全 文 ：第 17 卷 第 4 期
Vol. 17 No . 4
草 地 学 报
ACTA AGRESTIA SINICA
2009 年 7 月
Jul. 2009
Characteristic Analysis of Seriphidium transillense ( Poljak. ) Poljak. Desert
Grasslands at Different Degraded Stages in the Northern Tianshan Mountains
LIU Hong- lai1, 2 , ZHU Jin-zhong1, 3, * , JIN Gu-i li1, 3 , FAN Yan-m in1, 3 , WANG Ca-i hong1
( 1. College of Grassland and E nvironmen tal Resources, Xinjiang Agricultural University, U rumqi, Xin jiang Uygu r Autonomous
Region 830052, Ch ina; 2. College of Animal Science and T ech nology, Ch ina Agricu ltural U niver sity, Bei jing 100193, China;
3. Key Laboratory of Xinjiang Grass lan d Resou rces and Ecology, Vygur, Xinjian g Vygur Autonomous Region 830053, C hina)
Abstract: In the mid-1980s, gr asslands ow ned by the state w ere contracted to families for 30 to 50 years,
and then increased liv estock populat ions impo sed heavy grazing pressure on Seriphid ium tr ansil l ense ( Po-l
jak. ) Po ljak. desert g rasslands. Three sites w ith differ ent deg radat ion degrees w ere chosen on the north-
ern slopes in T ianshan Mountains in w estern China to analyze the character ist ics of plant populations, soil
propert ies, and their relat ionships w ith dif ferent deg radat ion stag es. T he results demonst rate that the
dom inant species shif ted aw ay from S . tr ansil l ense into Petr osimonia sibi r ica ( Pall. ) Bunge which adopted
r-st rategy to produce innumerable seeds and then fo rmed large population and consequent ly protected the
grassland from completely co llapse. T he f requency , cover, yield, and density of S . tr ansil lense gr adually
decreased while P. sibir i ca increased and the feeding value of grassland became w orse along w ith the deg-
r adation degrees. T he composit ional shif t tow ar ds smaller, unpalatable and more ephemeral species w as
indicative of g rassland degr adation. The deg radation o f grassland vegetat ion was not in accordance w ith
soil deg radation. The surface soil had a t rend of becoming coarse w ith the degraded degr ees; how ever, the
soil org anic mat ter and total nit rog en steadily increased and reached the climax in over-degradat ion areas
w here the dung accumulation was the main reason for high so il nut rient content .
Key words:S er ip hidium tr ansi l lense ( Poljak. ) Poljak. desert ; Deg raded grassland; Soil propert ies; Vege-
tat ion characterist ics; N orthern T ianshan M ountains
天山北坡伊犁绢蒿荒漠不同退化阶段草地特征分析
刘洪来1, 2, 朱进忠1, 3, * , 靳瑰丽1 , 范燕敏1 , 王彩虹1
( 1. 新疆农业大学草地与资源环境学院, 乌鲁木齐 830052; 2. 中国农业大学动物科学技术学院, 北京 100193;
3.新疆草地资源与生态重点实验室, 乌鲁木齐 830052)
摘 要: 对天山北坡不同退化阶段伊犁绢蒿( S er ip hid ium trans illense ( Poljak. ) P oljak. )荒漠草地植被特征和土
壤特性进行分析,以期为该地区的草地恢复提供理论依据。结果表明:随着草地退化程度的加剧,伊犁绢蒿荒漠的
优势种逐渐被叉毛蓬(Petr osimonia sibir ica ( Pall. ) Bunge)替代;地带性植被伊犁绢蒿的频度、盖度、生物量和密度
随着草地退化程度的加剧逐渐降低,而叉毛蓬却呈相反趋势; 叉毛蓬种群在扩繁时采用/ r0对策, 产生大量种子迅
速占领领地,阻止了草地的完全崩溃。随着伊犁绢蒿荒漠退化, 草地的饲用价值劣化, 草地植物种类向低矮、适口
性差、短命和类短命植物过度。草地的植被退化与土壤退化不完全一致, 随着草地退化, 土壤表层颗粒有变粗趋
势,然而土壤有机质和全氮却呈现稳定升高的趋势;极度退化地区的高养分含量主要是由家畜排泄物导致。
关键词: 伊犁绢蒿荒漠; 特征分析;土壤特性; 植被特征
中图分类号: S812 文献标识码: A 文章编号: 1007-0435( 2009) 04-0419-09
收稿日期: 2008-03-19;修回日期: 2009- 01-04
基金项目:新疆维吾尔自治区重点实验室新疆草地资源与生态实验室开放课题( XJDX0209-2004-03)
作者简介:刘洪来( 1979- ) ,男,吉林人,博士研究生,研究方向为草地资源与生态学, E-mail: lh onglai@ 126. com; * 通讯作者 Author for
cor resp on dence, E- mail: x jauz jz@ 126. com
草 地 学 报 第 17卷
Gr assland is one of the most important natural
resources because it is the crucial barrier of ecolo g-
ical env ir onment and the basis of the sto ckbr eeding
development . Many types of grassland thr oughout
the w orld, especially in arid and sem-i arid reg ions,
have experienced a shift in dominant vegetat iv e
composit ion fr om perennial g rasses to shrubs or
bare soil w hich coincides w ith desert if ication [ 1~ 5]
and roughly 3. 6 @ 109 hm2 on 6 cont inents w er e
consider ed under deser tif icat ion[ 2] . China is one o f
the most severely desert if icat ion countries in the
w orld w ith up to 3. 3 @ 108 hm2 deser tif icat ion
land
[ 6~ 9]
, w hich maybe cause seriously environ-
mental problems. In 2001, sandstorm originating
from Xinjiang inf luenced Beijing, Japan, ev en U-
nited States w hich w as carried out through high a-l
t itude airf low . In recent years, grassland degrada-
t ion, w hich is the main form of desert ificat ion, has
become a major issue, att racting w idespread at ten-
t ion in China, especially in the arid and sem-i arid
climat ic zones.
