obviously very few fish got through to the dam in the winter of 1964. This is also reflected in the catch above the dam. Besides reaching its peak period as early as January, it was exceptionally small compared with the winter total for the lake. Comparisons between the numbers caught in the different years should be made with caution, in view of the varying strength of the year classes. The winter fishing at Bustrom also started exceptionally early in 1964 (Table 5). As only a fraction of the migrating population succeeded in getting down to the dam, a record catch was anticipated, but the expecta­ tions were not fulfilled. After a short peak period in February, which as usual coincided with a draw-down of 8—12 metres, the daily catches deteriorated in late winter as never before. As a result, the total winter catch at Bustrom was well below the average for the period 1960/61— 1966/67. Without doubt the unexpected outcome was due to a fishex-y THE WINTER MIGRATIONS OF CHAR 21 1960-61 1961 -62 1962-63 1963 - 64 1964 - 65 1965- 66 1966 - 67 Fig. 10. Winter water-level fluctuations in Tunhovdfjord 1960/61—1966/67. Vertical lines indicate lowering of the reservoir by 5, 10 and 15 metres. developing at a place called Turrsnippen, situated 5—6 kilometres to the north of Bustrom. Here the lake narrows, and as a result of the great low- water flow a current was formed earlier in the season than usual. In previous winters the fishing at this place had been insignificant, the total catches amounting to only a few hundred fish. But now the char concen­ trated in great numbers in the strait. In consequence many fishermen were attracted, and the total catch in 1964 was estimated at about 13,000 char (Table 6). Many fish tagged as spawners in Tunhovdvik were recaptured 22 PER AASS Table 6. Tunhovdfjord. Estimated numbers of char caught at Turrsnippen in winters 1960/61—1966/67. Year Dec. Jan. Feb. March April May Total 1960/61 ................... 100 100 200 1961/62 ................... — — — 100 300 100 500 1962/63 ................... • --- — — — 250 — 250 1963/64 ................... — 200 3,100 7,400 2,300 — 13,000 1964/65 ................... — — 200 2,900 2,100 — 5,200 1965/66 ................... — — — 1,500 3,000 300 4,800 1966/67 ................... — — — 700 1,600 200 2,500 at Turrsnippen (lower part of Table 4). Winter recaptures for 1964 are indicated in Fig. 11, and the distribution should be compared with winters with ordinary draining, such as 1961/62—1962/63, in Fig. 5. Undoubtly a great proportion of the char accumulating at Turrsnippen in 1964 belonged to the northern spawning populations, which normally concentrate near the dam and at Bustrom. Once halted, the char behaved just as at the other currents and stayed on for the rest of the winter, only a minority moving further down the lake (Fig. 12). Once more the interdependence of the fishing places is demonstrated. Just as the concentration of fish at Bustrom Fig. 11. Winter recaptures 1964 of char tagged as spawners at Tun- hovdvik October 1963. THE WINTER MIGRATIONS OF CHAR 23 starts the decline in the fishing at the Tunhovd Dam, so the accumulation at Turrsnippen marks the turning point for the Bustrom fishery (lower part of Fig. 4). In the winter of 1965 the draining followed the traditional pattern until the middle of January, when the reservoir was lowered faster than ever before (Fig. 10). Near the Tunhovd Dam the fishing started and developed as usual, the draining apparently having occurred too late to stop the early migrants arriving from Bustrom. Then followed a quick deterioration, and though the yield at Bustrom did not suffer to such a degree as in 1964, the distribution of recaptures indicated that the char stopped at Bustrom rather early in the season. Once again a considerable fishery developed at Turrsnippen, but this time at a later date and the total catch was smaller than the year before. Since the rebuilding of the dam, the draining has followed the normal pattern. Owing mainly to a great increase in fishing intensity, aroused by the two good years, the catch at Turrsnippen still exceeds the earlier years. There is, however, a distinct downward trend. The facts mentioned suggest a close connection between currents and accumulations of char, and this is confirmed by a further circumstance. The local fishermen complain of a regular reduction in catches during Saturdays and Sundays, when draining is reduced because industries need Fig. 12. Winter recaptures 1964 of char tagged at Turrsnippen in March 1964. 24 PER AASS Table 7. Tunhovdfjord. Mean number of char caught per hour winters 1960—63. Experimental angling at Bustrom. Period Number of hours Mean number, Mean number, weekdays weekends 22/2—31/3 1960 ............. 354 9.0 7.7 20/2—29/3 1961 ............. 326 6.5 4.9 21/2—30/3 1962 ............. 374 4.4 3.5 22/2—29/3 1963 ............. 349 3.4 3.3 less power. For many years a control fishing with spoon was carried out at Bustrom six days a week, but only in the years 1960—63 were Sundays included. Every year the same two anglers were employed, the fishing duration, time of day, and locality being kept unchanged. The fishing results during the peak period when according to taggings the char make a permanent stop are outlined in Table 7. The catches per effort are signi­ ficantly smaller during weekends, thus bearing out the fishermen’s complaint. The reduction may be due to an increased total fishing effort, since during weekends more anglers take part than on weekdays. But the holiday fishers are of course less efficient than the local fishermen, as they are inexperi­ enced and fish shorter hours. The fishing area is adequate, and it is doubtful whether the weekend influx had any effect on the control fishery. If it did not, the results suggest a weaker affinity between fish and current during weekends. Apparently the fish do not leave the current; had that been the case char tagged at Bustrom in March would have been recaptured all round the lake the same winter, and this was not the case. When the water velocity is reduced, the char possibly disperse more freely in the water volume. During heavy draining the char tend to concentrate at a depth of about 8—10 metres, well above the bottom. Speed of migration The slow increase in the number of char caught, tagged and untagged, in the currents during early winter, imply that the char do not set out in one big shoal. The migration takes place over a period of several months, with maximum arrival in the southern region about early March. It is more difficult to establish when the bulk of the char leave the northern basin. Winter recaptures in Tunhovdvik of spawners tagged at the place in autumn are too few to show any trend, and for that purpose winter tagging was carried out on two occasions. In April 1957 and February/March 1963, respectively, 255 and 240 char were tagged. In all, 22 were recaptured during the winter they were tagged, but in each case close to the tagging THE WINTER MIGRATIONS OF CHAR 25 locality. The majority of the migrating fish had probably left the shores already, i.e. not later than the middle of February. This means that the fish take at least 2—3 weeks on their way to the southern region, a linear distance of 15—20 kilometres. Another indication of the travelling speed is given by the changing yield in the southern part of the lake. When all the moving char stop at Bustrom, it takes barely a week before the effect is felt at the Tunhovd Dam. The distance between the places is 5—6 kilo­ metres. At Stornesvik on the western shore near the middle of the lake, 200 char were tagged both in February 1961 and February 1962; 24 were recovered near the tagging place, in most cases fairly soon after being released, 3 had crossed the lake and 8 were recaptured at Bustrom, 4—9 weeks after tagging. The small number of recaptures suggests that at this place, too, only a minority of the migrating fish had not left by the middle of February. Size of migrating char The existence of large-scale winter migrations can be demonstrated with certainty only for the tagged size group, i.e. fish measuring 20 centimetres or more. If the migrating population consisted of a well-defined size group, its departure and arrival could be observed by analysing the local catches, provided that the moving fish regularly entered into the catches. Regular length measurements of catches during the winter season were made at the various currents between 1955 and 1968. The results differ little from place to place and from year to year, char sized 20—25 centimetres con­ stituting the bulk. The percentage length distribution at Bustrom, where in all 82,610 char were measured, is listed in Table 8. To judge from this distribution, it is mainly the bigger fish that move into the currents. If this is the case, their departure should be observable as a reduction of this size group in the remaining population. Measurements made in March 1963 at Tunhovdvik and Bustrom — Tunhovd Dam show a distinct difference in size distribution in north and south (Table 9). Small fish predominate in the northern bay, bigger ones in the currents. The result may be due to an uneven distribution of young fish belonging to the rich year class 1960. The investigation was repeated in 1966—68, when in all 2,499 char were measured in Tunhovdvik. In these years there was no significant difference in the mean length of northern and southern catches. This would mean that fish of all sizes left the northern basin. The selectivity of the gear, however, renders an interpretation of the result difficult. The small fish are not afraid of the spoon — on the contrary it is often difficult to avoid catching large numbers. But it can be clearly observed through the ice that the big fish dominate the small ones, which have little chance of being hooked until their superiors are reduced in number. As long as there exists 26 PER AASS Table 8. Tunhovdf jord. Length distribution in per cent of char caught at Bu- strom in winters 1955—68. 195.-) 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 Number measured . . 345 1,218 2,270 2,921 2,611 22,998 25,534 5,577 3,308 3,164 4,972 3,023 2,408 2,261 10—10.9 cm 0.1 11—11.9 „ — — — — — — — — 0.2 — — 0.1 — 0.4 12—12.9 „ — 0.2 0.1 — — 0.1 — 1.5 1.0 0.2 0.6 0.2 0.3 1.0 13—13.9 „ 0.3 0.2 0.3 0.2 0.1 0.2 0.1 0.9 1.4 0.5 1.0 0.7 0.9 1.2 14—14.9 „ 0.3 1.3 0.4 0.4 0.4 2.0 0.3 1.5 2.8 0.9 1.8 1.0 0.9 0.8 15—15.9 „ 0.9 2.7 2.6 1.0 1.0 4.3 0.6 1.7 3.1 1.5 3.2 1.5 1.2 1.3 16—16.9 , 2.3 6.7 7.0 1.6 1.4 4.1 1.0 1.8 3.5 2.3 7.2 4.2 1.7 2.6 17—17.9 „ 4.4 7.3 11.9 4.1 3.8 3.7 2.4 1.8 3.6 4.0 8.2 8.4 4.6 4.4 18—18.9 „ 4.4 8.8 10.5 8.3 5.8 3.7 2.9 1.7 3.1 6.3 13.5 9.8 7.9 7.2 19—19.9 „ 10.7 5.9 10.6 16.8 12.7 6.7 5.6 1.4 2.2 7.9 9.6 12.0 12.3 8.7 20—20.9 „ 6.1 6.6 11.6 23.1 21.9 13.2 8.8 1.5 2.2 8.8 10.2 13.6 16.2 11.5 21—21.9 „ 4.4 8.7 8.9 15.5 27.5 21.3 12.0 3.2 3.1 8.9 10.1 15.6 20.3 19.8 22—22.9 „ 4.1 9.0 8.2 11.1 17.1 25.8 25.4 13.0 4.8 10.6 10.6 13.2 18.1 20.3 23—23.9 „ 7.8 10.2 6.8 7.7 5.4 11.1 24.0 31.9 12.3 10.9 10.5 7.2 9.3 13.7 24—24.9 „ 12.6 8.6 6.8 4.4 1.3 3.2 11.6 26.1 26.7 13.1 7.2 6.2 3.7 4.7 25—25.9 „ 10.1 9.0 6.7 2.8 0.9 0.5 3.7 10.0 19.1 12.1 3.5 3.8 1.9 1.7 26—26.9 „ 13.3 6.7 3.5 1.7 0.4 0.1 1.1 1.6 8.0 7.4 1.9 1.5 0.5 0.4 27—27.9 „ 8.7 4.7 2.2 1.0 0.2 — 0.3 0.3 2.2 3.4 0.7 0.8 0.2 0.2 28—28.9 „ 5.8 1.7 1.4 0.3 0.1 — 0.2 0.1 0.6 0.9 0.2 0.2 — — 29—29.9 „ 2.6 1.1 0.5 — — — — — 0.1 0.2 — — — — 30—30.9 „ 1.2 0.4 — __ — — — — — 0.1 — — — — 31—31.9 „ — 0.2 — — — — — — — — — — — — Table 9. Tunhovdfjord. Length distribution in per cent of char caught by angling in March 1963. Tunhovd- vika Bustrom Tunhovd Dam Number measured ... . 907 3,308 1,025 10--10.9 .......................................... 0.8 — — 11--11.9 .......................................... 1.8 0.2 0.2 12--12.9 .......................................... 5.3 1.0 1.8 13--13.9 .......................................... 4.8 1.4 1.8 14--14.9 .......................................... 5.5 2.8 2.0 15--15.9 .......................................... 12.1 3.1 3.4 16--16.9 .......................................... 12.8 3.5 4.4 17--17.9 .......................................... 11.5 3.6 4.0 18--18.9 .......................................... 10.2 3.1 2.4 19--19.9 .......................................... 7.8 2.2 2.4 20--20.9 .......................................... 8.6 2.2 3.3 21--21.9 .......................................... 3.1 3.1 2.9 22--22.9 .......................................... 3.6 4.8 6.8 23--23.9 .......................................... 4.1 12.3 13.1 24--24.9 .......................................... 3.7 26.7 27.9 25--25.9 .......................................... 2.1 19.1 14.4 26--26.9 .......................................... 1.5 8.0 6.5 27--27.9 .......................................... 0.7 2.2 2.0 28--28.9 .......................................... — 0.6 0.4 29--29.9 .......................................... — 0.1 0.2 30--30.9 .......................................... — — 0.1 THE WINTER MIGRATIONS OF CHAR 27 a sufficient supply of big fish, these will constitute the greater part of the catch, thus concealing the presence of a supernumerary population of small fish. If this kind of selection occurs in the currents, the proportion of small fish ought to increase in late winter when the number of big fish is supposed to have been reduced by fishing. A test of catches made at Bustrom in 1960—63, reveals no significant difference in mean length from January to April. But there is a slight tendency for smaller fish of 2—3 years age to be caught in April. The result does not imply a great influx of small fish to the currents, but the possibility of a winter movement directed towards the outlet cannot be excluded. In that case the young fish occupy the bottom, not being able to stay in the stream. The winter movements of the young char have not yet been explained. The extent of migration With few exceptions, the char accumulating in the currents must have moved into these from other places. To what extent the different populations migrate certainly depends on the situation of the spawning grounds in relation to the currents. The movements of the norhernmost spawning populations are best known, and it is evident from the distribution of the recaptures that a high proportion of the population must leave for the currents in winter. But owing to the variable arrival of the fish during winter and the division of the migration population between several places, it is difficult to make a quantitative estimate of the stock involved. Table 10 shows the percentage winter recapture in the Bustrom—Tunhovd Dam region of spawners tagged in Tunhovdvik during the preceding autumn. Immediate recaptures are deducted. The average for the years 1959—65, Table 10. Tunhovdfjord. Winter recapture in the Bustrom—Tunhovd Dam region of spawners tagged at Tunhovdvika the preceding autumn. Males only, with the exception of the 1958—60 material, when a few females were included. Recaptures made before the winter season have been deducted. Number Autumn of tagging tagged Winter of Percentage(after deduction) recapture recapture 1958 .................................. 306 1959 8.8 1959 .................................. 498 1960 10.4 1960 .................................. 492 1961 8.9 1961 ................................... 130 1962 13.8 1962 .................................. 193 1963 23.3 1963 ................................... 147 1964 12.4 1964 .................................. 135 1965 9.0 1965 .................................. 174 1966 12.6 1966 .................................. 279 1967 11.5 28 PER AASS which are supposed to give the most reliable results, is 12.3. This is a rather high rate of exploitation, but spawning populations belonging to the currents or nearby localities are possibly more drastically reduced, being exploited throughout the season. Only once has it been possible to tag a greater number of spawners at Bustrom. Of the 237 tagged in autumn 1965, nearly one hundred were recaptured before the lake was frozen over. The remaining are supposed to have behaved and dispersed in just the same way as the early arrivals from the north. The winter recapture of these fish in 1966 in the Bustrom—Tunhovd Dam region was 26.8 per cent, compared with the 12.6 per cent recapture of fish tagged at Tunhovdvik in autumn 1965. The difference in the returns is assumed to be due to the northern group being only partly present. The proportion suggests that about 47 per cent of the tagged population had moved southwards this winter. For three years char were tagged and released at Bustrom early in January, just when fishing had started, and for several years in March at the time when maxi­ mum arrival from north was supposed to occur. The recapture percentages in the Bustrom—Tunhovd Dam region in the winter of tagging are listed in Table 11. When combined with the percentage recapture of fish tagged in Tunhovdvik given in Table 10, the proportion of fish migrating from Tunhovdvik is estimated to be from 40 to 47 per cent when fish released in January are made the basis of the estimate. The corresponding figures when the March taggings are used are 51 and 86 per cent. The results in 1964 are not included, the whole migration rhythm being then upset by the unusual draining. The discrepancy is due to the migrating population not being exploited to the same degree as the fish released at Bustrom. The majority of the migrators are present during only part of the season. The higher fishing mortality of the local fish tagged early in season results in an unduly low estimate of the proportion leaving the northern bay. Conversely, the fish released at peak arrival are less exploited than the migrating stock, part of which arrives earlier in the season. If this group Table 11. Tunhovdfjord. Winter recapture in the Bustrom—Tunhovd Dam region of char tagged at Bustrom the same winter. January March Year Number tagged Percentagerecapture Number tagged Percentage recapture 1959 1960 1961 1962 1963 1964 1965 — — 300 11.0 — •— 230 13.5 200 19.0 200 17.5 — — 250 16.0 — — 250 38.8 250 30.8 250 9.2 200 22.0 208 13.5 THE WINTER MIGRATIONS OF CHAR 29 is used as a basis, the resulting estimate will be too high. As a rough estimate the proportion of spawners leaving the upper end of the lake may perhaps be put at 50—60 per cent. Upstream migration For years the recaptures showed only a downstream winter migration, and the location of the fishing places confirmed this impression. Recently a winter fishery has come into existence at the upper end of Tunhovdfjord, below the tunnel outlet from Pålsbufjord. One year the catch amounted to about 3,000 char, and as a whole it is quite insignificant compared with the output at the other end of the lake. The fishing starts late in winter, after the downstream migration has begun, and the catch consists mostly of small fish. The outlet of the tunnel is situated only 2 kilometres from the tagging place in Tunhovdvik, yet recaptures here are few. In the winters 1965—67 only 3 out of 134 recaptures of spawners tagged at Tunhovdvik were made at this place, and the movement towards the tunnel outlet must be of small extent. It is unlikely that emigration from Pålsbufjord contri­ butes much to this fishery. No winter-tagged char from Pålsbufjord have been recaught the same season in Tunhovdfjord. A few fish tagged as spawners in Pålsbufjord in the preceding autumn have been recaptured in the upper end of Tunhovdfjord near the outlet in winter. Probably they emigrated through the bottom gates in the Pålsbu Dam already in the late autumn, and stayed in the pool below the outlet during the winter (Fig. 13). In most years Pålsbufjord is drained through the bottom gates until early winter, and the tunnel between the lakes is in use from the middle of January to May. As long as the water level in Tunhovdfjord exceeds 727.5 metres the fish can get at the outlet directly; at lower water they have to pass over a small rapid. The tunnel is about 1,300 metres long and the fall is 1:600. At full reservoir in Pålsbufjord the flow is 30 cubic metres/sec. Under these conditions the tunnel is impassable for char migrating upstream. When approaching minimum regulated level the lower end of Pålsbufjord is transformed into a river, which empties into the tunnel. The water flow is then reduced to about 5 cubic metres/sec. In this period char may have a chance to pass through the tunnel. Yet of a total of approximately 13,000 char tagged in Tunhovdfjord, immediate recaptures not deducted, only 5 have with certainty been recaptured in Pålsbufjord. In addition some cases not verified are reported. The upstream migration is insignificant, whether because of difficulties in entering or leaving the tunnel, or of the reduced capacity of the fish to move against the stream in winter. Char is, however, not reluctant to pass through tunnels upstream if the circumstances make it possible. The migration from Lake Breivatn to Lake Bonsvatn, county of Telemark, offers a striking example. The tunnel be- 30 PER AASS PÅLSBU MAX. REG. LEVEL NORMAL ------- MIN. REG.-------- WINTER FISHING PLACE TUNNEL PÅLSBU DAM WITH BOTTOM GATES REG LEVEL REG LEVEL NHOVDFJORD WINTER FISHING PLACE Fig. 13. Pålsbu Dam and the adjacent parts of Pålsbufjord and Tunhovdfjord. Pålsbufjord is drained through the bottom gates in the dam until approaching the old normal water level, then the tunnel is taken into use. tween the lakes is 15.3 kilometres long, with an average fall of 1: 4635 and a maximum water velocity in summer of 1 metre/sec. A mixed population of trout and stunted char occurred in Breivatn, while trout was the only species in Bonsvatn. The tunnel was brought into use in the winter of 1958 and during the following years many char were caught in Bonsvatn. The population is now well established and outnumbers the trout. V. Winter migrations in Lake Pålsbufjord The migration pattern is much the same as in Tunhovdfjord, but deviates in some details. After spawning, the spent char approaches shallow water, THE WINTER MIGRATIONS OF CHAR 31 but the stay may be short and the fish move to a greater extent. To attain maximum power production, the water level of Tunhovdfjord is kept as high as possible and the lowering of Pålsbufjord starts early, mostly in November (Fig. 14). The fish are continually compelled to change location, and in so doing some also move along the Pålsbu Dam. If the bottom gates are open, a downstream migration to Tunhovdfjord takes place, but it involves only a minority of the population. The shorebound movement gives 1960-61 1961 - 62 1962- 63 1963-64 1964-65 1965-66 1966 - 67 Fig. 14. Winter water-level fluctuations in Pålsbufjord 1960/61—1966/67. Vertical lines indi­ cate lowering of the reservoir by 10, 15 and 20 metres. 32 PER AASS Fig. 15. Winter recaptures following tagging experiments on spawning char at Asodden October 1958—60 and Bjornsrud October 1965—67. Upper and lower left, recaptures in the periods December—March and April—May after the Asodden taggings. Upper and lower right, recaptures in the same periods after tagging at Bjornsrud. rise to a fishery of short duration at depths of only 1—2 metres. In December the char return to deeper water, but the mobility is still rather high. Fig. 15 (upper part) shows the distribution of recaptures in December—March of fish tagged as spawners during the preceding autumn at Asodden and Bjornsrud. The reason for this greater spreading is certainly the fact that a current strong enough to accumulate the char does not develop until late in winter. When the reservoir approaches minimum regulated level the lake narrows to a sound, situated about 1,200 metres north of the tunnel intake. In years with normal lowering this happens in March—April, and for the rest of the season the fishing is largely concentrated in this place (Fig. 13). The accumulation of char can only be due to a gradual approach, and the migration involves fish from the whole of the lake, even the remotest part. Judged by the fishery statistics and the distribution of recaptures, a great proportion of the total population must stay in this restricted region in late winter (Fig. 15, lower part). Once having arrived, the char remains firmly attached to the current for the rest of the winter. For several years char have been tagged at the place, but not even once in a year has a recapture outside the current been made in the tagging season, and — when this occasionally happens — always in the lower end of Pålsbufjord (Fig. 16). Of the 2,346 char tagged in the lake in winter, nearly one half were released in the current. Seemingly nothing prevents the char from moving downstream to Tunhovdfjord, but THE WINTER MIGRATIONS OF CHAR 33 Fig. 16. Winter recaptures 1963 of char tagged near the tunnel intake in Pålsbu- fjord March—April 1963. no certain recapture has been made in this lake indicating a winter migra­ tion between the two lakes. The gradual increase in yield suggests that the char enters the current over a period of a couple of months. The time for their leaving the upper regions for the current may vary greatly. In the period 1954—63 char were tagged about March 20th every year at Åsodden in the middle of the lake. In some years the recaptures the same winter were concentrated at the current; in other years the recaptures were mostly made at the tagging place or close to it, only stray fish being recovered from the current. The variations are apparently due to water-level differences. The time of max- MAX.REG.LEVEL t | MIN. REG. LEVEL RES.DOWN 0 4 8 12 16 20 24 METRES Fig. 17. Relation between water level and number of returns made near the tunnel intake in Pålsbufjord during the winters of 1956—68. The char were tagged as spawners at Âsodden and Bjornsrud. 3 34 PER AASS Table 12. Pålsbufjord. Winter recapture above the Pålsbu Dam of spawners tagged at Åsodden and, Bjornsrud the preceding autumn. Males only, Recaptures made before the winter season have been deducted. Tagging Autumn of Number tagged (after deduc­ tion) Winter of Percentage location tagging recapture recapture Åsodden ............................ 1961 209 1962 6.7 1962 194 1963 7.2* 1963 94 1964 0 1964 117 1965 0 Bjornsrud ....................... 1965 175 1966 9.1 1966 212 1967 7.1 1967 250 1968 2.8 imnm arrival at the current of this and other northern spawning populations is easier to establish. Of spawners tagged in the upper half of the lake, in all 226 have been recovered in the current, including the winter of 1968. The rate of recapture, and most probably of arrival, is highly dependent on the water level. Apart from a few stray fish, recoveries were made when the reservoir was lowered 18 metres below maximum regulated level. The majority were recaught when the reservoir was lowered 21—23 metres, i.e. close to the regulated minimum level. Fig. 17 shows the relation between water level and recapture. The yield has a trend similar to the rate of recapture. In winters when the reservoir is not fully exhausted, the importance of the water level for the long-distance migrations is clearly demonstrated. In 1964 and 1965 the reservoir was lowered only about 18 and 14 metres respectively (Fig. 14), and no recaptures of the tagged northern spawning populations were made at the current in these winters. Table 12 shows the winter recapture percentages at the current calculated after deducting immediate recoveries near the tagging place. The downstream movements of the more local populations were certainly also influenced. Consequently the catch in the current declined, and its share of the total yield dropped (Table 13). But owing to the late concentration of char, the fishing does not depend on the current to the same degree as in Tunhovdfjord. On the whole, ice quality and yearclass fluctuations are as important for the fishery as is the formation of a current. The extraordinarily high yield in the remaining lake in 1964 was not due to the char concentrating at substitute places. The catch was made early in winter, before the current normally develops. Because there was little snow and the lowering was unusually slow it was possible to drive cars on the ice. This drew crowds of anglers to the lake and made the fishing more efficient. Based on the rates of recapture in the current of fish released there and THE WINTER MIGRATIONS OF CHAR 35 Table 13. Pålsbufjord. Estimated numbers of char caught in winters 1960/61—1966/67. Place Year Near tunnel intake Rest of lake Total 1960/61 .................. 4,500 8,300 12,800 1961/62 ................. 6,700 5,800 12,500 1962/63 ................. 4,900 6,000 10,900 1963/64 .................. 3,600 15,700 19,300 1964/65 .................. 1,700 8,400 10,100 1965/66 .................. 8,000 9,100 17,100 1966/67 .................. 11,500 8,900 20,400 spawners tagged at Åsodden, the proportion of the spawning population from Åsodden entering the current may be estimated at 35—45 per cent for the years 1959—60 and 1962—63. In spite of the autumn tagging places being situated in the upper half of the lake, no recaptures were made in winter at the outlet of the River Skurdalselv, which empties further up in the lake. However, a winter fishing of some importance takes place at the outlet of the main tributary, the River Numedalslågen, in the middle of the lake. In most years some of the tagged spawners are recaptured at this place, and the accumulation may consist mainly of fish migrating from the northern regions. Outside the current, winter taggings in the southern and largest basin are few, but the results do not indicate an upstream migra­ tion of any importance. VI. Discussion Large-scale winter migrations of inland char in natural lakes have not been reported, but their existence cannot be ruled out. Lindström (1954) reports of a winter fishery close to a current in an unimpounded Swedish lake. The char must migrate to the place from the spawning regions. In hydro-electric reservoirs, however, migrations occur regularly, commending as soon as the impoundments are accomplished. The characteristic hydro- logical difference between Scandinavian lakes and reservoirs in wintertime is that the latter are drained, resulting in a water flow of unusual strength for the season. In the case of Tunhovdfjord the mean winter water flow at the outlet is 50—60 cubic metres/sec. Retween the middle of November and the end of April approximately 750—800 million cubic metres flow through the lower end of this narrow lake. Currents develop upstream of the tunnel intakes and, depending on the topography of the reservoirs, may also arise far from the outlets when the lakes are lowered. From a long­ term production point of view, the reservoirs are characterized by a reduced 36 PER AASS bottom fauna production and the increased importance of the zooplankton as fish food (Nilsson 1961, Nilsson & Andersson 1967, Grimas 1965, Aass 1969). Jointly, the two factors could explain the winter migrations, which finally end with the stay in the currents. The shorebound movements in early winter take place after spawning, when the spent fish exert a great demand for food. At this time the water level generally gains its maximum for the year (Fig. 2). This makes it possible for the char to exploit the food production of areas difficult to approach during most of the year: banks, moss islands and sheltered shallow regions with some organogenic sedi­ ments and a scarce vegetation, mostly consisting of Ranunculus reptans L. and Littorella uniflora (L) Asch. The drained areas of Tunhovdfjord and Pålsbufjord are still covered with stumps from the former forest, and the shores are littered with drifts of forest debris. The remnants are used as a food basis for the bigger insect larvae (Grimås 1964). During a short period char feeds well on organisms typical of shallow water. The stomach contents consist mostly of semi-terrestrial Chironomidae larvae of the Diamesini group (det. M. Hirvenoja) and Macrocyclops albidus Jurine. A few Megacyclops viridis Jurine and Eucyclops serrulatus Fischer also occur (Cyclops det. Kåre Elgmork). In Pålsbufjord and Tunhovdfjord the char mostly approach the shore close to their spawning sites. Recaptures show that stray spent fish are few. The majority possibly stay along the shores until the food resources are exhausted or the drainage forces the fish towards the barren stone and gravel bottom of the deeper impounded zone. The mature char of these crowded populations is no true bottom dweller and possibly turns to the pelagic zone in search of food when the bottom supply is short. In so doing the fish may be attracted by the water flow that passes along the entire whole length of the lakes, carrying the organic drift from the watercourse above. Following the drift, the char move down to the currents, and make them their wintering places. This may be due to the drift being concentrated in the currents by turbulence; the accumulation of char in the mid-water layers is a conspicuous feature. In the Swedish reservoir Lake Jormsjön, Runnström (1964) showed that a considerable migration of spent fish took place from the only important spawning site to the main feeding ground. In this lake the two localities are distinctly separated. Owing to natural underwater barriers the feeding ground is not entirely drained during the winter and the char are not compelled to leave the region. In Jormsjön too, char accumulate near the outlet in winter and to a lesser degree in a sound dividing the lake into two parts. But it is not established by tagging experiments that the spent char, having once arrived at the feeding place, leave for the currents in winter. The remaining Scandinavian char-tagging experiments in reservoirs give little information about the immediate post-spawning THE WINTER MIGRATIONS OF CHAR 37 Table 14. Tunhovdfjord. Relation between water velocity and number of Bosmina coregoni in char stomachs. Tunhovd Dam January—March 1960. Number of Bos- January Current February Current March Current mina slow 1 strong slow strong slow 1 strong 0 9 6 9 10 11 6 0— 5 4 1 4 3 3 5 6— 10 1 2 3 5 1 1 11— 25 1 4 2 1 4 4 26— 50 1 3 3 3 2 4 51—100 4 1 1 0 1 0 >100 7 13 2 2 2 5 Number of fish examined 27 30 24 24 24 25 Other 1 Cyclops 1 Cyclops 2 Chironom. 