ArticlePDF Available

Strong Disturbance Impact of Tropical Cyclone Lionrock (2016) on Korean Pine-Broadleaved Forest in the Middle Sikhote-Alin Mountain Range, Russian Far East

Authors:

Abstract and Figures

Tropical cyclones (hurricanes and typhoons) cause large-scale disturbances in forest ecosystems all over the world. In the summer of 2016, a strong tropical cyclone named Lionrock created windthrow patches in the area of more than 400 km 2 on the forested eastern slopes of the Sikhote-Alin Range, in the Russian Far East. Such large-scale forest destruction by wind had never been recorded in the area prior to this event. We examined the tropical cyclone impact upon the forest composition, structure and tree mortality rates on two study sites (1 ha and 0.5 ha in size)-a contiguous windthrow patch site, and a site with partial canopy damage. Korean pine (Pinus koraiensis Siebold and Zucc.), Manchurian fir (Abies nephrolepis Trautv.) and Dahurian larch (Larix cajanderi Mayr.) were the primary tree species represented in the affected forest communities. Combined with the partial canopy damage, 7.7% of trees were blown down by the disturbance event. We determined that this one event mortality rate nearly equaled the average mortality rate for a ten year period for these forests (8.5 ± 4.0%) under normal conditions (no large-scale disturbances). Within a contiguous windthrow patch, tree mortality was determined to be 52.6%, which is significantly higher than the cumulative tree loss for the previous 50 years (42.4%). A substantial portion of thinner-stemmed trees (DBH (diameter measured at breast height) < 30 cm) were wind snapped, and those with larger diameters (DBH > 60 cm) were uprooted. Our results indicate that the probability of tree loss due to catastrophic wind loads increases as a result of the decrease in local density. We believe that tree loss estimates should include the impacts within contiguous patches of windthrows, as well as the patches with only partial tree canopy damage. Strong wind impact forecasting is possible with accounting for species composition within the stand sites and their spatial structure.
Content may be subject to copyright.
Article
Strong Disturbance Impact of Tropical Cyclone
Lionrock (2016) on Korean Pine-Broadleaved
Forest in the Middle Sikhote-Alin Mountain
Range, Russian Far East
Anna S. Vozmishcheva 1, 2, *, Svetlana N. Bondarchuk 3, Mikhail N. Gromyko 3,
Dmitriy E. Kislov 1, Elena A. Pimenova 1, Michail A. Salo 3and Kirill A. Korznikov 1
1
Botanical Garden-Institute of the Far Eastern Branch of the Russian Academy of Sciences, Vladivostok 690024,
Russia; kislov@easydan.com (D.E.K.); pimenova_garden@mail.ru (E.A.P.); korzkir@mail.ru (K.A.K.)
2Institute of Ecology and Geography, Siberian Federal University, Krasnoyarsk 660041, Russia
3Sikhote-Alin State Nature Biosphere Reserve n.a. K.G. Abramov, Terney 692150, Russia;
bonsal@mail.ru (S.N.B.); gromyko.zap@yandex.ru (M.N.G.); salo_mihail@mail.ru (M.A.S.)
*Correspondence: vozmishcheva@inbox.ru
Received: 6 October 2019; Accepted: 12 November 2019; Published: 13 November 2019


Abstract:
Tropical cyclones (hurricanes and typhoons) cause large-scale disturbances in forest
ecosystems all over the world. In the summer of 2016, a strong tropical cyclone named Lionrock
created windthrow patches in the area of more than 400 km
2
on the forested eastern slopes of the
Sikhote-Alin Range, in the Russian Far East. Such large-scale forest destruction by wind had never
been recorded in the area prior to this event. We examined the tropical cyclone impact upon the
forest composition, structure and tree mortality rates on two study sites (1 ha and 0.5 ha in size)—a
contiguous windthrow patch site, and a site with partial canopy damage. Korean pine (Pinus koraiensis
Siebold and Zucc.), Manchurian fir (Abies nephrolepis Trautv.) and Dahurian larch (Larix cajanderi
Mayr.) were the primary tree species represented in the aected forest communities. Combined with
the partial canopy damage, 7.7% of trees were blown down by the disturbance event. We determined
that this one event mortality rate nearly equaled the average mortality rate for a ten year period for
these forests (8.5
±
4.0%) under normal conditions (no large-scale disturbances). Within a contiguous
windthrow patch, tree mortality was determined to be 52.6%, which is significantly higher than the
cumulative tree loss for the previous 50 years (42.4%). A substantial portion of thinner-stemmed
trees (DBH (diameter measured at breast height) <30 cm) were wind snapped, and those with larger
diameters (DBH >60 cm) were uprooted. Our results indicate that the probability of tree loss due to
catastrophic wind loads increases as a result of the decrease in local density. We believe that tree loss
estimates should include the impacts within contiguous patches of windthrows, as well as the patches
with only partial tree canopy damage. Strong wind impact forecasting is possible with accounting for
species composition within the stand sites and their spatial structure.
Keywords:
forest structure; windthrow; wind; disturbance; mortality; tropical cyclone; Korean
pine; Lionrock
1. Introduction
Impact on the forest ecosystem caused by strong winds during typhoons (referred to as hurricanes
in the Western Hemisphere) have been reported in many studies [
1
,
2
]. Wind disturbances are recognized
to produce a key eect on forest dynamics and stability [
3
8
]. Strong winds are the cause of tree crown
damages that lead to tree death in temperate and boreal forests.
Forests 2019,10, 1017; doi:10.3390/f10111017 www.mdpi.com/journal/forests
Forests 2019,10, 1017 2 of 14
As a consequence, changes in forest structure, as well as tree stand and understory species
composition follow, accompanied by an increase of fire hazards and a decrease of carbon storage
capacity after salvage logging [
9
11
]. In contrast, tropical forests are located in the areas where
typhoons are more common, and tree mortality rates caused by strong wind are comparable to baseline
mortality in temperate and boreal forests. This can be explained by the complex structure of tropical
forest ecosystems adapted to regular heavy wind loads, where wind disturbances play a significant
role in the formation and maintenance of the forests’ high biodiversity [12,13].
The Russian Far East is located in the area aected by cyclones originating in the tropical parts
of the Pacific Ocean. Their biggest impacts fall upon the coast continental and island territories
of East Asia including China, Korea and Japan [
14
]. Gradually losing their intensity, the most
powerful tropical cyclones are capable of reaching 56–58 degrees north. In spite of the wind speed
reduction at high latitudes [
15
], where these typhoons are termed “tropical depressions” as per World
Meteorological Organization (WMO) classification, they are still able to significantly impact boreal
forest ecosystems [
16
]. Strong winds will be able to make considerable eect on boreal forest ecosystems
in the future, since frequency and poleward migration of tropical cyclones increased considerably
during the last 50 years [17,18] due to global climate changes [19,20].
Massive windthrow events in the Middle Sikhote-Alin Mountain Range (Primorsky Region,
Russia) were caused by tropical cyclone Lionrock that began its activity on 15 August, 2016 in the
western part of the Pacific Ocean. Gradually increasing in strength, the tropical depression approached
the east coast of Japan on 19 August. The typhoon changed its trajectory twice, and then crossed the
territory of Japan entering from the Pacific coastal side on the 29th of August, causing much destruction
and economic loss. On the night of 31 August (local time, Greenwich Mean Time +10), tropical cyclone
Lionrock reached the coastline of the continent (the eastern macroslope of the Middle Sikhote-Alin
Range, Russian Far East) with wind gusts of 30 m
·
s
1
(108 km
·
h
1
or 67 mph) and subsequently moved
further to Northeast China, transforming into an extratropical cyclone [
21
]. Cumulative precipitation
for the cyclone event amounted to 70 mm [
22
]. According to various estimates, the overall area of
windthrows on the eastern macroslope of the Sikhote-Alin Range adds up to about 400 km
2
. Wind
disturbances of such scale had never been recorded in the Primorsky Region.
Recently, a number of authors reported on the probable shift of tropical cyclone trajectories to the
North [
18
,
23
25
], which is attributed to global climate change. Thus, the probability of strong wind
impact on temperate and boreal forest dynamics in the south of the Russian Far East, previously free of
such extreme events, will possibly increase in the near future.
In this paper, we report on our study to estimate the impact of tropical cyclone Lionrock on the
structure of the zonal forests of the region, based on observations within the windthrow patch. The core
of these analyses was an evaluation of four hypotheses related to forest impacts as follows: (1) Tree
death probability is species dependent; (2) tree death probability is size dependent; (3) hollowed and
rotten trees are more likely to fall; (4) windthrow distribution is grouped into patches. To test these
hypotheses, we analyzed the spatial structure of a tree stand on a one ha permanent plot which was
established at a location of strong wind disturbance. We also provided a quantitative assessment of the
cyclone-damaged trees.
2. Materials and Methods
2.1. Study Area
This study was conducted in the Sikhote-Alin Nature Reserve (45
20
0
N, 136
09
0
E) located in
the central part of the Sikhote-Alin Range, Russian Far East (Figure 1and Figure S1). The climate of
this region is monsoon with a cold dry winter, rainy summer and seasonal changes in wind direction.
For the period 1981 to 2019, the mean annual temperature was +3.9
C. January was the coldest month
(
12.3
C average) and August was the warmest (+18.0
C average). Mean annual precipitation was
828 mm, most of which fell from May to September [26].
Forests 2019,10, 1017 3 of 14
Forests 2019, 10, x FOR PEER REVIEW 3 of 15
°C average) and August was the warmest (+18.0 °C average). Mean annual precipitation was 828 mm,
most of which fell from May to September [26].
According to the KöppenGeiger climate classification, the climate type here is humid
continental, characterized by a warm summer (Dwb) [27]. The bioclimate type is temperate/boreal
marine [28].
