ArticlePDF Available

Tree Damage in Mechanized Uneven-aged Selection Cuttings

  • Ramboll Finland Oy

Abstract and Figures

The amount of selection cuttings in uneven-aged forest stands is supposed to increase in Finland with the new Forest Act. In uneven-aged management, it is estimated that the cutting could be repeated every 15–20 years with the removal of around 100 m3/ha, depending on the site type and stand growth. This interval and volume highly depend on survival of lower canopy trees in cuttings. The number of these trees (2.5–15 m in height) is typically limited. Felling larger trees from above means a high damage risk for smaller trees and also restricts the harvesting outside the heavy frost period due to top damage risk. Damage to trees taller than 2.5 meters was studied in three selection cutting stands. Mechanized harvesting (harvester–forwarder) was carried out in late winter with no frost, which is the optimal time for selection cuttings. On the average 21.5% of the remaining trees were damaged. The percentage of damage to smaller (2.5–10 m) trees was highest, 28.4%. Stem damage and breakage were the most common types of injury. A logistic mixed model was used to model the probability of tree injury (uninjured/injured). Distance from the nearest removed tree, harvested basal area within 25 m of the tree and diameter of the tree were the explanatory variables taken into the model. The model discrimination ability by the ROC curve was 72.2%. With a classification cutpoint of 0.5 for the model fitted injury probabilities, the rate of correct classification was 79.1%. There is a need to develop optimal working practices for mechanized selection cuttings. Information on the stand structure, practical operator tutoring and knowledge of the goals of the forest owner are needed for successful harvesting implementation.
Content may be subject to copyright.
Original scientic paper
Croat. j. for. eng. 36(2015)1 33
Tree Damage in Mechanized Uneven-aged
Selection Cuings
Mai Sirén, Juha Hyvönen, Heikki Surakka
The amount of selection cuings in uneven-aged forest stands is supposed to increase in
Finland with the new Forest Act. In uneven-aged management, it is estimated that the cuing
could be repeated every 15–20 years with the removal of around 100 m3/ha, depending on the
site type and stand growth. This interval and volume highly depend on survival of lower
canopy trees in cuings. The number of these trees (2.5–15 m in height) is typically limited.
Felling larger trees from above means a high damage risk for smaller trees and also restricts
the harvesting outside the heavy frost period due to top damage risk. Damage to trees taller
than 2.5 meters was studied in three selection cuing stands. Mechanized harvesting (har-
vester – forwarder) was carried out in late winter with no frost, which is the optimal time for
selection cuings. On the average 21.5% of the remaining trees were damaged. The percentage
of damage to smaller (2.5–10 m) trees was highest, 28.4%. Stem damage and breakage were
the most common types of injury. A logistic mixed model was used to model the probability
of tree injury (uninjured/injured). Distance from the nearest removed tree, harvested basal
area within 25 m of the tree and diameter of the tree were the explanatory variables taken into
the model. The model discrimination ability by the ROC curve was 72.2%. With a classica-
tion cutpoint of 0.5 for the model ed injury probabilities, the rate of correct classication
was 79.1%. There is a need to develop optimal working practices for mechanized selection
cuings. Information on the stand structure, practical operator tutoring and knowledge of the
goals of the forest owner are needed for successful harvesting implementation.
Keywords: uneven-aged forest management, selection cuing, mechanized harvesting, tree
damage, spatial analysis.
sions. Even more than a half of the forest owners are
satised with the current forest management practic-
es, one of six forest owners feels unsatised especially
with clear cuings, lack of management alternatives,
soil preparation and damage caused by heavy machin-
ery. The aitude towards uneven-aged forest manage-
ment is positive. Near half of forest owners see a po-
tential for uneven-aged management, and 27% of
forest owners are ready to try it at least in a part of
their forest property (Kumela and Hänninen 2011).
Since 2010 there has been a renewal process of the
Finnish Forest Act. The new Forest Act (hp://www.
nlex.//laki/smur/1996/19961093) was set in the
beginning of 2014, leaving more freedom and choices
in forest management. The demand for alternative for-
est management is likely to increase, and forest own-
1. Introduction
Current forest management in Finland is based on
even-aged management. The use of alternative forest
management methods including selection cuings has
been marginal concentrating to urban forests, land-
scape protection areas, valuable habitats, riparian and
other buer zones (e.g. Hyvän metsänhoidon suosi-
tukset 2006). The amount of selection cuings in un-
even–aged forests has been only some thousands of
hectares per year (Selvitysrapori metsälain 6 §: n
2003). The structure of private forest ownership has
changed and today 36% of forest owners live outside
the municipality, where their forest property is located
(Hänninen and Karppinen 2010). Many forest owners
are not highly dependent of forest income anymore
and emphasize multiple values in management deci-
M. Sirén et al. Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42)
34 Croat. j. for. eng. 36(2015)1
ers as well as professionals need more information on
eects and implementation of alternative manage-
ment practices. Compared to even-aged management,
our knowledge on alternative methods is vague. This
lack of knowledge covers stand development and pro-
duction, economy, ecology as well as harvesting.
Kuuluvainen et al. (2012) present a thorough re-
view of even-aged and uneven-aged forest manage-
ment in Boreal Fennoscandia. They conclude that, al-
though the number of relevant studies has increased
in recent years, the ecological and economic perfor-
mance of alternative management methods still re-
mains insuciently examined. The uneven-aged for-
est management consists of a range of methods, in
which the forest cover is only partially removed. In
this denition of uneven-aged management, the tree
age distribution does not necessarily conform to the
reverse J-shaped form and an uneven-aged forest can
also consist of spatially segregated groups of tree age
classes created by the group selection method. This
broader group of uneven-aged management is also
referred to as continuous cover forestry (Pommering
and Murphy 2004).
Pukkala et al. (2010) optimized the structure and
management of uneven-aged stands in Finland. The
post thinning diameter distribution of stands was op-
timized with 20-year cuing cycle when aiming at
maximum economic protability. Spruce stand opti-
mizations were done for fertile and medium sites. The
optimal post-thinning distributions had a truncated
reverse J shape. The harvesting removal on fertile site
was 145 m3/ha and 95 m3/ha on medium site. Optimi-
zation of economic result meant removing all log-
sized trees at 20-year intervals. Increasing discounting
rate and decreasing site productivity improved the
relative performance of uneven-aged management
compared with even-aged management.
Near 100% of harvesting carried out by the forest
industry in Finland is done with the mechanized cut-
to-length method. In selection cuings, harvesting car-
ried out by the forest owner would be a proper alterna-
tive, making also the cuings with small removals
possible. However, the ability of forest owners living
in towns to carry out cuings themselves is limited. If
selection cuings are carried out on a larger scale,
mechanized cuing is the main alternative. The e-
cient use of machinery needs sucient removals, at
least 70–100 m3/ha. In selection cuings, the removal
mainly consists of larger trees. Felling and processing
of these trees means a high risk of damage to smaller
trees and saplings.
In uneven-aged stands, future development, as
well as harvesting conditions, depend on the structure
of the stand, harvesting intensity, forest regeneration,
but also on the amount of damage in harvesting. In
Finland, Norway spruce (Picea abies (L.) Karst) stands
have the highest potential for uneven-aged manage-
ment on a commercial scale (Valkonen and Maquire
2005, Lähde et al. 2002), but they also have high risk of
pathogen infections following harvesting damage
(Hakkila and Laiho 1967, Isomäki and Kallio 1974).
There is also a high risk of Heterobasidon root rot in
all tree size classes in uneven-aged stands (Piri and
Valkonen 2013). Knowledge on harvesting damage in
selection cuing in Scandinavia is limited, and it has
mainly focused on small saplings. Granhus and Fjeld
(2001) found that the injury probability of saplings
depends both on stand and operational characteristics,
the most important factor being the interaction be-
tween these two variables. Sapling height and spatial
distribution of saplings relative to the strip roads and
larger trees of the residual stand represented stand
characteristics, whereas operational characteristics
were described by the operational method and har-
vesting intensity.
Surakka et al. (2011) studied injuries on 0.5–2.5 m
saplings. Depending on the stand, the percentage of
injured saplings varied between 17.6–61.0%. The dis-
tance of the sapling to the nearest strip road, sapling
height, harvested basal area within a distance of 25 m
from the sapling and sapling distance to the nearest
remaining tree explained the probability of injury. Sap-
lings near the strip road and taller saplings were more
prone to damage than saplings located further away
from the strip road and small saplings.