1. 2 Chiio- 2 Cyclops 6 Cyclops organisms 1 Chironom. I. 1 Planorbidae nom. I. 1 Planorbidae 2 Chiro- 1 Pisidium 4 Trichoptera l. nom.l. 47 Hydracarina 2 Pisidium movements. Renewed contact with the fish after tagging at the spawning places is mostly established later in winter, after accumulation in currents (Aass 1963). Runnström (1963) states from Borgasjön that a pronounced downstream migration of mature fish takes place in the non-spawning period, November—August, but he does not specifically mention winter movements. A pause in recoveries between spawning and winter concen­ tration also applies to taggings performed in other seasons (Hanell & Holmberg 1968). Runnström (1964) maintains that at the outlet of a reservoir the organic drift attains its maximum during winter, when the main draining takes place. The erection of dams and draining through deep-set tunnels or bottom gates has, however, extinguished the original filter fauna of the shallow outlets. Judged by the stomach contents the amount of drifting plankton must also be very scarce in winter. Up to the end of March the drift in Tunhovdfjord consists almost entirely of small individuals of Bosmina coregoni Baird. Table 14 shows the number of Bosmina in the stomachs of char caught near the tunnel outlet in January—March 1960. Most of the stomachs were empty or contained only small numbers of the crustacean, and none was quite full. Towards the end of winter the amount of Bosmina in the stomachs showed a downward trend. True bottom organisms are consumed only in small numbers, and it is questionable whether they form part of the drift. Only a few fish had eaten the odd Pisidium or Planorbidae. One stomach containing 47 Hydracarina and 4 Trichoptera larvae was a marked exception. A few observations suggest that Bosmina disappear in late winter, being replaced by Cyclops sp. as the main 38 PER AASS food of the char. Of 19 fish caught at Bustrom in April 1960, 12 had eaten Cyclops in small numbers, the stomachs of the remaining 7 being empty. When 62 stomachs of Pålsbufjord char, caught near the outlet in March— April 1967, were examined, 39 were empty, while 16 contained small num­ bers of Bosmina; 3 fish had eaten molluscs and chironomid larvae and 4 fish had, in addition to crustaceans, also consumed molluscs or chironomid larvae, the total number of these organisms amounting to 50 and 10 respec­ tively. The food consumption of char staying in the currents is obviously very small. Table 14 also shows there is no significant difference between the amounts eaten during heavy and slight draining, i.e. in periods with strong and weak currents, respectively. The déconcentration of fish during week­ ends is thus difficult to explain in terms of reduced food supply. But the digestion in winter may be so slow as to disguise any short-time fluctuations in consumption. The stomach analyses indicate that another factor than the food supply may be active in concentrating the char in the currents. The winter migra­ tions of char in reservoirs may be interpreted as a partly active, partly passive movement. The arrival at and departure from the shores in early winter would then be the active phase. But once the fish have passed into the pelagic zone, the transport down the lakes could be as passive movement with the gentle flow. The assumed speed of the fish, about 1 kilometre every 24 hours, does not imply an active migration. The fish moving in deep water under the ice are not provided with visual clues and navigation must be difficult. The char can only halt when the water attains a velocity which is perceived by the fish. This happens in the currents, and without such an instinctive reaction, nothing would prevent the char from drifting out of the lakes, with an extinction or severe reduction of the population as the result. The Scandinavian mountain reservoirs, being mostly small or medium­ sized, rarely allow of migrations exceeding 20—30 kilometres a linear distance. This will not prevent the char from the most remote spawning place arriving at the outlets during winter, even if the movement is a passive one. The reluctance of the char to move down the rapid and open parts of the currents resembling rivers is established by the low rate of downstream recapture after winter taggings (Fig. 6, 12 and 16) and by gill-net fishing in the open water. This behaviour may be due to several factors, e.g. the in­ creasing energy requirements, the often observed reluctance of the char to leave the protecting ice cover, but most plausibly by the disposition of the char to halt when the bottom appears to rush away. This behaviour is more easily observed in summertime when char are swept down through tunnels or bottom gates in dams. Eric Fabricius, in a personal communica­ tion, was the first to explain the halt as a reaction of the char to the visual clue provided by the bottom rushing away. THE WINTER MIGRATIONS OF CHAR 39 The brown trout behave quite differently and are often observed in the open parts during winter, and, to judge by the catch, they rarely appear in the deep and slow currents. In the winters of 1966 and 1967 the propor­ tion between trout and char in the catches made at Bustrom and Tunhovd Dam was about 1: 400, against 1: 50 for the whole year. Aquarium observa­ tions show little or no territorial behaviour in the young stages of char. In still- water tanks the fry swim about when feeding starts after the yolk sack is finished, but they keep close to the bottom (Fahricius 1953, Frost 1965). Direct observations from the ice at Tunhovdfjord and Pålsbufjord confirm this. The char appear to be benthic in their fry and parr stages. The change in body colour, the dark sides with parr marks shifting to a silvery hue, is accompanied by a change in habitat. The char become more pelagic and form bigger shoals. In this behaviour they resemble salmon rather than trout. The change from a benthic to a pelagic habitat takes place in the two lakes when the char are 3—4 years old. If the migration towards the outlet is a passive movement with the water, the moving populations would naturally consist mainly of fish in the pelagic stage, as these most easily come in contact with the heaviest flow. This may explain why the migrating populations apparently consist of the bigger fish. In Tunhovdfjord and Pålsbufjord this impression is based on the composition of the winter catches (Table 8). But if it may be assumed that the scanty number of char leaving a current for further downstream migration is representative of the population migrating into it, then a check undertaken in the outlet should give some data about the size of the moving fish. Such a check has been made in the impounded Lake Limingen, situated in the county of Nord-Trondelag and close to Lake Jormsjön. During the observation period the lake vas drained in winter only, and in the seasons 1954/55 to 1958/59 a grating was erected some hundred metres down in the drainage tunnel, barring the whole cross section. Char down to a length of 10 centrimetres were detained, but apparently the grating worked effectively only on fish of 13 centimetres or longer. The fish had no opportunity to return to the lake, and every 24 hours the water was cut off and the char were caught in a trap. In winter char accumulate near the tunnel inlet, and the catch consists mainly of mature and larger immature fish, just as in Tunhovdfjord and Pålsbufjord. The size and age of the fish differ, however, the char in Limingen being bigger and older. The seasonal catch in the tunnel trap varied between 1,701 and 3,739 char. Fish of 20 cen­ timetres or more constituted the great majority, confirming the impression that the moving fish chiefly belong to the pelagic stage. But smaller fish occurred relatively more frequently than in the ice fishing catches. This may imply that more small fish take part in the downstream winter migra­ tions than is generally supposed. By keeping close to the bottom they escape exploitation by the fishing that takes place in mid-water, concen- 40 PER AASS trating on the bigger fish. Lake Limingen has an area of about 95 sq. kilometres and char is the main species. The number emigrating must be considered insignificant. The total recapture in the trap of spawners tagged at different localities in the lake was only 1 per cent (Aass 1963). This small winter emigration accords with the results obtained by tagging in Pålsbu- fjord and recapture in Tunhovdfjord. VII. Fish management applications Generally, impoundments will alter the balance between populations of trout and char. As a result of the reduced bottom-fauna production in reservoirs, the segregation between the species will increase (Aass 1957, Nilsson 1965). The trout, mainly a bottom feeder, will be hardest hit, and in many cases reduced spawning facilities will accelerate the decline of the populations. Conversely, char of all stages may increase their consump­ tion of planktonic crustaceans, the summer abundance of which is almost unchanged when the short-term impoundment effect is terminated (Löt­ marker 1964). Improved spawning possibilities in the eroded zone may theoretically lead to an increased recruitment of char, provided that the reservoirs are drained slowly. Data of char recruitment are, however, generally lacking. As a final result of impoundments, the mature char decrease in size and the traditional autumn fishing with gill-nets suffers a setback. In some places the decline is aggravated by the long-term difficulties of net fishing over drained areas (Jensen & Aass 1958). Although the char utilize the food production of subarctic and high boreal reservoirs better than do the trout, stocking with char has not been recommended, one of the objections being the difficulty of exploiting char stocks. A satisfactory catch after impoundment is dependent on a change in fishing methods. In many reservoirs the winter migrations and the subsequent accumulations of char in the currents render this possible. The favourable conditions created make them outstandingly good fishing places, and ice fishing with baited hooks becomes the most important method, also when the reservoirs are situated far from densely populated districts. In addition to the local population, which derives a direct income from fishing, visiting anglers spend considerable amounts on building or hiring cabins, and everybody has to pay for a fishing licence. In Tunhovdfjord about 4,000—5,000 daily visits are paid to the Bustrom—Tunhovd Dam region alone in an average winter. Of the total catch of char in Tunhovdfjord during the years 1961—67 (Table 15), winter fishing with hooks was responsible for 87.8 per cent of the number caught, and as much as 74.8 per cent of the total was captured in the currents. The corresponding figures for the catch in Pålsbufjord THE WINTER MIGRATIONS OF CHAR 41 Table 15. Tunhovdfjord. Total catch of char in numbers 1961—67. Year Winter fishing in currents | rest of lake Summer and autumn fishing Total 1961........................... 59,900 7,300 12,500 79,700 1962 ........................... 44,650 6,100 10,000 60,750 1963 ........................... 34,050 5,050 8,000 47,100 1964 ........................... 36,900 7,600 6,900 51,400 1965 ........................... 47,700 10,600 9,300 67,600 1966 ........................... 73,800 12,500 7,500 93,800 1967 ........................... 71,100 14,600 5,800 91,500 Total 1961—67 . . 368,100 63,750 60,000 491,850 (Table 16) were 71.9 and 27.8 per cent. In this lake there is only one current and no substitute for this exists if draining is late or incomplete, as it was in 1964—65. The average yearly rate of exploitation during the winter fishing in the currents in the Bustrom—Tunhovd Dam region varies be­ tween 25 and 13 per cent for the local and the most distant spawning popula­ tions respectively. In Pålsbufjord an average of 6—7 per cent of fish tagged in the middle of the lake is recaptured near the outlet in winter. Summer and autumn fishing being insignificant, the winter rates could be con­ siderably increased without endangering the recruitment. A higher exploita­ tion of the remote spawning populations is especially desirable, as this will have the greatest effect on the total yield. Theoretically the migrations may be speeded up or delayed by varying the velocity and level of the water, and the starting time and duration of the fishing season thus influenced. The char may be concentrated at one fishing place or distributed among several if possibilities for this exist. This seems to be a general effect, and should be taken into consideration by those who manage char reservoirs. In the case of Tunhovdfjord the lowering should start in December, and the reservoir ought to be 5 metres below maximum regulated level in the Table 16. Pålsbufjord. Total catch of char in number 1961—1967. Year Winter fishing Summer and autumn Total in currents rest of lake fishing 1961........................... 4,500 8,300 7,300 20,100 1962 ........................... 6,700 5,800 5,200 17,700 1963 ........................... 4,900 6,000 5,900 16,800 1964 ........................... 3,600 15,700 5,900 25,200 1965 ........................... 1,700 8,400 6,900 17,000 1966 ........................... 6,400 9,100 6,250 21,750 1967 ........................... 11,500 8,900 2,200 22,600 Total 1961—67 . . 39,300 62,200 39,650 141,150 42 PER AASS second half of January. This is somewhat earlier than the general practice and should speed up the start of the longdistance migrations. After a period of gentle lowering, allowing a large proportion of the char population to reach the dam, the reservoir should be rapidly lowered to 8—10 metres below maximum from the middle of February. This will halt the char at Bustrom, and create conditions for a long fishing season at the two places where the char are most effectively exploited. Further lowering should take place slowly and as late in the season as possible. Rapid drainage at this stage will disperse the fish to several different places, but the total catch is unlikely to increase. The ideal management of the reservoir from a fisheries point of view is likely to clash with the interests of power production. The latter is favoured by a high water level and early lowering is not desirable. Later in the season the conflict of interests should be of less importance. For a maximum yield, Pålsbufjord should be lowered almost to the minimum regulated level early in winter. But for two reasons it is impossible to speed up the draining as desired. First, the management of the reservoir is dependent on the water level of Tunhovdfjord; this must always be lower than in Pålsbufjord if draining is to he possible. Secondly, the existing tunnel is not dimensioned for a rapid lowering of Pålsbufjord. Such an obstacle is hardly met with in modern hydro-electric developments. But even if a rapid lowering had been possible, its accomplishment is not desirable every year, because of the consequences on the recruitment of the char. Many spawning places are situated in the impounded area, and early lowering will kill a large proportion of spawn and alevins, making the year class a poor one (Aass 1964)). Every third or fourth year the draining should be slow or incomplete. VIII. Summary 1. During the years 1951 and 1954—68, in all 21,073 char were tagged in the two lake reservoirs and the connecting river. Of these 10,475 were tagged in winter, 1,884 in spring and summer and 8,714 in autumn as spawners. 2. Post-spawning and early winter movements were shorebound feeding migrations, mostly covering short distances. 3. Later in winter, char accumulate in great numbers in the currents near the intakes of the drainage tunnels and in narrow sounds which come into existence with the lowering of the reservoir. The movements are nearly always downstream. 4. Char from neighbouring areas enter the currents first, the populations of distant regions chiefly arriving in middle and late winter. 5. There is a close connection between water level and dispersion of char. THE WINTER MIGRATIONS OF CHAR 43 When a current attains a certain threshold velocity the char interrupt their migration and keep to the current for the rest of the winter. 6. The catch in the currents consists mostly of fish in the pelagic stage, that is mature or older immature char. It is not proved to what extent the younger fish take part in the migrations. 7. The downstream migrations involve a great part of the total population. Of the more distant spawning populations in Tunhovdfjord, more than 50 per cent take part. A fish may participate in several subsequent years. 8. The food supply in the currents is negligible, and the downstream move­ ments make little sense as active feeding migrations. The lowering of the water level brings about a movement of the pelagic char towards deeper water and in contact with the flow through the reservoirs. The further migration is possibly a passive downstream displacement. 9. Important winter fishing arises in the currents. In Tunhovdfjord, the annual rate of exploitation varies between 12—25 per cent for distant and local spawning populations respectively. On average, 75 per cent of the total char catch in Tunhovdfjord is made by angling in the currents. The corresponding figure for Pålsbufjord is 28 per cent. 10. The affinity of char to currents may be exploited to improve the fishing. When managing the reservoirs, the distribution of char most favourable to the fishing should be considered. IX. Acknowledgements My thanks are due to all who have assisted in the gathering of informa­ tion and in the catching and tagging of the fish. Particularly I should like to mention Oskar Hvamb, Bjarne Myhre, Hans Rodberg, the late Henrik Aasen and the late Knut Lislegård. Financial support has been given by the Agricultural Research Council of Norway, Reguleringsforeningenes Landssammenslutning and Reguleringsindustriens konsesjonsavgiftsfond. X. References Aass, P. 1957. Fiskeriundersokelsene i Pålsbufjord og Tunnhovdfjord 1949—1956. Års- beretning for fiskeriundersokelser i reguterte vassdrag. Inspektoren for ferskvannsfisket. 36 pp. (Mimeographed in Norwegian.) — 1963. Limingenreguleringens virkninger på fisket. 40 pp. (Mimeographed in Norwegian.) —■ 1964. Reguleringer og royerekruttering. Norges .leger- og Fiskerforbunds tidskr. 93: 378—381. — 1965. Rödingvandringar i Pålsbufjord och Tunnhovdfjord. Inform. Sötvattenslabora- toriet, Drottningholm (4). 17 pp. (Mimeographed in Swedish.) — 1969. Crustacea, especially Lepidurus arcticus Pallas, as brown trout food in Norwegian mountain reservoirs. Rep. Inst. Freshm. Res. Drottningholm 49: 183—201. 44 PER AASS Caelin, B. 1955. Tagging of Salmon Smolts in the River Lagan. Rep. Inst. Freshw. Res. Drottningholm 36: 57—74. Dahl, K. 1920. Studier over roje i orretvand. Norges leger- og Fiskerf orb und s tidskr. 49: 233—248. — 1926. Undersokelser ved Tunhovdfjorden angaaende fiskens næringsforhold for og efter reguleringen. Medd. Norges Vassdrags- og Elektrisitetsvesen. Foss. 1. 1—18. — 1932. Influence of water storage on food conditions of trout in Lake Paalsbufjord. Det Norske Vidensk. Akad. Oslo, 1, Mat.-Naturv. Kl. 1931. 4: 1—53. Fabricius, E. 1953. Aquarium Observations on the Spawning Behaviour of the Char, Salmo alpinus. Rep. Inst. Freshw. Res. Drottningholm 34:14—48. Fossedirektdren. 1932. Nore kraftanlegg. Medd. Norges Vassdrags- og Elektrisitetsvesen. Foss. 3. 1: 199. Frost, W. E. 1965. Breeding habits of Windermere charr, Salvelinus willughbii (Günther), and their bearing on spéciation of these fish. Proc. Roy. Soc. B. 163: 232—284. Grimas, U. 1964. Studies on the bottom fauna of impounded lakes in southern Norway (Tunnhovdfjord, Paalsbufjord and Rodungen). Rep. Inst. Freshw. Res. Drottningholm 45: 94—104. — 1965. Effects of impoundment on the bottom fauna of high mountain lakes. Acta Univ. Upsaliensis Abstr. Uppsala Diss. Sei. 51: 1—24. Hanell, L. & R. Holmberg 1968. Rödingvandringar i Kultsjön. Information från Söt- vattenslaboratoriet, Drottningholm (7). 10 pp. (Mimeographed in Swedish.) Huitfeldt-Kaas, H. 1935. Der Einfluss der Gewässerregelungen auf den Fischbestand in Binnenseen. Oslo. 105 pp. Jensen, K. W. & P. Aass 1958. Bunnforholdene i regulerte sjoer. Norges leger- og Fislcer- forbunds tidskr. 87:44—55. Lindström, T. 1954. Non-reproductive Migrations in the Char. Salmo alpinus, L. Rep. Inst. Freshw. Res. Drottningholm 35: 118—132. Lötmarker, T. 1964. Studies on planktonic Crustacea in thirteen lakes in northern Sweden. Rep. Inst. Freshw. Res. Drottningholm. 45: 113—189. Nilsson, N.-A. 1961. The effect of water-level fluctuations on the feeding habits of trout and char in the lakes Blåsjön and Jormsjön, North Sweden. Rep. Inst. Freshw. Res. Drottningholm 42:238—261. — 1965. Food segregation between sahnonoid species in North Sweden. Rep. Inst. Freshw. Res. Drottningholm 46: 58—78. — & G. Andersson 1967. Food and growth of an allopatric brown trout in northern Sweden. Rep. Inst. Freshw. Res. Drottningholm. 47: 118—127. Runnström, S. 1963. Rödingens lekvanor och vandringar i Borgasjön före och efter sjöns reglering. 12 pp. (Mimeographed in Swedish.) — 1964. Kontroll av Jormsjöns lekrödingbestånd i Blåsjöälven genom en fiskspärr samt studier över dess vandringar genom märkningsförsök. Inform. Sötvattenslaboratoriet, Drottningholm (4). 17 pp. (Mimeographed in Swedish.) On the ecology of the ide Idus idus (L.) in the River Kävlingeån, south Sweden1 By Plutarco Gala Department of Animal Ecology, University of Lund, Lund, Sweden 2 Contents I. Introduction ................................................................................................................................... 4g II. Techniques ..................................................................................................................................... 47 III. Study area........................................................................................................................................ 47 IV. Physico-chemical environments .............................................................................................. 49 Water temperature ................................................................................................................ 51 Colour ........................................................................................................................................ 51 PH .............................................................................................................................................. 51 Hardness, Ca + Mg ................................................................................................................ 51 Specific conductivity ............................................................................................................ 52 Chloride ion contents ............................................................................................................ 52 Oxygen contents .................................................................................................................... 52 V. Pollution ......................................................................................................................................... 52 VI. Death of fish ................................................................................................................................ 54 VII. Qualitative composition of benthic fauna............................................................................ 55 VIII. Production of plankton organisms......................................................................................... 55 IX. Fish fauna and fishery.............................................................................................................. 5g X. Notes on the macrophytic vegetation.................................................... 61 XI. Migration and spawning season of the ide ......................................................................... 61 Spawning time ....................................................................................................................... 63 Spawning results ................................................................................................................... 65 XII. Growth and age of the ide ....................................................................................................... 65 Embryonal development...................................................................................................... 65 Sac-fry growth ....................................................................................................................... 66 Growth of voung-of-the-year.............................................................................................. 67 Age and increase in body length ....................................................................................... 69 Characteristics of the scale annulus................................................................................. 70 Time of annulus formation ............................................................................................... 70 Regeneration of scales.......................................................................................................... 71 Body-scale relation ................................................................................................................ 75 Length distribution ................................................................................................................ 78 1 This paper was prepared from a portion of a thesis submitted in partial fulfillment of the requirements for the degree of Ph. D. (FL) in the Department of Animal Ecology, university of Lund. This investigation was aided by fellowships from the Banco de la Republica, Colombia, and from SIDA (Swedish International Development Authority). 2 Present address: Departamento de Biologia, Universidad Nacional de Colombia, Apartado Aéreo 7495, Bogota. 46 PLUTARCO CALA Growth in length ...................................................... Length-weight relation ........................................... Calculated growth in weight .................................. XIII. Sex ratio on spawning ground...................................... XIV. Food habits ......................... .......................................... Composition of the diet in relation lo the fish size Food utilization and food abundance................... Seasonal changes ................................................... Previous investigations ......................................... XV. Summary ....................................................................... XVI. Acknowledgements ....................................................... XVII. References .................................................................... 79 81 86 87 88 89 89 92 93 94 97 97 I. Introduction Various aspects of the ecology of the ide have been investigated previously. Studies concerning its growth were carried out by Brofeldt (1917), JÄÄSKE- LÄINEN (1917), Rosén (1920), Otterstrom (1930—31), Segerstråle (1933, 1950), Hochman (1956), Popescu et al. (1960), Penåz (1961), and Balon (1962). Sundewall (1851, 1855), Stankovitch (1921), and Segerstråle (1926) îeported on the youngest stage of development, Balinsky (1948) described the development of certain specific characters of the species. Lenkiewicz (1964) investigated the temperature preferendum. Popescu et al. (1958) treated artificial fertilization, and Popescu et al. (1960) described fecundity, Schneider (1901, 1920), Collett (1905), Huitfeldt-Kaas (1917), Jääske läinen (1917, 1921), Otterstrom (1930—31), and Popescu et al. (1960) worked on the feeding habits. Other papers treating in some way ecological questions of the ide are those by Ekström (1831), Sierold (1863), Lillje- bgrg (1891), Smitt (1895), Stuxberg (1895), Ekman (1922), Komonen (1948), Berg (1949), Pliszka (1953), Hochman (1959), Duncker (1960), Mann (1961), Alm (1964), and Bogan (1964). Further, Popescu et al. (1960) and Balon (1962) reviewed several works dealing with ecological questions of the ide in the U.S.S.R. Lundberg (1899) and Berg (1949) described the geographical distribution of the ide. There are a few more important papers on the life history of the species in the Nordic Countries, viz. in Finland, Schneider (1901), Brofeldt (1917), Jääskeläinen (1917, 1921), Rosén (1920), Segerstråle (1933, 1950), in Denmark, Otterstrom (1914, 1930—31), and in Norway, Collett (1905) and Huitfeldt-Kaas (1917). There are few studies of the ecology of the ide in Sweden. Sundewall (1851, 1855), and Segerstråle (1926) studied its early development while Ekström (1831), Lilljeborg (1891), Stuxberg (1895), and Ekman (1822) just mentioned spawning and food. