Georgraphical variation in Sitka spruce productivity and its dependence on environmental factors
MetadataShow full item record
The objective of this study was to investigate altitudinal and geographical variation in the productivity of Sitka spruce in upland Scotland, to relate this to environmental variables and to use the information to develop a basis for predicting Sitka spruce yield form site factors. A total of 188 0.04 ha temporary sample plots were established in 15 to 50 year old Sitka spruce stands at 37 sites in Scotland and northern England, mostly spanning the upper 200m elevation range of plantations. At each plot estimates of General Yield Class (GYC) were made and the following site factors were assessed; elevation, geomorphic shelter (topex), aspect, slope, soil type and rooting depth. In addition, estimates of wind-climate, mean summer temperature (June-September), mean annual accumulated temperature > 5.6°C and annual rainfall were made by extrapolation of forestry Commission "tatter flag" records and Meteorological Office data. GYC declined by about 3.2 - 4.0 m3 ha -1 yr -1 per 100 m increase in elevation due to the effects of increasingly adverse climatic and edaphic conditions, GYC was fairly closely correlated with elevation at the individual sites, but there was considerable site to site variation. GYC values at specific elevations were higher in inland and southern areas than in coastal and northern ones. The geographical pattern of the relationship between GYC and elevation was strikingly similar to the distributions of growing season temperatures and wind-climate. Correlation and multiple regression analysis demonstrated that GYC was well correlated with extrapolated values of accumulated temperature and tatter rate, these two variables accounting for up to 78 per cent of the variation in GYC in the best multiple regression models. GYC also proved to be correlated with aspect, topex, soil type and crop age. Productivity was highest on north and east-facing aspects and increased with greater levels of geomorphic shelter. Differences in soil type only accounted for a small amount of variation in GYC (2-3 per cent). GYC was significantly negatively correlated with crop age, probably as a result of improved standards of silvicultural treatment. The best multiple regression models accounted for 78-86 per cent of the variation in GYC and were associated with confidence limits of ±2.2 - 2.6 m3 ha -1 yr -1. The mean figure for predicting GYC for a single site (acquisition) was calculated to be ± m3 ha -1 yr -1, and this figure was confirmed by results of a validation survey. The information presented could easily be adapted for predicting productivity and assessing suitable upper planting limits in practical forestry.