Figure 1. Of the permanent plots PP1 (first permanent plot)and PP2 (second permanent plot)(b) in
the Sikhote-Alin Nature Reserve (SANR). The dashed line (c) stands for tropical cyclone Lionrock
trajectory. Forest areaswith windtrow disturbance are denoted as black (a) (based on data provided
by the GlobalForestWatch: https://www.globalforestwatch.org/).
Zonal forest communities are represented by the Korean Pine-broadleaved forest (Pinus
koraiensis Siebold and Zucc.). Mountain forests within the Sikhote-Alin Nature Reserve boundaries
have not suffered catastrophic disturbances in recent times and therefore, the old-growth forest
ecosystems are dominated by Pinus koraiensis. Other species, including Picea ajanensis Fisch. and Carr.,
Abies nephrolepis (Trautv.) Maxim., Larix cajanderi Mayr,Tilia amurensis Rupr., Quercus mongolica Fisch.
ex Ledeb., Acer mono Maxim. ex Rupr. and Betula costata Trautv. are also present. Understory species
include Corylus mandshurica Maxim., Ribes maximoviczianum Kom., Lonicera chrysantha Turcz. ex
Ledeb, L. maximowiczii (Rupr.) Regel, Rosa acicularis Lindl., Spiraea betulifolia Pall., and Actinidia
kolomikta (Maxim. andRupr.) Maxim.
2.2. Field methods
The first permanent plot (PP1, 100 m × 100 m) was established in 2017 within the windthrow
patch of the mountain old-growth Korean Pine-broadleaved forest at 270 m a.s.l. on a slope with
Figure 1.
Of the permanent plots PP1 (first permanent plot)and PP2 (second permanent plot) (
b
) in
the Sikhote-Alin Nature Reserve (SANR). The dashed line (
c
) stands for tropical cyclone Lionrock
trajectory. Forest areaswith windtrow disturbance are denoted as black (
a
) (based on data provided by
the GlobalForestWatch: https://www.globalforestwatch.org/).
According to the Köppen–Geiger climate classification, the climate type here is humid continental,
characterized by a warm summer (Dwb) [27]. The bioclimate type is temperate/boreal marine [28].
Zonal forest communities are represented by the Korean Pine-broadleaved forest (Pinus koraiensis
Siebold and Zucc.). Mountain forests within the Sikhote-Alin Nature Reserve boundaries have not
suered catastrophic disturbances in recent times and therefore, the old-growth forest ecosystems
are dominated by Pinus koraiensis. Other species, including Picea ajanensis Fisch. and Carr., Abies
nephrolepis (Trautv.) Maxim., Larix cajanderi Mayr,Tilia amurensis Rupr., Quercus mongolica Fisch. ex
Ledeb., Acer mono Maxim. ex Rupr. and Betula costata Trautv. are also present. Understory species
include Corylus mandshurica Maxim., Ribes maximoviczianum Kom., Lonicera chrysantha Turcz. ex Ledeb,
L. maximowiczii (Rupr.) Regel, Rosa acicularis Lindl., Spiraea betulifolia Pall., and Actinidia kolomikta
(Maxim. andRupr.) Maxim.
2.2. Field Methods
The first permanent plot (PP1, 100 m
×
100 m) was established in 2017 within the windthrow
patch of the mountain old-growth Korean Pine-broadleaved forest at 270 m a.s.l. on a slope with
southern exposure. The study was conducted in 2017–2018. After marking the permanent plot with
laser range finder TruPulse 200 (USA), we determined local coordinates within 0.01 m (1 cm) of all
trees having a diameter measured at breast height (DBH) greater than or equal to 6 cm. For dead trees,
DBH was also measured and the type of mortality noted as uprooted or snapped. Locations of the
uprooted trees were determined at the center of their root pits, while the coordinates for the snapped
trees were taken at the base of the trunk. Both height and DBH were measured for each living tree
with the tree health status (healthy or damaged). The trees killed prior to cyclone Lioncrock were
Forests 2019,10, 1017 4 of 14
recorded separately. For Pinus koraiensis and Larix cajanderi trees, both living and dead, core samples
were collected using increment borers (Haglof, Sweden). Each sample was inspected visually for stem
rot. Cores were collected only from medium to large canopy trees to be sure that these trees fell due to
the direct tropical cyclone influence, and not because of the impact of other falling trees.
The baseline mortality rate of Korean Pine-broadleaved forest was estimated on the second 0.5 ha
permanent plot (PP2, 50 m
×
100 m) established in 1967 and located at a 100 m distance from the
windthrow patch used as the described above permanent plot. There was no massive tree disturbance
by tropical cyclone Lionrock in this part of the forest. We used available data on tree mortality and
recorded diameters (DBH
6 cm) observed every 5–10 years for the 50 years period 1967 to 2016
inclusive, as well as most recent data collected in 2017. Then, we calculated mean baseline mortality
as loss of stem density and total basal area values per each 10 year relative to total stem density and
total basal area at the time of measurement in 1967 (data converted to one ha). Subsequently, the mean
10-year baseline mortality rate was calculated and accepted as a natural disturbance regime without
large-scale windthrow events and later compared with tropical cyclone-induced mortality on PP2.
2.3. Statistical Analyses
We analyzed the size structure of each tree species (DBH distribution) with a sucient number of
individuals: Abies nephrolepis (n=383), Pinus koraiensis (n=246), Larix cajanderi (n=35). Broadleaved
species (Acer mono,Betula costata,Quercu smongolica and Tilia amurensis Mill., n=81) were combined,
and their overall size structure was also calculated. Stem density and basal area of the tree stand before
and after the disturbance were also taken into account. Significance of dierences between mean
values was estimated using t-test. Before applying t-test, all data samples were tested against normality
using the Kolmogorov-Smirnov test. Further, using t-test, we compared pre- and post-disturbance tree
populations, healthy and damaged ones, and finally, uprooted and snapped tree populations.
Chi-squared test (
χ
2) was applied to test dependence between tree diameter and stem rot presence.
Also, we used this same chi-square test to discover the relationship between the local density of spatial
tree distribution (kernel density estimation was used) and the probability of a tree to be either uprooted
or snapped. Trees were separated into tree size classes depending on their diameter: small (6–29 cm),
medium (30–60 cm), and large (>60 cm). Based on the Bernoulli trial scheme, tree mortality probability
(P
±
), as dependent on size class and species, was calculated with or without the type of death
categorization [29].
Probability assessments of tree fall during the cyclone event and its dependency on local
environmental conditions were based on logistic regression [
30
]. The following set of predictors were
examined: (1) tree diameter at breast height (DBH); (2) tree height (H); (3) local density of tree stand
composition (more precisely, probability density estimation of tree presence) (P); (4) presence of a
hollow in wood (W). Missing height values (snapped trees) were estimated using a Curtis model
which describes the relationship between tree diameter and its height [
31
]. The nonlinear Curtis
regression model was fitted to all available data obtained by combining datasets from both PP1 and
PP2. To estimate the local density of tree distribution, we applied a 2D kernel density estimator (kde2d
function) from the MASS package available in R [
32
]. Before fitting logistic regression, we tested data
against multicollinearity using the omcdiag function from the mctest R-package. To get assessments
on spatial tree structure we calculated Ripley’s L(r)- and L
12
(r)-functions [
33
]. The first is Ripley’s
K-function normalized to be equal to zero for the Poisson point process, that was chosen as a model
for complete spatial randomness. The second, the L
12
(r)-function, is a generalization of the first one,
allowing researchers to make decisions regarding the mutual aggregation and segregation of points
belonging to dierent virtual types 1 and 2. If L
12
(r) tends to be small, points of type 1 and 2 spatially
Forests 2019,10, 1017 5 of 14
avoid each other, and otherwise they are mutually aggregated. Ripley’s L(r)- and L
12
(r)-functions were
used to describe spatial univariate and bivariate patterns of the trees according to Equations (1) and (2):
ˆ
L(r)=sA
n2Pn
i=1Pn
j=1δriji,j
πr(1)
ˆ
L12(r)=v
tA
npnqPnpnq
i=1Pn
j=1δriji,j
πr(2)
where Ais the permanent plot area, nis the number of trees, n
p
and n
q
are the number of points in
class 1 and class 2,
δ
(r
ij
)then is the indicator function of the mean number of neighbors within a circle
around each tree with the distance rbeing the radius, and rij is the distance from tree ito tree j.
The Monte-Carlo method with 999 generations was used for the construction of confidence limits
(null hypothesis acceptation area), null hypothesis bias estimation and to a get significant value of
1%. Functions L(r) and L
12
(r) were calculated using Programita software (http://programita.org/), [
34
].
The Null hypothesis of complete spatial randomness (univariate analysis) was tested to compare tree
spatial structure before and after cyclone Lionrock impact (pre-disturbance and post-disturbance tree
populations). Further on, the spatial patterns of uprooted and snapped trees were analyzed. Function
value L(r) within confidence limits indicates the random distribution of trees within the analyzed
distances. Values above the upper confidence limit indicate tree aggregation, while those below the
lower confidence limit show a segregation of trees. Distance values between trees, critical to the
downfall of living trees caused by the fall of dead trees and resulting in considerable destructions,
were estimated using antecedent conditions hypothesis (bivariate analysis). The function value L
12
(r)
within confidence limits indicates the independent interaction between the spatial distribution of trees
belonging to nominal categories “1” and “2”. If the L
12
(r) value falls above the upper confidence limit,
positive correlation takes place (trees belonging to nominal types “1” and “2” tend to be distributed
close to each other). If this L
12
(r) value falls below the lower confidence limit, trees avoid each other
in their spatial distribution. In our study, we investigated spatial relationships between uprooted
(category “1”) and snapped (category “2”) trees.