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 47 The main purposes of the present investigation were to study the ide population, in the River Kävlingeån in southwestern Sweden with regard to age and size at onset of maturity, age and growth, body and scale rela­ tionship, length and weight relation, growth rate in relation to sex and food. II. Techniques This study is based on data from ide collected in the River Kävlingeån, in the section Högsmölla to the mouth (Fig. 1). The studies were carried out during a 4-year period (1963—1967). Adult ide were collected mostly by means of seine, bow-nets, and weire-netting used by commercial fisher­ men in the water course. This sampling gear captured, almost exclusively, fish of age group II or older. Young-of-the-year or fry were collected by electro-fishing. During electro-fishing, the fish were collected with a hoop-net of a mesh size of ca 1 mm to catch fry and detached bottom fauna at the same time. Fish younger than one year were preserved — like the bottom fauna — in 80 % alcohol. The total length of the fish (tip of the nose to tip of the tail, lobes com­ pressed) was determined to the nearest millimetre. Representative samples of scales were taken. Fresh weights were determined to the nearest gram (or 0.1 g of some of the fish). Sex was determined by gross inspection of the gonads, or in immature fish by microscope examination. Sex could not be determined of age group 0 or age group I until near the end of the second growth season. Stomachs of adult ide were removed and fixed in 80 % alcohol for studies of food and feeding habits. In 1964 and 1965 quantitative collections of plankton were made and preserved in formalin. The methods used are described in more detail in the relevant sections. III. Study area The River Kävlingeån, which in its lower part is called Löddeå, runs from the Lake Vombsjön and falls out into the Lomma Ray in the straits of Öre­ sund about 15 km north of Malmö (Fig. 1). The length of the river is 46 km. The river runs through one of the most cultivated parts of the country, and only some of the tributaries run through woodland. The bottom is over­ laid with a layer of silt that has been eroded from the fields. Intensive cultivation has caused nearly all the land to be drained, ditched and ploughed. From 1850 to 1900 about 28,000 ha. of bog and lake were drained in the basin of the River Kävlingeån. In addition, large parts of the old 48 PLUTARCO CALA HÅSTAD MÖLLA ) kàvlingil_-7A ^rààS LÖDDEKÖPINGEV^ÖGSMÖLLA ÖRTOSFT^ lödde åA^SiThstävie} i LOMMA BAY\ ÖRE SUND :e vombsjon V björkaån SCALE 1150000 LINGAVÄLSÅN Fig. 1. The water system of River Kävlingeån. Collecting sites I, II, III, IV, V, VI, A. grazing grounds were put under cultivation, a total of about 19,000 ha. Thus, the rain water, to a large extent, passes into drain-pipes and canals together with much topsoil which is lost out to sea. An analysis of the water of Käv­ lingeån shows that 1 m3 of water contains about 50 g of humus and 100 g of minerals (Wolf 1960). The principal tributaries are the streams Bråån, Klingavälsån and Björka- ån, the continuation of the river upstreams of the lake Vombsjon (Fig. 1). The tributaries arise between 130 and 170 m above the sea level, while Lake Vombsjon is only 19.5 m above the sea level. The catchment area amounts to about 1,230 km2 of which 26.2 km2 consist of lake surface. The largest lake, Vombsjon has a surface area of 12.2 km2. Before intensive drainage was carried out during the latter half of the 19lh Century the water-holding area amounted to 29 %> of the total catchment area. Today this area has decreased to 3.4%. This has undoubtedly given rise to a greater variation in the water level of the river at different times of the year than in natural conditions (Weijman-Hane 1964). SMHI (Meteoro­ logical and Hydrological Institute of Sweden) estimate the water flow at the river mouth as follows: Highest waterflow ...................... 122 m3/sec Normal high water flow .... 55 Normal water flow ....................... 11 Normal low water flow.......... 1.8 Lowest low water flow.......... 0.6 The critical years of 1939 and 1947 a very low water flow was measured at Kävlinge, viz 0.24 nrVsec and 0.39 m3/sec respectively. Lake Vombsjon and the river downstream of Vombsjon are exploited for water supply, Vombsjon by Malmö city and the river by several industries including Kävlinge community as seen from Table 1. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 49 Table 1. Exploiters of the River Kävlingeån and their future water require­ ments (after Weijman-Hane 1964). Exploiter Future water requirements mean maximum l/sec l/sec Furulund Community ................... Mab & Mya, Ltd., Furulund Swedish sugar factory, Kävlinge Kävlinge community ................... Kävlinge weaving-mill ................... Leather factory, Kävlinge............. Scans’ butchery, Krutmöllan Swedish sugar factory, Örtofta . . Harlösa milk industry................... Total .................................................... 15 15 83 200 76 100 35 35 7 17 46 190 17 17 55 55 56 61 390 690 From Table 1 it is seen that the future water need of 0.69 m3/sec con­ siderably exceeds the lowest low water flow. With the present water flow it is not possible to maintain a satisfactory water supply by returning as waste water to the river a lower volume of water than was taken from it. At the same time this means that high demands on the cleanness of the waste water must be set. The investigation was carried out at six main collecting stations (vide Fig. 1), situated at the following distances — as the crow flies — from the mouth of the river: 6.2 km (I), 5.9 km (II), 4.8 km (III), 4.4 km (IV), 2.3 km (V), and 1.2 km (VI). A complementary station was selected at 7.2 km (A). IV. Physico-chemical environments Samples of surface water for analysis were collected during June to December, 1965, in the lower reaches of the River Kävlingeån, at six main sites (Fig. 1). The laboratory analysis of the water comprised determination of specific conductivity, pH, oxygen, alkalinity, chloride, Ca + Mg, and colour (Table 2). The methods are based on Karlgren’s (1962). During 1963 there occurred several periods of low water flow. At the beginning of June the flow was only 1.6 m3/sec at Kävlinge. This was con­ siderable lower than in the corresponding period in 1961 and 1962 (3.4 and 4.6 nrVsec respectively). At the end of November there was an extreme high flow of 34 m3/sec. In 1964 the water flow, from the end of May to the end of November, was almost constant, ranging between 2 and 4 m3/sec. The lowest water flow was 1.4 m3/sec at the end of May and beginning of June and 1.6 m3/sec at the end of August. Compared with previous years the flow was lower than 4 50 PLUTARCO CALA Table 2. Water chemical data from the lower part of the River Kävlingeån, 1965. Site Surface Colour mg Pt/1 02 Cl mg/1 Ca + Mg mmolA Total Date (vide map 1) tempera­ ture PH mg/1 percent « 20.106 alkalinity mequiv./l Jun. 14 I 20.2 50 6.26 9.54 108 504 2.7 2.80 --- II 20.2 50 6.70 6.33 72 473 — 2.6 3.68 — III 20.4 40 6.20 4.79 56 483 — 2.6 2.88 — IV 20.0 70 6.26 10.38 115 483 — 2.4 2.80 w --- V 21.6 70 5.50 11.12 129 1,239 — 3.2 3.00 „ — VI 21.8 60 6.20 12.08 140 945 — 3.0 2.88 Jul. 1 I 18.1 70 6.00 9.10 99 483 39 2.2 3.00 » --- II 18.4 60 5.80 12.19 133 483 40 2.5 3.40 » --- III 18.0 40 6.36 6.54 71 483 38 2.5 3.04 — IV 18.2 50 6.00 9.53 106 473 •--- 2.5 3.12 — V 18.0 40 6.16 8.25 89 1,533 — 3.4 3.16 „ — VI 18.4 40 5.80 11.12 123 2,835 — 4.1 2.80 Jul. 20 I 22.0 70 8.34 13.57 158 462 37 2.4 3.08 _ II 19.9 60 7.80 8.46 95 483 38 2.6 3.36 — III 20.6 60 7.60 5.90 67 483 40 2.5 3.16 — IV 20.5 60 7.60 5.27 60 483 33 2.4 3.20 5j --- V 20.5 50 7.64 6.76 77 609 120 2.5 3.12 » — VI 20.7 40 8.04 10.06 112 1,512 550 3.4 2.84 Jul. 30 I 16.0 50 7.20 3.67 38 546 57 2.4 2.92 M --- II 15.8 40 7.36 3.10 32 546 56 2.5 3.04 » --- III 16.2 60 7.46 2.50 26 546 52 2.4 3.20 — IV 16.2 70 7.50 3.03 32 546 53 2.4 3.16 M -- V 16.0 60 7.60 5.80 61 588 90 2.5 3.32 » VI 16.2 50 7.66 4.09 43 672 120 2.5 3.32 Aug. 6 I 16.2 25 7.30 3.88 41 525 52 2.5 3.16 Aug. 24 I 18.5 40 6.86 3.45 38 536 57 2.3 3.12 Sep. 9 I 14.0 40 6.86 4.63 46 525 55 2.2 2.80 Oct. 20 I 9.6 50 7.50 6.22 56 525 55 2.0 2.80 » --- V 9.8 40 7.50 6.44 58 1,554 480 3.0 2.88 » --- river- mouth 9.8 40 7.00 5.80 53 3,990 2,700 6.1 2.88 Nov. 9 in 7.2 35 7.06 4.73 40 583 52 2.5 3.32 M --- V 7.1 35 7.60 6.44 55 553 54 2.6 3.36 „ — A 7.1 40 7.40 5.27 45 618 51 2.7 3.36 Dec. 7 Lödde- 0.8 40 7.06 10.38 75 420 23 2.2 2.72 köpinge „ — A 0.8 60 7.50 10.91 78 467 40 2.2 2.80 in 1961—63 but considerable more steady. High flow occured only on a few occasions and had very short duration. In 1965 the lowest flow was 1.9 m3/sec during some days in June and one day at the end of August. In 1966 the lowest flow, 3.6 m3/.sec, was at the beginning of June. During winter and spring there occurred very high flow with a maximum value of 63 m3/sec in late February, which coincided with emptying the ponds from Örtofta sugar factory. During summer and ON THE ECOLOGY OF THE IDE ID US IDUS (L.) IN THE RIVER KÄVLINGEÅN 51 autumn the water flow was relatively constant (ca 4 m3/sec), and excep­ tionally low water did not occur. The data on the water flow were taken from Weijman-Hane & Almestrand (1963—1966). Water temperature Temperature records are based on the daily measurements at Håstad Mölla by the Salmon and Trout Association (1954—66). The winter of 1954—55 was very hard: minimum air temperature —24.5°C at the end of February and —14°C in the middle of March. The winters 1956—57 and 1960—61 were mild, and most of the time the air temperature was above zero. As early as 9 March, the water temperature was over 5°C which usually occurs three to four weeks later and above 13°C from the beginning of May until 9 October. In 1962 the winter was fairly mild but long. The lowest air temperature at Håstad Mölla was —8.2°G in March and most of this month the air tem­ perature was below zero. The water temperature of the river reached 5°C as late as 12 April and was above 13°C from 5 June until 7 September. Thus 1962 differed considerably from 1961. The winter was cold and long in 1963 and very cold in 1966. In 1963 the water temperature at Håstad Mölla reached 5°C on 3 April only for three days, then decreased to about 1°C between 12—19 April; it was 5°C on the 25 April in 1966. Usually the water of the river reached 5°C at the end of March and be­ ginning of April and was above 13°C from the middle of May to September, i.e. there is a long summer without high temperatures. The coldest month is usually February and the highest water temperature occurs in July. Colour Colour values vary between 25 (6 Aug.) and 70 (14 June & 1, 20, 30 July) mg Pt/1. No significant differences in the variation between the different sites seem to occur. pH The pH varies around 7, as is shown in Table 2. On 1 July the lowest value, 5.50, was recorded. The highest value, 8.34 (one further value, 8.04, is above pH 8) was recorded on 20 July. The latter are probably temporary peaks caused by phytoplankton photosynthesis. The lowest pH values (some­ times below 5) (Weijman-Hane 1964) may be caused by the introduction of acid industrial wastes. Hardness, Ca+Mg The hardness varies mainly around 2.5 mmol/1. The highest value, 4.1 mmol/1, was recorded on 1 July. The lowest value, 2.0 mmol/1, was recorded 52 PLUTARCO CALA on 20 October. There is a high value, 6.1 mmol/1 from the river-mouth on 20 October connected with influx of sea water. Specific conductivity The water of the River Kävlingeån has a x-value of 473—945 X10~6 and increases successively downstream (Table 2). Occasionally there occurs an inflow of sea water giving the lower part of the river the character of an estuary (below Löddeköpinge). Thus the highest specific conductivity of 2.835X10 6 was measured on 1 July at site VI, and 1.512 X10" on 20 July, x-values of 1,554 X10-6 on 20 October, and 1,533 X10~6 on 1 July were recorded at site V.A x-value of 3,990 X10-6 was recorded on 20 October at the river mouth which regularly has a mixohaline water. Chloride ion contents The Cl- content varies around 50 mg/1 in the upper and central reaches of the river (Table 2). On 9 December the lowest value, 23 mg/1, was recorded at Löddeköpinge. The high values are from the lower reaches of the river where there is an inflow of sea water. There, high values of 550 mg/1, 480 mg/1 and 120 mg/1 were recorded on 20 July and 30 July. The value of 2,700 mg/1 was recorded at the river mouth on 20 October. Oxygen contents The oxygen content varies considerably during summer (Table 2). The highest value was 13.57 mg/1 measured on 20 July (site I) and the lowest was 2.50 mg/1 on 30 July (site III). Low oxygen concentration occurs from late July until September. In spite of the complicated relations in the River Kävlingeån, there is usually a good correlation between the oxygen content and the water flow. Thus, with a water flow of 2.0 m3/sec follows an oxygen content of about 1.0 mg/1 (Weijman-Hane 1964). V. Pollution River Kävlingeån is severely polluted. The pollution discharged into the river was estimated to about 400,000 Pe (Pe = the equivalent of pollution from one person) (Weijman-Hane 1964). The pollution accumulated down­ stream Vombsjön was estimated to about 300,000 Pe with the following distribution: ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 53 Urban districts....................... 25,000 Pe Sparsely populated areas . . 105,000 „ Industries ................................ 170,000 „ Most of the pollution enters the river near the inflow of the River Bråån (vide Fig. 1). Örtofta sugar factory during the working season — October and November — brings heavy pollution to the river which results in large development of the fungi Sphaerotilus natans and Leptomitus lacteus, which can be found as far downstream as Löddeköpinge. Swenander as early as 1910 concluded that the major polluters of the River Kävlingeån were the sugar factories. Thus, there are two critical periods: from the end of June to September — low water time — and during October to December when the sugar factories are in action. The importance of the latter has changed in the last few years. In 1964 a provisory storage-capacity regulation of the Lake Vombsjön was carried out by the city of Malmö, which garantees a water flow of at least 1.5 m3/sec at Kävlinge. This condition and storing of the spill water from Örtofta sugar factory in large ponds improved the water quality of the river downstream Kävlinge. The fungi were thereby removed from the river. This was the first autumn since the 1940’s that there had been no fungi. But in the following spring when the Örtofta sugar factory started emptying its ponds the development of the fungi began again. In 1966, the Örtofta sugar factory began to drain the ponds on 18 February and stopped on 7 March as a result of complaint about the development of large quantities of Leptomitus lacteus and Sphaerotilus natans. The former disappeared almost immediately downstream Örtofta. During this time 83,700 m3 effluent water were emptied from these ponds into the river. The biochemical oxygen demand was very high: mean BOD5 = 1,025 mg/1. Later in the summer the sugar factory emptied the ponds after the water had been treated so that it did not give rise to development of the above mentioned fungi. Certain bottom sediments are very oxygen consuming and may play an important role in the oxygen cycle of the river. These bottom sediments may be primary, i.e. consist of substances which through incomplete purification were released directly into the water course. They may also be secondary, i.e. caused by bioflocking or by dead organisms. This phenomenon is very common in the River Kävlingeån and may periodically give rise to oxygen deficiency and mud drift. Another kind of pollution is the mineral oil that from time to time is emptied into the river. When the sugar factories are in action, i.e. August—November and in the last few years discharge of waste water from dams in spring, the lowest 54 PLUTARCO CALA concentration of oxygen usually occurs at Kvarnvik, some kilometres down­ stream Örtofta sugar factory. At this time it is often difficult to obtain a correlation between the concentration of oxygen and the water flow. This may partly depend upon the large pollution. In 1959 at a water flow which for a long time was about 3.0 m3/sec, the concentration of oxygen was 0.42 mg/1 at Kvarnvik. In 1961 oxygen con­ centration was 2.0 mg/1 at a flow of 7.2 m3/sec, in 1962 at 5.2 m3/sec 3.6 mg/1, in 1964 0.75 mg/1 at 3.4 m3/sec and in 1966 0.7 mg/1 at flow of 18 m3/sec. Much has been done to purify the water of the river, but it seems that the discharge of polluted water tends to increase more rapidly than the measures of purification. With this tendency the present water supply from the Lake Vombsjön is not enough to maintain conditions satisfactory, particu­ larly so at low water flow. VI. Death of fish Swenander (1910) mentioned that for a long time people have com­ plained about the heavy fish death in the river as soon as the sugar factories began their activities in the autumn. In 1907 at that season, large quantities of fish of different species died in the lower part of the river. He suggested that large quantities of Sphaerotilus and Leptomitus — occurring down­ stream of Örtofta sugar factory — indicate strong pollution and scarcity of oxygen, and concluded that the fish death was caused either by poisoning or oxygen deficiency or by both. Since Swenander’s report, death of fish has occurred many times in the river. A few more recent cases will be dealt with. In 1947, death of fish occurred in the lower reaches of the river. The rainfall was very low and the water flow at Kävlinge as low as 0.4 nrVsec. In combination with a rela­ tively high water temperature and heavy pollution of the river this was the cause of the death of fish (Weijman-Hane 1964). The largest mass death of fish in the river occurred in late October 1951 when a pond dam broke down at Örtofta sugar factory discharging stored waste water into the river and causing extensive death of fish downstream as far as the river mouth (Wolf 1965). In October 1959, the Swedish Salmon and Trout Association (cited from Weijman-Hane & Almestrand 1960) reported death of fish at Krutmöllan as a result of low oxygen content and pollution caused by water coming from the stream Bråån and from Örtofta sugar factory. The same phenomenon occurred at Håstad Mölla in 1960. On 18 June 1962, upstream Furulund, fish were observed gasping for air at the water surface. The oxygen situation was found to be critical in the river from Kävlinge to Löddeköpinge. In spite of the water flow being ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVEK KÄVLINGEÅN 55 5 m3/sec the pollution from Kävlinge in combination with high water tem­ perature resulted in rapid oxygen consumption which initiated fish death (Weijman-Hane & Almestrand 1962). In autumn 1962, large quantities of fish died in the lower reaches of the river. Wolf (1965) estimated the dead fish to ca 75 tons. The cause was the low content of oxygen, as low as 0.5 mg/1 at Högsmölla, due to pollution by waste water of different kinds (Weijman-Hane & Almestrand 1962). At the end of August and beginning of September 1964, the situation was very critical between Rinnebäck and Högsmölla. Gasping fish were observed. The oxygen content was slightly over 1 mg/1 between Rinnebäck and Kävlinge urban district, but at Högsmölla only 0.8 mg/1 (Weijman-Hane & Alme­ strand 1964). Dead fish were found. Thus, present conditions are more or less the same as those described by Swenander in 1910. VII. Qualitative composition of benthic fauna The collections were made when electro-fishing in May to November 1965 and 1966, in the lower reaches of the river. The fauna was collected with a net of ca 1 mm mesh size, and preserved in 80 % alcohol and identified to species or the lowest possible taxon (Table 3). The samples from each month were added since the organisms in question were found to be distributed through the study area, except such marine or brackish water species as Neomysis integer which reach only short dis­ tances upstream the river mouth. Most samples were taken near Stävie. The sampling was intended to give a simple survey of bottom-living animals which might be of importance as fish food. The Ephemer opter a were specifically identified by Dr. Staffan Ulf- strand. Fil. kand. L.-E. Persson helped with the identification of the Gastropoda. The most important taxa were Gastropoda, Chironomidae, Corixidae, Ephemeroptera (especially Cloëon dipterum), Asellus aquaticus, Hirudinea (especially Piscicola geometrd) and Oligochaeta (cf. Table 3). VIII. Production of plankton organisms Plankton was collected from April to November in 1964 and 1965. Three samples each of 10 1 of water were filtered through a net with a mesh size 100 p (in 1964) and 145 p (in 1965). The samples were taken between 10.00 hr and 17.00 hr at the surface (not deeper than 30 cm) and in the same area where fish fry were collected by electro-fishing. The stations are described above (p. 8). 56 PLUTARCO CALA Table 3. The occurrence of bottom and vegetation fauna in the lower part of the River Kävlingeån (xxx = common, xx = less common, x = rare).-'■()* num­ ber of samples. April May June July August Sept. Oct. Nov. 1966 (2)* 1966 (5)* 1966 (7)* 1966 (5)* 1966 (2)* 1966 (7)* 1966 (8)* 1966 (4)* 1965 (4)* 1965 (R* 1965 (4)* 1965 (2)* HYDROZA ................................................................. _ X X XX £ X X X X TURBELAR1A .................................................... - X - - X XX X _ X _ X X NEMATODA .......................................................... XX - - - - - _ _ _ _ __ OLIGOCHAETA ................................................. XXX XX X X XXX X XX _ _ _ XXX X TURBELARIA .................................................... - X - X XX X - -X _ X X HIRUDINEA Piscicola geometra (L.) .................. - X - _ X X X _ XX X XXX X Other leeches ............................................ XX X - X XXX XX XXX - XXX X XXX XX COPEPODA Argulus foliaceus (L.) ........................ - - XX - XX XX - - - - _ - MYSIDACEA Neomysis integer (Leach). . . . - - - - - - - XXX - - - - ISOPODA Asellus aquaticus L ............................... - XX X XX XXX X XXX - XXX X XXX XXX AMPHIPODA Gammarus pulex L ............................... - - - - X X X - XXX - XXX _ HYDRACARINA ................................................. X X X X XXX X XX X X _ XX _ EPHEMEROPTERA Baetis rhodani (Pict.) .................. - X _ XXX _ XX _ _ _ _ _ _ B. fuscatus (L.) ...................................... - X - - - X XX _ XXX _ _ X B. buceratus Etn........................................... - - - - _ X _ _ _ _ _ _ B. vernus Curt.................................................. _ _ _ — _ _ XX _ _ _ _ XX Cloeon dipterus (L.)............................... - X XX _ XXX X XXX _ XXX _ XXX X Caenis robusta Etn. . .. _ _ _ _ X _ _ _ _ _ X Centroptilum lutecium (Müll.) - - X ODONATA Larvae ...................................... - X - - X - X - - _ XX XX HEMP1TERA Notonectidae .............. _ _ _ _ X _ X _ X Corixidae ............................................................... X XXX X XX XXX XX XXX XXX XXX _ XXX XXX NEUROPTERA Larva (Sisyridae?) - - - X TRICHOPTERA ................................................. - X - - XX X X X XXX XX XX XX COLEOPTERA Dytiscidae ................................................... - _ _ _ XXX _ X _ _ _ X _ Other beetles ................................................. - X - X XX - X - XX - X XX DIPTERA Simuliidae ....................................................... XX XX _ X _ XX X X XX _ XX Chironomidae.................................................... XXX XXX XX XXX XXX XX XXX XX XXX XX XXX XXX Geratopogonidae ...................................... X X _ _ X _ X _ _ _ _ _ Other Diptera ............................................. X X X - XX _ XX - XX - - - GASTROPODA Lymnaea auricularia (L.) .... - _ _ _ XX X X _ _ _ _ X L. stagnalis (L.) .................. - — _ _ X _ X _ XX _ X _ Phgsa fontinalis (L.) ......................... - _ _ X XX X _ _ XX _ XX _ Planorbis carinatus Müll. . _ _ _ _ _ _ _ _ _ _ X _ Anisus vortex (L.) ............................. _ _ _ _ _ X XX X Bathyomphalus contortus (L.) — _ X X _ X _ _ _ X Gyraulus spp................................................... - X - — XX X XX _ XX _ XXX XXX Viviparus fasciatus Leach .... - - - _ _ XX _ _ _ _ _ _ Acroloxus lacustris (L.) ................. - X - - - - - - X - X - ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE HIVER KÄVLINGEÅN 57 Table 3. (Continued) April May] June July August Sept. Oct. Nov. 1966 (2)* 1966 (5) * 1966 (7)* 1966 (5)* 1966 (2)* 1966 (7)* 1966 (8)* 1966 (4)* 1965 (4) * 1965 (1)* 1965 (4) * 1965 (2)* Bithynia tentaculata (L.) .... X XX XX X XX B. leachi (Sheppard) ................... - - X - XX X - - X - X - Hydrobia sp ........................................ - - - - - X - - - - - - Valvata macrostoma (Steenbuch) ................................... _ _ _ _ _ _ X _ _ X V. Piscinalis (Müll.) ................... - X - - XX - X - X - XXX XX PELECYPODA Anodonta piscinalis Nilsson .. XX Sphaerium corneum (L.) .... - - - - X - - - - - XX - Mytilus edulis L.................................. - - - - X - - - - - - - Rotatoria and nanoplanktonic forms which pass through the meshes of the net, were of subordinate importance for the present study of the diet of the ide fry and its relation to the plankton, since quantities of nanoplanktonic organisms are eaten by fish only during the very early stages of life. The plankton was studied after preservation in formalin. The occurrence of plankton algae was noted only in 1965. Planktonic organisms in the river were recorded by Swenander (1910) and Weijman-Hane & Almestrand (1958—1966). They treated especially those species in some way connected with pollution, e.g. the ciliates Colpidium colpoda, Paramaecium bursaria Ehrenberg, Carchesium lachmanni, Chilo- don cucullulus, and the flagellates Euglena viridis Ehrenberg, and Chlamy- domonas. They also recorded the rotifer Rotifer vulgaris. The mass development of Protozoa and the rotifers is largely dependent upon the extent of water pollution and they may be used as indicators of pollution. The lower reaches of Kävlingeån are rich in planktonic organisms typical of an eutrophic environment. Phytoplanktonic organisms of particular impor­ tance in Kävlingeån are the green algae Cladophora and Stigeoclonium tenue. The latter is typical for environments under self-purification. Further, the diatoms Navicula gregaria and Nitzschia palea, are common. Swenander (1910) lists five crustacean zooplankters from the river, viz. Bosmina coregoni Baird, Chydorus ovalis Kurz, Daphnia hyalina Leydig and the copepods Diaptomus graciloides Lilljeborg and Cyclops sp. B. core­ goni was the most common species and Ch. ovalis the least common of the crustacean zooplankters. In general Cladocera were more abundant than Copepoda. My results (Table 4) confirm that Cladocera are more common than Copepoda, 58.7 °/o compared with 41.3 % of the total samples in 1964. 8.8 %> were Cladocera and 13.2 % Copepoda in 1965. 58 PLUTARCO CALA Table 4. Quantités of crustacean zooplankters in the lower part of the River Kävlingeån with notes on planktic algae in 1965. CLADOCERA COPEPODA ALGAE Date of collec­ ting Number of station Number of specimens per litre of water %> of the total planktic crusta­ ceans per sample Number of specimens per litre of water °/o of the total planktic crusta­ ceans per sample XXX = common xx=less common x = rare 9.V.64 la _ 9 100.00 lb — — 4 100.00 Ic 1 8.33 11 91.67 28.V.64 la 111 25.00 333 75.00 lb 54 2.95 1,777 97.05 Ic 111 6.45 1,611 93.55 Id 834 37.52 1,389 62.48 16.VI.64 la 142 92.21 12 7.79 lb 24 92.31 2 7.69 w Ic 12 85.71 2 14.29 Id 334 96.53 12 3.47 Ie 6 100.00 — — 6. VII.64 la 1 50.00 1 50.00 lb 1 100.00 — — 3.IX.64 la 2,069 99.28 15 0.72 n lb 430 50.00 430 50.00 n Ic 3,846 100.00 — — 30.X.64 I 1 50.00 1 50.00 13.XI.64 la — — — — lb 2 75.00 1 25.00 n Ic — ----- - 1 100.00 14.VI.65 II 213 84.52 39 15.48 X III 313 73.13 115 26.87 — IV 624 76.28 194 23.72 — V 25 100.00 — — X VI 285 96.94 9 3.06 XXX 1.VII.65 I 159 98.76 2 1.24 XXX n II 296 96.42 11 3.58 X III 415 98.34 7 1.66 — IV 322 97.28 9 2.72 X V 157 96.32 6 3.68 X VI 70 90.90 7 9.10 XX A 15 88.24 2 11.76 — 20.VII.65 I 224 82.96 46 17.04 XXX II 535 90.07 9 9.93 — III 157 74.41 54 25.59 X IV 91 84.26 17 15.74 XXX V 169 91.85 15 8.15 XXX w VI 33 75.00 11 25.00 XX A 13 54.17 11 45.83 — 30.VII.65 I 56 81.16 13 18.84 — n II 828 95.50 39 4.50 — III 15 31.25 33 68.75 — IV 9 81.82 2 18.18 XX V 217 75.09 72 24.91 X VI 52 89.66 6 10.34 — A 4 66.67 2 33.33 — 6.VIII.65 I 213 75.53 69 24.47 XX 15.IX.65 la 11 84.62 2 15.38 — lb 24 58.54 17 41.46 — 18.X.65 la 33 45.83 39 54.17 — 9.XI.65 III 37 100.00 _ — — » A 74 91.36 7 8.64 — ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 59 When studying the table we have to remember that although the sampling technique allowed for diurnal variations in the plankton density, no parti­ cular regard was taken to possible patchiness of plankton distribution. IX. Fish fauna and fishery The fish fauna of the River Kävlingeån is comparatively rich and diverse. In order of frequency the families represented are: Cyprinidae, Percidae, Anguillidcie, Esocidae, Gasterosteidae, Pleuronectidae, Salmonidae, Gadidae and Petromyzontidae. The cyprinids were represented by roach Rutilus rutilus (L.), ide Idus idus (L.), bream Abramis brama (L.), tench Tinea tinea (L.), crucian carp Carassius carassius (L.), bleak Alburnus alburnus (L.), white bream Blicca bjorkna (L.), rudd Scardinius erythrophthalmus (L.), minnow Phoxinus phoxinus (L.), and common carp Cyprinus carpio L. The Percidae were represented by perch Perea fluviatilis (L.), pike-perch Lucioperca lucioperca (L.) and the ruff Acerina cernua (L.). The sticklebacks present were the three-spined stickleback Gasterostens aculeatus L., the ten- spined stickleback Pungitius pungitius (L.), and the fifteen-spined Spinachia spinachia L. The salmonids of the river are atlantic salmon Salmo salar L. and the sea- trout Salmo trutta L. Further species represented are as follows: the eel Anguilla anguilla (L.), the pike Esox lucius L., the flounder Pleuronectes flesus L., the gadids burbot Lota lota (L.), the atlantic cod Gadus morhua L., and finally the river lamprey Lampetra fluviatilis (L.) and the sea lamprey Petromyzon marinus L. All these species were collected by me in the lower part of the river, from Högsmölla to the river-mouth. Gadus morhua and 5. spinachia occurred near the river mouth where also Coitus scorpius L., Labrus berggylta L. and Zoarces viviparus L. were captured by the Swedish Salmon and Trout Association’s fishermen. They also captured the cyprinid Abramis vimba L. in the lower reaches of the river. I have caught Pleuronectes flesus up to Högsmölla, i.e. about 10 km upstream, especially juveniles when electro­ fishing. In June 1966, 3 adult specimens of Petromyzon marinus were cap­ tured. Rare fish species in the lower part of the river are Salmo salar, S. trutta, Lota lota, Acerina cernua, Alburnus alburnus and the Petromyzontidae. The river has had no natural stock of salmon during the last hundred years and no stock of sea-trout during the last fifty years, mainly because of the bad pollution. Moreover mill-dams in the lower part of the river during periods of low water flow also formed obstacles preventing the migration of the fish to the spawning ground lying further up. 60 PLUTARCO CALA Table 5. Amount in kg of the most important commercial species of fish captured in the lower reaches of the River Kävlingeån, from Högsmölla to the mouth, during a period of 12 years (data from the Svenska Lax- och Lax- öringsföreningen u.p.a. 1955—1966). Year Pike Perch Eel Bream Roach & ide Other fish 1955 . .. 