SEARCH

SEARCH BY CITATION

FilenameFormatSizeDescription
gcb2733-sup-0001-TableS1.docxWord document16KTable S1. Description of the variation in elevation (metres above sea level) within and among the focal areas. All values are based on the are falling within 100 elevation bands (e.g. from 1000–1099 m asl, from 1100–1199 m asl etc.). Mean values are based on the area falling within each elevation band, not only the maximum and minimum values. >90% indicates the elevation range where more than 90% of the area falls within.
gcb2733-sup-0002-TableS2.docxWord document17KTable S2. Overview of the meteorological stations used for climate data in each of the focal areas. Station identification number from the Meteorological office in Norway is given behind the station name (No). The density of observation stations reporting precipitation and snow is higher than for stations reporting temperature. We therefore used different observation stations for temperature and precipitation/snow, in each case choosing the nearest station reporting data for the extension of the study period. Lat = latitude; lon = longitude; m asl = metre above sea level. Temperature measurements were not available in Møsstrand II in 1999 and this year was therefore excluded from the analyses.
gcb2733-sup-0003-TableS3.docxWord document17KTable S3. P-values resulting from simple linear regressions with the North Atlantic Oscillation Index (NAO) explaining the yearly variation in local climate variables in the four focal mountain areas in Southern Norway. The PC based NAO for December, January, February and Mars is used to explain snow (maximum snow depth the previous winter) while the PC based NAO for Mars, April and May is used to explain spring temperature, GDD and precipitation (see Table 1). GDD is Growing Degree Days (adjusted for elevation in Setesdal). Temperature is average temperature, in May. Significant effects included in the final analyses are in bold.
gcb2733-sup-0004-TableS4.docxWord document18KTable S4. AIC values for different linear mixed models explaining lamb autumn body mass in the four focal areas in Southern Norway by use of different NDVI values (at the grazing area or study area spatial scale) and temperature measures (top) and for different linear mixed models explaining spring NDVI values by use of different temperature measures. As temperature measures we used mean temperature and growing degree days (GDD) in May (GDD adjusted for elevation in Setesdal). We included all individual based covariates in the models explaining lamb autumn body mass (see Table 2). Grazing area was included as random factor in these models. The contrast between the two spatial scales of NDVI values was not tested in Setesdal as the lambs could not be assigned to a particular grazing area. The best variables (lowest AIC value) are indicated in bold.
gcb2733-sup-0005-TableS5.docxWord document19KTable S5. AIC values for different linear regression models explaining the yearly variation in spring NDVI (see Table 1) in the four focal areas in Southern Norway. As North Atlantic Oscillation Index in winter (NAOw) we used the PC based NAO for December, January, February and Mars while for spring NAO (NAOs) we used the PC based NAO for Mars, April and May. Snow is maximum snow depth the previous winter while average temperature (Temperature) and precipitation is measured during May.
gcb2733-sup-0006-TableS6.docxWord document24KTable S6. AIC values for different linear mixed models explaining lamb autumn body mass in the four focal areas in Southern Norway. NDVI values are from second half of May and for the entire study area except for Hardangervidda West where NDVI values for each grazing area proved to best explain the variation in lamb autumn body mass (Table S3). As winter NAO (NAOw) we used the PC based NAO for December to Mars while for spring NAO (NAOs) we used the PC based NAO for Mars to May. Snow is the maximum snow depth the previous winter. Spring temperature precipitation is measured in May. Since average temperature was a better variable for explaining variations in NDVI in all focal areas Growing Degree Days was not included in the models (Table S3). We included all individual based covariates in all models (see Table 2). Grazing area was included as random factor in all models in all focal areas but Setesdal where the lambs could not be assigned to a particular grazing area.
gcb2733-sup-0007-Figure1.docxWord document15KFigure S1. Seasonal development of NDVI in the four focal areas in Norway. Plots are based on 9 years of data (1999–2007) in Setesdal and 16 years of data (1992–2007) in the other areas, i.e. the entire study period for each focal area.
gcb2733-sup-0008-Figure2.docxWord document15KFigure S2. Plot of the sampling chains of NDVI and grazing pressure in Setesdal resulting from three chains of 5000 MCMC permutations where the first 2500 permutations were discharged as burn-in and from the remaining only every fifth value was used, to avoid autocorrelation, leaving 1500 samples in the plot.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.