Greek fir (
Fir (
The phytosociological research of Greek fir forests started with
Although Greek fir can be found on soils originating from different bedrock including gneiss, serpentine, flysch, schist, limestone and dolomite, the species shows no association with a specific soil type (
Drought is a complex environmental factor having both climatic and soil components. Drought-related variables have been successfully used to explain the distribution of vegetation types on local (
To test such hypothesis, we chose a representative mountain range in central Greece covered by extensive fir forests, to analyze and quantify the influence of drought on their floristic differentiation. Our objectives were:
to describe the Greek fir plant communities found in the study area;
to identify the components of drought that determine the floristic gradients in the Greek fir forests and;
to quantify the influence of drought by determining a drought threshold among plant communities.
Because tree species composition of a forest is more likely to be influenced by human activities, we analyzed the ground vegetation composition. It has been demonstrated that ground vegetation can serve as an indicator of site conditions (
The study area is located in the Oxia-North Vardousia mountain system in south-central Greece (Sterea Ellas). The area is characterized by three mountain ranges (Kokalia, Kokkinias and North Vardousia) reaching 1923 m a.s.l. (summit of Saradena). The ranges lie between 38° 39′ 36″ and 38° 57′ 0″ N of latitude and 21° 50′ 24″ and 22° 9′ 0″ E of longitude, and cover an area of 414 km2 (
The most abundant forest type in the region is the montane Greek fir coniferous forest, which covers 44.6% of the study area (
The dominant bedrock on the Oxia-North Vardousia mountain system is flysch, while Jurassic limestones and scree slope (or terrace) deposits occur to a small extent (
There are two basic concepts of drought: the first considers it as a temporary climatic aberration (meteorological anomaly), characterized by a “prolonged and abnormal moisture deficiency” (
Three different approaches for drought quantification were followed in this study (
the classical approach using simple climatic variables: the main climatic factors related to drought,
the second approach using an index obtained through the combination of two or more climatic factors related to drought: we used a modified version of the Transeau’s humidity index (HI -
the third approach (called “site water balance approach”) considers also the soil components of drought, along with the above climatic factors: in this study both reference actual evapotranspiration (AETref) and water deficit (D) were included in the water balance; water deficit is defined as the difference between PETref and AETref (
To calculate the monthly mean of daily Rs, the “r.sun” model (
The methods used here for drought quantification took into account the spatial variation of fine-scale physiographic features such as elevation, slope, exposition and topographic shadowing effects. Any changes of these features within a distance of less than 100 m were considered as “micro-scale” changes, whereas those greater than 100 m were considered as “local-scale”.
The humidity index (HI) was used to spatially quantify drought and to capture the whole range of drought intensity over the study area. The whole area was classified and mapped into four strata based on equal intervals of HI during the vegetation period (HI: 0.28-0.39, 0.39-0.51, 0.51-0.62, 0.62-0.74). The underlying assumption was that the four strata represent forest site conditions, each with a different drought intensity. Based on the map of HI intervals obtained (
In each plot, the species composition (including vascular plants and mosses growing on mineral soil or humus) was recorded. Structural information was obtained by assessing four layers of vegetation cover (moss, herb, shrub and tree). The moss layer included only mosses, while the herb layer included herbaceous and woody species up to 50 cm in height. The woody species between 50 cm and 5 m tall were included in the shrub layer, and those above 5 m in the tree layer. For all species and layers, the cover-abundance was estimated using the extended (9-point) Braun-Blanquet scale (
In each of the 45 sampling plots a pit was dug to expose the soil profile up to 1 m in depth or until the parent material was reached. The rooting depth was measured and the soil skeleton content was estimated. Undisturbed soil samples were taken using soil sample rings of 100 and 163 cm3, and analyzed for the determination of the soil water retention curve (
The vegetation data was classified based on the floristic composition and species cover values. A hierarchical agglomerative cluster analysis based on Bray-Curtis distance (
For the identification of the diagnostic taxa of vegetation units, fidelity values were calculated for each species using a modification of the Φ index (phi coefficient of association -
Non-Metric Multidimensional Scaling (NMDS -