Earlier Fjeld and Granhus (1998) compared the ef-
fect of two operating systems (motor-manual cuing
followed by cable skidding and one-grip harvester fol-
lowed by forwarding) and three harvest intensities on
the injury rate in multi-storied Norway spruce stands.
The average injury rate was higher in mechanized
than in motor-manual harvesting. The largest dier-
ences were at high harvest intensities in densely
stocked stands. The average injury rate was 13% for
small trees (diameter under 10 cm) and 7.5% for larg-
er trees.
The long term future of an uneven-aged stand de-
pends on ingrowth, survival and height growth of
small trees. The growth of small trees is very low, and
with average growth rates, it takes about 60 years for
a spruce germinant to achieve 1.3 m in height (Ee-
rikäinen et al. 2014). Thus shorter term harvesting pos-
sibilities are highly based on survival of smaller and
medium-sized trees in cuings. If a signicant per-
centage of these trees is damaged in harvesting, re-
peated cuings every 15–20 years are not possible.
Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42) M. Sirén et al.
Croat. j. for. eng. 36(2015)1 35
2. Aim of the study
The aim of this study was to evaluate the amount,
type and quality of damage to trees taller than 2.5 m
in mechanized selection cuing (cut-to-length meth-
od, harvester – forwarder) of uneven-aged Norway
spruce stands and to construct a model for the descrip-
tion of damage.
3. Materials and methods
3.1 Study stands and measurements
The harvesting experiments were carried out in
three Norway spruce dominated stands located in
Sounenjoki, Northern Savonia region. The sites repre-
sented the submesic Myrtillus type and the mesic Ox-
alis – Myrtillus in terms of Cajander (1909). The stands
(A, B and C, 1.08, 0.85 and 0.42 hectares, respectively)
were earlier harvested motor-manually in 1987. Aer
that, the stands A and B were harvested again in 1999,
A mechanically with a one-grip harvester and B mo-
tor-manually. All these harvesting operations aimed
to an uneven-aged structure. The whole strip road net-
work was already in place aer these previous entries,
with a few additions. The average distance between
strip roads was 25, 27 and 23 m in stands A, B and C,
Before harvesting, trees of commercially valuable
tree species (i.e. Norway spruce, Scots pine (Pinus syl-
vestris L.), silver birch (Betula pendula Roth), downy
birch (B. pubescens Ehrh) and aspen (Populus tremula
L.) with a height of > 2.5 m were mapped (x and y
coordinates) and measured for diameter at breast
height (d). Tree heights were measured from a sample
of 75–125 trees per stand. The sample trees were used
to estimate the tree heights for the rest of the trees and
the stem volumes for all the trees. The tree heights
were estimated with the models by Näslund (1937),
and the stem volumes with the models by Laasase-
naho (1982) and Kärki et al. (1999).
Trees to be felled were selected in three phases
(Surakka et al. 2011): 1. trees located on the strip roads,
2. trees of weak health or poor technical quality, 3.
Single tree selection from the remaining stand using a
computerized tree selection procedure, with a classical
negative exponential distribution (de Liocourt 1898)
as the structural framework.
The target basal area aer harvesting was set at
20 m2/ha, and the trees to be removed were marked
before cuing. Stand parameters before and aer har-
vesting are summarized in Table 1 and diameter dis-
tributions are presented in Fig.1.
Table 1 Stand characteristics before and after cuttin
Stand A B C
Volume, m3/ha 289 296 295
Basal area, m2/ha 27.9 29.2 31.1
Stems/ha, h>2.5 m 766 1021 1262
Volume, m3/ha 185 200 173
Basal area, m2/ha 18.3 20.3 18.7
Stems/ha, h>2.5 m 625 891 947
Volume, m3/ha 104 96 122
Basal area, m2/ha 9.6 8.9 12.4
Stems/ha, h>2.5 m 141 130 315
Cuing was carried out with a one-grip harvester
Ponsse Ergo HS16 in late March and forwarding in
early April 2007 with a Ponsse Bualo forwarder. The
temperature varied from –1 to +13 °C, snow depth
from 0 to 20 cm, and visibility was good during the
cuing. Two skilled harvester operators carried out
the cuings. The harvester and forwarder drivers were
instructed to prevent damage to both trees and sap-
lings. The harvester operators were instructed to fell
the marked trees away from the strip roads. Aer cut-
ting, the damage caused by the harvester was marked
and inventoried to keep it separate from the damage
caused by forwarding.
Aer harvesting, the total number of remaining
trees higher than 2.5 m was 1808. Injuries were as-
sessed for each remaining tree. Trees were classied
into a) uninjured, b) injured, will survive or c) fatally
injured. Fatally injured consists of perished small
trees, felled trees and trees broken near the ground.
The damage type of a tree could be one of the follow-
ing (for injury b) types 1–6 and for injury c) types 2 or
7 were possible): 1) stem damage including root collar
damage, 2) stem breakage, 3) root damage, 4) crown
damage, 5) tilt, 6) several types of damage or 7) disap-
peared. A stem or root damage was recorded, if the
damage area was at least 1 cm2. Damage size was mea-
sured at two dimensions, lengthwise and crosswise.
Distance from root collar to the beginning of the dam-
age was measured for stem and root damage. Stem
and root damages were further divided into bark dam-
age (bark removed) and wood damage (wood
smashed), depending on how deep the damage was.
For stem breakage, the height of the breaking point
was measured. Crown damage was recorded if the
green crown loss was noticeable (more than 10% of
M. Sirén et al. Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42)
36 Croat. j. for. eng. 36(2015)1
Fig. 1 Diameter distributions of study stands
green crown volume). A tilt was recorded if a tree
tilted at least 10 degrees from its original (vertical) po-
sition. Trees that were not found aer harvesting were
entered as »disappeared«. Usually they were smaller
trees oen under piles of slash or logs. For the model-
ing purposes, we simplied the injury classication
into two groups: uninjured and injured.
3.2 Modelling
Variables aecting the probability of a tree being
injured y (a tree j sampled from a stand i) were ex-
plored using a logistic mixed model:
y= 1, if a tree was injured (living or dead),
probability = p
y = 0, if a tree was uninjured, probability = 1–p
y ~ Binomial (1,)
( )
0 1 1i 2 2ij 3 1i 2ij i
ln ...
pX X X X u
pb b b b
 
= + + + × + +
 
 
where p is the modelled injury probability, ln is the
natural logarithm function, β0, β1, β2, are xed eects
parameters to be estimated, X1i, X2, … are stand-spe-
cic (i) or tree-specic () explanatory variables (con-
Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42) M. Sirén et al.
Croat. j. for. eng. 36(2015)1 37
tinuous or dummy), »×« denotes an interaction eect,
and ui is a random stand eect, ui ~ Normal (0, σ2u).
Adding the stand as a random, categorical variable
to the model takes into account the possible correla-
tion of the tree observations within the stands. In this
data, all the explanatory variables were tree-specic
and continuous except the categorical variable tree
species group (coniferous or broadleaved). To observe
beer the eect of spatial variation on a tree injury,
several tree-specic explanatory variables were per-
formed (Table 2) from the mapped tree data to de-
scribe the remaining stand around the tree. Various
transformations of the continuous explanatory vari-
ables or interactions of the explanatory variables were
also checked in candidate models.
The model t was assessed by the discrimination
and the correct classication of the data, using the
xed part of the model to predict injury probability
p. From the model (1) we get:
( )
( )
( )
( )
0 1 1i 2 2ij 3 1i 2ij
0 1 1i 2 2ij 3 1i 2 ij
exp ...
1 exp ...
pX X X X
b b b b
b b b b
+ + + × +
=+ + + + × +
where exp is the exponential function.
The area under the ROC curve is a measure of the
discrimination ability of a statistical model for a bi-
nary response variable: it is the probability (or the per-
centage) that, for a randomly selected pair of an in-
jured and uninjured tree, the model p is greater for
the injured one. The probability 0.5 was used as a cut-
point value for the model classication of an observed
tree: if the model pij 0.5, the tree was classied
injured (value 1 is closer), otherwise uninjured (value
0 is closer). The cutpoint 0.5 does not usually give the
best correct classication result, but it tries to be an
objective cutpoint. Aer the model classication the
rate of correct classication could be counted by cross-
tabulating observed and predicted tree injuries.
The analyses were carried out by the SAS statistical
soware, version 9.3. (SAS Institute Inc. 2014). The
GLIMMIX procedure was used for the model estima-
tion, the LOGISTIC procedure for the ROC curve calcu-
lation and the FREQ procedure for the cross-tabulation.
