Research on Lca and Forest Structure

research on LCA and forest Structure

in Kastania Pierias (greece)

Introduction

The study described here is an investigation into forest structure parameters of a typical oak coppice forest (“Greek Quercus frainneto woods”) as defined by directive 92/43/EU (cod. Corine 41.B or 9280 NATURA 2000). The work forms part of a larger project, concerned with the application of LCA to assess the environmental impact of alternative forest management systems. The research is funded by the National Agricultural Research Foundation of Greece and the British Council and is being carried out in a private forest called “Notio” (South) in the area of Kastania, municipality of Kolindros, prefecture of Pieria (GREECE).

Two Life Cycle Assessment (LCA) groups; at the Forest Research Institute (Greece) and at ImperialCollege (Great Britain), have co-operated on forest research for the last 2 years. To date, the principles of LCA have been most often applied to forestry in north European countries, where forest production is focused on wood products. There are few studies applying LCA to forestry in Mediterranean countries, where various services provided by forests can be more important than timber products. This report is part of a joint project, applying the principles of LCA to the management of a typical Mediterranean coppice forest in northern Greece. Data collection is being conducted principally by the Greek researchers in collaboration with the Greek forest owner and the Forest Service of Pieria, which supervise the forest (Photo collection in Annex 1). The British group is supporting the programme with models and sharing of experience and knowledge.

It has always been understood and clearly stated that LCA is an iterative process. The scope, boundaries and even the goals of a study frequently have to be refined based upon information that only comes to light later in the process. Inventory data must frequently be corrected, expanded or added during the impact assessment or interpretation. In the case of an LCA with the explicit goal of addressing forestry land-use and other management-related issues, this process may be even more fundamental.

As has been clearly identified in the work of WG2 of COST E9, the setting of boundaries, the application of appropriate process models and the assessment of end-points for land-use impacts are all intimately related. Furthermore, appropriate models are largely missing or not yet identified for application in LCA. In this project, therefore, whilst we have a clear, predefined aim, we are in the unusual position, for most scientific exercises, of having neither a fixed set of end points, nor a pre-set methodology for arriving at them. This position is, in fact, not unusual for Life Cycle Assessments, and quite common for LCA of forestry.

In order to deal logically with these problems, our approach has been to begin in the middle, and work forwards and backwards from there. Specifically, we have begun by seeking to establish what inventory measures and models are available, appropriate and can be implemented for the forestry system in question, and the major environmental concerns identified for it. The goals achievable, boundaries required and impacts included are being determined iteratively by what is achievable.

The functional unit of the LCA is the maintenance and management of an area of productive forest. The management variables to be assessed through scenario testing include: varying the period between cutting, the area of uncut (nurse) stands and the number of uncut (standard) trees in cut areas - within and between site classes.

The parameters to be used to investigate the effects of these changes include both directly measured and modelled effects. The modelled quantities, to be described in later papers, include: nutrient flows and pools, bulk soil movements, the movement of carbon between compartments and the effects of removals through grazing.

This report describes part of the fieldwork component of the inventory process - the development of measures of forest structure and quality. As data are accumulated, and in parallel with simultaneous research into the availability of appropriate environmental models and data, they are assessed for potential relevance and value in the possible modelling approaches.

This kind of scientific 'fishing expedition' is highly unusual, and fraught with difficulties. It is believed, however, that the unusual nature of the problem, to define and then include these issues in LCA, requires and justifies a novel approach.

Site selection

The specific forest type was selected because:

• more than 25% of Greek forests consist principally of oaks and especially the broad leaf Hungarian oak (Ντάφης, 1966),
• the local landscape is dominated by this forest type
• soil protection, biodiversity and watershed are modulated by them
• the economy of nearby settlements depends upon products of this forest type (e.g. goats, honey, fuelwood, charcoal, timber, etc.)

The specific research forest area was selected for three principal reasons:

• it has been managed continuously under the same system for several decades, meeting the requirements of Greek forest legislation
• the borders of the area are well-defined, and all activities can be divided by administrative and land planning units.
• Levels of cooperation in the study area are high.

