Diversity and Distribution Pattern of Bryophytes and Vascular Plants in a Boreal Spruce Forest

Small scale pattern of bryophyte communities is one of the remarkable sources of diversity in species-poor boreal forests. By means of correlation and general linear model approaches, the relationships between bryophyte vegetation and upper layers, as well as the response of ground and field layer species to several environmental factors, was analyzed in a boreal spruce forest in South-East Estonia. Of the studied factors, the strongest influence on the diversity and spatial distribution of ground and field layer species was found for ̒ distance from nearest treeʼ. Species from different layers react differently to the proximity of trees. Species richness of bryophytes is higher further from trees, whereas more vascular plant species prefer to grow in the vicinity of tree trunks. For bryophyte species richness, the pH of the decay horizon is also important; fewer bryophyte species occur in more acid conditions.


Introduction
According to numerous authors (e.g.Greig-Smith 1979, Pielou 1984, Legendre & Fortin 1989), organisms are distributed in nature neither uniformly nor randomly.As a rule they are aggregated as patches or are distributed continuously, depending on habitat conditions.Due to the universal character of this phenomenon, investigation of the heterogeneity of populations and communities on various spatio-temporal scales is one of the central questions in plant ecology and plays a crucial role in many ecological theories (Legendre et al. 1985, Addicott et al. 1987) and in the maintenance of species diversity (Mladenoff 1994, Frisvoll andPrestø 1997 etc.).
In boreal forests, bryophytes are important structural components of ground vegetation as well as of the whole community (Lai Roi and Stringer 1976, Longton 1984, Økland 1994, Esseen et al. 1996).It has been demonstrated that the spatial aggregation of species in communities represents habitat partitioning corresponding to microhabitat heterogeneity (Whittaker and Levin 1977, Greig-Smith 1979, 1983, Beatty 1984, Legendre and Fortin, 1989, Slack 1990, Frego and Carleton 1995, Gillet and Gallandat 1996).Several authors have affirmed that in forests, besides the factors of the abiotic environment, the composition and structure of the undergrowth vegetation depends mainly on the tree layer above it, which determines directly or indirectly the conditions for the field and ground layers (Abolin 1974, Kotov 1983, Beatty 1984, Hytteborn et al. 1987, Gustafsson and Hallinbäck 1988, During and Verschuren 1988, Jonsson and Esseen 1990, Saetre 1999).As already stated by Cajander (1926), trees modify to a great extent the light conditions and water and nutrient availability for the understorey, and affect the composition, abundance and spatial distribution of species in the forest understorey.In that way, the plants of the ground and field layers are influenced by the same microenvironment conditions, but, as argued by Scott (1971), Herben (1987), and Slack (1990), due to their specific anatomy, bryophytes react to changes of the microenvironment more rapidly than vascular plants.Studies in grassland communities, where big trees are sparse or missing, have also shown that vascular plants and bryophytes react differently to changes in habitat conditions (Herben 1987, Ingerpuu et al. 1998).How bryophytes and vascular plants share the space in forests, has been discussed only briefly (e.g.Aude andLawesson 1998, Saetre 1999).
In studies of plant-environment relationships, or mutual dependency between different layers of the plant community, it is very important to consider also the spatial scale used for data collection, since the character of these relations may change crucially at different scales (Stowe and Wade 1979, Greig-Smith 1979, McCune and Antos 1981, Hanski 1983, Maslov 1989, Reed et al. 1993).In the current study we analyse the relationships between field and ground layer species and the environment on a decimeter scale.
In the current study our basic object of investigation is the bryophyte vegetation in a typical boreal forest and its relationship with the upper layers and some other environmental factors.The study is addressed at the two main questions: 1) how are bryophytes and vascular plants correlated in forest floor vegetation and, 2) how does ground and field layer structure react to different environmental factors, particularly to the spatial pattern of trees in the same forest community.

