免费文献传递   相关文献

Population Ecological Characteristics of the Rare and Endangered Plant Camellia rhytidophylla from Guizhou

贵州珍稀濒危植物皱叶瘤果茶的种群生态特征研究



全 文 :贵州珍稀濒危植物皱叶瘤果茶的种群生态特征研究∗
刘海燕1ꎬ 杨乃坤2ꎬ 邹天才3∗∗ꎬ 李媛媛1ꎬ 洪  江3
(1 贵州省植物园ꎬ 贵阳  550004ꎻ 2 贵州大学林学院ꎬ 贵阳  550025ꎻ 3 贵州科学院ꎬ 贵阳  550001)
摘要: 皱叶瘤果茶 (Camellia rhytidophylla) 是颇具经济价值的特有濒危植物ꎬ 分布于贵州高原亚地区 (Ⅲ
D10d) 开阳县花梨镇洛旺河流域海拔 573~920 m 的常绿阔叶落叶混交林中ꎬ 研究其种群生命特征与保育
利用具有重要意义ꎮ 选择皱叶瘤果茶疑似分布区 (6 km2) 开展踏查ꎬ 在密集分布区设置样地进行详查ꎬ 分
析其种群结构、 动态和空间分布格局ꎮ 结果表明: 皱叶瘤果茶种群结构为增长型ꎬ 幼龄树在种群中占的比
重达 46􀆰 38%ꎬ 种群密度大小为幼龄树>中龄树>成年树ꎬ 种群存活曲线为 Deevey ̄Ⅲ型ꎬ 死亡率曲线和消失
率曲线分别在Ⅰ龄期和Ⅳ龄期出现 2个高峰ꎬ 随后又同时在Ⅲ龄期和Ⅵ龄期出现 2个低谷ꎮ 2个样地内的
幼树种群在所有尺度下均呈集群分布ꎬ 中龄树种群在小尺度上呈集群分布ꎬ 在大尺度上则表现为随机分
布ꎬ 成年树种群由于人为活动干扰和生境异质性而使空间分布格局明显不同ꎮ 各发育阶段的空间分布格局
有较大差异ꎬ 关系不密切ꎬ 均表现为负相关或不相关ꎮ 皱叶瘤果茶种群空间分布格局是其物种生物学特
性、 生境异质性及人为干扰等因素共同作用的结果ꎬ 自然繁殖率极低是限制种群扩散的关键因素ꎬ 生境异
质性、 山丘阻碍种子散布以及人类活动的干扰是其种群狭限分布的主要原因ꎮ
关键词: 皱叶瘤果茶ꎻ 特有植物ꎻ 种群生态特征ꎻ 空间分布格局ꎻ 静态生命表
中图分类号: Q 948          文献标志码: A                文章编号: 2095-0845(2015)06-837-12
Population Ecological Characteristics of the Rare and Endangered
Plant Camellia rhytidophylla from Guizhou
LIU Hai ̄yan1ꎬ YANG Nai ̄kun2ꎬ ZOU Tian ̄cai3∗∗ꎬ LI Yuan ̄yuan1ꎬ Hong Jiang3
(1 Guizhou Botanical Gardenꎬ Guiyang 550004ꎬ Chinaꎻ 2 College of Forestryꎬ Guizhou Universityꎬ
Guiyang 550025ꎬ Chinaꎻ 3 Guizhou Academy of Sciencesꎬ Guiyang 550001ꎬ China)
Abstract: Camellia rhytidophylla is endemic and endangered and with important economic speciesꎬ from the Ⅲ
D10d in eastern Asiaꎬ which distributed at 573-920 m in evergreen broad ̄leaved deciduous mixed forest in Luowang
river valley of Kaiyang county. Soꎬ it is significant to study the population characteristicsꎬ conservation and utiliza ̄
tion. We took general survey in suspected distribution area about 6 square kilometersꎬ set up plots in dense areaꎬ
and analyzed the population structureꎬ development and spatial distribution pattern. The results showed that the
structure of C􀆰 rhytidophylla populations were increasing and the proportion of young tree in population was 46􀆰 38%.
The size of population density was young shrubs>middle ̄aged shrubs>adult shrubs. The survival curve of population
was Deevey ̄Ⅲ model. There were 2 peaks in the Ⅰand Ⅳ age ̄classes on the mortality rate curve and disappearance
rate curve respectivelyꎬ then there are 2 troughs in the Ⅲ and Ⅵ age  ̄classed at the same time. The spatial distribu ̄
tion pattern of C􀆰 rhytidophylla significantly differentiated at different stages of developmentꎬ the young individuals
were aggregated at all spatial scales while the middle ̄aged individuals were aggregated at small spatial scales and
randomly distributed at larger scales. Differences in the distribution of adult individuals could be attributed to arti ̄
植 物 分 类 与 资 源 学 报  2015ꎬ 37 (6): 837~848
Plant Diversity and Resources                                    DOI: 10.7677 / ynzwyj201515105

∗∗
Funding: National Natural Science Foundation of China (31360075ꎬ 31560097)
Auther for correspondenceꎻ E ̄mail: 1211111951@qq􀆰 com
Received date: 2015-06-30ꎬ Accepted date: 2015-09-25
作者简介: 刘海燕 (1982-) 女ꎬ 副研究员ꎬ 主要从事植物资源学研究ꎮ E ̄mail: liuhaiyan301@163􀆰 com
ficial disturbance and habitat heterogeneity. The spatial pattern was not close in different stages of developmentꎬ
which all showed negative or no correlation. The spatial distribution pattern of C􀆰 rhytidophylla was the interaction
of the factorsꎬ such as its biological characteristicsꎬ habitat heterogeneityꎬ and artificial disturbanceꎬ etc.. The
key factor limiting population development was low natural reproduction rateꎻ the primary factors causing its en ̄
demic distribution mainly included habitat heterogeneityꎬ topographical constraints on seed dispersal and artificial
disturbance.
Key words: Camellia rhytidophyllaꎻ Endemic plantꎻ Population ecological characteristicsꎻ Spatial distribution pat ̄
ternꎻ Static life table
  Population ecology examines characteristics of
individualsꎬ populationsꎬ communitiesꎬ and connects
various fields of ecology ( Zhang et al.ꎬ 2007a).