Parameter selection

It has been reported that biodiversity is higher in multiple-use forests than in single use, and that the specific management system also has a significant effect ( Salwasser, 1998; Beese, 1998; Romane et.al.,1998; Buriánek, 1998; Wikstróm, 1998). Biodiversity evaluation is usually based on a combination of: structure indicators, plant richness and function indicators (Larson Tor-Bjoern et.al., 2001) that may be applied at the national, regional or stand level.

In this study the evaluation of the status and impacts of coppice forest, and of a range of management systems, is based on a number of quantitative parameters:

• area remaining unharvested (mother stands)
• number and status of standards remaining in harvested areas
• percentage of bare soil
• density and mortality of coppice stools in regenerating areas
• number and density of sprouts, root-suckers and seedlings

• biodiversity measured as species (vascular plant) richness.

Key qualitative parameters related to these indicators are:

• soil protection by means of plant coverage
• presence of protected key-species
• degree of natural regeneration
• quality of regeneration (growth characteristics)

Methods

Site classification

The current sub-division system of the forest in terms of management units is based on physical and technical borders such as forest roads. This system was established as part of the first management plans for the forest, produced in the 1950's.

The forest as a whole (1,917 ha) is divided into 9 sections and 24 sub-sections (stands). Within individual geographic divisions there is a range of ecological types, of site capacities (in terms of timber productivity) and of production processes used.

The current division system makes the analysis of ecological parameters and the application of ecologically based management very difficult.

For this study we have therefore re-classified and divided the forest area according to ecological principles, based upon research by the Greek Ministry of Agriculture - Forest Commission. In particular, the soil maps in the ‘Kolindros’ series: 'Land resource and land capability for forestry'; and Forestry Commission laboratory analysis data.

Based upon these data, we have replaced the old division system with three new site types (or classes) called, S1, S2 and S3. The first new site class (S1) covers 321.4 ha (24,22%) and represents lower slopes and valley bottoms with low relief and deep soils. The second site class (S2) covers 873.1ha (65,79%) and covers slopes and ridges. The third site class (S3) is 132.7 ha (10%) and includes degraded forest areas with difficult site conditions around the local village. It includes agricultural land, part of which has been rather unsuccessfully reforested as a plantation, and is characterised by very low productivity.

Sampling

A stratified random sampling method (Barbour, G.M. et.al.,1998) was used to establish three experimental plots in the highest quality site-type (S1) and three more in the intermediate quality site-type (S2)(Map 1). S3 sites were not included at this stage.

These forest sites range from 1 to 4 years since last cutting. The main characteristics of the experimental plots are shown in the following table.

Table.1. Main characteristics of the experimental plots established at the private forest in Kastania Pierias (GREECE)

Experimental plots / Altitude
(m.) / Aspect / Year of cutting / Site type (S1,S2)
1 / 380 / Flat / 1998 / S2
2 / 360 / East / 1997 / S2
3 / 240 / South-East / 1997 / S1
4 / 260 / East / 1999 / S1
5 / 260 / Flat / 2001 / S1
6 / 320 / East / 1999 / S2

Four square (10m X 10m) sampling sub-plots were established in each main rectangular (40m X 25m) experimental plot. Data of standards were recorded in each experimental plot and inventory data of natural regeneration (sprouts and ground flora) were selected within each sub-plot.

Scheme 1. Plot (40M Χ 25m) and subplots (10Χ10m2) of the experimental area

The evaluation of forest structure is based on parameters including:

• the shape of standards,
• the ratio of living to dead stools
• the composition and status of natural regeneration.

The IUFRO method, modified for coppice forest use, was also used for the evaluation of standards in cut areas (Leibundgut, 1984). Evaluation for the shape of standards used three different classes: A (good), B (moderate) and C (bad). Additionally, the evaluation of cut areas included measurement of stool density, mortality (number of dead sprouts), maximum height and dominant height (average of three highest sprouts).

Evaluation of the protective function of the forest is based on:

• proportion of bare soil
• percentage of general coverage and distribution among other categories (regeneration, ground flora, litter, dried branches)
• Frequency of “key” species (potential and existing)

These potential key species, listed in Table 2, are protected according to international and/or local directives (red lists, native and rare taxa etc.). Sampling is also to be carried out in stands of both site-types (S1, S2), 20 years after last cutting.

