免费文献传递   相关文献

Multivariate analysis, description, and ecological interpretation of weeds and alien invasive weeds in the autumn tea gardens of Jiangsu Province, China

江苏茶园秋季杂草群落及外来入侵杂草的数量生态学分析



全 文 :植物保护学报 Journal of Plant Protectionꎬ 2015ꎬ 42(5): 848 - 858 DOI: 10􀆰 13802 / j. cnki. zwbhxb. 2015􀆰 05􀆰 022
基金项目:国家公益性行业(农业)科研专项(200903004ꎬ200903004 ̄43)ꎬ江苏省农作物有害生物种类与发生危害特点研究项目
∗通讯作者(Author for correspondence)ꎬ E ̄mail: jtszbz@ 163. comꎬ Tel: 0519 ̄80189051
收稿日期: 2014 - 11 - 18
Multivariate analysisꎬ descriptionꎬ and ecological
interpretation of weeds and alien invasive weeds in
the autumn tea gardens of Jiangsu Provinceꎬ China
Zhang Haiyan1   Sun Guojun1ꎬ2∗   Li Fenhua1   Han Min1   Yuan Fang1   Zhang Chunyun3
(1. Plant Protection and Quarantine Station in Jintan District of Changzhouꎬ Jintan 213200ꎬ Jiangsu Provinceꎬ Chinaꎻ
2. School of Horticulture and Plant Protectionꎬ Yangzhou Universityꎬ Yangzhou 225009ꎬ Jiangsu Provinceꎬ Chinaꎻ
3. Yizheng Plant Protection and Quarantine Stationꎬ Yizheng 211400ꎬ Jiangsu Provinceꎬ China)
Abstract: To investigate the effects of herbicide useꎬ tillageꎬ tree ageꎬ slopeꎬ and canopy density on the
weed species composition and community structure in autumn tea gardens in Jiangsu Provinceꎬ 90
representative sampling sites of Yizhengꎬ Jintanꎬ and Yixing were selectedꎬ and weed surveys were
conducted in October 2013. The 90 sampling sites were classified into three groups based on the
dominant weed species and weed abundance according to canonical correspondence analysis (CCA) and
principal component analysis ( PCA ). The Group Ⅰ consisted of Arthraxon hispidusꎬ Geranium
carolinianumꎬ Oxalis corniculataꎬ Setaria viridisꎬ Rubus hirsutusꎬ and Ophitopogon japonicumꎬ and these
species were present in all sample sites from the tea gardens with lower frequencies of tillage and
herbicide applicationꎬ older treesꎬ and higher canopy densities. This community had more perennial
weeds and weed species and a lower weed density. The Group Ⅱ consisted of Digitaria sanguinalisꎬ
Echinochloa crus ̄galli var. austro ̄japonensisꎬ Conyza canadensisꎬ G. carolinianumꎬ and S. viridis. All
the sample sites in this group relied primarily on chemical control of weedsꎬ and therefore more annual
grassy malignant weedsꎬ herbicide ̄tolerant weedsꎬ and prominent dominant weed species were present.
The Group Ⅲ contained D. sanguinalisꎬ E. crus ̄galliꎬ Eleusine indicaꎬ G. carolinianumꎬ O.
corniculataꎬ and Leptochloa panacea. Because the tea gardens of this group were intensively managedꎬ
fewer dominant weed species were presentꎬ and weed density was lower. Moreoverꎬ analyses of alien
invasive weed species and environmental factors indicated that herbicide application and tillage highly
significantly affected the occurrence and distribution of alien invasive weeds in autumn tea gardens in
Jiangsu Province. The results indicated that different management practicesꎬ e. g. ꎬ herbicide application
and tillage practiceꎬ play decisive roles in the formation of weed communities.
Key words: tea gardenꎻ weed communityꎻ alien invasive weedꎻ environment factorꎻ multivariate analysis
江苏茶园秋季杂草群落及外来入侵杂草的数量生态学分析
张海艳1   孙国俊1ꎬ2∗   李粉华1   韩  敏1   袁  方1   张春云3
(1. 常州市金坛区植保植检站ꎬ 江苏 金坛 213200ꎻ 2.扬州大学园艺与植物保护学院ꎬ 江苏 扬州 225009ꎻ
3.仪征市植保植检站ꎬ 江苏 仪征 211400)
摘要: 为明确除草剂使用、翻耕、树龄、坡度和郁闭度等杂草管理措施及环境因子对江苏茶园秋季
杂草群落和外来入侵杂草物种组成、群落结构的影响ꎬ于 2013 年 10 月分别对江苏茶叶主产区仪
征、金坛和宜兴 3 地共 90 个样地进行了杂草调查ꎮ 通过主成分分析和典范对应分析ꎬ90 块样地可
以依据优势杂草种类和杂草丰富度被分成 3 个类群ꎮ 类群Ⅰ为荩草、野老鹳草、酢浆草、狗尾草、蓬
蘽、麦冬杂草群落ꎬ该类群翻耕和化学除草频次少ꎬ树龄较大ꎬ郁闭度高ꎬ田间多年生杂草多ꎬ杂草种
类多ꎬ田间杂草密度低ꎻ类群Ⅱ为马唐、小旱稗、小飞蓬、野老鹳草、狗尾草杂草群落ꎬ该类群以化学
除草为主ꎬ田间一年生禾本科恶性杂草和抗除草剂杂草多ꎬ杂草优势种突出ꎻ类群Ⅲ为马唐、小旱
稗、牛筋草、野老鹳草、酢浆草、虮子草杂草群落ꎬ该类群茶园管理较为精细ꎬ田间优势杂草种类少ꎬ
杂草密度低ꎮ 对外来入侵杂草与环境因子关系的数量分析发现ꎬ化学除草剂的使用和翻耕频率显
著影响外来入侵杂草在江苏茶园的发生和分布ꎮ 研究表明ꎬ不同管理措施ꎬ如化学除草剂和翻耕频
率ꎬ对杂草群落的形成起决定性作用ꎮ
关键词: 茶园ꎻ 杂草群落ꎻ 外来入侵杂草ꎻ 环境因子ꎻ 多变量分析
    Weedsꎬ as a component of biodiversity in tea Ca ̄
mellia sinensis L. ecosystemsꎬ compete with tea plants
for resources such as waterꎬ nutrientꎬ and lightꎬ reduc ̄
ing tea yield and influencing tea quality significantly
(Xu et al. ꎬ2010). Weeds also change the structure of
the tea plantation ecosystem directly by affecting the
diversity and quantity of insects and microorganism
species in tea gardensꎬ reducing the yield and quality
of tea (Zhou et al. ꎬ2012). Thereforeꎬ weed control is
important for tea production. Because of the migration
of populations from the tea districts to the citiesꎬ the
increasing cost of labourꎬ and elimination of natural
predatorsꎬ herbicides has gone up indiscriminately to
achieve immediate control on weeds. Tea is a unique
agricultural product that is directly processed without
washing after harvesting and is often steeped in water
before consuming. It is thus necessary to reduce or a ̄
void the application of herbicides to insure the safety of
tea. Thereforeꎬ it is urgently needed to develop effec ̄
tive and safe methods (Jaggi et al. ꎬ2001ꎻSood et al. ꎬ
2004). The studies on the composition and structureꎬ
and the influencing factors of weed communities would
be beneficial for the development of weed management
strategy. In previous reportsꎬ multivariate analyses of
weeds in crops have demonstrated that agronomic man ̄
agement practices and agro ̄ecological environmental
factors significantly influence the distribution of weeds
in croplands (Qiang & Liꎬ1990ꎻQiang & Huꎬ1999ꎻ
Qiangꎬ2005). These discoveries furnished insights to
support prevention and control measures.