175 1,775 720 17,400 100 1956 . .. 241 1,659 730 645 22,900 127 1957 . .. 60 1,191 655 338 9,015 115 1958 . .. 90 1,580 1,100 508 10,485 511 1959 ... 209 2,348 1,040 534 16,900 376 1960 . .. 178 3,225 540 190 11,220 140 1961 . .. 217 1,867 680 398 10,455 202 1962 . .. 322 2,158 835 490 14,210 1,132 1963 . . . 177 830 875 31 6,615 110 1964 . .. 307 1,040 360 310 10,350 165 1965 . .. 293 520 785 165 10,750 245 1966 . .. 186 350 1,800 175 6,350 400 Up to latter half of the 19th Century there was still good fishing (Wolf 1960). A good supply of perch, pike, eel, roach, ide and crayfish existed. In the lower reaches of the river, about fifty men got a livelihood from fishing. The yield in these parts has been estimated at several hundred tons annually. In the upper parts of the river fish production was also high. Salmon has not been caught since the 1850’s. Sea-trout, on the other hand, were caught in small numbers within the lower reaches of the river up to the beginning of the 20th Century. During the present century, the fish yield from the river has steadily decreased. Between 1945 and 1950 about two million fry af salmon and sea-trout were implanted by the Swedish Salmon and Trout Association in small, unpolluted tributaries of the river. It was found that these small water courses had a considerable productive capacity: brooklets only half a meter to one meter wide yielded more than 300 kg/ha (Wolf 1960). According to Wolf, a large part of the migrant fish died each year in the seriously polluted lower reaches of the river. A number survived, however, and returned after their growth period in the sea. Returns were first regis­ tered in 1949 for Salmo trutta and in 1950 for S. salar. Since 1950 salmon and trout have been caught every year. Adult salmonids are taken in a trap at Hållstad Mölla when the fish, after having spawned, migrate to the sea. Some thousand S. salar and S. trutta have migrated up the river and spawned during the last few years (Wolf 1960). It is not yet certain, however, whether the stock of S. salar can be maintained as a naturally reproducing population in the river. According to Wolf (1965) S. trutta fry have not been planted since 1946 in the water system, while between 1962—1964, 18,000—21,000 S. salar fry ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 61 have been annually implanted in the tributary Bråån. The migration down­ stream increased from 300 smolts in 1963 to 800 specimens in 1964 (average length 155—160 mm and the largest 198 mm). Some of these salmon were marked and captures were reported from the Norwegian Westcoast and from Dueodde on Bornholm but so far no specimens have been caught in the river. Juvenile S. trutta migrate downstream every year, the number fluctua­ ting from 100 to 700 specimens per year. Evidently, a reproducing stock of S. trutta exists in the water system. Table 5 gives the biomass in kg of the most important commercial species captured in the lower reaches of the river, from Högsmölla to the mouth, during the last 12 years (data from the Svenska Lax- och Laxöringsföreningen u.p.a. 1955—1966). The eels were collected at Håstad Mölla. X. Notes on the macrophytic vegetation In the lower part of the river there is a rich vegetation of floating and submersed plants, which in the summer develop so that they may cover the entire width of the river (Fig. 2). The vegetation gradually diminishes downstream except for Phragmites communis. Common aquatic plants are Potamogeton pectinatus, P. lucens, P. natans, Ceratophyllum demersum, Lemna minor, L. polyrrhiza, Scirpus lacustris, and Nymphaea alba, Nuphar luteum, and Phragmites communis. XI. Migration and spawning season of the ide The ide penetrates the River Kävlingeån water system upstream as far as to Lake Vombsjön and the streams Bråån and Klingavälsån (vide Fig. 1). The largest movements of ide occur in the lower reaches of Kävlingeån, between the mouth and Högsmölla where the migrating fishes encounter the first mill-dam. Upstream of Högsmölla there are mill-dams and industrial constructions affecting the water flow at Kävlinge, Rinnebäck, Krutmölla, Harriemölla, Besmölla, Kvarnvik and Håstadmölla. Serious obstacles to fish movement occur at Rinnebäck and Kvarnvik where a cement wall across the river diverts all the water except at high flow. Salmo trutta and S. salar pass all these obstacles and reach their spawning places. Ide en route downstream are captured at Håstadmölla which shows that the population in the upper reaches is based upon fish migrating up­ stream for spawning. Most of the large captures of downstream-migrating ide at Håstadmölla, occur during spring in connection with the large migrations of full-grown ide in the lower reaches, coinciding with spawning movements and high 62 PLUTARCO CALA Fig. 2. View of the River Kävlingeån (station II). Dense vegetation of Potamogeton spp. covering the entire width of the river (9 Aug. 1966). water flow. Of 57 ide migrating downstream at Håstadmölla 36 were cap­ tured in April—June (Svenska Lax- & Laxöringsföreningen u.p.a. 1954— 1966). According to Ekström (1831), Smitt (1895) and Alm (1964) ide may ascend even large rapids, jumping over obstacles like the salmon. The ide spends the first year of its life in the river and then joins older migrating fish to the sea. During the summer (June—August) ide older than age group 0 disappear almost completely from the river, spending the summer in the sea. However, some specimens not older than group II remain in the lower part of the river and from time to time large ide pass the river mouth and proceed into the lower reaches of the river when weather is rough. In autumn the ide return to the river, the smallest ones in September and as the autumn advances the schools increase in number and size. Late in November, or before that time depending on water temperature, all age groups are represented in the lower part of the river. The fish remain during winter at the river mouth and in the low reaches (cf. also Otterstr0M 1930—31). During autumn and spring the fish stay in the lower reaches. At times ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 63 large schools are easily observed on the move. The mature fish go to the spawning place. When spawning is complete they leave for the sea, followed by subadult fish and finally by the small individuals. Generally, the largest ide leave the river earlier (April to May) and return later (October to November). Thus fish larger than 26—30 cm (older than age group IV—V) spend much more of the year in the sea than do the small fish. Spawning time The spawning of the ide in the River Kävlingeån during the 4-year study period occurred from 29 March to 17 April. In 1964 it started on 11 April (peak of spawning on 12 and 13). In 1965, spawning began on 1 April (peak of spawning on 2, 3 and 4) and in 1966 on 6 April (peak of spawning on 6 and 7). In 1967 the ide spawned on 29 March (peak of spawning on 30 and 31). The ide spawns in shallow streams at a depth of 0.5—1.5 m. The eggs adhere to stones, plants, etc. The largest fish spawn first. Spawning began 2 to 3 days after the daily water temperature had reached and remained above 5°C (Table 6). Most ide spawned during 2 to 3 days when the water temperature rose to about 7°C and the weather was fine. This was the case in 1964. But if the water temperature drops after the initial period of spawning, additional spawning takes place when the temperature rises again, as in 1966. Then spawning started on 8 April, on 10 April the water tem­ perature dropped to 2°C and spawning ceased. It recommenced on 25 April when the temperature returned to 5°C (vide Table 6). According to Lilljeborg (1891) spawning is generally earlier in southern Sweden than further north and varies according to the water temperature. According to Stuxberg (1895) the spawning period extends from the end of March in South Sweden to June in North Sweden. At a water temperature of 8 to 10°C the spawning is reported to last only 3 to 4 days, occurring day and night. Smitt (1895) reports spawning of the ide in Central Sweden to occur about the end of April. The males arrive earlier to spawning place than the females, the latter arriving when the water temperature was about 9°C. In Denmark Otterstrdm (1914) describes spawning at a water temperature of 8—10°C. Later (1930—31) he reported spawning in March and April, saying that the ide seemed to have begun spawning while the Lake Arreso had a tempera­ ture of about 6°C. Popescu et al. (1960) stated that the spawning of the ide starts when the water temperature has risen to 7—9° G and lasts 10 days and that it is interrupted by large changes of temperature or by flooding. Ekström (1831) found spawning to occur at the end of April in the islands round Mörkö — Central Sweden — and reports that the peak did not last 64 PLUTARCO CALA Table 6. Daily mean water temperature in °C in the River Kävlingeån at Håstadmölla in relation to the spawning of the ide in 1964—1967. Date 1964 1965 1966 1967 March 20 ..................... — 3.1 2.8 3.6 2! _ ___...... — — 3.2 4.2 22 ..................... — 2.9 2.8 4.6 23 ..................... — 2.9 2.8 5.0 24 ..................... — 3.5 2.6 — 25 ..................... — 3.8 2.3 3.1 26 ..................... — 3.3 2.0 — 27 ..................... — 3.2 2.5 5.2 28 ..................... — 3.4 3.4 5.7 29 ..................... — 4.4 2.6 4.8 30 ..................... — 5.5 2.4 5.0* 31 ..................... — 6.0 2.4 5.0* April 1 ..................... — 6.0 3.0 5.1* 2 ..................... — 5.8* 3.9 — 3 ..................... — 5.9* 5.3 5.1 4 ..................... — 5.5* 5.2 5.4 5 ..................... — 5.6 5.0 5.8 6 ..................... 2.8 6.2 4.6* 5.8 7 ..................... 3.9 6.6 5.0* 5.5 8 ..................... 3.7 5.9 5.0* 5.3 9 ..................... 4.6 4.4 4.2 — 10 ..................... 5.3 3.2 2.0 5.7 11 ..................... 6.0 3.0 — 6.0 12 ..................... -- - 4.6 — 6.2 13 ..................... 6.7* 5.2 — 7.2 14 ..................... 7.0* 5.5 — 8.1 15 ..................... 7.0 6.1 — 8.2 16 ..................... 7.7 6.1 — 8.7 17 ..................... 8.3 6.8 — 8.2 18 ..................... 9.0 6.8 1.8 6.2 19 ..................... 10.8 6.5 — 6.3 20 ..................... 11.5 5.8 — 6.8 21 ..................... 12.5 7.2 — 8.2 22 ..................... 11.8 8.2 4.2 6.2 23 ..................... 11.1 8.5 — 6.4 24 ..................... 9.8 8.0 4.2 6.8 25 ..................... 8.9 — 5.1 7.1 26 ..................... 9.0 7.8 5.9 7.7 27 ..................... 8.8 8.0 6.4 8.5 28 ..................... 9.2 7.6 7.3 * water temperatures at peak of spawning. 8.8 more than 3 days and that the males arrived first to the spawning place and that after the spawning males and females return together. Stuxberg (1895), Kröyer (cited by Otterström 1930—31) and Duncker (1960) agreed that the males arrived first in the spawning place and that the females are first to leave. Mühlen (c/. Otterström 1930—31) in Estonia (Lake Wirz-Järm) described the spawning of ide above 1 kg from the end of March to the end of April and of the small ide after middle of April. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 65 Åström (1859) reports spawning around 20 March in the River Höjeå in southern Sweden, Siebold (1863) from April to May in middle Europe, and Duncker (1960) from March to May in Nordmark (Germany). Spawning results Under natural conditions percentage of fertilization is very high. The rate of hatching is usually also rather high (Pliszka 1953). The important factors influencing the success of the spawning in the River Kävlingeån are: — 1) hatching is likely to show considerable losses if the water temperature drops after spawning and is below 5°C for some time (vide p. 32). Thus, in 1966 the water temperature dropped to about 1°C after the spawning and remained below 5°G for 16 days from April 8 to 24 (see Table 6). — 2) Deficiency in oxygen supply during the early development of the eggs, caused by pollution (cf. e.g. p. 13). — 3) Predation: gulls and other birds and certain species of fish feed on the eggs. In 1965 after a very successful spawning gulls were observed taking eggs at the spawning place at Stävie when the water level was rather low. XII. Growth and age of the ide The first year of life was divided into three stages, viz. 1) embryonic, from time of fertilization to hatching of the eggs, 2) sac-fry, the time during which the young fish obtain their nourishment entirely from the yolk sac, and 3) older fry which stage begins when the fry start to feed actively and continues to the end of the season’s growth. Embryonal development The length of time required for the eggs to hatch is related to water temperature. Fertilization of eggs from females with sperm from 2—3 males was carried out in pans at the river side. The eggs were thereafter transported to the laboratory in a pan with 3—5 1 of river water and was kept at 4—5°G. Batches of eggs were then distributed in small beakers to aquaria filled with 20—30 1 tap water in which the eggs adhered to small branches or to plastic netting. Material from two spawning periods was examined (Table 7). In 1965 the eggs began to hatch after 5.5 days at a mean temperature of 13.8°G (tem­ perature sum 94) : the length of the fry-sac at hatching ranged between 5.8—6.6 mm. In 1967, one batch of eggs began to hatch efter 12.8 days at a mean water temperature of 10°C (temperature sum 129), another after 5 66 PLUTARCO CALA Table 7. Development of embryo and sac-fry stages. Sample Embryonic stage (from fertilization to hatching of eggs) Sac-fry stage during the firs! (1965) week 2 days to hatch Temperature °C Tem­ perature sum Days Mean tempe­ rature 2 speci­ mens Mean length in mm Growth rate in mm per daymean range 1965 5.5 13.8 6.5—18.0 (16—18)1 94 0 18 10 6.0 1967-1 12.7 10.1 8.3—16.8 (8&11)1 129 1 18 11 7.1 1.1 1967-11 17.0 10.1 4.1—12.6 (11)1 172 2 18 12 8.2 1.1 1967-III 23.0 9.3 4.3—10.0 (9—10)1 213 3 18 17 8.5 0.3 4 18 18 8.8 0.3 6 18 17 9.0 0.2 1 Modal temperature 17 days at a mean of 10.1°C (172), and a third after 23 days at a mean water temperature of 9.3°C (213). When eggs were kept in water at 4—5°C, they did not hatch. Thus, low temperatures retard hatching, and below 5°C there is no hatching. Popescu et al. (1958) gave 9 days as the hatching time of the ide at a mean temperature of 12.6°C (temperature sum 110). According to Krijanovskii (cited by Popescu et al. 1958) the hatching period is longer than 17 days at 10—12°C. Sundewall (1855) and Stankovitch (1921) report 14—18 days and Segerstråle (1926) 8—12 days, but they do not give the water tem­ perature. Sac-fry growth In the laboratory in 1965, the length of the fry at hatching ranged be­ tween 5.8 and 6.4 mm (mean, 6 mm) (Table 7). The growth rate was 1.1 mm per 24 hours during the first 56 hours and than it decreased. On the 6th day it was only 0.2 mm per 24 hours. The mean growth during the first week was 0.5 mm per day, the fry having only the contents of the vitelline sac as source of food. According to Sundewall (1855) a newly hatched larva measured 5 mm and a three days old larva somewhat above 6 mm. Probably the measure­ ments refer to larvae fixed in alcohol. Stankovitch (1921) gave a length of 7 mm at hatching and 10 mm in 15—20 days old larvae after the vitelline sac was absorbed. Finally, Popescu et al. (1958) give the length as 6.5 mm at hatching, 7 mm after 12 hours and 7.5 mm after 48 hours. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 67 Table 8. Total length in mm of the young-of-the-year ide collected in 1963—1966 in the River Kävlingeån. Date 2 fish Mean Range 1963 4.VI ........................ 4 16 14—17 19.VI........................ 39 27 20—33 3.VII ..................... 229 34 27—45 5.VIII..................... 56 47 35—60 1964 28.V......................... 42 16 14—17 16. VI ....................... 21 25 23—27 30.IX........................ 1 69 — 11.XII ..................... 9 66 52—73 1965 8.1 ......................... 12 68 59—78 3.II ........................ 19 76 57—86 3.VI ........................ 26 13 12—15 14.VI........................ 161 18 14—21 l.VII ..................... 8 28 26—30 20.VII ..................... 3 38 32—44 30.VII ..................... 7 59 48—66 15.IX ........................ 4 57 55—58 28.IX........................ 20 61 53—71 18.X ......................... 6 64 58—73 7.XI........................ 17 68 59—79 9.XI........................ 21 65 57—74 1966 24. V......................... 56 10 8—11 2.VI ........................ 82 11 9—13 5.VII ..................... 43 32 26—37 12.VII ..................... 10 35 33—38 20.VII ..................... 53 40 33—46 9.VIII ................... 47 55 42—64 29.VIII ................... 12 68 54—79 25.IX........................ 32 73 57—87 7.XI ........................ 17 67 55—78 24.XI........................ 23 76 54—94 Growth of young-of-the-year The fry remain in individual broods, usually mixed with roach fry, for a period of about two months after hatching. The length at the end of this period is 2—3 cm. The total length of 1,081 young-of-the-year, collected at intervals of 2—3 weeks during 1963, 1965 and 1966, was measured to study the early growth (Table 8 and Fig. 3). In 1965 and 1966 samples were taken at seven main sites in the lower part of Kävlingeån covering an distance of about 10 km (vide Fig. 1). All the fish collected in one day were treated as a single sample, since there was no significant difference in mean length. The young-of-the-year were measured after fixation in 80 % alcohol. Total 68 PLUTARCQ CALA J I L NUMBER OF DAYS Fig. 3. Increase of total length of young-of-the-year ide, Kävlingeån, 1963, 1965 and 1966. Broken line connects the mean length at hatching with the first sampling each year. The material is listed in Table 8. body length was measured to the nearest mm. Shrinking due to fixation was corrected for (Cala). The ide fry were separated from the fry of other cyprinids, mainly that of roach, by means of the pigmentation in the ventral aorta (Balinsky 1948) or by analysing the pharyngeal teeth (Spillmann 1961). Considerable variation occurred in the growth patterns between the dif­ ferent years. The total length at the end of 2-week intervals after hatching was estimated from Fig. 3 (Table 9). The greatest total increase in one season was observed in 1963 and 1966, while the least growth was observed in 1965. The individual variation in total length was considerable after brood dispersal to deeper water. The inherent variation is probably increased by the delay and heterogenity of the spawning conditions. In order to compare growth in different years, the rate of growth during each 2-week period was calculated from the above material (Table 9). The growth rate was highest during periods 1—3 and then decreased. Growth rates were high for a somewhat longer time in 1966. It was concluded from the results of the year 1965 and 1966 (vide Table 9) ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 69 Table 9. Total length in mm of ide fry, calculated for successive 2-week periods, and growth rate during each period, in 1963, 1965 and 1966. The growt rate is expressed as percentage increase of length (in brackets). The material is based on Fig. 3. Week no 1963 1965 1966 2.............................. 12.0 11.0 13.0 (100.0) (83.3) (116.7) 4.............................. 19.5 17.5 20.0 (62.5) (59.1) (53.8) 6.............................. 27.5 24.0 27.0 (41.0) (37.1) (35.0) S.............................. 34.5 31.0 35.5 (25.5) (29.2) (31.5) 10.............................. 40.5 37.5 44.5 (17.4) (21.0) (25.4) 12.............................. 45.5 43.5 53.5 (12.4) (16.0) (20.2) 14.............................. — 49.0 61.0 (12.6) (14.0) 16.............................. — 53.5 67.5 ( 9.2) (10.1) 18.............................. — 58.0 71.5 ( 8.4) ( 5.9) 20.............................. — 61.5 74.0 ( 6.0) ( 3.5) 22.............................. — 64.0 75.5 ( 4.1) ( 2.0) 24.............................. — 65.5 76.0 ( 2.3) ( 0.7) 26.................. ........... — — 76.0 ( 0.0) that the growth season of the young-of-the-year in Kävlingeån ended in early November. The fry had then attained a mean total length between 6.8 and 7.6 cm for the respective years, and an individual length ranging from 5.5 to 9.5 cm. Stankovitch (1921) gave a total length of 10 mm for 15—20 days old fry and 13.5 mm for 6 week-old fry. Segerstråle (1933) reports a mean total length (on 6 October) of 4.4 cm with fry ranging from 3.6 to 5.1 cm. Age and increase in body length The age of ide was determined by examination of the scales. Fish were collected from March 17 to April 4, 1966 (294 specimens) and from November 11 to December 5, 1966 (61 specimens). In 1967, 79 ide were captured from February 27 to April 4. Collections were made mostly by seining or by electro-fishing. Some fish were collected in trap nets. Scales were collected from freshly captured fish and the total length and the body weight re­ gistered. Sex was determined by gross inspection or, in immature fish, by microscopic examination of the gonads. 70 PLUTARCO CALA Characteristics of the scale annulus The annulus consists of the usual interrupted ridges or circuli followed by a complete circulus which is laid down as growth resumes in the spring or early summer (Fig. 4). The annuli are easier to recognize in the cranial or anterior part of the scales. The first annuli are easier to recognize than the later ones (Fig. 5), which are especially difficult to recognise in scales from fish older than 10 years (Fig. 6). A growth field is made up by a series of complete circuli followed by a series of incomplete circuli that gradually shorten as growth slows down (Fig. 4). The first circulus that appears in the following growth season cuts across these shortened circuli. This pattern of growth is most apparent in the cranial part of the scales. Time of annulus formation Scanty material was available for the study of annulus formation (Table 10). The data suggest that annulus formation usually occur around early May and is completed about the end of June. Fig. 4. Scale of ide in age group II (specimen from 18.4. 1966), total length 163 mm, showing the “cutting over” of the circuli at annulus formation. ON THE ECOLOGY OF THE IDE IDUS IDTJS (L.) IN THE RIVER KÄVLINGEÅN 71 Fig. 5. Scale of ide in age group VII (specimen from 27.4. 1966), total length 440 mm, showing rather easily recognizable annuli. It seems that older ide, especially mature fish, form annuli earlier than the younger ones. Yasuda (1959) stated that the time of scale ring forma­ tion varies considerably. In a single species scale ring formation is influenced by geographical region, age, and degree of maturity. Yasuda concluded that the scale ring generally seems to be formed in an early period of reproduc­ tive maturation and/or in connection with spawning. According to Balon (1962) the annulus in scales of the Czech Danube ide may appear in April, but the main formation is in May and June. Regeneration of scales The regeneration of scales may be affected by environmental factors such as water temperature and food supply (Yamada 1961). 72 PLUTARCO CALA Fig. 6. Scale of ide in age group XII, (specimen from 1.5. 1966), total length 545 mm, showing annuli poorly recognizable in the marginal area. In 1966 yearlings were captured on 6 May and kept in the laboratory at about 18°C until the start of the experiment on 17 July. Primary scales adjacent to each other were removed from a selected area between the dorsal fin and the lateral line. Five fishes were kept at a mean temperature of about 18°C (warm water), and four fishes at a water temperature constantly changing, by means of a heater, between 4—11°G (cold water). The fish were fed with dry food and Cladocera. They were temporarily immobilized with ethyl m-aminobenzoate when measurements and observations were made. The mean length of the fish in warm water at the beginning of the experi­ ment was 69 mm and the mean weight 2.5 g. The mean length of the fish in cold water was 71 mm and mean weight 2.8 g. On 21 August the values Ta bl e 1 0. Id e s pe ci m en s w ith co m pl et e a nn ul i a t v ar io us co lle ct in g d at es in 19 66 an d 1 96 7. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 73 i|nuuB }no -qiiAV qsij s X i{nuuB q;iAV qsij g * UnuuB ino -xtiiAV qsij s qnuuB qiiAV qsij £ ijnuuB ino -qqAV qstj £ > qnuuB qqAV qsij £ qnuuB ;no -qqAV qsij £ > qnuiiB qsij x > qnuuB ino -qqAV qsij £ qnuuB qqAV qsij £ > qnuuB jno -qqA\ qsij £ qnuuB m* qsij s > qnuuB jno -qqAV qsij £ qnuuB qqAV qsij £ qnuuB jno -qiiAv qsij s l-H qnuuB q«* qsij £ qnuuB jno -qqM qsij £ qnuutf qqAV qsij £ qnuuB jno -qqAV qsij £ qnuuB qsij £ CO I HPÎ CtfT3 T3Öcö ueS>- 001>-!M^D(MTb(MO5CM tH (M i-H (M LO ffii Û, « d 74 PLUTABCO CALA Fig. 7. Regenerated scale, 6 weeks after the removal of the ontogenetic scales, showing circuli in the outer of the central area. were 72 mm and 2.8 g in warm water and 72 mm and 2.9 g respectively in cold water. Regenerated scales were observed two weeks after the removal of the primary scales in fish in warm water. The regenerated scales were thin and flexible and had just started the formation of circuli at their margin. Three weeks later 3—5 circuli were present at the scale margin. No regeneration of scales in cold water was observed. On 21 August (after 6 weeks) the regenerated scales in warm water were similar in appearence to primary scales and had about eight circuli in the outer part of the central area (Fig. 7). No regenerated scales were observed in fish maintained in cold water. On 5 September the regenerated scales from fish in warm water was observed to be very similar, even in size, to the primary scales. Since regeneration of scales in cold water did not occur, one of the fishes from this experiment was transferred to water with a temperature of about 18°C. This fish, like the others kept at the same temperature showed rege­ neration of scales after 2 weeks. When fish remaining in cold water did not show any regenerated scales on 1 November (about 16 weeks after start of the experiment) the temprature was increased and kept at about 13°C. Two weeks later regenerated scales had developed. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 75 Fig. 8. Regenerated scale from the same fish as in Fig. 4 showing a wide central area and numerous grooves. The regenerated scales are characterized by 1) a broad central area with­ out circuli but with a grooved network, 2) first circuli were set at broad intervals, and 3) numerous radial grooves in the cranial and caudal area of the scale, connected with the central network (Fig. 8). Body-scale relation Scale samples were taken from the side of the fish between the dorsal fin and the lateral line. Scales were impressed on plastic slides by a roller press (Smith 1954), and were examined with a microprojector at X30. Distance from scale focus to margin, along a selected cranial radius was registered and plotted against the body length (Fig. 9). The body-scale rela- 76 PLUTARCO CALA JUVENILE d’d1 ?? c© o ©' F5 160 0 80 160 240 320 400 480 560 TOTAL LENGTH IN mm Fig. 9. Relation between scale radius and total body length of material of 433 ide from River Kävlingeän, in 1966—1967. Table 11. Relation between total body length of fish and scale radius of ide from River Kävlingeån, 1966 and 1967. Length intervals in mm Number of fish Mean total body length in mm Mean scale radius X30 54--73 .............. 8 64 14 74--93 ............... 2 77 20 94--113 ............... 2 106 29 114--133 .............. 1 116 36 134--153 ............... 10 140 50 154--173 .............. 15 168 59 174--193 ............... 58 186 72 194--213 ............... 54 201 82 214—233 ............... 30 222 95 234--253 .............. 17 242 108 254--273 ............... 26 263 126 274 -293 .............. 16 283 135 294--313 ............... 13 304 152 314 -333 ............... 14 320 170 334--353 ............... 7 343 173 354--373 ............... 4 364 186 374--393 .............. 4 383 204 394--413 .............. 5 403 219 414--433 ............... 17 426 225 434--453 ............... 32 445 241 454—473 ............... 40 465 248 474--493 ............... 35 483 251 494--513 ............... 12 499 258 514--433 ............... 3 522 267 ON THE ECOLOGY OF THE IDE IDUS ID US (L.) IN THE RIVER KÄVLINGEÅN 77 -106.37 + 2.3236 S L = 80,77 + 1.4720 S L = 35.92 + 2.1014 S 160 240 320 400 TOTAL LENGTH OF FISH IN mm Fig. 10. Relation between total body length and scale radius (X 30) of ide from River Kävlingeån, in 1966 and 1967. tionship was obtained by plotting the means of total fish lengths (by 2 cm intervals) against the means of the scales measurements (Table 11). The size-class means (Fig. 10) indicate the body-scale relation to be des­ cribed adequately by three straight lines intersecting at a length of about 16.6 and 40.3 cm. Thus, the body-scale curve may be subdivided into 3 components: 1) a linear portion representing ide up to 15.3 cm long, 2) a linear portion representing fish between 15.4 and 40.3 cm long, and 3) the linear portion representing larger ide (40.4—53.3 cm). The straight lines fitted by least squares to the means of fish lengths and scale radii had these equations : L= 35.92 + 2.1014S (LC15.3 cm), L= 80.77 + 1.4720S (15.340.3 cm), 78 PLUTARCO CALA Table 12. Length distribution of 433 age-grouped ide from the River Ivävlingeån, in 1966—1967. Total length Age Group in mm I II III IV V VI VII VIII IX X XI XIII XIV 56— 75 8 _ ____ ____ ____ ____ 76— 95 2 — — — — — — — — — ■--- — — 96—115 1 1 — — — — — — — — ---- — — 116—135 — 1 — — — — — — — — — — — 136—155 --- 8 2 — — — — — — — — — — 156—175 — 3 18 — — — — — — — — — — 176—195 — — 64 1 — — — — — — — ---- — 196—215 — — 52 — — — — — — — — — — 216—235 — — 22 2 — ■---- — — — — — — — 236—255 — — 14 4 — — — — — — — — — 256—275 — — 16 11 — — — — — — — — — 276—295 — ---- 4 11 — — — — — — ---- — ---- 296—315 — — — 15 1 — — — — — — — — 316—335 — — — 8 3 1 — — — — — — — 336—355 — — — 4 6 — — — — — — — — 356—375 — — — — — 1 — — — - - * — — — 376—395 — — — — 2 3 — — — — — — — 396—415 — — — — — 2 4 1 — 1 — — — 416—435 — — — — — «J 6 2 6 1 — — — 436—455 — — — — — 2 7 13 6 3 ---- — — 456—475 — — — — — — 3 18 8 7 4 — — 476—495 — — — — — 1 4 11 6 7 4 1 — 496—515 — — — — — — 1 2 2 4 1 — ■--- 516—535 — — — — — — — — — — — — 1 2 fish 11 12 192 56 12 13 25 48 28 24 9 1 1 °/o of total 2.5 2.8 44.3 12.9 2.8 3.0 6.0 11.1 6.5 5.5 2.1 0.2 0.2 Mean length 70 143 206 288 348 413 443 462 467 472 480 470 530 where L equals total length of fish in mm and S equals length of anterior scale radius (X30) in mm. Length distribution The length distribution of the 433 ide whose ages were determined, is shown in Table 12. The values are lengths at capture when the season’s growth had not yet started. There was a wide amplitude in length in each age group and extensive overlapping between adjacent age groups. The greatest difference between the longest and the shortest fish in any age group was 17.9 cm (groups IV and VI). Age groups VII, VIII and X had range of 11.9 cm. The well-represented age group III had a range of 15.9 cm. The well-represented size classes contained fish from 2 to 7 age groups. The largest ide (54.5 cm in total length) registered during the four years of study was a male captured on 2 May 1966. It is evident that the body length is a poor indication of age of ide from the River Kävlingeån. Similar conditions were illustrated by Balon (1962) for ide from the Danube. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 79 Table 13. Calculated total lengths of male ide from River Kävlingeån, 1966 and 1967. The last value in each group is the average length at capture. Length in mm at the end of the respective year of life l|2|3|4|5|6|7|8|9|l0|ll III ................... 15 74 139 200 — — - — — — — — IV ................... 21 86 163 227 292 — — — — — — — V 7 64 154 237 297 351 — — — — — — VI 7 80 171 251 321 370 404 — — — — — VII 14 77 181 263 326 374 414 454 — — — — VIII 19 79 169 245 307 356 397 435 467 — — — IX 11 85 170 250 312 356 390 426 455 477 — — X 6 82 167 230 277 319 358 396 424 450 467 — XI 1 94 176 235 287 337 354 368 402 444 464 480 Age group Number of fish Grand average calculated length 80 165 237 302 347 386 416 437 457 465 480 Increment of average 80 85 72 65 45 39 30 21 20 8 15 Grand average increment of length 80 85 72 60 48 34 33 31 30 18 16 Sum of average increment 80 165 237 297 345 379 412 443 473 491 507 Growth in length The growth under previous years of life was calculated from the relation of scale measurements to total length (vide Fig. 10). For each fish (433 specimens) the following measurements were made: 1) distance from the focus of the scale to each annulus, 2) length of scale radius, and 3) total length of the fish at the time the scale sample was taken. The body length at the end of any year of life for a specimen was cal­ culated from Fig. 10. The body length at the time of sampling gives an expected value for the scale radius. The ratio “expected” (R) : “observed” (S) value is used as a correction factor ( — ) to adjust the radii observed for different years. The corrected scales are referred to the corresponding lengths which are thus the estimated body length of the fish at the end of the respective year of life. The calculated average lengths for males (101 specimens) and females (132 specimens) showed no significant differences (Tables 13 and 14). Data for males and females of the whole studied ide material (433 specimens) were combined in Table 15. Calculated lengths exhibited a successive change in growth with increasing age. The discrepancies among calculated growths of the various age groups are approximately randomly distributed. The discrepancies between the older age groups may be attributed to small numbers of fish. The estimates of the general growth in length at the bottom of Tables 13, 80 PLUTARCO CALA Table 14. Calculated total lengths of female ide from River Kävlingeån, 1966 and 1967. The last value in each group is the average length at capture. Age group Number Length in mm at the end of the respective year of life of fish 1 1 2 3 1 4 1 5 6 7 8 9 10 11 12 13 14 III ..................... 14 82 144 207 IV ..................... 22 87 177 238 297 __ __ __ __ __ __ __ __ __ __ V ..................... 5 72 146 221 285 344 __ __ __ __ __ __ __ __ — VI ..................... 6 78 161 240 312 373 423 __ __ VII ..................... 12 88 186 256 312 367 401 429 — — — — — — — VIII ..................... 28 86 187 256 315 362 402 434 460 — — — — — — IX ..................... 17 81 164 230 290 335 377 411 436 460 — — — — --- X ..................... 18 82 161 232 291 339 378 408 431 454 473 __ __ __ — XI ..................... 8 77 146 211 262 312 358 397 422 443 463 480 __ __ __ XIII ..................... 1 68 101 141 173 223 257 300 346 389 420 441 463 470 — XIV ..................... 1 99 188 243 272 302 340 384 418 446 473 495 506 519 530 Grand average calculated length 82 160 225 281 328 367 395 419 438 457 472 484 494 530 Increment of average 82 78 65 56 47 39 28 24 19 19 15 12 10 36 Grand average increment of length 82 78 65 54 49 40 36 30 28 24 20 16 10 11 Sum of average increment 82 160 225 279 328 368 404 434 462 486 506 522 532 543 Table 15. Calculated total lengths of ide from River Kävlingeån, 1966 and 1967. The last value in each group is the average length at capture, as the season’s growth had not begun. Age group Number Length in mm at the end of the respective year of life of fish 1 2 1 3 4 5 6 7 8 9 10 11 12 13 14 I ..................... 11 70 II ..................... 12 71 143 — — — — — — — — — — — — Ill ..................... 192 79 144 206 — — — — ---- — — — — — — IV ..................... 56 83 165 227 288 — — __ — — — — — — — V ..................... 12 67 151 231 292 348 _ _ _ _ _ __ __ __ — VI ..................... 13 79 166 246 317 371 413 _ _ __ __ _. — __ — VII ..................... 26 82 183 260 320 371 408 443 VIII ..................... 48 83 179 252 312 360 399 434 462 — — — — — — IX ..................... 28 83 167 238 300 345 383 417 444 463 __ _ __ — --- - X ..................... 24 82 163 231 287 334 373 405 429 453 472 __ __ — __ XI ..................... 9 79 150 213 265 315 357 394 420 443 463 480 — — — XIII ..................... 1 68 101 141 173 223 257 300 346 389 420 441 463 470 — XIV ..................... 1 99 188 243 272 302 340 384 418 446 473 495 506 519 530 Grand average calculated length 79 158 226 283 330 366 397 420 439 457 472 484 494 530 Increment of average 79 79 68 57 47 36 31 23 19 18 15 12 10 36 Grand average increment of length 79 79 66 54 48 39 37 31 27 24 20 16 10 11 Sum of average increment 79 158 224 278 326 365 402 433 460 484 504 520 530 541 ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 81 Fig. 11. Increase in length (accumulative) of ide in River Kävlingeån (com­ plete line) and yearly in­ crement (broken line) ; material (433 specimens) collected in 1966—1967. i -y--—» YEAR OF LIFE 14 and 15 are based on the grand average calculated lengths and on the successive summation of the grand average increments of length. Lengths estimated by the two methods were nearly identical for ages 1 through 8 for males and ages 1 through 6 for females. The sums of the grand average increments were used to demonstrate the growth of the ide in Fig. 11. Male and female ide show their greatest annual increase in length (85 mm and 78 mm respectively) during the second year of life. After the second year the increments decreased to 30 mm (males) and 28 mm (females) in the 9th year and 16 mm (males) and 20 mm (females) in the 11th year. The low representation of male ide in the older age groups may indicate different mortality of the sexes. In the present material there is a tendency towards more rapid growth of the ide males in the River Kävlingeån than the females. Balon (1962) stated that the ide males in Little Danube River grew more quickly, but from the 5th year onwards the females grew more rapidly. Length-weight relation This study is based on data from 989 young-of-the-year ide collected from May to November during 1963—1966 (vide Table 8), and from 428 yearling and older fish collected during the winter months. In 1966, 266 ide 6 82 PLUTABCO CALA O 1200 160 320 480 TOTAL BODY LENGTH (mm) Fig. 12. Length-weight relation for males (o) and females (•) of ide. were caught from March 17 to April 17, and 82 from November 11 to December 5. In 1967, 80 fishes were collected between February 27 to April 4. Collections were made mostly by seining and trap nets, or by electro­ fishing especially for young fish. Lengths were measured to the nearest mm and weight of whole fish to the nearest gram, and to the nearest 0.01 g for the young-of-the-year. Measurements of the fish were made as soon as possible after their being collected with exception of the young-of-the-year which were fixed in 80 % alcohol. Shrinkage and weight lose, caused by the fixative, was studied and correction made (Cala). Maximum size re­ corded for females was 53 cm total length and 2.33 kg (one specimen ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 83 Fig. 13. Length-weight relation of ide (excepting young-of-the-year) from Riv­ er Kävlingeån. The curve represents the calculated weights and the dots the empirical weights. 2000 - .21200 $ 800 160 320 480 TOTAL BODY LENGTH (mm) 20 40 6C FISH LENGTH (mm) Fig. 14. Length-weight relationship of the young-of-the-year ide from River Kävlingeån, 1963—1966. The curve rep­ resents the calculated weights and the dots the empirical weights. The material is listed in Table 16. 80 84 PLUTARCO CALA Table 16. Length-weight relation of young-of-the-year ide from River Kävlingeån, 1963—1966. (The lengths and weigths are the true means for fish in 2-mm groups). Total length intervals (mm) Number of fish Mean total length (mm) Mean weight 10--11 ..................... 98 10.2 0.004 12--13..................... 44 12.1 0.01 14--15..................... 24 14.2 0.03 16--17..................... 98 16.4 0.03 18--19..................... 80 18.2 0.05 20--21..................... 10 20.2 0.06 22--23..................... 5 22.9 0.09 24--25..................... 15 24.7 0.13 26--27..................... 33 27.5 0.18 28--29..................... 27 28.9 0.21 30--31 ..................... 34 31.1 0.27 32--33..................... 46 33.1 0.35 34--35..................... 65 35.2 0.40 36--37..................... 57 37.4 0.48 38--39..................... 55 39.2 0.55 40--41..................... 29 40.2 0.62 42--43..................... 19 42.3 0.78 44--45..................... 15 44.4 0.88 46--47..................... 13 46.4 1.00 48--49..................... 20 48.7 1.23 50--51 ..................... 7 50.7 1.40 52--53..................... 13 52.5 1.38 54--55..................... 13 54.5 1.57 56--57..................... 16 56.5 1.75 58--59..................... 19 58.7 1.77 60--61 ..................... 14 60.6 1.95 62--63..................... 23 62.9 1.96 64--65..................... 10 64.3 2.22 66--67..................... 22 66.8 2.40 68--69..................... 12 68.8 2.61 70--71 ..................... 11 71.1 2.95 72--73..................... 13 73.2 3.05 74--75..................... 12 74.7 3.28 76--77..................... 4 76.8 3.50 78--79..................... 6 79.2 3.78 80--81 ..................... 4 80.9 3.88 82--83..................... 1 83.6 5.00 84--85..................... 5 84.6 4.58 captured on 31 March 1967). The largest male measured 52 cm total length and weighed 1.82 kg (one specimen captured on 18 April 1966). Examination of length-weight data (Fig. 12) in relation to sex, sexual maturity or state of gonads revealed no differences. Determination of the length-weight relation for ide older than age group 0 according to Fig. 13 may be possible only for certain times of the year. Ide of a given length captured during winter — December to March — seem to be consistently heavier than fish taken at other times of the year. Periods of such conditions in advance of spawning have been established for coregonid species (Deason and Hile; and Van Oosten and Hile, cited by Bailey 1964) . ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVEH KÄVLINGEÅN 85 Table 17. Length-weight relation of ide from River Kävlingeån, 1966—1967. (The lengths and empirical weigths are the true means for fish in 2-cm groups; the calculated weights are from the equation shown in Fig. 13). Total length intervals (mm) Number of fish Mean total length (mm) Weight (g) empirical calculated 54— 73 .............. 8 64 2 2 74— 93 ............... 14 82 4 4 94—113 .............. 3 102 8 8 114—133 .............. 1 116 13 — 134—153 ............... 7 142 25 26 154—173 .............. 15 168 44 45 174—193 .............. 58 185 62 62 194—213 ............... 55 201 77 82 214—233 .............. 32 222 108 114 234—253 .............. 17 242 153 152 254—273 .............. 26 262 197 198 274—293 .............. 17 283 257 256 294—313 .............. 13 304 341 326 314—333 .............. 14 320 405 386 334—353 .............. 6 344 405 492 354—373 .............. 5 365 446 599 374—393 .............. 4 383 707 704 394—413 .............. 5 403 785 834 414—433 .............. 18 426 1,018 1,004 434—453 ............... 29 445 1,217 1,161 454—473 .............. 36 465 1,354 1.346 474—493 .............. 32 482 1,444 1,516 494—513 .............. 10 499 1,641 1,702 514—533 .............. 3 522 2,008 1,980 The length-weight relation of ide was linear when both variables were transformed to common logarithms. But it seems that the length-weight rela­ tion for ide fry below 13 mm in average length (shortly after hatching to the time of yolk-sac absorption, or beginning of feeding) is different (Fig. 14), although the material does not adequately prove it. For this reason, fish below 12 mm in total length were not included in the following computation of the length-weight relation. The general parabola, W = cln, where W=weight, L = total length, and c and n are empirically estimated constants has been used to describe the length-weight relation of numerous species of fish. The least squares regres­ sion of the logarithmic transformation (Y = a + bX in which Y = logio W, a = logio c, and X = logio L) was used for estimating values of c and n. The length-weight equation, determined by fitting a straight line to the logarithms of the lengths and weights listed in Tables 16 and 17, are: log W=—5.11749 + 3.03175 log L (young-of-the-year) and log W=—5.77664 + 3.33853 log L (yearling and older ide) Where W = body weight in g and L = total length in mm. The value of n indicates that the body weight increases approximately as the cube of the 86 PLUTARCO CALA weights and annual increments o the River Kävlingeån. Year of life Calculated weight (g) Increment 1 ............ .. . 4 4 2 .................. 37 33 3 .................. 117 80 4 .................. 241 124 5 .................. 411 170 6 .................. 599 180 7 .................. 827 228 8 .................. 1,060 233 9 .................. 1,298 238 10 .................. 1,537 239 11 .................. 1,760 223 12 .................. 1,953 193 13 .................. 2,082 129 14 .................. 2,229 147 length for young-of-the-year longer than 12 mm in length and more rapidly than the cube of the length for yearling and older ide. The equations make possible the calculation of the unknown weight for a known length and vice versa. It becomes evident that the total length of an ide at a weight of 1 g is 52 mm and at a weight of 1 kg 453 mm. These results essentially agree with those of Balon (1962) who found no difference of length-weight relations between the juvenile specimens and males and females. Balon gives the standard length of an ide of 1 g as 39 mm and of 1 kg 361 mm. Calculated growth in weight The data of Table 17 indicate a generally good agreement between empiri­ cal and calculated weights. The greatest discrepancies (72,61,56 and 49 g) were at 474—493, 494—513, 434—453 and 394—413 mm, respectively, i.e. among the larger fish. The calculated weights of Table 18 were computed by applying the cal­ culated lengths (sum of the average increments) of Table 15 to the length- weight equation. The annual increments of weight thus increased to a maximum of 239 grams in the 10th year of life, despite the fact that the greatest annual in­ crements of length were attained during the 2nd year. Annual weight incre­ ments declined to 129 g in the 13th year. In the River Kävlingeän, ide require 8 years to reach a body weight of about 1 kg. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 87 Table 19. Sex ratio of mature ide taken on the spawning grounds, in the River Kävlingeån, during 1964. Date Number of fish c3(3 ! 95 1 Total 26.III............................ 15 15 30 l.IV ............................ 78 90 168 5.IV ............................ 5 12 17 6.IV ............................ 237 262 499 9.IV ............................ 53 60 113 15.IV ............................ 31 11 42 16.IV ............................ 129 66 195 24.IV ............................ 57 61 118 2.V ............................ 16 9 25 2 ................................... 621 586 1,207 «/o ................................ 51.4 48.6 XIII. Sex ratio on spawning ground A sample of 1207 adult ide taken just before and during the spawning season in 1964 between 1 April and 2 May was composed of 621 (51.4 %>) mature males and 586 (48.6%) mature females (Table 19). These numbers 120 xcn u_ Lu O cr 1x1 CD 80 40 JÜ£ 38 D MALES I FEMALES JLfl. JL 42 46 LENGHT IN cm 50 54 Fig. 15. Sex ratio and length distribution of 1208 mature ide captured in the lower reaches of the River Kävlingeån just before and during spawning in 1964. 88 PLUTARCO CALA Table 24. Food items taken by the ide in the River Kävlingeån, 1963—1967 ROTATORIA INVERTEBRATE EGGS ANNELIDA OLIGOCHAETA H1RUD1NEA CRUSTACEA CLADOCERA OSTRACODA COPEPODA ISOPODA Asellus aquaticus L. AMPHIPODA Gammarus spp. ARACHNID A HYDRACARINA INSECTA EPHEMEROPTERA (nymph) ODONATA (nymph) HETEROPTERA CORIXIDAE TRICH OPTERA (larva) COLEOPTERA DIPTERA CH1R0N0MIDAE (larva, pupa and imago) CERATOPOGONIDAE (larva) SIMULIIDAE (larva and pupa) OTHER DIPTERA MOLLUSCA GASTROPODA Lymnaea spp. Physa fontinalis (L.) Planorbis carinatus MÜLL. Anisus vortex (L.) Bathyomphalus contortus (L.) Gyraulus spp. Viviparus fasciatus Leach Acroloxus lacustris (L.) Bithynia tentaculata (L.) B. leachi (Sheppard) Hydrobia sp. Valvata macrostoma (Steenbüch) V. piscinalis (Müll.) PELECIPODA Mytilus edulis L. OSTEICHTHYES Perea fluviatilis (L.) (young) Rutilus rutilus (L.) (young) Idus idus (L.) (egg and young) ALGAE (Cladophora, etc.) HIGHER PLANTS PLANT TISSUE Potamogeton spp. Lemma minor Other higher plants SEEDS (Nymphaea alba, Carix, etc.) DETRITUS, MINERAL MATTER, ETC. do not deviate significantly from a 1 : 1 ratio (x2 = 0.30). The male spawning population had a mean length of 45.6 cm, ranging from 39 to 52 cm (mode 47 cm, n = 99), and the female 45 cm, ranging between 38—51 cm (mode 45 cm, n= 108) (Fig. 15). XIV. Food habits 1,163 ide from the lower reaches of the River Kävlingeån — collected June 1963 to December 1966 — were examined with regard to the contents of the upper part of the S-shaped digestive tract, from the pharynx to the first loop. The food items in the rest of the intestine are strongly digested with excep­ tion of crustacean zooplankters in fry ide which can be easily recognizable in the whole digestive tract. The young-of-the-year were fixed in 80 °/o alco­ hol and later dissected for analysis. The intestines of juvenile and adult fish were preserved in 80 °/o alcohol as soon as possible after capture. Total body length was registered consistently. The occurrence method was used in assessing the food. After the percent­ age of the number of fish examined containing the individual food items is found, the number of occurrences of the food items is summed to give the percentage composition of the diet. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 89 Fish such as cyprinids chew their food (Hartley 1947, cited from Hynes 1950) and so make counting impossible. The presence of large volumes of mucus prevents accurate measurement of weight or volume. The investigator can therefore only use the occurrence method, which although apparently the least satisfactory was shown to give substantially the same results as theoretically better methods (Hynes 1950). Organisms represented in the intestines are shown in Table 24. The composition of the food based on occurrence, of the different size groups and of the total number of fish are seen from Table 25. Composition of the diet in relation to the fish size In the present material there were definite changes in food habits with increase in size which became evident at about 20 mm total length, and at about 140 mm (Table 25). Crustacea were most important in the first group (10 to 20 mm total length), but insects predominated in the second group (20 to 140 mm), especially chironomid larvae. Rotifers were taken by a considerable number of fish, but only by fry below 12 mm. Cladocera were caught by fish between 10—80 mm and Copepoda until 60 mm. If any zooplankter is taken by larger fish it seems to be by mere chance, e.g. when higher plant tissue or algae are eaten. Medium-sized or large organisms such as Asellus aquaticus, molluscs and Oligochaeta were rarely found in young-of-the-year ide (less than 80 mm total length). Higher plant tissue and detritus were also rare in this size group. Fish did not appear except in ide larger than 240 mm total length. In age groups just larger than 140 mm several food items became impor­ tant such as higher plants. Asellus aquaticus, and Oligochaeta. Insecta con­ tinued to be an important group. It seems, however, that some Insecta (Ephemeroptera, Coleoptera and Ceratopogonidae) disappeared from the diet and Chironomidae become less dominant as the fish grow and at above 260 mm they are replaced by higher plant tissue (Table 25). Food utilization and food abundance Simultaneous collections of fry and plankton were made during 1964 and 1965 at the time when the crustacean zooplankters were the main food items of ide fry from about 10 to 25 mm long. Collecting methods and habitats are described above (vide pp. 47, 55, and Fig. 1). Estimated numbers of Cladocera and Copepoda varied from 6 to 1,834 per litre water (Table 26). Relative numbers of these groups varied consi­ derably, with Cladocera being the most abundant (vide also Table 4). The occurrence of these groups in the fish intestines (Table 26) shows that the ide in its early life does not specialize on either Cladocera or Copepoda, but 90 PLUTARCO CALA Table 25. Food of 1,163 ide from River Kävlingeån, 1963—1966, listed containing the various food items. II. Total length interval in mm N um be r o f fi sh % em pt y s to m ac hs M et ho d o f as se ss m en t Ro ta to ria V er te br at e eg gs O lig oc ha et a H iru di ne a Cl ad oc er a O str ac od a Co pe po da As el lu s a qu at ic us L . G am m ar us spp . H yd ra cc tri na Ep he m er op te ra (n ym ph ) 10— 20 ......................... 150 6.0 I 7.3 1.3 75.3 24.7 II 5.8 1.1 — — 59.8 — 19.6 — — — — 21— 40 ......................... 325 55.1 I — — — — 7.1 — 1.2 — — — 0.9 11 — — — — 14.1 — 2.4 — — — 1.8 41— 60 .......................... 144 31.9 I — 1.4 1.4 — 4.9 — 2.1 0.7 — — 2.1 II — 1.4 1.4 — 4.9 — 2.1 0.7 — — 2.1 61— 80 ......................... 117 35.0 I — — 8.5 — 0.8 — — 3.4 ---- 0.8 0.8 II — — 10.0 — 1.0 — — 4.1 — 1.0 1.0 81—100 ......................... 8 62.5 I — — — — — — — — — — — 101—120 .......................... 6 33.3 I — — — — 121—140 .......................... 25 36.0 I __ __ 4.0 __ __ __ __ 20.0 __ __ 4.0 II — — 2.7 .--- — — — 13.5 — — 2.7 141—160 .......................... 23 26.1 I — — 8.7 — — — — 30.4 4.3 — — II — — 5.1 — — — — 17.9 2.6 — — 161—180 ....................... .. 45 17.8 I — — 13.3 — 2.2 — — 26.7 — — 2.2 II — — 7.8 — 1.3 — — 15.6 — — 1.3 181—200 ......................... 56 28.6 I — — 5.3 1.8 — — — 28.6 — — 1.8 II — — 3.0 1.0 — — — 16.2 — — 1.0 201—220 ......................... 40 27.5 I — — 5.0 — — — — 12.5 2.5 — 2.5 II — — 3.4 — — — — 8.6 1.7 — 1.7 221—240 .......................... 29 31.0 I — — 6.9 — 3.4 — — 20.7 — — — II — — 4.3 — 2.1 — — 12.8 — — — 241—260 .......................... 18 55.5 I — ---- 5.5 ---• — — — 11.1 — — — 11 — — 6.7 — — — — 13.3 — — — 261—280 ......................... 49 55.5 I — — 8.2 — — — — 10.2 — — — II — — 10.8 — — — — 13.5 — — — 281—300 ......................... 18 50.0 I — 5.5 16.7 — — — — 16.7 5.5 — 5.5 II — 5.3 15.8 — — — — 15.8 5.3 — 5.3 301—320 ......................... 15 53.3 I — 6.7 20.0 — 6.7 — — 6.7 — — — II — 6.7 20.0 — 6.7 — — 6.7 — — — 321—340 ......................... 5 0.0 I — — — — — — — — — — — II — — 341—360 .......................... 2 — 1 — — 4- — — — — + — — + II — — + — — — — + — — + 361—380 ......................... 1 — I — — — — — — — — — — — 381—400 .......................... 2 — I — — — — — — — — — — — 401—420 ......................... 7 28.6 I __ __ 14.3 __ __ __ — — — — — II — — 12.5 ---- — — — — — — — 421—440 .......................... 11 45.4 I — — — — — — — — — — — 441—460 .......................... 18 50.0 I _ _ 11.1 _ _ 5.5 _ 16.7 _ 5.5 ____ II — — 8.7 — — 4.3 — 13.0 — 4.3 — 461—480 .......................... 29 55.2 I — 6.9 13.8 ---- — — ---- — — — — II — 10.5 21.0 — — — — — — — — 481—500 .......................... 19 63.1 I — — 10.5 — — — — 5.3 — — — II — — 11.8 — — — — 5.9 — — — 501—520 .......................... 3 — I — — — — — — — — + — — II — — — — — — — — ~r — — Total I 0.9 0.7 4.0 0.1 12.6 0.1 3.8 6.2 0.3 0.2 1.1 10—520 ....................... 1163 37.9 11 1.0 0.7 4.1 0.1 12.8 0.1 3.8 6.3 0.3 0.2 1.1 ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 91 according to the occurrence method (Hynes 1950). I. Percentage of fish Percentage composition of the food. d C 2 03a3 C Cy pr in id s (ro ac h & ide ) *-< % t-Ta; O do na ta (n ym ph ) Co rix id ae Tr ic ho pt er ci (la rv a) Co le op te ra Ch iro no m id c (la rv a, pu pa an d im ag o) Ce ra to po go n (la rv a) Si m ul iid ae (la rv a, a nd j O th er Ins ec t G as tro po da (o ne oc cu rre M yt ilu s ed ui •S'S3 c ÛH Eg gs ide Al ga e C3 § C s s Se ed s H ig he r pl an ! m at er ia l D et rit us , m in er al ma tt U ni de nt ifi ed — an im al — __ _ __ __ 14.0 __ __ 0.7 __ __ __ __ __ __ __ 2.7 __ — — — — 11.1 — — 0.5 — — — — — — — 2.1 — — — — — 1.5 34.5 — 2.5 1.5 — — — — — — — 0.6 0.3 — — — — 3.1 68.7 — 4.9 3.0 — — — — ---- — ■--- 1.2 0.6 — — 1.4 — 8.6 39.4 — 14.1 20.4 0.7 — — — 0.7 — — 0.7 1.4 — — 1.4 — 8.6 39.4 — 14.1 20.4 0.7 — — — 0.7 — — 0.7 1.4 — ---- ---- — 4.3 28.2 0.8 5.1 15.3 — — — — — — 3.4 0.8 6.0 4.3 — — — 5.2 34.0 1.0 6.2 18.5 — — — — — — 4.1 1.0 7.2 5.2 — — 12.5 — 37.5 — — — — — — — — — — — 12.5 — — — 20.0 — 60.0 — — — ---- — — — ---- — — — 20.0 — 16.7 — — 16.7 16.7 — 16.7 83.4 — — — — — — — — — — 11.1 — — 11.1 11.1 — 11.1 55.5 — — — — — — — - - - — ---- 4.0 ---- 8.0 32.0 — 8.0 24.0 — — — — — 8.0 4.0 24.0 8.0 — — 2.7 — 5.4 21.6 — 5.4 16.2 — — — — ---- 5.4 2.7 16.2 5.4 — — — 4.3 4.3 39.1 4.3 4.3 — 4.3 — — — 4.3 13.0 4.3 26.1 8.7 8.7 — — 2.6 2.6 23.1 2.6 2.6 — 2.6 — — — 2.6 7.3 2.6 15.4 5.1 5.1 — 11.1 — 2.2 28.9 — — 6.7 4.4 — — — — 8.9 6.7 22.2 17.8 17.8 — 6.5 — 1.3 16.9 — — 3.9 2.6 — ---- — — 5.3 3.9 13.0 10.4 10.4 — 1.8 1.8 5.3 32.1 — 5.3 7.1 14.3 — — — — 5.3 — 37.5 26.8 1.8 — 1.0 1.0 3.0 18.2 — 3.0 4.0 8.1 — — — — 3.0 — 21.2 15.1 1.0 — — 2.5 2.5 37.5 — 7.5 7.5 7.5 — — 2.5 — 7.5 ---- 35.0 7.5 5.0 — — 1.7 1.7 25.9 — 5.2 5.2 5.2 — — 1.7 — 5.2 — 24.1 5.2 3.4 — — 3.4 — 55.2 3.4 17.2 3.4 6.9 — — — — ---- — 20.7 20.7 — — — 2.1 — 34.0 2.1 10.6 2.1 4.3 — — — — — — 12.8 12.8 — — — — — 11.1 — ---- 5.5 — — 11.1 — 5.5 — — 11.1 11.1 11.1 — — — — 13.3 — — 6.7 — — 13.3 — 6.7 — — 13.3 13.3 13.3 — — 2.0 2.0 6.1 — 2.0 — 4.1 — — — 2.0 2.0 ---- 20.4 16.3 — — — 2.7 2.7 8.1 — 2.7 — 5.4 — ---- — 2.7 2.7 — 27.0 21.6 ---- — 5.5 — — 5.5 — 5.5 5.5 5.5 — 5.5 — — — — 22.2 — — — 5.3 — — 5.3 •--- 5.3 5.3 5.3 — 5.3 ---- — — — 21.0 — — — — — — 13.3 --- - 6.7 — 6.3 6.7 13.3 — — — — 6.7 6.7 — — — — — 13.3 — 6.7 — 6.3 6.7 13.3 — — — — 6.7 6.7 — — — — — 20.0 — 20.0 — — — — — — — — 80.0 20.0 — — — — — 14.3 — 14.3 — — — — — — — — 57.1 14.3 — — — + ---* + + 1 1 — + — — — — — ~r — — + — — — — — — — — — + + — — — + + 28.6 — — — — — — 14.3 — 14.3 — — — — 28.6 — — 14.3 — — — — — 12.5 — 12.5 — — — — — 25.0 — — 25.0 12.5 — — — — — 9.1 — 18.2 — — — — — — — — 36.4 18.2 9.1 — — — — 10.0 — 20.0 — — — — — — — — 40.0 20.0 10.0 — 5.5 5.5 — 11.1 — — 5.5 16.7 — — 11.1 — — 27.8 — 5.5 — 4.3 4.3 — 8.7 — — 4.3 13.0 — — — 8.7 — — 21.7 — 4.3 — — — — 6.9 — 6.9 — — — 3.4 3.4 — — 3.4 13.8 — 6.9 — — — — 10.5 — 10.5 — — — 5.3 5.3 — — 5.3 21.0 — 10.5 — 5.3 — — 15.8 — 5.3 5.3 10.5 — 5.3 — 5.3 — — 21.0 — — — 5.9 — — 17.6 — 5.9 5.9 11.8 — 5.9 — 5.9 — — 23.5 — — — — — — + — — -1- + 4- — — — — — + + — — — — — + — + 4- + + + — 0.1 1.0 0.7 3.0 28.4 0.3 5.4 7.2 2.3 0.2 0.8 0.2 0.9 1.4 0.9 9.8 5.3 2.1 0.1 1.0 0.7 2.8 28.3 0.3 5.2 7.0 2.4 0.2 0.8 0.2 0.9 1.4 0.9 9.8 5.4 2.1 92 PLUTARCO CALA Table 26. Crustacea, Copepoda and Cladocera in quantitative plankton samples compared with amounts in intestines of fry ide collected at the same time and same place. Length of fry ide 10—20 mm. River Kävlingeån, localities described on p. 49. Total number of Crustacea Per cent Date Locality water (per litre) intestine Cladocera Copepoda water intestine water intestine 196 A 28.V lb 1,831 191 2.95 7.8 97.05 92.2 16.VI la 154 11 91.6 100.0 8.4 0.0 — lb 26 0 92.3 — 7.7 — „ — Ie 6 1 100.0 100.0 0.0 0.0 1965 14.VI II 252 228 84.5 100.0 15.5 0.0 w --- IV 818 228 76.3 100.0 23.7 0.0 M --- V 25 397 100.0 100.0 0.0 0.0 — VI 294 198 96.9 100.0 3.1 0.0 l.VII II 307 22 96.4 100.0 3.6 0.0 20.VII II 544 18 90.1 100.0 9.9 0.0 there seems to be a tendency for fry to feed more heavily on Cladocera when these organisms are abundant. As stated above (p. 55), collections of benthic fauna were made when electro-fishing, in order to survey bottom-living animals which might be of some interest as fish food. The material is listed in Table 3. The most frequent organisms were Chironomidae, Gastropoda, Corixidae, Asellus aqua- ticus, Ephemer opter a, Hirudinea, Oligochaeta, Simuliidae, Coleoptera and Hydracarina. The importance of the vegetation was discussed on page 61. The major food items of the ide in the lower part of the river (cf. Table 25) were in order of importance Chironomidae, higher plant tissues, A. aquaticus, lnsecta (excluding specified groups), detritus, Simuliidae, Oligochaeta, Cole­ optera, Gastropoda, unidentified animal matter, Lemna minor, Ephemerop- tera and Corixidae. There seemed to be no definite preference in the food choice. Some groups, viz. Corixidae, Hirudinea and Hydracarina were well represented in the bottom samples but not in the intestines. The reason seems to be that the Corixidae concentrated in shallow marginal water with dense vegetation where ide rarely occur. Seasonal changes There are small seasonal variations of the feeding habits of the ide. Plant material as Lemna minor, Potamogeton and seeds are eaten mainly in sum­ mer and early autumn, when the vegetation is luxuriant. Oligochaeta seem ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 93 to be more frequent in the diet in winter (November to March). It should be noted that large ide spend most of the summer (May to September) in the sea (see pp. 62—63) where their diet consists of marine matter. Summing up, the ide is almost omnivorous in the River Kävlingeån, taking the most frequent food material that it encounters. Previous investigations Ekström (1831) mentioned that the ide in Lake Mälaren, Central Sweden fed especially on vegetation, but in fish from deeper water he found remains of insects, Crustacea and sometimes small fish. Schneider (1901) concluded that in the Finnish Archipelago, the younger fish feed on insects and molluscs, larger specimens mainly on molluscs. For the west of the Baltic Sea, the diet of the ide was stated to contain Crustacea, fish and plant matter, and for the Finnish Bay molluscs and insects and, in freshwater bays plants. Collett (1905) recorded that some ide collected in June in brackish water in Norway contained algae and Hydrobia minuta in large quantities, in larger fish from the same locality where exclusively flower buds of Potamogeton spp. In fish from Grue (May) he found only Odonata nymphs. Specimens from Vestfossen (May) had eaten chironomid larvae. Win­ ter specimens from Lundevand in Eidanger had fed on young lumbricids; fish from the same locality in autumn contained only small copepods (Cyclops, Diaptomus, etc.). Huitfeldt-Kaas (1916) studied ide from Lake Mjosen (Norway) and found that fish up to 0.5 kg had taken only Characeae. Besides plant matter, medium-sized ide fed on animals, especially insect larvae, while larger fish took much small fish, feeding on pelagic schools of Osmerus eperlanus (L.) and Coregonus albula L. Jääskeläinen (1917, 1921) stated that the Finnish ide fed to a great extent on molluscs but in its mature life regularly was a predatory fish. He summed up the diet as follows: 1) in Ladoga, especially Mollusca (Lymnaea), insect larvae, fish, (less common food: insects), in older ide fish is the most frequent food item, 2) in the Finnish Bay: mainly Mollusca (Tellina, Mytilus, Cardium, Lymnaea, Bythinia) and Crustacea (Chiridothea), occasionally Gammarus, Asellus and insect larvae. According to Otterstrom (1930—31) of 18 ide captured in April—June (37—41 cm long) in Arreso (Denmark), 8 specimens had eaten fish and 10 empty intestines. Schneider (1920) records ide in Lake Wirz-Järm (Estonia) feeding on nymphs of Ephemera vulgata L., Tropicoris rufipes L., bottom molluscs, remains of plants, Trichoptera larvae and chironomid larvae. Smglian (1920) records as food items bottom animals, e.g. insect larvae and molluscs, and plankton. Berg (1949) stated as food insect larvae, mayflies and other adult insects, Oligocliaeta, fish spawn, small molluscs, diatoms and 94 PLUTARCO CALA filamentous algae. Segerstråle (1950) concluded that the better growth of larger ide in the Baltic sea at Åland than east of Nyland might be due to the rich source of Mytilus edulis L. in Aland water and the lack of this mussel in the Nyland area. Finally, Popescu et al. (1960) concluded that the ide in the lower Danube River is omnivorous, with tendence to rapacity. It takes insects, Crustacea, zooplankton, Gastropoda, and plant matter such as diatoms and filamentous algae, detritus and higher aquatic plant material. The dominant food items are Naucoris cimicoides, Plea minutissima, Corixa spp„ diatoms and remains of plants. In spring (March—April) the predominant food consisted of numerous larvae of Lepidoptera, Chironomidae, Coleoptera, numerous Lum- bricidae, sporadic molluscs, spawn and fish [Cyprinidae, Pungitius platygas- ter). During July and August the dominating items were chironomid larvae, Gastropoda (Lymnaea sp.), macrophytic plant matter, Cladocera, Copepoda, Insecta and sporadic fishes (Pungitius platygaster). They stated that generally young-of-the-year ide between 3—5 cm long, feed on Cladocera, larvae of Chironomidae, diatoms and filamentous algae. From 6—8 cm they take molluscs, adult insects and sporadic larvae of Chaoborus. Finally Popescu et al. (1960) mention that the ide start feeding at a water temperature of 1—2°C. According to Mann (1961) plant matter is the main food source of ide in the River Alster (Germany), so fish species which are omnivorous or have an animal diet do not compete with it. XV Summary 1. This study is based on data from ide ldus idus (L.) collected in the lower part of the River Kävlingeån-Högsmölla to the mouth (Fig. 1). The studies were carried out during a 4-year period (1963—67). 