To determine the influence of each drought related variable on vegetation patterns, recursive partitioning based on species composition was used to derive classification trees for the categorical response variable (
The classification of the vegetation plots analyzed revealed two distinct forest communities (indicated as A and B in
General appearance: the community is comprised of pure stands of
Distribution: this community occurs on Mt. Oxia and Mt. Vardousia in the meso-mediterranean and the lower part of the supra-mediterranean zones between 690 and 1360 m a.s.l. It covers the upper, middle and lower sections of gentle to steep slopes (25-90%) in all expositions (
Sub-communities: two sub-communities could be distinguished (A1 and A2 -
General appearance: this community is made up of pure
Distribution: this community can be found on Mt. Oxia and Mt. Vardousia at higher altitude (1130-1580 m a.s.l.) as compared with community A (
Sub-communities: two sub-communities may be distinguished (B1 and B2 -
Multivariate analysis on floristic data allowed to identify two main vegetation gradients in the study area. The first floristic gradient was depicted by the horizontal axis (NMDS1) of the ordination diagram displayed in
The underlying ecological gradients were detected by fitting the different environmental variables onto the ordination scores (
Differences in topography (exposition and inclination -
Recursive partitioning based on the HI during the driest period (HI.dry) revealed the existence of two groups of plots (
The
The
In contrast to the aforementioned authors (
A syntaxonomic synopsis for the communities analyzed in this study is given below:
Class:
Order:
Alliance:
1. Community:
Sub-community: with
Sub-community: with
2. Community:
Sub-community: with
Sub-community: with
The spatial scale of the analysis should be taken into consideration for any ecological interpretation. Indeed, it is important to distinguish the environmental factors with relevance only at particular scales of analysis from other variables showing an explanatory power across all scales of analysis (
At the local scale, the floristic differentiation of Greek fir forest communities depends largely on climate. Elevation was the main factor affecting the regional climatic water balance, thus determining the patterns of floristic composition within the Greek fir forest. Elevation does not have any direct ecological or physiological impact on species (
The HI and other related moisture indexes have been repeatedly used to express the climatic water balance and explain the vegetation distribution ranging from regional (
A principal differentiation of the Greek fir forest vegetation into mesophytic and xerophytic plant communities could be found in our study area.
At the micro-scale, the floristic differentiation depends on micro-environmental interactions (micro-climate and soil conditions), since the local water balance is further affected by site-specific variations in topography (exposition, inclination) and available soil water storage capacity (ASWSC). The division of the two main
Although historic records can be very difficult to find and interpret, such information would help to clarify the interactions of drought with grazing, logging and fire and their subsequent impact on vegetation patterns. The differences in the tree layer cover between the two sub-communities A1 and A2 of the
The floristic variation within the Greek fir forest vegetation of the study area reflects a principal differentiation between mesophytic and xerophytic forest communities. This pattern, mainly driven by climatic factors related to elevation, appears on a local scale. On the Oxia-North Vardousia mountain system the wide altitudinal distribution of Greek fir follows a drought gradient influenced by an increase in precipitation and a decrease in potential evapotranspiration. Greek fir follows a similar elevation gradient in all mountains of southern and central Greece. This suggests that the pattern from mesophytic to xerophytic plant communities is likely typical of all Greek fir forests.
The modified Transeau’s humidity index calculated over the driest season appears to be the most suitable variable for the quantification of drought and the climatic water balance on a local scale in central Greece. It is also suitable to delineate the distribution threshold between the mesophytic and xerophytic Greek fir forest communities and to predict their occurrence.
The field work was partly funded by the Deutscher Akademischer Austausch Dienst (DAAD) through the IKYDA-programm. We are grateful to Bernd Künemund, Rodrigo Vargas, Cristabel Durán, Osvaldo Vidal and Carl Höcke for their help in the field. Special thanks go to Nikos Alexandris for helping with the climatic analysis, to Günter Gottschlich for verifying or identifying the specimens of the genus
Map of the study area. Coordinate reference system: WGS 84.
Stratification of the study area based on HI.veg values (humidity index of the vegetation period). Each stratum is indicated with a different color. The black dots indicate the locations of the 45 sampling plots.