Alien plant invasions aggravate the severity and
complexity of weed infestation. Currentlyꎬ most studies
of alien invasive plants focus primarily on non ̄agricul ̄
tural habitats. Howeverꎬ the agroecosystem is one of
the most vulnerable ecosystem types to exotic plant in ̄
vasion because of human activity interference and rich
resources. Qiang et al. ( 2010 ) reported that more
than 115 species of alien invasive plants exist in Chi ̄
naꎬ and these weed species exhibited different patterns
of occurrence and distribution in different types of
farmland. Howeverꎬ the lack of research on alien inva ̄
sive plants has strongly retarded efforts to develop
methods for effective and long ̄term prevention and con ̄
trol of alien invasive plants in tea gardens. As a peren ̄
nial crop that receives abundant labour resource input
and strongly artificial interferenceꎬ tea is easily infested
by pioneering alien invasive species such as Conyza
spp. and Veronica spp. (Xu & Qiangꎬ2011). Addi ̄
tionallyꎬ a relatively low frequency of weeding and are ̄
as that are unsuitable for ploughing cause certain pe ̄
rennial invasive plants to become a larger problem
(Qiangꎬ2001).
Although numerous quantitative studies on weeds
in croplands have been performedꎬ few studies on tea
gardens have been conducted (Guo & Liꎬ1998ꎻQiang
& Huꎬ1999ꎻQiang et al. ꎬ2003). Moreoverꎬ only a
few of these studies have examined the effects of factors
such as agronomic management practices and the eco ̄
logical environment on alien invasive weeds (Xiang et
al. ꎬ2009ꎻXu et al. ꎬ2010). Although the quantitative
ecology of the understory weed community and environ ̄
mental factors associated with tea gardens were studied
in Xiaoshegouꎬ Shanxi Provinceꎬ by Yu et al.
(2013)ꎬ few studies have addressed the effects of ag ̄
9485 期 Zhang Haiyanꎬ et al. : Multivariate analysis of weed comminities in the Jiangsu autumn tea gardens
ronomic management practices and the ecological envi ̄
ronment on weed community compositionꎬ diversityꎬ
and the presence of alien invasive weeds in tea gar ̄
dens. Thereforeꎬ surveys of weed speciesꎬ densityꎬ
and frequency were conducted in October 2013 in rep ̄
resentative sites of the three main tea ̄producing areas
of Jiangsu Province. The objectives of this study were
to quantitatively analyse the effects of different agro ̄
nomic management practices and the ecological envi ̄
ronments on the community composition and diversity
of weeds and alien invasive weeds which would provide
further insight into the scientific prevention and control
of autumn weeds in tea gardens of Jiangsu Province.
1 Materials and Methods
1􀆰 1 Materials
Jiangsu Provinceꎬ Chinaꎬ is situated between 30°
45′ and 35°07′Nꎬ and 116°22′ and 121°55′E. Tea
gardens occupy 32 500 hm2 of landꎬ primarily in the
hilly and mountainous areas of Nanjingꎬ Zhenjiang and
Yangzhou and in the sloping lands of the Lake Taihu
Basin in Jiangsu Province. The areas devoted to tea
growing are 8 467 hm2 in Nanjingꎬ 7 800 hm2 in Chan ̄
gzhouꎬ 5 733 hm2 in Wuxiꎬ 4 467 hm2 in Zhenjiangꎬ
2 400 hm2 in Suzhouꎬ 2 333 hm2 in Yangzhouꎬ 1 400
hm2 in Lianyungangꎬ and 333 hm2 in Huai’an. Nine ̄
ty ̄five percent of the tea ̄producing areas are distribu ̄
ted in hilly areas along the Yangtze River in southern
Jiangsu and northern Jiangsu. Three cities along the
Yangtze Riverꎬ including Yixing and Jintan in southern
Jiangsu and Yizheng in northern Jiangsuꎬ were selected
as the study sites.
1􀆰 2 Methods
1􀆰 2􀆰 1 Investigations of environmental factors
The research was conducted in October 2013 prior
to autumn weeding. Thirty tea gardens were selected as
sampling sites in each of the three selected representa ̄
tive sitesꎬ yielding a total of 90 sampling sites. For
each sampling siteꎬ tea tree ageꎬ inter ̄row canopy den ̄
sityꎬ slopeꎬ tillage frequencyꎬ and herbicide applica ̄
tion frequency were recorded as environmental factors.
The sampling sites were coded from 1 to 90 (Yixing:
1 - 30ꎻ Yizheng: 31 - 60ꎻ Jintan: 61 - 90).
1􀆰 2􀆰 2 Weed investigations
The number and species of weeds at each sam ̄
pling site were investigated in October 2013 in nine
0􀆰 25 m2quadrats that were placed according to the in ̄
verted ̄W 9 ̄point sampling method ( Thomasꎬ1985).
The heights of all weeds in each quadrant were also
measured.
The importance values ( IV) of all weed popula ̄
tions that occurred in tea gardens were calculated using
the following formula: IV = (RA% + RF% + RH%) / 3ꎬ
Where: relative abundance (RA) represents the rich ̄
ness of the weed species in the survey plots. The weed
species with the highest relative abundance value was
considered the primary weed species adapted to the local
ecosystem (Qiangꎬ2001ꎻWang et al. ꎬ2008). Relative
abundance (RA) was calculated using the following for ̄
mula: RA = (RD% + RF% + RU%) / 3. Relative fre ̄
quency ( RF)ꎬ relative density ( RD)ꎬ and relative
height (RH) of a weed species were calculated by ap ̄
plying the following formulas: RF = (Absolute frequency
value for a species / Total of absolute frequency value for
all species) ×100%ꎻ RD = (Absolute density for a giv ̄
en species / Total absolute density for all species) ×
100%ꎻ RH = (Mean height of a given species / Sum of
the mean height of all species) ×100% .
1􀆰 3 Data analyses
The IVs of all weed species with frequencies of
greater than 3􀆰 00% in all 90 sample sites were inte ̄
grated into a species data matrixꎬ and the environmen ̄
tal variables were integrated into an environmental data
matrix. The IVs of all alien weed species were integrat ̄
ed into another species data matrix. The matrixes were
analysed using CANOCO for Windows 4􀆰 5 ( ter Braak
& Smilauerꎬ2002)ꎬ a statistical and ecological soft ̄
ware packageꎬ to obtain two ̄dimensional scatterplots of
the PCA (principal component analysis) and the CCA
(canonical correspondence analysis) . The CCA crea ̄
ted an ordination that maximised the correlation be ̄
tween environmental factors and weed communities. A
Monte Carlo permutation test was used to test the null
hypothesis that the observed variations in weed commu ̄
nity composition were unrelated to variations in the en ̄
vironmental data.
058 植  物  保  护  学  报 42 卷
    SPSS 16􀆰 0 was used to analyse the correlations
(Pearson correlation coefficient) between the IVs ( or
frequency) of invasive weed species and the environ ̄
mental factors by the two ̄tailed test.
2 Results
2􀆰 1 Weed flora
One hundred and seven weed species were recor ̄
ded in the autumn tea gardens in Jiangsu Provinceꎬ
and these species represented 96 genera and 42 fami ̄
lies. Nineteen species in 18 genera were members of
the Compositaeꎬ which accounted for 17􀆰 76% and
18􀆰 75% of the total weed species and generaꎬ respec ̄
tively. Thirteen species in 13 generaꎬ representing
12􀆰 51% and 13􀆰 54% of the total species and generaꎬ
respectivelyꎬ were weeds in the family Gramineae.