In the last 20 years, grasslands have been se-
r iously demo lished by intensiv e disturbance related
to human activit ies and some g rasslands close to
the tenured farms have been in over-degr adation.
Over grazing by livestock has play ed a major r ole in
grassland degradat ion[ 2, 3, 10, 11] , consequent ly, the
quality and quantity o f the g rassland decline and
the contradiction betw een g rassland and stock-
breeding sharpens. M ost important , the deteriora-
t ion o f biolo gical and economic po tential is thought
to be largely irreversible
[ 3, 10, 12]
.
Xinjiang has 5. 73 @ 107 hm2 natural gr assland
including 1. 08 @ 107 hm2 spring-autumn pasture[ 13]
and Ser ip hidium transi ll ense ( Poljak. ) Po ljak.
desert is the main spring-autumn pastur e w hich be-
long s to m iddle Asiat ic climate desert that concen-
trates in no rthern Xinjiang plains and low
mountains, especially on the no rthern slopes of the
Tianshan M ountains. In the t ime sequence, the
plant community composit ion, st ructure as well as
some life-econom ic t rait s of m iddle Asiat ic climate
desert is mo re complex than that o f centr al Asiat ic
climate deser t[ 14] . Ow ing to the ecolo gical sensit iv-
ity of desert g rassland and long- term unreasonable
ut ilizat ion by lo cal people, the S . t r ansi l lense des-
ert had been ser iously degraded, our research
aimed at explo ring the characterist ic changes of
vegetat ion populat ions, soil propert ies, and their
relat ionships in order to provide informat ion fo r
resto rat ion o f the deg raded grassland ecosy stem.
Materials and Methods
Study area
T he study area ( 43b49cN ~ 43b56cN, 87b02cE
~ 87b05cE) , an open flat alluv ial plain belong ing to
the spring-autumn pastures of Ashili village, is lo-
cated on the middle part o f no rthern slope of the
T ianshan M ountains in Xinjiang U ygur Autono-
mous Reg ion, China ( F ig. 1) . It is approx imately
16 km northeast of Chang ji city, 40 km east to
U rumqi city and 32 km southeast to the A sian con-
t inent geog raphy center ( ACGC) . T he elev at ion
ranges betw een 750 and 950 m. T his area has
str ong f lorist ic aff inity to the middle Asia g eo-
gr aphic region that st retches f rom w estern Asia
through the T ianshan M ountains to w ester n M on-
go lia in terms of bo tanics. S. tr ansil l ense desert is
the zonal vegetat ion w ith ephemeral species in
spring and annuals form ing dominant synusia in
summer. The study area has an ext remely ar id cl-i
mate w hich belong s to the middle Asian desert cl-i
mate, w ith a mean annual rainfall of 180 to 190
mm ( 70% occurring in spring and summer) , a po-
tent ial annual evapor at ion > 1760 mm, and dr ying
index of 4 to 10. U nlike other ar id zones, this area
has an average annual cover period of 103 days,
f rom late November to mid-March of the fo llow ing
year
[ 15]
. T he mean annual temperature is 6. 5 e
w ith a hot summer and a cold w inter; the f rost- free
per iod in one year is about 160 to 190 day s. The
soil is classif ied as g ray deser t soil and the soil par-
ent material is the loess- like material.
Experimental design
T he research area is used by Kazak herdsmen
as grazing gr assland w ith seasonal cont inuous gr az-
ing most ly. The g razing act ivit ies occur primar ily
f rom late April to June, and September to Novem-
ber. Due to a policy change in mid-1980s, g rass-
lands ow ned by the state w ere contracted to fam-i
lies for 30 to 50 years
[ 16]
. The change in land-use
policy dramatically increased animal product iv ity,
420
第 4 期 LIU Hong- lai et al: Character istic Analysis of Ser ip hidium tr ansillense ( Poljak. ) Poljak. Desert Grasslands at Different Degraded Stages in the Northern Tianshan Mountains
but caused severe deg radat ion of g rassland areas.
Along w ith the liv estock populat ion increasing , the
grassland quality was in serious decline under the
huge st resses of overg razing. According to the ac-
tual condit ion, the gr assland of this zone can be d-i
v ided into three deg raded stag es, namely mid-deg-
r adation ( Ò ) , heavy-deg radat ion ( Ó) and over-
degradat ion ( Ô) [ 17] . The heavy-deg raded area w as
closer to resident ial area than m id-deg raded area
and had been distur bed w ith high f requency . T he
over-degraded ar ea w as chosen near the Santun riv-
er used as w atering place by livestock and had very
high density of domest ic livestock populat ion. A
permanent enclosing area lo cated near Junhu Farm
which had been enclosed beyond 30 years w as a-
dopted as control sample and defined as non-degra-
dation ( Ñ) ( F ig. 1) .
Fig. 1 The sketch map of study area in the northern Tianshan
Mountains of Xinjiang Vygur Autonomous Region
Measurements
The f ield invest ig ation w as carried out in early
April, 2006, w ithout corr esponding gr azing. At
each selected degraded area, 3 sample sites o f
1000 m2 wer e set up ( the distances among each site
w ere beyond 30 m ) and 5 census quadrats w ere a-
dopted in each site. Within each census quadrat ,
plant name, height, coverage, density , and fre-
quency w ere recor ded and the aboveg round biomas-
ses w ere measur ed. T he coverage w as measured by
recording all the species contacts made by a sharp
st ick for a total 100 points in each census quadr at .