2. A list of the literature about the ide’s biology in different countries, with especial reference to Scandinavia, has been produced. 3. A physico-chemical analysis (water temperature, colour, pH, hardness: Ca + Mg, specific conductivity, chloride ion contents and oxygen con­ tents) of the study area was realized. Large changes, from time to time, of some of these physical and chemical properties of the river water was noted. 4. Death of fish in the River Kävlingeån have been noted since 1910 (Swenander) as a result of low oxygen content, combined with high temperature and pollution caused by discharging of water coming from Örtofta sugar factory and other industries and urban districts. 5. Qualitative notes on the biomass (benthic fauna, plankton, fish and fishery) of the lower part of the River Kävlingeån has been described. 6. The larger movement of the ide occurs in the lower reaches of Kävlinge- ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 95 ån. The ide spends the first year of its life in the river and then joins with other migrating fish to the sea. During summer (June—August) ide older than age group 0 disappear almost completely from the river, spending the summer in the sea. In autumn the ide return to the river, the smallest ones in September and as the autumn advances the schools increase in number and size. Generally, the largest ide leave the river earlier (April to May) and return later (October to November). 7. The spawning of the ide in the River Kävlingeån during the 4-year (1964—67) study period occurred from 29 March to 17 April. The ide spawn in shallow streams at a depth of 0.5—1.5 m. The eggs adhere to stones, plants, etc. The largest fish spawn first. Spawning began 2 to 3 days after the daily water temperature reached and remained about 5°C (Table 6). Most ide spawned during 2 to 3 days when the water tem­ perature rose to about 7°C and the weather was fine. But if the water temperature drops after the initial period of spawning, additional spawning takes place when the temperature rises again (Table 6). 8. Eggs from ide, in the laboratory, began to hatch after 5.5 days at a mean temperature of 13.8°C and after 23 days at a mean temperature of 9.3°C. Thus, low temperature retards hatching, and below 5°C there is no hatching (Table 7). The length of the fry-sac at hatching ranged be­ tween 5.8—6.6 mm (Table 7). 9. It has been realized a growth analysis of 1081 young-of-the-year (Fig. 3) and 433 ide specimens older than age group 0 of the lower reaches of the River Kävlingeån. It was concluded from the results of the years 1965 and 1966 (Table 9) that the growth season of the young-of-the-year ended in early November. The fry had then attained a mean total length between 6.8 and 7.6 cm for the respective years. The size class means (Fig. 10) indicate the body-scale relation, of ide older than age group 0, to be described by three straight lines. The straight line fitted by least squares to the means of fish lengths and scales radii had these equa­ tions : L= 35.92 + 2.1014S (L<15.3cm), L= 80.77 + 1.4720S (15.340.3 cm), where L equals total length of fish in mm and S equals length of anterior scales radius (X30) in mm. The body length at the end of any year of life for a specimen was calculated from Fig. 10. The calculated average lengths for males (101 specimens) and females (132 specimens) showed no significant differences (Table 13 and 14). Data for males and females of the whole studied ide material (433 specimens) were combined in Table 15. Males and females ide show their greatest annual increase in length during the second year of life. 96 PLUTARCO CALA 10. Annulus formation usually occurs around early May and is completed about the end of June. It seems that older ide, especially mature fish, form annuli earlier than the younger ones. The regulation of scales was affected by the water temperature. 11. Length-weight relation is based on data from 989 young-of-the-year ide, and from 428 yearling and older fish. The length-weight relation in relation to sex, sexual maturity or state of gonads revealed no differences Fig. 12). The length-weight relation of ide was linear when both vari­ ables were transformed to common logarithms, except fry below 13 mm in average length. The general parabola W=cLn, where W = weight, L = total length, and c and n are empirically estimated constants has been used to describe the length-weight relation. Thus, the length-weight equation, determined by fitting a straight line to the logarithms of the lengths and weights listed in Tables 16 and 17 are: log W=—5.11749 + 3.03175 log L (young-of-the-year), and log W=— 5.77664 + 3.33853 log L (yearling and older ide), where W = body weight in g and L = total length in mm. The value of n indicates that the body weight increases approximately as the cube of the length for young-of-the-year longer than 12 mm in length and more rapidly than the cube of the length for yearling and older ide. The equa­ tion makes possible the calculation of the unknown weight for a known length and vice versa. It becomes evident that the total length of an ide at a weight of 1 g is 52 mm and at a weight of 1 kg 453 mm. The greatest annual increment in weight of 239 g occurred in the 10th year of life, despite the fact that the greatest annual increments of length were attained during the 2nd year. 12. A sample of 1207 adult ide taken just before and during the spawning season in 1964 between 1 April and 2 May was composed of 621 (51.4 %>) mature males and 586 (48.6%) mature females (Table 19). These num­ bers do not deviate significantly from a 1 :1 ratio (x2 = 0.30). 13. 1163 ide from the lower reaches of the River Kävlingeån were examined with regard to the stomachs’ contents. In the present material there were definite changes in food habits with increase in size which became evident at about 20 mm total length, and at about 140 mm (Table 25). Crustacea were most important in the first group (10 to 20 mm total length), but insects predominated in the second group (20 to 140 mm), especially chironomid larvae. Rotifers were taken by a considerable number of fish, but only by fry below 12 mm. Cladocera were caught by fish between 10—80 mm and Copepoda until 60 mm. In age groups just larger than 140 mm several food items became important such as higher plants, Asellus aquaticus, and Oligochaeta. Insecta continued to be an important group. Chironomidae become less dominant as the fish ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE KIVER KÄVLINGEÅN 97 grow and at above 260 mm they are replaced by higher plant tissue (Table 25). Table 26 shows that the ide in its early life does not speialize on either Cladocera or Copepodci, but there seems to be a tendency for fry to feed more heavily on Cladocera when these organisms are abun­ dant. There seemed to be no definite preference in the food choice by ide. There are small seasonal variations of the feeding habits of the ide. Plant material is eaten mainly in summer and early autumn, when the vegetation is luxuriant. It should be noted that large ide spend most of the summer in the sea (see pp. 62—63) where their diet consists of marine matter. XVI. Acknowledgements The work was carried out at the Department of Animal Ecology, University of Lund. To the head of the department, Professor P. Brinck, I wish to express my sincerest gratitude for his help and interest during the investiga­ tion; to Professor E. Dahl, Lund, and to my previous teacher Professor G. Dahl, Bogotâ, for their encouragement in various circumstances and help in getting the fellowships and Professor G. Svärdson, head of the Institute of Freshwater Research, Drottningholm, for allowing me to use the equipment for examination of fish scales at that Institute. I further thank Michael Moon, B. Sc., for correcting my English, Miss Anita Warnholtz for type-writing and proof-reading, Mr. Axel Andersson for help in the field, Mrs Lena Eriksson for technical assistance and dr awnings. The work has been supported by grants from SIDA (Swedish International Development Authority), Faculty of Mathematics and Science, University of Lund, and the Royal Fishery Board of Sweden. XVII. References Alm, G. 1964. In K. A. Andersson (Ed.). Fiskar och fiske i Norden. Vol. II: Fiskar och fiske i sjöar och floder (3rd ed. Stockholm) : 445—446. Arsredogörelse för Svenska Lax- och Laxöringsföreningen u.p.a. Malmö. 1954—62. Åström, C. P. 1859. Några iakttagelser rörande de vertebrerade djur, som förekomma i trakten af Lomma. Lund. Bailey, M. 1964. Age, growth, maturity, and sex composition of the American smelt, Osmerus mordax (Mitchill), of Western Lake Superior. Trans. Amer. Fish. Soc. 93(4): 382—395. Balinsky, B. I. 1948. On the development of specific characters in cyprinid fishes. Proc. Zool. Soc. Lond. 118(11) : 335—344. Balon, E. K. 1962. The growth’s legality of the Danube ide Leuciscus idus (L.). Prdce Laboratoria Rybårstva 1:117—151, 16 photos. (English summary.) Berg, L. S. 1949. Freshwater fishes of the U.S.S.R. and adjacent countries. 11:96—99. (Translated from the Russian ed. by Israel Program for Sei. Transi. Jerusalem 1964.) 7 98 PLUTARCO CALA Bogan, F. E. 1964. Expansion of the range and some biological peculiarities of the ide in the upper Pechora. Zool. Zh. XLIII (4) : 616—619. (English summary.) Brofeldt, P. 1917. Bidrag till kännedom om fiskbeståndet i våra sjöar. Finlands Fiskerier 4: 173—212. Cala, P. Effects of alcohol on length and weight of young-of-the-year ide Mus idus (L.) and roach Rutilus rutilas (L.). Caldasia 51 (in press.) Collett, R. 1905. Meddelelser om Norges fiske i Aarene 1884—1901. Fork. Vid. Selsk Christiania, 8: 1—173. Duncker, G. 1960. Die Fische der Nordmark. Hamburg. Ekman, S. 1922. Djurvärldens utbredningshistoria på skandinaviska halvön. Stockholm. 612 pp. Ekström, C. U. 1831. Fiskarne i Mörkö skärgård K. Svenska Vetensk. Akad. Handl. för år 1930: 149—152. Hochman, L. 1956. O rustu jelce jesena (Leuciscus idus L.) v rece Dyji. Zool. Listy 5 (XIX) 4: 345—354. (German summary.) —• 1959. Zkusenosti s chovem jesenu v rybnicich. Csl. Rybårst. 14(5) : 69. Huifeldt-Kaas, H. 1917. Mjosens fisker og fiskerier. K. norske vidensk. Selsk. Skr. 2: 1—257. Hynes, H. B. N. 1950. The food of fresh-water Sticklebacks (Gasterosteus aculeatus and Pygosteus pungitius), with a review of methods used in studies of the food of fish. /. Anim. Ecol. 19(1) : 36—58. JÄÄSKELÄINEN, V. 1917. Om fiskarna och fisket i Ladoga. Finlands Fiskerier 4: 250—332. —- 1921. Über die Nahrung und die Parasiten der Fische im Ladogasee. Ann. Acad. Sei. Fenn., Serie A, XIV (3) : 1—55. Karlgren, L. 1962. Vattenkemiska analysmetoder. Uppsala. Komonen, A. 1948. Vuoksen pohjoisen suuhaaran kalastosta. Luonnon Tutkiga, Helsinki, 52: 110—113. Lenkiewicz, Z. 1964. Temperature preferendum of some freshwater fishes. Folia Biol. 12(1): 95—140. Lilljeborg, W. 1891. Sveriges och Norges fiskar. Vol. III. Uppsala. 830 pp. Lundberg, R. 1899. Om svenska insjöfiskarnas utbredning. Medd. K. Lantbruks styr. 58: 1—87. Mann, H. 1961. In H. Caspars and H. Mann: Bodenfauna und Fischbestand in der Ham­ burger Alster. Ein quantitativ-ökologischer Vergleich in einem Stadtgewässer. Abh. Natur. Ver. Hamburg N. F. 5: 89—110. OTTERSTR0M, C. V. 1914. Fisk. II. Blodfinnefisk. Danmarks Fauna 15. Copenhagen. 351 pp. — 1930—31. De danske skallearter (Leuciscus rutilus L., L. grislagine L., L. idus L. og L. erythrophthalmus L.). Vidensk. Medd. Dansk Naturh. Foren. 90:85—311. PenAz, M. 1961. Bemerkungen zum Wachstum des Alands (Leuciscus idus L.) im Thaya- Fluss. Zool. Listy X(XXIV), 3:231—241. (German summary.) Pliszka, F. 1953. The effect of spawning conditions in lakes on the survival rate of juveniles. Polsk. Arch. Hydrobiol. I(XIV): 165—188. (English summary.) Popescu, E., V. Ziemiankowski, and A. Rotaru. 1958. Citeva observatii asupra fecundatiei artificiale si dezvoltarii embrionare si postembrionare la vaduvita (Leuciscus idus (L)). Bui. Inst. Cere. Pise. XVII (2) : 57—64. — 1960. Contributii la cunoasterea biologiei vaduvitei (Leuciscus idus (L)) din cursul inferior al Dunarii. Stud. Cercet. 2(5): 7—33. (French summary.) Rosén, N. 1920. Om Norrbottens saltsjöområdes fiskar och fiske. Medd. K. Lantbruksstyr. 225: 1—94. Schneider, G. 1901. Ichthyologische Beiträge. II. Fortsetzung der Notizen über die an der Südküste Finlands vorkommenden Fische. Acta Soc. Fauna Flora fenn. 22(4) :1—57. ON THE ECOLOGY OF THE IDE IDUS IDUS (L.) IN THE RIVER KÄVLINGEÅN 99 — 1920. In M. von zur Mühlen and G. Schneider: Der See Wirzjerw in Livland. Biologie und Fischerei. Dorpat. SegerstrAle, C. 1926. Från en fiskbiologs arbetsfält. I. Idfiske på Sarvlax. Allas Krönika, Illustrerad Veckoskrift, 23(27): 685—686. — 1933. Über scalimetrische Methoden zur Bestimmung des linearen Wachstum bei Fischen. Insbesondere bei Leuciscus idus L., Abramis brama L. und Perca fluviatilis L. Acta Zool. fenn. 15: 1—168. — 1950. Om braxens (Abramis brama L.) och idens (Leuciscus idus L.) tillväxtart i södra Finlands kustvatten samt i vattnen kring Åland. Memor. Soc. Fauna Flora fenn. 25: 156—165. Sierold. C. Th. E. v. 1863. Die Süsswasserfische von Mitteleuropa. Leipzig. 430 pp. Smith, S. H. 1954. Method of producing plastic impressions of fish scales without using heat. Progr. Fish Cult. 16(2): 75—78. Smitt, F. A. 1895. Scandinavian fishes. Vol. II (2nd ed.). Stockholm. Smolian, K. 1920. Merkbuch der Binnenfischerei, I—II. Berlin. 1228 pp. Spillman, Ch. J. 1961. Fauna de France. 65. Poissons d’eau douce Paris, pp. 303. Stankovitch, S. 1921. Etude sur la morphologie et la nutrition des alevins de poissons cyprinides. Trau. Lab. Piscicult. Univ. Grenoble XIII: 1—182. Stuxberg, A. 1895. Sveriges och Norges fiskar, deras lefnadssätt, födoämnen, lektid, före­ komst och utbredning. Göteborg. 678 pp. Sundewall, C. J. 1851. Fisk-ungars utveckling, öfversigt K. Svenska Vetensk. Akad. Förh. 8(6): 161—169. — 1855. Om fiskyngels utveckling. K. Svenska Vetensk. Akad. Handl. Ny följd. 1:1—24. Svenska Lax- & Laxöringsföreningen u.p.a., Malmö. Årsbok. 1963—1966. Swenander, G. 1910. Biologiska undersökningar i af vissa fabriksanläggningar förorenade vatten. Södra Sveriges Fisk. Fören. skr. 6: 7—80. Weijman-Hane, G. 1964. Kävlingeån som vattentäkt och recipient. Skånes Natur 51:9—28. and A. Almestrand. 1958—1966. Redogörelse för företagna undersökningar beträffande Kävlingeåns förorening. (Mimeographed in Swedish.) Wolf, Ph. 1950. Fiskeribiologiska undersökningar i Kävlingeån. Lund. 72 pp. — 1960. Land drainage and its dangers as experienced in Sweden. Altrincham. 73 pp. (See also Salm. Trout. Mag. 1960—62.) — 1965. Kävlingeån — en fråga om ädelfisk eller smutsvatten. Skånes Natur 52: 73—80. Yamada, J. 1961. Studies on the structure and growth of the scales in the goldfish. Mem. Fac. Fish. Hokkaido 9(2): 181—226. Yasuda, H. 1959. Implication of the scale ring-II. Ring formation. Bull. Jap. sei. Fish. 25 (2): 95—99. Fishing pressure, growth and recruitment in a small high mountain lake By T. Lindström, Å. Fagerström and K.-J. Gustafson I. Introduction Lake Långbjörsjön in the Jämtland high mountains has been described in two previous papers (Fagerström et al, 1969, Gustafson et al, 1969) where the methodological aspects of tagging and the distribution of the two existing species — char and trout — over the different habitats in the lake are treated. These studies have been used for the continued analysis in the present paper. In this small lake, one tenth of a square kilometre in area, a fishing ex­ periment was started in 1958 to investigate whether an increase in the fishing could change the length distribution in the trout population of a small high mountain lake in such a way that the lake would provide the fly fishermen with a more valuable fishing. The immediate effect, that of a reduction of the old and big fish, had been anticipated in, e.g., a memorandum for the management of the sport fishery in the region, written by B. Tägtström (1952). The question was whether the recruitment could also be reduced to a point where the individual growth in the remaining population would increase spectacularly. Consequently the length distribution (Table 11), the growth (Section IV, Fig. 1) and the recruitment of trout (Section VII, Table 8, Fig. 2) have been studied. The fishing has taken both char and trout, but the growth and recruitment Table 1. Catch in a full set of gill nets (15 nets, Table 6), fished three nights with a regular distribution over the ten sections and first and second position in gangs. The data have been converted into catch per set. Mesh-size designation: 5.00 3.75 3.33 3.00 1 2.50 2.14 1.67 Total 1961, Aug. 9-—11 a, number of trout . 0.3 0 2.0 3.0 14.3 16.7 33.3 70 b, number of char . . 0 0 0.7 1.0 24.0 22.3 38.3 86 b:a .............................. — — — 0.3 1.7 1.3 1.2 1.2 1965, July 9—11 a, number of trout . 0 0.3 1.7 2.3 16.0 22.0 42 b, number of char . . b:a .............................. 0 0.3 0 0.3 0.1 7.0 0.4 67.0 3.0 75 1.8 1966, July 1—3 a, number of trout . 0.3 1.3 2.0 1.7 13.7 7.0 17.0 43 b, number of char . . 0 1.0 1.3 1.0 6.3 29.3 47.3 86 b:a .............................. — — — 0.6 0.5 4.2 2.8 2.0 FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 101 of char has not been investigated, for methodological reasons. As an indica­ tion of the char-trout proportion, the proportion in gill nets, set along the shores in a regular manner, is shown in Table 1 (cf. also Gustafson et al., 1969). The comparison starts with 1961, the year of the most intense fishing during the experiment. The density of char in different size categories, like that of trout, is of course related to age distribution and variations in recruitment between years; thus the char-trout ratio was as high as 1.8: 1 in 1960 in spite of a distribution of the gill nets over the lake that favoured the catching of trout. A change in the balance between the species would have been a very plausible outcome of a fishing experiment, but nothing definite can as yet be stated in this case; the fishing pressure has unintentionally been directed more against trout. II. Total Yield from Lake Långbjörsjön In Table 2 only those fish are included that were not put back into the lake, There may be some illegal fishing that could not be reported, but the known yield is in itself fairly high for a high mountain lake of about 0.1 square kilometre. “Experimental netting” means netting to obtain data on recap­ tures, etc. “Other fishing” means mainly sport fishing and (1962) seining, but in this group, too, a fair number of the fish have been caught, or checked by the biologists, and scales obtained or tagging noted; this relates specifi- Table 2. Total yield from Lake Långbjörsjön. Year Trout Char Experi­ mental netting, number Other fishing, number Total number weight Experi­ mental netting, number Other fishing, number Total number weight 1951 . 25 4.6 3 0.4 1952 . — — 186 29.3 — — 74 7.0 1953 . — — 115 19.6 — — 105 10.7 1954 . — — 101 18.1 — — 46 3.5 1955 . — — 89 17.4 — — 38 4.2 1956 . — ---- 137 22.1 ---- — 3 0.7 1957 . — — 97 16.6 — — 3 0.4 1958 . 112 150 262 37.4 79 4 83 6.9 1959 . 264 60 324 33.2 402 15 417 35.1 1960 . 176 63 239 — 320 5 325 — 1961 . 380 176 556 — 441 38 479 — 1962 . — 264 264 — — 255 255 — 1963 . 65 190 255 33.1 115 36 151 — 1964 . 46 101 147 20.6 157 25 182 13.5 1965 . 127 123 250 28.4 224 27 251 20.3 19661 . 129 102 231 25.4 267 11 278 20.6 1 Only part of the season. 102 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON Table 3. Mean size of trout in the sport fishermen’s catches, kg. 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 0.18 0.16 0.16 0.18 0.21 0.16 0.17 0.16 0.13 0.15 0.14 0.13 0.14 0.15 0.10 0.13 cally to 1962 as no gill nets were set in that year, but three-quarters of the yield was caught by the experimental fishing staff. Dates for the netting are recorded in Table 4 with the exception of the netting October 5—7 in 1961. “Other fishing” was performed in June, July and August. III. Mean size of trout in catch The mean size of trout in the sport fishermen’s catches decreased in 1959, the second year of the experiment, and did not regain the original level after­ wards, Table 3. Data for fishing places, hook sizes, etc, are not available in full, but the selection through hook size makes mean size in the population — or mean size in the experimental gill nets, too, for that matter — of less relevance for the experience of mean size that a sport fisherman can obtain. Length distribution in the population is not available at all, but in the experimental gill netting the size class above 24 cm could be studied in the size distribu­ tion, Table 11, in order to obtain a view of the quality of the water for fly fishing. The number in this size category also drops from 1958 to 1959 and the lowest value is obtained in the last year, 1966, Table 4. As appears from the table, the date of the gill netting varies from year Table 4. Number of fish 25 cm or longer in one gill net set of standard com­ position in 1958—59 and 1961—66 (Table 6). The last row is the minor portion of the trout that was caught in the nets with a knot-to-knot measure from 5.00 to 3.00 cm and gives some suggestion of the correction that should be applied to the 1960 trout values; in 1960 5.00—3.00 mesh size nets were not in use, only nets with 1.67—2.50 cm mesh size. Year 1958 1959 1960 1961 1963 1964 1965 1966 Date ............................................................. July July July 14 Aug. July June July July 10 14—25 -Aug. 14 9—11 4 17 9—11 1—3 Number of trout .................................... 38 17 (11) 16 23 24 22 10 Number of char....................................... Number of above trout caught in nets with 3.0—5.0 cm knot-to-knot mea- 16 13 (13) 16 ? 7 7 5 sure ............................................... 7 3 — 3 ? 1 3 1 FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 103 to year, and so does the share of the total catch which preceded the date of gill netting. The gill netting itself, however, takes an important share of the yearly yield, (Table 2). The distribution of the gill nets between the different sections of the lake and between “first” and “second” in each gang (from the shore) is best in 1961, 1965 and 1966. A table of the number of trout 30 cm or longer shows almost the same pattern as Table 4. IV. Growth of trout Since 1953 scales from trout have been collected from Lake Långbjörsjön and the growth rates have been ascertained by “back calculation”. The diffi­ culties of achieving an exact result have already been discussed (Fagerström et al, 1969) ; nevertheless, although some mistakes are inevitable, the results mm Fig. 1. Growth of trout in Lake Långbjörsjön. 104 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON Table 5. Mean length of trout in Lake Långbjörsjön before and after the beginning of the experiment. Age Before After Increase in cm Increase in °/o 1 ............... 4.9 cm (452) 5.1 cm (508) 0.2 4.1 2 ............... 9.1 „ (365) 10.7 „ (591) 1.6 17.6 3 ............... 14.2 „ (263) 16.2 „ (561) 2.0 14.1 4 ............... 19.1 „ (180) 20.9 „ (418) 1.8 9.4 5 ............... 23.1 „ (102) 24.8 „ (158) 1.7 7.4 (The figures in brackets show the number of trout examined.) should give a mainly correct view of what has happened regarding the growth rates of trout. Table 5 and Fig. 1 show the growth rate of trout from Lake Långbjörsjön during the years 1950—64, as ascertained from scales collected from 974 fishes. As early as 1917 Dahl showed what happens when it is attempted to reduce the population of trout in a small lake by removing about 50 per cent of the stock yearly. The old trout were gradually removed and the population then consisted of young fish with a good growth rate. However, a sport fisherman is as a rule not interested in the age of the trout he catches but only in their length and weight. At present it does not seem probable that the better growth rate has neutralized the rejuvenation in Lake Långbjörsjön. The fishing manager is, on the other hand, very interested in what happens to the whole stock. This table shows that the growth for the trout has increased considerably since the beginning of the experiment. This is especially interesting as regards the trout during their second year. The increase may show that they spend the whole of their second year in the lake. This agrees with the experience of the sport fishermen, who exceptionally catch trout smaller than 10 cm. There remains another possible explanation for the good growth in the second year: some of the more recent year classes have been rather small, as will be shown in section VII, and the competition within the year class correspondingly reduced on the nursery grounds, wherever they are. As a matter of fact it has not been possible to ascertain where the trout from the lake spawn. The inlets are very small and look rather unimportant with the exception of the main inlet. This is, however, blocked by a very big flat stone close to the lake and it seems unlikely that the trout can pass this obstacle. The outlet has an impassable rapid at its very beginning. The trout might spawn in Lake Långbjörsjön but the material is not at all con­ clusive. FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 105 V. Variations in growth and fishing; preliminaries for the two following sections The population structure will be judged by the catch in a gill net set of standard composition. The year classes will be judged mainly by the number in age group 3+ (Table 8), as sport fishing rarely takes such small trout. This age group is generally catchable with the standard set. The increased growth treated in the last section does not interfere seriously even if the falling number of nets in different mesh sizes in a standard set (Table 6) is taken into account; a detailed comparison between Fig. 1 of growth and length distribution in the most fine meshed gill nets shows that it is the year classes 1955 and 1956 at the age of 3+ that may be underestimated by the catch in a standard set as they were not yet fully catchable and this only strengthens the conclusions put forward (cf. section VII on recruitment). Year class 1955. Caught for the first time in 1958 at 3+ years mainly with the 1.67 cm meshed gill nets. All fine-meshed nets closest to the shore and coarse-meshed predominantly as second in the gang; distribution otherwise not specified in 1958. Year class 1956. Caught for the first time in 1959 at 3+ years with mesh sizes 1.67 and 2.14 cm. Eastern shore less fished than the western, most fine-meshed nets closest to the shore and most coarse-meshed as second in the gang in 1959. Year class 1957. Caught for the first time in 1960 at 3+ years with mesh sizes 1.67 and 2.50 cm. Gill nets fairly well distributed over the different sections of the lake but twice as many nets in position closest to the shore as in position second net in the gang in 1960. (No coarse-meshed nets.) Year class 1958. Some eight fish were caught as 2+ already in 1960 by gill nets but in the main this year class, too, was untouched by fishing until the age of 3 +, in 1961. In that year, however, much of the sport fishing preceded the gill netting and scale sampling which, as an exception, were performed in August. With the good growth established for this year class some fish may have been caught by sport fishing in July 1961, but on the other hand the year class was certainly fully available to several mesh sizes in the set of gill nets (1.67, 2.14 and 2.50 cm), so the conclusion is that the year class could hardly have been a very good one. Distribu­ tion of the gill nets over the different sections was good, and equal numbers on first and second position from the shore in 1961. Year class 1959. The first that could hypothetically have been influenced by the increased fishing — fewer spawners left in the autumn of 1958. Owing to good growth and late netting in 1961 the year class was caught already in 1961 in fair numbers at 2+ years, mostly in the finest mesh size. In 1962 there was no gill netting, but the year class predominated in the scale samples obtained with hook fishing, etc. In 1963 they were caught at 4+ years, still in good numbers (distribu­ tion of the nets not as regular as in 1961). A rough estimate for this year class as 3 + in 1962 would surpass all other 3+ estimates in Table 8. Year class 1960. Some few may have been caught by the seining and hook fishing by the biologists in 1962; there were 36 trout below 20 cm in the catch but then no 2+ was obtained in the scale sample. Otherwise this year class, too, was unfished until the age of 3+ (in 1963). Year class 1961. Caught as 3+ in the standard set in June 1964. Fine-meshed gill 106 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON nets twice as often in first as in second position from the shore and slightly irregu­ larly distributed over the sections of the lake (in 1964). Year class 1962. This year class may have been influenced by the heavy fishing in 1961 — fewer spawners left to the autumn of 1961. The fish were caught for the first time in gill nets in 1965, mesh sizes 1.67 and 2.14 cm, and then in an increased number at 4+ years in 1966, mesh sizes 1.67—2.50 cm, both years with good distribu­ tion of gill nets over the ten sections of the lake and equal numbers of gill nets in first and second position. Judged by this method it is the smallest year class in the series. Year class 1963. Some eight caught in gill net 1965 at 2+ years, otherwise caught for the first time in 1966 at 3+ years, mainly in the most fine meshed gill nets (1.67 cm). VI. Effect of fishing on the population structure of trout The increased growth in all age groups of trout indicates that recruitment of trout and char has at least not increased to a degree that exploits the whole food production that was “released” when the fishing mortality of the two species increased. A picture of the population structure of the trout has been obtained from the number of trout in different age groups caught in gill nets in one set of standard composition (Table 6). Several corrections have to be applied. The first of these are explained in Table 7, Le. a raising factor when all trout in the catch were not aged and a recalculation to catch in one set of standard composition when the fishing was otherwise performed. The corrected density values are listed in Table 8. In 1960 coarse-meshed nets were missing altogether and no correction for this has been applied and the density estimates for different age groups and year classes in Table 8 will be correspondingly low this catch year ; the coarse-meshed nets contribute, however, very little to the catch of young age groups. A third correction relates to the distribution of the gill nets. It is not only catch years 1961, 1965 and 1966, with the best distribution of the gill nets round the lake that are included in this analysis, but particular attention should perhaps be paid to these three catch years. In the catch years 1958, 1959, 1960 and 1964 the fine-meshed nets were more often set close to the shore than as second net in the gangs and the corresponding correction of young-trout values is explained in next section. A correction for old-trout values and the irregular distribution of mesh size 2.50 cm. nets in the catch Table 6. Composition of a standard set of gill nets. Mesh size, knot-to-knot measure, cm .... 