Content uploaded by Zhu Jiao-jun
Author content
All content in this area was uploaded by Zhu Jiao-jun on Jan 14, 2018
Content may be subject to copyright.
Effect of gap size created by thinning on seedling emergency,
survival and establishment in a coastal pine forest
Jiao-jun Zhu
a,b,*
, Takeshi Matsuzaki
b
, Feng-qin Lee
a
, Yutaka Gonda
b
a
Institute of Applied Ecology, Chinese Academy of Sciences, Wenhua Road 72, Shenyang 110016, PR China
b
Faculty of Agriculture, Niigata University, Niigata 950-2181, Japan
Received 16 April 2002; received in revised form 26 November 2002; accepted 3 February 2003
Abstract
It is desirable and necessary to preserve the continuity of a coastal forest through reasonable management because it can
provide many shelter benefits through altering the wind behavior along the shore. Thinning is an undoubtedly important measure
for the continuity of forests as it provides suitable conditions for natural regeneration; however, thinning increases the risk of
wind damage immediately after thinning in the coastal areas. Therefore, few thinning study related to regeneration in a coastal
forest has been made. In order to test whether coastal forest of Japanese black pine (Pinus thunbergii Parl.) requires a specific gap
size created by thinning for regeneration and to compare seedling establishment among four thinning treatments, observations of
emergence, survival and establishment of P. thunbergii seedlings, together with soil water content, litter, wind and light regime
were made. The observations were conducted over four growing seasons in three sizes of circular gaps (the gap diameter to stand
height ratios for the four gap sizes were 0.5, 1.0, 1.5 and 0.0, the control) corresponding to the four thinning treatments in Niigata
shore, Japan. Results indicated that density of seedlings older than 1 year increased with gap size or canopy openness (OP).
Seedling establishment was greater in 1.5 gap sizes than in any other gap sizes, while seedlings peaked near the west and north
edges of the gaps but not in the gap centers exposed to direct solar radiation. Seedling growth in 1.5 gap sizes was also
significantly higher than that in any others. A tendency of seedling height increasing from east to west edge and from south to
north edge across the gap was observed. Only 1- or 2-year-old seedlings occurred in gap sizes of 0.0 and 0.5, therefore
establishment in both gap sizes was considered as failing. The results imply that although P. thunbergii seeds can germinate in
small gaps, even in under canopy, the seedlings are unable to survive. The seedlings apparently require a minimum gap size
1.0, or OP >30% in order to survive, and may require at least gap size 1.5, or OP >40% for further development into sapling.
These results can be explained by the changes of microclimates, i.e. increase of light, soil water and airflow exchange, decrease
of litter and canopy cover, and alleviation of the competitions for water in gaps created by thinning. Therefore, thinning strategy,
especially patch-pattern thinning is potentially a viable silvicultural measure in management for the coastal pine forest. These
results provide references for establishment and management of coastal P. thunbergii forests.
#2003 Elsevier Science B.V. All rights reserved.
Keywords: Gap size; Pinus thunbergii Parl.; Natural regeneration; Coastal forest; Patch-pattern thinning
1. Introduction
Japanese black pine (Pinus thunbergii Parl.) is one
of the most important tree species in coastal forests of
Japan islands (Murai et al., 1992). These coastal
Forest Ecology and Management 182 (2003) 339–354
*
Corresponding author. Tel.: þ86-24-23843192;
fax: þ86-24-23843313/23843192.
E-mail addresses: zrms29@yahoo.com, jiaojunzhu@iae.ac.cn
(J.-j. Zhu).
0378-1127/$ – see front matter #2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0378-1127(03)00094-X
forests, through altering the wind behavior along the
shore, simultaneously provide shelter benefits such as
the protection for the local environment, recreation
and nature conservation, etc. Therefore, it is necessary
to preserve the continuity of their shelter functions
through reasonable management. Thinning, as one of
the most important management techniques for plan-
tations, is undoubtedly indispensable for the sustain-
able development of forest (Hong et al., 1994; Seiwa
and Kikuzawa, 1994; Jiang et al., 1994, 1997; Ooishi
et al., 1998; Krauchi et al., 2000). Thinning creates
canopy openness (OP) or gaps, which are critical in the
community dynamic of many types of forest (Gray and
Spies, 1996). Particularly, the processes of gap creat-
ing and filling are thought to play a central role in
species coexistence and regeneration in a variety of
forests (Brokaw, 1987; Lertzman, 1992; Brett and
Kilnka, 1998; Dalling et al., 1998). Research indicates
that even the shade-tolerant tree species require
canopy gaps to reach the canopy in old-growth con-
iferous forest (Spies and Franklin, 1989; Gray and
Spies, 1996; Gobbi and Schlichter, 1998). For exam-
ple, Tsuga (Tsuga sp.), the most shade-tolerant among
temperate forest species, can regenerate without
canopy gaps, but may eventually require one or more
small openness to survive and grow into the overstory
(Spies and Franklin, 1989). Variations in physical
environment within gaps provide opportunities for
species that could not establish under a closed canopy.
Differentiation of the responses of the species to gap
sizes has significant implications for general model of
forest dynamics (Canham, 1989). The importance of
small gap-forming disturbances has emerged as a com-
mon theme in research on forests dynamic and natural
regeneration from a variety of forests worldwide (Lertz-
man, 1992; Gray and Spies, 1996; Mclaren and Janke,
1996; Jennings et al., 1999). Gaps created by the normal
thinning process are generally small-scale and ephem-
eral (Spies and Franklin, 1989), while in forests where
stand-destroying disturbances are rare, the small-scale
disturbances that create gaps, play key roles on the
development of forest structure, and on forest floor
conditions such as light regime, soil water content
and litter decomposition (Brokaw, 1987; Lertzman
et al., 1996; Myers et al., 2000). The importance of
gap forming and filling has been recognized for a period
and most research on gap dynamics focuses on the old-
growth phase of development (Gray and Spies, 1996;
Meer et al., 1999), but the role of canopy gaps in
development of special forests is little understood. In
particular, the study on tree seedling establishment in
relation to canopy gaps within a coastal forest is poorly
understood because of its peculiarities, i.e. the planta-
tion system close to the sea, which provides many
shelter benefits, is vulnerable to disturbances. However,
thinning (one kind of disturbance) is the most important
silvicultural measure to promote natural regeneration
and maintain the continuity of shelter benefits. There-
fore, it is necessary to understand the relationships
between thinning and regeneration in a coastal forest.