Table 2 Statistics of tested possible explanatory variables for the injury model
Variable nMinimum Median Mean Maximum
Tree species 1808 0.00 1.00 0.82 1.00
Diameter 1808 1.00 10.80 14.28 90.00
Height 1808 2.54 11.60 12.73 35.50
Basal 1808 0.00 6.21 6.08 12.73
Stems 1808 0.00 96.77 105.78 325.95
Distance 1808 0.10 4.14 6.05 33.96
Distance_2 1808 0.65 5.93 7.51 34.95
Tree species – tree species group (1 = coniferous tree, 0 = broadleaved tree)
Diameter – diameter of tree at breast height, cm
Height – height of tree, m
Basal – harvested basal area at distance of 25 m from the tree, m2/ha
Stems – harvested number of trees at distance of 25 m from the tree, stems/ha
Distance – distance of tree to the nearest removed tree, m
Distance_2 – distance of tree to the centre of the nearest strip road, m
Fig. 2 Percentages of uninjured and injured trees by stands
M. Sirén et al. Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42)
38 Croat. j. for. eng. 36(2015)1
Fig. 3 Percentages of uninjured and injured trees by height classes
Fig. 4 Percentages of injured trees in damage types by height
4. Results
4.1 Amount, type and severity of injury
Fig. 2 presents the percentage of trees of the stands
A–C divided into three classes: uninjured, injured, will
survive fatally injured. The percentages of trees in dif-
ferent height classes are shown in Fig. 3, and percent-
ages of dierent damage types in Fig. 4.
Stem damage and stem breakage were the most
common damage types, together aecting near 70% of
the damaged trees. Stem breakage or disappearing oc-
curred mostly for lower canopy trees, and these trees
formed the injury class »fatally injured«. The percent-
age of root damage was under 10% for all damaged
trees, but nearly 50% for damaged trees taller than 20 m.
Near 15% of the damaged trees had several types of
Table 3 Distributions percentiles (minimum, lower quartile, median, upper quartile and maximum) of stem damage area (cm2) and the distance
to lowest damage point from root collar (dm, in parenthesis) by damage type and height of tree (m)
Height of
tree, m nMinimum Lower
quartile Median Upper
quartile Maximum
2.5–10 84 1 (0) 8 (2) 15 (5) 45 (8) 380 (35)
10–20 62 2 (0) 8 (5) 20 (10) 48 (18) 1600 (78)
20–35.5 19 5 (0) 21 (4) 100 (9) 600 (27) 4200 (70)
2.5–10 22 15 (0) 36 (2) 120 (5) 210 (7) 300 (33)
10–20 10 6 (0) 60 (2) 155 (5) 320 (16) 560 (42)
20–35.5 3 20 (1) 20 (1) 75 (16) 400 (70) 400 (70)
damage. Most damage, 88.4% of all damage, was
caused in cuing. Forwarding caused 11.6% of dam-
age, and near all root damage was caused in forward-
ing. The distribution percentiles of size distributions
and locations of stem and root collar damage are pre-
sented in Table 3 and distribution percentiles of stem
breakage heights in Table 4.
The amount of damage was also calculated accord-
ing to the classication system of the Forest Act, where
only trees with d1.3 ≥ 7 cm are included and for super-
cial stem damage the minimum wound size is 12 cm2
under d1.3 or 30 cm2 in the whole tree (Fig. 5).
Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42) M. Sirén et al.
Croat. j. for. eng. 36(2015)1 39
Table 4 Distribution percentiles (minimum, lower quartile, median, upper quartile and maximum) of stem breakage height from root collar
(dm) by height of tree (m)
Height of
tree, m nMinimum Lower
quartile Median Upper
quartile Maximum
2.5–10 100 0 0 0 20 95
10–20 30 0 4 53 90 200
20–35.5 1 200 200 200 200 200
Fig. 5 Percentages of uninjured and injured trees by stands with
injury classification according to the Forest Act
degrees of freedom). The area under the ROC curve
for the model was 0.722 (72.2%), which is considered
acceptable discrimination ability by Hosmer and Lem-
eshow (2000). With a classication cutpoint of 0.5 for
the model ed injury probabilities p using the for-
mula (2), the rate of correct classication was 79.1%
(uninjured 99.2%, injured 5.4%).
The inuence of the explanatory variables to the
injury probability of a tree in the model is presented
in Fig. 6.
5. Discussion
The study material used in this study was collected
in 2007 and was earlier used in modelling damage to
0.5–2.5 m saplings (Surakka et al. 2011). Almost the
entire strip road network of the study stands was es-
tablished in earlier cuings, and thus there was very
lile removal from strip roads. Trees to be removed
were marked before cuing, which is not typical in
practical harvesting operation, and the harvester op-
erators were instructed to fell the marked trees away
from the strip roads. As the felled trees were mainly
large, the possibility to li them under felling to the
direction of the strip roads would have been limited.
However, utilizing the strip road openings in fell-
ing could have decreased the amount of damage to
some extent. Mieinen (2005) simulated the eect of
harvester working technique on damage risk for sap-
lings in selection cuing on three permanent study
plots (Eerikäinen et al. 2007), where the locations of
larger trees and saplings had been mapped. A working
method where all trees were felled into the stand away
from the strip roads was compared with a method
where trees nearer than ve meters from the strip road
were directed towards the strip road, also liing them
on the strip road when possible. When felling trees
into the stand without utilization of the strip roads,
45% of saplings were exposed to damage. When strip
roads were used, the percentage of damaged saplings
was somewhat lower, at 38%.
Aer harvesting, injuries were assessed for each re-
maining tree. This kind of total inventory, where all
remaining trees on the area of 2.35 hectares were as-
4.2 Model for spatial variations in the probabil-
ity of injury
In the modelling, there were 1420 uninjured and
388 injured trees. The following model was used to
explain the injury probability of the tree (3):
ln 0.685 0.195
0.040 0.117
 
= − ×
 
 
− × + ×
DISTANCE distance of tree to the nearest re-
moved tree, m
DIAMETER diameter of tree at breast height, cm
BASAL harvested basal area at distance of 25
m from the tree, m2/ha
All the explanatory variables were clearly signi-
cant in the model: p < 0.0001 (F-test with 1 and 1802
M. Sirén et al. Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42)
40 Croat. j. for. eng. 36(2015)1
sessed, is a laborious operation. However, when areas
with suitable uneven-aged structure for the study were
hard to nd, total measurement of quite limited areas
allowed collecting a proper material for modelling.
Typically the damage inventories are made on sample
plots. When quality of harvesting is followed by Finnish
Forest Centre, circular sample plots of 100 m2 are used.
When studying harvesting quality and tree damage in
even-aged stands, Sirén (1998) used 240 m2 rectangular
sample plots divided into eight 30 m2 measuring zones.
In the study of selection cuing damage by Fjeld and
Granhus (1998), the post-harvest inventory was based
on 4 m wide and 24 m long rectangular sample plots
stretching from strip road centre to strip road centre.
Trees were classied into uninjured, injured, will sur-
vive and fatally injured. For small lower canopy trees
the damage was oen fatal, and the class fatally injured
consists of small trees with severe damage. In most
damage inventory methods (Eriksson 1981, Björheden
and Fröding 1986, Sirén 1998), tree damage is moni-
tored only for trees with commercial value.
In this study, 78.5% of the trees had no damage,
while 15.7 trees were damaged but will survive and
5.8% were fatally damaged. Near 90% of the damage
was caused by the cuing operation. The trees with very
severe damage were mainly 2.5–10 m tall lower canopy
trees, which had stem breakage or had disappeared. In
the study of Fjeld and Granhus (1998), the percentage
of damaged trees was 8.7–13.7%. In their study the in-
jury rate increased with harvesting intensity, and was
higher in mechanized than in motor-manual cuing.
The injury rates were highest near strip roads, where the
injury rates with both methods exceeded 20%.
The probability of damage was explained by dis-
tance to the nearest removed tree, harvested basal area
within 25 m from the tree and diameter of tree. The
distance to the nearest removed tree best predicted the
damage probability. The lower canopy trees have
higher injury probability than larger trees. These ex-
plaining variables are logical. With increasing amount
of work per area unit, the risk for damage increases.
When larger trees are felled and processed, smaller
nearby trees are at high risk for damage. The model
for injury probability was able to correctly classify
79.1% of the trees as injured or not injured. In the
model for sapling damage (Surakka et al. 2011), the
rate of correct classication was 73.0%, and with mod-
el presented by Granhus and Fjeld (2001) 70.5%.
