Most of the plant diversity in European forests is found in the understorey, and despite its relatively low biomass compared to the forest overstorey, its functional importance is high (nectar source for pollinators, habitat for small mammals, affects tree regeneration, invasion resistance, decomposition and nutrient cycling (Gilliam 2007). Relationships between overstorey and understorey diversity and composition will be tackled. At the Satakunta sites in Finland, plots of 20m x 20m were established in 1999 in which the desired species mixture was achieved by mixing the species tree by tree (so no pattern of small monocultures next to each other) at planting. Vegetation was first monitored in 2003 by the team of Julia Koritcheva. In each plot 3 frames of 1m x 1m (= subplots)were randomly spread and the cover of each species within these frames was recorded. The exact location of these subplots was NOT determined nor marked. This experimental set-up was repeated in July 2011 by Evy Ampoorter and Kris Ceunen: 3 frames of 1m x 1m (vegetation subplots) were randomly located within each plot (however, big rocks were avoided) and following characteristics were recorded: percentage cover per species in the understorey layer (< 1.3m), total cover by understorey layer, other cover (litter, bare soil moss). A biomass sample was also taken: within the frame of 1m x 1m a smaller frame of 0.5m x 0.5m was placed at a location where vegetation was relatively abundant and composition was representative for the bigger frame (biomass subplot). All aboveground biomass within this smaller frame was clipped. Dead vegetation was removed from each biomass sample. For each first subplot per plot the biomass sample was also further subdivided into following groups: Epilobium angustifolium, Rubus idaeus, dwarf shurbs, grasses, seedlings, ferns, other herbs. After removal of death vegetation and eventual subdivision into groups, the biomass samples were dried for 24h at 70°C (samples were already dry when collecting). Then they were weighed and analysed for P, Ca, K, Mg (Forest & Nature Lab, Ghent University) and C/N (Institut für Walbau, Universität Freiburg). In each plot at 6 randomly spread locations a soil sample was also taken from the upper 10cm of the mineral layer (if possible; namely, very rocky soil and often very thick layer of accumulated organic material that could not be removed so that mineral soil layer could not be reached ; soil sampler could only sample to a depth of 10cm). These 6 samples per plot were thoroughly mixed and dried at 40°C for about a week (very wet!). Then they were analyzed for pH, K, Ca, Mg, C, N,S total P and Olsen P (Forest & Nature Lab, Ghent University).

Three frames of 1m x 1m were randomly located within each plot (however, big rocks were avoided) and following characteristics were recorded: percentage cover per species in the understorey layer (< 1.3m), total cover by understorey layer, other cover (litter, bare soil moss). A biomass sample was taken: within the frame of 1m x 1m a smaller frame of 0.5m x 0.5m was placed at a location where vegetation was relatively abundant and composition was representative for the bigger frame. All aboveground biomass within this smaller frame was clipped. Dead vegetation was removed from each biomass sample. For each first subplot per plot the biomass sample was also further subdivided into following groups: Epilobium angustifolium, Rubus idaeus, dwarf shurbs (such as Vaccinium sp.), grasses, seedlings, ferns, other herbs. After removal of death vegetation and eventual subdivision into groups, the biomass samples were dried for 24h at 70°C (samples were already dry when collecting). Then they were weighed and analysed for P, Ca, K, Mg (Laboratory of Forestry, Gontrode, Belgium) and C/N (Institut für Walbau, Universität Freiburg). Preparation of biomass samples for chemical analyses: dried at 70°C, grinded using a sieve of 1mm (for analyses of P, Mg, Ca en K content) and afterwards further grinded with a ball mill (MM301, Retsch, Hahn, Germany) in Freiburg (for analysis fo C/N) . Before analyses of Ca, Mg, K en P-content, grinded biomass samples are 'destroyed' with HNO3/HClO4, diluted with demineralised water and filtered according to the method of Varian Instruments at work, AA-24 (McKenzie T., 1982. Automated multielement analysis of plant material by flame atomic absorption spectroscopy. Varian Instruments at work, AA-24). Analysis for Ca-content: destruction of biomass sample with the above mentioned method. Then the sample is diluted, 10% CsCI is added as modifer on the AAS (Varian SpectrAA 220) and calibrated to 5mg Ca/l and atomized in a nitrous oxide-acetylene flame, absorption is measured at 422.7nm. The analytical quality of the analysis is verified by repeated measurement of blanco and control samples. Analysis for K-content: destruction of biomass sample with the above mentioned method. Then the sample is diluted, 10% CsCI is added as modifer on the AAS (Varian SpectrAA 220) and calibrated to 1mg K/l and atomized in a air-acetylene flame, absorption is measured at 766.5nm. The analytical quality of the analysis is verified by repeated measurement of blanco and control samples. Analysis for Mg-content: destruction of biomass sample with the above mentioned method. Then the sample is diluted, 10% CsCI is added as modifer on the AAS (Varian SpectrAA 220) and calibrated to 0.5mg Mg/l and atomized in a air-acetylene flame, absorption is measured at 258.2nm. The analytical quality of the analysis is verified by repeated measurement of blanco and control samples. Analysis for P-content: destruction of biomass sample with the above mentioned method, then reaction of the destroyed biomass sample with metholsulphite, ammoniumheptamolybdate and sodiumacetate solution. Finally, measurement of the total P-content with a spectrophotometer type Varian Cary 50 at 700nm (Chemical analysis of plants and soils. Cottenie, Verloo,... Laboratory of analytical and agrochemistry, RUG). Analyses for C/N: C/N analysis is done by dry combustion with an elemental analyzer leco truspec™ (Leco, St. Joseph USA). 100-200 mg finely ground sample is wrapped in tin foil and burned with pure oxygen at 950°C. Tin foil rises temperature additionally throughout the burning process. Carbon is detected as CO2 with an infrared measuring cell, nitrogen is reduced to N2 and detected with a thermal conductivity cell. For prediction of C and N content with near infrared spectroscopy spectra are taken from finely ground material with a Tensor 37 spectrometer (Bruker, Ettlingen, Germany). The wavenumber range is 12000 to 3800 cm-1, Resolution is 16 cm-1 with 64 scans for each spectrum. Five spectra are measured for each sample. For at least 100 samples the dry combustion reference analysis is done. With Opus 6.5 software (Bruker, Ettlingen, Germany) spectral information is combined with the corresponding reference value and models to predict unknown samples are developed using PLS (partial least square) regression.

Satakunta,Finland: 3 sites: areas 1 and 2 near Pomarkku, area 3 near Lankoski (vegetation was only monitored at sites 1 and 3)

Summer 2011

understorey vegetation: woody and non-woody vascular species (herbs, seedlings) < 1.3m

Overall very warm, sunny weather. In 2011, no mowing activities took place. In 2010, mowing of non-sown tree species took place.