Content uploaded by Yechun Wang
Author content
All content in this area was uploaded by Yechun Wang
Content may be subject to copyright.
1 23
Annals of Forest Science
Official journal of the Institut National
de la Recherche Agronomique
(INRA)
ISSN 1286-4560
Volume 68
Number 6
Annals of Forest Science (2011)
68:1127-1141
DOI 10.1007/s13595-011-0128-5
Shrub island effects on a high-altitude
forest cutover in the eastern Tibetan
Plateau
Yechun Wang, Weikai Bao & Ning Wu
1 23
Your article is protected by copyright and
all rights are held exclusively by INRA and
Springer Science+Business Media B.V.. This
e-offprint is for personal use only and shall
not be self-archived in electronic repositories.
If you wish to self-archive your work, please
use the accepted author’s version for posting
to your own website or your institution’s
repository. You may further deposit the
accepted author’s version on a funder’s
repository at a funder’s request, provided it is
not made publicly available until 12 months
after publication.
ORIGINAL PAPER
Shrub island effects on a high-altitude forest cutover
in the eastern Tibetan Plateau
Yechun Wang &Weikai Bao &Ning Wu
Received: 17 December 2010 / Accepted: 10 March 2011 / Published online: 2 September 2011
Abstract
&Context The roles of woody-plant islands are well
documented in low-altitude regions, but research related
to such shrub effects in high-altitude regions is scant.
&Aims Four common shrub species (Cerasus trichostoma,
Ribes glaciale,Rosa omeiensis and Salix sphaeronymphe)
in a high-altitude forest cutover of the eastern Tibetan
Plateau, were chosen to evaluate the effects of both species
and size of shrub islands on microhabitats, herbaceous
communities and woody seedling regeneration.
&Methods Total 86 shrubs with different sizes were investi-
gated; The shrub size, herb community structure and species
composition, litter, soil nutrient and microclimate parameters
beneath the shrub canopies were also measured.
&Results All shrubs significantly ameliorated microcli-
mates, increased content of soil organic matter and total
nitrogen, both grass and forb species richness, and litter
cover and biomass, and promoted woody seedling recruit-
ment (richness and number), but decreased cover and
biomass of the herbaceous community beneath them. These
effects were greater for larger shrubs, and also varied
among shrub species with different crown architectures. We
also found differences in species-dependency of the shrub
effect for the responses of the herbaceous and woody
seedling species, suggesting that shrubs also indirectly
facilitate forbs and seedling regeneration through competi-
tion release of grasses. We conclude that shrub-island
effects are size- and species-dependent. In order to
accelerate natural succession and restoration in alpine
cutovers, shrub island preservation and their effective
utilization as reforestation microhabitats should be integrat-
ed into vegetation management procedures.
Keywords Shrub island effect .Alpine cutover.Tibetan
Plateau .Herbaceous community.Progressive succession
1 Introduction
Shrub-island effects are widely recognized in many
ecosystems (Maestre and Cortina 2005;Duarteetal.
2006; Endo et al. 2008), and can generally be grouped into
two categories. One is the microhabitat effect, which
includes mainly microclimate amelioration (Franco and
Nobel 1989;Endoetal.2008) and soil nutrient enrich-
ment (Garner and Steinberger 1989; Throop and Archer
2008). The other is the nucleation succession effect, which
involves mainly the replacement of herbaceous species
and the acceleration of woody seedling regeneration
(Pugnaire et al. 1996;Duarteetal.2006). Cuesta et al.
(2010) found that the shrub Retama sphaerocarpa not
only directly facilitated late-successional Quercus ilex
seedlings by reducing seedling photoinhibition and water
stress,butalsoimprovedseedlinggrowthindirectlyby
reducing the competitive capacity of herbs. In most cases,
these two kinds of shrub-island effects (including direct
and indirect effects) occur simultaneously and are equally
important, especially for ecosystems in stressed environ-
Handling Editor: Gilbert Aussenac
Y. Wang :W. Bao (*):N. Wu
Key Laboratory of Ecological Restoration,
Chengdu Institute of Biology, Chinese Academy of Sciences,
No. 9, Section 4, Renming South Ave, P.O. Box 416, Chengdu,
Sichuan 610041, People’s Republic of China
e-mail: baowk@cib.ac.cn
Y. Wang
Graduate School of the Chinese Academy of Sciences,
Beijing 100039, People’s Republic of China
Annals of Forest Science (2011) 68:1127–1141
DOI 10.1007/s13595-011-0128-5
#INRA and Springer Science+Business Media B.V. 2011
Author's personal copy
ments (Brooker et al. 2008). Indeed, shrub-island effects
are managed by a complicated interaction among biotic
and environmental factors, with direct and indirect, and
positive and negative interaction mechanisms, and vary
across spatial and temporal scales (Levine 1999;Reisman-
Berman 2007; Brooker et al. 2008). Therefore, a better
understanding of shrub island effects across spatio-
temporal scales will provide an important pathway
towards understanding plant–environment interaction
mechanisms.
