Miljøovervåkingsprogram for Ormen Lange landanlegg. Overvåking av hjorteviltbestanden 2010 - 2012 - sluttrapport

Abstract

Ormen Lange landanlegg på øya Gossa ble satt i drift i 2007 med A/S Norske Shell som oper-atør. Anlegget er plassert i et område som ifølge Norsk Hydro sin egen konsekvensutredning (2002) er i konflikt med hjortetrekket til og fra fastlandet. Spesielt gjelder dette et viktig trekk som er konsentrert rundt Steinsvika i den nordøstlige delen av landanlegget. Konsekvensutredningen konkluderte forøvrig med at anlegget er plassert i et område som hadde middels betydning for hjorteviltbestandene (Norsk Hydro konsekvensutredning 2012). For å kunne dokumentere even-tuelle effekter av landanlegget på hjorteviltbestandene ble det i 2008 gjennomført en grunnlags-undersøkelse som skulle danne utgangspunktet for et overvåkingsprogram av hjorteviltbestan-dene i årene 2010–2012. Denne rapporten presenterer resultatene fra overvåkingsprogrammet. I overvåkingsprogrammet var det et hovedmål å få bedre estimater på utvikling i bestandstetthet og hvis mulig estimater på bestandsstørrelse. Det var også viktig å evaluere hvorvidt og i så fall hvordan hjorteviltet ble påvirket av landanlegget og virksomheten der. For å oppnå dette var det nødvendig å forbedre metodikken til mer komplekse tilnærminger. I denne studien er det derfor anvendt direkte og indirekte metoder for bestandsestimeringer som kan gi noen svar på disse spørsmålene. På Gossa er det livskraftige bestander av både hjort og rådyr. Transekttellinger viser at kjønns- og aldersstrukturen til begge bestandene er forholdsvis lik de man finner i tilsvarende bestander som høstes gjennom jakt. Estimeringen av bestandstettheter viser at det som forventet var langt mer rådyr enn hjort på øya. Det var rimelig god overensstemmelse mellom viltnemndas og mine tellinger, noe som indikerer at man ved forholdsvis enkle tellinger av dyr langs faste ruter kan registrere trender i hjorteviltbestander. Dersom det er behov for å estimere antall individer i be-stander, kreves det mer tidkrevende metoder som distance sampling og møkkdungetellinger slik som i denne studien. Det var stor forskjell mellom tetthetsestimatene basert på distancetilnær-mingen og møkkdungetellingene. Som gjennomsnittsverdi for studieperioden foreslo distancetil-nærmingen at det var 5,4 rådyr/km2 mens møkkdungetellingene tilsa 12,5 rådyr/km2. Tilsvarende tettheter for hjort var hhv. 0,7 hjort/km2 og 3,4 hjort/km2. Estimatforskjellene mellom metodene kan skyldes flere faktorer, men tett kantvegetasjon langs transektene reduserte siktbarheten svært mye slik at transektlengden ble veldig stor i forhold til antall observasjoner. Estimatene fra distancetilnærmingen ble derfor sannsynligvis kunstig lave. Imidlertid synes det rimelig sikkert at rådyrbestanden har minket etter 2010. Analysen antyder at også hjortebestanden har minket i løpet av studietiden, men dette må tolkes med forsiktighet på grunn av lite datamateriale. For-søksvis anvendte jeg mitt begrensede datamateriale fra møkkdungetellingene til å estimere to-talbestander av hjort og rådyr på Gossa. Estimeringen foreslår at rådyrbestanden minket fra rundt 700 til 400 dyr i løpet av studietiden. Tilsvarende tall for hjort var en reduksjon fra rundt 200 til 80 individer. Disse tallene er meget usikre, men de rimer forholdsvis bra overens med fellingsstatistikken. For å vurdere eventuelle effekter av virksomheten ved landanlegget på hjorteviltet ble Gossa delt inn i to soner, en østlig og en vestlig del. Bestandstetthetene av hjortevilt i den østlige delen ved landanlegget var ikke forskjellig fra den vestlige delen. Med andre ord, dyretettheten ved landanlegget var ikke forskjellig fra andre steder på Gossa. Gossa, Aukra, Møre og Romsdal fylke, oljeindustri, prosessanlegg, Nyhamna, Ormen Lange, hjorteviltovervåking, transekt, habitat-seleksjon, distance sampling, Monte Carlo simulering, møkkdunge-tellinger, bestandstetthet, rådyr, hjort