Population dynamics are the core of this branch of e ̄
cologyꎬ and changes in the number and spatial dis ̄
tribution of populations are the focus points of popu ̄
lation ecological research. Life tables and survival
curves are important tools for statistical analysis of
dynamic changes in populations (Xu et al.ꎬ 2005ꎻ
Yan et al.ꎬ 2001). Analysis of spatial patterns is an
important method for studying interactions between
populations and relationships between populations
and the environment. Different spatial patterns re ̄
flect the status of resource used by populations and
can reveal the rules behind population growth and
succession (Condit et al.ꎬ 2000ꎻ Zhangꎬ 1998).
The species of the genus Camellia (family Thea ̄
ceae)ꎬ which have high economic and ornamental
valueꎬ are an important component of subtropical ev ̄
ergreen broad ̄leaved forests. Many species of the tea
group (such as C􀆰 sinensis and C􀆰 assamica) are plan ̄
ted long for tea production in China and provided re ̄
nowned beverages in the international marketꎬ and
now widely cultivated in many parts of the world.
More than 60 species of the sect. Oleifera Chang and
sect. Camellia ( L.) Dyerꎬ which have seeds with
high oil contentꎬ are excellent materials for food and
industrial uses. In additionꎬ some species of genus
Camellia and several other genera of Theaceae are
important horticultural species with high ornamental
valueꎬ such as elegant shapeꎬ abundant flowersꎬ and
long flowering period (they bloom in winter after oth ̄
er plants have senesced) (DCPꎬ 2014ꎻ Zhang and
Renꎬ 1998ꎻ Li et al.ꎬ 2005b). These plants are pop ̄
ular in flower markets and possess strong internation ̄
al reputations with their long history of cultivation.
Guizhou Province is located in the northern edge
of the subtropical zone and has abundant Camellia
germplasmꎬ including C􀆰 rhytidophylla Y􀆰 K. Li et
M􀆰 Z. Yang. C􀆰 rhytidophylla is endemic to the Guizhou
Plateau Subregion (IIID10d) (Central China phyto ̄
geographic areaꎬ Eastern Asiatic Region) (Wu et
al.ꎬ 2010) and is known as the “ iron filings tree”
(DCPꎬ 2014). The species occurs in the understory
of mixed evergreen and deciduous broadleaved for ̄
ests in warmꎬ humid habitats at elevations between
570 and 920 m.
C􀆰 rhytidophylla is an evergreen shrub that typi ̄
cally grows to 2􀆰 0-5􀆰 0 mꎬ and was identified and
named in 1987 (Li and Yangꎬ 1987). This species
is rare and is only distributed in Kaiyang Countyꎬ
Guizhou Province. Its dark green leavesꎬ white flow ̄
ersꎬ and aesthetic shape have made it to be an ideal
species for afforesting and beautifyingꎬ with cluster
plantings in courtyards and gardens. With its resist ̄
ance to sulfideꎬ planting this species in mining areas
that have severe air pollution might help to reduce
sulfide contamination (Zouꎬ 2001). Systematic ex ̄
ploration of C􀆰 rhytidophylla will promote the deve ̄
lopment and protection of germplasm resources. Va ̄
rious researchers have studied the introduction and
cultivation of C􀆰 rhytidophylla and have reported on
its leaf anatomy and chemical content ( Liu et al.ꎬ
2011bꎻ Longꎬ 2013ꎻ Zhang et al.ꎬ 2010ꎻ Zouꎬ 2001ꎻ
Zou and Louꎬ 1995). Howeverꎬ systematic ecologi ̄
cal research is lacking for this speciesꎬ and no popu ̄
lation ecology studies have been published to date.
In this studyꎬ we conducted systematic field investi ̄
838                                  植 物 分 类 与 资 源 学 报                            第 37卷
gations of the population structure and dynamics and
spatial distribution of C􀆰 rhytidophylla. We discuss
the ecological characteristicsꎬ endangered statusꎬ
environmental adaptability of C􀆰 rhytidophylla popu ̄
lationsꎬ and the factors that limit survival and devel ̄
opment of its populationsꎬ and we attempt to provide
a scientific basis for its effective protection and the
rational use of its germplasm. Through this workꎬ we
also hope to provide a reference for further studies of
floristic composition and population characteristics of
other plants in the Guizhou Plateau as well as re ̄
search on endemic and rare plants in other regions of
the world.
1  Materials and methods
1􀆰 1  Study sites
We examined the reported natural distribution
areas of C􀆰 rhytidophylla (Li and Yangꎬ 1987) and
selected a study area in the evergreen broad ̄leaved
deciduous mixed forest region in the Luowang River
valley of Huali Townꎬ Kaiyang Countyꎬ Guizhou Prov ̄
inceꎬ China. C􀆰 rhytidophylla was distributed at ele ̄
vations between 570 and 920 m from 27°06′51″ to
27°08′03″ N and 107°04′08″ to 107°05′31″ E. We
surveyed the 4􀆰 14 km2 region in which C􀆰 rhytidoph ̄
ylla was thought to be distributed and conducted a
detailed investigation within a 1􀆰 05 km2 area where
it formed dense populations. These areas occurred in
a mountain ̄and ̄hill zone ( erosion landform) with
scattered flat areas interlaced with streams and val ̄
leys. The climate was subtropical humid monsoon
with average annual rainfall of 1 230 mmꎬ average
annual humidity of approximately 70%ꎬ maximum
and minimum temperatures of 35􀆰 4 ℃ and -10􀆰 1 ℃
(annual average = 13􀆰 6 ℃)ꎬ and cumulative annual
temperature of 5 000-5 100 ℃ . The majority (75%)
of rainfall occurs in May ̄October ( Agricultural re ̄
gional planning committee of Kaiyang countyꎬ 1989ꎻ
Liang and Zhangꎬ 2014).
C􀆰 rhytidophylla grows individually or in clusters.