Shrubs with different size and crown architectures
differ in how they affect the microhabitat (Li et al.
2007; Throop and Archer 2008). These differences may
lead to the formation of different assemblages of herba-
ceous and woody plant establishment beneath the shrub,
which will affect patterns of nucleation succession,
involving mainly the replacement of herbaceous species
and the acceleration of woody community regeneration
(Pugnaire et al. 1996;Duarteetal.2006; Endo et al.
2008). In addition, different plant species, such as grasses
and forbs, may be affected differently by shrub islands;
forbs usually exhibit morphological, life historical and
ecophysiological characteristics that contrast sharply
with those of grasses, e.g., their intrinsically higher
photosynthetic capacity and resource-use efficiency
compared to grasses (Turner and Knapp 1996). They
can therefore, be expected to react differently to shrubs if
shrubs affect light intensity and soil nutrient levels. Alpine
plants have shorter growth periods, and have to endure
higher radiation and lower temperatures when compared
with low elevation regions. Thus, alpine plants face more
rigorous challenges and experience different growth
processes, and have developed various adaptation strate-
gies to cope with those environmental restrictions (Körner
2003; Dona and Galen 2007). Investigating the roles of
alpine shrub islands would help our understanding of the
mutual interaction of plants and vegetation dynamics in
the unique alpine environment.
The roles of woody-plant islands in forest restoration
have been well documented in low-altitude regions (Mattson
and Putz 2008). However, relevant research in high-altitude
destroyed regions is scant (Dona and Galen 2007), and the
relationships between species, island size and their effects in
these habitats are not well understood.
Alpine cutovers are found extensively on the eastern
Tibetan Plateau, as a consequence of excessive logging
alpine forests near the timberline (3,400–3,900 ma.s.l.) (Wang
et al. 1995). Currently, most recent cutovers (< 25 years)
remain in the early succession stage, still covered by
herbaceous communities with scattered shrubs (Bao
2004). Although traditional reforestation has been carried
out after clear felling for the past 30 years, most such
efforts have not succeeded, due mainly to the harsh
climate, environmental degradation and yak grazing
(Wang et al. 1995;Bao2004). Developing ways to
effectively promote restoration of alpine cutovers has
been being a pressing problem related to the establish-
ment of regional ecological safety barriers. Previous
work in the alpine region has found that cultivating
spruce seedlings beside tall shrubs can allow better
survival and growth, implying that microhabitat improve-
ment by shrubs is probably one of the mechanisms
responsible for this (Wang et al. 1995;Bao2004), but this
has not yet been tested. It is well known that vegetation
succession on cutovers after forest logging proceeds from
herbs to shrubs and finally to the forest stage, and that
shrubs provide a key link from herbs to the woody stage,
but there is little information on succession mechanisms
relating to the role of shrubs on alpine clear-cut areas.
In the present study, therefore, we chose four
common shrub species to evaluate the effects of shrub
size and species on a high-altitude cutover in the eastern
Tibetan Plateau. The following questions were
addressed. (1) How do shrubs affect microhabitats in
the high-altitude cutover? (2) How do shrub islands
affect the herbaceous community in the understorey,
especially grasses and forbs? (3) How do shrub islands
affect natural recruitment of woody seedlings? (4) How
do these effects vary with shrub size and species?
2 Methods
2.1 Study site
The study area is located in Rangtang County, northwest
Sichuan Province, China (32°19′N, 100°48′E). It is a
typical high-altitude forest–grassland ecotone of the
eastern Tibetan Plateau. It lies in a plateau monsoon
climate area with an annual average temperature of 4°C
and an annual average rainfall of 700–800 mm. The
growing season for vascular plants is about 90–105 days
per year, from late May to early September (Wang et al.