5.00 3.75 3.33 3.00 2.50 2.14 1.67 Number of nets in the set.......................... 1 1 2 2 3 3 3 FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 107 do * .S ^ Jh 03 CD c/3 > ÖoO ÉH o«4H O d) 5 ^ çd ^ fH Cö CÖ «_2 ö Ç C3 i/3 a; G o T5 Sh a TJ Ö 0) tlD Ö CÖ .M ci T3 T3— a> uCD GS £■< qj sh 03 * 3 o « o rQ £ 3 H< . «5>*-> 05 6ß r-H G g O CD coSh o 'S oÿ co CD ’S 2 ö ^ § Æ2 öo . tt s ■ fH <3 CD a; rC O a A03 H 3 Ö 03 ,COHH •XXX’!Ohh C^l CO CD CM co co co «iO pH pH Tt coX ^ XXH pp x-^ÇO CM CD ^ CM pH CO CM s. v 00 CD O CO CM CM iC5 co CM uO CO CO co CO CO pH pH XXXX« 03 tH Dl CM •o w CM \ pH CO co O \ pH CO 03 CO 00 CO co rH UO co co CM t-1. pH T\ co^ X i X ^ ^ x^-^ uO CD CD 'rH !>■ 03 X I XT pH UO ) 03 : g ö i T3 rrt> g !G3 h > M cn' a; g I s s fe fe i2 «? CA W3 ^ 4) d D 33 'S w » -g ’S * w ^ 0) 03 <» <» 03 03 (U 03 43 03 03 £ tn frn 03 03 ^ h.h -G A G G G G fehHCOHH Q0DOHC0710© kOUOCDCOCDcOCOCO 03 03 03 03 03 03 03 03 108 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON Table 8. Density estimates for different year classes and their different age groups. The number caught in a certain mesh size and classified as belonging to a certain age at the scale analysis is multiplied by the factor in Table 7. Corrected values for one age group are then added together from different mesh-size values in one standard set from a certain catch year. The different catch years are shown diagonally in the table with 1958 at the top and 1966 at the bottom and no values for catch year 1962 (no net fishing that year). Year class Age group 1 + 1 2+ 1 3 + 4 + 5 + 6 + 7 + ! 8 + 1950 ........... _ _ _ 3 1951 ........... — — — — — — 3 1 1952 ........... ■-- -- - — — — - 12 1 0 1953 ........... — — — — 42 12 0 0 1954 ........... — — — 59 12 1 0 — 1955 ........... — — 27 12 11 2 — 0 1956 ........... — 2 19 13 6 — 0 0 1957 ........... 0 0 10 14 — 3 1 0 1958 ........... 0 5 13 — 3 2 0 0 1959 ........... 0 35 — 38 22 6 0 — 1960 ........... 1 — 17 9 12 1 — — 1961 ......... — 4 13 14 12 — — — 1962 ......... 0 0 7 11 — — — — 1963 ......... 0 3 14 — — — — — 1964 ......... 0 5 — — — — — — 1965 ......... 0 years 1958, 1959 and 1964 must be a little more uncertain, as the total number caught is less but a maximum reduction by 0.25 (1958) or 0.15 (—-59) times the values in Table 8 was tentatively carried out1. (Catch year 1960 could not be used for the old age groups. Whichever reduction is made for catch years 1958, and —-59, the numbers of old trout (5+ and older) in these years exceed the numbers in 1961, 1965 and 1966 (with regular net distribution), indicating the reducing effect of increased fishing pressure on old age groups. The number of old trout in catch year 1964 occupies an intermediate position, and is highly influenced by the strong year class 1959. It is theoretically possible that a decreasing natural mortality during the experiment has to some degree masked the effect of increased fishing mortality. Little information is available on age and length of spawning trout (Table 9). The material suggests that some of the trout are mature at the age of 5+ (Table 9 compared with Fig. 1). Studies of maturity in the gill net catch 1 The factors were arrived at by studying the catch and the relevant distribution only of nets 2.50 cm knot-to-knot measure. 55 trout>25 cm were caught by 14 efforts with the nets closest to the shore while 28 such trout were caught by 13 efforts with the nets as second net in the gangs. Due regard to the more coarse-meshed nets and their small catch would tend to reduce the applied corrections. FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 109 Table 9. Distribution of spawning (s) and non-spawning (n) trout in a sample, October 1961. Total length, cm 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Mesh size 1.67 s n 2.14 s n 2.50 s n 3.00 s n 3.33 s n Totals, spawners Totals, nonspawners 4 3 4 6 4 —3 ------ — 1 7 8 2 -----2-------- 2 1 0 0 0 0 0 0 0 4 3 6 7 11 10 6 1 2 —--------------------- — 2 3 12—1 — ----- 1 — 1 — 1 2 6 3 7 3 4 4 0 0 10 112 5 11 4 9 3 5 4 1 2 — 2----------------------------------- 2 2 1 3 0 0 0 1 3 110 0 10 0 July 1966 (129 trout) confirm this conclusion. The adult population was large in 1958 and gave rise to the important year class 1959, while the adult population was small in 1961 and gave rise to the small year class 1962. This will be further discussed in the following section. VII. Trout recruitment estimates If the density indices for age 3+ from Table 8 are plotted against the mean air temperature in May—June and July—August in the adjacent Storlien, the indices of year classes 1958, 1962 and 1963 (catch years 1961, 1965 and 1966) show a reasonable relation to mean temperatures of either kind of year of hatching, Fig. 2. Using the bottom figures in Table 7, Gustafson et al. (1969) for catch in mesh sizes 1.67 and 2.14 cm knot-to-knot, it can be estimated that a net closest to the shore catches 3.4 times the number of trout caught in the second net where the size groups here relevant are concerned. Density indices from Table 8 for year classes 1955, 1956, 1957 and 1961 can then be converted according to the known distribution of the nets between first and second place in the gang (reported in section V) using the conver­ sion factor 0.65 and 0.85 respectively, and these estimates are also introduced in Fig. 2 together with the original values from Table 8 and the rough density estimate 37 for year class 1959 as 3 + . The conversion results in a fairly good concurrence between density indices and mean July—August temperature while the year classes 1955 and 1959 do not fit in the relation between year class indices and May—June temperatures. The lake is normally ice-covered from the middle of October up to the end of May. Some guesses can be made about the year classes 1953 and 1954 though there is no density estimate for these trout in the 3+ age. As is shown by 110 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON 531 154 20 - JO---- 572 ro u n JULY-AUGUSTMA Y-JUNE Fig. 2 a. Recruitment estimates for trout and mean air temperature at Storlien during May— June and July—August. The arrows indicate the corrections (reductions) discussed on page 109. 55, 56 etc. indicate year class. Only mean temperatures are inserted for year classes 1953 and 1954. Table 8 the indices for 3 + and 4 + are generally both on the top of a dome- shaped curve. The indices 42 (age 5 + ) and 59 (age 4 + ) from Table 8 should at least not correspond to much lower indices for year classes 1953 and —54 as 3+; the indices would appear very high up in the diagram, Fig. 2. The corresponding mean temperatures are inserted at the top of the diagram. This material does not strengthen the impression of a simple temperature relation. From the possible composite effects of both temperature and increased fishing mortality, the most promising seems to be a positive correlation between May—June temperatures and recruitment, combined with a reduc­ tion of the spawning population in 1959 and subsequent years, due to in­ creased fishing mortality and consequently reduced recruitment level begin­ ning with year class 1960 (cf. Tables 3, 4 and 8 and p. 108). The netting during the first year of the experiment (1958) was in any case not enough to bring down the spawning population to a critical level. FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 111 Fig. 2 b. Ditto, mean air temperature during May to September inclusive. o---------- > LO IS? 59 >55 560 \|/ Ç57 P O56 J6 60 O N06 1 3 06, 6 7 6 9 W H °c A positive correlation between the size of the spawning population and the size of the resulting year class does conform to the general ideas of population control (cf. Frost and Brown 1967, Northcote 1969) provided the fishing pressure is high enough to reduce the number of spawners to a point below a critical level where the number of spawners becomes a regula­ ting factor, and perhaps this point is more easily reached for trout than for many other fish. In Sweden their spawning streams are often protected against fishing during spawning time. Once above this critical level, on the other hand, environment or self-régulating processes can make the popula­ tion size oscillate around a saturation level or even bring the “recruitment curve” downwards at higher population levels (Ricker 1958 p. 237, Bever- ton and Holt 1957 pp. 55—61). Within the salmonoid family the parent and progeny fractions of a pink salmon stock seem to be very effectively isolated from each other, the progeny escaping to the sea before the next spawning run (Manzer and Shepard 1962, Gilhousen 1962). In Lake Jölstervann in Norway where lake spaw­ ning is confirmed the different fractions of the trout population co-exist in a very deep lake (Jensen and Senstad, 1962, Klemetsen 1967). The self­ regulation of the population should be more effective, when the shelter of 112 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-.T. GUSTAFSON Table 10. Population estimates by the Petersen method (within one season). Increased fishing mortality of tagged fish in the recapture period and in- Species Trout Marking Fincutting Tagging Tagging Fincutting Year and date of marking............................ 1959, July 2—11 1960, July 4—10 1961, July 3—16 1961, July 3—16 Total length, latitude, cm Number Marked (M) . . . . 21— <20 25 >25 54 78 45 21— <20 25 >25 31 85 46 21— <20 25 >25 27 86 55 21— <20 25 >25 2 19 3 Recapture period July 14—25 July 14—24 August 14 August 9—11 August 9—11 Total length, latitude, cm Number recaptured (R) Catch, number (C) ........... Population Estimate 21— <20 25 >25 11 8 9 151 65 49 21— <20 25 >25 2 11 6 69 77 30 22— <21 26 >26 5 11 7 126 52 28 22___ <21 26 >26 0 2 1 126 52 28 p M'C R 741 634 245 1070 595 230 680 407 220 — 494 84 1620 1895 1307 Total length, latitude, cm Number recaptured (R) Catch, number (C) ........... — — — 22— <21 26 >26 4 10 5 83 75 18 23— <22 27 >27 5 15 3 134 51 21 23— <22 27 >27 0 3 0 134 51 21 Population estimated P = _. _ — 643 638 166 724 292 385 — 323 1447 1401 the young, pre-recruit stages against interactions from other parts of the population is incomplete. According to our present knowledge a third ex­ ample, the trout population in the shallow Lake Långbjörsjön, may be a relevant case of strong selfregulation (when the fishing pressure is released), but so far there is not enough material to substantiate this hypothesis. A permanent influence on the recruitment level of trout should also be exerted by the char population in the lake. This effect is, however, not in­ vestigated at all in the present study, and the gap necessitates caution in conclusions about the trout population control. VIII. Appendix: Population size estimates from marking recapture data To get a rough impression of the population size in this 0.1 square kilo­ metre lake, where the described population structure changes occurred, the FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 113 creased natural mortality of small tagged fish are suspected to operate (Fagerström et al 1969, Sections III and IV and Tables 4 and 7). Species Char Marking Fincutting Tagging Tagging Fincutting Year and date of marking....................... 1959, July 2—11 1960, July 4—10 1961, July 3—16 1961, July 3—16 21 _ 21 _ 21_ 21_______ Total length, latitude, cm <20 25 >25 <20 25 >25 <20 25 >25 <20 25 >25 Number Marked (M) .. 33 35 2 13 38 4 13 107 11 3 19 0 Recapture period July 14—-25 July 14—24 August 14 August 9—-11 August 9 —11 21- 21- 21- 21- Total length, latitude, cm <20 25 >25 <20 25 >25 <20 25 >25 <20 25 >25 Number recaptured (R) 2 9 2 0 3 3 1 19 4 0 3 0 Catch, number (C) .... Population Estimate 120 264 17 132 160 28 94 140 25 94 140 25 „ M • C P- R 1980 1026 17 — 2027 37 1222 788 69 — 887 — 3023 2079 22- 22- 22- Total length, latitude, cm — — — <21 26 >26 <21 26 >26 <21 26 >26 Number recaptured (R) — — — 1 4 1 3 19 2 0 3 0 Catch, number (C) .... Population estimated 166 144 10 125 122 12 125 122 12 „ M-C P- R — — — 2158 1368 40 542 686 66 — 773 — 3566 1294 marking data from Fagerström et al. (1969) will be used. In the marking analysis it was shown that tags were shed, probably fairly soon after the summer of tagging, and that small fish (below 20 cm) were not recaptured to the same extent as larger ones, not even during the summer of tagging. To use for population estimate tagged fish above 20 cm and recaptures during the first summer — as is carried out in some columns of Table 10 — is, however, not enough to secure a good population estimate, as tagged fish are likely to be somewhat overrepresented in gill-net catches in the present experiment (Fagerström et al. 1969, p. 31). An attempt to check this popula­ tion estimate is therefore made in Table 10, using the data of fin-cut and recaptured fish from the marking experiment. Some of those were below 20 cm too, and they have been entered in separate columns in Table 10; different length fractions can be separated but individual recognition is impossible. The number of small, fin-cut fish is low but should not give such a biassed estimate of the number of small fish as tagged small fish do. The best population estimate would therefore be obtained by adding the 8 114 T. LINDSTRÖM, Â. FAGERSTRÖM AND K.-J. GUSTAFSON estimate from those below 20 cm to the estimate from those between 20 and 25 cm and the estimate from those above 25 cm at marking, using the fin-cut series. As stated in Gustafson et al. (1969) — corrected by “errata” in that volume — the density falls by something like three or four catchable fish (char plus trout) per ten metres square. IX. Summary An experiment was started in a small, high mountain lake to investigate whether the recruitment could be reduced to a point where the individual growth would increase spectacularly. Mean size of trout in the sport fisher­ men’s catches has decreased and the number of trout and char above 24 cm in a standard set of gill nets has decreased (Tables 3 and 4) although the individual growth of trout has increased (Fig. 1). From the possible com­ posite effects of both temperature and increased fishing mortality on recruit­ ment of trout, the most promising seems to be a positive correlation be­ tween May—June temperatures and recruitment, combined with a reduction of the spawning trout population in 1959 and subsequent years (due to in­ creased fishing mortality) and a consequently reduced recruitment level beginning with year class 1960 (Fig. 2). Rough population size estimates are presented in the Appendix. X. References Beverton, R. J. H. & S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Lond. (2), 19: 1—533. Dahl, K. 1917, 2nd edition 1943. 0rret og orretvann. Studier o forsok. Oslo 182 pp. (In Norwegian.) Fagerström, Â., K. J. Gustafson & T. Lindström. 1969. Tag shedding, growth and differ­ ential mortality in a marking experiment with trout and char. Rep. Inst. Freshw. Res. Drottningholm 49: 27—43. Frost, W. E. & M. E. Brown. 1967. The Trout. London. 286 pp. Gillhousen, P. 1962. Marine factors affecting the survival of Fraser River pink salmon. In Symposium on pink salmon, N. J. Wilimovsky (Ed.), MacMillan Lect. Fish., Univ. Brit. Columbia: 105—109. Gustafson, K. J., T. Lindström & A. Fagerström. 1969. Distribution of trout and char within a small Swedish high mountain lake. Rep. Inst. Freshw. Res. Drottningholm 49: 63—75. Jensen, K. & C. Senstad. 1962. 0rret som gyter på stille vann. Jakt-Fiske-Friluft. 91:202, 203, 232. (In Norwegian.) Klemetsen, A. 1967. On the feeding habits of the population of brown trout (Salmo trutta L.) in Jölstervann. Nytt Mag. Zool. 15: 50—67. Manzer, J. & M. P. Shepard. 1962. Marine survival, distribution and migration of pink FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 115 salmon (Oncorhynchus gorbuscha) off the British Columbia coast. In Symposium on pink salmon, N. J. Wilimovsky (Ed.), MacMillan Lect. Fish., Univ. Brit. Columbia: 113—121. Northcote, T. G. (Ed.) 1969. Symposium on salmon and trout in streams. MacMillan Lect. Fish., Univ. Brit. Columbia. 388 pp. Ricker, W. E. 1958. Handbook of computations for biological statistics of fish popula­ tions. Bull. Fish. Res. Bd. Can. 119: 1—300. Ta bl e 1 1 A . Le ng th dis tri bu tio n o f tr ou t in di ffe re nt ki nd s o f g ea r. (M es h s iz e d es ig na tio n 2 .1 c = co tto n n et s, t he on ly ne ts o f th is m at er ia l). Ca tc h i n s ta nd ar d s et is ca lc ul at ed fro m th e o th er en te re d g ill ne t c at ch es . D at es of fis hi ng in Ta bl e 4 . T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON uio ‘q}8u9i pjjox »Okco>0>0k0k0i0i0.k£0u0 0 T-irHiMiM'cOCO^^iC»OCOCÛt>t^OOo6o50iodiHTH rH i-t tH tH tH tH tH t-M tH tH t—( tH y—( tH 04 04 04 04 19 59 , pe rio d o f re ca pt ur es G ill ne ts sta nd ­ ar d se t - I> tH © (2 .1 c) S' 2. 1 to 2. 5 iO ! p p id od 30 19 59 , m ar ki ng A ll ki nd s of g ea r Se in in g 19 58 G ill ne ts sta nd ar d se t T—( 1. 7 04 2. 1 04 2. 5 04 1 O O ud cd 04 Pe rio d G ea r, f or gil l ne ts a lso m es h s iz e de sig na tio n (c f. T ab . 6) N um be r o i o i cd I q I co d I c 1 6 1 oi 1 co tH I |> I CM. I I 00 04 ‘ rH ‘ u0 * ffd * CO 04 04 04 ÎÜ C0I>Q0^Ol'<-i00 04^ ! ■tH05 04 04 04CDi004tH ! i-H CO ^ H 04 t> CO ! 05 I ZD 04 I 04 I H 04 CO 04 04 04 iquqiqiqiqiquouOuOiOtco rHH04oicOCO^^»OiÔcO CD I> 00 QO 05 05 O Ö FH rtrtHHHHHHHHHHHHHHHHNOlOlO) uio lq}8u0i iBjox FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 117 MlMCOWrHTjHiOiOCDCDC'^OOOOOîOîOOTHiHWWfOCO^^ifliOCÛCOtNlX» CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCO cm 05 Ö 00 © © CO © CM I I CM rfi I CO I CM 1-H iß TH CO HHCOHCO I CM iß ICO I© loo I C© I TH I IO I iß I CM CO ICO ; CM I Iß I I> I 00 I 00 |C0 ICO CM iH i-h I H I cm I h I N I I H CO CM TH CO >ß I TH CM I CO iß CO CO CM CM CM ! I I I I ! *£J UJ U.S UJ Iß —„ _» CMCMC0c6^ni»ßiOC0CDI>l>000005C5©©!Hi-JcMCMC0c6,*^TH*ßißC0cdl>,l>00 CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCO iß © 00 rH CM 05 CM CM © CM CM tH CO o _ H Ö fics <3 OJ —< 118 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON Total length, em cocoTfH-rHiOiocOcor^r^ooodosoioö^tHOicvicocd HHHHHHHHHHHrtHHNWWNNCINN 19 61 , m ar ki ng Fi n cu tti ng al l ki nd s of ge ar Ta gg in g juag jo spun] lie ‘spnox suoods .I-OK) sail! J8110 guiqsij ' a' LI guiupg 19 60 , re ca pt ur e p er io d G ill ne ts In co m pl et e sta nd ar d se t A ug . - 3 rH CM CM CM tH U0 CM CM 19 60 , ta gg in g SPROX sand ipiAV .10110 gmqsij Am guiurag Pe rio d G ea r, f or gil l ne ts a lso m es h s iz e de sig na tio n (c f. T ab . 6 ) N um be r CO I (Q I CD CO|CO|CO|C3|CDIC:|N CM I H I CO I CO i CM I CM CM|h|^|0|C1|0|iO|^|Ö5 CO CO »C CO CO CM CM 00 00 o I H I cm' I h I -r-t I CO I CO I CO I CM* I CM j CM ■»—1 ! uOCOCOgOtHCO^-^CDCO^CM^COtHCM I tH I H COOCMt>CMCOCOCOCO 'CM I H H CM H ^ I> Ttl|-rH|u0|l0|O|0lj^jL0 CM I CO I rH i Cx 1 uO CO I CM I CM I HO I CO ^ I O I H I CD I CM I N iC»iOLiOkOiOLiOiO»OLOiOuO cow^rjîi£50ôcdt^i> 00,00 oîciod^HCMcicoco rtHHHHHHHHHHHHHCqCMCMCMClCMClCM rao ‘qjguaj ib}ox FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 119 uOH5uonOHO^qHO*qiqHOHOHOHO ^^io^cDcDt>^ooo6©0}odrHHOidM255!2!2525SC:WiMWCvliMiMWiMiMCMWMfOCOfOCQWCQCOCOCOCOMCOCOfOCQ uo - 1 50 1 - 1 ^ Il IIl~ 1 1 1 (III 1 1 1 1 1 1 1 24 CO* CM Ci 18 16 1 2 15 1 00 1 7t< 1 SO 1 5. 3 3. 2 2. 5 uorpas siqj m snoiias jsoui S}0U 1 p9qs0ui-9SJBOO jo 00U0sqy ->■ 00 CO CM TH 1 ^ tH CM il il il il 1 M- 1 1 1 1 1 1 1 72 00* 5 3 2 rtH h ! 1 1 1 « Il II - 1 II ! II 1-1 1 1 1 1 50 COCM HO HÎ 4 3 4 3 ’* 1 |N 1 |Hî,NC,'| 1 1 1 1 1 1 1 1 1 1 54 rH* CM 00 CM 1 *> I CM |T*|rt<|r^|«0|C0|CM| 1 1 7—1 1 I I CM CO CM 1 2 1 1 TH 1 1 1 1 1 1 ! 1 I ! 1 1 CD t-H CM CM O i ^ I CM I » lI|w|CO|CO|||||| 1 tH I 1 1 001 1 1 III 1 1 1 1 1 1 1 1 1 l 1 1 HO CM 00 1 CO 1 CM 1 1 |CM|th|th|i—(|th|tH| I I 1 1 1 j CO CO*1 1 1 1 1 1 1 1 1 1 1 1 CO CM uO 00 1C1 1 CO ! HO |CM|CM|CO|CM|CM|tH| I 1 1 1 1 1 l> CO* CM £ "So c Ö ho HO HO u0 HO HO HO HO HO HO *q HO HO 13 § V tH HO io cd © r> I> 00 00 05 05 O Ö H T-i CM cm’ CO CO* ^ tH* HO HO CD CD oH ^5 03 CM CM CM CM CM CM CMCMCMCMCMCOCOCOCOCOCOCOCOCOCO CO CO CO CO CO 120 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON T* X © i—1 G ill ne ts sta nd ­ ar d se t - 1. 7 co 2. 1 co 2. 5 CO 1© © lO co X CM O 05 Se in in g 19 61 , au tu m n G ill ne ts r- CM © 2. 5 © 1© © vd cd X tH 19 61 , re ca pt ur e p er io d G ill ne ts sta nd ­ ar d se t - I> rH © 2. 1 © 2. 5 © 1© © >d co X ▼H Pe rio d G ea r, f or gil l ne ts a lso m es h s iz e de sig na tio n (c f. T ab . 6 ) N um be r Total length, cm tT-ic'jc'icocô'ïHTjÎLftLocoçor'-r^ocodcscioo’-i^CNic'ico Cl I CM I CM I I I CO I CM I I ^ uOiOiOuOLiO»LiOLqLO»OuOLOuO iHCMCjfOCO^^LOiOCDCON^XOOOcioOHHCJClW IHHHHHHHHHHHHrtHrtHHNCMlMCMCKNCM ma ‘q}8ua| pqox FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 121 lOHOHOHOHOHOHOHOuOiOuOHOnOHOHOio COrHrHiOiO®OM>OOod05050di-lH(MCsiwc6Ttlt^OOOÖ05 CMCMCMCMCMCMCMCMCM I CM I CO I CO I CM I CM I ^ X|t^|X|l>.|cOll>|CM CO CO CM tH CO CO CM CM CO I CO I CM I CO » I HO I CO I CO nO I HO I HO I CO I hOuOlOhOhOhOUOhOhOhOhOhOhOhOhOHO POTH^kCiOCD!Dt>f>'oOo6c5CjodHHCldcOCO^^iOiOCOcDM>Xo6c5 CMCMCMCMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCO o> X d CM rH Sa © 9961 1/3 lO^O^O CMCiwcO'^’^tOtÔcOcdt^o'oOOOCîCiOOH'H hhhhhhhhhhhhhhhhNNCMÎM mo oi)Suo[ pïto.i. FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 123 iOhOhOhOhOhOiOhOhOhDiOlOiO (MWCOWTfi*^iCR5COOI>l>OOOCC50iod»- C5 © CO I> h i tjh i h i © I lo I co I ci rtt IhHtHiO |CMCM^COOOCOI>^ MM ! 10 ! CO I N I Oi MM H I CO I CO I O I N CO I © I -rH I 05 I CM CM I tH I CO I CD I 00 I rJH ! I CMCMCOCOTjHTHuOidcOCDl>l>OOod05C>OÖ rao ‘qiSioi IBJOX 13 2 /4 18 3 /4 7 7 1 1 8 48 8 10 .1 FISHING PKESSURE, IN A SMALL HIGH MOUNTAIN LAKE 125 HOHOiOHOHOkOnOLOnOHOHOHOLOHO ^T-i(MdfO«^^iCiO«Ocdl>t>’oOo6C!5oioÖHHiMoiM«Ttir)5iÖ Wd(M(MWlMlM(MfM(M(M!M(MiMC^CMCl(MCOCOCOCOWW«COMMCO r* CO HO j cm I th I CO I> I>; rJH H GO co I cm j ö i th i © l> TH CD CM CO CO CO CO CO ■>-< CO CO HO I O h Cl CO tH (M vH I H 05 CM CO O CO l> ' I O H C5 Cl C h CM Ci I h CO I CO CM I 00 I uo I TtH I CO CM HO 05 © CM CO I rfH ! I ! I I S3 HO d CO HO H CM CM I C1H I ! I CO I I ho cm’ I I CM CM I> 126 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON Total length, cm BC !hö &.a = T3 « s ’S .g »o o .reaS jo spm>[ [ju ‘SIBJOX suoods i ! 1 ! AV ^'l KJ S9!TJ UI! « .ioiio Suiqsij ÂJ.1 §umi9§ a> w B cö O M0 cg spnox s^o°H §niui9S t® O g O «D O CÖ Ä öSdH O s ä «,'U iO *q 05 0500’ icMC'icOCO-' ’^^^OOCßOt^r^OOOOOSCi© I i INI eo © »q »q © ci © © i ! i i i i I I I'M ! i ! i i N I iq iq iq ci J © I ^ J J CO oo cq oo • w i> 1-Î I k£5 I l> I rJH I 1C I © USWvOOOWtNCOOSOOOCOrH UI3 ‘qjSuOI IBjOX 20 .5 FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 127 oioicioiCkOLOLOiLCLOicursiLOiOiOiO HHiMiMCOW^^iCifîCOCÛ^t^OOOOOîOjoÔHHCKM’cOW^^iOldcDCÛtN CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCOCOCOCOCOCO cm i a i ho CO I 00 I 00 | I uO I CM CO I I CO I HO HO HO O CO* I I CM I ^ I CO I CM* CM I> 00 CO CO ho © I © I © I © I CD I HO CO CO tH CM CO CM I CO O0505H0h0C0C01>h0C0C0tH' COCOiOCO^OOOOlO^OOiOiM' OJJIQ -> uoipas siq; ui snouas jsoui sj9u ■<— p9qS9UI-9SJBOO JO 90U9Sqy -> O I « I O I TP I H I CO I - Il i II ! •> «x«i uv U.J ix«; uj HO lcj tes rirîSff'WîP^ÿiOiOCOCÔt^ts.'oOOÔOSOâOÔw^HCSICM’cOCO'TH^îiOàdçDcd^ CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCOCOCOCOCOCO M CM* C O r HO hO CO CO* !>■ I> 00 CO CO cq CM 0/D - G03 OD O ^ 128 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON rao ‘qi§uai pqoj. a o o 'S «C 50cS <« O o »O CO ÖJD .S°s o o CO i-rj e " C3 ^3 lOiOtfS^oi-OiOiOLq 05 CN|c4cOCOTHTHiC54f5COcdl>I> 00 00O5O5 03 \ I » I «O o © iO CO *5i> o « oo(/} •—( "3 " ctf-ön J» (M < ^ CO ^ I> co I co I co I i-i I a co co\ — I 05 lOOLO^O'OuOuOkO 05 03c4cOCOrt-I>.’GOod0505 ‘qjäuai p3)0X 3 2 /3 — 1 — 3 6 — — — 18 29 1/ 2 FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 129 kO kO kO kO kO kC5 kO kO kO kC uO OC'HHiMCM'fOCO^rHkOkOcC'COt^tNOOOÖOJOiodH tH CM CM CM CM CM CM 2 I 5 I §};S I ^ I " Or^TtHOs-^CMT-i^ J I CO t-h CM kO kO CM CO I I i-t I CM I CO CM CO CM I CM CM CM I CO I t> I O I CO CO CO CO CO CO CO ▼H CM tH tH tH ^“1 CD|O[00|CM|00|h>|^|CMI CO I CM I 00 I kO kO I rff ! 00 loo loo I t>- I CO "sf I CO I kO I CO I O I CD I kO CO CO CO CO T-l CM t-I tH kOkOkOkOkOkOkOkOkOkOkO OOHHiMM'cOCO’+^klîidcDcdtNo’ûCCOQCiod’H T-H CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCOCOCO ^ 00 CM 00 CD CM CM q cd CM «2 . OiD « § § c-1 9 130 T. LINDSTRÖM, A. FAGERSTROM AND K.-J. GUSTAFSON Total length, cm O c3T3 a) G & CÖ O O O cô CjT3 ' Gc3 CM* o O O CO uoioioiOLOiOLqioiotq COCO^T|JiOiOCCiCDl>l> XodGcioOHrHiMCvi ▼H tH tH tH tH t—I t-H ▼—I tH tH CM CM CM CM CM CM CO îO CD CD îD uO ^ ^ ^ CM |CM|rtllCO|CMlCO|eO iMkOCO^COuOOO^'HCOGClCO < » CM ^ CM ! I i co co co + + CO + + + + + iOdOuOlOuOiOuO^OIlOI-O C0C0rJi'!jîiCi0Gcdl>t>*’ 00 0dc50iodHHiMiM‘ HHHHHHHHtHHHHHHCMCMCMCMWCM Total length, cm 11 1/ 3 2 1 10 5 11 3/ 6 FISHING PRESSURE, IN A SMALL HIGH MOUNTAIN LAKE 131 ioiouOloioiouoilOvoioilOuo COCOTtl>-’oOOo"C505©©THi-4cMCM*COCO,'^TjHiO (MWfMWWWiMWOlWlMWlMlMCOMCOMCOCOMMMeOCO CD CD Tt< l> I CD CD CO CM CO CO CM 05 Th 1-1 CD O CD CM CO i I 00 cm I 00 CM CO CM tH CM h I h © 05 Tjî i-H CM 2 1e0!0 + + + CO CM rH + + + CD I O I CM 00 05 CD A 132 T. LINDSTRÖM, Å. FAGERSTRÖM AND K.-J. GUSTAFSON Table 12. Growth of trout from Lake Långbjörsjön and numbers of examined specimens (figures within brackets). Year Black-calculated length in centimetres at age: class 1 1 2 1 3 1 4 5 1950 4.7 (10) — 1951 4.8 (19) 9.2 (10) 1952 4.7 (35) 9.1 (19) 1953 4.8 (52) 9.1 (34) 1954 4.8 (74) 8.8 (52) 1955 4.8 (76) 9.0 (74) 1956 5.0 (100) 8.8 (76) 1957 4.9 (86) 9.6(100) 1958 5.3 (73) 9.5 (86) 1959 5.2 (189) 10.5 (73) 1960 5.0 (63) 11.4(189) 1961 5.0 (85) 10.6 (62) 1962 5.0 (48) 10.6 (83) 1963 5.2 (50) 10.8 (48) 1964 — 11.3 (50) 14.3 (10) — — 14.4 (17) 18.6 (10) — 14.4 (34) 19.7 (17) 22.6 (10) 13.7 (52) 19.5 (34) 24.0 (17) 14.9 (74) 18.5 (51) 23.9 (27) 13.8 (76) 19.2 (68) 22.5 (48) 15.0 (99) 18.3 (57) 22.9 (16) 15.8 (86) 20.0 (69) 23.4 (29) 16.8 (65) 21.4 (69) 24.2 (16) 17.21(120) 22.2 (32) 24.9 (16) 16.4 (62) 22.1 (78) 26.1 (5) 16.2 (81) 21.4 (45) 25.5 (40) 15.9 (48) 21.2 (68) 25.9 (36) Spatial distribution of pelagic fishes in Lambarfjärden (Mälaren, Sweden) with particular reference to interaction between Coregonus albula and Osmerus eperlanus 1 By T. G. Northcote and H. Rundberg British Columbia Fish and Wildlife Branch, Victoria and Institute of Freshwater Research, Drottningholm Contents I. Introduction ................................................................................................................................ 133 II. Methods ....................................................................................................................................... 134 Physical and chemical limnology.......................................................................................... 134 Gill net sets ................................................................................................................................. 135 Echo sounding .......................................................................................................................... 135 Dynamite explosions ................................................................................................................ 135 III. Some limnological features of Lambarfjärden ................................................................ 137 IV. Species and size composition of netted fish...................................................................... 142 V. Seasonal changes in vertical distribution ......................................................................... 143 Distribution indicated by netting.......................................................................................... 143 1. Cisco .......................................................................................................................................... 143 2. Smelt.......................................................................................................................................... 143 3. Perch.......................................................................................................................................... 143 4. Roach ........................................................................................................................................ 144 Distribution indicated by Echo-sounding .......................................................................... 146 1. Interpretation of echo traces ........................................................................................... 146 2. Results from echo-sounding.............................................................................................. 148 Distribution indicated by Dynamite Explosions ............................................................. 150 VI. Diel changes in vertical distribution .................................................................................. 150 VII. Seasonal and diel changes in horizontal distribution.................................................... 