Soil seed banks of pure spruce (Picea abies) and adjacent mixed species stands

Plant and Soil 264: 53–67, 2004.© 2004 Kluwer Academic Publishers. Printed in the Netherlands.

53

Soil seed banks of pure spruce (Picea abies) and adjacent mixed speciesstands

Torsten W. Berger1,3, Bing Sun2 & Gerhard Glatzel11Institute of Forest Ecology, Universität für Bodenkultur, Peter Jordan-Strasse 82, 1190 Vienna, Austria. 2Instituteof Botany, Universität für Bodenkultur, 1190 Vienna, Austria. 3Corresponding author∗

Received 26 March 2003. Accepted in revised form 22 December 2003

Key words: Carex pallescens, Fagus sylvatica, forest conversion, Juncus effusus, Picea abies, soil seed banks

Abstract

The soil seedbank of long living seeds of herbs, graminoids and shrubs can survive several decades in the soiland germinate after disturbances like windthrow or clearcutting. The main goal of this study was to evaluate therisk of weeds, which may limit the success of conversions of secondary pure spruce stands (Picea abies) to mixedspecies stands. In a first step, germination experiments were performed in the greenhouse on soil samples collectedunder adjacent pure spruce and mixed species stands (mainly mixtures of spruce and beech – Fagus sylvatica) ontwo different soil substrates (Flysch: nutrient rich, basic soil; Molasse: nutrient poor, acidic soil). Seedling densityand species richness were higher on the nutrient rich soil on Flysch. Comparisons between seedlings that emergedfrom soil samples collected at the end of the vegetation period and in spring justify the statement of the hypothesisthat mixed spruce-beech stands advance the transient seed bank while pure spruce stands stimulate the persistentseed bank. In a second step, the seed banks of different soil horizons down to 35 cm soil depth were studied ina multivariate statistical design for the most dominant species J. effusus, C. pallescens and R. idaeus, which areknown to form long-term persistent seeds. Effects of bedrock material (Flysch, Molasse), species composition(pure spruce, mixed species) and treatment (control, nitrate) were tested. The total sum of these three species wassignificantly higher on Flysch than on Molasse. However, species composition indicated no significant differences,although there was a trend of higher amounts of germinating seeds under pure spruce. Nitrate treatments didnot promote germination of viable buried seeds, indicating that the number of emerged seedlings is a realisticindicator of the seed bank density for the studied stands. It is concluded that overstorey tree species compositionis not an important controlling factor for seed germination of the studied species after disturbances. The majorityof emergents are the graminoids J. effusus and C. pallescens which were not present at all in the abovegroundvegetation. Viable seeds were found down to 35 cm soil depth, although most seeds were concentrated in the upper10 cm soil. Hence, care should be taken if management strategies create conditions that are generally favorable togermination. The success of forest regeneration or a conversion of pure spruce to mixed species stands could beendangered by any disturbance, which causes an immediate increase of light levels.

Introduction

Monospecific stands of Norway spruce (Picea abies)below elevations of 700 m within Central Europemust be considered manmade on zonal soils (Mayer,1974). These secondary pure spruce (Picea abies)stands are prevalent in many parts of Austria (FBVA,

∗FAX No: +43-1-4797896. E-mail: [email protected]

1998), because high productivity coupled with goodtimber prices are tempting for the short term eco-nomic success. On the long term, however, forestsite degradation (Schmidt-Vogt, 1986) and low sta-bility of such stands which results among others inenhanced risks from wind throw (Rottmann, 1989) andpests (Schwerdtfeger, 1981) may cause lower profita-bility of secondary spruce stands than of natural mixedforests. That is why conversion of secondary spruce

54

stands to mixed species stands is a current issue inEurope.

Soil seed banks have been defined as those seedsthat can remain dormant for a period of time in the sur-face soil until their germination is triggered by an en-vironmental change (Archibold, 1989; Simpson et al.,1989). The amount of seeds in soil progressively de-clines with the development of closed-canopied forestsbut increases can occur when stands enter the oldgrowth stage when the forest canopy begins to breakup (Buckley et al., 1997; Augusto et al., 2001). Allstudies have demonstrated that the soil seed banks ofthese temperate forests include mostly early succes-sional ‘intolerant’ species (e.g., Betula spp., Carexspp., Juncus spp., Rubus spp.) that differ from theexisting composition (Archibold, 1989; Pickett andMcDonnel, 1989; Ashton et al., 1998; Kostel-Hugheset al., 1998; Onaindia and Amezaga, 2000).

According to Thompson et al. (1997) studies ofthe impact of different forest management practices(e.g., coniferous versus mixed coniferous-deciduousstands) on seed viability and longevity in differenttypes of soil are rare and worthy topics of futureresearch. Several distinct effects of the overstoreyspecies composition (pure spruce versus mixed spe-cies), which may influence species composition anddensity of the soil seed bank, are reported for thestudied stands: physical top soil parameters (Bergerand Hager, 2000), soil temperature and moisture re-gime (Hager and Weinmeister, 2001), soil chemistry(Berger et al., 2002), soil fauna (Berger, 2001) androot distribution (Schmidt and Kazda, 2001). Addi-tional discriminating factors may include seed rain,secondary dispersal, predation, pathogens, burial, andmicrosite conditions (Simpson et al., 1989). Earth-worms prefer to forage at the surface of high pH soils(e.g., mixed species stands; Neirynck et al., 2000) andmay form deep permanent burrows (Edwards and Bo-hlen, 1996), resulting in greater vertical mixing of thesoil and the seeds/spores therein.

The size of the buried seed pool reflects the type,intensity, and frequency of disturbances (Archibold,1989) which are different between uniform monospe-cific spruce and mixed coniferous-broad-leafed stands(e.g., Rottmann, 1989). Long living seeds of herbs,graminoids and shrubs can survive in the soil for sev-eral decades and germinate after disturbances like,e.g., windthrow or clearcutting, effecting the successof forest restoration. Additionally, soil disturbances inrecent forests may lead to the germination of early suc-cessional, highly competitive or ruderal species, which

may hamper the establishment of more stress tolerantforest plant species (Honnay et al., 2002). E.g., the nat-ural regeneration of forest trees is negatively affectedby dense ground cover of Rubus spp. in mountainousforests of southwestern Germany, causing notoriousdifficulties for forest management practices (Schreineret al., 2000). Most of the species represented in the soilseed bank under forests in southwestern Germany aretypical for the edge of forests, forest roads and clear-cutting, dominated by the genera Carex and Juncus(Ludemann, 1994). Both the number of individualsand species germinating in the soil of lowland planta-tions and woods in southern England were stimulatedby cutting treatments (Buckley et al., 1997). Hon-nay et al. (2002) conclude that in order to decreasethe cover of these often light demanding, competi-tive species, disturbances should be avoided. Undercontinuous canopy closure, the comparative advan-tage of forest plant species in shaded environments isincreased. Van der Valk and Pederson (1989) state thatvegetation management, based on the exploitation ofseed banks, may be considered a useful silviculturaltool.

Nitrate is the major naturally occurring inorganicsoil component that stimulates seed germination (Hil-horst and Karssen, 2000; and references therein).However, the ecological significance of nitrate for ger-mination in the field cannot be treated in isolationfrom other biotic and abiotic factors, such as water,temperature, light, responsiveness of seeds and otherchemical soil constituents. In fact, nitrate is mosteffective in stimulating germination when combinedwith other stimuli such as light, alternating tempera-tures, and ethylene (Cavers and Benoit, 1989). It hasbeen shown that soil disturbances may also result inconsiderable release of nitrate ions from the soil (e.g.,Johnson et al., 1995). This phenomenon is probablya significant factor in the promotion of weeds (Hil-horst and Karssen, 2000). Destruction of vegetationby clear-cutting rapidly accelerates the process of ni-trification, increasing nitrate concentrations in the soilsolution (e.g., Bormann and Likens, 1994). Hence,this fact may affect soil emergent densities, if forestsare clear-cut or converted.