Table 2. Taxa proposed (potential) for use as “key” species in Hungarian Oak forest areas (Theodoropoulos, 1991; Petermann J, 1999)

PLANT (KEY) SPECIES / PROTECTION CRITERIA / REFERENCE
Anacamptis pyramidalis / CITES / Theodoropoulos K.
Anthemis sibthorpii / Native / Theodoropoulos K.
Asperula purpurea subsp. apiculata / Other (criterion) / Theodoropoulos K.
Campanula spatulata / Native / Petermann J.
Cephalanthera longifolia / CITES / Theodoropoulos K. & Petermann J.
Crocus veluchensis / Other (criterion) / Petermann J.
Dactylorhiza sulphurea subsp. pseudosamkucina / CITES / Theodoropoulos K.
Digitalis ferruginea / Other (criterion) / Petermann J.
Digitalis viridiflora / Other (criterion) / Theodoropoulos K.
Epipactis helleborine / CITES / Theodoropoulos K.
Fritillaria pontica / ΠΔ / Theodoropoulos K.
Galium laconicum / Other (criterion) / Petermann J.
Ilex aquifolium / Other (criterion) / Theodoropoulos K.
Neottia nidus-avis / CITES / Theodoropoulos K.
Orchis laxiflora subsp. palustris / Other (criterion) / Theodoropoulos K.
Orchis mascula subsp. mascula / CITES / Theodoropoulos K.
Platanthera bifolia / Other (criterion) / Petermann J.
Platanthera chlorantha / Other (criterion) / Theodoropoulos K.
Rosa arvensis / WCMC / Theodoropoulos K.
Trifolium pignantii / Other (criterion) / Theodoropoulos K. & Petermann J.
Valerianella locusta / Other (criterion) / Theodoropoulos K.

In the present study, biodiversity is expressed as the number of species in a stand, with the number of dead or decaying wood (sprouts) also recorded.

Figures 1 & 2: Clear cut area with standards in the background, just after cutting (summer 2001) (Left). Steep forest edges after clear cutting (Right)

Results and Discussion

Soil coverage

The results of the soil coverage assessment are shown in Fig 3. This shows that in the next year after clear-cutting (2001) it was seen bare soil cover 11% and a high percentage of litter and branches with another 35% cover. By contrast, vegetation cover averaged 75% in four years old cuts, which means rapid establishment of natural vegetation (regeneration and ground flora) and satisfactory soil protection.

Fig.3. Means of different soil coverage categories of six (6) experimental plots

Uncut (Mother) stands

Standards

The average density of standard trees in the experimental areas was sixty per hectare. Quercus frainetto (Hungarian oak) is the dominant species, comprising 75% of standards,followed by Fraxinus ornus (Ash) with 12.5%, and the remaining 12.5% is composed of a mixture of Acer campestre, Acer hyrcanum, Pirus malus, Sorbus domestica and Sorbus torminalis. Mean diameter at breast height (DBH) was 8.5cm, with a mean height of 7m.

17.5% of the standards were classified as of “Good” shape, and the mean classification according to the IUFRO classification system was 122/255, meaning a healthy over-storey tree with moderate stem, canopy and silvicultural value.

Stools

Forest regeneration produces 4,320 stools/Ha (Table 3), with Hungarian oak as the dominant species. Stool density was not found to be related to site type.

An average of 67 dead stools per hectare (1.5% of the total) were found and statistical analysis showed no relationship between stool mortality and site type. Mortality in the poorest stands was more than twice the average.

Table 3. Mean number of stools per 100m2 and site type quality.

Stand parameter / Statistical parameters / Site type (quality) / TOTAL
S1 / S2
Dead stools (Ν/100m2) / MEAN / 0,42a* / 0,92a / 0,67
Ν / 12 / 12 / 24
(stdDEV) / 0,67 / 0,79 / 0,76
Living stools (Ν/100m2) / MEAN / 46 a / 40,5 a / 43,2
Ν / 12 / 12 / 24
(stdDEV) / 16,3 / 9,2 / 13,2

* the same letter a indicate no significant statistical difference (T-test, significance level 0.05)

Secondary stand-regeneration

The number of sprouts per stool (sprouting capacity) was positively correlated with site type (Table 4). In contrast, the mean mortality of sprouts, which was 1% for the stands examined, is not related to site type. Once more our results agree with findings by other authors, for both the specific cutting season and age of stand (Chatziphilippidis 1998 ).