3. Results
We detected 364 live trees of 12 species and 404 (52.6%) trees that were fallen on the PP1 permanent
plot (Table 1). Most of them were Abies nephrolepis,Pinus koraiensis (by the level of stem density); Pinus
koraiensis and Larix cajanderi (by the level of basal area). Considerable damages to tree crowns and
stems of 13.8% trees (6.85 m
2·
ha
1
) were also revealed, which will contribute to the total basal area loss
in the near future, bringing it up to 66.4% from the today’s 52.6% due to damaged tree mortality.
Of the 404 dead trees, 157 (39.2% mortality) can be attributed to having been snapped, and 247
(60.8%) to being uprooted trees (Table 2). Fall probability showed no correlation with trees’ DBH.
Mean DBH with standard deviation (SD) of all living and killed trees was found to be 25.5 cm and
23.6 cm, respectively (t-test, p=0.08). At the same time, the mean size of uprooted trees (30.2 cm) was
considerably higher (t-test, p0.01) than the mean DBH of those snapped trees (18.3 cm).
Forests 2019,10, 1017 6 of 14
Table 1.
Overall tree density and basal area of the trees alive prior to tropical cyclone Lionrock, as well as those damaged during the tropical cyclone activity, and
cyclone-killed trees in the old-growth Korean Pine-broadleaved forest, Middle Sikhote-Alin.
Species Density, Stem·(ha1)Density, % of Overall Basal Area, m2·ha1Basal Area, % of Overall
Overall Damaged Cylone-killed Damaged Cylone-killed Overall Damaged Cylone-killed Damaged Cyclone-killed
Betula costata 16 4 11 0.5 1.4 0.23 0.03 0.15 0.1 0.3
Pice aajanensis 28 6 7 0.8 0.9 0.91 0.15 0.34 0.3 0.7
Pinus koraiensis 246 33 140 4.3 18.2 33.84 4.54 19.63 8.8 37.9
Quercus mongolica 15 7 4 0.9 0.5 2.06 1.04 0.61 2.0 1.2
Tilia amurensis 26 8 7 1.0 0.9 1.43 0.47 0.26 0.9 0.5
Larix cajanderi 34 0 24 0.0 3.1 7.15 0.00 5.38 0.0 10.4
Abies nephrolepis 381 38 204 4.9 26.6 5.92 0.41 3.92 0.8 7.6
Acer mono 6 1 4 0.1 0.5 0.02 0.00 0.01 0.0 0.0
Acer ukurunduense 9 4 1 0.5 0.1 0.06 0.04 0.00 0.1 0.0
Betula platyphylla 1 1 0 0.1 0.0 0.10 0.10 0.00 0.2 0.0
Prunus
maximowiczii 5 3 2 0.4 0.3 0.07 0.05 0.02 0.1 0.0
Sorbus amurensis 1 1 0 0.1 0.0 0.01 0.01 0.00 0.0 0.0
Total 768 106 404 13.8 52.6 51.80 6.85 30.32 13.2 58.5
Overall—live trees at the time of the tropical cyclone Lionrock, and cyclone-killed—uprooted and snapped trees together.
Table 2.
Number, DBH (tree diameter at breast height ) (mean
±
SD (standard deviation)) and probability estimation (P
±
) of snapped, uprooted and total killed
trees on PP1 (first permanent plot) per 1 ha of strong disturbance resulting from tropical cyclone Lionrock, Middle Sikhote-Alin.
Size Class (cm) Snapped Trees Uprooted Trees Total Killed Trees
No DBH P±No DBH P±% DBH P±
10–30 133 13.9 ±4.9 0.17 ±0.03 114 13.3 ±5.5 0.15 ±0.02 32.0 13.6 ±5.2 0.32 ±0.03
30–60 21 35.8 ±7.8 0.03 ±0.01 109 40.7 ±9.2 0.14 ±0.02 16.9 39.9 ±9.1 0.17 ±0.03
>60 5 66.9 ±13.5 0.006 ±0.005 24 63.3 ±6.8 0.03 ±0.01 3.8 63.9 ±8.2 0.38 ±0.03
Total 159 18.3 ±12.7 0.21 ±0.03 247 30.2 ±18.6 0.32±0.03 52.7 25.5 ±17.5 0.53 ±0.03
Forests 2019,10, 1017 7 of 14
Results of chi-square test of the tree size classes distribution showed deviation from uniform for
the killed, snapped and uprooted trees (p
0.05). There was no statistically significant dierence found
in tree size distribution in pre-disturbance and post-disturbance tree populations. The same statement
is true for healthy and damaged trees, as well as for all species considered together and each taken
separately (Figure 2; Kolmogorov-Smirnov two-sample test, p<0.05). However, the size distribution
of uprooted trees diered from that of snapped trees (Kolmogorov-Smirnov two-sample test, p<0.05).
In addition, the frequency of snapped trees was higher within the small size class DBH <30 cm, while
the frequency of uprooted trees was higher within the large size class DBH >60 cm (Table 2). Snapped
trees of the understory Abies nephrolepis layer were larger in diameter (DBH =15.2 cm) than the mean
value for the species (12.5 cm, t-test, p
0.01). Diameter of the uprooted A.nephrolepis (DBH =12.9 cm)
did not show considerable dierence from the overall mean for the species (t-test, p=0.59). The DBH
of snapped and uprooted trees of Pinus koraiensis and Larix cajanderi, as well as the broadleaved species,
did not dier from the tree size distribution before the Lionrock cyclone’s impact (t-test, p0.05).
The baseline mortality level during the 50-year period prior to tropical cyclone Lionrock in the
Korean Pine-broadleaved forest was 42.4% of stem density or 26.5% of total basal area (Table 3). Mean
baseline mortality per 10 years increased insignificantly during the study period, not exceeding the
values of 13.7% of stem density and 7.2% of total basal area per 10 years. The mortality rate resulting
from tropical cyclone Lionrock equals that for a 10 year period.
Table 3.
Baseline mortality rates on PP2 over 50 years before the Lionrock cyclone’s disturbance and tree
loss caused by the tropical cyclone in old-growth Korean Pine-broadleaved forest, Middle Sikhote-Alin.
Year Density Basal Area
Stem·(ha1) Percent·10 year1m2·ha1Percent·10 year1
1967–1976 46 3.6 1.2 2.4
1977–1986 88 6.9 2.4 5.0
1987–1996 88 6.9 2.4 5.0
1997–2006 156 13.7 3.4 7.2
2007–2016 128 11.2 3.2 6.9
2017dead trees 88 7.7 3.0 6.4
2017_damaged trees 20 1.8 1.4 2.8
Mean (1967–2016) 101 ±42 8.5 ±4.0 2.5 ±0.9 5.3 ±2.0
Total (1967–2016) 506 42.4 12.4 26.5
After the analysis of 147 trees of Pinu skoraiensis and Larix cajanderi (65% of total number of
medium and large size trees for these species, before the Lionrock cyclone’s impact), we detected that
dead trees were less aected by stem rot in comparison with living trees, regardless of damage type
and species (Table 4).
Table 4.
The ratio of living to killed trees with and without stem rot (hollow) in the old-growth Korean
Pine-broadleaved forest after the Lionrock cyclone’s impact, Middle Sikhote-Alin.
Tree Number Proportion (%)
Healthy Hollow Healthy Hollow
Larixcajanderi
alive 5 3 62.5 37.5
snapped 1 0 100.0 0.0
uprooted 6 1 85.7 14.3
Pinuskoraiensis
alive 40 21 65.6 34.4
snapped 4 1 80.0 20.0
uprooted 46 19 70.8 29.2
Total
alive 45 24 65.2 34.8
snapped 5 1 83.3 16.7
uprooted 52 20 72.2 27.8
Forests 2019,10, 1017 8 of 14
Figure 2.
Tree size distribution of pre-disturbance and post-disturbance populations on PP1 in the
old-growth Korean Pine-broadleaved forest, Middle Sikhote-Alin. The distribution of trees of both
disturbance types (healthy and damaged trees) is shown, as well as the ratio of killed (uprooted
and snapped together) and post-disturbance living trees (healthy and damaged together) for each
tree species.
A logistic regression model was used to examine the influence of a set of parameters on tree
mortality probability caused by tropical cyclone Lionrock. Using the omcdiag helper function, it was
found that the covariance matrix determinant significantly (minimum significance level 0.05) diers
from zero in all studied cases (Table 5).
Table 5.
Logistic regression model of cyclone-induced tree mortality. Accuracy evaluation was
performed using leave-one-out cross-validation (LOOCV) scheme.
Logistic Regression Coecients Estimate
Species Intercept DBH PH W LOOCV Killed
Trees (%)
All trees case
Pinus koraiensis 6.29*** 0.17*** 599.0* 0.51*** - 0.85 0.7
Abies nephrolepis 0.37 0.21*** 998.3*** 0.12 - 0.66 0.63
Larix cajanderi 0.56 0.006 1595* 0.064 - 0.54 0.71
Broadleaved species 4.24*** 0.12*** 1676* 0.36 - 0.77 0.71
Total 0.29*** 0.12*** 610.0*** 0.26*** - 0.71 0.66
Hollow presence trees case
Pinus koraiensis 6.82*** 0.22*** 581 0.61*** 0.27 0.86 0.66
Larix cajanderi 80.037 533 0.33 1.9 0.31 0.5
Total 5.28*** 0.19*** 648 0.49*** 0.13 0.8 0.64
Predictors: DBH—tree diameter at breast height; H—tree height; P—local probability density estimation;
W—existence of a hollow in a tree wood. Significance: ***p<0.001, *p<0.05.