1995). The main vegetation types in the region are
primary spruce forests, alpine scrublands and meadows.
Large-scale harvesting of the original forests occurred
from 1976 to 1998, and left a sequence of cutovers of an
average size of about 5 ha. Most cutovers are dominated
by herbaceous communities with sparse shrubs. Our study
was undertaken on an 18-year-old cutover (area: 5.3 ha;
slope: 23°; aspect: NW 26°; elevation: 3,650 ma.s.l.),
where herbs and shrubs account for about 85 and 15% of
total cover, respectively.
1128 Y. Wang et al.
Author's personal copy
2.2 Focal shrub species
Four deciduous shrub species, Cerasus trichostoma,
Ribes glaciale,Rosa omeiensis and Salix sphaeronymphe
(hereafter referred to as Cerasus, Ribes, Rosa and Salix,
respectively), were selected because they are common on
the high-altitude cutovers, and have different crown
architecture and fruit type: Ribes (low-branching, fleshy
fruits), Cerasus and Rosa (medium-branching, fleshy
fruits), and Salix (high-branching, dry fruits). In August
2008, Cerasus, Ribes, Rosa and Salix of different sizes
were chosen and the corresponding area, age and height
were investigated in the field (Table 1). Crown cover was
estimated from crown projection diameter as the shrub
area. The distance from the ground to the mean height of
most sprout stems were determined by tape measure as the
shrub height. The age of the shrub was determined by
counting the growth rings of the thickest branch.
2.3 Microclimate measurements
To investigate the microclimate effect of shrub size and
species, we chose 48 additional shrubs randomly, and
then classified them into three classes according to area
size: small (0.5–1.0 m
2
)medium(2.0–3.0 m
2
)andlarge
(5.0–7.0 m
2
) shrubs, with four shrubs for each of the three
classes. Light intensity, air temperature and relative
humidity at 5 cm above the ground on the down-slope
middle position of the shrub canopy were measured from
12:00–14:00 pm on two sunny days (17–18 August 2008).
We presumed that the selected summer sunny days at noon
could represent the common typical situation beneath
shrub islands and could allow better comparison of
microclimate differences. Light intensity was measured
using a TES-1339 Light Meter Pro (TES Electrical
Electronic, Taiwan). Air relative humidity was calculated
from readings of dry and wet bulb temperatures of a
psychrometer (Red Star Instrument, Hebei, China), and
the dry bulb temperature regarded as air temperature.
Moreover, meadow without shrub cover on the cutover
was selected as a control check (CK) and the same climate
variables were measured at the same time. For each
variable, we took at least three readings and used the
mean value for the statistical analysis. Light transmission
was derived from the ratio of light intensity under the
shrub and the CK, and vapor pressure deficit (VPD) was
calculated (Jones 1992) from air temperature and relative
humidity.
2.4 Community survey
Firstly, the projected canopy area for each of the 86
shrubs was regarded as a plot. The plot was divided
into four parts (north, south, east and west) in which all
herbaceous species, and number and species richness of
woody seedlings were recorded. We then set one subplot
of 50 cm×50 cm (0.25 m
2
) in the center of each part to
investigate coverage of the total herbaceous community,
grasses, forbs and litter. We also measured above-ground
herbaceous biomass and litter mass by clipping all the
herbaceous plants at ground level and collecting the litter,
respectively. We regarded the part as one subplot if it was
toosmalltosetupa50cm×50cmsubplot.Therewere
only 14 such subplots (parts) of insufficient sampling size
(i.e., 50 cm×50 cm) in 11 small shrub islands. In addition,
we established 70 plots of 1 m×1 m in the meadow
between shrub islands as a control (CK), where all the
parameters investigated for shrubs were measured. Sam-
ples of plants and litter were transferred into the
laboratory, dried at 70°C for 12–13handweighedfor
dry mass. Plant individuals found during the survey were
identified in the field when possible or collected for later
identification at the herbarium. We calculated the average
cover value by four subplots.
