ANACAT-Theme session T May stocking programs effect the predator stocks and decrease the survival of the wild Atlantic salmon juveniles ?

Stocking of Atlantic salmon juveniles is carried out in many rivers to increase the production of smolts. In many cases, the number of stocked juveniles exceeds the assumed carrying capacity of the river, and represents a conciderable and instantaneously increase of the fish biomass. Normally, the stocked juveniles suffer from high predation mortality, presumably higher than for native fish. Therefore, the introduction of stocked naive prey may influence the number and size of predators. When the stocked juveniles has been depleted, the predation upon their wild conspecifics might increase. In this paper we will discuss possible relationship between the survival of the wild juveniles and the introduced biomass, the time of stocking and its influence on the stock of predators. Introduction Stocking of juvenile stages of Atlantic salmon in rivers has been performed throughout this century and in a high number of rivers to enhance stocks of Atlantic salmon. Initially, alevins at the yolk sac stage was the most frequently used for stocking. Following improved culture techniques and investments in local hatcheries, there has been a clear trend towards stocking with older juveniles and occasionally smolts ready to migrate to the sea. Stocking of Atlantic salmon in Norway is performed to maintain fisheries in regulated rivers and where the original stock is lost or heavily reduced due to acidification (Hesthagen & Hansen 1991) or due to infestations of the parasite fluke Gyrodactylus salaris (Johnsen & Jensen 1991). The stocking in regulated rivers is usually carried out to enhance fisheries under the assumption that the natural production has been lowered. A release of hatcery-reared young salmon into the river environment and the effect on vertebrate predators may in some sense be compared to a rapid build up of short-lived, cyclical prey populations in terrestrial environments. Examples of such terrestrial prey-predator relationships are: snoeshoe-hare-fox, moose-wolf, ptarmigan-fox, small rodents-birds of prey among many others. The common features of these prey populations are rapid growth, very high reproductive rate, short livfespan, and highly fluctuating population densities. The predators are longer lived and their response to a temporary increase in prey availability is an increase in number and survival of their offspring leading to an enlarged predatory stock. When the temporary prey is depleted or reduced due to other densitydependent factors, the increased number of predators are forced to switch their feeding to other prey which experience an increased predation pressure. We have chosen an approach based on the examples from West Norwegian watercourses hosting populations of anadromous Atlantic salmon and brown trout. In the part of the river inhabited by anadromous fish, Atlantic salmon and brown trout are the dominating fish species, in addition European eel may be found. In lakes situated at the parts accessable for anadromous fish, brown trout and Arctic cham are normally the dominating original fish species, followed by European eel and occassionally populations of three spined stickleback. In contrast to the almost obligate anadromous Atlantic salmon, the populatons of brown trout consist of both anadromous and resident individuals (Jonsson 1985,1989), but there are little or no genetic difference between sympatric populations of anadromous and resident trout (Skaala 1992, Hindar et. al. 1991). Young fish in the rivers are predated upon by conspecifics, however to an unknown extent. Other predators are mink, otters, heron and mergansers, but river resident brown trout is regarded as the most important predator in larger rivers. The present study deals with stocking of young Atlantic salmon of the year during August and November in rivers where natural recruitment of Atlantic salmon takes place, i.e. supportative breeding. At the time of release, the high number of predator naive fish increases the total number of young fish considerably and one should expect increased level of aggression, territorial conflicts and probably some state of territorial chaos between stocked and wild conspecifics. Also, there might be an increased mortality through cannibalization of wild specimen (Evans and Willox 1991) and indirect effect through the build up of predatory species in response to large scale releases (Wood 1986, Beamish et. a1 1992). In this paper we will describe possible effects ecological of stockings. The knowledge on interactions between the stocked fish and structural changes in the predatory stocks is however limited. This discussion will therefore be based upon, and discuss the following assumptions that may be fulfilled in some rivers: 1) The carrying capasity in the river is reached during late winter. 2) The production of wild fish is at, or close to, carrying capasity. 