Study Area
The study was carried out in the southeastern part of Estonia, in the landscape reserve of the Ahja river valley, near the village of Valgemetsa (58°08ʼN, 26°59ʼE).Phytogeographically this area belongs to the East-Baltic subprovince, where moderately humid boreal forest is typical.Mean annual temperature in winter is -7.0°C and in summer 17.5°C; there are about 200 rainy days per year, most of them occur in the springtime (Laasimer 1965).According to Moen (1999) the area is located in the indifferent oceanic-continental section of the boreonemoral vegetation zone.
An Oxalis-Vaccinium myrtillus site type spruce forest (Paal 1997), representing a species-poor community with simple structure and abundant ground layer, was chosen for the analysis.The dominating trees of Picea abies are about 100 years old.Some trees of Pinus sylvestris also occur in the tree layer; composition of the tree layer is 9 parts Picea abies and 1 part Pinus sylvestris.The diameter at breast height of all trees occurring in the studied forest was at least 24 cm.The canopy coverage is approximately 70%.Natural regeneration is scattered, consisting only of Picea abies.The shrub layer is represented by very sparsely growing Sorbus aucuparia, Frangula alnus and Salix cinerea.

Data Collection and Analyses
The data was collected in September 1996.To avoid the biasing of data due to topographical heterogeneity, a homogenous area without obvious variation in microtopography was chosen.A rectangular grid for 500 plots of 20 × 20 cm was placed down on the ground.In every plot the total cover of bryophytes and vascular plants, and then the cover of each species, was estimated according to the scale 0.1, 1, 2, 3, 5, 7, 10, 15, 20, 25, 30, …, 100 % (Suominen 1969).The tree juveniles, that occurred in plots were included in the field layer.The nanotopographical form (height interval up to 1 dm) was visually evaluated and classified as a concavity, plain, or a hump.The distance from the centre of every plot to the nearest tree trunk was measured in centimetres.From the centre of each plot, the thickness of the decay horizon was measured.Samples of the decay horizon were taken for the estimation of pH in the laboratory.All samples were air-dried and weighed.Before measuring, the soil samples were mixed with 2.5 times their own weight of distilled water, according to Økland and Eilertsen (1996).pH was measured with the pH/Cº-meter E6121A.
Since field and ground layer species form and modify the environment in their surrounding, some plant-related factors such as the number of vascular plant and bryophyte species in a sample plot, and the total cover of vascular plants and bryophytes in a sample plot were treated as additional environmental factors in data processing.
The frequency and mean cover for all species registered in studied plots were calculated.Correlation between cover values of bryophyte and vascular plant species was evaluated on the basis of Spearmanʼs rank correlation coefficients.The general linear model (GLM) was used for analyzing the significance of the environmental factors on the ground and field layer species-richness and on the cover values of the dominant species.In that model, only plots with total cover values greater than zero and species with a relative frequency greater than 20% were included.To describe the effect of environmental factors on species-richness in both considered layers, cover values of the dominant species of the corresponding layer were also included.To approximate distribution to the normal, the cover values were square root transformed before analysis.The nanotopography was treated as a categorical factor with three levels: 1) a concavity, 2) a plain and 3) a hump.When nanotopography had a significant effect on the cover values, the Tukey HSD test was used to determine significant differences between group means.The distance from nearest tree trunk, thickness of decay horizon, and pH of decay horizon were treated as continuous factors in the model.All statistical analyses were performed using the statistical package Statistica 6.0 (StatSoft, Inc. 2001).

Results
In total, twelve bryophyte and thirteen vascular plant species occurred in the ground and field layers of the studied forest.Only one vascular species (Vaccinium myrtillus) and three bryophyte species (Hylocomium splendens, Pleurozium schreberi and Ptilium crista-castrensis) had a frequency of occurrence more than 70% (Table 1).Four bryophyte and four vascular plant species occurred in less than five plots and were treated as occasional species for this community.
Not one of bryophyte species cover values had a significant correlation with the cover of the dominating vascular plant species Vaccinium myrtillus (Table 2).The cover of Pleurozium schreberi had a weak but significant positive correlation with the cover of Maianthemum bifolium and Oxalis acetosella and a negative correlation with the cover of Vaccinium vitis-idaea, while the cover of Hylocomium splendens and Ptilium cristacastrensis had just the opposite relation with the cover of these vascular plant species.The cover of Ptilium crista-castrensis, Aulacomnium palustre and Brachythecium oedipodium was weakly positively correlated with the cover of young regrowth of Picea abies.The latter species tended also to prefer the neighbourhood of Vaccinium vitis-idaea.
According to the general linear model, distance from the nearest tree trunks had significant effect both on the ground and field layer speciesrichness (Table 3).Bryophyte species number increased with distance from the tree trunks (Fig. 1), whereas the species number for vascular plants was higher close to the trees (Fig. 2).Speciesrichnesses of different layers also had opposite regression parameter signs for the increase of total cover of the corresponding layer -the number of    on the vascular plant species-richness.Vascular plants species richness was higher in the nanotopographically lower parts (Fig. 2).The same tendency could be observed for bryophytes, but for them the topographical effect was not significant.
Additional GLM modelling by species revealed that distance from nearest tree had a significant effect on the cover values of almost all dominant ground (Table 4) and field layer species (Table 5).Cover of Ptilium crista-castrensis was higher at a greater distance from trees, while cover of Pleurozium schreberi and Dicranum polysetum increased near tree trunks (Fig. 3).Cover of Hylocomium splendens did not have a significant relationship with the distance from trees.Ptilium crista-castrensis grew better in plots with a thicker decay horizon.At the same time, pH of the decay horizon had no influence on the cover of the considered bryophyte species.Of bryophytes, only Hylocomium splendens was dependent on nanotopography.According to the Tukey HSD test, Hylocomium splendens was more abundant on little humps compared to concavities.
Of vascular plants, the distance to the nearest tree trunks had a significant negative effect on the cover values of Vaccinium myrtillus and a positive effect on the cover values of Luzula pilosa (Table 5).Cover of Vaccinium myrtillus depends significantly also on nanotopography, having a higher value on humps.The Tukey HSD test recorded a significant difference in the cover value in concavities and on humps.