The f requency values w ere est imated by throw ing
of 0. 25 m2-circle f rame 100 t imes into the sample
site and r ecording how of ten each species occurred.
The aboveg round biomass of each census quadrat
w as co llected by harvest ing, and all stems and
leaves above the ro ot co llar and the f resh w eights
w ere measured respectively.
T hr ee so il prof iles were dug in each selected
deg raded ar ea and so il samples fr om 0~ 10 cm, 10
~ 20 cm, and 20~ 30 cm depths w ere col lected.
Af ter air-drying, al l samples w ere g round to pass
0. 15 mm sieve opening and stor ed in plast ic bot t les
for physical and chem ical analy sis using standard
methods as described in the So il Analy sis book[ 18]
and Methods of Soil Analy sis[ 19] . Soil part icle size
w as classified into sand ( > 0. 05 mm) , silt ( 0. 002
~ 0. 05 mm) and clay ( < 0. 002 mm) . Soil part icle
size distr ibution was determined for the 0~ 10 cm,
10~ 20 cm , and 20 ~ 30 cm soil lay er s using a
Kohn-type pipette analyzer. Soil bulk density w as
measured by the co re method ( 100 mm diameter) .
Soil samples w ere pret reated w ith hydrogen
per oxide pr io r to digest ion test for soil or ganic
mat ter ( SOM) concentration. T he SOM ex tract ion
processes w er e conducted as fo llow ing : 10 g o f air-
dried soil w as put into a 200 mL pr opy lene f lask,
100 mL o f ext ractant ( 0. 5 N NaOH) w as added
and the air in the f lask w as displaced by N 2 , and
shook fo r 24 h at ro om temperatur e, then the dar k-
colored supernatant solut ion w as separated f rom
the residual soil by centrifug ation ( 10000 r pm fo r
10 m in) , the soil residue w as suspended in 50 m l
of dist illed w ater to separate the phases by centrif-
ugat ion as befo re, the w ashing s w ere added to the
super natant .
Soil pH and T otal salt ( TS) w ere measured in
a 1 B 5 ( soil B water) ex t ract using pH elect rode
and a conduct ivity meter, respectively. To tal n-i
t rog en ( TN ) and total pho spho rus ( T P) concen-
trat ions w ere determined on sub-samples using an
Alpkem autoanaly zer ( AA2, TECHNICON Com-
pany, USA) . T N w as analy zed using the standar d
Kjeldahl acid- digest ion method. TP concentr at ion
w as analy zed using acid digest ion method: mix ed
2. 0 g air- dried soil w ith 30 mL of 60% HClO4 in a
250 mL vo lumetric flask, and cont inue heated the
m ix tur e at the boiling temperature 20 m in longer,
cooled the mix ture, and add dist illed w ater to ob-
tain a volume of 250 mL. T otal Potassium ( T K)
concentrat ion w as analyzed using the digest ion
421
草 地 学 报 第 17卷
techniques ( HF-HClO 4 ) ; the ext racted f rom the
soil w as determined by atom ic absorpt ion spec-
trophotometer (AA-670, SHIMADZU Company, Japan) .
Statistical analysis
The stat ist ical analy sis w as carried out by
SPSS 11. 0 fo r Window s Package. Signif icant ly dif-
ferent means ( P < 0. 05) w er e ident ified by the
least signif icant dif ference ( LSD) af ter the conf ir-
mat ion of signif icant ef fects ( P< 0. 05) by one-w ay
ANOVA.
Results
1 Vegetation characteristics
There w ere obvious species composit ional
shift s among different degradat ions. 25 species
w hich belong to 15 sections w ere found in study ar-
ea and the species of Chenopodiaceae and Compos-i
tae accounted for 40% of the total species number.
T he species names and their impo rtant values in
each deg raded area are listed in Table 1. Species of
Chenopodiaceae and Compositae accounted fo r 40%
in the community at non-degraded area, and S.
t ransi ll ense was the dominant species ( T able 1) .
Due to the domest ic liv esto ck g razing , the impo r-
tant v alue of S. tr ansil l ense decr eased dr amat ically
in deg raded g rassland and it consequent ly lost the
dominant status and released mor e resources to
other plants leading to incr easing species number.
T here w ere 2 to 4 geophy tes occur red in each de-
gr aded stage in comparison w ith non-degradat ion
areas because these geophytes could successfully a-
void the t rampling of animals. At over-degraded
stag e, P. sibi ri ca replaced S. t ransi ll ense as the
dominant species, and the important value of T.
arcuata increased g radually along w ith the increas-
ing of degr adation deg ree. All these facto rs con-
tributed to the species composit ional shif t.