Five species in five generaꎬ representing 4􀆰 67% and
5􀆰 21% of the total species and generaꎬ respectivelyꎬ
were members of the Labiatae. Five weed species in
four genera were found in the Amaranthaceae. Be ̄
sidesꎬ there were five weed species in three generaꎬ
respectivelyꎬ in each of the Euphorbiaceaeꎬ Polygo ̄
naceaeꎬ and Cyperaceae. The Liliaceae and Legumino ̄
sae were each represented by three species in three
genera. In the Solanaceaeꎬ three weed species in two
genera were found. Two weed species in two genera in
each of the Rubiaceaeꎬ Rosaceaeꎬ Umbelliferaeꎬ
Caryophyllaceaeꎬ Malvaceaeꎬ Scrophulariaceae and
Convolvulaceae were identified. The Violaceae and
Plantaginaceae were each represented by two species in
one genus. The other families were each represented
by one weed species in one genus.
2􀆰 2 Principal component analysis of weed communities
In totalꎬ 85 weed species with frequencies of grea ̄
ter than 3􀆰 00% in the autumn tea gardens of Jiangsu
Province were used as variables of principal component
analysis (PCA) (Table 1). Among these weed spe ̄
ciesꎬ the frequency of D. sanguinalis was the highest
of 95􀆰 56% in tea garden with the highest IV of
13􀆰 65% . The frequencies of weed species G. carolin ̄
ianumꎬ C. canadensisꎬ O. corniculataꎬ and E. crus ̄
galli var. austro ̄japonensis were more than 50% with
the higher IVs of 3􀆰 75% ꎬ 3􀆰 08% ꎬ 2􀆰 81% and
5􀆰 80% ꎬ respectively. Among theseꎬ G. carolinianumꎬ
C. canadensis and O. corniculata were alien species
which indicated that the invasion of alien species were
seriously.
A two ̄dimensional plot of the PCA ordination was
drawn (Fig. 1) based on the IVs of these 85 weed spe ̄
cies (Table 1).
Table 1 IVs and frequency of 85 weed species in the autumn tea gardens of Jiangsu Province %
No. Family Weed species IV Frequency No. Family Weed species IV Frequency
1 Cyperaceae Cyperus iria 1􀆰 31 28􀆰 89 44 Convolvulaceae Pharbitis purpurea∗ 0􀆰 48 3􀆰 33
2 Scrophulariaceae Lindernia crustacea 0􀆰 62 14􀆰 44 45 Labiatae Mentha arvensis 0􀆰 73 8􀆰 89
3 Amaranthaceae Amaranthus retroflexus∗ 1􀆰 04 22􀆰 22 46 Graminae Setaria viridis 2􀆰 60 37􀆰 78
4 Euphorbiaceae Acalypha australis∗ 1􀆰 05 27􀆰 78 47 Umbelliferae Daucus carota∗ 0􀆰 33 10􀆰 00
5 Asteraceae Gnaphalium affine 0􀆰 17 5􀆰 56 48 Pteridaceae Pieris multifida 0􀆰 45 5􀆰 56
6 Portulacaceae Portulaca oleracea 0􀆰 33 6􀆰 67 49 Liliaceae Ophitopogon japonicum 0􀆰 89 7􀆰 78
7 Geraniaceae Geranium carolinianum∗ 3􀆰 75 61􀆰 11 50 Solanaceae Solanum nigrum 1􀆰 00 26􀆰 67
8 Compositae Conyza canadensis∗ 3􀆰 08 62􀆰 22 51 Acanthaceae Rostellularia procumbens 1􀆰 47 28􀆰 89
9 Amaranthaceae Amaranthus ascendens 0􀆰 43 7􀆰 78 52 Polygonaceae Polygonum perfoliatum 1􀆰 18 3􀆰 33
10 Liliaceae Allium macrostemon 0􀆰 45 7􀆰 78 53 Geraniaceae Echinochloa crus ̄galli
var. austro ̄japonensis
5􀆰 80 54􀆰 44
11 Rosaceae Rubus hirsutus 1􀆰 09 17􀆰 78 54 Geraniaceae Cynodon dactylon 0􀆰 30 3􀆰 33
12 Commelinaceae Commelina bengalensis 1􀆰 09 25􀆰 56 55 Liliaceae Smilax glabra 0􀆰 95 5􀆰 56
13 Vitaceae Cayratia japonica 0􀆰 96 17􀆰 78 56 Leguminosae Glycine soja 0􀆰 79 4􀆰 44
14 Graminae Eleusine indica∗ 2􀆰 74 40􀆰 00 57 Solanaceae Solanum lyratum 0􀆰 82 8􀆰 89
15 Compositae Erigeron annuus∗ 0􀆰 93 28􀆰 89 58 Amaranthaceae Alternanthera philoxeroides∗ 0􀆰 77 8􀆰 89
16 Gramineae Digitaria sanguinalis 13􀆰 65 95􀆰 56 59 Rosaceae Duchesnea indica 0􀆰 30 3􀆰 33
17 Amaranthaceae Achyranthes bidentata 1􀆰 54 26􀆰 67 60 Sterculiaceae Melochia corchorifolia 0􀆰 81 16􀆰 67
18 Polygonaceae Polygonum multiflora 1􀆰 06 4􀆰 44 61 Graminae Leptochloa panicea 2􀆰 68 22􀆰 22
1585 期 Zhang Haiyanꎬ et al. : Multivariate analysis of weed comminities in the Jiangsu autumn tea gardens
Table 1 (continued)
No. Family Weed species IV Frequency No. Family Weed species IV Frequency
19 Polygonaceae Polygonum lapathifolium 0􀆰 52 12􀆰 22 62 Aizoaceae Mollugo stricta 0􀆰 52 12􀆰 22
20 Polygonaceae Polygonum longisetum 1􀆰 50 28􀆰 89 63 Graminae Panicum bisulcatum 0􀆰 55 4􀆰 44
21 Rubiaceae Galium aparine var. tenerum 0􀆰 21 8􀆰 89 64 Violaceae Viola japonica 0􀆰 16 3􀆰 33
22 Rubiaceae Paederia scandens var. tomentosa 1􀆰 37 22􀆰 22 65 Asteraceae Cephalanoplos segetum 0􀆰 41 8􀆰 89
23 Asclepiadaceae Metaplexis japonica 0􀆰 71 4􀆰 44 66 Plantaginaceae Plantago virginica∗ 1􀆰 48 10􀆰 00
24 Asteraceae Youngia japonica 0􀆰 43 12􀆰 22 67 Boraginaceae Trigonotis peduncularis 0􀆰 88 25􀆰 56
25 Oxalidaceae Oxalis corniculata∗ 2􀆰 81 55􀆰 56 68 Euphorbiaceae Euphorbia humifusa∗ 0􀆰 57 17􀆰 78
26 Caryophyllaceae Malachium aauaticum 1􀆰 02 20􀆰 00 69 Labiatae Lamium amplexicaule 0􀆰 19 4􀆰 44
27 Gramineae Arthraxon hispidus 1􀆰 46 14􀆰 44 70 Asteraceae Sonchus oleraceus∗ 0􀆰 41 17􀆰 78
28 Leguminosae Vicia sativa 1􀆰 08 31􀆰 11 71 Asteraceae Xanthium sibiricum 0􀆰 54 10􀆰 00
29 Dryopteridaceae Dryopteris chinensis 0􀆰 72 8􀆰 89 72 Solanaceae Physalis angulata∗ 0􀆰 32 5􀆰 56
30 Lygodiaceae Lygodium japonicum 0􀆰 88 7􀆰 78 73 Malvaceae Sida acuta 0􀆰 74 4􀆰 44
31 Euphorbiaceae Phyllanthus urinaria 1􀆰 10 11􀆰 11 74 Asteraceae Chrysanthemum indicum 0􀆰 72 3􀆰 33
32 Apocynaceae Trachelospermum jasminoides 0􀆰 68 8􀆰 89 75 Asteraceae Solidago canadensis∗ 0􀆰 52 2􀆰 22
33 Cucurbitaceae Melothria indica 0􀆰 75 3􀆰 33 76 Convolvulaceae Calystegia hederacea 0􀆰 58 7􀆰 78
34 Asteraceae Hemistepta lyrata 0􀆰 45 15􀆰 56 77 Cyperaceae Fimbristylis miliacea 0􀆰 29 5􀆰 56
35 Brassicaceae Capsella bursa ̄pastoris 0􀆰 17 4􀆰 44 78 Asteraceae Eclipta prostrata 0􀆰 75 13􀆰 33
36 Asteraceae Taraxacum mongolicum 0􀆰 41 13􀆰 33 79 Asteraceae Carpesium abrotanoides 0􀆰 56 3􀆰 33
37 Phytolaccaceae Phytolacca americana∗ 0􀆰 58 12􀆰 22 80 Compositae Conyza sumatrensis∗ 0􀆰 42 2􀆰 22
38 Euphorbiaceae Euphorbia helioscopia 0􀆰 27 7􀆰 78 81 Malvaceae Hibiscus trionum∗ 0􀆰 22 3􀆰 33
39 Caryophyllaceae Cerastium caespitosum∗ 1􀆰 48 25􀆰 56 82 Scrophulariaceae Veronica didyma∗ 0􀆰 17 3􀆰 33
40 Violaceae Viola philippica 0􀆰 16 6􀆰 67 83 Compositae Lapsana apogonoides 0􀆰 14 3􀆰 33
41 Asteraceae Emilia sonchifolia 1􀆰 04 15􀆰 56 84 Cyperaceae Cyperus rotundus∗ 0􀆰 42 3􀆰 33
42 Compositae Crassocephalum crepidioides∗ 0􀆰 54 7􀆰 78 85 Euphorbiaceae Euphorbia hirta∗ 0􀆰 38 7􀆰 78
43 Menispermaceae Cocculus orbiculatus 1􀆰 06 17􀆰 78
    The weed species with “∗” were alien invasive plant species.