Several studies have explored the effects weedshave on competition for light, nutrients and water, sug-gesting that weed control may accelerate young foreststand development (e.g., Millbacher, 1987, 1992;Romagosa and Robinson, 2003). Hence, the objec-tive of this study was to evaluate the risk of weeds,which may limit the success of conversions of sec-

55

Table 1. Soil properties under the pure spruce and the mixed species stand at the experimental sites Kreisbach(bedrock: Flysch) and Frauschereck (bedrock: Molasse): stores of organic carbon (Corg in kg m−2 per

horizon) and nitrogen (Ntot in g m−2 per horizon), effective cation exchange capacity (CEC, µmolc g−1),base saturation (%) and pH (H2O)

Soil horizon Corg Ntot CEC Base sat. pH

pure mixed pure mixed pure mixed pure mixed pure mixed

Kreisbach

Forest floor 0.7 0.3 17 7

0-5 cm 1.2 1.0 82 82 113 142 48 97 4.3 6.2

5–10 cm 1.0 0.9 81 85 106 108 49 88 4.4 5.5

10–20 cm 1.8 1.6 172 158 98 102 58 79 4.8 5.5

20–30 cm 1.4 1.3 138 129 96 98 58 92 4.9 5.8

30–40 cm 1.1 0.8 116 94 96 107 61 94 4.9 5.8

40–50 cm 0.9 1.0 98 127 95 103 64 86 5.0 5.6

sum 8.1 6.9 714 682

Frauschereck

Forest floor 2.1 1.9 79 70

0–5 cm 2.4 4.1 94 165 148 153 17 26 3.8 3.6

5–10 cm 0.8 1.9 66 85 122 142 10 8 3.9 3.6

10–20 cm 2.6 2.5 73 117 81 120 8 6 4.3 4.0

20–30 cm 1.5 1.9 76 97 33 45 13 7 4.4 4.5

30–40 cm 0.9 1.0 54 49 21 32 19 12 4.6 4.5

40–50 cm 0.6 0.8 45 59 20 26 22 16 4.6 4.5

Sum 10.9 14.1 483 642

ondary pure spruce stands to mixed species stands.We hypothesize that species composition, density anddepth distribution of the soil seed bank is affected by(1) overstorey tree species composition (pure spruceversus mixed species stand), (2) soil type (nutrient richversus nutrient poor soil) and (3) nitrate treatmentswhich promoted germination of viable buried seeds inother studies (e.g., Auchmoody, 1979).

Materials and methods

Study sites

Two pairs of secondary pure spruce stands (Piceaabies) and adjacent mixed species stands on compar-able sites were selected for this study on two differ-ent bedrock materials for soil formation: Flysch andMolasse. Former species compositions of all studysites were mixtures of spruce and beech.

Study sites on Flysch (Kreisbach)The Flysch zone is a narrow strip in the foothillsof the Northern Limestone Alps from west to east

throughout the country. The study sites are south ofSt. Pölten (15◦39′ E, 48◦05′ N), Lower Austria, atan elevation of 480 m. Flysch consists mainly of oldtertiary and mesozoic sandstones and clayey marls.Nutrient release from this bedrock material is higher incomparison to Molasse (see Table 1; methods of soilcollection and chemical analysis are given in Bergeret al., 2002) and consequently the prevalent humusforms are mull (mixed stand) to intermediate typesbetween mull and moder (pure spruce), indicatingquick turnover of the forest floor (0.8 cm thicknessunder mixed species to 1.4 cm under pure spruce). pHvalues of the upper mineral soil (0–5 cm soil depth)indicate top soil acidification under pure spruce dueto sequestration of base cations in the forest floor(4.3 versus 6.2). The soils of these study sites wereclassified as pseudogley (Scheffer and Schachtschabel,1998; FAO classification: stagnic Gleysol), since hori-zons with a high fraction of fine material (loam, clay)cause temporary waterlogging.

Stand characteristics are given in Table 2. Adja-cent (eastward) to the pure spruce stand there is a100 m wide transition zone which is characterizedby increasing admixture of beech until the overstorey

56

Table 2. Forest stand characteristics of a pure spruce and an adjacent mixed species standat the experimental sites Kreisbach (bedrock: Flysch) and Frauschereck (bedrock: Molasse)

Parameter Kreisbach Frauschereck

Pure Mixeda Pure Mixed

spruce beech spruce spruce beech spruce

Age (years) 53 65 – 58 89 89

Stems (N ha−1) 1012 960 – 1264 266 148

Growth volume (m3 ha−1) 567 588 – 432 244 140

Basal area (m2 ha−1) 57 47 – 51 31 11

Mean height (m) 25 26 – 20 29 26

aAdjacent (eastward) to the pure spruce stand at Kresibach there is a 100 m wide transitionzone which is characterized by increasing admixture of beech until the overstorey consistsof pure beech. Soil samples were collected from the pure beech stand, approximately 150 meast of the pure spruce stand. Since this pure beech stand rather represents a mixture ofspruce and beech (see text for details), it is still called a mixed species stand throughout thispaper.

consists of pure beech. Soil samples were collectedfrom the pure beech stand, approximately 150 m eastof the pure spruce stand to avoid direct impacts ofthe monoculture. Since this pure beech stand ratherrepresents a mixture of spruce and beech in regard toseed rain and dispersal, it is still called a mixed speciesstand throughout this paper. Soil chemical data of thetransition zone (not shown in Table 1) are betweenpure spruce and pure beech, but very similar to thepure beech stand, the so-called mixed stand of thisstudy (see Neubauer, 2002). This is in accordancewith Rothe et al. (2002), who found non-additiveeffects of tree species (beech vs. spruce) comparedto the monocultures. The natural forest vegetation ofmixed species stands on Flysch is Asperulo odoratae-Fagetum (Mucina et al., 1993; see Table 3). The totalcover of the herb layer is 5% (pure) to 20% (mixed).

Study sites on Molasse (Frauschereck)These study sites are located near Mattighofen(13◦08′ E, 48◦07′ N), Upper Austria, in a forestedlandscape, called Kobernausserwald at an elevation of700 m. Parent material for soil formation are tertiarysediments (so-called ‘Hausruck-Kobernausserwald’gravel), which consist mainly of quartz and othersiliceous material. Because of this acidic bedrock ma-terial with low rates of nutrient release (see Table 1),the soil types for both stands were classified assemi-podzol (Scheffer and Schachtschabel, 1998; in-termediate soil type between cambisol and podzol;FAO classification: dystric cambisols). Humus formis moder and the thickness of the forest floor variesbetween 6 and 8 cm. pH values (H2O) of the uppermineral soil (0–5 cm soil depth) are between 3.6 and

3.8 (see Table 1). In general, soils on Molasse se-quester more organic carbon and are more acidic, lesssupplied with nutrients (compare nitrogen stores andbase saturation) and more sandy than soils developedon Flysch (see Table 1).

Stand characteristics are given in Table 2. The nat-ural forest vegetation is Luzulo nemorosae-Fagetum(Mucina et al., 1993; see Table 3). The total cover ofthe herb layer is 10% (pure) to 15% (mixed).

Soil sampling

Soil cores were taken with a core sampler of 70 mmdiameter in June (only at Kreisbach) and September(only at Frauschereck) 1998, in March 1999 and inFebruary 2000 during a warm weather period. Therewere five randomly distributed replications at eachsite. In the years 1998 and 1999 the soil was dividedinto the forest floor and geometric horizons of 1-cmwidth down to 4 and 6 cm soil depth, respectively.In 2000, the soil was separated into the followinghorizons: forest floor, 0–5, 5–10, 10–15, 20–25 and30–35 cm soil depth.