The purpose of this paper is to report the results of
four growing-season observations in natural regenera-
tion of the coastal P. thunbergii forest in relation to
thinning ratios. The relationships between natural
regeneration and microclimate conditions, whether
P. thunbergii requiring a specific canopy gap size
for regeneration, and comparison of seedling estab-
lishment in various sizes of canopy gaps created by the
thinning are examined. Some implications of thinning
to establishment and management of the coastal P.
thunbergii forest are also discussed.
2. Material and methods
2.1. Site description
The experiment was carried out at the middle of the
shoreline along the Japan Sea. It is located at Aoyama
coastal area, Niigata prefecture (37852041.300N,
138856016.800E). The annual precipitation is 1778 mm,
annual average temperature is 13.2 8C, minimum tem-
perature is 13.0 8C; the first frost is on 24 December,
and the last frost is on 30 March (National Astronomical
Observatory, 2000). The coastal forest is composed of
pure P. thunbergii on sandy soil, which stretches along
the shoreline in a width ranging from 100 to 200 m. It
was planted about 40 years ago at an initial stocking
density of approximately 4500 stems ha
1
, the current
preserved rate is averaged at about 73%. About 50-m
wide zone of young pine trees exists behind the sand
dune close to the sea. There is a 15-m road in about 308
in E–W (about 608from the true north) between the
young pine trees and the coastal forest (Fig. 1). The
micro-site factor and micro-topography in the experi-
mental area are almost the same in a wide range.
340 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
Eco-physiological studies point out that P. thunbergii
is relatively light demanding, and the micro-site
factors of coastal sandy dune accelerate the senes-
cence of P. thunbergii trees there (Ogasawara, 1986,
1988). For a dense coastal pine forest on the sandy
dune, the natural regeneration appears to be limited by
light conditions and competition. Therefore, gaps are
required for its natural regeneration (Murai et al.,
1992; Gobbi and Schlichter, 1998). While the coastal
plantation close to the sea is vulnerable to wind risk
(Perry, 1994; Zhu et al., 2001a), any large-scale
disturbances, which affect the shelter function of
the coastal forest, are undesirable.
2.2. Thinning treatments
In order to examine the responses of wind vulner-
ability of coastal forest stand to thinning intensity and
the effects of thinning on forest regeneration, the
coastal forest stand was thinned by patch-pattern in
December 1997 with random sampling techniques.
The thinning (patch-pattern) treatments were set as 20,
30, 50% (thinned) and 0.0% (unthinned), which are
referred to treatments 1, 2, 3 and 4, respectively
(Fig. 1). The effective area of each treatment reached
40 m 50 m. The mean characteristics of stand before
and after thinning are shown in Table 1.
2.3. Gap characteristics
A canopy gap is defined as a hole extending through
all levels of the canopy to within 1 (Myers et al., 2000)
or 2 m (Lawton and Putz, 1988; Gray and Spies, 1996;
Yamamoto, 2000) of the forest ground. The gap
definition as Myers et al. (2000) suggest, i.e. 1 m
above the ground was used in this study. Three sizes
of circular gaps created by thinning, with two repli-
cates each, were selected corresponding to the thin-
ning treatments in the coastal forest. The gap size was
determined by following steps.
Before measuring the gap diameter, the gap center
was determined using a digital camera equipped with a
monitor (Nikon, Coolpix 910, Japan, f¼7–21 mm).
The gap center was fixed as the midpoint of the gap
diameter, then, the gap length was measured from the
edge of the upper crown on one side, to the edge of the
upper crown on the opposite side. Four diagonal con-
jugate lengths were measured for calculating the gap
diameter. Gap diameter, which was calculated from the
measurements, was scaled to the average height of the
stand as suggested by Gray and Spies (1996). Four gap
sizes, i.e. the ratios of gap diameter to stand height,were
0.0 (a control plot, referred to as 0.0 gap size, equal in
area to the 1.0 gap size), 0.5 (small gap, 0.5 gap size),
1.0 (middle gap, 1.0 gap size) and 1.5 (large gap, 1.5 gap
Fig. 1. Layout of the experimental location (Aoyama coastal area, Niigata prefecture, Japan, 37852041.300 N, 138856016.800E).
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 341
size). The large gap was selected from treatment 3 (50%
thinned), the middle gap was from treatment 2
(30% thinned), the small one was from treatment 1
(20% thinned) and the control, i.e. treatment 4
(unthinned area). To avoid disturbing the regeneration
in gaps, microclimate conditions were investigated in
other positions outside the experimental gaps. The
positions are similar to the gaps in each thinned treat-
ment because the various sizes of gaps were corre-
sponding to the thinned treatments. Namely, the
microclimate conditions in treatments 1, 2, 3 and 4
were used to represent the ones in corresponding to gap
sizes of 0.5, 1.0, 1.5 and 0.0, respectively.
2.4. Measurements of micro-site conditions
2.4.1. Light environment
In order to compare the light regime in each treatment
and each gap, the canopy openness was estimated from
the silhouettes of hemispherical photographs using a
digital hemispherical camera (Nikon, Coolpix 910,
Japan, f¼7–21 mm) with 1808fish-eye converter
(Nikon, FC-E8, f¼8–24 mm). The hemispherical
photographs were taken by placing the camera on a
tripod at a height of 1.0 m at seven points for each
treatment (Fig. 2) and at the center of gaps. The direct
light (S
dir
) and diffuse light (S
dif
) were estimated
according to the process developed by Steege (1994)
using CanopOn Program (Ver1.09, Takenaka’s Web)
from the hemispherical images.