    On the x ̄axis (Fig. 1)ꎬ P. scandens var. tomen ̄
tosa ( 0􀆰 696 )ꎬ R. hirsutus ( 0􀆰 600 )ꎬ C. japonica
(0􀆰 548)ꎬ and A. hispidus (0􀆰 531)ꎬ which are mostly
shade ̄tolerant or shade ̄preferring perennial and vege ̄
tative propagated weed speciesꎬ were located on the far
right. The species were primarily distributed in older
tea gardens that had older treesꎬ less soil tillageꎬ and
higher inter ̄row canopy density. On the left side of
the x ̄axisꎬ D. sanguinalis ( - 0􀆰 500)ꎬ E. crus ̄galli
( - 0􀆰 489)ꎬ Cyperus iria ( - 0􀆰 450)ꎬ and Eleusine
indica ( - 0􀆰 448) were found. These are mostly light ̄
requiring annual weed species of Gramineae that dis ̄
play large seed production and high resistance to till ̄
age. The species were the dominant weed species of
mature or young tea gardens with loose and fertile soil
and low inter ̄row canopy density.
On the y ̄axis ( Fig. 1)ꎬ L. panacea (0􀆰 427)ꎬ
E. crus ̄galli (0􀆰 420)ꎬ P. longisetum (0􀆰 403)ꎬ and
Achyranthes bidentata (0􀆰 401)ꎬ which were primarily
distributed in flat or gently sloped tea gardens with
Fig. 1 Two ̄dimensional scatter plot of principal
component analysis (PCA) ordination for
85 weed species in autumn tea gardens
of Jiangsu Province
The number corresponding to each weed species in this
figure is shown in Table 1.
 
258 植  物  保  护  学  报 42 卷
higher tillage frequency and lower herbicide applica ̄
tionꎬ were located in the upper region ( Fig. 1). In
contrastꎬ in the lower region of the y ̄axisꎬ E. humifusa
( -0􀆰 563)ꎬ P. virginica ( - 0􀆰 561)ꎬ L. apogonoides
( -0􀆰 518)ꎬ A. macrostemon ( - 0􀆰 499)ꎬ T. peduncu ̄
laris ( - 0􀆰 490)ꎬ and C. caespitosum ( - 0􀆰 469) were
found. These species were mainly distributed in deeply
sloped tea gardens where tilling was limited and weed
control mainly relied on herbicides.
Based on the two ̄dimensional PCA scatter plotꎬ
the 90 sample sites were artificially divided into three
groups ( Fig. 2). Group I included 28 sample sites:
1 - 5ꎬ 7ꎬ 9 - 11ꎬ 13ꎬ 14ꎬ 20ꎬ 25 - 30ꎬ 32ꎬ 35ꎬ 52ꎬ
55ꎬ 64ꎬ 77 - 80ꎬ and 87. Most of the sites in Group
Ⅰ were in Yixing ( 18 sites )ꎻ the others were in
Yizheng and Jintan ( sites 4 and 6ꎬ respectively ).
Sample sites of Group Ⅰ had lower frequency of tillage
and herbicide application and featured older trees and
a higher inter ̄row canopy density than the sites in
Groups Ⅱ and Ⅲ. Although various weed species (80
weed species in 44 genera and 37 families)ꎬ mostly
perennial (34􀆰 52% )ꎬ were presentꎬ the weed density
was relatively lowꎬ and the dominant species was not as
prominent as in Groups Ⅱ and Ⅲ. Most of the 80
weeds in Group Ⅰ were Compositaeꎬ including 17 spe ̄
cies in 17 genera. Of theseꎬ P. virginica was the most
abundant weed species ( IV = 4􀆰 41)ꎬ followed by A.
hispidusꎬ G. carolinianumꎬ and O. corniculata ( IV =
3􀆰 82ꎬ 3􀆰 67 and 3􀆰 49ꎬ respectively). The weed com ̄
munity represented by Group Ⅰ was named the P. vir ̄
ginicaꎬ A. hispidusꎬ G. carolinianumꎬ O. corniculataꎬ
C. caespitosum community.
Group Ⅱ consisted of 28 sample sites: 12ꎬ 17ꎬ
22ꎬ 31ꎬ 33ꎬ 41ꎬ 42ꎬ 44ꎬ 45ꎬ 47ꎬ 49ꎬ 51ꎬ 53ꎬ 54ꎬ
56ꎬ 61ꎬ 62ꎬ 66 - 68ꎬ 73ꎬ 82 - 86ꎬ 88 and 90. Most
of these sites were located in Yizheng and Jintan (12
and 13 sitesꎬ respectively). Tillage and herbicide ap ̄
plication were the primary weed control methods at
these sites. In these tea gardensꎬ 77 species in 70 gen ̄
era and 35 families were recorded. Most of the weeds
were Compositae (14 species in 14 genera) followed
by Gramineae (nine species in nine genera). Among
these weed speciesꎬ most are annual and biennialꎬ and
only a few are perennial. Overwhelmingly dominant
species with high plant densities were found in this
groupꎻ most of these were annual malignant grass
weeds with potential herbicide tolerance. D. sanguina ̄
lis was the dominant speciesꎬ with an IV of 23􀆰 75ꎻ
other abundant species were E. crus ̄galliꎬ L. panacea
and E. indica ( IV = 7􀆰 64ꎬ 4􀆰 99 and 4􀆰 31ꎬ respec ̄
tively). The weed community of Group Ⅱ was named
the D. sanguinalisꎬ E. crus ̄galliꎬ L. paniceaꎬ E. in ̄
dica community.