Boxplots of the environmental variables in the different vegetation units. Median, mid-spread and range of variables are displayed. For the mid-spread (interquartile range) the 25% and 75% quartiles were used. Whiskers extend up to 1.5 times the interquartile range. Outliers are indicated with small circles. Rockiness (f) is the cover of exposed rocks and stones. The exposition to the north (c) is dimensionless, ranging from -1 (south) to 1 (north).
Distribution of the vegetation units along the two axes (NMDS1, NMDS2) of the ordination (NMDS). The different symbols indicate site groups (vegetation units) formed by the cluster analysis. The envelopes comprise the two main plant communities A and B. The stress for the solution with two axes is equal to 13.6.
Projection of the environmental variables (as vectors) on the ordination axes (NMDS). The direction and strength of the gradients is represented by the direction and length of the vectors respectively. For the abbreviations of the environmental variables and their coefficients of determination see
Classification tree for the two fir forest communities (A:
Climatic parameters and climatic-soil components of water balance used in the ordination (as fitted environmental vectors) and the recursive partitioning analysis (as explanatory variables). The variables were calculated for different time periods: monthly, driest period (the four driest months of the year - June-September), vegetation period (beginning of April until end of October), growing season (daily mean temperature remains above 6 °C), and annual period.
Group | Variable | Abbreviation | Units |
---|---|---|---|
Climatic parameters | Mean air temperature | Tav | °C |
Minimum air temperature | Tmin | °C | |
Maximum air temperature | Tmax | °C | |
Precipitation | P | mm | |
Global (total) radiation | Rs | Mj/m2 | |
Reference potential evapotranspiration | PETref | mm | |
Humidity index | HI | dimensionless | |
Climatic-soil components of water balance | Reference actual evapotranspiration | AETref | mm |
Water deficit | D | mm |
Topographic, structural and environmental/abiotic conditions (mean ± standard deviation) of each vegetation unit.
Variable | Community A | Sub-comm.A1 | Sub-comm.A2 | CommunityB | Sub-comm.B1 | Sub-comm.B2 |
---|---|---|---|---|---|---|
Elevation (m a.s.l) | 973 ± 154 | 928 ± 127 | 1005 ± 168 | 1375 ± 143 | 1401 ± 152 | 1303 ± 85 |
Inclination (%) | 52 ± 16 | 45 ± 17 | 56 ± 14 | 46 ± 17 | 50 ± 15 | 33 ± 19 |
Cover of tree layer (%) | 65 ± 10 | 71 ± 11 | 61 ± 8 | 67 ± 14 | 65 ± 16 | 73 ± 6 |
Cover of shrub layer (%) | 19 ± 13 | 27 ± 14 | 13 ± 7 | 11 ± 10 | 12 ± 11 | 7 ± 3 |
Cover of herb layer (%) | 31 ± 24 | 44 ± 30 | 22 ± 14 | 34 ± 22 | 31 ± 23 | 42 ± 18 |
Cover of moss layer (%) | 29 ± 28 | 28 ± 31 | 31 ± 27 | 8 ± 7 | 8 ± 7 | 7 ± 9 |
Total vegetation cover (%) | 80 ± 10 | 89 ± 8 | 74 ± 7 | 81 ± 9 | 80 ± 9 | 85 ± 5 |