Table 1 Species names and important values in each degraded area
Species nam e
Degradat ion d egree
Ñ Ò Ó Ô
S er ip hid ium transi l len se ( Pol jak . ) Poljak. 0. 55 0. 36 0. 06 < 0. 01
P et rosimonia sibi ri ca ( Pall. ) Bunge 0. 16 0. 05 0. 04 0. 41
T rig one lla arcu ata C. A. Mey er 0. 03 0. 03 0. 05 0. 12
Ce ratocar p us arenarius L . 0. 07 0. 08 0. 03 0. 03
T ara xacum al taicum schischk. 0. 04 0. 02 0. 07 0. 01
Ma lcormia af ri cana ( L. ) R. Br. 0. 03 0. 01 0. 02 ―
Ce ratoce ph alus orthoceras DC. 0. 08 0. 01 ― 0. 02
G agea bulbi f era ( Pall. ) Roem. et S chu lt . ― 0. 36 0. 55 0. 05
T ul ipa i l ien si s Regel ― 0. 01 0. 01 0. 02
G eranium p ret ense L. ― ― 0. 08 0. 27
R ap eseri s sp. ― 0. 01 0. 01 ―
A st rag alu s v i car ious ― 0. 03 0. 02 ―
K ochia p r ost rat e ( L. ) Sch rad. ― ― 0. 01 ―
Sal sola col lina Pal l. ― 0. 01 ― 0. 05
K oe lp inia l ineari s Pall . ― ― 0. 04 0. 01
T ra gop og on kasahstanicu s S. Nikit i. 0. 01 ― ― ―
A lyssum sibi ri cum Willd. 0. 02 ― ― ―
C onvolvu lus ar v ensis L. < 0. 01 ― ― ―
L epid ium p er f ol ia tum L. 0. 01 ― ― ―
A nchu sa ovata Lehm. ― 0. 01 ― ―
Ch enop odium album L. ― ― < 0. 01 ―
Ce ratoide s lat eens ( J. F. Gm el. ) Reveal et H olmgren ― ― 0. 01 ―
P lantag o lessingi i Fisch. et M ey. ― ― ― 0. 01
A l l ium ch ry santhum Regel ― ― ― < 0. 01
P oa bulbosa L. ― ― ― 0. 01
Note: Ñ , Ò , Ó an d Ô repres ent non-degradat ion, mid-degradat ion, heavy-d egradat ion an d over-d egradat ion, respectively; the number
m eans the plant exi st ing in this s tage and it s importan t valu e ( IV) and / ―0 means the species does not exist in the stage and it s IV is zero, IV
= ( RC+ RY+ RD+ RF) / 4, RC, RY, RD, an d RF mean s relat ive coverage, biomass , density, and f requency, respectively; sam e as b elow
422
第 4 期 LIU Hong- lai et al: Character istic Analysis of Ser ip hidium tr ansillense ( Poljak. ) Poljak. Desert Grasslands at Different Degraded Stages in the Northern Tianshan Mountains
S. tr ansil l ense was the dom inant species in
non-degradat ion and its dominant status decreased
great ly w ith degraded degrees ( Table 2 ) and its
frequency, coverage, yield, and density changed
congr uously. In non-deg radat ion, S. tr ansi l lense
grew to more than 10 cm high, how ever, it became
shorter at degr aded areas. At over-degr aded stage,
no S. tr ansi l lense was found in quadrats, but it
w as found at low- lying spots during fr equency
sampling . T here w as few T . ar cuata in non-degra-
dation areas, but it increased cont inually w ith de-
gr aded deg ree and accordingly became a main spe-
cies in over-deg radat ion areas; it s fr equency, cov-
erage, y ield, and density increased w ith the degree
of degradat ion w hile it s height became sho rter.
Grassland feeding value decreased w ith degradat ion
stag es, although the species number of palatable
vegetat ion increased in comparison w ith non-de-
gr aded areas ( T able 3) . The yield of unpalatable
species increased steadily f rom mid- to over-deg ra-
dat ion stages.
Table 2 Characteristics of S. transillense, T. arcuata, and P. sibirica at different degradation stages
Characterist ic parameter Species
Degradat ion degree
Ñ Ò Ó Ô
Frequen cy ( % ) S. t ransi ll ense 72 82 42 8
T. arcuata 19 28 58 68
P . sibi ri ca 49 43 67 93
C overage ( % ) S. t ransi ll ense 34. 7 9. 1 1. 8 0
T. arcuata 0. 4 0. 9 1. 6 5. 1
P . sibi ri ca 1. 7 1. 8 0. 3 9. 9
Yield ( g m- 2) S. t ransi ll ense 102. 11 18. 69 3. 96 0. 00
T. arcuata 0. 94 0. 91 2. 21 7. 86
P . sibi ri ca 3. 94 0. 86 0. 09 8. 80
Den sity ( no. m- 2 ) S. t ransi ll ense 35 17 4 0
T. arcuata 6 9 16 60
P . sibi ri ca 81 69 9 454
H eight ( cm) S. t ransi ll ense 10. 4 1. 9 2. 4 0. 0
T. arcuata 1. 8 0. 7 0. 6 0. 5
P . sibi ri ca 1. 4 0. 6 0. 5 0. 5
Table 3 Change of grassland feeding value in different degradation
Forage select ion rat ing
Ñ Ò Ó Ô
Yield
( g m- 2 )
Cover
( % )
Yield
( g m- 2)
Cover
( % )
Yield
( g m- 2 )
Cover
( %)
Yield
(g m- 2)
Cover
( % )
Palatable 106. 1 35. 8 27. 7 22. 9 32. 5 33. 0 7. 7 4. 9
S . tr ansil l ense 102. 1 34. 7 20. 6 11. 5 4. 4 1. 4
T . arcua ta 0. 9 0. 4 0. 5 0. 7 1. 4 1. 3 6. 3 3. 2
T a. al tai cum 2. 5 0. 6 0. 7 0. 9 4. 6 1. 8 0. 2 0. 5
Ce ratoce . orthoceras 0. 6 0. 1 < 0. 1 0. 4 < 0. 1 0. 1
G . bulbi f e ra 3. 9 9. 2 15. 7 25. 7 0. 9 0. 9
T u. i li ensi s 0. 2 0. 1 0. 2 0. 5 0. 2 0. 1
K oe . l ineari s 3. 5 2. 0 0. 1 0. 1
A st rag alu s vi car ious 1. 8 0. 5 2. 1 0. 2
K oc . p rostr ata 0. 6 0. 1
Unpalatable 4. 9 2. 2 1. 2 1. 1 2. 7 2. 7 24. 3 27. 2
P. sibir ica 3. 9 1. 7 0. 3 0. 3 0. 1 0. 4 13. 4 13. 9
Ce ratoca. arenarius 0. 8 0. 5 0. 5 0. 7 0. 2 0. 2 0. 6 0. 9
M. af ri cana 0. 2 < 0. 1 0. 3 0. 1 0. 8 < 0. 1
G e. pr aten se 1. 6 2. 1 8. 9 11. 3
Sa. coll ina 0. 1 < 0. 1 1. 4 1. 1
423
草 地 学 报 第 17卷
2 Soil properties
2. 1 Bulk density and soil texture
So il bulk density and soil texture o f 0~ 10 cm,
10~ 20 cm, and 20~ 30 cm depths among dif ferent
degradat ion areas are show n in T able 4. T he soil
bulk density of each layer in degraded areas w as
low er than the counterpart lay er o f non-deg raded
ar ea. Soil bulk density of 0~ 10 cm gr adually de-
creased, although it did not dif fer significant ly;
never theless the average bulk density of 0~ 30 cm
increased w ith deg radation. From non-degradat ion
to over-degradat ion, the sand content steadily in-
cr eased from 42. 94% to 62. 40% indicat ing a rep-
resentat ive phenomenon of desert ificat ion. T he silt
content w as higher in deg raded areas than non-de-
gr aded areas w hile the clay content w as low er.