Greenhouse methods

The soil seed bank was estimated using the seedlingemergence method, in which the samples are subjectedto conditions that are generally favorable to germina-tion (Simpson et al., 1989). Soil samples were layeredin plastic flats no more than 5 mm deep on top of 5 mmsterilized sand (90% SiO2, 6% Al2O3 and ca. 4%K2O + Na2O). Shallow layering of soil samples hasbeen shown to enhance overall germination (Kostel-

57

Table 3. Species in the aboveground soil vegetation (according to a 1997 survey within an area of 50 × 50 m; nomenclature ofspecies according to Encke et al., 1994) under a pure spruce and a mixed species stand at the experimental sites Kreisbach andFrauschereck. Bold letters indicate species that were also identified in the seed bank (seedling emergence method). Combinedvalues for abundance and dominance are given according to Braun-Blanquet (1964): r, solitary, small cover; +: few, smallcover; 1: numerous, < 1/20 cover; 2: 1/20 to 1/4 cover

Kreisbach Frauschereck

Pure Mixed Pure Mixed

Abies alba, + Acer pseudoplatanus, r Abies alba, + Abies alba, 1

Acer pseudoplatanus, r Carex pendula, + Bazzania trilobata, + Carex brizoides, +Atrichum undulatum, + Carex sylvatica, + Deschampsia flexuosa, 2 Carex pilulifera, +Carex sylvatica, 1 Dryopteris filix-mas, r Dicranodontium denudatum, 1 Deschampsia flexuosa, +Carpinus betulus, + Epipactis helleborine, r Dicranum scoparium, 1 Dicranodontium denudatum, +Dryopteris dilatata, + Fagus sylvatica, + Dryopteris dilatata, 1 Dicranum scoparium, +Fagus sylvatica, + Fissidens taxifolius, + Fagus sylvatica, + Dryopteris dilatata, 2

Fraxinus excelsior, + Fraxinus excelsior, + Leucobryum glaucum, + Fagus sylvatica, 2

Galeopsis speciosa, + Galium odoratum, 2 Polytrichum commune, 2 Picea abies, 1

Galium rotundifolium, 1 Geranium robertianum, + Vaccinium myrtillus, 1 Polytrichum commune, 1

Galium odoratum, + Impatiens noli-tangere, + Rubus fruticosus, +Lamiastrum montanum, + Lamiastrum montanum, 2 Vaccinium myrtillus, +Mycelis muralis, + Lysimachia nemorum, +Oxalis acetosella, + Mercurialis perennis, 1

Picea abies, + Oxalis acetosella, 2

Prunella vulgaris, + Paris quadrifolia, +Prunus avium, r Sambucus nigra, +Rubus idaeus, + Scrophularia nodosa, +Salvia glutinosa, + Veronica montana, +Sambucus nigra, 1 Viola reichenbachiana, 1

Scrophularia nodosa, +Stachys sylvatica, 1

Viola reichenbachiana, 1

Hughes et al., 1998). Since the subsoil is needed toprovide the nutrients necessary for the emerged seed-lings to grow, the bottom of the flats below the sandwas first filled with fertilized (N, P2O5 and K2Oi ineach case 50–125 mg L−1) peat which was sterilizedin the oven for 2 h at 180 ◦C. In the years 1998 and1999 we used 10 × 20 cm flats which were largeenough to take up the whole sample per 1-cm horizon.In 2000, two subsamples of each 5-cm horizon wereplaced on flats of 90 mm diameter and the ratio (massof subsample/mass of whole geometric horizon) wasused for upscaling.

Flats were watered automatically several timesduring the day. In addition, the flats were wateredby hand once a day to take care of different amountsof water needed to reach field capacity of sandy(Frauschereck) and clayey (Kreisbach) soils. Naturallightening was supplemented daily with artificial light-ening from 6:00 a.m. till 9:00 p.m. Seedling emer-

gence was monitored daily. Seedlings were identifiedand removed from the flats to prevent competition forlight and nutrients with new seedlings. If they werenot yet identifiable, each was transplanted to anotherpot for future identification. Only in 2000, we werenot interested in identifying other seeds than Carexspp., Juncus spp. and Rubus spp. (the most domi-nant species of the 1998 and 1999 experiments), whatreduced the amount of work substantially. Since noother species than C. pallescens, J. effusus and R. id-aeus of the genus Carex, Juncus and Rubus emergedin the 2000 experiment, we will focus on these spe-cies only. The soil in the flats was stirred at theend of the second month to stimulate germination ofdeeper buried seeds. Each experiment was stoppedafter ca. 3–4 months, since the database of Thompsonet al. (1997) reveals that up to 3 months was judgedadequate for most viable seeds to germinate from re-duced, thin layered soil samples, as is the case in this

58

study (whereas up to one year was necessary when thebulk of the samples was not reduced).

In 2000, one split of all soil samples was treatedwith a calcium nitrate solution (Ca(NO3)2 ×4 H2O) tostimulate germination (compare Auchmoody, 1979).The flats were irrigated twice (end of February andmiddle of March) with a solution of 100 mg NO3per liter deionized water. A total of 4 L was usedfor all 120 flats (4 stands × 5 replications × 6 hori-zons = 120 flats). Assuming that 20 g soil of eachhorizon was used to fill the flats (the ratio massof sample/mass of geometric horizon was used forupscaling) this treatment roughly represents a nitro-gen fertilization with 12 kg N ha−1 on Flysch and6 kg N ha−1 on Molasse (due to a lower bulk density)for the 0–5 cm soil horizon. The second split (addi-tional 120 flats) was treated in exactly the same waywith deionized water only.

Results and discussion

Seed bank size and composition in 1998 and 1999

Numbers of species emerging from soil samples col-lected in June 1998 (forest floor and mineral soil downto 4 cm depth) and March 1999 (down to 5 cm soildepth) are given for the experimental site Kreisbach(Flysch) in Table 4. Total seedling density was es-timated 15176 seedlings m−2 under the mixed speciesstand and 9043 under the pure spruce stand in 1998. In1999, densities were lower with a higher number underpure spruce (7485 versus 1663 seedlings m−2). Jun-cus effusus (up to 89%), B. pendula (up to 12%) andC. pallescens (up to 8%) were the dominant species.Species richness amounted between 2 (March 1999,mixed stand) and 13 (June 1998, mixed stand).

Our results show that the time of soil sample col-lection (spring versus summer or autumn) has a verystrong impact on emergent densities. The observed de-crease of emergents and species richness from summerto spring samples of the following year at the experi-mental site Kreisbach (see Table 4 and Figure 1) washigher under mixed species than under pure spruceand may be partly interpreted by increased loss dueto predation, since, e.g., birds and rodents are moreattracted by mixed species than by spruce monocul-tures. Own observations within a study of nutrientfluxes at the same sites yielded higher bird droppingsin throughfall samplers and higher mice damages oftubings (lysimeters for collection soil solution) and

wires (soil hydrological instruments) in the mixedstand than in the pure spruce stand (compare Louda,1989 and DeGraaf et al., 1998). Earthworms whichprefer higher soil pHs (Neirynck et al., 2000), as mea-sured under the mixed species stands (see Table 1),are handled as seeds consuming animals as well(Louda, 1989). Another part of this loss simply reflectsphysiological death of transient seeds, which persistfor less than one year (Thompson et al., 1997). Hence,the sampling period in early spring reflects the per-sistent viable seed bank after stratification of dormantseeds during the winter period and before the inputof fresh seeds. Based on these results the last detailedexperiment of 2000 was performed with soil samplescollected in early spring (February), since persistentseeds are more likely to affect the success of forestregeneration after, e.g., a sudden breakdown of sec-ondary spruce stands or silvicultural conversions frompure to mixed stand.