2.5 Soil sample collection and nutrient analysis
Soil samples (0–20 cm layer) under shrub canopies and
in the meadow field were collected by soil auger for the
determination of nutrient content. From the center of
Table 1 Basic characteristics of Cerasus (C. trichostoma), Ribes
(R.glaciale), Rosa (R.omeiensis) and Salix (S.sphaeronymphe)
sampled on a high-altitude cutover, eastern Tibetan Plateau
Species Area (m
2
) Age (years) Height (cm)
Cerasus n22 22 22
Minimum 0.57 8 133.32
Maximum 11.10 19 375.80
Mean ± SE 3.51± 0.52 12.23 ± 0.54 246.89±12.33
Ribes n23 23 23
Minimum 0.56 7 124.82
Maximum 7.32 20 228.36
Mean ± SE 2.36± 0.32 12.91 ± 0.73 185.37±6.80
Rosa n21 21 21
Minimum 0.52 5 119.00
Maximum 9.94 20 290.27
Mean ± SE 3.05± 0.52 11.05± 0.88 202.95± 10.43
Salix n20 20 20
Minimum 0.61 7 140.71
Maximum 12.35 21 495.62
Mean ± SE 4.64± 0.93 12.79 ± 0.86 295.50±20.11
Shrub island effects on a high-altitude cutover 1129
Author's personal copy
each part, three to five sub-samples were combined on-
site into a composite sample for one shrub island. All
soil samples were air-dried, sieved through a 2-mm
mesh and analyzed for soil organic matter (SOM)
content by the Walkley–Black method, for total nitrogen
content (TN) by the Kjeldahl method, and for total
phosphorus content (TP) by Mo–Sb spectrophotometry
(Liu 1996).
2.6 Statistical analyses
Each microclimate variable for each plot of shrubs was
an average of five replicated measurements. Differences
in light transmission, air temperature and VPD among
Cerasus, Ribes, Rosa and Salix between each shrub-
area class and between the three area-classes of one
shrub species were tested separately by one-way
ANOVA and post hoc LSD tests. For each shrub
species, we performed a series of linear regression
analyses to determine the effects of shrub area on soil
nutrients (SOM, TN and TP), on the herbaceous
community (richness and cover of total herbaceous
plants, grasses and forbs; and herbaceous plant bio-
mass), on the litter (cover and biomass), and on woody
seedling recruitment (richness and number). All the
shrub area data was natural log-transformed to achieve
normality. Differences in the intercept and slope of
linear regression equations for one variable among
Ribes, Cerasus, Rosa and Salix were then tested by
analysis of covariance (generally known as ANCOVA)
and the shrub area, species and the evaluated parameter
were considered as the covariate, fixed factor and
dependent variables, respectively.
Mann–Whitney non-parametric tests were used to
identify differences in TP among Cerasus, Ribes, Rosa,
Salix and CK by considering all samples from the same
shrub species as replicates, because TP within the
canopies did not vary with patch size for each shrub
species. A t-test was used to determine differences in
richness and number of woody seedlings between Ribes
and CK, since both did not vary with Ribes shrub area. All
statistical analyses were carried out using SPSS 16.0 for
Windows (SPSS, Chicago, IL). Data of cover of total
herbaceous community, forbs, grasses and litter, and
herbaceous biomass were derived from the subplot level;
and data of richness of herbaceous plants, grasses and
forbs, and richness and number of woody seedlings were
derived from plots or shrub islands.
3 Results
3.1 Microclimate
The light transmission, ground air temperature and
VPD decreased with increasing shrub area for all four
species (Table 2); however, there were no significant
differences in ground air temperature and VPD among
Cerasus, Ribes and Rosa, regardless of size. Only Salix
islands (compared to the other species) had significantly
higher values of air temperature for the medium area class
andinVPDforthetwolargerarea classes. Salix islands
had significantly higher values for light transmission, in
the two smaller area classes, namely small and medium
class. In general, there were lower light transmission
intensities under Ribes islands compared with the other
three species.