Group Ⅲ consisted of 34 sample sites: 6ꎬ 8ꎬ 15ꎬ
16ꎬ 18ꎬ 19ꎬ 21ꎬ 23ꎬ 24ꎬ 34ꎬ 36 - 40ꎬ 43ꎬ 46ꎬ 48ꎬ
50ꎬ 57 - 60ꎬ 63ꎬ 65ꎬ 69 - 72ꎬ 74 - 76ꎬ 81ꎬ and 89.
All the sites were finely managed tea gardens with rela ̄
tively flat terrainꎬ fertile soilꎬ and high frequency of
tillage. A total of 79 weed species in 71 genera and 38
families were found at these sample sites and more an ̄
nual (56􀆰 96% ) than biennial (11􀆰 39% ) weed spe ̄
cies were presented. Most of the weeds were Composi ̄
taeꎬ (14 species in 14 genera) followed by Gramineae
(nine species in nine genera). Among these weed spe ̄
ciesꎬ E. crus ̄galli was the dominant species ( IV =
9􀆰 61)ꎻ the other abundant species were D. sanguina ̄
lisꎬ G. carolinianum and C. canadensis ( IV = 7􀆰 71ꎬ
6􀆰 70 and 4􀆰 97ꎬ respectively). The weed community
represented by Group Ⅲ was named the E. crus ̄galliꎬ
D. sanguinalisꎬ G. carolinianumꎬ and C. canadensis
communityꎬ respectively.
2􀆰 3 Canonical correspondence analysis (CCA) be ̄
tween weed species and environmental factors
The species ̄environment correlations with the 1stꎬ
2ndꎬ 3rdꎬ and 4th axes were 0􀆰 749ꎬ 0􀆰 744ꎬ 0􀆰 742ꎬ
and 0􀆰 680ꎬ respectively. Significant negative correla ̄
tions were found between the 1st axis and tillage fre ̄
quency and herbicide applicationꎬ with inter ̄set correla ̄
tions of -0􀆰 582 and -0􀆰 603ꎬ respectively. Significant
positive correlations were found between the 1st axis and
tree age and inter ̄row canopy densityꎬ with inter ̄set
correlations of 0􀆰 298 and 0􀆰 528ꎬ respectively. The 2nd
axis had significant positive correlations with herbicide
application and slopeꎬ with inter ̄set correlations of
0􀆰 391 and 0􀆰 543ꎬ respectively (Table 2).
P. multiflorumꎬ D. indicaꎬ P. multifidaꎬ and L.
japonicum were positively correlated with tree ageꎬ
whereas the group had negative correlations with tillage
3585 期 Zhang Haiyanꎬ et al. : Multivariate analysis of weed comminities in the Jiangsu autumn tea gardens
      Table 2 Inter ̄set correlations of the environmental factors with the first four species axes
of the CCA of the weed survey data
Factor Axis 1 Axis 2 Axis 3 Axis 4
Tillage - 0􀆰 582∗∗ 0􀆰 061 - 0􀆰 150 - 0􀆰 177
Herbicide - 0􀆰 603∗∗ 0􀆰 391∗∗ 0􀆰 028 - 0􀆰 177
Tree age 0􀆰 298∗∗ - 0􀆰 128 - 0􀆰 593∗∗ - 0􀆰 207
Canopy density 0􀆰 528∗∗ 0􀆰 096 - 0􀆰 252∗ - 0􀆰 412∗∗
Slope 0􀆰 163 0􀆰 543∗∗ - 0􀆰 263∗ 0􀆰 354∗∗
Eigenvalues 0􀆰 201 0􀆰 105 0􀆰 102 0􀆰 072
Cumulative percentage variance of species data 3􀆰 700 5􀆰 600 7􀆰 500 8􀆰 800
Cumulative percentage variance of species ̄
environment correlations
38􀆰 300 58􀆰 300 77􀆰 800 91􀆰 500
Species ̄environment correlations 0􀆰 749 0􀆰 744 0􀆰 742 0􀆰 680
    ∗ꎬ ∗∗ in the same column indicate significant difference at P < 0􀆰 05 or P < 0􀆰 01 level by the Monte Carlo permutation testꎬ respectively.
Fig. 2 Two ̄dimensional scatter plot of PCA ordination
for the 90 sample sites in autumn tea
gardens of Jiangsu Province
1 - 30ꎬ 31 - 60ꎬ and 61 - 90 are the Yixingꎬ Yizheng
and Jintan sample sitesꎬ respectively.
 
frequency and herbicide application. A. macrostemonꎬ
G. affineꎬ C. segetumꎬ and C. iria were positively
correlated with tillage frequency but negatively correla ̄
ted with canopy density. M. corchorifoliaꎬ P. bisulca ̄
tumꎬ P. virginicaꎬ and E. humifusa were positively
correlated with herbicide application. S. glabraꎬ G.
sojaꎬ and X. sibiricum were positively correlated with
slope. P. scandens var. tomentosaꎬ V. japonicaꎬ and
S. lyratum were positively correlated with canopy den ̄
sity (Fig. 3).
Based on the site ̄environment CCA ordination
(Fig. 4)ꎬ the 90 sample sites were artificially divided
into three groups. The results showed that tillage and
herbicide application were the most important factors
Fig. 3 CCA of 85 weed species illustrating the
relationships between weed species and tillageꎬ
herbicideꎬ slopeꎬ canopy densityꎬ and
tree age vectors in the ordination
The numbers 1 - 85 correspond to the weed numbers
1 -85 in Table 1. The weeds shown in red are alien invasive
weeds.
 
that influenced the composition of the weed communi ̄
ties in the sample sites of Group Ⅱ. In the sites of
Group Ⅰꎬ the weed compositions were greatly affected
by tree age and canopy density. The weed communities
that had significant positive correlations with slope were
primarily distributed in the sample sites located in
steeper areas in Yixing and Jintan. In contrastꎬ the
weed communities that had significant negative correla ̄
tions with slope were primarily distributed in the sam ̄
ple sites of Group Ⅲ in Yizheng and Jintan with low
and smooth terrain.
458 植  物  保  护  学  报 42 卷
Fig. 4 CCA illustrating the relationships between sample
sites and tillageꎬ herbicideꎬ slopeꎬ canopy densityꎬ
and tree age vectors in the ordination
1 - 30ꎬ 31 - 60ꎬ and 61 - 90 are the Yixingꎬ Yizheng
and Jintan sample sitesꎬ respectively.
 
2􀆰 4 Quantitative analyses of the relationships between
alien invasive weeds and environmental factors
Twenty ̄three alien invasive (Table 1) weed spe ̄
cies in 19 genera and 14 families were recorded in the
autumn tea gardens of Jiangsu Province. The highest
number of species was in the Compositaeꎬ with six spe ̄
cies in five genera. Among these 23 speciesꎬ G.
             
carolinianum had the highest IV of 3􀆰 75 with a fre ̄
quency of 61􀆰 11% ꎬ followed by C. canadensis ( IV =
3􀆰 08) with a frequency of 62􀆰 22% ꎬ and O. cornicula ̄
ta ( IV = 2􀆰 81) with a frequency of 55􀆰 56% .