At the experimental site Frauschereck (Molasse)seedling densities (forest floor and mineral soil downto 5 cm depth) were much lower than on soils de-veloped on Flysch (Table 5). Seedling densities es-timated from soil samples collected in March 1999were between 260 (pure spruce) and 364 (mixed stand)seedlings m−2 (no significant differences). Emergingseedlings from samples collected in September 1998were significantly lower (P < 0.01) under purespruce (156 seedlings m−2) than under mixed species(4573 seedlings m−2) due to relatively high abund-ances of J. effusus (73%) and R. idaeus (20%).

The number of species within the same overstoreyspecies composition was not much different betweensoil samples collected in September and March of thefollowing year at the experimental site Frauschereck.Under the mixed species stands slightly more spe-cies (4–5) emerged than under the pure spruce stand(1–2). In 1998 and 1999 the amount of J. effusus seed-lings emerged from soil samples under pure sprucewas zero. However, under mixed species a relativelyhigh number was estimated for J. effusus on samplescollected in September 1998, but again no seedlingswere counted after spring sample collection in 1999.Hence, this particular high seedling density (3326J. effusus seedlings m−2) under the mixed speciesstand in September 1998 was probably caused by freshseed rain or some kind of disturbance which favoredthe germination of dormant seeds. In accordance withLudemann (1994) fresh seed rain is only likely fromthe surrounding vegetation, since none of the speciesare listed in Table 3 (aboveground soil vegetation).

59

Figure 1. Depth distribution of emerging Juncus effusus (above) and Carex pallescens (below) from soil samples collected in June 1998 (left)and March 1999 (right) at the experimental site Kreisbach. Absolute numbers of emergants are plotted as seedlings m−2 per horizon (means of5 replications; error bars are SE), and the relative contribution of each horizon is given as percentage of the total amount of seedlings down to4 cm (June 1998) and down to 6 cm (March 1999), respectively.

He concludes that most of the species represented inthe seed bank of forests (different combinations ofspruce, fir, beech, maple within the Black Forest ofGermany) are typical for the edge of forests and forestroads or for clearings; in the present vegetation of theforests these species are rare or completely missing. Inagreement with our results Ludemann (1994) recordedespecially frequently seedlings from the genera Carex,

Juncus and from the species Rubus idaeus. We do notknow of any disturbances within the 50 × 50 m plots,however, forest management activities are likely tohave a long lasting impact on the seed bank character-istics of our study sites; e.g., soil compaction causedby logging or thinning favors seedling establishmentof Juncus effusus (Sun, 2000). Berger et al. (2004)suggests that disturbances, which increase light and

60

Table 4. Species emerging from soil samples collected in June 1998 (forest floor and mineral soildown to 4 cm depth) and March 1999 (forest floor and mineral soil down to 5 cm depth) at theexperimental site Kreisbach (seedlings m−2). Differences (number of emergents) between the purespruce and the adjacent mixed beech-spruce stand were tested and the level of significance is shownas ∗∗∗: P < 0.001; ∗∗: P < 0.01; ∗: P < 0.05; ns: P > 0.05

Species name Seedbank typea June 1998 March 1999

Pure Mixed Pure Mixed

Betula pendula 1, 2 1039 156 ∗∗ 728 0 ∗∗Carex pallescens 2(3) 728 156 ∗ 468 0 ns

Cirsium arvense 1-4 0 52 ns

Epilobium roseum 1 52 52 ns

Fragaria vesca 1(2) 104 0 ns

Galium rotundifolium – 52 0 ns

Gnaphalium sylvaticum 2(4) 0 156 ns 52 0 ns

Hieracium sylvaticum 1 0 52 ns

Hypericum maculatum 1-4 52 364 ns

Hypericum perforatum 1-4 156 0 ns

Juncus effusus 1-4 6756 13564 ∗∗ 5405 1507 ∗∗∗Oxalis corniculata 4(2) 104 208 ns

Poa nemoralis 1(2) 52 0 ns

Rubus fruticosus 1-4 52 52 ns

Salix caprea 1 104 156 ns

Scirpus sylvaticus 1-4 416 0 ∗∗Sonchus oleraceus 1-4 52 104 ns

Taraxacum officinale 1-4 52 208 ns 52 0 ns

Verbascum thapsus 3(2) 0 52 ns

Veronica officinalis 1-4 52 0 ns

Sum of emergents 9043 15176 ∗ 7485 1663 ∗∗∗Number of species 11 13 10 2

aSeedbank type is given according to Thompson et al. (1997): 1: transient, seeds which persist for lessthan one year; 2: short-term persistent, seeds which persist in the soil for at least one year; 3: long-termpersistent, seeds which persist in the soil for at least five years; 4: present: seeds present but cannot beassigned to one of the three seed bank types.

temperature regime and consequently mobilization ofnutrients, sequestered temporarily in the forest floor,are more frequent in mixed species stands than in purespruce stands, as indicated by dendrochemistry.

According to the discussions above it seems jus-tified to assume that the germination experiments of1998 (June and September) represent the viable seedbanks shortly after fresh seed rain (sum of transientand persistent seeds) while the experiments of 1999(March) reflect persistent seeds only. Based upon thestated assumption and the fact that the seedling dens-ity was significantly higher under the mixed speciesoverstorey in 1998 but similar (Frauschereck) or sig-nificantly lower (Kreisbach) than under pure spruce in1999, we postulate the following hypothesis: mixedspruce-beech (coniferous-broadleaf) stands enhancethe viable transient soil seed bank but pure spruce

(pure coniferous) stands increase the viable persistentsoil seed bank. We suggest that the enhanced viabletransient soil seed bank under mixed stands is causedby more frequent disturbances (Berger et al., 2004)and consequently higher seed rain from the surround-ing vegetation (increased by higher wind dispersalwithin the stands due to a lower leaf area, espe-cially during early spring), while the reduced numberof viable transient seeds (relative increase of persist-ent seeds) under pure spruce is attributed to a morenegative soil climate (see Table 1).

Seed banks of Carex pallescens, Juncus effusus andRubus idaeus in 2000

Mean numbers of emerging J. effusus, C. pallescens,R. idaeus (species are ordered by decreasing num-

61

Table 5. Species emerging from soil samples (forest floor and mineral soil down to 5 cmdepth) collected in September 1998 and March 1999 at the experimental site Frauschereck(seedlings m−2). Differences (number of emergents) between the pure spruce and the adjacentmixed beech-spruce stand were tested and the level of significance is shown as ∗∗: P < 0.01;∗: P < 0.05; ns: P > 0.05

Species name Seedbank typea September 1998 March 1999

Pure Mixed Pure Mixed

Angelica sylvestris 1(2) 0 208 ns

Carex pallescens 2(3) 156 0 ns

Epilobium spp. 1 0 52 ns

Juncus effusus 1-4 0 3326 ∗Rubus fruticosus 1-4 0 52 ns

Rubus idaeus 1-4 0 935 ns 156 156 ns

Salix alba 3(1) 104 0 ns

Scrophularia nodosa 3,4 0 104 ns

Sambucus spp. 1-3 0 52 ns

Taraxacum officinalis 1-4 0 52 ns

Sum of emergents 156 4573 ∗∗ 260 364 ns

Number of species 1 4 2 5

aSeedbank type is given according to Thompson et al. (1997): 1: transient, seeds which persistfor less than one year; 2: short-term persistent, seeds which persist in the soil for at least oneyear; 3: long-term persistent, seeds which persist in the soil for at least five years; 4: present:seeds present but cannot be assigned to one of the three seed bank types.

bers) and the sum of these three species (N m−2

per soil horizon) from soil samples collected in Feb-ruary 2000 at the experimental sites Kreisbach andFrauschereck are given in Table 6. A 2 × 2 × 2 AN-OVA (Backhaus et al., 1994) was performed to testdifferences between means of the grouping variablesbedrock material (Flysch, Molasse), species compos-ition (pure spruce stand, mixed species stand) andtreatment (control, nitrate) for each soil horizon (usedsoftware package: SPSS for Windows, Release 10.0.7,June 2000, Standard Version). Seed bank densities arehigher for the nutrient rich soil on Flysch than for thenutrient poor, acidic soil developed on Molasse. Thesum of all three species within the upper 15 cm soildepth are 9732 (Flysch) versus 2728 seedlings m−2

(Molasse; P < 0.01; see Table 6).