Using fewer numbers of predictors and all possible combinations did not improve the model. Use
of the stem rot infection parameter allowed for a 5% change in model precision for Pinus koraiensis
(accuracy changes from 15% to 20%), and did not change it for Larix cajanderi.
The spatial structure of trees present before the disturbance event is determined to be grouped
at all distance levels up to 38 m (Figure 3). Uprooted trees resulting from tropical cyclone Lionrock
Forests 2019,10, 1017 9 of 14
were also grouped. The maximum distance of the uprooted tree distribution patterns was 39 m (about
4800 m
2
). The spatial structure of snapped trees is more regular than the distribution of uprooted trees.
The results of bivariate analysis showed the impact of uprooted trees on snapped and damaged trees
on small scales (up to 3 m and 8 m, respectively).
Forests 2019, 10, x FOR PEER REVIEW 4 of 15
Figure 3. Ripley’s L(r)-function patterns for pre-typhoon, typhoon-killed and typhoon-damaged tree
populations in the Middle Sikhote-Alin. Dashed lines of 999 randomly generated processes indicate
99% confidence limits.
4. Discussion
On the eastern macroslope of the Sikhote-Alin Mountain Range, we estimated the total area of
windthrow patches equalled 400 km
2
, of which, 20% occurred in the Korean Pine-broadleaved forest
[35]. On the Sikhote-Alin Nature Reserve territory, tropical cyclone Lionrock formed a large-scale
mosaic of complete windthrow patches (several ha in size) and forested areas with the fall of
individual trees.
Tropical cyclones in the high latitudes of northeastern Asia are typically rare, because they are
blocked by the cold air masses of the Sea of Okhotsk. As a result, when a rare event such as this occurs
it may play a much more significant role in the forest ecosystem dynamics of the mainland coastal
and insular areas of the Russian Far East. Although the results of our study represent an isolated and
first example of a tropical cyclone impact (manifesting itself in strong wind gusts and heavy
precipitation within a short time period) on the forest ecosystems of the Sikhote-Alin Mountain
Range, it provides a valuable general understanding of the possible impact levels capable of causing
disturbance events in this region.
Factors that play a key role in tropical cyclone impact on forest ecosystems differ from each other
depending on the scale of the study. Wind speed and total precipitation are considered to be the main
factors at the regional scale [36,37], while at the landscape scale, it is the overall topographic site-
specific characteristics and plant community-related factors that play the key role [38]. Tree size,
species and tree health can influence the rate and type of mortality at the community scale [39,40].
During tropical cyclone Lionrock, the maximum wind speed recorded at the nearest weather station
(Ternei) was 30 m·s
1
. However, our study plots were located 500 m higher in elevation and 10 km
away from the station. A number of studies [41] indicate that wind strength increases with elevation,
therefore making it impossible to extrapolate the wind speed registered at the weather station to the
actual study areas.
The results of our study show that the impact of the tropical cyclone Lionrock led to the
appearance of windthrow patches, decreasing tree stem density by 52.6% and basal area by 58.5%.
Figure 3.
Ripley’s L(r)-function patterns for pre-typhoon, typhoon-killed and typhoon-damaged tree
populations in the Middle Sikhote-Alin. Dashed lines of 999 randomly generated processes indicate
99% confidence limits.
4. Discussion
On the eastern macroslope of the Sikhote-Alin Mountain Range, we estimated the total area of
windthrow patches equalled 400 km
2
, of which, 20% occurred in the Korean Pine-broadleaved forest [
35
].
On the Sikhote-Alin Nature Reserve territory, tropical cyclone Lionrock formed a large-scale mosaic of
complete windthrow patches (several ha in size) and forested areas with the fall of individual trees.
Tropical cyclones in the high latitudes of northeastern Asia are typically rare, because they are
blocked by the cold air masses of the Sea of Okhotsk. As a result, when a rare event such as this occurs
it may play a much more significant role in the forest ecosystem dynamics of the mainland coastal and
insular areas of the Russian Far East. Although the results of our study represent an isolated and first
example of a tropical cyclone impact (manifesting itself in strong wind gusts and heavy precipitation
within a short time period) on the forest ecosystems of the Sikhote-Alin Mountain Range, it provides a
valuable general understanding of the possible impact levels capable of causing disturbance events in
this region.
Factors that play a key role in tropical cyclone impact on forest ecosystems dier from each other
depending on the scale of the study. Wind speed and total precipitation are considered to be the
main factors at the regional scale [
36
,
37
], while at the landscape scale, it is the overall topographic
site-specific characteristics and plant community-related factors that play the key role [
38
]. Tree size,
species and tree health can influence the rate and type of mortality at the community scale [
39
,
40
].
During tropical cyclone Lionrock, the maximum wind speed recorded at the nearest weather station
(Ternei) was 30 m
·
s
1
. However, our study plots were located 500 m higher in elevation and 10 km
away from the station. A number of studies [
41
] indicate that wind strength increases with elevation,
Forests 2019,10, 1017 10 of 14
therefore making it impossible to extrapolate the wind speed registered at the weather station to the
actual study areas.
The results of our study show that the impact of the tropical cyclone Lionrock led to the appearance
of windthrow patches, decreasing tree stem density by 52.6% and basal area by 58.5%. Everham and
Brokaw [
10
] reported that trees significantly damaged by strong winds die o. In this study, we found
that the proportion of standing cyclone-damaged trees was 13.8%. Thus, taking into account direct and
indirect “losses”, we come to the conclusion that the Lionrock cyclone’s impact on the forest ecosystem
was quite large, with more than 70% loss of pre-disturbance living trees within windthrow patches.
We established that the tree mortality rate corresponds to the similar previously estimated
disturbances from strong winds, storms, hurricanes and typhoons found in temperate and boreal
forest ecosystems. Foster [
4
] provides information about the 75% mortality rate in New England
forests in the northeastern United States. Sheeld and Thompson [
42
] recorded a 66% mortality in
South Carolina (U.S.) after Hurricane Hugo. Mortality levels ranged from 23.3% to 63.4% of stem
density and from 29.5% to 86.8% of basal area as a result of the 1999 catastrophic storm in boreal
forests in northeastern Minnesota [
40
]. Detailed studies of forest ecosystems in northeastern Poland
identified tree mortality of 49% of stem density and 48% of basal area caused by the 2002 hurricane [
43
].
Death rate of trees caused by typhoons is lower in tropical rainforests [
44
,
45
]. Thus, Bellingham [
46
]
found 8% tree mortality in Jamaican forests; Whigham, et al. [
47
] registered 11.2% in Mexico forests
23 months after Hurricane Gilbert, and Walker [
44
] discovered 7% of tree deaths in Puerto Rico during
a one year period following Hurricane Hugo.
In this study, we found that the baseline mortality of the forest under discussion was 4%–14%
of stem density and 2%–7% of basal area per 10 year period over the past 50 years. Similar baseline
mortality rate, determining the mode of natural disturbances, is typical of coniferous–hardwood forest
ecosystems of temperate and boreal forest zones [
48
]. Higher tree basal mortality rates in temperate
and boreal forests are explained by simpler spatial structure and homogeneous species composition.
Unlike tropical forests, the above forest ecosystems are not adapted to frequent disturbance events
caused by strong winds. According to recent data, mixed composition of tree stands in rainforests is
the reason for high forest sustainability [49].
A number of studies have shown that the key factor influencing tree susceptibility to wind is
the tree size; large trees being more prone to fall [
50
]. Mortality of small trees may be low due to
the shielding eect of big trees [
51
]. On the other hand, there is evidence that strong winds cause
most damage to mid-size trees, since small trees are shielded by the canopy and large trees are more
resistant to gusts [
9
]. According to our analysis, the overall tree loss, resulting from tropical cyclone
Lionrock, impacted all size classes of trees rather equally. Overall windfall frequency of large trees
diered insignificantly from that of smaller size trees.
It is known that tree disturbance type (uprooted and snapped) is species and tree size
dependent [
52
]. In this study, the ratio of uprooted trees was generally much higher than of
snapped trees. Large trees (DBH >60 cm) were recorded into the uprooted category more (P=0.55
±
0.15), while small trees’ damage was more frequently of the snapped category (P=0.27
±
0.04). Similar
results were obtained when analyzing wind damage complexes in eastern North America [
52
], the
Caribbean [
53
,
54
] and southeastern Slovenia [
55
]. We are inclined to assume that the uprooting eect
damage caused by tropical cyclone Lionrock is the result of strong wind, heavy precipitation combined
withthin the mountain soil layer, and the shallow surface root systems of the largest conifer trees Pinus
koraiensis and Larix cajanderi. Furthermore, large trees have a larger crown profile. Damages resulting
in tree-snapping could be considered the result of the mechanical action of larger falling trees.
Stem rot or hollow in trees is considered by a number of researchers to be a factor increasing tree
death probability due to its vitality deterioration and increased susceptibility to strong wind [
50
,
55
].
Our data does not allow us to make such conclusions for the Korean Pine-broadleaved forest, Middle
Sikhote-Alin. Among the surviving Pinus koraiensis and Larix cajanderi trees, the ratio of trees with rot
or hollow is significantly higher than among those killed during the tropical cyclone Lionrock. We
Forests 2019,10, 1017 11 of 14
assume that this result is the consequence of smaller crowns in trees with rot, which reduces their
windfall probability. Testing this hypothesis does not appear possible in our case due to the lack of
ability to identify dead tree crown size and aliation during field work.
According to the presently available evidence, patterns of canopy formation can vary significantly
in cases of partial mortality of trees (group distribution) and after catastrophic events accompanied by
strong winds (random distribution) [
48
]. On the other hand, the spatial distribution of uprooted trees
resulting from tropical cyclone impact may depend on their initial spatial patterns in the community [
37
].