In Forestry recommendations (Äälä et al. 2014),
the amount of tree damage is one element of silvicul-
tural harvesting quality. Finnish Forest Centre mea-
sures annually the quality of more than 200 thinning
stands, and the percentage of damaged trees in 2013
was 3.6 (Korjuujälkitarkastukset 2013), which is quite
typical in long run. If the percentage is over 15%, it
exceeds the limit of the Forest Act. Although there is a
large variation between stands, the limit of the Forest
Act is seldom exceeded in even-aged stands.
In this study, the percentage of damaged trees aer
the classication of the Forest Act was 13.8%. As the
amount of damage in good conditions carried out with
skilled operators was near to exceed the limit, we can
nd that in selection cuings it can be very dicult to
reach low damage levels typical for even-aged stands.
Also, new results (Hämäläinen 2014) on the amount
of tree damage in uneven-aged stands show quite high
Fig. 6 Influence of harvested basal area at distance of 25 m from the tree (A = 5 and B = 10 m2/ha), diameter of tree at breast height (cm)
and distance of tree to the nearest removed tree (m) on injury probability of tree
Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42) M. Sirén et al.
Croat. j. for. eng. 36(2015)1 41
damage numbers and high damage risk especially in
lower canopy trees.
High percentages of damaged trees with mecha-
nized cut-to-length method have also been presented
outside Scandinavia in even-aged stands. In Germany
12.6% of remaining trees were damaged (Sauter 1995).
In North-America Beinger and Kellogg (1993) re-
ported very high damage numbers. In their study near
40% of remaining trees were damaged.
Košir (2008) modelled the amount of tree damage
with motor-manual and cut-to-length methods in Slo-
venia. With motor-manual method, the share of injured
trees in thinnings was 17–19% and with cut-to-length
method 13–15%. During a 160-year rotation period 10
thinnings take place, and the total number of damaged
trees continuously grows and reaches 90% at the end of
rotation. The modelled percentage of damaged trees is
very high, but a high share of damaged trees (64–70%)
has also been recorded in old stands (Košir 1998).
There are many interesting study topics in selec-
tion cuing operations and working methods. As
stand structures and goals of forest owners vary sub-
stantially, the harvester operator needs information on
stand structure, slopes and soil bearing capacity. There
are new interesting ways to utilize multi source infor-
mation and operator tutoring in harvesting (Räsänen
et al. 2014, Väätäinen et al. 2013). Selection cuings are
a challenging but interesting scene to test these tools.
6. References
Beinger, P., Kellogg, L.D., 1993: Residual stand damage from
cut-to-length thinning of second growth timber in the Cas-
cade Range of western Oregon. Forest Products Journal 43(11
12): 59–64.
Björheden, R., Fröding, A., 1986: Ny rutin för praktisk gall-
ringsuppföljning. (A new routine for checking the biological
quality of thinning in practice) Sveriges lantbruksuniversitet,
Institutionen för skogsteknik. Uppsatser och Resultat 48. 14
Cajander, A., 1909: Über Waldtypen. (Forest site types) Acta
Forestalia Fennica 1(1): 175.
De Liocourt, F.,1898: Deaménagement des sapinieres.
(Spruce stand management) Bull. Soc. Franche-Comet Bel-
Eerikäinen, K., Miina, J., Valkonen, S., 2007: Models for the
regeneration establishment and the development of estab-
lished seedlings in uneven-aged, Norway spruce dominated
forest stands of southern Finland. Forest Ecology and Man-
agement 242(2–3): 444–461.
Eerikäinen, K., Valkonen, S., Saksa, T., 2014: Ingrowth, sur-
vival and height growth of small trees in uneven-aged Picea
abies stands in southern Finland. Forest Ecosystems 2014(1:5),
10 p.
Eriksson, L., 1981: Stickvägar och körskador i gallringsbe-
stånd. Resultat från Riksskogstaxeringens inventering åren
1978–1979 (Strip roads and damages caused by machines
when thinning stands. Results from the Swedish National
Forest Survey for 1978 and 1979). The Swedish University of
Agricultural Sciences. Department of Operational Eciency.
Report No 137, 44 p.
Fjeld, D., Granhus, A., 1998: Injuries aer selection harvesting
in multi-storied spruce stands – the inuence of operating
systems and harvest intensity. Journal of Forest Engineering
9(2): 33–40.
Granhus, A., Fjeld, D., 2001: Spatial distribution of injuries to
Norway spruce advance growth aer selection harvesting.
Canadian Journal of Forest Research 31: 903–913.
Hakkila, P., Laiho, O., 1967: On the decay caused by axe marks
in Norway spruce. Communicationes Instituti Forestalis Fen-
niae 64(3): 1–34.
Hosmer, D.W., Lemeshow, S., 2000: Applied logistic regres-
sion, 2nd edition. Wiley series in probability and statistics, New
York, 392 p.
Hyvän metsänhoidon suositukset, 2006. (Forestry recommen-
dations) Metsätalouden kehiämiskeskus Tapio, 100 p.
Hämäläinen, J., 2014: Poimintahakkuun nykykäytännöt: työ-
ohjeistus, ajanmenekki ja korjuujälki (Practices in selection
cuings: operator tutoring, time consumption and harvesting
damage) Pro gradu – work. University of Helsinki, Depart-
ment of Forest Sciences, 46 p.
Hänninen, H., Karppinen, H., 2010: Metsänomistusrakenteen
muutos ja puuntarjonta. (Change in forest ownership and
wood supply) In: Hänninen, R., Sevola, Y., (eds.). Finnish For-
est Sector Economic Outlook 2010–2011: 52–55.
Isomäki, A., Kallio, T., 1974: Consequences of injury caused
by timber harvesting machines on the growth and decay of
spruce (Picea abiea (L.) Karst). Acta Forestalia Fennica Vol.
136: 1–25.
Korjuujälkitarkastukset 2013. (Harvesting quality survey
2013) Finnish Forest Centre bulletin.
Košir, B., 1998: Damage to mountain spruce stands due to
harvesting. Conference proceedings »Gorski gozd«, Ljublja-
na. Biotechnical Faculty, Department of Forestry and Renew-
able Forest Resources: 95–107.
Košir, B., 2008: Modelling stand damages and comparison of
two harvesting methods. Croat. j. for.eng. 29(1): 5–13.
Kumela, H., Hänninen, H., 2011: Metsänomistajien näkemyk-
set metsänkäsielymenetelmien monipuolistamisesta. (Forest
owners´opinions on alternative forest management practices)
Working Papers of the Finnish Forest Research Institute 203,
76 p.
Kuuluvainen, T., Tahvonen, O., Aakala, T., 2012: Even-aged
and uneven-aged forest management in Boreal Fennoscandia:
a review. AMBIO 41(7): 720–737. Royal Swedish Academy of
Kärki, T., Eerikäinen, K., Heinonen, J., Korhonen, K. T., 1999:
Harmaalepän (Alnus incana) tilavuustaulukot. (Volume func-
tions for grey alder) Metsätieteen aikakauskirja, 1/1999: 39–49.
Laasasenaho, J., 1982: Taper curve and volume functions for
pine, spruce and birch. Communicationes Instituti Forestalis
Fenniae 109, 74 p.
Lähde, E., Laiho, O., Norokorpi, Y., Saksa, T., 2002: Develop-
ment of Norway spruce dominated stands aer single-tree
M. Sirén et al. Tree Damage in Mechanized Uneven-aged Selection Cuings (33–42)
42 Croat. j. for. eng. 36(2015)1
selection and low thinning. Canadian Journal of Forest Rese-
arch 32: 1577–1584.
Mieinen, A., 2005: Paikkatietoanalyysien soveltaminen eri-
ikäisrakenteisten metsien hakkuiden tutkimuksessa. (Spatial
analysis in uneven-aged forestry research) Pääötyö, paikka-
tiedonhallinnan erikoistumisopinnot, Hämeen ammaikor-
keakoulu, 10 p.
Näslund, M., 1937: Skogsföranstaltens gallringsförsök i tall-
skog. (Thinning trials in pine stands) Meddelanden från Sta-
tens Skogsförsökanstalt 29, 169 p.
Piri, T., Valkonen, S., 2013: Incidence and spread of Heteroba-
sidion root rot in uneven-aged Norway spruce stands. Cana-
dian Journal of Forest Research 43(9): 872–877.
Pommering, A., Murphy, S.T., 2004: A review of the history,
de nitions and methods of continuous cover forestry with
spe cial aention to aorestation and restocking. Forestry
77(1): 27–44.