Seed bank size and composition compared with otherstudies

The methods used in a seed bank study can affectdensity and should be taken into consideration whencomparing studies (Simpson et al., 1989). Kostel-Hughes et al. (1998) compared seed bank studies ofdeciduous forests and pointed out the following con-cerns: (i) Direct counting with viability testing versusseedling emergence, although all studies in his review

utilized the emergence method; (ii) Duration of ger-mination trials ranged from five weeks to three years,although most germination (> 90%) occurred in thefirst year; (iii) Differences in total area sampled andsample depth.

In general, estimated mean emergent densitiesare roughly within the range of other studies. Meanseed densities of soil seed banks in temperate de-ciduous forests in the northeastern US were summar-ized by Kostel-Hughes et al. (1998) ranging from91 to 9651 seedlings m−2. Kellmann (1970) coun-ted > 1000 seeds m−2 in soil cores collected from a100-years old Pseudotsuga menziesii and Tsuga het-erophylla forest site in British Columbia. Seedlingemergence in conifer-dominated sites in New Bruns-wick (Moore and Wein, 1977) ranged from 580 seed-lings m−2 for a Picea mariana and Pinus strobus standto 180 seedlings m−2 for a Larix laricina stand. Au-gusto et al. (2001) studied seed banks of forests innortheastern France and recorded 874 seedlings m−2

under a beech stand and 3537 and 1517 seedlings m−2

under two different spruce stands at soil pH 3.7 and4.8, respectively (pH water in 0–5 cm soil depth).At pH 3.7 Juncus spp. amounted 71% but at pH 4.8only 2% of the total seedling density. The numberof dormant, viable seeds under a Melico-Fagetum

62

Table 6. Mean number of emerging Juncus effusus, Carex pallescens, Rubus idaeus and the sum of these three species (N m−2 per soilhorizon) from soil samples collected in February 2000 at the experimental sites Kreisbach and Frauschereck. A 2 × 2 × 2 ANOVA wasperformed to test differences between means of the grouping variables bedrock material (Flysch - basic soil, Molasse - acidic soil), speciescomposition (pure spruce stand, mixed species stand) and treatment (control, nitrate) for each soil horizon. Significant interactions betweengrouping variables indicate that these variables can not be tested individually but affect the dependent variable jointly. Significance of eachfactor is shown as n.s., P > 0.10; (∗), P < 0.10; ∗, P < 0.05; ∗∗, P < 0.01. Total number of replications = 5 replications per variant×2 × 2 × 2 × 6 horizons = 240. A Scheffe’s multiple range test was performed to test differences between soil depths (forest floor andindividual 5-cm horizons) within each group: different letters within a single row indicate differences; P < 0.05, only significant resultsare given; a represents the lowest mean

Horizon Bedrock material (1) Species composition (2) Treatment (3) Interaction

Flysch Molasse pure mixed control nitrate

Juncus effusus

Forest floor 0a 0 ns 0 0 ns 0a 0a ns ns

0–5 cm 3001ab 590 ∗ 2154 1437 ns 3161b 430ab ∗ ns

5–10 cm 4107b 311 ∗∗ 2885 1533 ns 1335ab 3082b ns ns

10–15 cm 985ab 470 ns 384 1071 ns 813ab 642ab ns ns

Sum 8093 1371 ∗∗ 5423 4041 ns 5309 4154 ns ns

20–25 cm 710ab 0 ns 0 710 ns 580ab 130a ns ns

30–35 cm 152a 0 ns 152 0 ns 152a 0a ns ns

Carex pallescens

Forest floor 0 510 ns 510 0 ns 0 510 ns ns

0–5 cm 321 225 ns 210 337 ns 546 0 ∗ ns

5–10 cm 390 0 (∗) 390 0 (∗) 132 258 ns (1) × (2)∗10–15 cm 397 0 ns 397 0 ns 269 128 ns ns

Sum 1108 735 ns 1507 337 (∗) 947 896 ns ns

20–25 cm 134 0 ns 134 0 ns 0 134 ns ns

30–35 cm 152 0 ns 152 0 ns 0 152 ns ns

Rubus idaeus

Forest floor 0 134ab ns 0 134 ns 134 0 ns ns

0–5 cm 254 488b ns 394 348 ns 487 255 ns (2) × (3)∗5–10 cm 127 0a ns 0 127 ns 127 0 ns ns

10–15 cm 150 0a ns 150 0 ns 150 0 ns ns

Sum 531 622 ns 544 609 ns 898 255 (∗) ns

20–25 cm 0 0a ns 0 0 ns 0 0 ns ns

30–35 cm 0 0a ns 0 0 ns 0 0 ns ns

Jun. + Car. + Rub.

Forest floor 0a 644ab (∗) 510 134 ns 134a 510ab ns (1) × (2) × (3)(∗)

0–5 cm 3576ab 1303b (∗) 2758 2122 ns 4194b 685ab ∗∗ ns

5–10 cm 4624b 311ab ∗∗ 3275 1660 ns 1594ab 3340b ns ns

10-15 cm 1532ab 470ab ns 931 1071 ns 1232ab 770ab ns ns

Sum 9732 2728 ∗∗ 7474 4987 ns 7154 5305 ns ns

20–25 cm 844ab 0a (∗) 134 710 ns 580a 264a ns ns

30–35 cm 304a 0a ns 304 0 ns 152a 152a ns ns

63

typicum and a Melico-Fagetum allietosum in Ger-many ranged between 2500 and 9000 seedlings m−2

within the upper 6 cm of the soil (Fischer, 1987).Hence, measured densities on Flysch (Kreisbach) tendto be within the upper range, due to a large per-centage of graminoids (J. effusus and C. pallescens),while densities on the nutrient poor soils on Molasse(Frauschereck) in 1998 and 1999 seem to encom-pass the lower range, especially if seedling densitieswithout J. effusus are compared with other studies.However, as was pointed out above, it may not be jus-tified at all to compare estimated mean densities of thesoil seed banks of the studied forest stands with otherstudies due to the application of different methods.Hence, this study will put more emphasis on relativecomparisons between different species compositions,bedrock materials and treatments.

Relationship between species in the seed bank and thevegetation

One of the more widely held generalizations aboutseed banks both in coniferous forests (Archibold,1998) as well as in deciduous forests (Pickett and Mc-Donnell, 1989) is their lack of correspondence withthe aboveground vegetation. Plant species exhibit awide range of reproductive strategies, especially withregard to seed production and dormancy, which tend todecline with successional age. Consequently it shouldnot be surprising that no seedlings of Picea abies orFagus sylvatica germinated in the soil samples (seeTables 4 and 5), which are the dominant forest specieswithin the overstorey of the studied stands.