The correlation analyses showed that the IVs of
alien invasive weeds were significantly positively corre ̄
lated with herbicide application and tillage but signifi ̄
cantly negatively correlated with canopy density and tea
tree age. Besidesꎬ the frequency of occurrence of alien
invasive weeds had significantly positive correlation
with tillage and herbicide application (Table 3). A ̄
mong these 23 weed speciesꎬ C. sumatrensisꎬ P. vir ̄
ginicaꎬ S. canadensisꎬ E. helioscopiaꎬ E. hirtaꎬ and
E. humifusa showed significant positive correlations
with herbicide applicationꎬ suggesting that they had
some level of herbicide resistance or tolerance. P. a ̄
mericanaꎬ D. carotaꎬ and C. crepidioides were signifi ̄
cantly positively correlated with canopy density due to
the height advantage of these species compared with the
tea tree. A. philoxeroidesꎬ V. didymaꎬ and P. angu ̄
lataꎬ which are adapted to an environment of low lightꎬ
humidityꎬ and fertile soilꎬ were positively correlated
with tree age and negatively correlated with slopeꎬ con ̄
sistent with their distribution in low ̄lyingꎬ old tea gar ̄
dens.
Table 3 Correlation analyses among five environmental factorsꎬ total important valuesꎬ
and frequencies of 23 alien invasive weed species
Variable Tillage Herbicide Tree age Canopy Slope
IV 0􀆰 233∗ 0􀆰 377∗∗ - 0􀆰 242∗ - 0􀆰 283∗∗ 0􀆰 050
Frequency 0􀆰 242∗ 0􀆰 365∗∗ - 0􀆰 206 - 0􀆰 186 - 0􀆰 090
    ∗ꎬ∗∗ in the table indicate significant correlation at P < 0􀆰 05 or P < 0􀆰 01 level by the two ̄tailed testꎬ respectively.
3 Discussion
Species diversity reflects differences in biotic com ̄
munity structure and typeꎬ organisation levelꎬ develop ̄
ment stageꎬ degree of stabilityꎬ and variety of habitatꎬ
which provide the most basic data and information for
the development and utilisation of plant resources
(Yang & Sun 2013ꎻZhang et al. ꎬ2013). At smaller
scales of the landscape and communityꎬ zonal environ ̄
mental factors such as the presence of specific microor ̄
ganismsꎬ soil typeꎬ and nutrient content determine the
distribution of vegetation ( Siefert et al. ꎬ 2012 ).
Yizhengꎬ Jintan and Yixing are not geographically dis ̄
tant from each other. The geographicꎬ environmentalꎬ
and natural climatic conditionsꎬ which are similar in
the three regionsꎬ are not the primary causes affecting
the weed community. Thereforeꎬ the different manage ̄
ment practices ( tillageꎬ fertilisationꎬ herbicides appli ̄
cation etc. )ꎬ soil texturesꎬ topographical conditionsꎬ
and the environmental pressures of competition have
played decisive roles in the distribution of weeds in the
three distinct cites ( Gu et al. ꎬ 2007ꎻ Xiao et al. ꎬ
2008).
Long ̄term herbicide application significantly de ̄
5585 期 Zhang Haiyanꎬ et al. : Multivariate analysis of weed comminities in the Jiangsu autumn tea gardens
creases the diversity of plant species ( Chen et al. ꎬ
2000). Under long ̄term inter ̄row tillage and single
herbicide applicationꎬ some weakly competitive weeds
disappearedꎬ and the weed community shifted domi ̄
nance by annual weed species such as D. sanguinalisꎬ
E. crus ̄galliꎬ C. iriaꎬ E. indicaꎬ and M. strictaꎬ due
to their shorter growth periodsꎬ higher seed yieldsꎬ and
stronger reproductive capacities. Frequent herbicide
applications usually provide the selection pressure to
weeds which might lead to the evolution of herbicide
resistance in weed (Owen & Zelayaꎬ2005). As a re ̄
sultꎬ weed species with stronger herbicide resistance or
tolerance such as M. corchorifoliaꎬ P. bisulcatumꎬ P.
virginicaꎬ and E. humifusa occurred frequently in tea
gardens that received frequent herbicide applications.
In additionꎬ a correlation between tree age and in ̄
ter ̄row canopy density was found in the current study.
Generallyꎬ the older tea trees were taller and their in ̄
ter ̄row canopy density was higherꎬ resulting in less
aeration and sunlight penetration and combing with
more difficult farming operations. Thusꎬ shade ̄tolerant
or shade ̄loving weed species such as P. scandens var.
tomentosaꎬ R. hirsutusꎬ C. japonicaꎬ L. japonicumꎬ
and P. multifidaꎬ were widely distributed in the older
tea gardens with taller treesꎬ higher inter ̄row canopy
densityꎬ and less soil tillage. Howeverꎬ heliophilous
weed species such as C. iriaꎬ A. macrostemonꎬ and
C. segetum were significantly negatively correlated with
canopy density.
Slope sizeꎬ height and aspect affected the struc ̄
tureꎬ moistureꎬ and organic matter content of the soilꎬ
which in combination with the influence of lightꎬ af ̄
fected the distribution of weed community ( Sternberg
& Shoshanyꎬ2001). Weed species such as S. glabraꎬ
G. sojaꎬ and X. sibiricumꎬ which were distributed on
the upper parts of slopesꎬ are mostly heliophilous spe ̄
cies with higher drought and exposure tolerance. In
contrastꎬ weed species such as A. philoxeroidesꎬ F.
miliaceaꎬ E. prostrataꎬ and H. trionumꎬ which were
located in the lower parts or at the bottom of slopesꎬ
are species that favour higher humidity and greater a ̄
mounts of fertiliser.
CCA provides the most effective ordination method
for the study on the relationship between plant distribu ̄
tion and the environment (Zhangꎬ1995). The tea gar ̄
dens were clustered into three significantly different ec ̄
ological groups according to the CCA in combination
with weed communities and environment factors in cur ̄
rent study. Most of the sites in Group Ⅰ were located
in hilly areas with significant fluctuations in slope. The
tea gardens in this group were mature or older with soil
of higher sand and rock contentꎬ older treesꎬ and high ̄
er canopy density. Thereforeꎬ weed density was lowerꎬ
and more weed species were perennial and shade ̄toler ̄
ant. Most sites in Group Ⅱ were located in the transi ̄
tion zone (gentle slopes)ꎬ that tea trees were usually
planted on the contours of the slopes. Because of the
high selection pressure exerted by the primary weed
control of herbicide combined with inter ̄row tillage in
the tea gardensꎬ the weed community of Group Ⅱ had
more annual malignant weeds of the Gramineaeꎬ more
herbicide ̄resistant weedsꎬ but fewer species and peren ̄
nial weeds. The tea gardens of Group Ⅲ were close to
other crop fields or were previously non ̄tea crop fieldsꎬ
thereforeꎬ their terrains were flatter and the soils were
looser and more fertile. These tea gardens were finely
managed with inter ̄row tillage along with herbicide ap ̄
plicationꎬ resulting in less dominant weed species and
weed infestation.
Biological invasions are one of the most serious
environmental problems that human society is facing in
this century ( Millennium Ecosystem Assessmentꎬ
2005). In our surveyꎬ 23 alien invasive weeds were
found in the autumn tea gardens. Among theseꎬ the
Compositae was the most highly represented. Correla ̄
tion analysis of the relationships between the 23 alien
invasive weeds and environmental factors showed that
the frequency of herbicide application affected the im ̄
portance values and occurrence frequency of alien
weeds mostly. In the tea gardens with more frequently
herbicide applicationꎬ the frequent occurrence of C.
sumatrensisꎬ S. canadensisꎬ C. rotundusꎬ and E. hu ̄
mifusa were found. Several alien invasive weeds had
significant correlations with canopy density. Represen ̄
ting these speciesꎬ V. didyma and D. carota are
shade ̄tolerant weed species that can complete their en ̄
tire life cycles in a weak light environment. Additional ̄
lyꎬ other species such as P. americana and
658 植  物  保  护  学  报 42 卷
C. crepidioidesꎬ although they cannot live in a weak
light environment for longꎬ had a height advantageꎬ
displaying climbing ability or more rapid growth at the
seedling stage that enabled them to reach above the tea
tree canopy for adequate light or improves the efficien ̄
cy of light use to maintain a high growth rate under
weak light conditionsꎬ e. g. P. americana (Dong et
al. ꎬ2014).
