Annex 6. Non-forest biomass survey

Slovenia is a biomass rich country, where the forest area is accompanied by other land uses that are often rich of woody biomass and by consistent areas of abandoned farmland that reverts to forest. While forests are studied in detail by forestry institutions and their production capacity is well known, little or no information was available concerning the amount of fuelwood produced outside forests and on the current and potential role of these areas in planning sustainable wood energy systems. On the other hand, it was understood that a significant part of fuelwood trade for household use is informal as fuelwood is either collected by farmers in their own lands and forests or bought locally.

To fill this information gap, and thus to complete the supply module, a specific survey of non-forest fuelwood sources was completed. The survey followed a two-phase approach based on (i) the new 2002 Land Use Map, used as stratification system (see table below), (ii) the available ortophotos coverage, used to estimate the crown cover of woody vegetation, and (iii) field sampling to relate crown cover to woody biomass stocking and increment.

Land use classes – Slovenia.

Land Use Map publication 2002. Approximate reference date 2000 (date of ortophotos)

Sampling scheme to assess non-forest biomass

In the first phase of the two-phase survey, a systematic sample of non-forest land units was analyzed using the digital ortophotos on which the 2002 land use map was based. The selected land units were identified on the digital ortophoto data set and interpreted to assess the average crown coverage of trees and lower woody vegetation, such as bushes and young trees.

Interpretation of cover types

The digital ortophotos offered an excellent base for the interpretation of vegetation cover and for the distinction of cover types. The national consultant and the RC defined the following cover types, in consideration of photo characteristics and of basic requirements for the subsequent field-measurement phase:

Code Cover type

1 Bushes and young trees (vegetation below 7 m height)

2 Intensive orchard

3 Extensive orchard

4 Young forest stand (up to the pole stand)

5 Middle-age forest stand (small to medium tree crown size)

6 Mature forest stand (medium to large tree crown size)

7 Individual (isolated) trees – crown area < 50 m² (diameter <8m)

8 Individual (isolated) trees – crown area > 50 m² (diameter >8m)

9 Lines of trees (e.g. roadside trees, hedges, etc.) with crown diameter < 8 m

10 Lines of trees (e.g. roadside trees, hedges, etc.) with crown diameter > 8 m

All interpretation was carried out visually on computer monitor. Each interpretation class was digitized as a closed polygon and coded according to cover type.

Cover types 1 and 4 referred to shrubs, bushes and young trees whose average height was estimated to be below 7 m. Type 1 referred to individual plants or small groups and 4 to larger formations.

The two orchard cover types referred to systematic layouts, usually composed by smaller crowns (2) and to older, irregular formations, usually composed by larger crowns (3). These two classes were found within or outside the classes defined as “orchards” in the land use map (in urban areas, for instance).

Dense stands were separated according to the size of tree crowns, into middle-age (5) and grow-up (6). A comparison with nearby forests allowed for the separation of stands that appeared younger from stands that appeared more mature.

Individual trees were divided into two main types: small to medium trees, with crown diameter below 8 m (7) and medium to large trees, with crown diameters above 8 m (8). Similarly, trees in lines were separated according to the width of the line. Where tree rows were wider, including two or more trees, they were classified as forest stands.

Other specifications

Minimum interpretation unit

The minimum interpreted crown area was 3 x 3 m (3 x 3 mm at scale 1:1000)

Sub-sample of large polygons

In case of large polygons (>100 ha) the interpretation was done on a representative fraction of the entire polygon area. The polygon was intersected with a 1 ha grid and the interpretation area was a systematic sample of the grid cells. The intensity of the systematic sample (1 on 2, 1 on 3, 1on 4, etc.) depended on the size of the polygon, with a minimum of 20 interpreted sub units (grid cells) well distributed over the polygon.

The class 3000 (Urban and built up, inclusive of roads) was practically one unique polygon over the entire Slovenia, since all urban centers are connected by thin stripes (roads). Therefore, in case of class 3000, and only in this case, the sampling unit was defined by the intersection between the class polygon and a circle of 250 m. radius around the selected 4x4 Km grid point.

Sample selection

The following table and figure show the distribution of the sample using a 4x4 Km grid (the same used by Slovenia Forestry Institute for permanent forest sample plots) and including only the classes of interest.

Summary of sample selection

Sample distribution map:

Example of ortophoto with land use class (red) and crown cover types (yellow). The land use class of the selected class polygon is 1410 (re-growth in abandoned farmland). The Crown Type classes are: 3, 4, 5, and 6

Field measurement work in phase 2:

In order to calculate the average amount of wood stock (and increment) for all the cover types described above, a subset of the 227 sampling units was randomly selected for field measurement.

Smaller polygons of individual woody cover types were measured entirely. In bigger polygons, the measurements were done in representative sub-plots of varying size, depending on the woody cover type concerned, as follows:

Sizes of sample plots for wood stock measurements in different woody cover types

The position of the plot inside the polygon was marked on the map and, in case of lines of trees, the measured part was marked.

In every land use unit, up to 5 individual trees of the most frequent woody cover types in the unit were measured.

In all woody cover types, trees and bushes from a diameter of 5 cm onwards were measured.

In order to determine the tariffs for calculation of wood stock (and also increment) in every polygon the height of three thickest trees were measured. For individual trees, the height was measured every 5 or 10 trees, specifying the diameter and species.

The calculation of the average wood stock for individual land use types (in m3/ha) was done in two modes:

Direct mode

Calculation of average wood stock value for individual non-forest land use types directly through shares of individual woody cover types inside individual land use types – on the basis of a known average wood stock (in m3/ha) of different woody cover types. Calculation was faster and simpler but it did not provide the calculation of precision or confidence interval of the average wood stock value for an individual non-forest land use type.

Indirect mode

Calculation of average wood stock value for individual non-forest land use types through calculated average wood stock values of individual land use units (from 227), which provides calculation of precision or confidence interval of the average wood stock value for an individual non-forest land use type.

The list and location of the land units selected for field measurements is given in the Table below:

Ljubljana:

Brežice:

Note: On one of plots 793 or. 917 measure two polygons of mature trees or two plots in mature tree stands (6).
On plot 793 measure 3 polygons of forest trees in lines with thick trees (10).

Celje:

Note: On one of plots 1545 or. 1672 measure two polygons of mature trees or two plots in mature tree stands (6).
On plot 1611 measure 3 polygons of forest trees in lines with thick trees (10)

Maribor:

Sežana:

Survey results

1 – CROWN COVER

Average - percent crown cover

lower veget = CT 1 and 4 tree cover = all other CT

2 - STOCKING AND INCREMENT