In the analyzed forest site, we were able to establish that the large, uprooted trees were the cause of the
snapped trees in the area of their fall. At the same time, a more even distribution of the snapped trees
in relation to the uprooted trees at the distances of 1 cm to 50 m can be explained by the chaotic fall of
the latter. Within the surveyed Korean Pine-broadleaved forest site, we found an increase of canopy
gaps of up to 4800m
2
(1200 m
2
in most cases), which is similar to the spatial distribution patterns in
this area. For comparison, Xi et al. [
36
] notes that the impact of Hurricane Fran in North Carolina led
to the average gap sizes of 1100 m2.
Forest disturbance caused by tropical cyclone Lionrock was the strongest on record in the
Sikhote-Alin region. Our detailed study of the damaged stand on one of the windthrow patches of the
zonal Korean Pine-broadleaved forest enabled us to determine that the tree mortality rate we found
exceeded the baseline tree mortality established over a 50-year period, which proves the strong eect of
the tropical cyclone. In addition, tropical cyclone Lionrock significantly impacted forest communities,
not only within the contiguous windthrow patches (PP1), but also in other parts of the forest ecosystem
(PP2), where it was determined that tree death comparable to a ten-year mortality level occurred
(Figure S2). We found that large trees were more susceptible to uprooting, while small trees were more
susceptible to wind breaking. The largest trees in each size class were found to be more likely to die
regardless of the type of death.
Based on logistic regression models defined by the coecients presented in Table 5, we concluded
that d the ecrease of tree density in the community led to an increased death probability due to the
tropical cyclone. At the same time, the absence of visible stem rot, larger tree diameter and smaller
height contributed to the likelihood of tree mortality in the canopy layer. We assume that these
geometric parameters are typical for healthy, large trees.
However, their near-surface root systems contribute to the insucient resistance to strong tropical
cyclones accompanied by heavy precipitation and strong wind loads. High fall resistance of trees
in old-growth forest ecosystems is likely related to their grouped spatial distribution. This can also
indicate a decrease of stand stability in the case of repeated strong tropical cyclone action due to lower
group density, including forest communities with partial canopy violation (single falls).
Trees less susceptible to wind-induced mortality were thin-stemmed and tall, with signs of damage
from stem rot. The lack of fit in the logistic model for Larix cajanderi, both with the use of single
predictors and their combinations, is explained by the insucient data set, while the Abies nephrolepis
mortality (the largest data set) is determined by external factors and not by the analyzed tree metric.
5. Conclusions
In our study, we firstly made an attempt to estimate the impact of strong wind due to a tropical
cyclone on the natural Korean Pine-broadleavedforest community, Middle Sikhote-Alin. It is possible
that at the landscape level, the degree of damage was determined by the high, short-term precipitation
and maximum observed gusts of wind during the Lionrock disturbance event). At the community
scale, the tree characteristics (diameter and height), as well as tree density, were most important.
The results of our study reveal a significant impact of tropical cyclone Lionrock, not only within
the windthrow patches, but also for the forest communities that sustained the least amount of damage.
In spite of the disturbance of this magnitude remaining a single recorded event in the region, climate
change may contribute to an increased frequency of such events, or even to their regularity in the
future. In future studies directed at the structure and dynamics of the catastrophic consequences of
Forests 2019,10, 1017 12 of 14
strong typhoons, we recommend that portions of the forest ecosystem that sustain partial canopy
damage be included in the assessment.
Supplementary Materials:
The following are available online at http://www.mdpi.com/1999-4907/10/11/1017/s1,
Figure S1: Windthrow forest in the Middle Sikhote-Alin Mountain Range, Russian Far East; Figure S2: The character
of wind-caused tree damage at PP1 (a) and PP2 (b)
Author Contributions:
Author Contributions: Conceptualization, A.S.V.; Data curation, A.S.V. and S.N.B.; Formal
analysis, A.S.V. and D.E.K.; Funding acquisition, K.A.K.; Investigation, A.S.V., S.N.B., M.N.G., E.A.P., M.A.S. and
K.A.K.; Methodology, A.S.V. and D.E.K.; Project administration, A.S.V.;Validation, A.S.V., D.E.K. and K.A.K.;
Visualization, A.S.V., D.E.K. and K.A.K.; Writing—original draft, A.S.V., K.A.K. and D.E.K.; Writing—review and
editing, S.N.B., M.N.G., E.A.P. and M.A.S.
Funding: The work was supported by the Russian Science Foundation (project No.18-74-00007).
Acknowledgments:
Authors are deeply grateful to Viktoria Chilcote and Mark Chilcote for the translation editing
and valuable comments. Authors are grateful to two anonymous reviewers for working with the first version of
the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Foster, D.R.; Boose, E.R. Patterns of forest damage resulting from catastrophic wind in central New England,
USA. J. Ecol. 1992,80, 79–98. [CrossRef]
2.
Veblen, T.T.; Kulakowski, D.; Eisenhart, K.S.; Baker, W.L. Subalpine forest damage from a severe windstorm
in northern Colorado. Can. J. For. Res. 2001,31, 2089–2097. [CrossRef]
3.
Canham, C.D.; Loucks, O.L. Catastrophic windthrow in the presettlementforests of Wisconsin. J. Ecol.
1984
,
65, 803–809. [CrossRef]
4.
Foster, D.R. Species and stand response to catastrophic wind in central New England, USA. J. Ecol.
1988
,76,
135–151. [CrossRef]
5.
Peterson, C.J.; Pickett, S.T.A. Treefall and resprouting following catastrophic windthrow in an old–growth
hemlock–hardwoods forest. For. Ecol. Manag. 1991,42, 205–217. [CrossRef]
6.
Turner, M.G. Disturbance and landscape dynamics in a changing world. Ecology
2010
,9, 2833–2849.
[CrossRef]
7.
Chi, C.H.; McEwan, R.W.; Chang, C.T.; Zheng, C.; Yang, Z.; Chiang, J.M.; Lin, T.C. Typhoon disturbance
mediates elevational patterns of forest structure, but not species diversity, in humid monsoon Asia. Ecosystems
2015,18, 1410–1423. [CrossRef]
8.
Sommerfeld, A.; Senf, C.; Buma, B.; D’Amato, A.W.; Despr
é
s, T.; D
í
az–Hormaz
á
bal, I.; Fraver, S.; Frelich, L.E.;
Guti
é
rrez,
Á
.G.; Hart, S.J.; et al. Patterns and drivers of recent disturbances across the temperate forest biome.
Nat. Commun. 2018,9, 4355. [CrossRef]
9.
Everham, E.M.; Brokaw, N.V.L. Forest damage and recovery from catastrophic wind. Bot. Rev.
1996
,62,
113–185. [CrossRef]
10.
Meigs, G.W.; Keeton, W.S. Intermediate-severity wind disturbance in mature temperate forests: Legacy
structure, carbon storage, and stand dynamics. Ecol. Appl. 2018,28, 798–815. [CrossRef]
11.
Royo, A.A.; Peterson, C.J.; Stanovick, J.S.; Carson, W.P. Evaluating the ecological impacts of salvage logging:
Can natural and anthropogenic disturbances promote coexistence? Ecology
2016
,97, 1566–1582. [CrossRef]
12.
Bellingham, P.J. Cyclone eects on Australian rain forests: An overview. Austral. Ecol.
2008
,33, 580–584.
[CrossRef]
13.
Tanner, E.V.J.; Rodriguez–Sanchez, F.; Healey, J.R.; Holdaway, R.J.; Bellingham, P.J. Long–term hurricane
damage eects on tropical forest tree growth and mortality. Ecology 2014,95, 2974–2983. [CrossRef]
14.
Lin, T.C.; Hamburg, S.P.; Lin, K.C.; Wang, L.J.; Chang, C.T.; Hsia, Y.J.; Vadeboncoeur, M.A.; McMullen, C.M.M.;
Liu, C.P. Typhoon disturbance and forest dynamics: Lessons from a northwest Pacific subtropical forest.
Ecosystems 2011,14, 127–143. [CrossRef]
15.
Ritchie, E.A.; Elsberry, R.L. Simulations of the extratropical transition of tropical cyclones: Phasing between
the upper-level trough and tropical cyclones. Mon. Weather Rev. 2007,135, 862–876. [CrossRef]
16.
Dvigalo, V.N.; Melekestsev, I.V. The geological and geomorphic impact of catastrophic landslides in the
Geyser Valley of Kamchatka: Aerial photogrammetry. J. Volkanol. Seismol. 2009,3, 314–325. [CrossRef]
Forests 2019,10, 1017 13 of 14
17.
Knutson, T.R.; Sirutis, J.J.; Zhao, M.; Tuleya, R.E.; Bender, M.; Vecchi, G.A.; Villarini, G.; Chavas, D. Global
projections of intense tropical cyclone activity for the late twenty-first century from dynamical downscaling
of CMIP5/RCP4.5 scenarios. J. Clim. 2015,28, 7203–7224. [CrossRef]
18.
Altman, J.; Ukhvatkina, O.N.; Omelko, A.M.; Macek, M.; Plener, T.; Pejcha, V.; Cerny, T.; Petrik, P.; Srutek, M.;
Song, J.-S.; et al. Poleward migration of the destructive eects of tropical cyclones during the 20th century.
Proc. Natl. Acad. Sci. USA 2018,115, 11543–11548. [CrossRef]
19.
Webster, P.J.; Holland, G.J.; Curry, J.A.; Chang, H.R. Changes in tropical cyclone number, duration and
intensity in a warming environment. Science 2005,309, 1844–1846. [CrossRef]
20.
Walsh, K.J.E.; McBride, J.L.; Klotzbach, P.J.; Balachandran, S.; Camargo, S.J.; Holland, G.; Knutson, T.R.;
Kossin, J.P.; Lee, T.C.; Sobel, A.; et al. Tropical cyclones and climate change. Nat. Geosci.