The plant communities in which C􀆰 rhytidophylla oc ̄
curs are species ̄rich and structurally complex and
include 21 families and more than 30 species. The
tree layer (5-12 mꎬ 60%-80% canopy cover) in ̄
cludes Cyclobalanopsis myrsinifoliaꎬ Cinnamomum
camphoraꎬ Idesia polycarpa var. vestitaꎬ Terminalia
amtayꎬ Podocarpus nagiꎬ and Ilex franchetiana Loesꎻ
C􀆰 myrsinifolia is the dominant canopy species. The
shrub layer (2-5 mꎬ 40% canopy cover) includes
Ziziphus jujuba var. spinosaꎬ Nandina domesticaꎬ Pa ̄
nax pseudoginseng var. notoginsengꎬ Rhapis excelsaꎬ
and Elaeagnus umbellata. The herbaceous layer (0􀆰 2
-1􀆰 2 mꎬ approximately 25% canopy cover) mainly
includes Pteridium aquilinum var. latiusculumꎬ Nephro ̄
lepis auriculataꎬ Pyrrosia linguaꎬ Selaginella unci ̄
nataꎬ Polygonum perfoliatumꎬ Delphinium delavayi
var. pogonanthumꎬ Aspidistra elatiorꎬ and Miscanthus
sinensis. Some interlaminar lianasꎬ such as Smilax
china and Sargentodoxa cuneataꎬ are also present.
1􀆰 2  Field study
Field surveys and data collection were conducted
during 2013-2014 within the selected sites in Dax ̄
ianggou Sectionꎬ Qingjiang Village (Guizhou Plateau
Subregion IIID10d). Based on the survey resultsꎬ
two representative sampling areas (A and B) were
chosen on the North Slope in Daxianggouꎬ where
C􀆰 rhytidophylla was densely distributed (Table 1).
We established one 50 m × 50 m plot ( one of the
plot boundaries was oriented perpendicular to the
slope and another was parallel to the slope) in each
sampling area. We measured the vertical distance of
each C􀆰 rhytidophylla plant to the edge of the sampling
Table 1  Geographical location and habitat characteristics of Camellia rhytidophylla sampling sites
Sampling
site Geographical coordinates
Elevation
/ m
Sample area
/ m2
Canopy
coverage Aspect Slope Soil type Soil pH
A 27°07′18″Nꎬ 107°05′06″E 673 2500 0􀆰 8 51° NE 76° Silty clay Slightly acidic
B 27°07′22″Nꎬ 107°05′00″E 666 2500 0􀆰 6 23° NE 68° Silty clay Slightly acidic
9386期      LIU Hai ̄yan et al.: Population Ecological Characteristics of the Rare and Endangered Plant Camellia 􀆺     
plot using a rangefinder and recorded the coordinatesꎬ
heightꎬ basal diameterꎬ and crown size (north ̄south) of
each plant. In each of the 2 500 m2 plotsꎬ we estab ̄
lished five 5 m × 5 m subplots to investigate species
composition and height and canopy coverage of the
shrub layer. We also established one 1 ̄m2 sample
plot at the center of each subplot to measure species
compositionꎬ heightꎬ and canopy cov ̄
erage of the herbaceous layer.
1􀆰 3  Population age ̄class determination
C􀆰 rhytidophylla are short tree with many lateral
branchesꎬ which made it difficult to obtain wood cores.
Furthermoreꎬ destructive sampling to determine the
age of individual shrubs would violate the principle
of biodiversity protection. Thereforeꎬ we analyzed po ̄
pulation age structure according to diameter class
rather than age class as the reported research (Wang
et al.ꎬ 2010ꎻ Zhang et al.ꎬ 2013)ꎬ basal diameter
(D) was used as a standard to categorize the diame ̄
ter classes. The first diameter class was defined as D
= 0-2 cmꎬ with additional diameter classes defined
in 2 ̄cm increments. We plotted diagrams of popula ̄
tion diameter ̄class structure from these data.
1􀆰 4  Life table preparation and curve plotting
We used the method of Feng et al. (2003) to
generate a time ̄specific life table for C􀆰 rhytidophylla
using the diameter ̄class data. The order of diameter
classes from small to large was considered as a re ̄
flection of age classes from young to oldꎻ the first di ̄
ameter class corresponded to the first ( youngest)
age class and so on. The numbers of plants in each
diameter class were counted and the data were stand ̄
ardized. Age (diameter) classes were plotted on the
horizontal axisꎬ and the standard survival number (lx)
was plotted on the vertical axis to obtain the survival
curve. Additionallyꎬ we used qx and Kx as the verti ̄
cal axis to plot the mortality and disappearance rate
curvesꎬ respectively.
1􀆰 5  Spatial distribution pattern analysis
Population spatial distribution patterns were an ̄
alyzed using spatial point pattern analysis ( SPPA)
(also called Ripley’s K ̄function) (Cetis and Frank ̄
linꎬ 1987ꎻ Zhang and Mengꎬ 2004ꎻ Li et al.ꎬ 2005a).
The basic formula for Ripley’s K is:
K^( t) = A
n2∑

j = 1


i = 1

Wij
It(uij)  ( i ≠ j) (1)
    Where A is plot areaꎬ n is the total number of
pointsꎬ t is distance scaleꎬ and uij is the separation
between shrub i and shrub j. When uij is < tꎬ It
(uij)= 1ꎻ when uij is>tꎬ It (uij)= 0. Wij is the ratio
of the length of the arc of a circle that centers at
point i and with radius uij and falls in area A to the
circumference of the circle. Wij reflects the probabili ̄
ty that a point (plant) would be observed (Diglleꎬ
1983ꎻ Haaseꎬ 1995).
For a simplified explanation of the resultsꎬ we
used the revised formula of Ripley’s ̄K:
L( t) = K^( t) / π - t (2)
for analysis of spatial distribution patterns. To evalu ̄
ate the significance of the deviation of L ( t) from a
random distributionꎬ we used a Monte Carlo method
and calculated the 99% confidence interval (CI) of
L (t) using 10 000 random spatial simulations (Zhangꎬ
1998). When L ( t) was greater than the upper limit
of CIꎬ the distribution was aggregatedꎻ when L ( t)
was within the CIꎬ the distribution was randomꎻ when
L ( t) was smaller than the lower limit of CIꎬ the
distribution was uniform.