Cover of litter (%) | 28 ± 23 | 37 ± 25 | 21 ± 20 | 43 ± 29 | 33 ± 22 | 72 ± 29 |
Cover of exposed rocks and stones (%) | 20 ± 21 | 23 ± 22 | 18 ± 22 | 22 ± 18 | 28 ± 17 | 4 ± 1 |
Height of highest trees (m) | 18 ± 3 | 20 ± 3 | 17 ± 3 | 21 ± 4 | 21 ± 4 | 22 ± 6 |
Diameter of highest trees (cm) | 45 ± 10 | 42 ± 8 | 47 ± 10 | 52 ± 23 | 51 ± 20 | 56 ± 32 |
Humus depth (cm) | 3 ± 2 | 3 ± 2 | 4 ± 2 | 4 ± 3 | 4 ± 3 | 2 ± 1 |
Soil depth (cm) | 66 ± 31 | 79 ± 33 | 57 ± 28 | 75 ± 27 | 78 ± 26 | 67 ± 32 |
ASWSC (mm) | 92 ± 65 | 119 ± 76 | 73 ± 52 | 128 ± 66 | 122 ± 56 | 146 ± 93 |
Synoptic table of the fir forest vegetation units (communities and sub-communities) in the study area based on cluster analysis (flexible beta with β = -0.25, Bray-Curtis distance). Frequency values of taxa are displayed. Only the diagnostic taxa are presented with their fidelity values (Φ) and their statistical significances. (*): p < 0.05, (**): p < 0.01, (***): p < 0.001. The frequency values of taxa considered as diagnostic for the vegetation units are reported in italic. Woody species occurring in different layers are referred to with a letter (t = tree; s = shrub; h = herb). Mosses are referred to with the letter m. Communities and sub-communities: (A):
Comm.(subcomm.) | Species | Form | A (n=22) | B (n=23) | Φ | ||
---|---|---|---|---|---|---|---|
A1 (n=9) | A2 (n=13) | B1 (n=17) | B2 (n=6) | ||||
(A) |
|
- | 67 | 92 | 6 | 0 | 0.778*** |
|
t | 44 | 23 | 0 | 0 | 0.451* | |
|
s | 56 | 54 | 6 | 0 | 0.571** | |
|
h | 100 | 100 | 24 | 33 | 0.746*** | |
s | 78 | 92 | 24 | 0 | 0.733*** | ||
m | 100 | 92 | 65 | 33 | 0.528** | ||
- | 67 | 85 | 29 | 17 | 0.526** | ||
|
m | 67 | 92 | 41 | 17 | 0.507** | |
|
- | 44 | 46 | 6 | 0 | 0.495* | |
- | 56 | 62 | 18 | 17 | 0.427* | ||
(A1) |
|
h | 67 | 0 | 0 | 0 | 0.775*** |
- | 56 | 0 | 0 | 0 | 0.696** | ||
|
s | 44 | 0 | 0 | 0 | 0.612** | |
s | 33 | 0 | 0 | 0 | 0.522** | ||
h | 56 | 15 | 0 | 0 | 0.572** | ||
- | 56 | 0 | 0 | 17 | 0.563** | ||
|
- | 33 | 0 | 0 | 0 | 0.522* | |
|
- | 44 | 15 | 0 | 0 | 0.477* | |
|
- | 33 | 0 | 6 | 0 | 0.457* | |
|
s | 22 | 0 | 0 | 0 | 0.42* | |
|
h | 33 | 8 | 0 | 0 | 0.439* | |
|
s | 56 | 15 | 6 | 17 | 0.439* | |
- | 22 | 0 | 0 | 0 | 0.42* | ||
- | 22 | 0 | 0 | 0 | 0.42* | ||
|
- | 22 | 0 | 0 | 0 | 0.42* | |
(A2) |
|
- | 22 | 100 | 0 | 0 | 0.87*** |
- | 11 | 85 | 0 | 0 | 0.821*** | ||
- | 0 | 85 | 12 | 0 | 0.817*** | ||
|
- | 0 | 69 | 0 | 0 | 0.792*** | |
|
- | 0 | 69 | 6 | 0 | 0.746*** | |
h | 11 | 69 | 0 | 0 | 0.708*** | ||
|
- | 44 | 100 | 29 | 0 | 0.658*** | |
- | 33 | 100 | 47 | 0 | 0.637*** | ||
|
- | 0 | 62 | 18 | 0 | 0.605** | |
|
- | 0 | 54 | 12 | 0 | 0.584** | |
|
- | 22 | 77 | 24 | 0 | 0.579** | |
|
- | 0 | 38 | 0 | 0 | 0.565** | |
|
- | 0 | 46 | 12 | 0 | 0.52* | |
- | 0 | 31 | 0 | 0 | 0.5* | ||
|
- | 0 | 31 | 0 | 0 | 0.5* | |
|
- | 0 | 31 | 0 | 0 | 0.5* | |
|
- | 0 | 31 | 0 | 0 | 0.5* | |
|
- | 0 | 31 | 0 | 0 | 0.5* | |
|
