Gaupas predasjon på sau – en kunnskapsoversikt.

Abstract

Gaupa er igjen å finne i store deler av Norge, og konfliktene rundt gaupe og sau er større enn de noen gang har vært. I følge erstatningstallene har en estimert bestand på mellom 44 og 92 familiegrupper siden 2000 årlig drept fra 6 125 til 10 093 sauer. Etter dagens lovverk skal alt tap til store rovdyr erstattes. Imidlertid blir kun en liten andel av de omsøkte tapene dokumentert gjennom en uavhengig kadaverundersøkelse, og den resterende andelen av de udokumenterte tapene blir erstattet etter en subjektiv vurdering gjort av forvaltere på fylkesnivå. Inntil nå er disse erstatningstallene ikke evaluert. For å redusere tapet av sau til gaupe er vi avheng av å forstå mekanismene bak gaupas predasjon på sau. I denne rapporten oppsummerer vi forskning på konflikten gjennomført de siste to tiårene på oppdrag fra nasjonal og regional forvaltning. Forskningsprosjektet Scandlynx har tallfestet hvor ofte gaupene dreper sau (drapstakt) i ulike deler av Norge, og hva som forklarer den observerte variasjonen i drapstakt. Vi benytter denne kunnskapen til å evaluere dagens ordning for erstatning av tap av sau på utmarksbeite og dagens forvaltning av konflikten. Tallfesting av gaupas valg av byttedyr og hvor ofte de dreper sau (drapstakt) ble gjort ved å følge 78 gauper med VHF- eller GPS-sender fra 1995 til 2013 i Sør- og Nord-Norge. I perioder er gaupene fulgt intensivt i beitesesongen. Metodene vi har benyttet har endret seg etter som ny teknologi ble tilgjengelig med årene. Store deler av de sørlige studieområdene har frittgående sau på beite i skog og fjellområder fra juni til september. I sør er rådyr det viktigste alternative byttedyret til sau. Studieområdet i Troms og Finnmark er dominert av tundra og fjellbjørkeskog. Fordelingen av sau er mer flekkvis enn i sør, tamrein er tilgjengelig i høye tettheter i hele området og rådyr er fraværende. Analyser av byttedyrvalg i de ulike områdene viser at gaupa foretrekker rådyr eller tamrein framfor sau hvis de er tilgjengelige. Vi ser videre at andelen av sau i sommerdietten minker med økende tetthet av alternative byttedyr kombinert med minkende sauetetthet. I Sør-Norge påvirker gaupenes kjønn, samt variasjon i tetthet av både rådyr og sau gaupenes drapstakt på sau. De høyeste drapstaktene fant vi hos hanngauper i områder med lave tettheter av rådyr og høye tettheter av sau. Når tettheten av rådyr øker og sauetettheten minker, så dreper gaupene færre sau per tidsenhet.

Following their recovery during the last 30 years Eurasian lynx can once again be found in large parts of Norway. Unfortunately, the conflicts caused by lynx depredation on sheep are greater than they have ever been, According to official data an estimated lynx population consisting of from 44 to 92 family groups (annual reproductions is an index of population size) has resulted in the payment of compensation for between 6 125 and 10093 sheep each year. Following the present legal frames, all losses to large carnivores should be compensated. However, only a small proportion of the compensation claims are based on animals that have been subject to an autopsy. The remaining undocumented losses are evaluated using subjective criteria by wildlife managers at the county level. These compensation numbers have never been evaluated. If there is any chance to reduce lynx depredation on sheep there is a need to understand the ecological mechanisms behind the conflict. In this report, we summarise the last 20 years of research that has been financed by regional and national wildlife management agencies. The Scandlynx research proiejct has quantified how often individual lynx kill sheep (kill rate) in different areas of Norway an dtried to explain variation in this parameter. We use this data to evaluate the present compensation system and other aspects of how the conflict is managed. Quantifying what lynx kill and how often they kill it has been done by following 78 individual lynx equipped with GPS and VHF collars (the technology has changed greatly during the course of our study) from 1995 to 2013 in south and north Norway. During intensive periods these lynx are followed closely, and personnel investigate areas where the lynx remain stationary for any periods that could indicate they have made a kill. In our southern study areas roe deer constitute the main prey of lynx, but unguarded and freeranging sheep are present in a large proportion of the forest and alpine tundra habitats from June to September. In our northern areas, semi-domestic reindeer are the main prey of lynx as roe deer are absent, and sheep are widespread, but occur in smaller patches than in the south, although they are also free-ranging and unguarded. The habitat consists of mountain birch forest and alpine tundra. Our results clearly indicate that lynx have a strong preference for killing roe deer (in the south) or reindeer (in the north) as opposed to sheep. We have also shown that the proportion of sheep killed decreases in areas with higher densities of their preferred prey and with lower sheep density. In south Norway we have seen that kill rates on sheep are influenced by the sex of the lynx and the variation in density of both roe deer and sheep. The highest kill rates are from male lynx in areas with low densities of roe deer and high densities of sheep. Lower kill rates were found in areas with high roe deer density and low sheep density. In northern Norway we did not find any difference between the sexes, although the chances of a lynx killing a sheep increased in area with fewer reindeer. Although we saw some individual variation in kill rates we did not find any evidence for the existence of ”problem individuals”. Most lynx in a population killed sheep at some stage. This implies that it is the size of the lynx population that will determine how many sheep are killed. The local effect of shooting a lynx will be very short lived in areas with a continuous lynx population, as the territory will be rapidly occupied by a new lynx. Our results also show that sheep grazed behind fences were almost never killed by our lynx, implying that fencing sheep will lower lynx–sheep encounters and thereby kill rates. Regional zonation of lynx management is challenging because lynx move over very large areas. Several of the present lynx zones do not fulfill the necessary criteria for successfully reducing depredation. This is either because the zones are too small, because of a lack of coordination between neighbouring management zones, or because of a failure to invest in effective protection measures in areas where lynx are given priority. We have estimated the number of sheep killed by lynx each year within each region by extrapolating from our data on kill rates and the current population estimates for the size of the lynx population. The analysis reveals a considerable gap between our estimates of what the lynx have killed and the compensation paid. In general, it appears that too much compensation is being paid. The discrepancy varies in both time (within a region) and between regions. One explanation could be that depredation caused by other carnivores is being wrongly assigned to lynx. It could also be that mortality from causes other than large carnivores (accidents, disease, smaller predators) is being wrongly attributed to them. We recommend that the compensation system could be improved by switching to one based on an objective assessment of depredation risk and kill rates specific to different landscapes. Sheep production has, and can continue to have, an important role in food production, cultural landscape conservation and rural livelihoods. We now have enough knowledge about what is necessary to reduce the conflict between lynx and sheep. Without a radical change in the form of husbandry it will be the regional population goals for lynx that determines the losses of sheep. A prerequisite for reducing the loss of sheep is to reduce the encounters between lynx and sheep in the zones where lynx are given priority. This can be achieved by either fencing sheep, moving sheep from forest to alpine grazing areas, or converting to other forms of livestock production. A change in husbandry practices will be difficult to achieve without changing to a form of compensation that includes suitable incentives to prevent losses and providing appropriate economic and technical assistance. Our studies show that a significant proportion of sheep mortality must be due to causes other than lynx, suggesting that it is important to initiate a study of all causes of sheep mortality.