Point ( or multi ̄point) pattern analysis of two
or more developmental stages examines the relation ̄
ships between developmental stages (Condit et al.ꎬ
2000ꎻ Zhang and Mengꎬ 2004). Its definition and
calculation principle is:
K^12( t) =

n1n2

n1
j = 1

n2
i = 1

Wij
It(uij)  ( i ≠ j)  (3)
    Where n1 and n2 indicate the number of indivi ̄
duals (points) in Stage 1 and Stage 2ꎬ respectivelyꎻ
Aꎬ It (uij)ꎬ and Wij are as in Eq.(1)ꎻ and i and j
represent individuals in Stages 1 and 2ꎬ respectively.
Similarlyꎬ we used the Monte ̄Carlo method to
examine the fitted L12( t) CIꎬ to determine if a sig ̄
048                                  植 物 分 类 与 资 源 学 报                            第 37卷
nificant correlation existed between Stages.
L12( t) = K^12( t) / π - t (4)
When L12( t) is greater than the upper limit of CIꎬ
the correlation is significantly positiveꎻ when L12( t)
is within the CIꎬ there is no correlationꎻ when L12(t)
is smaller than the lower limit of CIꎬ the correlation
is significantly negative. All analyses were performed
using the ecological software package ADE ̄4 (Li et
al.ꎬ 2013).
2  Results
2􀆰 1  Distribution rangeꎬ population densityꎬ and
number of C􀆰 rhytidophylla
Through the field surveysꎬ we identified that the
natural distribution region of C􀆰 rhytidophylla covered
a 4􀆰 14 km2 area (2􀆰 3 km × 1􀆰 8 km)ꎬ which consis ̄
ted of a dense distribution in a 1􀆰 05 km2 (1􀆰 5 km ×
0􀆰 7 km) section and a sporadic distribution in the
remaining area (Fig􀆰 1). There were approximately
86 900 C􀆰 rhytidophylla individuals in the dense dis ̄
tribution region (The unweighted geometric mean to ̄
tal density in this area was 828 shrubs / haꎬ with the
highest survey ̄area mean density being 1 368 shrubs /
ha and the lowest being 444 shrubs / ha.)ꎬ and 6 000
plants in the remaining area (average population den ̄
sity = 20 shrubs / ha). The natural distribution regions
were narrowꎬ the population was smallꎬ and the pop ̄
ulation distribution was extremely uneven ( density
varied from 1 368 to <20 plants / ha).
2􀆰 2  Diameter ̄class distribution and age struc ̄
ture of C􀆰 rhytidophylla
The diameter (D) ̄class distribution of C􀆰 rhytido ̄
phylla had a pyramidal formꎻ young shrubs were most
abundantꎬ and there were fewer adult shrubs (Fig􀆰 2).
We examined a total of 414 individuals in the two
50 m × 50 m plots. Of theseꎬ 192 (46􀆰 38%) were
young plants (D=0-2􀆰 0 cm)ꎬ 138 (33􀆰 33%) were
middle ̄aged (D= 2􀆰 0-6􀆰 0 cm)ꎬ and 84 (20􀆰 29%)
were adult shrubs D > 6􀆰 0 cm. For shrubs with
ground diameter ≤6􀆰 0 cmꎬ the number of individual
plants in the population rapidly decreased as the di ̄
ameter increasedꎬ but this trend slowed for shrubs
with ground diameter >6􀆰 0 cmꎬ which indicated that
the age structure of C􀆰 rhytidophylla took a “ pyra ̄
mid” shape and that its population was expanding.
Fig􀆰 1  Location and size distribution of endemic Camellia rhytidophylla population. Yellow dotted line indicates the overall
distribution regionꎻ red dashed line indicates the dense distribution regionꎻ A and B are the locations of study plots
1486期      LIU Hai ̄yan et al.: Population Ecological Characteristics of the Rare and Endangered Plant Camellia 􀆺     
Fig􀆰 2  Histogram of diameter (D) classes of
Camellia rhytidophylla population
2􀆰 3  Time ̄specific life table of C􀆰 rhytidophylla
We established a time ̄specific life table for the
C􀆰 rhytidophylla population (Table 2)ꎬ which showed
that life expectancy increased for shrubs in the first
three age classesꎬ decreased slightly in the fourth age
classꎬ and increased in the fifth age class againꎬ and
then sharply declined. The life ̄expectancy trend showed
a “low ̄high ̄low” pattern that reflected differences in
survival rate over different age classes. Besidesꎬ al ̄
though there were large numbers of young shrubsꎬ the
survival rates of the first two age classes were signifi ̄
cantly lower because of high mortality and disappear ̄
ance rates that occurred during the first and second age
classes. These results indicated that there were very few
individuals that could survive to the next age class.
2􀆰 4  Survivalꎬ mortalityꎬ and disappearance rate
of C􀆰 rhytidophylla
Survivalꎬ mortalityꎬ and disappearance rate curves
for the C􀆰 rhytidophylla population were plotted (Fig􀆰 3).
The Deevey classification system recognizes three types
of survival curves: Deevey ̄Iꎬ Deevey ̄IIꎬ and De ̄
evey ̄III ( Jiangꎬ 1992)ꎻ the survival curve of the
C􀆰 rhytidophylla population was the Deevey ̄III type
(Fig􀆰 3A). Over the lifetime of C􀆰 rhytidophyllaꎬ the
mortality and disappearance rate curves showed two
peaks and two troughs (Fig􀆰 3B). In addition to nat ̄
ural mortalityꎬ we observed that 22􀆰 5% of shrubs
were loggedꎬ as evidenced by the remaining stumps.
2􀆰 5   Spatial distribution pattern of populations
in different age classes
Point distribution patterns of C􀆰 rhytidophylla shrubs
at three developmental stages (diameter classes) reve ̄
aled differences in density among the stages (Fig􀆰 4).
Plot A included 520ꎬ 316ꎬ and 180 youngꎬ middle ̄
agedꎬ and adult shrubs / haꎬ respectively. Plot B in ̄
cluded 248ꎬ 236ꎬ and 156 youngꎬ middle ̄agedꎬ and
adult shrubs / haꎬ respectively. Young shrubs were
most abundant in both plotsꎬ followed by middle ̄
aged shrubsꎬ and adult shrubs were least abundant.