Discussion
The structure of vegetation is determined by the synergetic effect of several biotic and abiotic factors.The most evident result of the current analysis is that, among the studied environmental parameters, the tree layer has the strongest influence on the distribution of field and ground layer species.This is in good accordance with the well-known theory of ʻsingle-tree influence cir-clesʼ (Zinke 1962) or ʻphytogenic fieldʼ (Uranov 1965(Uranov , 1968)), according to which all plant specimen, particularly big and dominating trees, affect their closest surroundings and change the environmental conditions there.Every tree has an influence circle approximately proportional to the size of the crown projection on the soil surface; outside the crown projection the influence decreases rather quickly.It has been shown that under the crowns of trees amount and chemical composition of throughfall, temperature, frost heaving, light flux, and several soil chemical properties (Beatty 1984, Boettcher andKalisz 1990), as well as soil microbial biomass and activity (Saetre 1999) can be rather different in comparison to gap areas.Therefore, in boreal forests the influence circles of trees may be superimposed on the environmental pattern, and they can have privilege in the forming of the understorey vegetation structure (Abolin 1974, Korchagin 1976, Beatty 1984, Boettcher and Kalisz 1990).Obviously, this is true only in the case of comparatively homogeneous microtopography.If the latter is more heterogenous, rather remarkable differences can be discovered between the ecological conditions of different microsites, and these factors may prove to be decisive for the formation of the horisontal pattern of the undergrowth (e.g.Kuusipalo 1984, Økland and Bendiksen 1985, Maslov 1989, Økland 1994, Frego and Carleton 1995).
Our results indicate a rather strong influence of the tree canopy on ground layer vegetation, which is in good accordance with other related investigations (e.g.Havas andKubin 1983, Økland et al. 1999 etc.).Both the species-richness and cover values of dominant species of the ground and field layers are influenced by the vicinity of trees.The number of vascular plant species is higher near tree trunks, whereas more bryophyte species occur farther from trees.The cover of the majority of dominant bryophyte species is negatively affected by the vicinity of trees.Of vascular plants, only the cover of Vaccinium myrtillus is significantly affected by trees.Phanerogams are in general larger and better adapted to microenvironment variation, whereas bryophytes, due to the absence of roots, a well developed vascular system and cuticle, are more sensitive to environment alternations (Lee andLa Roi 1979, Herben 1987).As shown in previous studies, the throughfall of precipitation is the most important source of moisture for bryophytes (Tamm 1953, Beier et al, 1993, Weibull, 2000).Since the throughfall of precipitation is higher near the crown perimeter, this explains the higher cover values of bryophytes at a greater distance from tree trunks.Kotov (1982Kotov ( , 1983) ) has asserted that if the structure of the tree layer essentially affects the  (Herben 1987).As it appears from our results, the correlation between the cover of bryophyte and vascular plant species is rather weak on a species level.Only Ptilium crista-castrensis has a stronger relationship with field layer species, while the absolute value of correlation coefficients between other bryophytes and vascular plants remains under 0.2 (Table 2).Thus, we can conclude that on the decimetre scale the patterns of the field-and bryophyte layers do not coincide.The modest negative effect of the total cover of vascular plants on dominant bryophyte species can be explained by the sharing of growth space.Competition between vascular plants and bryophytes is considered less important (Carleton 1990, Økland andEilertsen 1996).For example, McCune and Antos (1981) and Diekmann (1994)  Besides the effect of trees and total cover of lower layers, the pH and thickness of the decay horizon, or nanotopography could be important in the formation of the richness and spatial pattern of ground and field layer species.Although Cox and Larson (1993) have argued that, due to the narrow range of changes within a habitat, pH does not have a significant effect on the vegetation composition on a community level, many authors have considered soil reaction to be one of the most important ecological factor influencing the distribution of vascular and bryophyte species (Robinson et al. 1989, Pärtel 2000, Virtanen et al. 2000).Our results show that the pH of the decay horizon at the decimeter scale has a significant effect on the bryophyte species richness: higher pH increases the species number.
The contrasting effect of total cover of the bryophyte layer and cover values of dominating species on bryophyte species-richness became evident from our results.Contrary to expectations, bryophyte species-richness is negatively influenced by the total cover of bryophytes, and positively, by cover of single dominant species.Treating the cover of bryophytes as an indicator of biomass, then, analogously to the biomassspecies-richness ʻhump-backʼ curve relation (Grime 1979), we could expect the decrease of bryophyte species-richness with an increase in the total cover of the ground layer.The cover of any single species was not found to constitute as much as half of the total cover (Table 1), and probably the increase in abundance of a single species promotes the species diversity of the layer.It has been shown that different factors may cause lower species-richness at higher cover levels, competition between species could be one of them (Grime 1979).Still, competition among bryophytes has been considered unimportant (Økland 1994) and it has been connected more with space than with resources (e.g.During 1990).If the role of competition is considered to be important in the formation of the species-richness pattern, we can speculate that, at a higher level of total cover, competition between bryophytes may reduce ground layer species richness.But, in order to verify this statement, detailed investigation of bryophyte competition is needed.
In conclusion, although distance to nearest tree has a remarkable effect on the distribution of ground and field layer species, others environmental factors are also significant predictors of the structure of the lower layers.Our results showed that there are different rules for the formation of spatial patterns and structure in different understorey layers in boreal forest.To maintain also small-scale species-richness on the forest floor, both plant-related, as well as different environmental factors should be considered in forest planning and management.