3.2 Soil nutrients
Both SOM and TN beneath the shrubs were increased
significantly with increasing area for all shrub species
Table 2 Difference in light transmisssion, air temperature, and vapor
pressure deficit (VPD) for the three area classes among Cerasus,
Ribes, Rosa and Salix islands on a high-altitude cutover, eastern
Tibetan Plateau. n=4 for each area class of each shrub species. Means
± SE were are shown. Different uppercase letters indicate significant
differences among small, medium and large shrubs within the same
species, and different lowercase letters indicate significant differences
between four shrub species within the same area class (ANOVA, LSD
test, P<0.05)
Light transmission (%) Air temperature (°C) VPD (kPa)
Small
d
Medium Large Small Medium Large Small Medium Large
Cerasus 16.2 ±1.0 Ab 6.2 ± 0.8 Bb 3.2 ± 0.5 Cab 25.6 ± 1.5 Aa 21.22± 0.6 Bb 19.2± 0.3 Ca 1.63±0.17 Aa 0.93±0.10 Bab 0.69 ±0.07 Cb
Ribes 15.8± 1.2 Ab 5.5± 0.7 Bc 3.1 ± 0.5 Cb 24.8± 1.3 Aa 20.68±0.5 Bb 19.2±0.2 Ca 1.57 ±0.20 Aa 0.83 ± 0.08 Bb 0.66± 0.06 Cb
Rosa 16.0± 0.9 Ab 6.3±0.9 Bb 3.4±0.4 Cab 25.5 ±1.3 Aa 20.68 ± 0.5 Bb 19.4 ± 0.3 Ca 1.62 ± 0.12 Aa 0.91± 0.09 Bab 0.68± 0.07 Cb
Salix 17.7± 1.3 Aa 7.6 ±0.9 Ba 3.3±0.5 Ca 25.9± 1.3 Aa 22.36±0.8 Ba 20.0 ± 0.6 Ca 1.68± 0.16 Aa 1.03± 0.12 Ba 0.78± 0.10 Ca
d
Small, 0.5–1.0 m
2
; Medium, 2.0–3.0 m
2
; Large, 5.0–7.0 m
2
;
1130 Y. Wang et al.
Author's personal copy
evaluated (Fig. 1). The intercepts from the regression
functions among the four shrub species were significantly
different: Ribes > Cerasus = Rosa > Salix (Table 3). For the
small shrub area class, the SOM was highest for Ribes,
intermediate for Cerasus and Rosa, and lowest for Salix.
The slope values among Ribes, Cerasus and Rosa islands
did not differ, but were all higher than that of Salix
(Table 3), indicating that SOM and TN accumulation rates
under Salix islands were the smallest. TP content, however,
did not vary significantly with area for each shrub species;
and also no significant differences were seen among
Cerasus, Ribes, Rosa, Salix and CK.
3.3 Herb species richness, cover and biomass
In total, 124 herb species (13 grasses and 111 forbs)
belonging to 89 genera and 32 families were found
under all shrub islands investigated. Only 69 herb
species (10 grasses and 59 forbs) belonging to 53
genera and 25 families were recorded in the contrast
meadow, all of which were also found under the shrub
islands. There were 65 species (3 grasses and 62 forbs)
recorded solely beneath the shrub islands investigated
(Appendix 1). For all shrub species, the richness of total
herbs, grasses and forbs all gradually increased with shrub
area (Fig. 2). There were no significant differences in
either the intercepts or slopes of linear regressions
between the four shrub species for the richness of total
herbaceous plants, grasses and forbs, except that the
intercept for grass richness under Salix was lower than
for the other species (Table 3).
Total herb cover and biomass decreased markedly
with increasing shrub area (Fig. 3). The intercepts of the
linear regressions for both the herb cover and biomass
differed significantly among shrub species: Salix > Cerasus ≈
Rosa > Ribes (Table 3), suggesting that herb cover and
biomass were highest under Salix, intermediate under
Cerasus and Rosa, and lowest under Ribes in the small
shrub area class. The absolute slopes for the total herb
cover and biomass for Ribes were significantly higher
than those for Cerasus and Rosa, which were in turn
higher than for Salix (Table 3). Thus, the rate of decrease
in total herb cover and biomass under Ribes was highest,
intermediate under Cerasus and Rosa, and lowest under
Salix.