With the g rassland deg radation, the surface soil
became coarse.
Table 4 Comparison of bulk density and soil texture (means ? S. E. ) among different degraded stages
Parameter Depth ( cm)
Degradat ion degree
Ñ Ò Ó Ô
Bulk densi ty ( g cm- 3 ) 0~ 10 1. 41 ? 0. 01a 1. 39 ? 0. 04a 1. 35 ? 0. 04a 1. 34 ? 0. 02a
10~ 20 1. 59 ? 0. 01a 1. 44 ? 0. 02a 1. 49 ? 0. 01a 1. 44 ? 0. 02a
20~ 30 1. 60 ? 0. 02a 1. 34 ? 0. 03b 1. 52 ? 0. 03ab 1. 49 ? 0. 01a
S oil texture Sand ( % ) 0~ 10 48. 06 ? 0. 20a 48. 14 ? 0. 79a 46. 69 ? 0. 57a 36. 88 ? 0. 59b
10~ 20 36. 14 ? 0. 08b 47. 78 ? 1. 08a 49. 39 ? 0. 35a 29. 07 ? 0. 67c
20~ 30 44. 32 ? 0. 29b 44. 03 ? 0. 09b 63. 67 ? 0. 23a 30. 63 ? 0. 33c
S ilt ( % ) 0~ 10 17. 33 ? 0. 13c 22. 55 ? 0. 33a 17. 60 ? 0. 62c 19. 50 ? 0. 18b
10~ 20 16. 82 ? 0. 17c 22. 40 ? 0. 86a 17. 97 ? 0. 43bc 19. 03 ? 0. 80b
20~ 30 10. 92 ? 0. 17d 28. 57 ? 0. 39a 12. 59 ? 0. 33c 22. 12 ? 0. 37b
Clay ( % ) 0~ 10 34. 62 ? 0. 32b 29. 31 ? 0. 58c 35. 71 ? 0. 95b 43. 32 ? 0. 56a
10~ 20 47. 04 ? 0. 25b 29. 82 ? 0. 37d 32. 64 ? 0. 14c 51. 37 ? 0. 31a
20~ 30 44. 76 ? 0. 16b 27. 41 ? 0. 32c 23. 74 ? 0. 32d 48. 12 ? 0. 85a
2. 2 Chemical properties
The results in Table 5 indicate the chemical
property changes among different deg raded stag es.
The soil pH did no t differ signif icant ly either at 0
~ 10 cm o r at 20~ 30 cm, the pH at 10~ 20 cm in
over-degradat ion areas differed significant ly ( P <
0. 05) w ith other ar eas. T here w as no signif icant
dif ference fr om each other in T S content at dif fer-
ent degr adation stages.
SOM content had an incr easing t rend w ith
degradat ion in each soil lay er, the content in over-
degradat ion w as 1. 5 times of that in non- degrada-
t ion ar eas and the content steadily decr eased w ith
deeper layers in prof ile in both degr aded ar eas and
non-degraded areas. T he SOM content of 20~ 30
cm in non-degradat ion and 0~ 10 cm in over-degra-
dation areas dif fered signif icant ly ( P< 0. 05) fr om
their counterpart layers in o ther deg radation areas.
The T N content o f 10~ 20 cm and 20~ 30 cm in
non-degradat ion dif fered signif icant ly ( P < 0. 05)
from the same soil layers in o ther degr adation a-
reas. In pro file, the T N content decreased g radua-l
ly. T he AN content of 0~ 10 cm in over-deg rada-
t ion and 20~ 30 cm in non-degradat ion areas dif-
fered signif icant ly ( P< 0. 05) from the same layers
in other deg radat ion areas.