Each of the developmental stages showed some de ̄
gree of aggregated distribution. We used spatial point
pattern analysisꎬ with Ripley’s K ̄functionꎬ to further
explore the relationship between distribution and scale
Table 2  Time ̄specific life table for the Camellia rhytidophylla population
Age class D / cm x Δx ax lx   dx   qx Lx   Tx   ex Sx Kx
I 0-2 1 2 192 1000􀆰 00 536􀆰 46 0􀆰 54 731􀆰 77 1656􀆰 25 1􀆰 66 0􀆰 46 0􀆰 77
II 2-4 3 2 89 463􀆰 54 208􀆰 33 0􀆰 45 359􀆰 38 924􀆰 48 1􀆰 99 0􀆰 55 0􀆰 60
III 4-6 5 2 49 255􀆰 21 67􀆰 71 0􀆰 27 221􀆰 35 565􀆰 10 2􀆰 21 0􀆰 73 0􀆰 31
IV 6-8 7 2 36 187􀆰 50 88􀆰 54 0􀆰 47 143􀆰 23 343􀆰 75 1􀆰 83 0􀆰 53 0􀆰 64
V 8-10 9 2 19 98􀆰 96 31􀆰 25 0􀆰 32 83􀆰 33 200􀆰 52 2􀆰 03 0􀆰 68 0􀆰 38
VI 10-12 11 2 13 67􀆰 71 15􀆰 63 0􀆰 23 59􀆰 90 117􀆰 19 1􀆰 73 0􀆰 77 0􀆰 26
VII 12-14 13 2 10 52􀆰 08 20􀆰 83 0􀆰 40 41􀆰 67 57􀆰 29 1􀆰 10 0􀆰 60 0􀆰 51
VIII 14-16 15 2 6 31􀆰 25 31􀆰 25 1􀆰 00 15􀆰 63 15􀆰 63 0􀆰 50 0􀆰 00 1􀆰 00
Dꎬ diameter class (basal diameter)ꎻ x: midpoint age (midpoint diameterꎬ cm)ꎻ Δxꎬ width of age (here as width of diameterꎬ cm. According to the
biological characteristics of the C􀆰 rhytidophylla populationꎬ 2 cm=one age)ꎻ axꎬ number of surviving individualsꎻ lxꎬ proportion of individuals survi ̄
ving to age x ( lx =ax / a0×1000)ꎻ dxꎬ number of dead individuals from age x to x+1 (dx =ax-ax +1)ꎻ qxꎬ mortality rate from age x to x+1 (qx =dx /
lx)ꎻ Lxꎬ mean number of individuals surviving from age x to x+1 (Lx = ( lx+lx +1) / 2ꎻ Txꎬ total number of individuals surviving from age x (Tx =∑
Lx)ꎻ exꎬ life expectancy at age x (ex =Tx / lx)ꎻ Sxꎬ age ̄specific survival (Sx= lx +1 / lx)ꎻ Kxꎬ age ̄specific mortality (Kx =ln [ lx]-ln [ lx +1]).
248                                  植 物 分 类 与 资 源 学 报                            第 37卷
(Fig􀆰 5). There were remarkable differences in the
distribution patterns of C􀆰 rhytidophylla according to
scale at different stages of development in plot A.
Young shrubs were significantly aggregated at all scales
(Fig􀆰 5A ̄1)ꎻ Middle ̄aged shrubs (Fig􀆰 5A ̄2) were
significantly aggregated from 0 to 7 m and had a ran ̄
dom distribution at larger scales (7-25 m)ꎻ Adult
shrubs (Fig􀆰 5A ̄3) were significantly aggregated in
the ranges of 0-3 m and 5-7 m and showed a ran ̄
dom distribution in the ranges of 3-5 m and 7-25 m.
Significant differences in distribution patterns of
C􀆰 rhytidophylla according to scale were also observed
in plot B for different developmental stages. Young
shrubs (Fig􀆰 5B ̄1) were significantly aggregated at
all scalesꎻ Middle ̄aged shrubs (Fig􀆰 5B ̄2) were sig ̄
nificantly aggregated within the scale ranges of 0-8
m and 19 - 25 m and were randomly distributed in
the 8-19 m rangeꎻ adult shrubs (Fig􀆰 5B ̄3) showed
a random distribution in the 1-4 m range and an ag ̄
gregation distribution at scales >4 m.
Fig􀆰 3  Survival curve lx (A)ꎻ mortality rate curve qx and disappearance rate curve Kx (B) of the Camellia rhytidophylla population
Fig􀆰 4  Point pattern of individual Camellia rhytidophylla plants of different growth stages
Panels A ̄1ꎬ A ̄2ꎬ and A ̄3 represent youngꎬ middle ̄agedꎬ and adult shrubs in plot Aꎬ respectivelyꎻ panels B ̄1ꎬ B ̄2ꎬ and B ̄3
represent youngꎬ middle ̄agedꎬ and adult shrubs in plot Bꎬ respectively. Axes represent plot dimensions (50 m × 50 m)
3486期      LIU Hai ̄yan et al.: Population Ecological Characteristics of the Rare and Endangered Plant Camellia 􀆺     
Fig􀆰 5  Point pattern analysis of different growth stages of Camellia rhytidophylla. Solid lines denote the L( t) (Functional value at scale t)
curves calculated from data and dotted lines denote the fitted envelop curves (99% confidence interval) . Panels A ̄1ꎬ A ̄2ꎬ and A ̄3 repre ̄
sent youngꎬ middle ̄agedꎬ and adult shrubs in plot Aꎬ respectivelyꎻ panels B ̄1ꎬ B ̄2ꎬ and B ̄3 represent youngꎬ middle ̄agedꎬ and adult
shrubs in plot Bꎬ respectively. We used the interval of t at 1 m for the analysisꎬ and the maximum value of t was equal to half of the length
of a sample plot (25 m). Distance (m) is shown on the x ̄axis
2􀆰 6  Spatial correlation analysis at different de ̄
velopment stages
We analyzed spatial correlations between the three
developmental stages of C􀆰 rhytidophylla using Ri ̄
pley’s K ̄function (Fig􀆰 6). In plot Aꎬ at the 0-9 m
scaleꎬ measured L12( t) values overlapped with the
lower envelop curveꎬ and at the 9-25 m scaleꎬ mea ̄
sured L12(t) values were between the upper and lower
envelop curves (Fig􀆰 6A ̄1). This showed a negative
or near ̄negative correlation between young and mid ̄
dle ̄aged shrubs from 0 to 9 mꎬ and no correlation at
larger scales. Young and adult shrubs were negative ̄
ly or near ̄negatively correlated at the 4-13 m scale
and had a significant negative correlation at smaller
and larger scales (Fig􀆰 6A ̄2). There was no obvious
correlation between middle ̄aged and adult shrubs at
any scale (Fig􀆰 6A ̄3). In plot Bꎬ young and mid ̄
dle ̄aged shrubs had a negative or near ̄negative cor ̄
relation at scales >7 m and had no correlation at oth ̄
er scales. Young and adult shrubs had a negative
correlation or near ̄negative correlation at all scalesꎬ
while middle ̄aged and adult shrubs showed no cor ̄
relation at any scale (Fig􀆰 6B).