153 VIII. Discussion and summary ...................................................................................................... 158 IX. Acknowledgements .................................................................................................................... 166 X. References ..................................................................................................................................... 166 I. Introduction Interaction between species of fish apparently sharing resources in the environment such as space or food has been the subject of considerable descriptive investigation, of much controversy and of some experimental 1 The first of a series of papers resulting from a reciprocal research exchange arranged between the British Columbia Fish and Wildlife Branch and the Swedish Institute of Freshwater Research during 1968 and 1969. 134 T. G. NORTHCOTE AND H. EUNDBERG research in ecology (see especially Nilsson, 1967). In addition to important basic problems in ecological theory raised by species interaction, there are significant implications for those concerned with management of fish popula­ tions utilized for human food, recreation or both. A thorough understanding of the interactive processes between cohabiting species may suggest ways of optimizing production in waters where they already occur or of most effecti­ vely utilizing environments into which various species combinations may be introduced. Although interaction between trout and char has been extensively studied in Swedish lakes (Nilsson, 1965) as has that between various whitefishes (Svärdson, 1949; Nilsson, 1958, 1960; Lindström and Nilsson 1962), the relationship between cisco (Coregonus albula) and smelt (Osmerus eperlanus) has received surprisingly little attention, despite the wide distribution and importance of these species. Both are pelagic, schooling and plankton-eating fishes present in many lakes, particularly in central Sweden (Rosén, 1956). General characteristics of the biology and fisheries for cisco and smelt in Mälaren are given in Svärdson (1956, 1966), Rundberg (1968) and Vallin (1969). Of 110 lakes north of Dalälven discussed by Svärdson (1966), 71 contained cisco, 51 smelt but only 13 both species. Their limited coexistence results in part from events and conditions associated with dé­ glaciation, but unsuccessful attempts to introduce cisco into some lakes con­ taining smelt suggest that interaction between the species may also be a factor. During the summer and early autumn of 1968, observations were made on the spatial distribution of pelagic fishes in Lambarfjärden, a semi-isolated basin of the large, multibasin Lake Mälaren on the eastern coast of central Sweden (Fig. 1). In addition, the vertical distribution of pelagic zooplankton was recorded and the feeding of cisco and smelt examined. Only informa­ tion relevant to spatial distribution of the common pelagic fishes in Lambar­ fjärden will be presented here in an attempt to determine if segregation may occur between cisco and smelt in the pelagic region of the lake. II. Methods Physical and chemical limnology Vertical temperature profiles were made at Station A (Fig. 1) with a recently calibrated Freshwater Biological Association (F.B.A.) probe and with an Institute of Freshwater Research (I.F.R.) thermistor TIV using a Tettex AB Bridge 8706. Vertical oxygen profiles were determined with the F.B.A. oxygen probe. Delay in obtaining an extension cable for the probe restricted both temperature and oxygen series to the upper 23 m until later in the sum­ mer, however the major changes in these parameters occurred within this depth zone. SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 135 Subsurface illumination was measured with a Model 514 M Photovolt photometer modified for underwater use. Combinations of neutral density filters were used to take readings over the wide light range encountered; a scotopic filter (human eye) and opal diffusing glass were also in place for all measurements. The instrument was calibrated using a N.B.S. calibrated lamp and all readings were converted to approximate lux units. The photo­ meter response did not fit a cosine function but approximated it reasonably well (Fig. 2). Gill net sets Monofilament nylon gangs of nets, each 30 m long and 10 m deep and containing equal lengths of 19 and 27 mm (knot to knot) green mesh, were set at Station A in Lambarfjärden at approximately two week intervals be­ tween late June and early September. Net gangs were suspended from the surface and at 10, 20, 30 and 40 m depths usually for 3 hour fishing intervals near midday (about 1030—1330 hours) and midnight (about 2230—0130 hours). Replicate gangs were set at the two depth zones which seemed best to capture cisco and smelt. A supplementary series of day and night sets was made once at selected depths in onshore areas at Station B. All nets were marked by narrow uncoloured portions of the webbing at 2 m depth intervals which facilitated recording the depth distribution of fish in the sets. Fish were removed from nets on the shore and their depth in the net noted. Usually the entire catch was measured (total length), sexual condition recorded, and stomachs of all cisco and smelt preserved in a 10 % formalin solution. Stomachs of other fish species were sampled to cover the size range of specimens caught. For graphical presentation, all catches have been weighted, where necessary, to represent totals from a 3 hour fishing period by two replicate gangs in each 10 m depth zone. Echo sounding During most fishing periods echo sounding tracings were made near the net sets at Station A and over a transect from Station A to Lambarudd (Fig. 1). A Furuno F-701 50 Kc/sec sounder was used at Gain 6 to make all tracings; the depth scale was adjusted to compensate for the position of the transducer 0.5 m below the water surface. No objects could be recorded in the upper 2.5—3 m because the zero band on the sounder obscured that depth interval. Dynamite explosions One series of fish collections was made at Station A by exploding 100 or 400 gm charges of dynamite (Nitrolit) at various depths. Each charge was suspended from the surface by a small float and detonated electrically from 17 ° 5 0' 136 T. G. NORTHCOTE AND H. RUNDBERG Fi g, 1. La m ba rfj är de n st ud y a re a sh ow in g d ep th s in me te rs ; in se t sh ow s lo ca tio n w ith in La ke M äl ar en . A — sta nd ar d m id w at er ne tti ng sta tio n; B — La m ba ru dd , ec ho sou nd in g an d ne tti ng are a; C — a n et tin g ar ea use d fo r co m m er ci al fis hi ng ; D — the net tin g ar ea us ed by I.F .R . fo r a n a nn ua l ci sc o a nd sm el t sa m pl e. SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 137 a boat a few meters away. Echo soundings were made as above over each area immediately prior to and following detonation. Charges at each depth were spaced at about 50 in intervals in an easterly direction from Station A. Only the fish which floated to the surface were collected. Seagulls took several large cisco and up to a hundred or more smelt from each of the 400 gm charges; fewer surfacing fish were lost in this manner from the 100 gm charges. III. Some limnological features of Lambarfjärden Lambarfjärden, one of the deepest basins in Lake Mälaren, formerly was subject to periodic intrusions of saline water from the Baltic which produced severe oxygen depletion of its bottom water (Svärdson, 1956). Damming of the outlet at Stockholm has prevented further intrusions and rarely now have its deeper layers been subject to serious stagnation. Depths within the main netting area at Station A were close to 50 m; only a few restricted portions of the basin are deeper than 60 m (Fig. 1). Thermal stratification was well established in the latter part of June (Fig. 3) when the thermocline lay at about 6 m. Thereafter it became progressively deeper, reaching nearly 14 m by early July and ranging be- Fig. 2. Approximation of Photovolt Model 514 M photometer response to a cosine function curve (after modification of the instru­ ment for underwater use). COSINE FUNCTION DEGREES OFF AXIS W A TE R TE M PE RA TU RE 138 T. G. NORTHCOTE AND H. RUNDBERG Fi g. 3. M id da y (j oi ne d o pe n p oi nt s) and mi dn ig ht (so lid po in ts) tem pe ra tu re an d o xy ge n pr of ile s at St at io n A, Lam ba rfj är de n du rin g su m m er and aut um n, 196 8. Cro ss es giv e da ta sup pl ie d b y D r. T . A hl fo r n ea rb y s ta tio n o n 2 4 Ju ne , 31 Ju ly , 12 Se pt , an d 2 5 O ct ., 19 68 . SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 139 tween 10 and 15 m throughout summer and early autumn. The thermocline had disappeared by late October when the basin was nearly isothermal. Evidence of internal thermal seich action was suggested by changes in mid­ day and midnight temperature profiles at Station A, but usually shifts in isotherm levels within 24 hours were only a few meters at most. No marked oxygen depletion was noted in Lambarfjärden throughout the summer (Fig. 3), and even in near bottom water at 50 m dissolved oxygen concentrations rarely fell below 7 mg/1. Surface layers periodically were supersaturated. At midday in Lambarfjärden, the lower limit of sensitivity of the photo­ meter (approximately 0.001 lux) was reached at a depth slightly in excess CO) (o) (O) (20) (o) IO'1 -•-----1'-»------------------ ----------------------------------- io-2 ---------------------- IO'3 MIDNIGHT (o) -• io-3 COcr ÜJ f— LU o IO z X I— a LÜ o 20 30 (50) (O) (50) (lOO) (30) flO) ( 40 ) MIDDAY 40 50 ] 30 JUNE j i IO 20 JULY 30 J L IO 20 AUGUST i 30 J IO SEPT Fig. 4. Midday and midnight levels of subsurface illumination at Lambarfjärden Station A during the summer of 1968. Numbers in parentheses give approximate percent cloud cover. LI G HT INT EN SI TY - L UX 140 T. G. NORTHCOTE AND H. RUNDBERG SMELT CISCO Fig. 5 a and b. Length-frequency distribution of the four species of fishes commonly taken by pelagic netting in Lambarfjärden during summer and autumn, 1968. SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 141 ROACH 8 10 12 14 16 10 20 22 24 26 20 30 PERCH 25 - 142 T. G. NORTHCOTE AND H. RUNDBERG of 20 m (Fig. 4). The rise in isolux levels evident in the mid and late July series probably was a combined result of cloud cover and very dense blue- green algal blooms. Midday surface levels of illumination were close to or slightly above 10,000 lux throughout summer. The 0.001 lux level at mid­ night occurred near 5 m in midsummer and virtually at the surface later in the summer. Midnight surface readings ranged from about 0.6 lux in June to less than 0.01 lux in late summer. Extensive studies of the general limnology of Lake Mälaren are carried out by the National Nature Conservancy Office. Limnological Survey (for­ merly the Lake Mälaren Research Group). Their numerous reports and publications may be consulted for further details. IV. Species and size composition of netted fish Of the nine species of fish taken in the netting series at Lambarfjärden Station A, only four were common throughout the season. These were cisco (Coregonus albula), smelt (Osmerus e perla mix), perch (Perea fluviatilis) and roach (Rutilus rutilus). A few bleak (Alburnus alburnus) were netted near the surface in June and a single bream (Abramis brama) there in July. Occa­ sionally a burbot (Lota lota) or a ruffe (Acerina cernua) was captured, usu­ ally at night near the bottom. Pike-perch (Lucioperca lucioperca) were rarely taken until August when a few were caught at night in the epilimnion; they were more frequently captured in autumn, and at greater depths. Two addi­ tional species known to occur in the pelagic region of Lambarfjärden but not taken in the netting series are threespined stickleback (Gasterosteus acule- atus) which are abundant and whitefish (Coregonus lavaretus) which are rare. Two size groups of cisco were netted (Fig. 5), the smaller gradually in­ creasing in length from early summer to late autumn while the modal class of the larger group remained between 250—260 mm throughout the season. On the basis of age determinations made from scales of cisco netted at Sta­ tion D (Fig. 1), most fish in the smaller size class would be 1+ year old, and those in the larger would be 2+ to 4+ years old. Most smelt netted ranged between 80 and 160 mm and would be 1+ to 3+ year old fish. The few larger individuals probably were 4 to 5 years old. Although distinct size classes of perch and roach were evident in the catch (Fig. 5), no attempt was made to estimate their age. SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 143 V. Seasonal changes in vertical distribution Distribution indicated by netting 1. Cisco In June and early July cisco were often taken during midday at depths less than 10 m in nets (Fig. 6). At midnight largest numbers also were caught near the surface, but many more fish were found in deeper layers than during midday. Later in July and in early August very few cisco were caught in the upper 10 m, either during midday or midnight. Most were netted at this period in the 10—25 m depth zone, and many were taken near the bottom (40—50 m). In late August and early September, cisco were again taken near the surface at midday, although they were still abundant in the 10—25 m zone and present in near-bottom levels, particularly during the day. By late October, cisco were broadly distributed over the entire 50 m, being most abundant in the 15—20 m zone during the day and somewhat deeper at night. No close association was evident between seasonal changes in the vertical distribution of cisco and thermal stratification, although their descent into deeper layers before mid July occurred at a time when the thermocline was becoming deeper and when temperatures were relatively high in near-surface layers (Figs. 3, 6). There appeared to be a considerable difference in vertical distribution of small (chiefly 1+ year old) and large (2+ to 4+ year old) cisco, particularly during mid summer and early autumn (Fig. 7). The former were rarely taken below 20 m while the latter were often caught in deeper layers, some near the bottom. 2. Smelt In contrast to the cisco, at no time during the summer were many adult smelt caught in the upper 5 m (Fig. 6). During July and August, most smelt were caught at night between 10 and 20 m. A gradual spread and deepening in their midday vertical distribution was suggested, since they were rarely taken in the 30—50 m zone in June or July but more frequently so in late summer and autumn. 3. Perch Large numbers of perch were taken in near-surface waters (above 15 m) throughout July and August (Fig. 8), particularly in midday sets. For most of the summer relatively few perch were caught at the surface (0—4 m) during midday, and most were concentrated between 6—10 m. In early and mid July many small perch were netted at night near the surface. Later in the summer they either were not present in the pelagic region at night or were not moving so as to be caught there. By autumn very few small or large perch were captured in the pelagic area at any time. DE PT H I N M ET ER S. k „ „ DEPT H IN ME TE RS 144 T. G. NORTHCOTE AND H. RUNDBERG MIDDAY 29 OCT.SEPT.22 AUGUST2 3 JULY 5 AUGUST13 JULY24JUNE20JUNE JULY MIDNIGHT CISCO (COREGONUS ALBULA ) MIDDAY 29 OCT.SEPT.5 AUGUST 22 AUGUST2 3 JULY2 JULY 13 JULY24JUNE20 JUNE MIDNIGHT SMELT (OSMERUS EPERLANUs) |>1FISH Seasonal changes in midday and midnight vertical distribution of cisco and taken in gill nets set at Station A, Lambarfjärden (catch combined for each 5 m depth interval). Open and solid circles show depth of the 15 and 10°C isotherms, respectively. 4. Roach Only recently have roach become abundant in the pelagic region of Lake Mälaren according to reports from commercial fishermen and others (Rund­ berg, 1968). In 1968 they were present in near surface waters there by EA RL Y S UM M ER M ID -S UM M ER EA RL Y AU TU M N LA TE AUT U M N SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 145 Fig. 7. Seasonal changes in the vertical distribution of small and large size classes of cisco and smelt taken in pelagic netting at Station A, Lambarfjärden, 1968. early June and many were caught in midday and midnight sets from late June to early August (Fig. 8). Unlike the perch, they were usually most abundant at the immediate surface (0—2 m) and rarely were taken below 6 m. In late summer and autumn very few were caught at Station A. 10 146 T. G. NORTHCOTE AND H. RUNDBERG i T \1 ! 'if1 ( : ! MIDDAY R - S f - — a - < r 1 ! --1 1 - : ■ 1 i ROACH ( Burnus BUIIUS) D-5 FISH Fig. 8. Seasonal changes in midday and midnight vertical distribution of perch and roach taken in gill nets set at Station A, Lambarfjärden (catch combined for each 2 m depth interval). Distribution indicated by echo-sounding 1. Interpretation of echo traces The discrete inverted v marks evident on the echo traces in Figure 9 are most probably fish. They undergo rapid vertical changes in depth distribu­ tion (as will be shown later) and appear very similar to those strongly SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJARDEN 147 Fi g. 9. M id da y (u pp er ) an d m id ni gh t (lo w er ) ech o tr ac es at St at io n A , La m ba rfj är de n ad ja ce nt to gi ll n et se ts s ho w n i n F ig ur es 6 an d 8 . 148 T. G. NORTHCOTE AND H. RUNDBERG implicated as fish in other lakes (Northcote et. al., 1964; Maitland, 1968; Dembinski, MS 1969). Even if all these markings could be ascribed solely to fish, it still would be difficult to determine with any certainty the species indicated. Nevertheless some progress can be made in this direction. Obviously the density of marks shown on the traces would suggest, if they are indeed fish, that they are quite numerous. Therefore, unless most are of a size too small to be taken in the extensive netting series, then only 4 species would seem to be primarily involved — i.e. cisco, smelt, perch and roach. Because perch and roach are largely restricted to the upper 10—15 m throughout the summer (Fig. 8), most of the marks below that depth prob­ ably represent cisco or smelt, both of which are known to be abundant in deeper layers of Lambarfjärden. Additional evidence that many of the marks below 15 m indicate cisco and smelt comes from echo tracings made adjacent to commercial and I.F.R. standard netting areas for these species (Fig. 10). On echo tracings made over a commercial fishing area for cisco (Fig. 10 A) a concentration of broad inverted v marks can be seen near the bottom at depths between 15 and 25 m. The total catch in sets made there was comprised largely of cisco and a few smelt. Similarly at the I.F.R. netting area, a somewhat less distinct layer of broad inverted v marks can be seen on the traces between 15 and 30 m (Fig. 10 B). Unfortunately the marks are obscured by static from the sounder which was malfunctioning at the time. Catches from gill nets set along the bottom in the region of the layer of marks again were dominated by cisco and smelt. Furthermore, the concentration of marks on the midday echo tracings at depths between 15 and 25 m suggesting schools of fish are very similar in appearance to the midwater layers recorded by Schärfe (1951) in Grosser Plöner See, by Lieder (1962) in Schweriner Aussensee, by Maitland (1968) in Lochmaben and by Dembinski (MS, 1969) in several Polish lakes. All interpreted these layers as being largely composed of schools of cisco or a similar species. Collections of fish from experimental dynamite charges (Figs. 11, 12) show that cisco are present and smelt abundant in the 15—30 m depth zone where the characteristic marks are evident on the echo tracings. 2. Results from echo-sounding Obvious seasonal changes are evident in the distribution of marks on midday and midnight traces from Station A, Lambarfjärden (Fig. 9). In June and early July, midday traces show concentrations of marks in two depth zones, above 7 m and 12—25 m. The upper zone probably results from cisco, perch and roach which are abundant then in that region (Figs. 6, 8), while the lower zone may be comprised of cisco and smelt, although only small numbers of the latter species were taken there by netting. Very few SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 149 A OFFSHORE""ONSHORE SET w 20 21351235 1240 2140 TOTAL CATCH (OBOO JULY2 — 0030 JULY3^) COREGONUS ALBULA ca 1500 (j50 kg^) OSMERUS EPERLANUS 75 Perca FLUVIATILIS I Rutilus RUTILUS O 1415 TOTAL CATCH 1400 JULY iO co IOOO JULY I 2j COREGONUS ALBULA _______ 317 OSMERUS EPERLANUS _______ 108 Perca FLUVIATILIS _______ very few, if any Rutilus rutilus ............... .......................... Fig. 10. A — Echo tracings adjacent to commercial cisco netting area (Station C, Fig. 1) and associated net catch. Broken lines show approximate depth of nets. B. — Echo tracings adjacent to I. F. R. standard cisco and smelt netting area (Station D, Fig. 1) and associated net catch. Broken lines show approximate location of 6 m deep nets fished on bottom. 150 T. G. NORTHCOTE AND H. RUNDBERG marks were recorded in the 30—50 ni zone and virtually no fish were taken in June and early July net sets there during midday. At night, particularly in early July there was an increase in marks recorded in the deeper layers which seemed to be reflected in catches of cisco and smelt (Fig. 6). High densities of marks in near-surface layers (above 10 m) at night, especially from June to mid July may be caused by small smelt and perch. From mid July to early August there is a progressive concentration of marks within the 12—25 m zone at midday as well as a marked increase in those recorded from 30—50 m. This fits reasonably well with seasonal changes noted previously (Fig. 6) in vertical distribution of cisco and smelt, particularly the former. Evidence of dense schools of fish, probably cisco and smelt, can be seen in many of the midday traces from July and August. These “schools” seem to disperse in the midnight traces which suggest a more scattered distribution of large numbers of individuals. General dark­ ening of the traces in August below 20 m may be caused by concentrations of smelt. Some evidence for this possibility will be given in the following section. Distribution indicated by dynamic explosions Cisco were not taken in the series of daytime explosions at near-surface depths (5, 10 m) but were collected from deeper charges (15, 20, 30 m) set at Station A (Figs. 11, 12). Results of the 100 gm series (Fig. 11) suggest that small smelt (40—65 mm) were concentrated in the 10—20 m zone whereas larger smelt (> 80 mm) were more abundant in deeper layers. Relative pro­ portions of the two size classes taken by 400 gm charges at 20 and 30 m also indicated such a trend (Fig. 12). Large numbers of smelt were collected from the 400 gm charges. Echo traces made several minutes after the explosions showed that much of the fine scattering evident at depths near the charges prior to detonation had disappeared (Fig. 12) but within an hour was re-established. Possibly the fine scattering shown on the echo traces at depths below 15 m was primarily caused by smelt, whereas the broad, heavy inverted v marks were chiefly cisco. If the above interpretation is valid, then one might infer from the tracings in Fig. 9 that in addition to the gradual deepening of cisco distribution throughout the summer there was a similar deepening and increase in abun­ dance of smelt in 20—50 m layer. VI. Diel changes in vertical distribution Attempts were made in early June, 1968 to follow diel changes in vertical distribution of fishes by gill netting at Station A, Lambarfjärden. This did D EP TH IN M ET ER S 151SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJARDEN DEPTH CISCO SMELT PERCH Fig. 11. Echo tracings associated with surface collections of fish rising from submerged explosions of 100 gm dynamite charges at Station A, Lambarfjärden, 0900—1200 hours, August 30, 1968. Minutes relative to detonation (time zero) given beneath each Irace. 152 T. G. NORTHCOTE AND H. RUNDBERG -6 +1 +1 4 +4 7 +7 2 Fi g. i2. Ec ho tra ci ng s as so ci at ed wi th sur fa ce col le ct io ns of fis h ri sin g fr om sub m er ge d exp lo sio ns of 40 0 gm dyn am ite cha rg es at St at io n A , La m ba rfj är de n, 140 0— 16 00 hrs , Au gu st 3 0, 196 8. Mi nu te s re la tiv e to det on at io n (ti m e ze ro ) giv en ben ea th eac h tr ac e; tota l nu m be rs of ea ch siz e cl as s o f ci sc o a nd sm el t gi ve n a t ri gh t of tra ce s, le ng th -fr eq ue nc y di str ib ut io n sh ow n fo r sm el t on ly ; up pe r an d lo w er se rie s sh ow res ul ts f ro m ex pl os io ns at 20 an d 3 0 m res pe ct iv el y. SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN 153 not prove feasible with equipment and staff available. Evidence of diel verti­ cal fish migration may be inferred from echo sounder tracings (Fig. 13), although they are difficult to interpret because marks of individual species cannot be recognized. Nevertheless most series from early to late summer indicated an evening ascent of fish from midwater layers (10—30 m) towards the surface, a concentration there prior to midnight and a dispersal and deepening of vertical distribution after midnight. By morning, a midwater concentration layer was reformed. Because perch and roach were largely confined to depths above 10 m, cisco and smelt were the species most likely responsible for the daytime midwater concentration layer and its apparent vertical nocturnal migration to and from the surface as indicated in the echo sounder traces. Throughout summer the concentration zone during midday was usually found from 15—20 m, where light readings ranged from 0.1 to 0.001 lux (Fig. 13). Midnight levels of illumination near the sur­ face were about 0.1 lux in June but dropped to less than 0.01 by late August. Suggestions of a dusk rise, a midnight dispersal, and a morning descent of fish are evident in several of the series (Fig. 13). More complete echo sounding traces were made in September (Fig. 14). Movement of marks (assumed to be fish) from the midwater concentration layer had started by 1830 hours on September 6 (Fig. 14 A), coincident with a rise in the illumination isopleths towards the surface. Dusk occurred some­ what earlier on September 9 due to cloud conditions; onset of fish movement towards the surface and rise of light isopleths also occurred earlier then than on the evening of September 6 (cf. Fig. 14 A and B). Early evening rise of fish from a 15—25 m concentration layer is clearly indicated in the series of closely spaced tracings made on September 16 (Fig. 14 C). “Surfacing” schools of fish thought to be adult and sub-adult cisco were first seen at 1635 hours and many groups were noted between then and dusk. Concurrent catches in nets suggest that cisco were present in the concentration layer and that they moved towards the surface at dusk, however, the pattern is con­ founded by the numbers of roach caught near the surface then. Additional netting in the autumn of 1969 confirmed that the surfacing schools of fish which appear in late afternoon are indeed cisco. Descent of fish to the concentration layer was indicated at dawn (Fig. 14 A) when light isopleths were also moving downwards. VII. Seasonal and diel changes in horizontal distribution Although the investigation was primarily concerned with vertical distri­ bution of fishes in the pelagic region of Lambarf järden, some information was obtained on their distribution between Station A and a prominent onshore point (Lambarudd), shown as Station B in Figure 1. One netting series was DEPTH IN METERS is Ju ly 12 u i6 is 20 22 24 02 04 06 08 10 Julyu SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJARDEN 155 DEPTH IN METERS p*> C/I g 03 £ 00- 3 4)aap 00 .S*Sh rO G 0)T3 f-l 03 £S c3 hJ g .S "3 c/D c3 0)X5 03 a GO 00 .a■'S GGO »3 O ,G 03 03 aO .*-4 en 00G O 03 (H 03 VJ 03 a X o C/2 .a 03 VJ 03 5 pd r0 3 CÖ CO ob E 156 T. G. NORTHCOTE AND H. RUNDBEKG DEPTH IN METERS SE PT . 6 H O U RS 157SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMBARFJÄRDEN o1 j&jçov .raste i Silllü « 2|p| 1 1 ÂÉIr i F .;■■ ■ '■ : V .' !Ë . -Ö^Stl 21 ■» 121<■> ! « I |n|o| I M ! I .5 u a «o 3 a £ Z. h C ^ -G •G "o 00 00 a01c/2 . -2 G 00 G o CO o 2 1 s' g S g■G ^G o) G Cfl GO G '—' CM GT5 GaiT3u :G s ü ’p *s *S -G G G- •G oi TO G pG s g tj03 G 'Is. 0.«3 •— f-H 0. •GTO G < G 2 t-G o ”aT G « £ *C o o) Gw 'S oq'—' m d Qi T5 7S 'G iO o I> Ocfl I co1 LOO CO-G 01 %*01 01 C g rGg G o 01 OS-t«♦1 "O-t G CO 00 01 c/3 V3 .2 tH *_» *ü O G £ G <4* g o 2 .Bf x .2 fe 3 DEPTH IN METERS 158 T. G. NORTHCOTE AND H. RUNDBERG made in midsummer in this area (Fig. 15). During midday, cisco (plus a few smelt and perch) were taken in the 20—30 m bottom set, a depth zone corresponding to the most dense portion of the concentration layer shown on the echo sounder traces. A few perch and pike-perch were taken in the 10—20 m bottom set, which may be represented by the scattered marks shown on traces in that region. Perch and roach were abundant in the onshore set at 10 m. Between evening and midnight the concentration zone on the echo traces had moved upward and onshore; at the same time cisco and smelt were netted in considerable numbers in the 10—20 m depth zone, a region where they were not captured during the day. Few perch were caught in any area whereas roach were confined to the onshore set. Results from this com­ bined netting and echo sounding series would indicate shoreward movement of cisco and smelt in the evening associated with their concomitant vertical migration. Seasonal changes in horizontal distribution of fish at Station B may be interpreted from echo sounding series (Fig. 16). As shown in traces at Station A, the midday concentration layer (probably largely cisco and smelt) gradu­ ally deepens and becomes denser throughout the summer. The concentration layer evident in the pelagic region at Station A extends to the corresponding depth onshore, and during ascent of the layer at dusk an onshore movement of fish forming the layer must occur. Although early morning traces are not shown for the summer period, those taken indicate that a corresponding offshore horizontal movement also occurs at dawn. An onshore-offshore pattern of movement associated with diel vertical migration of fish forming the midday concentration layer is evident in the more complete autumn echo sounding series (Fig. 17). VIII. Discussion and summary The inability of nets to adequately record distribution and abundance of smelt, the problems encountered in interpreting and distinguishing species on echo-sounding traces, and the limited use of dynamite explosions for fish collection all make it difficult to determine the extent to which cisco and smelt may be spatially segregated in the pelagic region of Lambarfjärden. Nevertheless differences were evident in the vertical distribution of these and other fishes there, and it would seem possible to indicate roughly the depth zones utilized by each species during summer and early autumn. Information obtained by netting, echo-sounding and dynamite collections will be used to develop a generalized picture of spatial distribution of the fishes. Because of obvious limitations of the data, the following outline can only be con­ sidered a tentative one and several features require much more detailed study. CI SC O SM EL T PE RC H RO AC H SPATIAL DISTRIBUTION OF PELAGIC FISHES IN LAMB ARE JÄRDEN 159 Ö 03 o 0)0) 03(M !1 O o02 0> Cß . „ P ns < ns 33 u 0) OS03 -Q 02 sc3’S hJ0 nsJZj P *5b 03 Oh 00)0203 ns xs t-1 :o3 a 4S*C3 pQo a 03 P x a 2 .2 o M % cS O ^ fi SO 4) -C 01 <73 r* £ S 'O -Q ÏC Ü C f3 O 0ß 03 two rt ö 2 O G 4) 02 -G g 5 IÜ C3 G ns -G rQ ÖO