Discussion
Our results agreed w ith the finding that Ar te-
misia f rigida Willd. biomass declined remarkably
w ith the increase o f g razing pressure af ter a 10-
year g razing t rial in Inner Mongolia desert
steppe
[ 20]
and disagreed w ith some researches
[ 16, 21]
,
w hich demonstrated that the dominant species S.
t ransi li ense changed litt le under heavy g razing
pressure. In the Li and Jiang. s study[ 21] , litt le bio-
mass change o f S. t ransi li ense under 4-year s of
heavy grazing w as perfo rmed because of the rela-
t ively high reg row th capacit ies of all palatable spe-
cies including the dominant S . tr ansi l iense in the
f irst stage of the community ret rog ression. On the
424
第 4 期 LIU Hong- lai et al: Character istic Analysis of Ser ip hidium tr ansillense ( Poljak. ) Poljak. Desert Grasslands at Different Degraded Stages in the Northern Tianshan Mountains
Table 5 Soil chemical properties ( means ? S. E. ) among different degraded stages
Parameters Depth ( cm) Ñ Ò Ó Ô
pH 0~ 10 8. 70 ? 0. 02a 8. 80 ? 0. 02a 8. 87 ? 0. 06a 9. 18 ? 0. 10a
10~ 20 8. 90 ? 0. 01b 8. 78 ? 0. 05b 8. 91 ? 0. 06b 9. 64 ? 0. 05a
20~ 30 8. 92 ? 0. 01a 8. 89 ? 0. 10a 9. 01 ? 0. 08a 9. 57 ? 0. 11a
T S ( g kg- 1) 0~ 10 1. 06 ? 0. 01a 0. 97 ? 0. 08a 0. 72 ? 0. 05a 1. 57 ? 0. 05a
10~ 20 1. 42 ? 0. 02a 2. 13 ? 0. 03a 1. 55 ? 0. 6a 2. 45 ? 0. 09a
20~ 30 0. 40 ? 0. 01a 1. 43 ? 0. 04a 1. 22 ? 0. 03a 2. 65 ? 0. 06a
SOM ( g kg- 1 ) 0~ 10 9. 17 ? 0. 14b 12. 22 ? 0. 06ab 12. 02 ? 0. 42ab 18. 04 ? 0. 10a
10~ 20 7. 86 ? 0. 17a 9. 17 ? 0. 07a 10. 30 ? 0. 18a 10. 66 ? 0. 03a
20~ 30 5. 78 ? 0. 08b 8. 77 ? 0. 02a 10. 10 ? 0. 03a 9. 37 ? 0. 06a
T N ( g kg- 1) 0~ 10 0. 95 ? 0. 01a 0. 78 ? 0. 03a 0. 75 ? 0. 03a 1. 22 ? 0. 03a
10~ 20 0. 44 ? 0. 01b 0. 61 ? 0. 04a 0. 77 ? 0. 03a 0. 74 ? 0. 01a
20~ 30 0. 35 ? 0. 01b 0. 51 ? 0. 01ab 0. 71 ? 0. 03a 0. 67 ? 0. 02a
AN ( mg kg- 1 ) 0~ 10 23. 7 ? 1. 0b 29. 1 ? 1. 9b 29. 7 ? 1. 2b 52. 1 ? 4. 0a
10~ 20 16. 1 ? 0. 9a 19. 8 ? 3. 6a 20. 1 ? 1. 5a 23. 1 ? 2. 9a
20~ 30 13. 2 ? 0. 8b 23. 5 ? 1. 0a 31. 7 ? 0. 7a 21. 9 ? 3. 9a
contrary, our study clarif ied a mar kedly declining
tr end of S. tr ansi l iense biomass w ith the aggrava-
t ion o f the degraded grassland caused by freely
heavy grazing. T he r eason w as the capacity for S.
tr ansi l iense reg row th had decr eased greatly and ac-
cumulated lit t le nourishment in the autumn for
surviving the severe w inter and the grow ing o f
nex t spring af ter long- term heavy g razing . T he
controversy among Li and Jiangcs[ 21] , Zhaocs [ 16] ,
and ours studies is probably caused by the follow-
ing reasons: L i and Jiangcs study site w as in ro ta-
t ional g razing area w hile ours in free g razing ar ea;
Zhaocs study area belongs to central Asia climate
and ours belong s to m iddle A sia climate.
Some r esearchers suggest that livestock graz-
ing has posit ive impact , particular in response to
light gr azing, on plant diversity[ 22, 23] and sheep sa-
l iv a enhances aboveg round net pr imar y productiv-i
ty[ 24] , w hile others provide more negat ive impacts
rang ing f rom simple lo sses in cover [ 25, 26] to specif ic
composit ional shift s
[ 27]
and change in plant life-
fo rms[ 28] . T he loss in vegetat ion cover and compo-
sitional shif t aw ay from zonal v egetat ion may cause
great eco logical problems in desert g rasslands, so
the ecolog ical funct ion is more important than
stockbreeding because once the desert gr assland
become over-deg radation, it is hard to get back to
the o riginal ecosystem.
The soil part icle size distr ibution analy sis ind-i
cates that the deg radat ion ef fects facilitated the re-
moval of clay part icles result ing in coar sening o f
soil tex ture w hich is also a signal of grassland deg-
radat ion. In our study , the sand part icle content
increased 24% and the clay part icle content de-
cr eased 27% w ith the grassland change f rom non-
deg radat ion into heavy-deg radation w hile the over-
deg radat ion had the highest clay part icle content
and the low est sand part icle content. Soil ero sion
by w ind is the main cause for changes in soil part-i
cle composit ion[ 29] . In our resear ch area, G. bulb-
if era occurr ed in large number in m id- and heavy-
deg raded ar eas, and its coverage w as 9. 2% and
25. 7% , respect iv ely . How ever, G. bulbif era took
lit t le effect against w ind erosion because o f its
slightness and short . Except for the coverage of G.
bulbif er a, the ef fective cover of vegetat ion w as
38. 2%, 14. 3% and 7. 2% from non-degradat ion to
heavy-degradat ion respectively. The decrease of ef-
fect iv e vegetat ion cover enhanced the probability of
w ind ero sion and then led to the soil coarsening.