Furthermore, grass and forb cover responded differ-
ently to increasing shrub area: the former decreased
significantly and the latter increased (Fig. 3b,c). The
intercepts of the regressions for grass cover under
Cerasus, Rosa and Salix did not differ significantly, but
were clearly higher than that of Ribes (Table 3). Thus the
grass cover under Ribes was the lowest for the small area
4
6
8
10
12
100
150
200
250
300
Ribes
Cerasus
Rosa
Salix
Soil organic matter (g kg-1)
Cerasus
Ribes
Rosa
Salix
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Total phosphorus (g kg-1)
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Ln Shrub area(m2)
Ln Shrub area(m2)
Ln Shrub area(m2)
Cerasus
Ribes
Rosa
Salix
CK
CK
CK
Total nitrogen (g kg-1)
a
b
c
Fig. 1 Contents of asoil organic matter (SOM), btotal nitrogen (TN)
and ctotal phosphorus (TP) in relation to shrub area for Cerasus,
Ribes, Rosa and Salix. Regression lines of total phosphorus are not
shown because they did not vary with patch size for each shrub
species. Parallel dashed lines Mean ± SE of control (CK) (n= 35).
Details of the linear equations are provided in Table 3
Shrub island effects on a high-altitude cutover 1131
Author's personal copy
Table 3 Results of linear regression analysis between shrub area and variable of herbaceous
community, litter and woody seedling on a high-altitude cutover, eastern Tibetan Plateau. SOM
Soil organic matter, TN total nitrogen, TP total phosphorus, HR herbaceous richness, GR grass
richness, FR forb richness, HC herb cover, HB herb biomass, GC grass cover, FC forb cover,
LC litter cover, LB litter biomass, RWS richness of woody species, NWS number of woody
seedlings. Different uppercase or lowercase letters indicate significant difference in intercepts
or slopes among Cerasus, Ribes, Rosa and Salix (ANCOVA, P<0.05), respectively
Variable Cerasus (n=22) Ribes (n=23) Rosa (n=21) Salix (n=20)
Intercept Slope r
2
PIntercept Slope r
2
PIntercept Slope r
2
PIntercept Slope r
2
P
SOM 176.96B 33.68a 0.66 <0.0001 196.09A 42.87a 0.57 <0.0001 183.31B 30.72a 0.67 <0.0001 171.08C 14.26b 0.62 <0.0001
TN 7.35B 1.31a 0.48 0.0003 8.12A 1.50a 0.56 <0.0001 7.46B 1.16 a 0.64 <0.0001 6.88C 0.45b 0.44 0.0001
TP 1.15 0.02 0.01 0.6130 1.20 0.05 0.12 0.1100 1.13 0.02 0.02 0.5600 1.14 <−0.01 0.00 0.9800
HR 3.18A 0.33a 0.66 <0.0001 3.07A 0.41a 0.68 <0.0001 3.2A 0.30a 0.68 <0.0001 3.23A 0.31a 0.71 <0.0001
GR 1.39A 0.29a 0.29 0.0090 1.47A 0.25a 0.24 <0.0001 1.28A 0.30a 0.59 <0.0001 1.03B 0.33a 0.3 5 <0.0001
FR 2.98A 0.34a 0.53 0.0001 2.83A 0.45a 0.67 <0.0001 3.04A 0.30a 0.59 <0.0001 3.11A 0.31a 0.64 <0.0001
HC 84.2B 5.23ab 0.31 0.0069 78.88C 9.98a 0.76 <0.0001 82.98B 5.9b 0.59 <0.0001 85.2A 2.4c 0.55 0.0002
HB 5.06B 0.12b 0.53 0.0001 4..87C 0.22a 0.74 <0.0001 5.03B 0.14b 0.67 <0.0001 5.09A 0.07c 0.58 <0.0001
GC 49.94A 10.71b 0.76 <0.0001 41.11B 14.10a 0.77 <0.0001 48.43A 10.47b 0.77 <0.0001 49.74A 6.07c 0.57 <0.0001
FC 34.99C 4.85a 0.25 0.0178 38.27A 4.47a 0.24 0.0165 35.1BC 4.32a 0.28 0.0140 36.56B 3.51b 0.27 0.0179
LC 17.757A 6.7a 0.35 0.0037 18.93A 6.29a 0.31 0.0059 18.95A 5.24a 0.26 0.0017 12.64B 3.14b 0.37 0.0048
LB 3.07B 0.23a 0.48 <0.0001 3.17A 0.21a 0.61 <0.0001 3.11B 0.27a 0.63 <0.0001 2.88C 0.12b 0.31 0.0060
RWS 1.15B 0.23b 0.23 0.0230 1.58 0.21 0.12 0.1126 1.15B 0.27b 0.35 0.0048 1.28A 0.85a 0.96 <0.0001
NWS 1.81A 1.05b 0.29 0.0100 2.15 0.46 0.10 0.1400 1.80A 1.21b 0.41 0.0017 1.41A 2.11a 0.71 <0.0001
1132 Y. Wang et al.
Author's personal copy
class. The species obviously differed in their absolute
slopes: Salix > Cerasus ≈Rosa > Ribes (Table 3),
demonstrating that the decreased rate of grass cover in
relation to shrub canopy area was highest for Ribes,
intermediate for Cerasus and Rosa, and lowest for Salix.