Pukkala, T., Lähde, E., Laiho, O., 2010: Optimizing the struc-
ture and management of uneven-sized stands of Finland: For-
estry 83(2): 129–141.
Räsänen, T., Hämäläinen, J., Lamminen, S., Lindeman, H., Sal-
mi, M.,ätäinen, K., 2013: Uudet informaatiolähteet puun-
hankinnan tukena. (Support of new information sources for
wood supply) Metsätehon rapori 279, 28 p.
SAS Institute Inc. 2014. SAS/STAT 9.3 User’s Guide. (hp://
Sauter, U., 1995: Competing long and short wood cuing sys-
tems using harvesters for thinning conifer stands. In: Kellogg,
L., Milota, G. (eds.): The way ahead with harvesting and trans-
portation technology. Proceedings of Iufro P3.04 meeting. Iu-
fro XX World Congress. Tampere, Finland, 70–76.
Selvitysrapori metsälain 6 §: Mukaisesta hakkuusta erityis-
kohteilla Maa-ja metsätalousministeriölle 8.12.2003. (Report
on cuings on specic areas aer Forest Act 6 §) Forestry De-
velopment Center Tapio and WWF Finnish Fund. 20 p.
Sirén, M., 1998: One-grip harvester operation, it´s silvicultural
result and possibilities to predict tree damage. Metsäntutki-
muslaitoksen tiedonantoja, 694, 179 p.
Surakka, H., Sirén, M., Heikkinen, J., Valkonen, S., 2011: Dam-
age to saplings in mechanized selection cuing in uneven-
aged Norway spruce stands. Scandinavian Journal of Forest
Research 26(3): 232–244.
The new Forest Act. (
Valkonen, S., Maguire, D.A., 2005: Relationship between seed-
bed properties and the emergence of spruce germinants in
recently cut Norway spruce selection stands in Southern Fin-
land. Forest Ecology and Management 210(1–3): 255–266.
Väätäinen, K., Lamminen, S., Ala-Ilomäki, J., Sirén, M., Asi-
kainen, A., 2013: Kuljeajaa opastavat järjestelmät koneelli-
sessa puunkorjuussa (Operator tutoring systems in mecha-
nized harvesting) Working Papers of the Finnish Forest
Research Institute 279, 24 p.
Äälä, O., Koistinen, A., Sved, J., Vanhatalo, K., Väisänen, P.
(eds.) 2014: Metsänhoidon suositukset (Forestry recommen-
dations) Metsätalouden kehiämiskeskus Tapion julkaisuja.
264 p.
Received: May 5, 2014
Accepted: September 2, 2014
Authors’ address:
Mai Sirén, PhD.*
e-mail: Mai.Siren@luke.
Natural Resources Institute Finland
Vantaa, PB 18
01301 Vantaa
Juha Hyvönen, BSc.
e-mail: Juha.Hyvonen@luke.
Natural Resources Institute Finland
Rovaniemi, PB 16
96301 Rovaniemi
Heikki Surakka, MSc.
e-mail: Heikki.Surakka@ramboll.
Ramboll Finland Oy, PB 25
02601 Espoo
* Corresponding author
... Current information on damage to residual trees following fully mechanised harvesterforwarder systems is primarily focused on young coniferous stands reporting 29.1-36.1% damaged trees, but with a mean wound area under 100 cm 2 in the thinning of a 12-year old Pinus taeda stand with a removal of 40% of trees in the stand [25], and 25% damaged trees in the first thinning of Pinus taeda stand with a removal of 50% of trees in the stand [26], to as low as 7.0% (with a 1 cm 2 bark damage threshold) in the first thinning with an intensity of 60% of the standing trees in Norway spruce dominated stand without previous marking of trees designated for felling, and 3.2% when prior marking of the trees was conducted [27]. Mechanised thinning of an uneven-aged Norway spruce dominated stand resulted in 21.5% damaged trees [28]. Late thinnings resulted in damage in 37.4% of the remaining trees in a mixed conifer stand [29] and shelterwood system regeneration felling (with 18.6 and 17% intensity) of a sessile-oak-dominated stands resulted in 20.47 and 23.36% damaged trees [30]. ...
... The share of trees with peeled bark, expressed as a percentage of the number of remaining trees after finishing harvesting operations, was 33% in Subcompartment 14b and 39% in Subcompartment 14c. When compared with previous research on residual stand damage in mechanised thinning, the results are similar to those reported for young coniferous stands (29.1-36.1% and 25% damaged trees) [25,26] and for late thinnings of a mixed conifer stand (37.4% of the remaining trees damaged) [29], but higher than 21.5% damaged trees in uneven-aged Norway spruce dominated stands [28] as well as 20.47% and 23.36% damaged trees in shelterwood-system regeneration felling of sessile oakdominated stands [30]. Previous research shows a wide range in the intensity of damage to the remaining trees in timber harvesting operations for the thinning of hardwoods (3-50%) [23]. ...
... With regard to the cause of peeled-bark injury, the harvester in Subcompartment 14c damaged 18% more trees than the forwarder, while in Subcompartment 14b the harvester caused as much damage as the forwarder. This is opposed to the research results on the thinning of young Douglas-fir stands, where the harvester damaged more than twice as many residual trees than the forwarder [15]; it is also opposed to results from unevenaged Norway spruce dominated stands, where 88.4% of the damage was caused by an harvester and only 11.6% by a forwarder [28]. However, it was to the results reported for mechanised thinning in young coniferous stands [25]. ...
Full-text available
This research was conducted to determine the cause, intensity and location of damage (stem, butt end, root collar, root) and the extent of damage to standing trees during felling and processing by an harvester and timber extraction by a forwarder (cut-to-length system). The research was conducted in the central part of the Republic of Croatia in the Management Unit (MU) “Bjelovarska Bilogora” during the thinning of Subcompartment 14b, area of 18.28 ha, in the stand of hornbeam (Carpino betuli—Quercetum roboris fagetosum Rauš 1975), age 70, and of Subcompartment 14c, area of 9.07 ha, in a stand of common beech (Carici pilosae—Fagetum Oberdorfer 1957) aged 79 years. The thinning intensity was 12.13% in Subcompartment 14b and 13.72% in Subcompartment 14c. Field measurements were carried out on sample plots—the first time in 2017 to determine the intensity and characteristics of the damage to standing trees with regard to the cause of the damage (harvester or forwarder), and the second time in 2018 to determine the overall intensity and features of the damage to standing trees after finishing harvesting operations. For all trees remaining in the stand after the harvesting operations, the following were determined: tree species, diameter at breast height (DBH), the position of the tree in the stand depending on the forest traffic infrastructure, and—if damaged—cause of damage, type of damage, the position of damage on the tree, and dimensions of damage. The intensity of the damage was expressed by the ratio of damaged and undamaged trees, with a detailed analysis of bark damage (squeezed-bark damage and peeled-bark injuries). The results of the research indicate the highest prevalence of peeled-bark injuries. In relation to the total number of standing trees, trees with peeled-bark injuries were more represented in Subcompartment 14c (39%) than in Subcompartment 14b (33%). In Subcompartment 14b, the harvester and the forwarder damaged an equal number of trees, while in Subcompartment 14c, the harvester damaged 59% of the damaged trees. In both subcompartments, an average of 83% of (peeled bark) injuries were up to 1.3 m above the ground. In both subcompartments, the most common (67%) were injuries up to 100 cm2 in size, for which many authors claim the tree can heal by itself. Given the increasing use of harvester-forwarder systems in deciduous stands and research results that indicate possible damage to standing trees, it is necessary to pay attention to all phases of planning and execution of timber harvesting operations, thus minimising negative effects.
... Additionally, the probably of damage to a remaining understory tree is related to a) the size of the closest harvested tree (Tatsumi et al. 2014), b) the distance to the closest harvested tree (Sawaguchi et al. 2001;Tatsumi et al. 2014;Sirén et al. 2015), and c) the total basal area harvested within a certain distance (Surakka et al. 2011;Sirén et al. 2015). These results strongly suggest that anthropogenic factors have a significant impact on the spatial pattern of understory damage. ...
... Additionally, the probably of damage to a remaining understory tree is related to a) the size of the closest harvested tree (Tatsumi et al. 2014), b) the distance to the closest harvested tree (Sawaguchi et al. 2001;Tatsumi et al. 2014;Sirén et al. 2015), and c) the total basal area harvested within a certain distance (Surakka et al. 2011;Sirén et al. 2015). These results strongly suggest that anthropogenic factors have a significant impact on the spatial pattern of understory damage. ...