2010
,3, 157–163.
[CrossRef]
21.
Shuo, L.; Deqin, L.; Han, S.; Li, T.; Ming, Z. The physical mechanism and strong precipitation in Northeast
China analysis during Typhoon “Lionrock” merging into extratropical cyclon. Plateau Meteorol.
2019
,38,
804–816.
22.
Nayak, S.; Takemi, T. Dynamical downscaling of Typhoon Lionrock (2016) for assessing the resulting hazards
under global warming. J. Meteorol. Soc. Jap. 2019,97, 69–88. [CrossRef]
23.
Kossin, J.P.; Emanuel, K.A.; Vecchi, G.A. The poleward migration of the location of tropical cyclone maximum
intensity. Nature 2014,509, 349–352. [CrossRef] [PubMed]
24.
Schurman, J.S.; Trotsiuk, V.; Bace, R.; Cada, V.; Fraver, S.; Janda, P.; Kulakowski, D.; Labusova, J.; Mikolas, M.;
Nagel, T.A.; et al. Large-scale disturbance legacies and the climate sensitivity of primary Picea abies forests.
Glob. Change Biol. 2018,24, 2169–2181. [CrossRef]
25.
Seidl, R.; Thom, D.; Kautz, M.; Martin-Benito, D.; Peltoniemi, M.; Vacchiano, G.; Wild, J.; Ascoli, D.; Petr, M.;
Honkaniemi, J.; et al. Forest disturbances under climate change. Nat. Clim. Change
2017
,7, 395–402.
[CrossRef]
26.
Gromyko, M.N. Climate. In Plants, Fungi and Lichens of the Sikhote-Alin Reserve; Dalnauka: Vladivostok,
Russia, 2016; pp. 14–20.
27.
Peel, M.C.; Finlayson, B.L.; McMahon, T.A. Updated world map of the Köppen-Geiger climate classification.
Hydrol. Earth Syst. Sci. 2007,11, 1633–1644. [CrossRef]
28.
Nakamura, Y.; Krestov, P.V.; Omelko, A.M. Bioclimate and zonal vegetation in Northeast Asia: First
approximation to an integrated study. Phytocoenologia 2007,37, 443–470. [CrossRef]
29. Oberle, B.; Ogle, K.; Zanne, A.E.; Wooda, C.W. When a tree falls: Controls on wood decay predict standing
dead tree fall and new risks in changing forests. PLoS ONE 2018,13, e0196712. [CrossRef]
30.
Walker, S.H.; Duncan, D.B. Estimation of the probability of an event as a function of several independent
variables. Biometrika 1967,54, 167–179. [CrossRef]
31.
Mehtätalo, L.; de-Miguel, S.; Gregoire, T.G. Modeling height-diameter curves for prediction. Can. J. For. Res.
2015,45, 826–837. [CrossRef]
32.
R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing:
Vienna, Austria. Available online: http://www.R-project.org/(accessed on 6 October 2019).
33. Besag, J. Contribution to the discussion of Dr. Ripley’s paper. J. R. Stat. Soc. 1977,B39, 193–195.
34.
Wiegand, T.; Moloney, K.A. Handbook of Spatial Point Pattern Analysis in Ecology; Chapman and Hall/CRC:
Boca Raton, FL, USA, 2014; p. 538.
35.
Gromyko, M.N. The first results of studying the catastrophic eect of typhoon Lyonrok on forest ecosystems
of the Sikhote-Alin Nature Reserve. In Proceedings of the XII Far Eastern Conference of Nature Conservation
Problems, Birobidzhan, Russian, 10–13 October 2017; pp. 35–37.
36.
Xi, W.; Peet, R.K.; Urban, D.L. Changes in forest structure, species diversity and spatial pattern following
hurricane disturbance in a piedmont North Carolina forest, USA. J. Plant Ecol. 2008,1, 43–57. [CrossRef]
37.
Mitchell, S.J. Wind as a natural disturbance agent in forests: A synthesis. Forestry
2013
,86, 147–157. [CrossRef]
38.
Boose, E.R.; Serrano, M.I.; Foster, D.R. Landscape and regional impacts of hurricanes in Puerto Rico.
Ecol. Monogr. 2004,74, 335–352. [CrossRef]
39.
Lin, S.Y.; Shaner, P.J.L.; Lin, T.C. Characteristics of old–growth and secondary forests in relation to age and
typhoon disturbance. Ecosystems 2018,21, 1521–1532. [CrossRef]
40.
Peterson, C.J. Within-stand variation in windthrow in southern boreal forests of Minnesota: Is it predictable?
Can. J. For. Res. 2004,34, 365–375. [CrossRef]
Forests 2019,10, 1017 14 of 14
41. Tan, F.; Lim, H.S.; Abdullah, K. The eects of orography in Indochina on wind, cloud, and rainfall patterns
during Typhoon Ketsana (2009). Asia Pac. J. Atmos. Sci. 2012,48, 295–314. [CrossRef]
42.
Sheeld, R.M.; Thompson, M.T. Hurricane Hugo: Eects on South Carolina’s Forest Resource; USDA Forest
Service: Asheville, NC, USA, 1992; p. 51.
43.
Szwagrzyk, J.; Gazda, A.; Dobrowolska, D.; Che´cko, E.; Zaremba, J.; Tomski, A. Tree mortality after wind
disturbance diers among tree species more than among habitat types in a lowland forest in northeastern
Poland. For. Ecol. Manag. 2017,398, 174–184. [CrossRef]
44.
Walker, L.R. Tree damage and recovery from Hurricane Hugo in Luquillo Experimental Forest, Puerto Rico.
Biotropica 1991,23, 379–385. [CrossRef]
45.
Bellingham, P.J.; Kapos, V.; Varty, N.; Healey, J.R.; Tanner, E.V.J.; Kelly, D.L.; Dalling, J.W.; Burns, L.S.; Lee, D.;
Sidrak, G. Hurricanes need not cause high mortality: The eects of Hurricane Gilbert on forests in Jamaica.
J. Trop. Ecol. 1992,8, 217–223. [CrossRef]
46.
Bellingham, P.J. Landforms influence patterns of hurricane damage: Evidence from Jamaican montane
forests. Biotropica 1991,23, 427–433. [CrossRef]
47.
Whigham, D.F.; Dickinson, M.B.; Brokaw, N.V. Background canopy gap and catastrophic wind disturbances
in tropical forests. In Ecosystems of Disturbed Ground; Elsevier Science: Amsterdam, The Netherlands, 1999;
pp. 223–252.
48.
Woods, K.D. Intermediate disturbance in a late-successional hemlock northern hardwood forest. J. Ecol.
2004,92, 464–476. [CrossRef]
49.
Jactel, H.; Bauhus, J.; Boberg, J.; Bonal, D.; Castagneyrol, B.; Gardiner, B.; Gonzalez-Olabarria, J.R.; Koricheva, J.;
Meurisse, N.; Brockerho, E.G. Tree diversity drives forest stand resistance to natural disturbances. Curr. For.
Rep. 2017,3, 223–243. [CrossRef]
50.
Canham, C.D.; Papaik, M.J.; Latty, E.F. Interspecific variation in susceptibility to windthrow as a function of
tree size and storm severity for northern temperate tree species. Can. J. For. Res. 2001,31, 1–10. [CrossRef]
51.
Imbert, D.; Labbe, P.; Rousteau, A. Hurricane damage and forest structure in Guadeloupe, French West
Indies. J. Trop. Ecol. 1996,12, 663–680. [CrossRef]
52.
Greenberg, C.H.; McNab, W.H. Forest disturbance in hurricane-related downbursts in the Appalachian
Mountains of North Carolina. For. Ecol. Manag. 1998,104, 179–191. [CrossRef]
53.
Basnet, K.; Likens, G.E.; Scatera, F.N.; Lugo, A.E. Hurricane Hugo: Damage to a tropical rain forest in Puerto
Rico. J. Trop. Ecol. 1992,8, 47–55. [CrossRef]
54.
Tanner, E.V.J.; Kapos, V.; Healey, J.R. Hurricane eects on forest ecosystems in the Caribbean. Biotropica
1991
,
23, 513–521. [CrossRef]
55.
Nagel, T.A.; Diaci, J. Intermediate wind disturbance in an old-growth beech–fir forest in southeastern
Slovenia. Can. J. For. Res. 2006,36, 629–638. [CrossRef]
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
... This indicates the formation of the secondary forest. Not only that, but it also leads to the loss of a lot of valuable timber in this area 64 . Therefore, it shows the economic value to the region (not rotten wood), and buck and forest investigate the environmental factors by using the data of the ODW (binary) 19,20 and the NDW (count) 21 to establish the GL, GWR, and GNNWR models, respectively. ...
... However, we can see from the range of the box that these are not the only relations (Fig. 7), which will vary according to different regions, which also shows the existence of non-stationarity in the spatial model 35 . In the E and W orientations, the ODW is high and the NDW is large 64 . In the N, NE and NW orientations, the ODW is high, but NDW not large. ...
Article
Full-text available
The natural forest ecosystem has been affected by wind storms for years, which have caused several down wood (DW) and dramatically modified the fabric and size. Therefore, it is very important to explain the forest system by quantifying the spatial relationship between DW and environmental parameters. However, the spatial non-stationary characteristics caused by the terrain and stand environmental changes with distinct gradients may lead to an incomplete description of DW, the local neural-network-weighted models of geographically neural-network-weighted (GNNWR) models are introduced here. To verify the validity of models, our DW and environmental factors were applied to investigate of occurrence of DW and number of DW to establish the generalized linear (logistic and Poisson) models, geographically weighted regression (GWLR and GWPR) models and GNNWR (GNNWLR and GNNWPR) models. The results show that the GNNWR models show great advantages in the model-fitting performance, prediction performance, and the spatial Moran’s I of model residuals. In addition, GNNWR models can combine the geographic information system technology for accurately expressing the spatial distribution of DW relevant information to provide the key technology that can be used as the basis for human decision-making and management planning.