As for forb cover, the intercept of the function was
greatest under Ribes of the four species (Table 3). The
slope of the regression for Salix was less than for Cerasus,
Ribes and Rosa (Table 3), indicating that the increased
rate of forb cover in relation to shrub canopy area under
Salix was the lowest.
3.4 Litter mass and cover
The litter cover and mass increased gradually with area for
all four shrub species (Fig. 4). The intercept and the slope
of the linear regression for litter cover under Salix were
lower than those under Cerasus, Ribes and Rosa islands
(Table 3), suggesting that litter cover under Salix was
lowest beneath the small shrub area, and that its accumu-
lation rate with increased shrub area was also lower. For
litter mass, the intercepts differed significantly: Ribes >
Rosa ≈Cerasus > Salix (Table 3), indicating that Ribes had
the highest litter mass and Salix the lowest under the small
area of the four species. However, the slope under Salix
was lower than for Cerasus, Ribes and Rosa, indicating that
the accumulation rate of litter mass with increased shrub
area was lowest under Salix.
3.5 Woody seedling recruitment
In total, 17 woody plant species seedlings belonging to
eight genera and five families were found under the 86
shrub islands investigated. However, only four woody plant
species seedlings (Rubus pungens,Rosa omeiensis,Spiraea
omeiensis and Salix cupularis)wererecordedinthe
meadow field (Appendix 2). There were 13, 11, 9 and 15
shrub species identified under the Cerasus, Ribes, Rosa and
Salix islands, respectively. No tree seedlings were recorded
under shrub canopies or in the meadow field. Moreover, 13
out of the 17 woody plant species were vertebrate-dispersed
(Appendix 2).
The species richness and number of woody seedlings
correlated positively with shrub area for all shrub species
except Ribes (Fig. 5; Table 3). The intercept and slope of
the linear function for the richness of woody seedlings were
higher under Salix islands than Cerasus and Rosa (Table 3).
This suggested that woody seedling richness under Salix
Ln Herbaceous richness
2.0
2.5
3.0
3.5
4.0
4.5
Ribes
Cerasus
Rosa
Salix
Ln Grass richness
0.5
1.0
1.5
2.0
2.5
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Forb richness
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Cerasus
Ribes Rosa
Salix
Cerasus
Ribes
Rosa
Salix
Cerasus
Ribes
Rosa
Salix
Ln Shrub area(m2)
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Shrub area(m2)
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Shrub area(m2)
a
b
c
Fig. 2 Species richness of aherbs, bgrasses and cforbs in relation to
shrub area for Cerasus, Ribes, Rosa and Salix. Details of the linear
equations are given in Table 3
Shrub island effects on a high-altitude cutover 1133
Author's personal copy
islands was the greatest of the four species for the small
area class, and that the accumulation rate was the highest.
For the number of woody seedlings, the intercepts showed
no difference among the four species (Fig. 5, Table 3);
however, the slope for Salix was significantly higher than
for Cerasus and Rosa (Table 3). ANCOVA test indicated
that all these slopes were noticeably higher than CK (P<0.05,
for all cases), suggesting that the rates of increase in the
number of woody seedlings in regard to shrub area under
these three shrub species were higher than for shrub-free
meadow (CK). The richness and number of woody seedlings
for Ribes were also greater than in CK (by t-tests, t=7.258,
P=0.001; t=4.331, P=0.017, respectively).
4 Discussion
4.1 Microhabitat effects
Our results show clearly that existing shrub significant-
ly reduces light transmission, air temperature and VPD
within islands; and that shrub development provided a
microclimate gradient according to shrub size (Table 2).
