Summary

 

 

A network of oak ( Quercus robur L.) chronologies established on 48 sites and containing 635 single-tree series was analysed to identify weather variables related to the annual dynamics of ring increment for the area of the southern Sweden (from 55 and 60 N and 12 to 19 E) for the period 1860-2002.

 

Growth of oak is positively correlated with summer precipitation in the current and in the previous growth seasons, and with the temperature in October of the previous season. On the sites with clay-rich soils correlation between growth and precipitation may become negative. Oak on such sites can therefore be at risk during the periods with excessive rains.

 

Periods with pronounced positive growth anomalies were around 1890-1900, 1920-1925, 1958-1964, 1978-1985. Negative growth anomalies occurred in 1885-1889, 1902-1907, 1914-1917, 1955-1957, 1965-1967, 1973-1977, and in 1992-1998.

 

Pointer years were defined as years with increment below lower 5% of ring-width distribution for a particular tree. Year 1965 was negative and the strongest pointer year in southern Swedish oaks. This year was recognized as a pointer year  in 32% of all trees sampled. Most of the negative pointer years were associated with exceptionally cold winters or springs (1868, 1940, 1956, 1965), and periods with spring-summer droughts (1869, 1947, 1992). Two positive pointer years (1882 and 1860) were not associated with neither temperature nor precipitation extremes. These years, however, had precipitation values reaching 83-94% of its distribution for at least two summer months.

 

Regional re-inventories of single oaks trees showed significant decline in the crown conditions from 1988 to 1999. Over two 5-year periods (1993-1997 vs. 1998-2002) the cumulative increment increased significantly for trees with healthy crowns, did not change in trees with moderately declining crowns, and significantly decreased in trees with heavily declining crowns. For trees with healthy crowns, this dynamics may represent growth recovery after 1992 drought. Instead, oaks with defoliation above 60% appear to reach a threshold in their ability to recover growth. At sites on nutrient-poor soils cumulative increments over 1998-2002 differed significantly among trees with different crown condition and no differences were observed at sites on nutrient-rich soils. Analyses and interpretation of the oak growth trends as recovered from tree-ring chronologies may be improved by controlling for the crown status of the trees sampled.

 

Condition of oak crowns at the moment of sampling did reflect trees’ growth at least over the preceding decade. Tree-ring increments were highest in trees with healthy crowns, intermediate in trees with moderately declined crowns, and the lowest in trees with heavily declining crowns.

 

Non-windfall mortality of oak can be a decade-long process. An analysis of growth patterns of dendrochornologically dated dead oaks shows that growth reduction in such trees begins already 14 years prior to the death date, staying at the lower 2.5 % limit of ring width distribution for about five following years. Death of an oak tree occurrs typically after 8-9 years of strong growth depression.