... Of these five explanatory variables, the former three values (c, t f and t y ) were generated with the physical contact model developed in GIS. Previous studies have used DBH to determine the relationship between individual size and injury risk (Tatsumi et al. 2014;Sirén et al. 2015). In this study, DBH was substituted by the maximum tree height in each grid. ...
Damage to the understory of shelterwood systems during generational succession is an issue that has been debated for a long time. In this study, we analyzed the spatial relationship between damaged understory trees and harvested trees during the final cutting in a shelterwood forest stand dominated by Chamaecyparis obtusa (cypress) in central Japan. We surveyed the positions of all cypress trees and observed the final cutting operation. By measuring the felling direction, the physical contact between the harvested trees and understory trees was reproduced using the Geographic Information System (GIS). Logistic regression was performed using the extracted frequencies of physical contact as explanatory variables and variables important for reproducing the spatial pattern of the damaged understory were selected. The model selection indicated that the damage to the understory could be explained by the contact between the understory trees and the crowns and trunks of felled trees. The understory height was relatively ineffective in explaining the spatial damage pattern. The model could reproduce the spatial distribution of damage rates more realistically than conventional models which depend on the distance from stumps or roads. It was quantitatively revealed that controlling the felling direction was an important factor in mitigating understory damage.
... The level of damage caused to residual trees during forest operations has been described by several authors and many factors influencing damage level have been indicated. As a summary of these factors from different findings, Siren et al. [76] quote in their publication that the level of damage depends on the forest characteristics, such as the amount of timber removed during harvesting, stand density and basal area. In addition, well designed access to the forest can limit the probability of damage, such as skid-trail spacing or road density [12]. ...
... It was also found that tree injuries depend on the season, and can be higher in summer [79], as well as affected by stand structure: in uneven-aged stands, damage can be more frequent, especially among younger trees [76]. Mechanised harvesting tends to cause less damage than using a chainsaw for felling and more damaged trees are observed near strip roads [80,81]. ...
... Damage to the remaining stand can be observed after thinning, selective cutting or another partial tree removal from a stand [76]. Damage is generally understood as bark and cambium removal with partial wood tissue disturbance, also a broken tree. ...
Full-text available
Purpose of Review Lowering the impact of forest utilisation on the forest environment is a part of the improvement in sustainable forest management. As part of forest utilisation, timber harvesting can also cause environmental implications. The main impact of forest operations is on the soil, on regeneration and on the residual stand. The aim of the present review was to identify the state of the art in forest utilisation, identifying how and how much forest operations affect forest soil, regeneration and the remaining stand. Particular attention was paid to the level of impact and potential to limit this. Recent Findings There are a large number of publications tackling forest harvesting, but most of them do not give a comprehensive framework and they mainly focus on one or very few aspects of forest damage. In order to improve general knowledge of the impact of forest operations, it was proposed that the scope of recent findings should be examined and a compilation of the available results from different regions should be presented in one paper. Summary It was found that the least impactful machine-based forest operations were harvester–forwarder technologies, while a larger scale of damage could be expected from ground-based extraction systems (skidders) and cable yarders. Animal power, if applicable, tended to be very neutral to the forest environment. A decrease in damage is possible by optimising skid trail and strip road planning, careful completion of forest operations and training for operators. The existence of legal documents controlling post-harvesting stand damage are rare and have been implemented in only two countries; there is no post-harvesting control on soil damage and natural regeneration.
... damage to trees that have grown in the patches cut in the first stage). Previous studies have found a negative relationship between risk of damage to a tree and several factors; distance to nearest tree cut (Granhus and Fjeld 2001;Sirén et al. 2015), size of the residual tree (Granhus and Fjeld 2001;Sirén et al. 2015) and distance to strip road . While saplings have been found to be exposed to a high probability of damage in felling of large trees (Sikström and Glöde 2000), the residual trees in subsequent treatments cannot be considered saplings. ...
... damage to trees that have grown in the patches cut in the first stage). Previous studies have found a negative relationship between risk of damage to a tree and several factors; distance to nearest tree cut (Granhus and Fjeld 2001;Sirén et al. 2015), size of the residual tree (Granhus and Fjeld 2001;Sirén et al. 2015) and distance to strip road . While saplings have been found to be exposed to a high probability of damage in felling of large trees (Sikström and Glöde 2000), the residual trees in subsequent treatments cannot be considered saplings. ...
Patch cutting is beneficial for many ecosystem services, but the effects of the management system on operations have not been analyzed. A two-machine system with harvester and forwarder is often used in mechanized cut-to-length operations. The aim of this study was to analyze differences in harvester and forwarder productivity in final felling and patch cutting, and estimate their effects on net revenues per harvested m3. Harvester time consumption was studied using automatic data collection from the machine computer. The data set comprised approximately 18,150 trees harvested during 48 shifts. Analyzes were based on shift level averages. In the observed interval of 0.30–0.60 m3 average tree volume, patch cutting productivity was 20–15% lower compared to final felling. Forwarding was analyzed in three steps. First, a GIS analysis of terrain transport distance found that patch cutting increased forwarding distance by 29%. Secondly, a time study found that loading and unloading times were 16% greater in patch cutting than in final felling. Thirdly, a theoretical analysis found that total forwarder time consumption was 16% higher in patch cuts than in final felling areas. Operational costs in patch cutting were 18% higher than in final felling, thereby reducing net revenues from harvesting operations by 4%. While operational costs were found to be higher in patch cutting than final felling, they are lower than the costs expected for other continuous cover forest management systems. Investigations of later stages of patch cutting are needed before full conclusions regarding the management system can be drawn.
... Lundqvist et al. (2007) reported mortality rates in two experimental stands including thinning from above as a treatment with relevance to selection harvesting. However, harvesting damage in singletree selection is well covered by several studies (Fjeld and Granhus 1998;Granhus and Fjeld 2001;Surakka et al. 2011;Sirén et al. 2015) and practical experience. ...
... If we correct the harvest loss level in the results by a factor of 1.33, the harvest loss would thus average 0.5 trees ha −1 a −1 in N and 0.04 m 3 ha −1 a −1 in V, representing 10% and 6% of the total mortality with an emphasis on the smaller trees Harvesting losses were probably less common in the experimental setting than could be expected in practical harvesting under similar conditions. The study by Sirén et al. (2015) was conducted under comparable conditions on practical harvesting sites. In a single entry, the proportion of fatally damaged trees was 5.8% of N, while in this study it was somewhere around 1-2%. ...
Full-text available
This study focused on tree mortality in spruce-dominated stands managed using the single-tree selection method in southern Finland. Together with regeneration and tree growth, mortality is one of the basic elements of the stand structure and dynamics in selection stands. The study was based on data acquired from a set of 20 permanent experimental plots monitored with repeated measurements for 20 years. The average mortality in the number of stems (N) was 4.45 trees ha−1a−1, in basal area (G) 0.07 m2 ha−1a−1, and in stemwood volume (V) 0.56 m3 ha−1a−1. In relative terms it was 0.50% of N, 0.30% of G and 0.27% of V, respectively. Wind and snow were the most common causes of mortality, while deaths by biotic causes (mammals, insects, pathogens) were extremely rare. Some 6–10% of the total loss in the number of stems and volume was attributable to the loss or removal of trees that sustained serious damage in harvesting. Most of the mortality occurred in the smallest diameter classes of up to 20 cm. Such a high mortality among small trees can have an adverse influence on the sustainability of selection structures if not successfully checked in harvesting and management.
... In addition to the intrinsic operational aspects of each situation, it is recommended to carry out studies in relation to other factors which may affect damage occurring, as their intensity may be related to the time of year, the location and the physiological conditions of the stand (for example). Cabral et al. (2018) observed that it is difficult to avoid damage occurring to the remaining trees of a stand submitted to thinning caused by the movement of wood and machinery; however, Sirén et al. (2015) described the need to plan thinning and develop better working techniques in order to reduce damage occurrence to trees in forest stands. ...
Full-text available
The objective of this study was to evaluate the damage caused to the remaining trees of a pinus stand submitted to mechanized thinning by a wheeled harvester. The data were obtained in the operational areas of timber harvesting in a forest company located in the city of Inácio Martins, in the state of Paraná, Brazil. The Pinus stand of the study was 11 years old, and was submitted to the first commercial thinning. The operation was performed by a harvester, characterized by the systematic removal of the 5th planting line to give access to the interior of the stand, followed by selective thinning in the individuals demarcated in the two lines adjacent to the traffic trails. The damage caused to the remaining trees of the stand regarding their numbers, dimensions and location in relation to the harvester’s operation track was evaluated, and the data were analyzed using the Profile Analysis multivariate analysis technique. It was observed that 25% of the remaining trees suffered some kind of damage, being considered a high value, but current to the literature. It was evidenced that the section of the tree in need of greater alert was the base because damages in greater intensity and dimensions were observed to it, which can bring future losses to the stand. Thus, a need for improvements in operational procedures was demonstrated.