3  Discussion
The results of survival analysis showed that the
survival curve of the C􀆰 rhytidophylla population was
a typical Deevey ̄III type and its population was ex ̄
panding. Neverthelessꎬ approximately half of the
young plants observed in our survey were seedlingsꎬ
448                                  植 物 分 类 与 资 源 学 报                            第 37卷
and the others sprouted from stumps after middle ̄
aged and adult shrubs were cut down or died natural ̄
ly. This indicates that its sexual reproduction was not
strong and that the expanding population is vulnera ̄
ble. The peak death rates occurred in the first and
fourth age classesꎬ which might be attributed to in ̄
traspecific competition and artificial disturbance
(e􀆰 g.ꎬ logging.) . C􀆰 rhytidophylla seedlings mostly
grow in clusters of four to six plants. Intraspecific
competition becomes more severe as a result of den ̄
sity ̄dependent constraints on the growth of young
shrubs ( Li et al.ꎬ 2013)ꎻ ultimatelyꎬ only one or
two of the seedlings in the cluster can surviveꎬ lead ̄
ing to relatively high mortality rates in young shrubs.
In additionꎬ local villagers continue to harvest trees
and shrubs ( especially those with basal diameter of
5-8 cm) for sellingꎬ which was a major reason for
the relatively high mortality and disappearance rates
in the fourth age class. Artificial disturbance and
competition decrease in older age classesꎬ resulting
in decreased mortality and disappearance rates. In ̄
creased mortality and disappearance after the sixth
age stage could be a result of limited nutrient supply
and predatory logging. Thereforeꎬ if effective protec ̄
tion measures are not taken in a timely fashionꎬ the
expanding population may stabilize or begin to con ̄
tract as young trees grow and intraspecific competi ̄
tion intensifies.
Fig􀆰 6  Point pattern analysis of different growth stage associations of Camellia rhytidophylla. Solid lines denote the L12( t) curves calculated
from data and dotted lines denote the fitted envelop curves (99% confidence interval) . Panels A ̄1ꎬ A ̄2ꎬ and A ̄3 show the correlation be ̄
tween young and middle ̄aged shrubsꎬ young and adult shrubsꎬ and middle ̄aged and adult shrubs in plot Aꎬ respectivelyꎻ panels B ̄1ꎬ B ̄
2ꎬ and B ̄3 show the correlation between young and middle ̄aged shrubsꎬ young and adult shrubsꎬ and middle ̄aged and adult shrubs in plot
Bꎬ respectively. We used the interval of t at 1 m for the analysisꎬ and the maximum value of t was equal to half of the length of a sample plot
(25 m). Distance (m) is shown on the x ̄axis
5486期      LIU Hai ̄yan et al.: Population Ecological Characteristics of the Rare and Endangered Plant Camellia 􀆺     
  Spatial distribution patterns of plant populations
are a resultꎬ in partꎬ of biological factors and limits
to seed dispersalꎬ habitat heterogeneityꎬ intraspecif ̄
ic and interspecific competitionꎬ and non ̄biological
factors (He and Duncanꎬ 2000ꎻ Li et al.ꎬ 2013ꎻ Lin
et al.ꎬ 2011ꎻ Liu et al.ꎬ 2011aꎻ Queenborough et
al.ꎬ 2007ꎻ Yuan et al.ꎬ 2012ꎻ Zhang et al.ꎬ 2007b).
The spatial distribution patterns of young and mid ̄
dle ̄aged C􀆰 rhytidophylla were similar in the two
sampling plotsꎬ with young shrubs aggregated at all
spatial scalesꎬ and middle ̄aged shrubs aggregated at
small scales and randomly distributed at larger (>8 m)
scales. These patterns could be related to biological
characteristics such as seed dispersal mechanismsꎬ
individual multiplicationꎬ and intraspecific competi ̄
tion (Wang et al.ꎬ 2010). We observed that C􀆰 rhy ̄
tidophylla fruit usually contained 4-6 seedsꎬ leading
to one or two seedlings in the clusterꎬ which could
explain the high level of aggregation (seedling clumps)
in young shrubs. Additionallyꎬ many young shrubs
sprouted from the stumps of middle ̄age and adult
plants after logging and often grew in clusters of
three to five plants. This is another important reason
for the relatively high level of aggregation in young
shrubs. Growth of young individuals and increased
intraspecific competition lead to high mortality (den ̄
sity ̄induced self ̄thinning)ꎬ resulting in less aggre ̄
gation at middle ̄aged stages and leading to random
distributions at some scales (Haaseꎬ 1995). Opposite
distribution patterns were observed in plots A and B
for adult C􀆰 rhytidophylla ( the shrubs were aggrega ̄
ted at small scales and randomly distributed at larger
scales in plot Aꎬ and the reverse in plot B). This
might have been a result of habitat heterogeneity and
artificial disturbance. Despite being located in the
same plant communityꎬ plots A and B had different
density and canopy coverage (Table 1). This would
result in small ̄scale heterogeneity of factors inclu ̄
ding soil water content and light conditions. In addi ̄
tionꎬ large ̄scale logging of middle ̄aged and adult
shrubs by local villagers was another important rea ̄
son for the differences in spatial distribution patterns
between the two plots.