Fig. 1 .
Fig. 1.Effect of significant environmental factors on the predicted values of ground layer species richness.

Fig. 2 .
Fig. 2. Effect of significant environmental factors on the predicted values of field layer species richness.

Fig. 3 .
Fig. 3. Influence of the distance to the nearest tree trunk on the expected values of dominant ground and field layer species cover.

Table 1 .
Relative frequency (Freq) of ground and field layer species and their mean cover values (Cov) in studied plots.

Table 2 .
Spearmanʼs rank correlation coefficients between the cover values of bryophyte and vascular plant species.Speciesʼ name abbreviations as in Table1.With * marked correlation coefficients are significant at p ≤ 0.01.

Table 3 .
Influence of environmental factors to the field and ground layer species richness according to general linear model.E-model parameter estimates; F -F-criterion.Notations for environmental factors: pH -pH of decay horizon; Distance -distance to the nearest tree; Nanotopography -nanotopography with three levels: concavity, plain, and hump; Thickness -thickness of decay horizon; Vas cover -total cover of vascular plants in plot; Vas no -number of vascular plant species in plot; Bry cover -total cover of bryophytes in plot; Bry no -no of bryophyte species in plot.Abbreviations of species included into model are given in Table1.The values marked by * are considered significant at the level p ≤ 0.0l.

Table 4 .
The effect of environmental factors to the cover values of dominant ground layer species according to general linear model.Notations and abbreviations for species names, environmental factors are given in Table1and 3 respectively.Abbreviations for model parameter estimate and F-criterion as in Table3.The values marked by * are considered significant at the level p ≤ 0.0l.

Table 5 .
The results of the effect of different environmental factors to the cover values of dominant field layer species according to general linear model.Notations and abbreviations for species names and environmental factors are given in Table1 and 3respectively.Abbreviations for model parameter estimate and F-criterion as in Table3.The values marked by * are considered significant at the level p ≤ 0.0l.