... Nearly all forest ecosystems throughout the world are shaped and influenced by natural disturbances, including cyclonic windstorms (hurricanes, cyclones, and tornadoes). Catastrophic storms cause immediate and long-term structural damage to individual trees, including massive defoliation and branch loss, a decline in stem density, basal area, diameter at breast height (DBH), and total height [1,2]. More specifically, the structure of the forest canopy can be molded by frequent wind disturbance, influencing the biophysical environment, tree physiology, atmospheric exchange, and biotic habitat [3,4]. ...
... Such time period selection considered the necessary degree of comprehension that is appropriate for the literature search questions that we would like to investigate through a preliminary online literature search. In addition to the identification of knowledge gaps about the reviewed topic, we also structured the present review into the following questions: (1) what is the proportion of studies for tropical versus subtropical and temperate regions?; (2) which effects on forests are mostly examined at each biological organization and forest functional type?; and (3) which methodological approaches were commonly used in determining the effects of windstorms on the forest?. The present work will provide us with an integrative understanding of the impacts of catastrophic windstorms on forests, particularly on structure, species composition, carbon storage and emission, and dynamics of tropical, subtropical, and temperate forests across levels of biological organization (i.e., molecular to landscape levels). ...
Article
Full-text available
Windstorm is one of the destructive natural disturbances, but the scale-link extent to which recurrent windstorms influenced forests ecosystems is poorly understood in a changing climate across regions. We reviewed the synergistic impacts of windstorms on forests and assessed research trends and methodological approaches from peer-reviewed articles published from 2000 to 2020 in tropical (TRF), subtropical (SUF), and temperate (TEF) forests/zones, based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Overall, the majority of the reviewed studies were conducted in TRF (i.e., 40%), intermediate in SUF (i.e., 34%), and the lowest in TEF (i.e., 26%). Among the four levels of biological organization, the species-population and community-ecosystem levels had the highest number of study cases, while the molecular-cellular-individual and landscape levels had the lowest study cases in all forest types. Most of the articles reviewed dealt largely on tree mortality/survival and regeneration/succession for TRF, tree mortality/survival and species composition/richness/diversity for SUF, and stem density, gap dynamics, and regeneration/succession for TEF. However, research on the effects of windstorms on mycorrhizal symbioses, population genetics, and physiological adaptation, element fluxes via litterfall, litter decomposition, belowground processes, biological invasion, and tree health are less common in all forest types. Further, most of the studies were conducted in permanent plots but these studies mostly used observational design, while controlled studies are obviously limited. Consequently, more observational and controlled studies are needed on the topic reviewed, particularly studies at the molecular-cellular-individual and landscape levels, to help inform forest management decision-making about developing sustainable and resilient forests amid climate change.
... Substrate stability is obviously important for structurally dependent epiphytes and structural failure accounts for a large proportion of deaths (Hietz 1997, Hietz et al. 2002, Sarmento Cabral et al. 2015, Schmit-Neuerburg 2002, Zotz 1998, Zotz et al. 2005. Under extreme conditions, that is, hurricanes (category 1 with wind speeds from 33 m s −1 ), forest canopies suffer severe damage (e.g., Hirsh & Marler 2002, Lugo et al. 1983, Vozmishcheva et al. 2019. The mechanical effect of wind on epiphytes themselves is hardly studied (but see Tay et al. 2021), although the possibility of epiphytes being dislodged by wind is repeatedly mentioned in the literature, albeit without quantitative data (e.g., Francisco-Ventura et al. 2018, Lowman & Linneroth 1995, Pett-Ridge & Silver 2002, Rodríguez-Robles et al. 1990). ...
Article
Full-text available
Several studies of hurricane damage on epiphyte communities implied that epiphytes might be in danger of being blown off their host when subjected to strong wind. There is very limited knowledge about the mechanical impact that wind may have on epiphytes. Using a wind-triggered camera set-up, we observed how epiphytic tank bromeliads are affected by wind. Despite offering a relatively large area of ‘attack’ to the airflow, bromeliads moved relatively little themselves. Rather than being directly moved by wind, the bromeliads in the upper crown of tall trees moved with the sway of the branches. Only when the substrate did not move, bromeliads with long broad leaves showed considerable disturbance due to wind. Our observations underline the complexity of the system and emphasise that our current understanding of the mechanical aspects of the epiphyte–host system is still very limited.
... regressions and more unpredictably destructive under global warming (Sun et al., 2017;Zhang et al., 2020). Recent unprecedently strong TC forest disturbances were observed in the north of our study area (Vozmishcheva et al., 2019). These changes of climatic systems, specifically the poleward migration and increasing intensity of TCs, might cause the shift of GTs towards the north along the latitudinal gradient (Altman et al., 2018). ...
Article
Full-text available
Aim Understanding how natural forest disturbances control tree regeneration is key to predicting the consequences of globally accelerating forest diebacks on carbon stocks and forest biodiversity. Tropical cyclones (TCs) are important drivers of forest dynamics in Eastern Asia, and it is predicted that their importance will increase. However, little is known about the impact of TCs on forest regeneration. Location Latitudinal gradient from south Korea (33° N) to the Russian Far East (45° N). Time period Last 300 years. Major taxa studied Quercus mongolica, Abies nephrolepis and Pinus koraiensis. Methods We explored the effects of TC activity on canopy accession strategies derived from long‐term tree radial growth patterns along a 1,500‐km latitudinal gradient of decreasing TC activity. We analysed canopy accession strategies for > 800 trees of three widely distributed tree species by dividing them into gap trees (GTs), which established immediately after gap formation, and released trees (RTs), which accessed the upper canopy after a period of competitive suppression. Results We found a substantial decrease in GTs and increase in RTs proportionally along the gradient of decreasing TC activity. Pinus koraiensis and A. nephrolepis exhibited high variability in the proportions of the individual canopy accession strategies along the latitudinal gradient, whereas it was more stable for Q. mongolica. We identified the gradient of TC activity as the main driver influencing canopy dynamics and thus changes in life‐history traits for P. koraiensis and Q. mongolica, whereas maximal growth rate was the main driver for A. nephrolepis. Main conclusions Flexibility in growth strategies enabled the studied species to cover extensive areas and indicates that they will be able to cope with shifts in disturbance regimes induced by the poleward migration of TCs and increasing TC intensity. Our results highlight the canopy accession strategy as an ecological indicator of past disturbance activity.
Article
Full-text available
Проведена оценка воздействия ветровала на структуру и фитомассу древостоев кедровников и березняков по материалам обследований постоянных пробных площадей (ППП) на территории Сихотэ-Алинского заповедника, на восточном макросклоне Сихотэ-Алиня, где вследствие тайфуна Lionrock произошло частичное выпадение древостоя. Ветровал 2016 г. – беспрецедентное в истории заповедника катастрофическое явление. Более 9% лесного покрова особо охраняемой природной территории перешло в сплошные ветровалы. Исходными материалами для анализа послужили ревизии пробных площадей, выполненные до и после ветровала. Межревизионный период у разных объектов составляет 8–18 лет. Запас надземной фитомассы определен по таксационным данным с использованием региональных аллометрических уравнений. Средний запас кедровников вследствие ветровала уменьшился с 552 до 298 м3/га, а березняков с 253 до 163 м3/га. Общий прирост для кедровников и березняков составил 6.2 и 5.5 м3/га год соответственно, а отпад 13.6 и 9.8 м3/га год соответственно. Средние значения запаса фитомассы кедровых насаждений – 291 т/ га, березовых насаждений – 210 т/га. Запасы фитомассы березняков уменьшились в результате ветровала на 35%, а кедровников – на 44%. Запас углерода 150 т С га можно считать максимальной углеродной емкостью фитомассы для кедровников среднего и верхнего высотного пояса, относительно которой следует вычислять депонирующий потенциал нарушенных лесов этой формации. Из хвойных пород менее устойчивы к ветровалу ель, пихта и кедр. Для осины и березы плосколистной действие тайфуна усиливает их естественное выпадение из состава древостоя. Наиболее устойчивыми породами оказались клены, липа амурская и лиственница.
Chapter
The Eastern Highlands of Zimbabwe is home to a unique floristic diversity, which include exotic plantations of pines and eucalyptus, natural moist forests and dry forests. We assessed the impact of tropical cyclone Idai on natural and plantation forests. Normalized difference vegetation index (NDVI) was used to determine the changes in forest conditions. We also conducted a questionnaire survey in the study area to elicit perceptions of key informants. Results show a decrease in the NDVI values during the month of the cyclone. Results from surveys indicated that the impact of the cyclone varied across locations, with the relatively low altitude (≤1400 m.a.s.l.) areas being the most affected compared to higher altitude (>1400) sites. The impact of the cyclone was more on pines compared to eucalyptus species. Pinus tecunumanii had the largest proportion (expressed as % of total volume damaged) of 37% followed by P. kesiya (16%) and lastly P. maximinoi (12.3%). Our results indicate that plantation species responded differently to the impact of the cyclone and that low elevation sites suffered more impact than high elevation areas. We conclude that silvicultural management could provide cues for mitigating the impact of cyclones in plantation species.