T he highest clay part icle and the low est sand con-
tent in over-degradat ion at t ributed to the vegeta-
t ion coverage o f 32. 6% and only 0. 9% accounted
by G. bulbi f er a, and the high bulk density par tly
avoided the chance of w ind erosion. In our study,
the average so il bulk density increased from 1. 39 g
cm
- 3
to 1. 46 g cm
- 3
f rom mid-degradat ion to over-
deg radat ion, because of soil compaction promoted
by livestock t rampling . In non-deg radat ion areas
w ithout disturbance for more than 30 year s, the
soil bulk density w as the highest and it fo rmed a
brow n t ight layers appeared as emplast ic mat ter
425
草 地 学 报 第 17卷
and iron mat ter f rom geo logy perspective.
The concentrat ion of SOM and TN are assoc-i
ated w ith the silt and soil part icles rather than
sand[ 30] . Therefo re, the r emoval of the majority o f
silt and clay part icles dur ing the g rassland degrada-
t ion process could direct ly result in deplet ion o f
SOM and TN. In the tr ansformation from poten-
t ial desert if icat ion land to ext remely deser tif icat ion
land in H orqin Sandy Land ( in no rthern China ) ,
the sand content at the 0 ~ 15 cm soil layer in-
creased from 69% to 93%, while org anic carbon
and T N contents decreased by 65% and 69% , re-
spectively
[ 31]
. In this paper, how ever , the sand
content increased 24% from non-degradat ion to
heavy-degr adation, the concentrat ions of SOM and
TN also increased w ith 32% and 31%. T his ab-
normal phenomenon gr eat ly ar oused our at tention.
Did the deg raded S. tr ansi l iense desert gr assland
cause the so il nut rient accumulation? The sour ces
of SOM in natural gr assland ecosy stem are rem-
nant plant leaf, stem, ro ot , soil life- forms and
livestock. s excrement. S . tr ansil l ense is per ennial
subshrub, leaf is slender and stem is the main bio-
mass; so the substance returned to environment
should be few . In deg raded g rassland, S . tr ansil-
lense was subst ituted by annuals and ephemeral
vegetat ion w hich had high density . Af ter grow ing
period, their above and under ground yields r e-
turned to soil becom ing the main source o f SOM
and ano ther essent ial sour ce w as the livestockc s
excr ement . The l iv esto ck populat ion had high
density in over- degraded area because of the w ate-
r ing place nearby , after drinking the livestock rest
in the neighboring g rassland and therefore , the
grassland w as subjected to great disturbance and
became over-deg raded. The excrement w as the
main source of soil fer tility in this area and the
high ef fect ive vegetat ion coverage w as a vital fac-
to r to avoid wind ero sion and pr eserv e more silt
and clay par ticles.
Conclusion
With grassland deg radat ion, the plant compo-
sition shif ted aw ay from S. tr ansi l iense, a perenn-i
al sub- shrub, to plants w ith annual and ephemeral
life histories including drought- resistant, t rample-
tolerant, or unpalatable species. The zonal plant
became subordinate component in the ear ly deg ra-
dat ion and dropped out gradually and f inally disap-
peared. P. sibi ri ca to ok the place of S. tr an-
si li ense and became the dom inant species in the o-
ver-degraded stage. From the aspect of grassland
feeding value, although the species number o f ten-
der palatable g enera increased, the to tal palatable
vegetat ion yield decreased dramat ically and conse-
quent ly led to the grassland quality w o rsening.
With the g rassland deterio rating fr om non-
deg radat ion to heavy-degradat ion, the sand part icle
content increased 24% caused by w ind erosion and
led to the so il coarsening . T he concentrat ions of
SOM and T N become higher w ith degr adation and
it w as no t in accor dance w ith the vegetat ion deg ra-
dat ion. T he livestock. s excrement accumulat ion
w as the main cause of the high soil nut rient.
Acknowledgements
Funding was provided by the pro ject / Degrad-
ed Succession and Restorat ion in Xinjiang Spring-
autumn Pastur e ( XJDX0209-2004-03) 0 from Xin-
jiang Uygur Autonomous Region Key Laboratory
of Grassland Resources and Ecolo gy. Authors are
g rateful to Dr. Yuguang Bai, ( Univer sity of
Saskatchew an, Canada) fo r his helpful comments
on the previous draf t of the manuscript . We also
express our sincer e thanks to anonymous reviewers
for their const ruct ive suggestions and rev ision of
the manuscript. Special thanks to Pro f. Shazhou
An ( Xinjiang Ag ricultur al University ) for his as-
sistance.
References
[ 1] Anderson P M L. Th e impacts of su stain ed heavy grazing on
plant diversity and com pos it ion in low land and upland habitat s
across the Kamiesberg mountain ran ge in the Su cculent Kroo,
South Africa [ J] . Journal of Arid Environments, 2007, 70:
686
[ 2] Daily G C. Restoring valu e to the w orlds d egraded lands [ J ] .
Science, 1995, 269: 350
[ 3] Van Auk en O W. Shrub invas ions of North American semiarid
426
第 4 期 LIU Hong- lai et al: Character istic Analysis of Ser ip hidium tr ansillense ( Poljak. ) Poljak. Desert Grasslands at Different Degraded Stages in the Northern Tianshan Mountains
grass lan ds [ J] . Annual Review of Ecology and Sys tematics,
2000, 31: 197
[ 4] S cheffer M , Car penter S, Foley J A, et al . Catast rophic shif t s
in ecosystems [ J] . Natur e, 2001, 413: 591
[ 5] Jackson R B, Banner J L , Jobbagy E G, e t al . E cosystem car-
bon los s w ith woody plan t invasion of grass lan ds [ J] . Natu re,
2002, 418: 623
[ 6] C hen G, Dong Z, Yan P. Desert ificat ion: international r e-
search topics an d research st rategies of China [ J ] . Explorat ion
of Nature, 1996, 15: 1-5
[ 7] Zha Y, Gao J. Ch aracterist ics of desert if icat ion an d it s rehabil-
itation in China [ J] . Journal of Arid E nvironmen ts, 1997, 37:
419
[ 8] Zhang D C. Dynamical E volu tion of S and S ea in Chin a [ M ] .
