Structural Diversity Results

This result group contains results that describe the Structural diversity of the trees in a treatment unit.

Variable name	Unit	Description
Deadwood-rich forest	boolean	Indicates if a treatment unit holds at least 20 m3/ha of dead wood with a diameter of at least 20 cm.
Even-aged Class	code	Even-aged type. Unknown: No information EvenAged: If at least 95% of the volume is within a 5-year range. MostlyEvenAged: If at least 80% of the volume is within a 20-year age range UnevenAged: Otherwise.
Forest with 60+ large trees	boolean	Indicates if a treatment unit holds at least 60 large trees per hectare: At least 45 cm in diameter for conifers, beech and oak. At least 35 cm in diameter for other species.
Mature broadleaf-rich forest	boolean	Indicates if a treatment unit holds forest older than 80 years in boreal regions or 60 years in boreonemoral and nemoral regions, and at least 25 % of the basal area is broadleaf.
Old forest	boolean	Indicates if a treatment unit holds forest older than 140 years in boreal regions or 120 years in Boreonemoral and nemoral regions.
Tree Size Diversity (Gini Coefficient)	0-1	The Gini coefficient is an equality index between 0 (=maximum equality, i.e. all trees have the the same size) and 1 (=maximum inequality). The index has been proposed to be used for forestry planning by Lexerød & Eid (2006).
Tree Size Diversity Class (Hugin def.)	code	Tree size diversity class according the Hugin system definition. Trees are grouped into four diameter classes, with class width = (dbhmax-dbbmin)/4. If the number of trees in classi > classi+1, the diameter class distribution is set as InverseJShaped, otherwise as Homogeneous.

Calculation of Gini coefficient in Heureka

The Gini coefficient is calculated with the formula for a discrete probability distribution (see https://en.wikipedia.org/wiki/Gini_coefficient).

First trees are sorted in ascending order so that g_i < g_{i+ 1}, where g_i = basal area of type tree i. Basal area is used to take into account that the stand volume is highly affected by the largest trees (see Lexeröd and Eid 2006).

$G=1-\frac{\sum _{i=1}^n(f(g_i))(S_{i-1}+S_i)}{S_n}$

where $S_i=\sum _{j=1}^{i}f(g_j)g_j$, and
S₀ = 0, and
f(g_i) = Frequency distribution, where g_i, i = 1..n, are the tree basal areas, indexed in increasing order (g_i< g_i+1)