Article
Full-text available
Since the Eighth International Workshop on Tropical Cyclones (IWTC-8), held in December 2014, progress has been made in our understanding of the relationship between tropical cyclone (TC) characteristics, climate and climate change. New analysis of observations has revealed trends in the latitude of maximum TC intensity and in TC translation speed. Climate models are demonstrating an increasing ability to simulate the observed TC climatology and its regional variations. The limited representation of air-sea interaction processes in most climate simulations of TCs remains an issue. Consensus projections of future TC behavior continue to indicate decreases in TC numbers, increases in their maximum intensities and increases in TC-related rainfall. Future sea level rise will exacerbate the impact of storm surge on coastal regions, assuming all other factors equal. Studies have also begun to estimate the effect on TCs of the climate change that has occurred to date. Recommendations are made regarding future research directions. Keywords: tropical cyclone, climate change, climate variability
Article
Full-text available
Typhoons are considered as one of the most powerful disaster-spawning weather phenomena. Recent studies have revealed that typhoons will be stronger and more powerful in a future warmer climate and be a threat to lives and properties. In this study, we conduct downscaling experiments of an extreme rain-producing typhoon, Typhoon Lionrock (2016) in order to assess the impacts of climate change on resulting hazards by assuming pseudo global warming (PGW) conditions. The downscaled precipitations over the landfall region in the present climate condition agree well with the Radar- Automated Meteorological Data Acquisition System (Radar-AMeDAS) observations. A typhoon track in the future climate similar to that in the present climate is successfully reproduced, with a stronger wind speed (by ~20 knots) and lower central pressure (by ~20 hPa) under the PGW condition. The changes in precipitation amounts associated with the typhoon under PGW condition are analyzed over 7 individual prefectures in the northern part of Japan. The typhoon in the warming climate produces more precipitation over all prefectures. Iwate, Aomori, Akita, Miyagi and Hokkaido are projected to have relatively more precipitation associated with the typhoon in the warming climate. The overall analysis suggests that Typhoon Lionrock under PGW may increase the risk of flooding, damages to infrastructures, and lives staying along the typhoon track.
Article
Full-text available
Increasing evidence indicates that forest disturbances are changing in response to global change, yet local variability in disturbance remains high. We quantified this considerable variability and analyzed whether recent disturbance episodes around the globe were consistently driven by climate, and if human influence modulates patterns of forest disturbance. We combined remote sensing data on recent (2001–2014) disturbances with in-depth local information for 50 protected landscapes and their surroundings across the temperate biome. Disturbance patterns are highly variable, and shaped by variation in disturbance agents and traits of prevailing tree species. However, high disturbance activity is consistently linked to warmer and drier than average conditions across the globe. Disturbances in protected areas are smaller and more complex in shape compared to their surroundings affected by human land use. This signal disappears in areas with high recent natural disturbance activity, underlining the potential of climate-mediated disturbance to transform forest landscapes.
Article
Full-text available
When standing dead trees (snags) fall, they have major impacts on forest ecosystems. Snag fall can redistribute wildlife habitat and impact public safety, while governing important carbon (C) cycle consequences of tree mortality because ground contact accelerates C emissions during deadwood decay. Managing the consequences of altered snag dynamics in changing forests requires predicting when snags fall as wood decay erodes mechanical resistance to breaking forces. Previous studies have pointed to common predictors, such as stem size, degree of decay and species identity, but few have assessed the relative strength of underlying mechanisms driving snag fall across biomes. Here, we analyze nearly 100,000 repeated snag observations from boreal to subtropical forests across the eastern United States to show that wood decay controls snag fall in ways that could generate previously unrecognized forest-climate feedback. Warmer locations where wood decays quickly had much faster rates of snag fall. The effect of temperature on snag fall was so strong that in a simple forest C model, anticipated warming by mid-century reduced snag C by 22%. Furthermore, species-level differences in wood decay resistance (durability) accurately predicted the timing of snag fall. Differences in half-life for standing dead trees were similar to expected differences in the service lifetimes of wooden structures built from their timber. Strong effects of temperature and wood durability imply future forests where dying trees fall and decay faster than at present, reducing terrestrial C storage and snag-dependent wildlife habitat. These results can improve the representation of forest C cycling and assist forest managers by helping predict when a dead tree may fall.
Article
Full-text available
Many properties of forest ecosystems, such as species composition and forest structure, naturally vary with forest age. However, in regions prone to cyclone disturbances, both forest age and cyclone severities can play a role shaping these properties. To evaluate potential effects of an altered cyclone regime on forest ecosystems, it is necessary to disentangle the roles of cyclones and forest age on different forest characteristics. In this study, we compared species composition and forest structure at plot level across sites with similar climate and topographic backgrounds, yet differing in age and typhoon severities in northeastern Taiwan. We found shorter tree stature, higher wood density, higher tree species diversity, and lower typhoon-induced tree mortality in the sites with more severe typhoon disturbances. On the other hand, regardless of typhoon severity, the sites of younger ages had a considerably smaller amount of woody debris, suggesting that it takes time for the accumulation of woody debris. More typhoon-induced canopy gaps at sites with more severe typhoon influences highlights a role of typhoons in canopy dynamics. However, the lack of gaps prior to typhoon disturbances in the less severely affected forest is likely related to the low background mortality associated with the relative young age of the forest. Our results indicate that frequent or severe typhoon disturbances can homogenize some of the structural differences among forests of different ages. If the frequency or severity of cyclones increase in the future, many forests, including old-growth forests, may gradually lose large trees.
Article
Full-text available
Determining the drivers of shifting forest disturbance rates remains a pressing global change issue. Large-scale forest dynamics are commonly assumed to be climate driven, but appropriately scaled disturbance histories are rarely available to assess how disturbance legacies alter subsequent disturbance rates and the climate sensitivity of disturbance. We compiled multiple tree-ring based disturbance histories from primary Picea abies forest fragments distributed throughout five European landscapes spanning the Bohemian Forest and the Carpathian Mountains. The regional chronology includes 11 595 tree cores, with ring dates spanning the years 1750 to 2000, collected from 560 inventory plots in 37 stands distributed across a 1000 km geographic gradient, amounting to the largest disturbance chronology yet constructed in Europe. Decadal disturbance rates varied significantly through time and declined after 1920, resulting in widespread increases in canopy tree age. Approximately 75% of current canopy area recruited prior to 1900. Long-term disturbance patterns were compared to an historical drought reconstruction, and further linked to spatial variation in stand structure and contemporary disturbance patterns derived from LANDSAT imagery. Historically, decadal Palmer drought severity index minima corresponded with higher rates of canopy removal. The severity of contemporary disturbances increased with each stand's estimated time since last major disturbance, increased with mean diameter and declined with increasing within-stand structural variability. Reconstructed spatial patterns suggest that high small-scale structural variability has historically acted to reduce large-scale susceptibility and climate sensitivity of disturbance. Reduced disturbance rates since 1920, a potential legacy of high 19th century disturbance rates, have contributed to a recent region-wide increase in disturbance susceptibility. Increasingly common high-severity disturbances throughout primary Picea forests of Central Europe should be reinterpreted in light of both legacy effects (resulting in increased susceptibility) and climate change (resulting in increased exposure to extreme events). This article is protected by copyright. All rights reserved.
Article
Determination of long-term tropical cyclone (TC) variability is of enormous importance to society; however, changes in TC activity are poorly understood owing to discrepancies among various datasets and limited span of instrumental records. While the increasing intensity and frequency of TCs have been previously documented on a long-term scale using various proxy records, determination of their poleward migration has been based mostly on short-term instrumental data. Here we present a unique tree-ring–based approach for determination of long-term variability in TC activity via forest disturbance rates in northeast Asia (33–45°N). Our results indicate significant long-term changes in TC activity, with increased rates of disturbances in the northern latitudes over the past century. The disturbance frequency was stable over time in the southern latitudes, however. Our findings of increasing disturbance frequency in the areas formerly situated at the edge of TC activity provide evidence supporting the broad relevance of poleward migration of TCs. Our results significantly enhance our understanding of the effects of climate change on TCs and emphasize the need for determination of long-term variation of past TC activity to improve future TC projections.
Article
Wind is one of the most important natural disturbances influencing forest structure, ecosystem function, and successional processes worldwide. This study quantifies the stand-scale effects of intermediate-severity windstorms (i.e., “blowdowns”) on (1) live and dead legacy structure, (2) aboveground carbon storage, and (3) tree regeneration and associated stand dynamics at four mature, mixed hardwood-conifer forest sites in the northeastern United States. We compare wind-affected forests to adjacent reference conditions (i.e., undisturbed portions of the same stands) 0-8 years post-blowdown using parametric (ANOVA) and nonparametric (NMS ordination) analyses. We supplement inventory plots and downed coarse woody detritus (DCWD) transects with hemispherical photography to capture spatial variation in the light environment. Although recent blowdowns transferred a substantial proportion of live overstory trees to DCWD, residual live tree basal area was high (19-59% of reference areas). On average, the initial post-blowdown ratio of DCWD carbon to standing live tree carbon was 2.72 in blowdown stands and 0.18 in reference stands, indicating a large carbon transfer from live to dead pools. Despite these dramatic changes, structural complexity remained high in blowdown areas, as indicated by the size and species distributions of overstory trees, abundance of sound and rotten downed wood, spatial patterns of light availability, and variability of understory vegetation. Furthermore, tree species composition was similar between blowdown and reference areas at each site, with generally shade-tolerant species dominating across multiple canopy strata. Community response to intermediate-severity blowdown at these sites suggests a dynamic in which disturbance maintains late-successional species composition rather than providing a regeneration opportunity for shade-intolerant, pioneer species. Our findings suggest that intermediate-severity wind disturbances can contribute to stand-scale structural complexity as well as development towards late-successional species composition, at least when shade-tolerant regeneration is present pre-blowdown. Advance regeneration thus enhances structural and compositional resilience to this type of disturbance. This study provides a baseline for multi-cohort silvicultural systems designed to restore heterogeneity associated with natural disturbance dynamics. This article is protected by copyright. All rights reserved.