2.4.2. Litter investigation
Four survey areas of 1 m 1 m were set up in each
treatment to determine the thickness and quantity of
litter (Fig. 2). All of the fallen leaves and branches
were collected in each area at the end of this observa-
tion. Litter thickness was measured in five points of
each area. After collecting the litter from the field, a
portion of litter was sampled to measure the oven-
dried weight in the laboratory.
2.4.3. Soil water content
Soil water content of layers 0–10, 20–30 and more
than 50 cm, were measured in each treatment to reveal
the changes of water content. After scarifying the
litters of dead leaves and branches, soil samples were
Table 1
Mean characteristics of stand in a coastal pine forest with four thinning intensities
Treatment
no.
DBH
a
(cm)
Clear bole
height (m) (h
0
)
Tree height
(H) (m)
Density
(stem ha
1
)
Basal area
(m
2
ha
1
)
h
0
/HThinning rate
by stem (%)
Thinning rate by
basal area (%)
Before thinning (December 1997)
1 9.2 3.9 7.5 3217 23.36 0.52 20.2 19.8
2 9.0 3.1 5.9 3167 21.42 0.53 31.6 32.5
3 10.1 4.2 7.3 3000 26.00 0.58 46.7 50.2
4 8.7 3.3 6.2 3600 23.15 0.53 0.0 0.0
After thinning (February 1998)
1 9.4 3.9 7.5 2517 18.75 0.52
2 9.1 3.2 5.9 2100 14.46 0.54
3 10.1 4.3 7.2 1483 12.94 0.60
4 8.7 3.3 6.2 3600 23.15 0.53
After thinning (January 2000)
1 9.8 4.0 8.5 2517 21.27 0.47
2 9.8 3.2 7.0 2100 16.91 0.46
3 10.8 4.1 8.2 1483 15.48 0.50
4 9.3 3.7 7.2 3600 26.13 0.51
After thinning (November 2001)
1 10.3 4.1 9.7 2517 22.66 0.45
2 10.4 3.2 8.3 2100 19.53 0.43
3 11.4 4.0 9.5 1483 17.82 0.42
4 9.8 3.9 8.8 3600 28.84 0.47
a
DBH: diameter at breast height, 1.3 m.
342 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
collected. Five cores (can volume ¼100 ml) at each
layer were taken during 2 h for all of the treatments.
The water content was calculated by the sample
weight before and after oven drying (105 8C). The
observations were conducted twice every month dur-
ing March and August, and once every month during
September and November of 2000 and 2001.
2.4.4. Wind measurement
One propeller anemometer (Tokyo Ota No. 111-T,
Kona Ltd., Japan) with a data logger (Kona DS-64K,
Kona Sapporo, Japan) was mounted at a height of 2 m
above the ground close to the sea for obtaining data
outside the coastal forest. In order to collect vertical
wind data in the coastal forest, two 10-m towers were
settled in the center of treatments 4 (unthinned) and 3
(50% thinned); 2-m steel pipes were set up in the
center of the other two treatments (see Zhu et al.,
2001b, 2003). Wind data inside the coastal forest were
collected during April 1999 and December 2001 using
one set of five-channel hot wire anemometer (Rion Tr-
Am-11, Rion Ldt., Japan). The sensors of the hot wire
anemometer were mounted on 1.5 m long arms on the
towers and 0.4 m long arms on the poles, respectively.
Intervals for wind collection were 2 and 10 min for
propeller anemometers, and 0.5 min for hot wire
anemometer. In order to improve the accuracy, wind
data obtained inside the coastal forest were selected to
meet the following criteria: (1) wind speed of interval
average outside the coastal forest was more than
3.0 m s
1
, (2) wind direction of interval average out-
side the coastal forest was approximately perpendi-
cular to longitudinal side of the coastal forest. If either
of these criteria were not met within the interval, the
data inside the coastal forest for the period were
discarded.
Wind profiles provide us the general wind distribu-
tion in each treatment; it can be modeled by expo-
nential function in Eq. (1) (Landsberg and James,
1971; Cionco, 1985; Amiro, 1990; Groß, 1993;
Peltola, 1996).
UinðzÞ¼UHexp½að1z=HÞ (1)
where zis the interest height in the coastal forest (m),
Htop height of canopy (m), U
H
wind speed (m s
1
)at
height H,U
in(z)
wind speed (m s
1
) at height z, and ais
called as attenuation coefficient (Amiro, 1990). The
coefficient a, which is related to drag coefficient and
shearing stress of momentum flux, represents the wind
reduction (Landsberg and James, 1971; Cionco,
1985).
2.5. Census of seedling regeneration
Five plots of 2 m 2 m were set up in each treat-
ment for monitoring the survival of seedlings (Fig. 2).
Fig. 2. Scheme for making canopy openness measurement (*), litter collection (&) and seedling census (through plot 1–5, area of plot 1–
5¼2m
2
) in the four treatments.
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 343
In each plot, the emergence, survival and establish-
ment of seedlings were followed twice every year after
the thinning, i.e. in spring and autumn.
Census of seedlings in gaps was conducted at the
end of the fourth growing season. Strips of 1-m width
through the gap center in east–west and south–north
transects were designed for the census. Gap positions
were delineated by dividing at drop line in the east–
west and south–north transects into sub-plots in inter-
val of 1 m from the gap center (*)inactinoid
directions (*!W, *!E, *!Sand*!N;
*, the center of the gap) (Fig. 3). All seedlings and
the corresponding age distinguished by seedling ver-
ticils (whorls) presented in each sub-plot were
recorded. The base stem diameter and height of
seedlings older than 1 year were measured. The height
of each seedling was determined by measuring the
distance from the forest floor (soil surface) to the
shoot tip, or the top part of the seedling. The emer-
gence was defined as the stage when the first above-
ground primary needles become visible, and estab-
lishment was defined as seedling that survived
throughout the study period as earlier suggested by
Xiong and Nilsson (1999).