Intraspecific correlations provide a static descri ̄
ption of relationships among individual plants in a
population for a given time period. This includes
spatial distribution and functional relationships with ̄
in populations (Wang et al.ꎬ 2010). In this studyꎬ
we observed negative or no spatial correlations in the
C􀆰 rhytidophylla population at all developmental sta ̄
gesꎬ which demonstrated that this species has differ ̄
ent spatial patterns at different growth stages. This
phenomenon can be attributed to natural thinning
processesꎬ disturbance patternsꎬ and habitat hetero ̄
geneity. For exampleꎬ soil water and nutrient condi ̄
tions affect the spatial distribution of woody plants
(John et al.ꎬ 2007ꎻ Zhang and Mengꎬ 2004). Neg ̄
ative correlationsꎬ or the absence of correlationsꎬ
between young C􀆰 rhytidophylla and middle ̄aged or
adult shrubs could be explained by various factors.
C􀆰 rhytidophylla populations have been affected by
loggingꎬ and middle ̄aged shrubs were affected most
severely. In additionꎬ plant propagules could be tran ̄
sported large distances from mother shrubs on the
steep hillsides where this species occurs.
The Guizhou Plateau Subregion (IIID10d) is a
typical mountain and river valley in eastern Asiaꎬ
characterized by widely distributed carbonate rocks
with dissected topography (Wu et al.ꎬ 2010)ꎬ high
endemic plant resourcesꎬ and distinct transitional
tropical ̄to ̄subtropical characteristics. Camellia spp.
are important components of subtropical evergreen
and deciduous broad ̄leaved forestsꎬ and are espe ̄
cially common in the subcanopy and the shrub layer
(Ying and Chenꎬ 2011). Differences in plant sur ̄
vival among valleys and watersheds are a result of
microclimate characteristics determined by topogra ̄
phyꎻ such characteristics are present in the C􀆰 rhyti ̄
dophylla habitat (Fig􀆰 1). Hills surrounding the riv ̄
er valley create a microclimate for C􀆰 rhytidophylla
and lead to the evolution of its distinct biological
characteristics. The hill environment and watershed
limit the spread of C􀆰 rhytidophylla seedsꎬ which re ̄
sults in a narrowꎬ sparseꎬ and uneven population dis ̄
648                                  植 物 分 类 与 资 源 学 报                            第 37卷
tribution. In additionꎬ low rates of seed set and high
incidence of pests and diseases (Zouꎬ 2001) limit
the expansion of the C􀆰 rhytidophylla population.
In conclusionꎬ a combination of geographic and
biological characteristics has resulted in the narrow
distribution of this endemic speciesꎬ and its popula ̄
tions are currently endangered as a result of anthro ̄
pogenic activity. Protection of rare and endangered
plants is an important priority of biodiversity preser ̄
vation. We found that the population structure of
C􀆰 rhytidophylla was relatively stableꎻ if protectedꎬ
this species could survive and proliferate. The key is
to effectively protect the native forest ecosystem in
which C􀆰 rhytidophylla is found. In additionꎬ artifi ̄
cial propagation and reintroduction can help to ex ̄
pand C􀆰 rhytidophylla populationsꎬ and horticultural
use of this species can provide an effective means of
conserving its germplasm.
The key scientific problem of conservation bio ̄
logy has always been the conservation and utilization
of the biodiversity of rare and endangered plants with
small populations. This paper showed that the distri ̄
bution of C􀆰 rhytidophylla was very narrow. Habitat
heterogeneity and artificial severe disturbance caused
some limitations for the seed dispersalꎬ population
self propagationꎬ diffusion and their stability. So we
suggest that: 1) C􀆰 rhytidophylla should be included
in the rare and endangered species protection list
Redbookꎬ so that more people know it and enhance
people’s protection awareness. We also should es ̄
tablish nature reserves for the species and protect the
ecological environmentꎬ population and its biodiver ̄
sity. 2) On the basis of in situ conservationꎬ we sug ̄
gest that more research about population characteris ̄
tics and the propagation rules should be done to
solve the propagation problemꎬ to expand the popu ̄
lation and its distribution area by cultivating seedling
reintroductionꎬ helping population growth and en ̄
hanced its reproduction and dispersal ability. At the
same timeꎬ we could plant in different places and do
production and application testꎬ promote the trans ̄
formation of the characteristic resources advantage
into economic advantageꎬ and strengthen the protec ̄
tion of germplasm resources in the production and
application. All of these have important significance
and practical value for the development of conserva ̄
tion biology and the construction of regional ecologi ̄
cal environment.
Acknowledge: The authors would like to thank Mr Li Jinhuai
and Zhai Shuaishuai for their assistance with field work and
post ̄processing.
References:
Agricultural regional planning committee of Kaiyang countyꎬ 1989.
Comprehensive Agricultural Regionalization of Kaiyang County [M].
Guiyang: Guizhou People′s Publishing House
Cetis Aꎬ Franklin Jꎬ 1987. Second ̄order neighborhood analysis of
mapped point patterns [J] . Ecologyꎬ 68: 473—477
Condit Rꎬ Ashton PSꎬ Baker Pꎬ 2000. Spatial patterns in the distri ̄
bution of tropical tree species [J] . Scienceꎬ 288: 1414—1418
DCP ( Scientific Database of China Plant Species)ꎬ 2014. http: / /
www. plants. csdb. cn / eflora / view / search / chs _ contents. aspx?
CPNI=CPNI ̄003 ̄10086 [OL]
Diglle Pꎬ 1983. Statistical Analysis of Spatial Point Patterns [M].
New York: Academic Press
Feng Lꎬ Hong Wꎬ Wu CZꎬ 2003. Study on the dynamics of the en ̄
dangered plant population of Tsuga tchekiangensis [ J] . Journal
of Wuhan Botanical Researchꎬ 21: 401—405
Haase Pꎬ 1995. Spatial pattern analysis in ecology based on Ripley’s
K ̄function: Introduction and methods of edge correction [ J] .