Full-text available
Considerable amounts of noncommercial materials generated from thinning treatments remain unattended on the site because the value of small-sized timber is lower than overall thinning operation costs in South Korea. In addition, thinning operations with conventional and mechanized harvesting systems often cause severe physical damage to residual trees. In this study, therefore, we compared and analyzed the harvesting productivity, cost, and residual stand damage between single-tree selection thinning (SST) and mechanized line thinning (MLT) systems on conifer plantation forests. For conventional SST, ground skidding (uphill/downhill) was performed using a tractor winch after manual felling and bucking. The MLT consisted of mechanized felling, downhill shovel logging, and processing with a small-scale grapple-saw for the fourth double row (MLT1) and the third row (MLT2) thinning section. The MLT system was more productive and cost-effective in performing thinning treatment and collecting thinning materials than SST. The MLT1 and MLT2 costs were 81.4% and 70.6% lower than the SST cost ($77.6/m3), respectively. The residual stand damages of the SST (3.4%) were lower than those of MLT1 (4.8%) and MLT2 (21.2%); however, there was no significant difference in residual stand damages between two thinning systems (p > 0.05). Therefore, forest managers should consider the use of MLT system to reduce thinning costs and efficiently produce thinning materials for their thinning operations. However, operators still need to be careful felling and extracting operations to reduce the residual stand damages for thinning treatments.
An analysis of the existing literature on the issue of damage to regeneration caused by timber harvesting, revealed that a great majority of results reported in those publications was obtained through laborious and time-consuming field research conducted in two stages. Field research methods for gathering data, employed by various authors, differed in terms of the manner of establishing trial plots, the accuracy of counting and evaluating the number of saplings growing on the investigated sites, classification systems used for distinguishing particular groups of regeneration based on quantitative (diameter at breast height, tree height) and qualitative features (biosocial position within the certain layer and the entire stand), classification systems used for identifying types of damage caused by cutting and felling, as well as transporting operations, and finally the duration of observation intervals and time spent on gathering data on the response of damaged saplings from both, the individual and collective perspectives. Obviously, the most reliable manner of gathering such data would be to count all damaged elements of the environment being a subject of interest of particular investigators at the certain point of time. However, due to time and work consumption of this approach, which is besides very costly, any research should be designed in such a manner as to reduce the above-mentioned factors. This paper aimed to (1) analyse the probability of occurrence of damage to regeneration depending on the form of timber assortments dragged from the felling site to the skidding routes, and timber harvesting technology employed in logging works, and (2) identify a method ensuring that gathered data is sufficient for performing reliable evaluation of share of damage to regeneration at acceptable accuracy level, without necessity to establish trial plots before commencing harvesting works. The scope of these studies enclosed a comparison between two motor-manual methods of timber harvesting in thinned stands, with dragging of timber in the first stage of skidding from the stand to landings. According to one of these methods, a classical one, operations of felling and delimbing of trees were carried out by sawmen at the felling site. Timber obtained using different methods was skidded by carters and horses, and operators of a light-duty cable winch, driven by the chainsaw’s engine, as well as operators of cable winches combined with farm tractors. In the latter, alternative method, sawmen performed only cutting and felling of trees. Delimbing and cross-cutting of trunks, dragged from the felling sites, was carried out by operators of processors combined with farm tractors, worked on skidding routes. The research was conducted in the years 2002–2010 in stands within the age classes II–IV mostly, located in the territories of Regional Directorates of State Forests in Krakow and Katowice, and in the Forest Experimental Unit in Krynica-Zdrój. In the course of a preliminary stage of investigations 102 trial plots were established in stands within early and late tinning treatments. As a result of the field research carried out in two stages, more than 3.25 thsd. circular sites were established and marked, on the surface of which over 25 thsd. saplings constituting the regeneration layer were inventoried. Based on the results of investigations and analyses it was revealed that regardless of the category of thinning treatment, the highest probability of occurrence of destroying P(ZN) to regeneration (0.24–0.44) should be expected when the first stage of timber skidding is performed using cable winches. Slightly lower values of probability (0.17–0.33) should be expected in stands where timber is skidded by horses, while in respect to processor-based skidding technology the probability of destroying occurrence oscillates between 0.12 and 0.27, depending on the particular layer of regeneration. P(ZN) values, very close to those of skidding technology engaging processors, were recorded for skidding performed using the light-duty cable winch driven by the chainsaw’s engine (0.16–0.27). The highest probability of damage P(USZK) to regeneration (0.16–0.31) can be expected when processors are used in the first stage of timber skidding. Slightly lower values of probability (0.14–0.23) were obtained when skidding was performed with the use of cable winches, whereas engaging horses for hauling of trunks results in probability of damage occnrrence oscillating between 0.05–0.20, depending on the particular layer of regeneration. With regard to the probability of occurrence of both, destroying and damage P(ZNUSZK) to regeneration (0.33–0.54), the highest values can be expected when cable winches are engaged in the first stage of skidding. Little lower (0.30–0.43) was the probability of their occurrence if processor-based technology of skidding was employed, while in respect to horse skidding these values oscillated between 0.27–0.41, depending on the layer of regeneration. The lowest values of probability of occurrence of damage P(USZK), and destroying and damage treated collectively P(ZNUSZK), within all layers of regeneration, were recorded in stands where thinning treatments were performed using the light-duty cable winch driven by the chainsaw’s engine. The models evaluated and respective equations, developed based on those models, for evaluating the number of destroyed saplings ZNha (tab. 40, 42, 44, 46, 48) could be used for determining the share of damage expressed as a percentage, upon conducting only one field research at the investigated felling sites, once the timber harvesting and skidding would have been completed. As revealed by the results of analyses, evaluation of statistically significant regression models was possible for all layers of regeneration (tab. 39, 41, 43, 45, 47). Nevertheless, the smallest part of these models that could be considered positively verified, were those for the natural young regeneration, although almost a half of them revealed to be significant. Within the medium-sized regeneration over three-fourths of all models could be considered positively verified, four of which explained more than 50% of variability. Within the high-sized regeneration almost two-thirds of evaluated regression models were statistically significant, five of which were verified positively, moreover, one of them explained more than 50% of variability. The most promising results were those obtained for the advance growth. Nearly 90% of the evaluated models revealed to be statistically significant, ten of which could be considered positively verified. Furthermore, four statistically significant models explained over 50% of general variability. With regard to the entire regeneration more than 80% of evaluated models were statistically significant. However, due to insignificant coefficients of regression, eight of them could be considered positively verified. At this point it should be stressed that in respect to logging technology employing the light-duty cable winch FKS it was impossible to evaluate statistically significant models of regression. Whereas, in the case of processor-based logging technology, firstly regarding the advance growth, and then the entire regeneration, all of the evaluated statistically significant models could be considered positively verified, in terms of both, all of the stands, and particular categories of thinning treatments individually. This latter case also revealed the highest degree of matching of evaluated models (R2 popr 0.73–0.76 for advance growth and 0.78–0.94 for the entire regeneration). A significant impact of the kind of form of hauled timber on the probability of damage occurrence P(USZK), mainly in early thinning treatments, could have been reflected in the results obtained for all stands (early and late thinning treated collectively). Moreover, due to an insignificant impact of the form of hauled timber and logging technology employed, on the probability of occurrence of damage in late thinned stands, and a significant impact of the above-mentioned variables on early thinned stands, it should be assumed that for performing an evaluation of destroying and damage caused by timber harvesting the both thinning treatment categories should be analysed separately. Furthermore, when evaluating the probability of occurrence of destroying and damage caused by timber harvesting, the layers of natural young regeneration and advance growth should be analysed separately. As proved by the results presented in this paper, varying values of probability computed for each of the layers of regeneration seem to indicate that when investigating damage to regeneration caused by timber harvesting, it would be reasonable and recommended to perform a separate analysis of damage to the highest saplings as well, namely individuals with diameter at breast height close to 7 cm. In respect to studies on damage to regeneration caused by logging technologies mentioned above, the evaluation of number of destroyed saplings within the advance growth can be carried out using the proportions of damaged and undamaged saplings per 1 ha of the stand. The numbers evaluated in this manner can be used to calculate the damage share expressed in relative values (percentage of damaged saplings compared with the entire number of saplings before commencing the logging works). However, one should keep in mind that this is true only if the field research have been carried out based on the methodology described in this paper.