Fig. 3. Sketch for seedling census in gaps created by thinning.
Table 2
The average quantity (dry weight) and thickness of litters for each
treatment in 1 m 1 m area
Litter
thickness
(cm)
Litter
weight
(g)
Litter
weight/stem
density
Treatment 1 (20%) 3.0
a
740.5
a
0.294
Treatment 2 (30%) 1.8
b
347.0
b
0.165
Treatment 3 (50%) 1.1
b
139.1
b
0.094
Treatment 4 (0%) 4.5
a
1112.1
a
0.309
Data not followed by the same letter are significantly different at
level P<0:05 according to the result of Kruskal–Wallis test with
Bonferroni-type multiple comparison.
Table 3
Mean values of canopy openness (%), direct light and diffuse light at height of 1.0 m in each treatment and each gap
Treatment or gap
Treatment 1
(20% thinned)
Treatment 2
(30% thinned)
Treatment 3
(50% thinned)
Treatment 4
(unthinned)
1 March 2000
Canopy openness (%) 15.7 18.9 33.1 8.5
Percentage of direct light (%) 6.3 16.4 43.8 10.9
Direct light (W m
2
) 51.4 133.8 357.4 88.9
Diffuse light (W m
2
) 780.5 940.9 1058.3 532.6
0.5 gap size 1.0 gap size 1.5 gap size 0.0 gap size
22 October 2000
Canopy openness (%) 15.9 27.9 46.3 8.2
Percentage of direct light (%) 19.5 16.7 31.3 12.5
Direct light (W m
2
) 159.1 136.3 255.4 97.9
Diffuse light (W m
2
) 557.1 789.3 1435.1 270.5
344 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
2.6. Data analysis
As the distributions of number of seedlings, litter
thickness and litter quantity were not normal, Krus-
kal–Wallis test (K–W test) was used to test the dif-
ference of observations among the four treatments.
Regression analysis was used to test whether relation-
ships existed among seedling density, height, age,
canopy openness and gap size or thinned treatment.
3. Results
3.1. Micro-site conditions
3.1.1. Leaf and branch litter
Statistically significant differences were not found
between treatments 1, and 4, and treatments 2, and 3
(Kruskal–Wallis test with Bonferroni-type multiple
comparison, P<0:05) (Table 2). There were 740.5,
Fig. 4. Soil water content at different depths for each treatment during May 2000 and November 2001. (A) 0–10 cm, (B) 20–30 cm, (C)
>50 cm.
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 345
347.0, 139.1 and 1112.1 g m
2
litter (total dry weight)
in treatments 1, 2, 3, and 4, respectively. The depth is
in the same ranking as quantity, i.e. 3.0, 1.8, 1.1 and
4.5 cm in treatments 1, 2, 3, and 4, respectively. The
ratio of litter quantity and stem density varied between
0.094 and 0.294 in the four treatments (Table 2).
3.1.2. Light regime
Canopy openness at 1.0 m above the ground in each
treatment was shown in Table 3. It indicates that if OP
is considered as 1 in treatment 4 (unthinned), the
canopy openness is 1.85, 2.22 and 3.89 in treatments
1, 2, and 3, respectively. The values of OP for 0.0, 0.5,
1.0 and 1.5 gap sizes were 8.2, 15.9, 27.9 and 46.3%,
respectively. The amount of open sky visible from the
hemispherical silhouettes is related to the quantity of
light reached the forest floor (Kobayashi and Kami-
tani, 2000; Myers et al., 2000). Both direct light and
Fig. 5. Mean wind profiles inside the four thinning treatments
U
in(z)
is wind speed (m s
1
) at height zin the forest stand, U
H
is
wind speed (m s
1
) at height H,His tree height (m).
Table 4
Survival of P. thunbergii seedlings in the four treatments for four growing seasons after thinning
Number surviving to year (seedlings 4 m
2
)
Recruited
in years
September
1998
April
1999
September
1999
May
2000
October
2000
May
2001
October
2001
Treatment 1 (20% thinned)
0–1 64 4930 37416127
1–2 7 215 11 9 710
2–3 2 1533
3–4000
4–5 0
Treatment 2 (30% thinned)
0–1 46 4143 34275326
1–26 315 1112914
2–3 3 3766
3–4323
4–5 0
Treatment 3 (50% thinned)
0–1 52 4548 45212717
1–2 11 9 20 16 25 22 12
2–399161513
3–48814
4–5 8
Treatment 4 (unthinned)
0–1 47 2735 11206531
1–2 9 15 3106
2–3 0 0000
3–4000
4–5 0
346 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
Fig. 6. Seedling density sorted by gap size and age in the fourth growing season after thinning.
Fig. 7. Distribution of seedling density at various positions in-gaps with different size gaps. (A) North–south transect, and (B) west–east transect.
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 347
diffuse light follow the same order to the stand den-
sities (Table 3).
3.1.3. Soil water content
Seasonal changes of soil water content by drying
soil samples of three-horizon indicated that the aver-
age of water content was correlated to the thinning
intensities positively, i.e. the highest in treatment 3,
and the lowest in treatment 4. The differences among
the four treatments became greater in each layer
during the drought period (July–August) (Fig. 4).
3.1.4. Wind profiles
The differences of the wind profiles among the four
treatments were quantitatively evaluated with the
attenuation coefficient ain Eq. (1).a¼3:22, 2.26,
1.98 and 1.81 for treatments 4 (unthinned), 1 (20%
thinned), 2 (30% thinned) and 3 (50% thinned),
respectively. This ranking also exactly follows the
ranking of stand densities for the treatments. There-
fore, it is concluded that greater attenuation of wind
speed is related to the canopy density in the various
thinning intensities (Fig. 5).
3.2. Seedling emergency, survival and
establishment in four growing seasons
Age composition of pine seedlings ranged from
1 year to the full 5 years among the four treatments.