Salix spp. and Betula spp., both of early succes-sional status, were the only trees present in the soilseed bank. As reported elsewhere (e.g., Pickett andMcDonnell, 1989; Kostel-Hughes et al., 1998), Rubusspp., another early successional species, was the soleor dominant shrub taxon in the seed bank.

From a total of 44 different species (see Table 3)which were present in the vegetation among all 4 sitesonly the following 3 species emerged from the seedbank samples: Galium rotundifolium, Rubus fruticosusand Rubus idaeus.

The majority of emergents are the graminoids J. ef-fusus (up to 89%) and C. pallescens (the only speciesthat emerged from soil samples under pure spruceat Frauschereck, collected in September 1998; seeTable 5), which were not present at all in the above-ground vegetation (see Table 3). Hence, care should be

taken if management strategies create conditions thatare generally favorable to germination.

Factors controlling seed bank size and composition

Soil depthSeedling densities given in the Tables 4 and 5 are themean (average of 5 soil cores) sums of the forest floorand 1-cm horizons down to 5 cm soil depth (exceptin June 1998 down to 4 cm soil depth only). As anexample, the depth distribution of the graminoids J. ef-fusus and C. pallescens, which emerged at Kreisbachin high numbers under both stand types, is plotted inFigure 1. While for the June 1998 samples the densityof J. effusus was higher under the mixed stand thanunder pure spruce the ratio was opposite for the March1999 samples. According to the given depth distri-butions (Figure 1) it is not justified to calculate totalseedling densities from samples down to 4 cm and6 cm soil depth respectively, since densities did notdecline with increasing depth.

This fact is in contrast to other studies that reporteda tendency for overall density of emergents to de-cline with depth (Moore and Wein, 1977; Grabner andThompson, 1978). In accordance with Kostel-Hugheset al. (1998) graminoids (in our case J. effusus andC. pallescens) even tend to increase with soil depthwithin the upper 6 cm (Figure 1). The 2000 experimentreveals that emergent densities start to decline below10 cm soil depth. While a few seedlings were countedfor J. effusus and C. pallescens even in 20 to 35 cm soildepth; no R. idaeus emerged below 20 cm soil depth.According to Thompson et al. (1997) Juncus spp. andCarex spp. are definitely classified as long-term per-sistent (seeds which persist in the soil for at leastfive years) within our study, since seeds are at leastas frequent in lower as in upper soil layers. In addi-tion, erroneous persistent records for transient speciesare much rarer than erroneous transient records forpersistent species (Thompson et al., 1997). Thus anyspecies with several type 3 records (see Tables 4 and5) probably is long-term persistent, even if type 1 and2 records are more numerous. Rubus idaeus is morefrequent in the upper mineral soil (0–5 cm) but presenttoo in lower soil layers (see Table 6); which wouldbe the definition for short-term persistent seeds (whichpersist for at least one year in the soil; Thompson et al.1977). However, if we consider the lower densities inthe forest floor than in 0–5 cm mineral soil depth andnumerous type 3 records, it seems more appropriate

64

to classify R. idaeus as long-term persistent as wellwithin the studied seed banks.

The sum of emerging J. effusus, C. pallescens andR. idaeus increased from 0 (forest floor) to 4624 (5–10 cm) and declined to 304 seedlings m−2 (30–35 cm)on Flysch (see Table 6 for statistical differences).However, on Molasse 644 seedlings were counted inthe forest floor and the number reached its maximumin 0–5 cm soil depth; no seedlings germinated be-low 20 cm soil depth. Significant differences of seeddensities between Flysch and Molasse for most hori-zons indicate that the soil substrate affects both theaverage and maximum depth of buried viable seeds.Greater vertical mixing of the soil and its seeds maybe caused by a higher abundance of earthworms inthe nutrient rich soils developed on Flysch (compareNeirynck et al., 2000). In addition, small seeds likethat of Juncus spp. may percolate through the soilwith rainwater, particularly through soils that werecracked after summer dryness (Bigwood and Inouye,1987). Such phenomena of a variable macropore sys-tem, which are common in loamy to clayey soils, aredocumented for the experimental site Kreisbach byHager and Weinmeister (2001).

Soil substrateSoil type is frequently mentioned as controlling seedbank composition, although this factor will often beconfounded with vegetation type (Pickett and McDon-nell, 1989). Both the data of all emerging species(1998 and 1999; Tables 4 and 5) as well as the resultsof the most common species J. effusus, C. pallescensand R. idaeus (2000; Table 5) reveal that the soil sub-strate affects the seedling density significantly. Seedbank densities are higher for the nutrient rich soil onFlysch than for the nutrient poor, acidic soil developedon Molasse with clear signs of carbon sequestration inthe top horizons (see Table 1). This decline of viableseeds within the soil on Molasse is probably causedby higher mortality rates due to a negative soil chem-ical climate: in accordance to Brown and Oosterhuis(1981) the studied soils with low pH support smallerseed banks, and in agreement with Moore and Wein(1977) organic soils support smaller seed banks thando mineral soils.

Species richness of seed bank composition washigher on Flysch than on Molasse (see Tables 4 and5). This is in accordance with Augusto et al. (2001),who found a higher species richness in the seed banksof soils with higher pH (4.8 vs. 3.7 in 0–5 cm soildepth) for pairs of oak, pine and spruce stands. How-

ever, seedling density under Norway spruce standswas higher on the acidic soil due to the dominantabundance of Juncus spp. (71% vs. 2%).

Overstorey species compositionOverstorey species composition (pure versus mixed)did not significantly affect the seed bank density ofthe persistent seeds (at the 5% level of significance;Table 6). However, there is a trend that densities ofJ. effusus and C. pallescens and consequently the sumof J. effusus, C. pallescens and R. idaeus within theupper 15 cm soil depth are higher under pure spruce(7474 seedlings m−2) than under mixed species (4987seedlings m−2). A similar trend is deducible frommore stratified means of the non treated samples (notshown in Table 5): Flysch: 13707 vs. 8631; Molasse:3620 vs. 2659 (pure vs. mixed; seedlings m−2 withinthe upper 15 cm soil depth). This trend is supportedby a significantly higher number of seedlings underpure spruce, which emerged from soil samples col-lected in March 1999 at Kreisbach (7485 vs. 1663seedlings m−2 in the upper 5 cm soil; Table 4).Augusto et al. (2001) measured significantly higherseedling densities in the upper 5 cm soil under a Piceaabies (3537 seedlings m−2) stand than under a Fagussylvatica stand (874 seedlings m−2) as well.

Our results are difficult to interpret since on the re-gional scale higher soil pHs and nutrient contents (e.g.,Flysch vs. Molasse) increase seed bank densities, onthe other hand, pure spruce, which is known to acidifythe top soil (Berger et al., 2002; see Table 1, study siteKreisbach), tends to stimulate a higher persistent seedbank density in comparison to adjacent mixed speciesstands. On this local scale, the fate of persistent seeds,once incorporated into the bank, is likely to contrib-ute to this difference. As discussed above, we suggestthat increased predation (compare higher loss of seedsfrom summer to spring sample collection under themixed species stands) and burial at excessive depth(compare chapter Soil depth) tend to increase densitiesof viable long-term persistent seeds under pure spruce.

Nitrate triggerAccording to Auchmoody (1979) the mechanism re-sponsible for germination has long been associatedwith forest disturbance, and hence with accompany-ing changes in light density and quality, mechanicalstirring of the forest floor, and different soil tempera-ture and moisture regimes. However, the same authorshowed that nitrogen fertilizers triggered germina-tion of dormant Prunus pensylvanica seeds naturally

65

buried in the forest floor of 60-year old Alleghenyhardwood stands. In accordance with similar studiesby Vegis (1964) and Hendricks and Taylorson (1974)he concludes that only nitrate-containing compoundsare effective in stimulating germination of dormantseeds, but that ammonium sources usually are not.