Recentlyꎬ considerable attention has been focused
on the protection of weed biodiversity because of the
roles of weeds in maintaining the ecological balance of
agricultural ecosystems (Chen et al. ꎬ2004aꎻbꎻYang et
al. ꎬ2007) and greater numbers of plant species led to
greater temporal stability of ecosystem (Tilman et al. ꎬ
2006). The protection of weed biodiversity should be
considered in the implementation of field management
practices so as to maintain a relatively stable ecosystem
and to prevent the outbreak of dominant or alien inva ̄
sive species. The present study clearly demonstrates
that there are differences in weed community diversity
in different habitats of autumn tea gardens in the pri ̄
mary tea ̄producing areas of Jiangsu Province. The re ̄
sults have practical significance in guiding the scientif ̄
ic control of weed prevalence and to protect weed
biodiversity in tea gardens.
Acknowledgements
We would like to thank Professor Qiang Sheng
( Nanjing Agricultural University ) and Dr. Zhang
Zhengꎬ Gao Pinglei (Nanjing Agricultural University)
for their polishing the manuscript. We also thank Zhu
Yeqin ( Jiangsu Plant Protection and Quarantine Sta ̄
tion)ꎬ Ji Min (Plant Protection and Quarantine Station
in Jintan District of Changzhou ) and Pan Yunfeng
(Yixing Plant Protection and Quarantine Station) for
their invaluable technical assistance.
References
Chen Xꎬ Tang JJꎬ Fang ZGꎬ Shimizu K. 2004a. Effects of weed
communities with various species numbers on soil features in a
subtropical orchard ecosystem. Agricultureꎬ Ecosystems & En ̄
vironmentꎬ 102(3): 377 - 388
Chen Xꎬ Wang ZQꎬ Tang JJ. 2000. The ecological functions of weed
biodiversity in agroecosystem. Chinese Journal of Ecologyꎬ 19
(4): 50 - 52 (in Chinese) [陈欣ꎬ 王兆骞ꎬ 唐建军. 2000.
农业生态系统杂草多样性保持的生态学功能. 生态学杂志ꎬ
19(4): 50 - 52]
Chen Xꎬ Yang YSꎬ Tang JJ. 2004b. Species ̄diversified plant cover
enhances orchard ecosystem resistance to climatic stress and soil
erosion in subtropical hillside. Journal of Zhejiang University
(Science Edition)ꎬ 5(10): 1191 - 1198
Dong ZYꎬ Bai XFꎬ Zhang JZꎬ Hou YPꎬ Bu QM. 2014. Adaptability
of an invasive plant Phytolacca americana to varied light envi ̄
ronment. Chinese Journal of Ecologyꎬ 33(2): 316 - 320 ( in
Chinese) [董周焱ꎬ 柏新富ꎬ 张靖梓ꎬ 侯玉平ꎬ 卜庆梅.
2014. 入侵植物美洲商陆对光环境的适应性. 生态学杂志ꎬ
33(2): 316 - 320]
Gu QZꎬ Yang XYꎬ Sun BHꎬ Zhang SLꎬ Tong YA. 2007. Weed
biodiversity in winter wheat field of loess soil under different fer ̄
tilization regimes. Chinese Journal of Applied Ecologyꎬ 18(5):
1038 - 1042 (in Chinese) [古巧珍ꎬ 杨学云ꎬ 孙本华ꎬ 张树
兰ꎬ 同延安. 2007. 不同施肥条件下黄土麦地杂草生物多样
性. 应用生态学报ꎬ 18(5): 1038 - 1042]
Guo SLꎬ Li YH. 1998. Significance and method of studies on weed
niche in crop fields. Acta Ecologica Sinicaꎬ 18(5): 497 - 503
(in Chinese) [郭水良ꎬ 李扬汉. 1998. 农田杂草生态位研
究的意义及方法探讨. 生态学报ꎬ 18(5): 497 - 503]
Jaggi Sꎬ Sood Cꎬ Kumar Vꎬ Ravindranath SDꎬ Shanker A. 2001.
Leaching of pesticide in tea brew. Journal of Agricultural and
Food Chemistryꎬ 49(11): 5479 - 5483
Millennium Ecosystem Assessment. 2005. Ecosystems and human
well ̄being. Washington: Island Press
Owen MDKꎬ Zelaya IA. 2005. Herbicide ̄resistant crops and weed
resistance to herbicides. Pest Management Scienceꎬ 61 (3):
301 - 311
Qiang S. 2001. Weed science. Beijing: China Agricultural Press.
( in Chinese) [强胜. 2001. 杂草科学. 北京: 中国农业出版
社]
Qiang S. 2005. Multivariate analysisꎬ descriptionꎬ and ecological
interpretation of weed vegetation in the summer crop fields of
Anhui Provinceꎬ China. Journal of Integrative Plant Biologyꎬ 47
(10): 1193 - 1210
Qiang Sꎬ Chen GQꎬ Li BPꎬ Meng L. 2010. Invasive alien species in
Chinese agricultural ecosystems and their management. Biodi ̄
versity Scienceꎬ 18(6): 647 - 659 (in Chinese) [强胜ꎬ 陈国
奇ꎬ 李保平ꎬ 孟玲. 2010. 中国农业生态系统外来种入侵及
其管理现状. 生物多样性ꎬ 18(6): 647 - 659]
Qiang Sꎬ Hu JL. 1999. Quantitative analysis of weed communities in
cotton fields in cotton ̄growing regions of Anhui Province. Acta
Ecologica Sinicaꎬ 19(6): 810 - 816 (in Chinese) [强胜ꎬ 胡
金良. 1999. 江苏省棉区棉田杂草群落发生分布规律的数
量分析. 生态学报ꎬ 19(6): 810 - 816]
Qiang Sꎬ Li YH. 1990. On the distribution pattern of weed commu ̄
nities of summer crop fields in river valley and hilly lands of An ̄
hui Province. Acta Phytoecologica et Geobotanica Sinicaꎬ 14
7585 期 Zhang Haiyanꎬ et al. : Multivariate analysis of weed comminities in the Jiangsu autumn tea gardens
(3): 212 - 219 (in Chinese) [强胜ꎬ 李扬汉. 1990. 安徽沿
江圩丘农区夏收作物田杂草群落分布规律的研究. 植物生
态学与地植物学学报ꎬ 14(3): 212 - 219]
Qiang Sꎬ Shen JMꎬ Zhang CQꎬ Shao GYꎬ Hu JLꎬ Wang FL. 2003.
The influence of cropping systems on weed communities in the
cotton fields of Jiangsu Province. Acta Phytoecologica Sinicaꎬ
27(2): 278 - 282 (in Chinese) [强胜ꎬ 沈俊明ꎬ 张成群ꎬ 邵
耕耘ꎬ 胡金良ꎬ 王凤良. 2003. 种植制度对江苏省棉田杂草
群落影响的研究. 植物生态学报ꎬ 27(2): 278 - 282]
Siefert Aꎬ Ravenscroft Cꎬ Althoff Dꎬ Alvarez ̄Yepiz JCꎬ Carter BEꎬ
Glennon KLꎬ Heberling JMꎬ Jo ISꎬ Pontes Aꎬ Sauer Aꎬ et al.