The Ormen Lange Onshore Processing Plant on the island Gossa started up in 2007 with A/S Norske Shell as the operator. The plant is situated in an area that partly will be in conflict with the annual red deer migration between the island and the mainland. This in particular applies to a migrating route concentrated around Steinsvika in the north-eastern part of the plant. A consequence analysis suggested that the plant is situated in an area of average importance for deer. A baseline monitoring program aiming to evaluate possible effects of the plant on deer populations on the island was completed in 2008. The present main monitoring study (2010 – 2012) is based on the experiences gathered in the baseline study. One important aim for the present study was to achieve better estimates for population densities and size if possible. Another goal was to evaluate if the plant and related installations somehow affected the deer populations and if so to detect the underlying mechanisms causing this. To deal with these questions, new approaches were necessary. Thus, indirect and direct methods for determining population densities and sizes were applied. The island of Gossa harbours healthy populations of roe and red deer. Counting of animals reveals that that the age- and sex structure resembles those of other populations that are harvested. As expected, the roe deer population was much larger than that of red deer. It turned out that the counting of animals done by the local authorities resembled mine results, suggesting that reasonable estimations of population trends can be achieved by relatively inexpensive and simple counts of animals along fixed routes. However, if estimates of population sizes are needed, more time consuming and expensive approaches like pellet group and distance sampling is needed. In the present study, there was a discrepancy between the density estimates obtained by distance and pellet group sampling. Distance sampling estimated 5.4 roe deer per km2 on average whereas pellet group counting suggested 12.5 roe deer per km2 . Equivalent estimates for red deer were 0.7 deer per km2 and 3.4 deer per km2 . This discrepancy may rely on several factors, but frequent stretches of dense vegetation along the roads used as transects reduced sight ability, thus rendering few observations compared to transect length. Estimates from distance sampling are therefore artificially low. However, it seems reasonably clear that the roe deer population has diminished after 2010. Red deer apparently has decreased too, but sample size is too small to conclude. By way of experiment I used my limited data material too estimate population sizes for roe and red deer. The estimates suggested that roe deer had decreased from approximately 700 to 400 individuals during the study, whereas corresponding numbers for red deer was a decrease from approximately 200 to 80 individuals. These numbers are certainly doubtful, even though they are reasonable if compared to the harvest statistics. During the study there was no difference between the densities of deer around the plant compared to other places on the island. However, deer were not evenly distributed in the vicinity of the plant. There were more animals near the plant than farther away from the plant, thus rejecting a random distribution of deer around the plant. It is therefore no reason to believe that the deer population on the island has any behavioral aversion towards the plant as such. Deer occurred close to and even on the inside of the fence encircling the plant, as long as the fence was open towards the seaside (it was permanently closed in 2012). This suggests that deer rapidly habituates to activities related to the plant. This corresponds to findings from other plant sites, deer habituates to buildings and scheduled activities. Hunting statistic is the only source for information about deer population densities before and under construction of the plant. Here I assume that hunting statistic reflects trends in deer pop-ulations. The statistics suggest that deer populations on Gossa increased rapidly during the construction of the plant up until 2010, thereafter populations decreased. I used roe deer as a case to scrutinize this trend more carefully. The roe deer population at Gossa fluctuated synchronously with the mainland population in Møre and Romsdal. However, roe deer at Gossa increased its population size 6 times faster than roe deer at the mainland. This difference can be explained by climatic factors, migration and nutritional variability, but one cannot overlook the possibility that the island population might have been affected by some local factor, mainly avoided by the mainland population. The plant and attached facilities occupies about 4-5 % of the total area of Gossa. During the construction of the plant, some deer had to leave the area and move to other unaffected places on the island where the roe deer population already was increasing. Thus, further compressing the population in these areas. Assuming a constant culling pressure, this could increase the harvest rate up to and including 2010. Thereafter, hunters could not manage to maintain the harvest rate due to a decreasing roe deer population. The population decline was therefore first of all a result of the harvesting regime. The increased harvest was intended because the population was generally considered to be too dense. To conclude, deer populations at Gossa did not exhibit any behavioral aversion towards the plant facilities. On the contrary, deer concentrations close to the plant was larger than further from the plant. Until the fence encircling the plant was permanently closed, deer moved in and out of the enclosure. This suggests that deer did not avoid the plant area. Deer populations at Gossa are mainly regulated by culling, habitat quality and migration. However, loss of habitat might affect the deer populations at Gossa. Fragmentation and loss of habitat is considered being a major problem regarding animal populations and diversity and as such no exception for Gossa. Given that habitat availability has decreased at Gossa, deer populations probably need to be adjusted somewhat to face a slightly changed carrying capacity. Therefore it is mandatory to carry out qualitatively well designed surveys to get good estimates of population sizes to be used as a guidance tool to achieve a sound management.