Height growth patterns of Scots pine and Norway spruce in the coastal areas of western Finland

Stem analysis data of 46 Scots pine (Pinus sylvestris L.) and 38 Norway spr uce (Picea abies (L.) Karst.) trees was used to construct height curves for naturally regenerated and cultivated Scots pine and for naturally regenera ted Norway spruce from the coastal areas of western Finland. The curves wer e compared with each other, with models for 28 naturally regenerated pine t rees off the coast and with general site index curves for pine and spruce i n southern Finland. The height growth pattern for coastal pine exhibited a strong stagnation over age compared to general site indexes. Spruce height development was dominated by a slow early development followed by an increa se in height increment compared to general index curves. There was a large variation in the height growth patterns of coastal spruce and a growth patt ern variable was needed in the functions in order to make reliable predicti ons. Even simple models provided relatively good predictions for coastal pi ne. It was argued that strong wind, temperature and moisture regimes caused the se differences between trees at the western coast and in the inner parts of Finland. The soil provides poor rooting conditions at the coast and this i ncreases trees' susceptibility to the climatic impact. The variation in hei ght growth patterns within the coastal area could partly be explained with location in relation to the sea and stoniness for pine, and humus layer thi ckness for spruce. The number of years trees had grown to breast height was used as a predictor, which explained the growth patterns very well. This v ariable did not describe causal relationships, since it depends both on the original status of site and of management intensity. It could be used to i mprove predictions when such data can is available. The early development s peed was correlated with the CIN ratio of the humus in spruce stands. This indicated that the growth patterns of coastal spruce may change along an ec ological gradient from dry, stony moraines with thick humus, to moist, dens e sorted soils with poor quality humus. The use of soil variables in predic ting height development is difficult, since they usually are time dependent and affected by stand characteristics and management. (C) 2000 Elsevier Sc ience B.V. All rights reserved.