Full-text available
A major component of sustainable forest management are the stands left behind after the logging operation. Large mechanized harvesting equipment involved in current forest management can inflict damage on residual trees; and can pose a risk of mortality from diseases, natural calamities, and/or degrade future economic value. The primary objective of this study was to evaluate the residual stand damage under different harvesting methods and silvicultural prescriptions i.e., crop tree release (CTR), diameter limit cut (DLC), and overstory removal (OSR). The second objective was to evaluate the intensity and frequency of damage occurring on the bole, canopy, and root at tree and stand level. The third objective was to document strategies adopted globally to minimize stand damage due to timber harvesting. Five harvest blocks implementing three silvicultural prescriptions, were selected as the treatments across two different industrial timberlands in central and northern Maine (Study Site (SS) I and II, respectively). A hybrid cut-to-length (Hyb CTL) and whole-tree (WT) harvesting method were employed for conducting the harvest in SS I and II, respectively. Systematic transect sampling was employed to collect information on type, frequency, and intensity of damages. The inventory captured 41 and 8 damaged trees per hectare with 62 and 22 damages per hectare from SS I and SS II respectively. Bole damage was the most frequent damage across all treatments. The Hyb CTL had lower damage density (damage per ha) and severity compared to WT. The average number of trees damaged per ha was higher for CTR prescriptions compared to DLC. There were no significant differences in the height of the damages from the ground level between treatments within each study site; however, there was a significant difference between the study sites. Species damaged was directly related to the residual trees left behind and was dominated by American beech, yellow birch, sugar maple, and eastern hemlock. Finally, the study provides strategies that can be adopted at different forest managerial phases to mitigate residual stand damage.
Full-text available
Forests provide a variety of ecosystem services and traditional forest management is largely based on the extraction of one product, wood. Multifunctional forestry, forest management aimed at benefitting multiple ecosystem services, has emerged as awareness has grown of other forest ecosystem services. Nature conservation management is a type of multifunctional forestry promoting ecosystem services other than harvest of wood, most commonly biodiversity and recreation. While the benefits of multifunctional forestry and nature conservation management is recognised, there are knowledge gaps regarding how to perform these operations. The overarching objective of this thesis is to increase knowledge and improve implementation of multifunctional forest operations in Sweden. This is addressed through four studies aiming at answering questions related to how forest operations can be implemented in multifunctional forestry. The findings indicate that many conservation values in forest land can be identified using commonly available GISdata. In most cases, nature conservation management operations are not complicated, but forest managers are disincentivised by conflicting goals and fear of high costs and criticism. The conclusion from detailed studies of operations is that costs in multifunctional operations are higher than conventional operations, but when the entire management system is analysed, effects on net revenues may be small. The general conclusion is that, in many cases, multifunctional forestry is not limited by the operations but rather a lack of clear goals and strategies for achieving goals and evaluating their attainment.
Full-text available
Notwithstanding Norway spruce (Picea abies (L.) Karst.) is highly prone to root rot caused by Heterobasidion parviporum Niemelä & Korhonen, but little is known about the epidemiology of Heterobasidion root rot in spruce stands applied to uneven-aged management. To get insight into the development of Heterobasidion infections in this type of forest, the size and spatial distribution of individual genets of H. parviporum were determined in five uneven-aged managed Norway spruce stands in southern Finland. In these stands, all tree size classes (regeneration, intermediate, and overstory trees) were infected by H. parviporum. The average number of trees and stumps infected by a single genet ranged from 3 to 6.3 (mean 4.4) among study plots. All Heterobasidion genets identified from overstory trees or stumps had spread to the younger tree generation. Secondary infection from overstory trees was the main way of infection (at least 85% of all infections) among the regeneration and intermediate trees. The results indicate that uneven-aged management strategies that maintain continuous spruce regeneration favour the secondary spread of H. parviporum between different tree size classes and may compromise the production of high-quality timber over successive generations.
Full-text available
The purpose of the study was to analyse height growth, mortality, and ingrowth of individual small-sized trees in uneven-aged spruce-dominated stands. It was based on experimental data from 16 stands for a 15-year observation period including four measurements with a 5-year interval. In the data of this study, the heights of small-sized trees varied from 0.1 to about 9 m. Results showed that the growth of small trees was rather slow, particularly in the smallest size classes. With average growth rates it would take about 60 years for a freshly emerged spruce germinant to achieve 1.3 m in height. The stand density, site quality and selection cuttings affected the growth of small-sized spruces. Average five-year mortality rates for spruce, birch and pine were 17.0%, 40.9% and 33.9%, respectively. Annual ingrowth rates with the threshold height of 1.3 m for the three species were on average 30.4, 2.8 and less than 0.1 trees per hectare, respectively. Even if, a selection cutting of modest intensity (25% of basal area removed on average) seemed to have accelerated the growth of small spruces, it is recommended that more intensive harvestings be applied to enhance the survival and growth of small spruces. It is also concluded that shade intolerant species like birch and pine do not seem to be capable of developing into viable undergrowth in spruce selection stands with their current levels of stand density.
This paper deals with the problem of tree damage in a remaining stand. Two models were used for assessment of stand damage over the entire production period. Damage accumulates on the trees and in the stand, which is why the total share of damage tends towards the limit 100% if the number of thinnings increases. Parameters used in the models were established partly by previous field measurements and partly by simulations measurements. Motor-manual and cut-to-length technologies were analysed and compared according to the total number of damaged trees and the structure of trees according to the number of injuries. It was found that motor-manual technology causes more damage to trees and results in worse tree structure, meaning more than one injury. Many unanswered questions provide plenty of opportunities for further research.
Injuries and mortality to advance growth (saplings) after selection harvesting was studied in 17 multistoried Norway spruce (Picea abies (L.) Karst.) stands. Harvest removals ranged from 33 to 67% of initial basal area. Four of the stands were harvested with a motor-manual method (chain saw and skidding with farm tractors; M-FT). The remaining stands were harvested with single-grip harvesters and forwarders (H-FW). In each stand, injury rates were evaluated on a 24 x 48 m plot, located between the centre lines of two parallel strip roads that were spaced 24 m apart. All logging teams had at least 5 years of experience in clear-cutting and thinning operations. The trees to be removed and the strip road centre lines were marked prior to harvest. Mortality varied from 5 to 51%, whereas total injury (injured + dead saplings) varied from 17 to 76%. Mortality and injury levels were generally highest on H-FW plots. Crown reduction and leaning stems were the most frequent types of injury, regardless of operating method. Injury rates increased with sapling height with the H-FW method, whereas the opposite was found on M-FT plots. Saplings without preharvest damage in the form of top or leader defects had a higher probability of being injured than saplings with such damage in stands harvested with the M-FT method. A similar difference was not found on H-FW plots. A logistic regression model shows that the spatial risk of injury depends on the interaction between forest condition factors and operational characteristics. Forest condition factors influencing the risk of injury are sapling height and the location of saplings relative to larger residual trees and strip roads. Corresponding operational characteristics are operating method and harvest intensity.
The aim of the study was to model the regeneration establishment, and the survival and development of established seedlings in uneven-aged, Norway spruce (Picea abies (L.) H. Karst.) forest stands in southern Finland. Models were based on a 5-year monitoring period of permanent sample plots established in forest stands managed by selective harvests. Separate models were obtained for spruce, birch (Betula pendula Roth and Betula pubescens Ehrh.) and aspen (Populus tremula L.), all of which were well represented in the data. Since the stand structure of uneven-aged forests is irregular, spatial variation in competition factors and species composition were considered in the modelling. Due to the spatially hierarchical correlation structures of the data, a mixed modelling technique was applied in the estimation of the model parameters. Establishment models were developed for the number and initial height of established seedlings that had attained the set threshold height limit of 10cm during a 5-year period and thereafter survived until the end of the period. Prediction models were also developed for the survival, height increment and diameter at breast height of juvenile trees that were at least 10cm tall at the beginning of the 5-year period. The latter model is needed only when juvenile tree models are replaced with models for advanced trees. Simulations conducted for demonstration and validation purposes revealed that the regeneration establishment and juvenile growth models can be applied to construct a simulation system for the optimisation of uneven-aged spruce stands in southern Finland.