Almost all of seedlings in treatment 4 (unthinned,
Fig. 8. Distribution of seedling density according to seedling age at various positions of 1.5 and 1.0 gap sizes. (A) West–east transect of 1.5
gap sizes, and (B) north–south transect of 1.5 gap sizes; (C) west–east transect of 1.0 gap sizes, and (D) north–south transect of 1.0 gap sizes.
348 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
0.0 gap size) were 1-year-old during the experimental
period and no seedlings could survive to the next
autumn after emergence (Table 4). In treatment 1
(20% thinned, 0.5 gap size), the emergence and sur-
vival were almost the same as in treatment 4, i.e.
nearly all of seedlings were 1-year-old although three
seedling of 2 years appeared during the experimental
period (Table 4). In treatment 2 (30% thinned, 1.0 gap
size), individual seedlings survived from 1 to 4 years,
most seedlings recruited in year 1 and 2. Seedlings
began in year 1 and extended through to year 5 in
treatment 3 (50% thinned, 1.5 gap size), individuals
survived from all years of recruitments (Table 4).
Most seedlings emerging in the spring disappeared in
the next summer for all of the four treatments, i.e. the
mortality rate was very high. Emergence of seedlings
showed no significant difference among the four treat-
ments (P<0:05) before May 2000 (Kruskal–Wallis
test and Bonferroni-type multiple comparison among
the four treatments). However, the number of older
seedlings was obviously different among the four treat-
ments (Table 4). Seedlings more than 5-year-old only
appeared in treatment 3 (50% thinned, 1.5 gap size).
3.3. Effects of gap size and within-gap position
on regeneration
3.3.1. Seedling density in different size gaps
The density of total seedlings did not vary consid-
erably among different gap sizes at the fourth growing
season after the gaps formed by thinning. Densities of
seedlings older than 1 year increased with an increase
of gap size (Fig. 6). Almost no seedlings older than 2
years in 0.0 gap sizes (control), and no seedlings older
than 4 years in 0.5, and 1.0 gap sizes could be
observed, while all age-class (1–5 years) seedlings
were observed in 1.5 gap sizes.
3.3.2. Seedling distribution related to
within-gap positions
Survival and establishment of seedlings differed
significantly among different sizes of gaps at the
end of the fourth growing season after thinning.
The distribution of seedlings within-gap positions
showed a similar tendency in the three gap sizes
(0.5, 1.0 and 1.5) (Fig. 7). Number of seedlings did
not peak at the gap center but near the north, west and
east edges of the gaps. This result seems to be con-
sistent with the fact that P. thunbergii is a light-
demand species. Distribution of seedlings divided
by age in 1.0 and 1.5 gap sizes plotted in Fig. 8
showed that 5-year-old seedlings only occurred near
the west and north edges of 1.5 gap sizes, but 2, 3 and
4-year-old seedlings occurred in each edge of both gap
sizes.
3.3.3. Seedling growth in gaps of different sizes
Seedling establishment had obviously failed in 0.0
and 0.5 gap sizes; therefore, the observation of seed-
ling growth was only conducted in 1.0 and 1.5 gap
Fig. 9. Growth of seedling base-diameter and height in 1.5 and 1.0 gap sizes (nis sample size. The vertical bar is standard error).
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 349
sizes. Height of seedling in 1.5 gap sizes varied
between 5 and 13 cm at year 2, 10 and 23 cm at year
3, 16 and 40 cm at year 4, and 32 and 60 cm at year 5,
while in 1.0 gap sizes, seedling height varied between
5 and 11 cm at year 2, 10 and 17 cm at year 3, 12 and
30 cm at year 4, and no seedlings extended to year 5
(Fig. 9). Both mean height and mean base diameter in
1.5 gap sizes were significantly higher than those in
1.0 gap sizes (P<0:05, t-test for independent samples
with unequal variance). Height of the seedlings varied
Fig. 10. Mean height of naturally regenerated seedlings older than 1 year (including seedlings of 2, 3, 4 and 5 years), showing the means and
the standard errors (vertical bar). (A) West–east transect, and (B) north–south transect.
350 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
greatly along both the north–south and west–east
transects (Fig. 10). The results showed a tendency
of increasing height across the gap from east edge to
west edge and from south edge to north edge.
4. Discussion and conclusions
4.1. Effects of micro-site conditions on
regeneration
Water and light availability, temperature, wind
regime, seed yield, pathogens, thigmo-morphogenesis
and seed-bed condition may potentially affect the pine
seed germination, seedling emergence and establish-
ment (Telewski and Jaffe, 1981; Ogasawara, 1988;
Gong et al., 1991; Futai and Nakai, 1993; Zeng et al.,
1996; Meer et al., 1999). Results from this observation
indicated that seedling density and growth (older than
1 year) increased significantly with increasing gap
size. The establishment succeeded in large gaps but
failed in small gaps and under canopy. The micro-site
conditions observed in this experiment may contribute
to this trend.
The amount of light received on the forest floor is
directly related to the size of the canopy opening. Gap
size is the essential factor in seedling establishment
(Meer et al., 1999). The canopy of unthinned treatment
(control) allows little sunlight to pass through (OP
8%), but the canopy of 50% thinned treatment
(treatment 3, 1.5 gap size) allows the maximum
amount of light to reach the forest floor. Consequently,
the micro-environment in large gaps is brighter than
that in small gaps and under canopy, which should be
favorable for the pine seed germination and seedling
establishment. However, the number of regenerated
seedlings did not peak at the gap center, indicating that
one possible reason is that strong light may restrain P.
thunbergii seeds to germinate in the center of the large
gaps (Kyereh et al., 1999).
Soil water conditions vary greatly according to gap
sizes (Ochiai et al., 1994). In this experiment, the soil
water content was higher in the most intensely thinned
plot (50% thinned, 1.5 gap sizes) than that in other plots,
especially in the growing period (May–August). The
water content in the top soil (0–10 cm), where seedling
root systems are distributed, may be the most significant
for seedling survival and growth. Observations from
this study showed that the water content of the top soil
was much higher in thinned plots than in unthinned plot
during the drier period (July–August) (Fig. 4). This
relatively moister soil in intensely thinned treatment or
large gaps is likely to increase the survival and growth
of seedlings. Evapotranspiration is considered as one of
the most important factors influencing the character-
istics of soil water conditions, while, because of the
reduction of stem density by thinning, evapotranspira-
tion was less in the stands with lower density. Although
the evaporation in the thinned plot with large gaps was
greater than in the unthinned treatment with small gaps,
the transpiration was much less.