m | 0 | 31 | 0 | 0 | 0.5* | |
|
- | 33 | 69 | 18 | 0 | 0.493* | |
- | 22 | 62 | 18 | 0 | 0.48* | ||
|
- | 33 | 69 | 24 | 0 | 0.469* | |
|
- | 11 | 46 | 12 | 0 | 0.442* | |
(B) |
|
- | 0 | 23 | 88 | 67 | 0.663*** |
- | 11 | 0 | 88 | 50 | 0.657*** | ||
|
- | 33 | 8 | 88 | 83 | 0.654*** | |
|
- | 11 | 0 | 47 | 83 | 0.624*** | |
|
- | 0 | 0 | 53 | 50 | 0.589** | |
|
- | 11 | 15 | 71 | 67 | 0.563** | |
|
- | 56 | 46 | 94 | 100 | 0.526** | |
- | 44 | 38 | 82 | 100 | 0.526** | ||
|
- | 0 | 0 | 35 | 50 | 0.521* | |
|
- | 22 | 15 | 76 | 50 | 0.452* | |
(B1) | - | 0 | 8 | 82 | 0 | 0.827*** | |
- | 0 | 15 | 59 | 0 | 0.598** | ||
- | 0 | 0 | 35 | 0 | 0.539** | ||
- | 56 | 31 | 100 | 50 | 0.48* | ||
|
- | 11 | 0 | 47 | 17 | 0.42* | |
(B2) | - | 22 | 8 | 12 | 100 | 0.78*** | |
|
h | 11 | 0 | 12 | 83 | 0.742*** | |
|
- | 0 | 0 | 6 | 50 | 0.6** | |
|
- | 0 | 0 | 29 | 67 | 0.576*** | |
|
s | 0 | 0 | 0 | 33 | 0.522* | |
|
- | 0 | 0 | 0 | 33 | 0.522* | |
- | 0 | 0 | 0 | 33 | 0.522* | ||
- | 0 | 0 | 18 | 50 | 0.51** | ||
|
- | 22 | 0 | 24 | 67 | 0.495* | |
(C) |
|
- | 100 | 77 | 53 | 0 | 0.671*** |
- | 78 | 100 | 82 | 17 | 0.657*** | ||
|
- | 44 | 77 | 82 | 0 | 0.588** | |
|
- | 56 | 85 | 59 | 0 | 0.574** | |
h | 56 | 85 | 41 | 0 | 0.526** | ||
- | 56 | 77 | 41 | 0 | 0.506* | ||
|
- | 33 | 100 | 82 | 0 | 0.747*** | |
|
- | 22 | 77 | 71 | 0 | 0.634*** | |
- | 0 | 46 | 29 | 0 | 0.483* | ||
- | 0 | 46 | 29 | 0 | 0.483* | ||
- | 0 | 31 | 35 | 0 | 0.445* | ||
|
- | 0 | 38 | 24 | 0 | 0.428* |
Relationships between the species ordination’s scores (NMDS) and the fitted environmental variables. Coefficients of determination (r2) and their significances assessed using 1000 random permutations. Only variables with significant coefficients are shown. (*): p < 0.05; (**): p < 0.01; (***): p < 0.001.
Variables | Abbreviations | r2 |
---|---|---|
Elevation (m a.s.l.) | ALTITUDE | 0.843 *** |
Inclination (%) | INCLINE | 0.205 * |
Exposition to the north | EXPO_N | 0.487 *** |
Cover of tree layer (%) | COV_TREES | 0.394 *** |
Available soil water storage capacity (mm) | ASWSC | 0.215 ** |
Solar radiation during growing period (Mj/m2) | Rs_growth | 0.522 *** |
Maximum air temperature during driest period (°C) | Tmax_dry | 0.857 *** |
Mean air temperature during vegetation period (°C) | Tav_veg | 0.856 *** |
Precipitation during vegetation period (mm) | P_veg | 0.852 *** |
Reference actual evapotranspiration during driest period (mm) | AET_dry | 0.462 *** |
Reference potential evapotranspiration during growing period (mm) | PET_growth | 0.753 *** |
Humidity index during driest period | HI_dry | 0.824 *** |
Humidity index during vegetation period | HI_veg | 0.839 *** |
Water deficit during driest period (mm) | D_dry | 0.690 *** |
Water deficit during vegetation period (mm) | D_veg | 0.696 *** |
Environmental variables fitted onto the ordination as vectors or factors.
Vegetation table of the Greek fir forest communities.