Journal of Vegetation Scienceꎬ 6: 575—582
He FLꎬ Duncan RPꎬ 2000. Density ̄dependent effects on tree survival
in an old ̄growth Douglas fir forest [J] . Journal of Ecologyꎬ 88:
676—688
Jiang Hꎬ 1992. Ecology of Picea asperata Population [M]. Beijing:
China Forestry Publishing Houseꎬ 8—32
John Rꎬ Dalling JWꎬ Harms KEꎬ 2007. Soil nutrients influence spa ̄
tial distributions of tropical tree species [ J] . Proceedings of the
National Academy of Sciences of the United States of Americaꎬ
104: 864—869
Li MHꎬ He FHꎬ Liu Yꎬ 2005a. Spatial distribution pattern of tree in ̄
dividuals in the Schrenk spruce forest northwestꎬ China [J] . Ac ̄
ta Ecologica Sinicaꎬ 25: 1000—1006
Li YLꎬ Zhang Jꎬ Pan YZꎬ 2005b. The germplasm resources of Camel ̄
lia plants and their utilization in landscape gardens in Sichuan
Provinceꎬ China [J] . Southwest Horticultureꎬ 33: 26—27
Li SFꎬ Liu WDꎬ Su JRꎬ 2013. Age structure and spatial distribution
patterns of Taxus yunnanensis population in Lanping Countyꎬ
Yunnan Province [ J] . Acta Botanica Boreali ̄Occidentalia Sini ̄
7486期      LIU Hai ̄yan et al.: Population Ecological Characteristics of the Rare and Endangered Plant Camellia 􀆺     
caꎬ 33: 0792—0799
Li YKꎬ Yang MZꎬ 1987. A news species of Camellia from Guizhou
[J] . Guihaiaꎬ 7: 13—14
Liang Fꎬ Zhang Wꎬ 2014. The preliminary study on landslide stability
in Qingjiang Village Kaiyang county of Guizhou province Flower
plow [J] . Shanxi Architectureꎬ 40 (4): 81—83
Lin YCꎬ Chang LWꎬ Yang KCꎬ 2011. Point patterns of tree distribu ̄
tion determined by habitat heterogeneity and dispersal limitation
[J] . Oecologiaꎬ 165: 175—184
Liu GFꎬ Ding Yꎬ Zang RGꎬ 2011a. Distribution patterns of Picea
schrenkiana var. tianschanica population in Tianshan Mountains
[J] . Chinese Journal of Applied Ecologyꎬ 22: 9—13
Liu HYꎬ Zou TCꎬ Zhou HY et al.ꎬ 2011b. Studies on cultivation and
propagation of twenty species native ornamental plants in Guizhou
[J] . Guizhou Agricultural Sciencesꎬ 39: 29—33
Long XQꎬ 2013. Exploitation and utilization of the woody edible oil re ̄
sources in Guizhou [J] . Resources Development & Marketꎬ 19:
243—247
Queenborough SAꎬ Burslem DFRPꎬ Garwood NCꎬ 2007. Habitat niche
partitioning by 16 species of Myristicaceae in Amazonian Ecuador
[J] . Plant Ecologyꎬ 192: 193—207
Wang Lꎬ Sun QWꎬ Hao CYꎬ 2010. Point pattern analysis of different
age ̄class Taxus chinensis var. mairei individuals in mountainous
area of southern Anhui Province [ J] . Chinese Journal of Ap ̄
plied Ecologyꎬ 21: 272—278
Wu ZYꎬ Sun Hꎬ Zhou ZKꎬ 2010. Floristics of Seed Plants from China
[M]. Beijing: Science Press
Xu XHꎬ Yu MDꎬ Hu ZHꎬ 2005. The structure and dynamics of Cas ̄
tanopsis eyrei populations in Gutian Mountain Natural Reserve in
Zhejiangꎬ East China [ J] . Acta Ecologica Sinicaꎬ 25: 645—
653
Yan GQꎬ Zhao GFꎬ Hu ZHꎬ 2001. Population structure and dynamics
of Larix chinenesis in Qinling Mountain [ J] . Chinese Journal of
Applied Ecologyꎬ 12: 824—828
Ying JSꎬ Chen MLꎬ 2011. Plant Geography of China [M]. Beijing:
Science Press
Yuan CMꎬ Meng GTꎬ Fang XJꎬ 2012. Age structure and spatial distri ̄
bution of the rare and endangered plant Alcimandra cathcartii
[J] . Acta Ecologica Sinicaꎬ 32: 3866—3872
Zhang JTꎬ 1998. Analysis of spatial point pattern for plant species
[J] . Acta Phytoecologica Sinicaꎬ 22: 344—349
Zhang HD (张宏达)ꎬ Ren SX (任善湘)ꎬ 1998. Theaceae [A] / /
Flora Reipublicae Popularis Sinicaeꎬ 49 (3) [M]. Beijing: Sci ̄
ence Press
Zhang JTꎬ Meng DPꎬ 2004. Spatial pattern analysis of individuals in
different age ̄classes of Larix principisrupprechtii in Luya mountain
reserveꎬ Shanxiꎬ China [J] . Acta Ecologica Sinicaꎬ 24: 35—40
Zhang QYꎬ Zhang YCꎬ Luo Pꎬ 2007a. Ecological characteristics of a
Sabina saltuaria population at timberline on the south ̄facing
slope of Baima Snow Mountainꎬ Southwest China [J] . Journal of
Plant Ecology (Chinese Version)ꎬ 31: 857—864
Zhang Jꎬ Hao ZQꎬ Song Bꎬ 2007b. Spatial distribution patterns and
associations of Pinus koraiensis and Tilia amurensis in broad ̄
leaved Korean pine mixed forest in Changbai Mountains [ J] .
Chinese Journal of Applied Ecologyꎬ 18: 1681—1687
Zhang Tꎬ Liu HYꎬ Zou TCꎬ 2010. Main chemical components in leav ̄
es of eight wild Camellia species in Guizhou [J] . Guizhou Agri ̄
cultural Sciencesꎬ 38: 78—80
Zhang XPꎬ Guo CYꎬ Zhang Qꎬ 2013. Point pattern analysis of Ptero ̄
celtis tatarinowii population at its different development stages in
limestone mountain area of north Anhuiꎬ East China [ J] . Chi ̄
nese Journal of Ecologyꎬ 32: 542—550
Zou TCꎬ Lou YLꎬ 1995. Studies on phytotomic characteristics in leav ̄
es of three wild species from Camellia [J] . Guizhou Scienceꎬ 13:
11—17
Zou TCꎬ 2001. Guizhou Endemic and Rare Spermatophyta [M]. Guiy ̄
ang: Guizhou Science and Technology Publishing Houseꎬ 228—
235
848                                  植 物 分 类 与 资 源 学 报                            第 37卷