Wind regime in the forest stand may be an indirect
factor modifying the micro-environment in gaps.
Firstly, an increase in wind speed often causes a decline
in transpiration rate (chilling effect) when leaves are
brightly illuminated (Grace, 1988). Secondly, wind
speed increases stem-flow volume in a rainfall event
(Kuraji et al., 1997). Therefore, the total input of rainfall
into canopy gap is higher than that under canopy.
Additionally, earlier study on responses of young trees
to wind suggests that wind action stimulates the growth
of seedlings (Grace, 1988; Stokes et al., 1995; Telewski,
1995), the relatively strong wind speed was observed in
the intensely thinned treatment (Fig. 5). Therefore,
seedlings in the large gaps should grow more strongly
than those in the small gaps.
Overall, the short-time effects of litter on regenera-
tion were mostly negative (Yajima, 1988; Kikuchi et al.,
1996; Toda et al., 1998; Xiong and Nilsson, 1999).
Litter on the forest floor in the unthinned stand was
much more abundant than that in the thinned stands
(Tab le 2). This means that the needles and branches
have longer life-spans under weaker light, less wind and
less moisture. Though the ratio of litter quantity to
standing stocks varied, it was higher in treatments 4 and
1 than that in treatments 3 and 2. Namely, the litter fallin
treatments 4 (unthinned) and 1 (20% thinned) are more
than in treatments 3 (50% thinned) and 2 (30% thinned).
It was the more litter in less thinned treatments that
impeded the survival and growth of seedlings.
4.2. Effects of gap sizes on regeneration
Gap size seems to be the main factor affecting the
growth of seedlings. The better seedling establishment
and growth in large gaps are likely to be the combined
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 351
consequences of (1) increased light and water avail-
ability and (2) the decreased litter accumulation or
quick decomposition. With an increase in gap size, the
stress of competitions for moisture and nutrient from
surrounding mature trees in the gap area is decreased.
This is also good for seedling growth in the gaps
(Madsen and Larsen, 1997). According to Wang et al.
(1998), regeneration in forest is controlled by both
above- and under-ground environmental conditions, of
which, light availability above ground and soil water
content under ground play important roles. The dis-
tribution patterns of seedlings in gaps supported the
Gap-partitioning hypothesis, i.e. the seedlings of light-
demand tree species in northern hemisphere distribute
in the north edge of the gaps. Additionally, the dis-
tribution patterns of P. thunbergii seedlings in gaps
also distributed in the west and east edges of the gaps.
The results of this investigation are consistent with
those of many other studies that have highlighted the
pronounced variation in light availability, which
occurs in many other forests (Lawton and Putz,
1988; Spies and Franklin, 1989; Lertzman et al.,
1996; Gobbi and Schlichter, 1998; Meer et al.,
1999; Myers et al., 2000). In this investigation, P.
thunbergii seedling of 1-year-old was not absent in
any size gaps or thinning treatments, but the seedlings
older than 1 year or more were absent both under
canopy and in small gaps. Gap size and within-gap
position had significant effects on seedling growth and
establishment. These results indicate that although P.
thunbergii seeds are clearly able to germinate in the
small gaps even under canopy, the seedlings appar-
ently require a minimum canopy gap size 1.0, or OP
>30% in order to survive and develop into seedling,
and may require at least canopy gap size 1.5, or OP
>40% for further developing into sapling. The experi-
ment indicates that treatment 3 (50% thinned) can
provide the best conditions for natural regeneration of
P. thunbergii. These findings therefore confirm the
view that P. thunbergii is light-demand tree species
(Ogasawara, 1986; Murai et al., 1992), and may
suggest that natural regeneration of P. thunbergii
demands gaps in the pine forest.
4.3. Implications and conclusions
It is often advocated that ecologically sound silvi-
cultural practices should be based on the knowledge in
natural forest processes (Lertzman et al., 1996; Meer
et al., 1999). Other earlier studies indicate that small-
scale forest disturbance plays an important role in the
natural tree regeneration of many forest types (Lawton
and Putz, 1988; Palik and Pederson, 1996; Gray and
Spies, 1996; Gobbi and Schlichter, 1998; Meer et al.,
1999; Myers et al., 2000; Vickers and Palmer, 2000).
These studies suggest that silvicultural systems for the
pine forests could include gap-cutting systems using a
variety of gap sizes, and the gap cutting silviculture
may enhance the development of a multi-layered
forest. In this coastal forest area, severely natural
disturbance regime, i.e. extreme wind, is at a relatively
low frequency (Quine, 2000; Zhu et al., 2001b).
Therefore, thinning (patch-pattern) strategy is a good
silvicultural technique as it produces openness or
gaps, and further provides suitable conditions for
natural regeneration. Another reason why thinning
strategy (patch-pattern thinning) may be a preferred
technique is that the shelter benefits of coastal forest
along the shoreline have not been affected, at least in
current thinning intensity (50%).
In conclusion, it is believed that thinning (patch-
pattern) strategy is potentially a viable silvicultural
practice in the coastal pine forest, and that the choice
for its application depends on the objectives of the
coastal forest management. From the view of regenera-
tion and protection, the patch-pattern thinning which
produces the small gaps (gap size <1.0) should be
avoided as seedling establishment is severely reduced
and chance of regeneration failure is increased. Thin-
ning intensity of 50% in patch-pattern is suggested in
current coastal pine forest. In addition, the further
investigations including vegetation changes, responses
of other species to the thinning and other environmental
conditions after the patch-pattern thinning need to be
addressed.