Nitrogen amounts, added on small soil volumes,are much lower than Auchmoody (1979) usedfor his treatments in the field (between 56 and336 kg N ha−1). However, we suggested that highnitrate concentrations rather than high total nitrateamounts trigger germination. The results of the treat-ments are given in Table 6. The nitrate trigger didnot cause significant differences of germination ratesof viable buried seeds, except for 0–5 cm soil depth,where the nitrate treatments reduced seedling densitiesof J. effusus and C. pallescens (P < 0.05). How-ever, the nitrate treatments caused an overall trend ofdeclining seedling densities (sum of all three specieswithin the upper 15 cm soil depth: control: 7154 seed-lings m−2; nitrate: 5305 seedlings m−2). It is verylikely that high nitrification rates occur at greenhouseconditions (stirring of soil, increased temperature andlight regime). Hence, any additional nitrate suppliedwith the treatment may toxify the soil sample.

The fact, that a known germination trigger did notincrease germination of buried seeds may justify theconclusion that the number of emerged seedlings isvery close to the real seed bank density. This means,that methods based on separation and identificationof seeds (which are very costly and time consuming)will not yield higher seedling densities than the de-scribed emergence method for this study. Sun (2000)applied exactly the same emergence method for ana-lyzing soil seed banks in Viennese grasslands andused a stereo microscope to test whether ungermin-ated seeds remained after cessation of the germinationexperiment. She concludes that only very few seedsremained ungerminated (without testing the viabilityof the detected seeds) supporting the reliability of theapplied emergence method.

Conclusions and management implications

The main goal of this study was to evaluate the riskof weeds (e.g., Millbacher, 1987, 1992; Schreineret al., 2000; Romagosa and Robison, 2003) whichmay limit the success of conversions of secondarypure spruce stands to mixed species stands. Accord-ing to Thompson (2000) no mature forest tree spe-

cies in north-western Europe has persistent seeds,for the simple reason that the lifespan of the domi-nant trees exceeds that of even the most long-livedseeds. In fact, no seeds of Picea abies or Fagus sylvat-ica, the dominant forest species within the overstoreyof the studied stands, were found in the soil seed bank,underlining the importance of this issue for forestrestoration management practices due to a possiblecompetition of tree species with weeds.

Overstorey species composition (pure versusmixed) did not significantly affect the seed bank dens-ity of the persistent seeds (hypothesis 1; see introduc-tion). However, there is a trend that densities of thehighly abundant graminoids J. effusus and C. palles-cens, which form long-term persistent seed, are higherunder pure spruce than under mixed species, prob-ably due to increased predation and burial at excessivedepth under the mixed species stand. Due to compar-isons between seedling densities estimated from soilsamples collected later in the vegetation period and inspring we hypothesize that mixed spruce–beech standsadvance the transient seed bank while pure sprucestands stimulate the persistent seed bank. Hence, theseason of the year may play an important role formixed species stands whether disturbances by thinningor clearing promote the risk of weeds.

The soil substrate affects seed bank density andmaximum depth of burial of viable seeds significantly(hypothesis 2). Seed bank densities are higher for thenutrient rich soil on Flysch than for the nutrient poor,acidic soil developed on Molasse with clear signs ofcarbon sequestration in the top horizons. Greater ver-tical mixing of the soil and its seeds may be caused bya higher abundance of earthworms in the nutrient richsoil on Flysch. In addition, small seeds like that of Jun-cus spp. may percolate through loamy to clayey soils(e.g., soils on Flysch) that were cracked after summerdryness.

Nitrate treatments did not promote germination ofviable buried seeds (hypothesis 3). This fact justifiesthe conclusion that the number of emerged seedlingsis a realistic indicator of the seed bank density for thestudied stands.

The majority of emergents are the graminoids J. ef-fusus and C. pallescens which were not present at allin the aboveground vegetation. Hence, care shouldbe taken if management strategies create conditionsthat are generally favorable to germination. Due toa possible competition of favorable tree species withgraminoids the success of forest regeneration or a con-version of pure spruce to mixed species stands could

66

be endangered by any disturbance, which causes animmediate increase of light levels.

Acknowledgements

This research was supported by the Austrian ScienceFoundation (FWF, Grant No. P15496 to T.W.B.) andby the Austrian Ministry of Agriculture and Forestry(BMfLuFw, Grant No. GZ: 56.810/06-VA2b/98 toG.G.). Gerhard Karrer participated in helpful discus-sions. We thank two anonymous reviewers for theircritical comments for the improvement of this paper.

References

Archibold O W 1989 Seed banks and vegetation processes in con-iferous forests. In Ecology of Soil Seed Banks. Eds. M A Leck,V T Parker and R L Simpson. pp. 107–122. Academic Press, Inc.,San Diego.

Ashton P M S, Harris P G and Thadani R 1998 Soil seed bank dy-namics in relation to topographic position of a mixed-deciduousforest in southern New England, USA. For. Ecol. Manage. 111,15–22.

Auchmoody L R 1979 Nitrogen fertilization stimulates germinationof dormant pin cherry seed. Can. J. For. Res. 9, 514–516.

Augusto L, Dupouey J-L, Picard J-F and Ranger J 2001 Potentialcontribution of the seed bank in coniferous plantations to the res-toration of native deciduous forest vegetation. Acta Oecologica22, 87–98.

Backhaus K, Erichson B, Plinke W and Weiber R 1994 Multivari-ate Analysemethoden, eine anwendungsorientierte Einführung.Springer-Verlag, Berlin Heidelberg, 594 pp.

Berger T W 2001 Impact of species composition on forest soil prop-erties of secondary spruce and mixed spruce-beech stands. Cbl.Ges. Forstw. 118, 193–216.

Berger T W and Hager H 2000 Physical top soil properties in purestands of Norway spruce (Picea abies) and mixed species standsin Austria. For. Ecol. Manage. 136, 159–172.

Berger T W, Neubauer C and Glatzel G 2002 Factors controllingsoil carbon and nitrogen stores in pure stands of Norway spruce(Picea abies) and mixed species stands in Austria. For. Ecol.Manage. 159, 3–14.

Berger T W, Köllensperger G and Wimmer R 2004 Plant-soil feed-back in spruce (Picea abies) and mixed spruce-beech (Fagussylvatica) stands as indicated by dendrochemistry. Plant Soil,264, 69–83.

Bigwood D W and Inouye D W 1987 Spatial pattern analysis of seedbanks: An improved method and optimized sampling. Ecology69, 497–507.

Bormann F H and Likens G E 1994 Pattern and Process in a ForestedEcosystem. Springer Verlag, New York, 253 pp.

Braun-Blanquet J 1964 Pflanzensoziologie - Grundzüge der Vege-tationskunde. Springer, Vienna, New York. 865 pp.

Brown A H F, Oosterhuis L 1981 The role of buried seed incoppicewoods. Biol. Conserv. 21, 19–38.

Buckley G P, Howell R and Anderson M A 1997 Vegetationsuccession following ridge edge management in lowland planta-tions and woods. 2. The seed bank resource. Biol. Conserv. 82,305–316.

Cavers P B and Benoit D L 1989 Seed banks in arable land. InEcology of Soil Seed Banks. Eds. M A Leck, V T Parker andR L Simpson. pp. 309–328. Academic Press, Inc., San Diego.

DeGraaf R M, Hestbeck J B and Yamasaki M 1998 Associationsbetween breeding bird abundance and stand structure in theWhite Mountains, New Hampshire and Maine, USA. For. Ecol.Manage. 103, 217–233.

Edwards C A and Bohlen P J 1996 Biology and Ecology ofEarthworms. Chapman and Hall, New York, 426 pp.