3) The mortality of stocked juveniles is considerable higher than that of the wild. 4) The biomass of juvenile salmon change seasonally with a maximum during late autumn. 5) The introduced biomass of stocked fish is high in relation to the wild biomass. 6) Because of their sizes, the stocked juveniles are available as prey for the trout. 7) The biomass of the stocked juveniles is sufficient to increase the number andlor size of predatory trout. 8) While the majority of the stocked juveniles are depleted, the enlarged stock of piscivore trout represent a long-term increase in the predation pressure upon survivors and the wild juveniles. 1) The carrying capasity in the river is reached during late winter. It may be anticipated that the carrying capacity of the river is basically determined by bottlenecks during winter. The mortality of young salmon and trout might be high during winter due to habitat compression and resticted overwintering area (Hvidsten 1993, Fjellheim et al. 1995, Gibson 1993). Winter mortality of larger fish, i.e. those that spend their last winter in the river before migrating to the sea as srnolts has been estimated to 30% during the last winter in freshwater (Gibson 1993). However, in the regulated river Orkla, the production of Atlantic salmon smolts increased after regulation. This was ascribed as an effect of higher and less variable discharge during winter and increased area of suitable habitats (Hvidsten 1993, Hvidsten and Johnsen 1995). It is not known whether the reduced natural mortality was caused by reduced competition or reduced predation. 2) The production of wild fwh is at, or close to, carrying capasity. We assume that the abundance will reach carrying capacity during winter irrespectively of the abundance in late autumn. Suggestively, some of the stocked fish will survive at the expence of wild conspecifics if the number of wild juveniles was high enough to reach winter capacity levels. The production of smolts varies a lot between rivers and there is a general trend that smaller second-order rivers are more productive than larger rivers with higher discharge (Gibson 1993). Within a river, the production of smolts varies less between years than the total density including all yearclasses of juvenile fish, which means that the pattern of mortality varies considerably between years depending on number of recruits (age-0) and the carrying capacity of fish of the size of smolts (Lacroix 1989, Gibson 1993, Saltveit 1993, Bagleni2re et al. 1994, Hvidsten & Johnsen 1995, Jensen et al. 1995, SaksgArd et al. 1995). 3) The mortality of stocked juveniles is considerable higher than that of the wild. There are very few studies on the survival of stocked juvenile Atlantic salmon compared to the wild ones and interactions between the effects of stocking and the survival of wild. Survival of stocked brown trout have been estimated in the regulated River Teigdalselva in Western Norway (Fjellheim et. a1 1995). A number of 70.000 brown trout parr (age-0) with average length of 6.4 cm were stocked in late June in 1992. The total density of juvenile trout increased more than tenfold after stocking with maximum densisty of 684 individuals per 100m2. Until 10 October 1992 the density of stocked fish was still very high at and near the stocking sites. During this period the stocked fish spread to a certain extent, and predominantly upstream, but they did not cross the river, showing that dispersal after stocking lasts for a long period. The next spring, in April 1993, the total density of brown trout juveniles was at the same level as in April 1992, prior to the stocking. The mortality from stocking until April 1993 was estimated to 99% for stocked fish and 79% for wild brown trout, all juvenile age-classes included. It is not known if survival of wild fish was affected negatively by the stocking, but stocked brown trout cannibalizated smaller young conspecifics (Fjellheim et. a1 1995). The result from this study clearly indicates that irrespecticve of density in autumn, there is density dependent mortality over the winter reducing the density to the level of the carrying capacity and also that the mortality was highest for the stocked fish for a long period after stocking. In a neighbouring, regulated river, the density of brown trout increased 10 times in the area upstream a weir during the years following the weir construction, presumbaly due to enlarged winter areas increasing survival (Raddum et. al 1989). Although these examples refer to brown trout, the main results might well apply for Atlantic salmon. 4) The biomass of juvenile salmon change seasonally with a maximum during late autumn The density and biomass varies conciderably within a river and between seasons. The biomass is lowest in spring and early summer after the smolts have left the river and is probably highest just in the end of the growing season, i.e. end of October. Due to winter mortality and low growth under low temperatures one should assume that there is a gradual reduction of fish biomass during winter, and then an abrupt decrease