Acknowledgements
We would like to acknowledge Professor Tomohiko
Kamitani for his valuable review and comment on this
manuscript. We are grateful to Prof. Dali Tao for his
checking the English manuscript and valuable com-
ments. We also thank the editors and referees for their
helpful comments and careful revision on this manu-
script. Thanks are also due to Mr. Jiuyi Zhu for his
352 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354
great help in the field investigation. This study was
supported by the innovation research project of the
Chinese Academy of Sciences and Monbusho of Japan
government.
References
Amiro, B.D., 1990. Comparison of turbulence statistics within three
boreal forest canopies. Boundary-Layer Meteorol. 51, 99–121.
Brett, R.B., Kilnka, K., 1998. A transition from gap to tree-island
regeneration patterns in the subalpine forest of south-coastal
British Columbia. Can. J. For. Res. 28, 1825–1831.
Brokaw, N.V.L., 1987. Gap-phase regeneration of three pioneer tree
species in a tropical forest. J. Ecol. 75, 9–19.
Canham, C.D., 1989. Different responses to gaps among shade-
tolerant tree species. Ecology 70, 548–550.
Cionco, R.M., 1985. Modeling windfields and surface wind profiles
over complex terrain and within vegetation canopies. In:
Hutchison, B.A., Hicks, B.B. (Eds.), The Forest–Atmosphere
Interaction. Reidel, Dordrecht, pp. 501–520.
Dalling, J.W., Hubbell, S.P., Silvera, K., 1998. Seed dispersal,
seedling establishment and gap partitioning among tropical
pioneer trees. J. Ecol. 86, 674–689.
Futai, K., Nakai, I., 1993. Influence of pine wilt disease on Pinus
densiflora and Pinus thinbergii regeneration. J. Jpn. For. Soc.
75, 287–292 (in Japanese with English abstract).
Gobbi, M., Schlichter, T., 1998. Survival of Austrocedrus chilensis
seedlings in relation to micro-site conditions and forest
thinning. For. Ecol. Manage. 111, 137–146.
Gong, Y.L., Swqine, M.D., Miller, H.G., 1991. Effects of fencing
and ground preparation on natural regeneration of native
pinewood over 12 years in Glen Tanar, Aberdeenshire. Forestry
64, 157–168.
Grace, J., 1988. Plant response to wind. Agric. Ecosyst. Environ.
22/23, 71–88.
Gray, A.N., Spies, T.A., 1996. Gap size, within-gap position and
canopy structure effects on conifer seedling establishment. J.
Ecol. 84, 635–645.
Groß, G., 1993. Numerical Simulation of Canopy Flows. Springer,
Berlin.
Hong, S.K., Nakagoshi, N., Song, H.K., 1994. Productivity of
secondary pine forests and their sustainable forest managementin
Korea. In: Sakai, H., Takahashi, I. (Eds.), Proceedings of IUFRO
International Workshop on Sustainable Management. The
University Forests, The University of Tokyo Press, pp. 81–89.
Jennings, S.B., Brown, N.D., Shell, D., 1999. Assessing forest
canopies and understorey illumination: canopy closure, canopy
cover and other measures. Forestry 72, 59–74.
Jiang, F.Q., Zhu, J.J., Zhou, X.H., Lin, H.M., 1994. Theory of
protective maturity and regeneration of shelterbelt. Chin. J.
Appl. Ecol. 5, 337–341 (in Chinese with English abstract).
Jiang, F.Q., Zeng, D.H., Zhu, J.J., 1997. Fundamentals and
technical strategy of sand-fixation forest management. Chin.
J. Desert 17, 250–254 (in Chinese with English abstract).
Kikuchi, J., Iido, S., Akemma, T., Futai, K., 1996. The role of litter
and humus layer as a mortality factor of seedlings of Japanese
red pine, Pinus desiflora Sieb. Et Zucc. J. Jpn. For. Soc. 71,
37–40.
Kobayashi, M., Kamitani, T., 2000. Effects of surface disturbance
and light level on seedling emergence in a Japanese secondary
deciduous forest. J. Veget. Sci. 11, 93–100.
Krauchi, N., Brang, P., Schonenberger, W., 2000. Forest of
mountainous regions, gaps in knowledge and research needs.
For. Ecol. Manage. 132, 73–82.
Kuraji, K., Tanaka, N., Shiraki, K., Karakama, I., Ohta, T., 1997.
Effects of wind speed on stem-flow volume in a mature
Cryptomeria japonica and Chamaecyparis obtusa stand. J. Jpn.
For. Soc. 79, 215–221 (in Japanese with English abstract).
Kyereh, B., Swaine, M.D., Thompson, J., 1999. Effect of light on
the germination of forest trees in Ghana. J. Ecol. 87, 772–783.
Landsberg, J.J., James, G.B., 1971. Wind profiles in plant canopies,
studies on an analytical model. J. Appl. Ecol. 8, 729–741.
Lawton, T.O., Putz, F.E., 1988. Natural disturbance and gap-phase
regeneration in a wind-exposed tropical cloud forest. Ecology
69, 764–777.
Lertzman, K., 1992. Patterns of gap-phase replacement in a
subalpine, old-growth forest. Ecology 73, 657–669.
Lertzman, K.P., Sutherland, G.D., Inselberg, A., Saunders, S.C.,
1996. Canopy gaps and the landscape mosaic in a coastal
temperate rain forest. Ecology 77, 1254–1270.
Madsen, P., Larsen, J.B., 1997. Natural regeneration of beech
(Fagus sylvatica L.) with respect to canopy density, soil
moisture and soil carbon content. For. Ecol. Manage. 97,
103–113.
Mclaren, B.E., Janke, R.A., 1996. Seedbed and canopy cover
effects on balsam fir seedling establishment in Isle Royal
National Park. Can. J. For. Res. 26, 782–793.
Meer, P.V.D., Digan, P., Savench, A.G., 1999. Effect of gap size on
seedling establishment, growth and survival at 3 years in
mountain ash (Eucalyptus regnans F. Muell.) forest in Victoria,
Australia. For. Ecol. Manage. 117, 33–42.