Encke F, Buchheim G and Seybold S 1994 ZANDER Handbuch derPflanzennamen. Ulmer, Stuttgart, 810 pp.

FBVA 1998 Österreichische Waldinventur (CD-ROM). ForstlicheBundesversuchsanstalt, Vienna.

Fischer A 1987 Untersuchungen zur Populationsdynamik am Be-ginn von Sekundärsukzessionen - Die Bedeutung von Samen-bank und Samenniederschlag für die Wiederbesiedlung vegeta-tionsfreier Flächen in Wald- und Grünlandgesellschaften. Diss.Bot. 110, Cramer, Berlin, Stuttgart, 234 pp.

Grabner R E and Thompson D F 1978 Seeds in the OrganicLayers and Soil of Four Beech-Birch-Maple Stands. U.S.D.A.,Broomall, PA. Cited in F Kostel-Hughes, T P Young and M JMcDonnell 1998 The soil seed bank and its relationship to theaboveground vegetation in deciduous forests in New York City.Urban Ecosyst. 2, 43–59.

Hager H and Weinmeister W 2001 Stand Climate and Forest Hy-drology and their Alterations by Measures of Forest EcosystemRestoration. In Special Research Program on Forest EcosystemRestoration, Final Report, Project Part F008-003. pp. 80–93.Univ. f. Bodenkultur, Vienna.

Hendricks S B and Taylorson R B 1974 Promotion of seed germin-ation by nitrate, nitrite, hydrozylamine, and ammonium salts.Plant Physiol. 54, 304–309.

Hilhorst H W M and Karssen C M 2000 Effect of chemical envir-onment on seed germination. In Seeds: The ecology of regener-ation in plant communities. Ed. M Fenner. pp. 293–309. CABIpublishing, Oxon, UK.

Honnay O, Bossuyt B, Verheyen K, Butaye J and Jacquemyn H 2002Ecological perspectives for the restoration of plant communitiesin European temperate forests. Biodivers. Conserv. 11, 213–242.

Johnson D W, Walker R F and Ball J T 1995 Lessons from lysi-meters: Soil N release from disturbance compromises controlledenvironmental study. Ecol. Applic. 5, 395–400.

Kellmann M 1970 The viable seed content of some forest soils incoastal British Columbia. Can. J. Bot. 48, 1383–1385.

Kostel-Hughes F, Young T P and McDonnell M J 1998 The soilseed bank and its relationship to the aboveground vegetation indeciduous forests in New York City. Urb. Ecosyst. 2, 43–59.

Louda S M 1989 Predation in the dynamics of seed regeneration. InEcology of Soil Seed Banks. Eds. M A Leck, V T Parker and R LSimpson. pp. 25–51. Academic Press, Inc., San Diego.

Ludemann T 1994 Zum Samenvorrat von Waldböden im Feld-berggebiet. Carolinea 52, 45–60.

Mayer H 1974 Wälder des Ostalpenraumes. Gustav Fischer Verlag,Stuttgart. 344 pp.

Millbacher E 1987 Blattflächen- und Biomassenentwicklung sowieNährstoffhaushalt einer Vergrasung mit Calamagrostis epigeiosin einer Fichtenaufforstung auf Pseudogley. Diplomarbeit, Univ.f. Bodenkultur, Vienna, 89 pp.

Millbacher E 1992 Der Einfluss einer Vergrasung auf den Wasser-haushalt von Fichtenkulturen. Forstliche Schriftenreihe der Univ.f. Bodenkultur, Band 6, Österr. Ges. f. Waldökosystemforschungund experimentelle Baumforschung, Vienna, 142 pp.

67

Moore J M and Wein R W 1977 Viable seed populations by soildepth and potential site recolonization after disturbance. Can. J.Bot. 66, 169–172.

Mucina L, Grabherr G and Ellmauer T 1993 Die Pflanzengesell-schaften Österreichs. Teil 3: Wälder und Gebüsche. Gustav-Fischer, Jena, Stuttgart, New York, 353 pp.

Neirynck J, Mirtcheva S, Sioen G and Lust N 2000 Impact ofTilia platyphyllos Scop., Fraxinus excelsior L., Acer pseudo-platanus L., Quercus robur L. and Fagus sylvatica L. on earth-worm biomass and physico-chemical properties of a loamytopsoil. For. Ecol. Manage. 133, 275–286.

Neubauer C 2002 Einfluss der Baumartenmischung auf Bodenzus-tand und Nährstoffflüsse. Ph.D. Dissertation, Univ. f. Bodenkul-tur, Vienna. 110 pp.

Onaindia M and Amezaga I 2000 Seasonal variation in the seedbanks of native woodland and coniferous plantations in NorthernSpain. For. Ecol. Manage. 126, 163–172.

Pickett S T A and McDonnell M J 1989 Seed bank dynamics intemperate deciduous forest. In Ecology of Soil Seed Banks. Eds.M A Leck, V T Parker and R L Simpson. pp. 123–148. AcademicPress, Inc., San Diego.

Romagosa M A and Robison D J 2003 Biological constraints onthe growth of hardwood regeneration in upland Piedmont forests.For. Ecol. Manage. 175, 545–561.

Rothe A, Kreutzer K and Küchenhoff K 2002 Influence of treespecies composition on soil and soil solution properties in twomixed spruce-beech stands with contrasting history in SouthernGermany. Plant Soil 240, 47–56.

Rottmann M 1989 Wind- und Sturmschäden im Wald: Beiträge zurBeurteilung der Bruchgefährdung, zur Schadensvorbeugung undzur Behandlung sturmgeschädigter Nadelholzbestände. Sauer-länder Verlag, Frankfurt (Main). 128 pp.

Scheffer F and Schachtschabel P 1998 Lehrbuch der Bodenkunde.Ferdinand Enke Verlag, Stuttgart. 494 pp.

Schmidt I and Kazda M 2001 Vertical distribution and radial growthof coarse roots in pure and mixed stands of Fagus sylvatica andPicea abies. Can. J. For. Res. 31, 539–548.

Schmidt-Vogt H 1986 Die Fichte - Band II/1 - Wachstum, Züchtung,Boden, Umwelt, Holz. Paul Parey, Hamburg und Berlin. 563 pp.

Schreiner M, Bauer E-M and Kollmann J 2000 Reducing predationof confier seeds by clear-cutting Rubus fruticosus agg. in twomontane forest stands. For. Ecol. Manage. 126, 281–290.

Schwerdtfeger F 1981 Die Waldkrankheiten. Paul Parey, Hamburg,Berlin. 486 p.

Simpson R L, Leck M A and Parker V T 1989 Seed banks: generalconcepts and methodological issues. In Ecology of Soil SeedBanks. Eds. M A Leck, V T Parker and R L Simpson. pp. 3–8.Academic Press, Inc., San Diego.

Sun B 2000 Qualitative and Quantitative Composition of SoilSeed Banks in Viennese Grasslands. Ph.D. Dissertation, Univ.f. Bodenkultur, Vienna, 130 pp.

Thompson K 2000 The functional ecology of soil seed banks. InSeeds: The Ecology of Regeneration in Plant Communities. Ed.M Fenner. pp. 215–235. CAB International, Oxon, UK.

Thompson K, Bakker J and Bekker R 1997 The Soil Seed Banksof North West Europe: Methodology, Density and Longevity.Cambridge University Press, Cambridge, 276 pp.

Van der Valk A G and Pederson R L 1989 Seed banks and themanagement and restoration of natural vegetation. In Ecology ofSoil Seed Banks. Eds. M A Leck, V T Parker and R L Simpson.pp. 329–346. Academic Press, Inc., San Diego.

Vegis A 1964 Dormancy in higher plants. Annu. Rev. Plant Physiol.15, 185–224.