The Finnish Forest Research Institute, PO Box 18, FIN-01301 Vantaa, Finland

Background and objectives
All trees probably hide various endophytic fungi in their phyllosphere. The endophytic fungi of trees have attracted increasing interest after the discovery that some of them affect insects attacking the phyllosphere. Tree density and their species composition probably affect the microclimatical conditions in the canopy and may thus be important factors directing the amount and diversity of phyllosphere endophytes.

The aim of this study was to characterize the overall infection rate and diversity of endophytic fungi of Norway spruce (Picea abies) needles in southern Finland and to find out how the tree species composition and density affects this fungal community. We also wished to find out whether the needle endophytic mycoflora of virgin spruce stands differs from that in managed stands.

Materials and methods
Norway spruce needles were sampled from two series of stand areas located in southern Finland. Both series consisted of five sampling areas in mature managed stands and one in a mature virgin stand. The proportion of spruce varied from 8 to 100% of the basal tree area and the major other species were pubescent birch (Betula pubescens) and Scots pine (Pinus sylvestris). From each sampling area (some of which consisted of several sites) 40 mature spruces were randomly chosen and healthy looking needles of the third age-class were sampled from heights of 5-8 ;m. Ten needles per tree were surface-sterilized and four slices of each needle were incubated on water agar for isolation of endophytic fungi. FAST-profiles (i.e. combined fatty acid and sterol profiles) were used for the classification of the fungal isolates into species and FAST-groups, i.e. operational chemotaxonomic units. Using FAST-profiles fungal isolates, including isolates of species not yet described, could be divided into groups (operational chemotaxonomic units) with an upper variation limit [1].

Results and conclusions
The majority of the isolates (1593 out of 1740) were identified by their FAST-profiles as Lophodennium piceae. Tiarasporella parca was a less common species and Thysanophora penicillioides, Rhizosphaera kalkhoffii, Sclerophoma pythiophila and Lirula macrosporawere found only occasionally. For calculation of fungal diversity, all isolates were classified by their FAST-profiles into 81 operational chemotaxonomic units.

The highest percentage of endophytically infected needles was found in pure spruce stands and dense virgin stands. Location of the stand, its proportion of spruce and total basal area of trees (i.e. tree density) explained 82% of the variation of the overall infection rate. The effect of location was probably a consequence of differences in air quality between the various sampling areas. No correlation could be found between recent growth rate or age of the spruces and the diversity or infection rate of the endophytic fungi of the needles.

Endophyte diversity, expressed as the number of FAST-groups per 40 spruces of each sampling area, correlated positively and statistically significantly with the percentage of needles infected or with the proportion of spruce in the stand. The highest endophytic diversity, expressed as FAST-groups per tree, was found in pure spruce and mixed virgin stands. Results are presented in detail by Muller and Hallaksela [2].

1. Muller MM, Hallaksela A-M, 1998a. Mycological Research 102, in press.
2. Muller MM, Hallaksela A-M, 1998b. Mycological Research 102, in press.