Murai, H., Ishikawa, M., Endo, J., Tadaki, R., 1992. The Coastal
Forest in Japan. Soft Science Inc., Tokyo (in Japanese).
Myers, G.P., Newton, A.C., Melgarejo, O., 2000. The influence of
canopy gap size on natural regeneration of Brazil nut
(Bertholletia excelsa) in Bolvia. For. Ecol. Manage. 127,
119–128.
National Astronomical Observatory (NAO), 2000. Rika nenpyo
(Chronological Scientific Tables). Maruzen Co. Ltd. Press (in
Japanese).
Ochiai, Y., Okuda, S., Sato, A., 1994. The influence of canopy gap
size on soil water condition in a deciduous broad-leaved
secondary forest in Japan. J. Jpn. For. Soc. 76, 308–314.
Ogasawara, R., 1986. The natural regeneration of the pine
forests (Pinus densiflora,Pinus thunbergii)intheWest-
Daisen state forest. Res. Bull. Tottori Univ. For. 16, 215–226
(in Japanese).
Ogasawara, R., 1988. Physiological and ecological studies on the
growth of Japanese black pine forests on sand dunes. Res. Bull.
Tottori Univ. For. 17, 37–137 (in Japanese with English
abstract).
J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354 353
Ooishi, T., Kikuchi, S., Yajima, T., Shimizu, O., Nakamura, F.,
1998. Effects of selection cutting on forest dynamics in the
Shiretoko national forest, stand structure and regeneration
pattern, 10 years after cutting. Res. Bull. Hokkaido Univ. For.
55, 349–368 (in Japanese with English abstract).
Palik, B.J., Pederson, N., 1996. Overstory mortality and canopy
disturbance in longleaf pine ecosystems. Can. J. For. Res. 26,
2035–2047.
Perry, D.A., 1994. Forest Ecosystem. The Johns Hopkins
University Press, Baltimiore.
Peltola, H., 1996. Model computations on wind flow and turning
moment by wind for Scots pines along the margins of clear-cut
areas. For. Ecol. Manag. 83, 203–215.
Quine, C.P., 2000. Estimation of mean wind climate and
probability of strong winds for wind risk assessment. Forestry
73, 247–258.
Seiwa, K., Kikuzawa, K., 1994. Optimum thinning ration in a
Japanese larch stand. In: Sakai, H., Takahashi, I. (Eds.),
Proceedings of IUFRO International Workshop on Sustainable
Management. The University Forests, The University of Tokyo
Press, pp. 237–245.
Spies, T., Franklin, J.F., 1989. Gap characteristics and vegetation
response in coniferous forests of the Pacific northwest. Ecology
70, 543–545.
Stokes, A., Fitter, A.H., Coutts, M.P., 1995. Responses of young
trees to wind: effects on root growth. In: Coutts, M.P., Grace, J.
(Eds.), Wind and Trees. Cambridge University Press, Cam-
bridge, pp. 264–275.
Steege, H., 1994. Hemiphot: A Programme to Analyze Vegetation
Indices, Light and Light Quality from Hemispherical
Photographs. The Tropenbos Foundation, Wageningen, The
Netherlands.
Telewski, F.W., Jaffe, M.J., 1981. Thigmo-morphogenesis: changes
in the morphology and chemical composition induced by
mechanical perturbation of 6-month-old Pinus taeda seedlings.
Can. J. For. Res. 11, 380–387.
Telewski, F.W., 1995. In: Coutts, M.P., Grace, J. (Eds.), Wind-
induced Physiological and Developmental Responses in Trees.
Cambridge University Press, Cambridge, pp. 237–263.
Toda, H., Sugisak, H., Haibara, K., 1998. Effect of water content on
the rate of nitrogen mineralization and immobilization in the A
0
layer of forest soils. J. Jpn. For. Soc. 80, 262–269 (in Japanese
with English abstract).
Vickers, A.D., Palmer, S.C.F., 2000. The influence of canopy cover
and other factors upon the regeneration of Scots pine and its
associated ground flora within Glen Tanar National Nature
Reserve. Forestry 73, 37–49.
Wang, Z.Q., Wang, Q.C., Chen, Q.S., 1998. Spatial heterogeneity
of soil nutrients in old-growth forests of Korean pine. J. For.
Res. 9, 240–244.
Xiong, S.J., Nilsson, C., 1999. The effects of plant litter on
vegetation: a meta-analysis. J. Ecol. 89, 984–994.
Yajima, T., 1988. A survey on the quantity of dead leaves and
culms of Sasa senanensis in a mixed forest. Res. Bull.
Hokkaido Univ. For. 45, 683–690 (in Japanese with English
abstract).
Yamamoto, K., 2000. Estimation of the canopy-gap size using two
photographs taken at different heights. Ecol. Res. 12, 203–208.
Zeng, D.H., Jiang, F.Q., Fan, Z.P., Zhu, J.J., 1996. Stability of
Mongolian pine (Pinus sylvestris var. mongolica) plantation on
sandy land. Chin. J. Appl. Ecol. 7, 337–343 (in Chinese with
English abstract).
Zhu, J.J., Matsuzaki, T., Gonda, Y., 2001a. Extreme wind over a
pine coastal forest established for sand-control at Aoyama of
Niigata in Japan. J. Erosion Contr. 54, 12–20.
Zhu, J.J., Matsuzaki, T., Gonda, Y., 2001b. Wind profiles in a
coastal forest of Japanese black pine (Pinus thunbergii Parl.)
with different thinning intensities. J. For. Res., The Jpn. For.
Soc. 6, 287–296.
Zhu, J.J., Matsuzaki, T., Gonda, Y., 2003. Optical stratification
porosity as a measure of vertical canopy structure in a Japanese
coastal forest. For. Ecol. Manage. 173, 89–104.
354 J.-j. Zhu et al. / Forest Ecology and Management 182 (2003) 339–354