Table 4. Mean number of sprouts per stand by site type.

Stand
Parameter / Statistical Parameters / Site type (quality) / TOTAL
S1 / S2
Number. of sprouts/stool / MEAN / 9,6 a* / 6,7 a * / 8,0
Ν / 12 / 12 / 24
stdDEV / 3,4 / 1,7 / 8,6

* the same letter a indicate no significant statistical differences (T-test, significance level 0.05)

Regeneration is predominantly through sprouts on coppiced stools (90%) with only 10% of root suckers and seedlings. Regeneration from seedlings was 3% in the first year after clear cutting. It is not possible to distinguish between root suckers and seedlings after the first year.

Fig.4. Maximum and dominant height of sprouts (1-4 years after clear cutting)

The dominant and maximum height of sprouts during the 4 years after clear cutting is shown in Fig.4.

The mean height of sprouts (1-4 years of age) was 1.6m. For both root suckers and seedlings the mean height was 0.43m. The height of seedlings may only be separately measured during the first year after clear cutting and was 0.07m (compared with 0.2m for sprouts). This major difference in growth rate, as well as observed heavy grazing, makes their participation in the final stand population uncertain.

No correlation between height and site type was found.

Table 5. Maximum heights of sprouts (regeneration) by site type

Maximum Height
of: / Statistical Parameters / Site type (quality)
S1 / S2
Sprouts / MEAN / 1,39 / 1,80
Ν / 423 / 435
stdDEV / 0,79 / 59
Root suckers and seedlings / MEAN / 0,40 / 0,52
Ν / 92 / 42
stdDEV / 0,62 / 0,36
Seedlings / MEAN / 0,07 / -
Ν / 33 / -
stdDEV / 0,02 / -

Biodiversity

The number of different tree and shrub species was recorded for each plot. Eighteen (18) species of trees and shrubs were present in the experimental area as a whole, fifteen (15) of which were found in site type S1 and 13 in site type S2 (Table 6.). Ten species were common to both site types, with four unique to S1 sites and three found only in S2 sites.

Species density was 5 and 4 number of tree and shrubs for the site types S1 and S2 respectively.

Table 6. Species of trees and shrubs recorded in the two site types S1 and S2

Species / Site type
S1 / Site type
S2
Acer campestre / + / +
Acer hyrcanum / +
Cornus mas / + / +
Crataegus laciniata / + / +
Crataegus monogyna / + / +
Carpinus orientalis / + / +
Fraxinus ornus / + / +
Juniperus communis / +
Pirus amygdaliformis / +
Pirus malus / +
Prunus spinosa / +
Quercus petraea ssp. Medwewii / +
Quercus frainetto / + / +
Quercus pubescens / + / +
Rhus corriaria / +
Sorbus domestica / + / +
Sorbus torminalis / + / +
Ulmus minor / +
TOTAL / 15 / 13

According to the ground vegetation inventory in spring 2002 129 species (taxa) were identified. The mean number of species/100 m2 was 26 and 29 for the site type S1 and S2 respectively, whereas the total number in each site was 97.

Finally, the dead wood in the form of dead sprouts in the two site types (S1, S2) was almost similar (7,4 and 11,1%, respectively).

Conlusions

The parameters used to study the coppice oak forest, and especially the current management system, included quantitative and qualitative ones. Four criteria (3 structural, 2 biodiversity, 1 environmental and 2 protective) and 8 indicators were distinguished (Table 7). The selected method was based first on the existing experience on forest management in Greece, second on the principles for sustainable forest management in Europe (Helsinki 1993) and thirdly on the sustainable timber certification in Greece. The results are also expected to support the application of the LCA method to the management of the coppice forest of Kastania.

The study showed that the current forest management system follows well known rules, such as are the 20 years rotation time, the forest regeneration by sprouts (coppicing) etc. The applied forest management regime to date practically ensures the sustainability of forest products and the financial success of the forest owner. Also, important silvicultural results came up from the study, such as are the right cutting season, the selection of the appropriate standards etc. that may improve the forest enterprise.