2012. Scale dependence of vegetation ̄environment relation ̄
ships: a meta ̄analysis of multivariate data. Journal of Vegeta ̄
tion Scienceꎬ 23(5): 942 - 951
Sood Cꎬ Jaggi Sꎬ Kumar Vꎬ Ravindranath SDꎬ Shanker A. 2004.
How manufacturing processes affect the level of pesticide resi ̄
dues in tea. Journal of the Science of Food and Agricultureꎬ 84
(15): 2123 - 2127
Sternberg Mꎬ Shoshany M. 2001. Influence of slope aspect on Medi ̄
terranean woody formations: comparison of a semiarid and an
arid site in Israel. Ecological Researchꎬ 16(2): 335 - 345
ter Braak CJFꎬ Smilauer P. 2002. CANOCO reference manual and Can ̄
oDraw for windows user’s guide: software for canonical community
ordination (version 4􀆰 5). New York: Microcomputer Power
Thomas AC. 1985. Weed survey system used in Saskatchewan for
cereal and oilseed crops. Weed Scienceꎬ 33(1): 34 - 43
Tilman Dꎬ Reich PBꎬ Knops JMH. 2006. Biodiversity and ecosys ̄
tem stability in a decade ̄long grassland experiment. Natureꎬ
441(7093): 629 - 632
van Couwenberghe Rꎬ Collet Cꎬ Lacombe Eꎬ Pierrat JCꎬ Gégout JC.
2010. Gap partitioning among temperate tree species across a
regional soil gradient in windstorm ̄disturbed forests. Forest E ̄
cology and Managementꎬ 260(1): 146 - 154
Wang YZꎬ Ji MSꎬ Qi ZQꎬ Gu ZMꎬ Wei SHꎬ Zhang Yꎬ Li XH.
2008. Occurrence of weeds in orchards in Liaoning Province.
Plant Protectionꎬ 34(4): 98 - 101 (in Chinese) [王英姿ꎬ 纪
明山ꎬ 祁之秋ꎬ 谷祖敏ꎬ 魏松红ꎬ 张杨ꎬ 李兴海. 2008. 辽
宁省果园杂草发生情况调查. 植物保护ꎬ 34(4): 98 - 101]
Xiang ZXꎬ Shan WXꎬ He QHꎬ Xiao RLꎬ Xu HQꎬ Chen Pꎬ Cheng
X. 2009. Effect of ecologically ̄based weed management strate ̄
gies on weed community and diversity in hilly tea plantations.
Chinese Journal of Eco ̄Agricultureꎬ 17 (5): 857 - 861 ( in
Chinese) [向佐湘ꎬ 单武雄ꎬ 何秋虹ꎬ 肖润林ꎬ 徐华勤ꎬ 陈
佩ꎬ 程孝. 2009. 两种生态控草措施对丘陵茶园杂草群落及
物种多样性的影响. 中国生态农业学报ꎬ 17 (5): 857 -
861]
Xiao RLꎬ Xiang ZXꎬ Xu HQꎬ Shan WXꎬ Chen Pꎬ Wang GXꎬ Cheng
X. 2008. Ecological effects of the weed community in tea garden
with intercropping white clover and straw mulching. Transactions
of the Chinese Society of Agricultural Engineeringꎬ 24 (11):
183 -187 (in Chinese) [肖润林ꎬ 向佐湘ꎬ 徐华勤ꎬ 单武雄ꎬ
陈佩ꎬ 王桂雪ꎬ 程孝. 2008. 间种白三叶草和稻草覆盖控制
丘陵茶园杂草效果. 农业工程学报ꎬ 24(11): 183 -187]
Xu HGꎬ Qiang S. 2011. China invasive alien species. Beijing: Sci ̄
ence Press (in Chinese) [徐海根ꎬ 强胜. 2011. 中国外来入
侵生物. 北京: 科学出版社]
Xu HQꎬ Xiao RLꎬ Xiang ZXꎬ Wen GYꎬ Xiong W. 2010. Effects of
different ecological management on the weed community biodi ̄
versity in tea plantation. Chinese Agricultural Science Bulletinꎬ
26(4): 283 - 286 (in Chinese) [徐华勤ꎬ 肖润林ꎬ 向佐湘ꎬ
文国宇ꎬ 熊文. 2010. 不同生态管理措施对丘陵茶园杂草生
物多样性的影响. 中国农学通报ꎬ 26(4): 283 - 286]
Yang LXꎬ Sun YZ. 2013. Biodiversity of understory vegetation in
different ̄aged Manchurian walnut plantations. Chinese Journal
of Ecologyꎬ 32(4): 807 - 812 ( in Chinese) [杨立学ꎬ 孙跃
志. 2013. 不同林龄胡桃楸林下植物多样性的差异. 生态学
杂志ꎬ 32(4): 807 - 812]
Yang YSꎬ Wang Hꎬ Tang JJꎬ Chen X. 2007. Effects of weed man ̄
agement practices on orchard soil biological and fertility proper ̄
ties in southeastern China. Soil and Tillage Researchꎬ 93(1):
179 - 185
Yu CXꎬ Li GRꎬ Zhou CHꎬ Zhu XQꎬ Dai ZX. 1995. Jiangsu soil.
Beijing: China Agriculture Pressꎬ pp. 1 - 68 (in Chinese) [喻
长新ꎬ 李桂荣ꎬ 周传槐ꎬ 朱向群ꎬ 戴志新. 1995. 江苏土壤.
北京: 中国农业出版社ꎬ pp. 1 - 68]
Yu Mꎬ Zhou ZYꎬ Kang FFꎬ Ouyang Sꎬ Mi XCꎬ Sun JX. 2013. Gradi ̄
ent analysis and environmental interpretation of understory herb ̄
layer communities in Xiaoshegou of Lingkong Mountainꎬ Shanxiꎬ
China. Chinese Journal of Plant Ecologyꎬ 37(5): 373 -383 (in
Chinese) [余敏ꎬ周志勇ꎬ康峰峰ꎬ欧阳帅ꎬ米湘成ꎬ孙建新.
2013. 山西灵空山小蛇沟林下草本层植物群落梯度分析及环
境解释. 植物生态学报ꎬ 37(5): 373 -383 ]
Zhang HYꎬ Sun GJꎬ Ji Mꎬ Li FHꎬ Han Mꎬ Xu YLꎬ Wan YC.
2013. Interspecies relationships and clustering of weed commu ̄
nities in tea gardens in southern hilly regions of Jiangsu Prov ̄
inceꎬ east China in autumn. Chinese Journal of Ecologyꎬ 32
(9): 2289 - 2297 (in Chinese) [张海艳ꎬ 孙国俊ꎬ 季敏ꎬ 李
粉华ꎬ 韩敏ꎬ 许映莲ꎬ 万玉成. 2013. 苏南丘陵茶园秋季杂
草种间生态关系及群落分类. 生态学杂志ꎬ 32(9): 2289 -
2297]
Zhang JT. 1995. Method of vegetation quantitative ecology. Beijing:
Science Pressꎬ pp. 144 - 154 ( in Chinese) [张金屯. 1995.
植被数量生态学方法. 北京: 科学出版社ꎬ pp. 144 - 154]
Zhou ZYꎬ Li CCꎬ Hu BJꎬ Xu LNꎬ Li XX. 2012. Investigation on
main species of weeds in tea gardens in Anhui Province. China
Teaꎬ (1): 18 - 20 (in Chinese) [周子燕ꎬ 李昌春ꎬ 胡本进ꎬ
徐丽娜ꎬ 李晓霞. 2012. 安徽省茶园杂草主要种类调查. 中
国茶叶ꎬ (1): 18 - 20]
(责任编辑:高  峰)
858 植  物  保  护  学  报 42 卷