Gansu Cultu re Pr ess , 1998.
[ 9] Zhong D C, Qu J J . Recent developm ent t rend and predicat ion
of san d deserts in Chin a [ J] . Journal of Arid Environments,
2003, 53: 317-329
[ 10] Laycock W A. Stable states and thresholds of rang e condit ion
on North American rangelan ds: a view point [ J ] . Journal of
Range Managemen t, 1991, 44: 427
[ 11] Fleis chner T L. Ecological cost s of livestock grazing in w estern
N orth America [ J] . Con servat ion Biology, 1994, 8: 629
[ 12] Valon e T J, Meyer M , Brow n J H , et a l . T imescale of peren-
nial grass recovery in deser tif ied arid grasslands follow ing l ive-
stock r emoval [ J ] . Conservat ion Biology, 2002, 16: 995
[ 13] Xu P. Xinjiang grassland res ou rce and ut ilizat ion [ M ] . Xin-
jiang H ealth Technology Publis her, 1993.
[ 14] Zhu J Z, Wu Y M. Analys is of economic character evolvem ent
of th e des ert grassland of Seriphidium transil len se in different
degen erate stages [ J] . Prataculural Science, 2005, 12: 1-6
[ 15] Li J, Zhao C, Zhu H , et al . Ef fect of plant species on sh rub
fertil e island at an oas is-d esert ecotone in the South Junggar
Basin, Chin a [ J] . Journal of Arid En vi ronmen ts , 2007, 71:
350-361
[ 16] Zh ao W Y, Li J L, Qi J G. Changes in vegetation diversity and
s t ructure in response to h eavy grazing pressu re in the northern
T ianshan Moun tains, China [ J ] . Journal of Arid Environ-
m ents, 2007, 68: 465-479
[ 17] Liu H L , Zhu J Z, Jin G L, et al . Divis ion on degraded su c-
cess ional series of Seriphidium transil len se desert grassland
[ J ] . Xinjiang Agricultur e Sciences, 2007, 44: 137-141
[ 18] S cience N I O S. Analysi s of soil ph ysic-ch emical features
[ M ] . Sh ang hai Science and T ech nology Press, 1980.
[ 19] America S S S O. M ethods of soil analys is [ M ] . Madison,
Wisconsin , USA, 1982.
[ 20] Li J H, Li Z Q, Ren J Z. Ef fect of graz ing intensity on clonal
morph ological plast icity and biomass al location patterns of Ar-
temisia f r ig ida and P otent il la acaul i s in the In ner Mongolia
steppe [ J ] . New Zealand Jou rnal of Agricul tural Research ,
2005, 48: 57
[ 21] Li J L, Jiang P. T he ef fect s of dif f erent rotat ional grazing in-
tensi ties on the soil, grassland and sheep product ions in the
northern Tianshan in China [ J] . Ar chivos de Zootecnia, 1997,
46: 301
[ 22] Noymei r I, Gutman M, Kaplan Y. Responses of Mediterra-
nean grassland plants to grazing an d protect ion [ J] . J ou rnal of
Ecology, 1989, 77: 290
[ 23] West N E. Biodiversity of rangelands [ J ] . J ou rnal of Range
Manag ement , 1993, 46: 2
[ 24] Zhan g Z, Wang S P, Jiang G M , et al . Resp on se of A r temisia
f rigid a Willd. ( Com pos itae) and L ey mus chinensi s ( Trin . )
T zvel . ( Poaceae) to sheep sal iva [ J] . Journal of Arid Envir on-
men ts , 2007, 70: 111
[ 25] Sh altout K H, E lhalaw any E F, Elkady H F. Consequences of
protection f rom grazing on diversity an d abundance of the
coastal low land vegetat ion in eastern Saudi Arabia [ J] . Biod-i
versi ty and Conservat ion, 1996, 5: 27
[ 26] Eccard J A, Wal ther R B, Milton S J. H ow l ivestock grazin g
af fects vegetation st ru ctures and small mammal dist ribu tion in
the s em-i arid Kar oo [ J ] . Journal of Arid En vi ronments , 2000,
46: 103
[ 27] T odd S W , Hof fman M T . A fence- line cont rast reveals ef fect s
of heavy grazing on plant divers ity an d community comp os iti on
in Namaqualan d, South Africa [ J ] . Plant Ecology, 1999,
142: 169
[ 28] Noy-M eir I, Oron T. E ffects of grazing on geophytes in Med-i
terranean vegetat ion [ J ] . Journ al of Vegetat ion Science,
2001, 12: 749.
[ 29] H eW Q, Gao W S , T uo D B, et al . Study on som e factors in-
fluencing soil erosion by w ind tu nnel experiment in north eco-
ton e betw een agriculture and pasture [ J] . Journal of S oil an d
Water Conversat ion, 2004, 18: 1- 8
[ 30] Lobe I, Amenlung W, Du P C C. Losses of carb on and n it ro-
gen with prolonged arable cropping f rom sandy soils of the
South African H ighveld [ J] . European Journal of Soil Science,
2001, 52: 93- 101
[ 31] Yong Z S , Ha L Z, Wen Z Z, e t al . Fractal featu res of s oil
particle si ze dist ribu tion and th e im plicat ion for indicat ing de-
sertif icat ion [ J] . Geod erma, 2004, 122: 43-49
(责任编辑 李 扬)
427