These findings in a high-altitude cutover agree with other
observations in low-altitude semi-arid and arid regions
(Endo et al. 2008). These results can be attributed mainly
to variation in foliation levels related to age-dependent
size for individual shrub species, since the shrub shading
40
50
60
70
80
90
Ribes
Cerasus
Rosa
Salix
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
Cerasus
Ribes
Rosa
Salix
Cerasus
Ribes
Rosa
Salix
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
10
20
30
40
50
60
Salix
Ribes
Rosa
Cerasus
10
20
30
40
50
60
70
Rosa
Cerasus
Salix
Ribes
Herbaceous cover (%)
Grass cover (%)
Forb cover (%)
Ln(Herbaceous biomass(g m
-2
)
Ln Shrub area (m
2
)
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Shrub area (m
2
)
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Shrub area (m
2
)
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Shrub area (m
2
)
CK
CK
CK
CK
ab
cd
Fig. 3 a Herbaceous cover, bgrass cover, cforb cover and dherbaceous biomass in relation to shrub area for Cerasus, Ribes, Rosa and Salix.
Parallel dashed lines Mean ± SE of CK (n=70). Details of the linear equations are given in Table 3
1134 Y. Wang et al.
Author's personal copy
level is positively related to foliation level (Reisman-
Berman 2007) and a shaded microenvironment reduces
ground air temperature in summer (Suzán et al. 1996;
Shumway 2000). Moreover, the lower temperature and
solar radiation would lead to lower VPD (Franco and
Nobel 1989) as occurred in the present study (Table 2).
We also found a significant difference in microclimate
effect among the shrub species (Ribes, Cerasus, Rosa and
Salix). The difference in branching angle of crown
architecture among shrub species may reasonably explain
this. Salix has a higher branch-angle (and low foliation
level) and is thus easier for sunlight to penetrate than the
other shrubs. Consequently, beneath Salix islands there
was higher light transmission, air temperature and VPD
than beneath Ribes (Table 2). Microclimate amelioration,
especially the reduction in light transmission and VPD
would be very important in high-altitude regions, since
mortality of transplanted spruce seedlings in this region is
attributed mainly to desiccation and heat stress (Wang et
al. 1995). Therefore, the four investigated shrubs of the
high-altitude cutovers could be promising nurse plants for
target tree seedling establishment and growth, supporting
previous results from spruce seedling reforestation exper-
iment (Wang et al. 1995;Bao2004).
Shrub islands on the high-altitude cutover signifi-
cantly improved SOM and TN contents, consistent with
results from low-elevation regions (Zhao et al. 2007).
Furthermore, SOM and TN contents beneath shrub
canopies were significantly and positively related to
shrub area, supporting the prediction of Maestre and
Cortina (2005). Shrubs can trap plant detritus more
effectively, both without and within a canopy, than the
Litter cover(%)
0
10
20
30
40
Ribes
Cerasus
Rosa
Salix
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln(Litter biomass(g m-2)
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Ln Shrub area(m2)
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
Ln Shrub area(m2)
Rosa
Salix
Cerasus
Ribes
CK
Rosa
Salix
Cerasus
Ribes
CK
b
a
Fig. 4 a Litter cover and blitter biomass in relation to shrub area for
Cerasus, Ribes, Rosa and Salix. Parallel dashed lines Mean ± SE of
CK (n=70). Details of the linear equations are given in Table 3
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
0
2
4
6
8
Sqrt(Woody species richness+1)
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ribes
Cerasus
Rosa
Salix
Ln Shrub area(m2)
Ln Shrub area(m2)
Rosa
Salix
Cerasus
Ribes
Rosa
Salix
Cerasus
Ribes
Sqrt(Number of woody seedlings+!)
CK
b
a
Fig. 5 a Richness of woody seedlings species, and bnumber of
woody seedlings in relation to shrub area for Cerasus, Ribes, Rosa and
Salix. Details of the linear equations are given in Table 3
Shrub island effects on a high-altitude cutover 1135
Author's personal copy
herb community in a meadow field (Zhao et al. 2007),
resulting in litter accumulation under shrub canopies
(Fig. 4). Differences in crown architecture may also be
responsible for the differences in SOM and TN among
shrub species. For instance, Li et al. (2007) found that soil
nutrients under Ta m a r i x spp. with their hemispheroidal
crowns were significantly higher than those under
Haloxylon ammodendron with their Y-shaped crowns,
suggesting that Y-shaped crowns were less capable of
capturing and maintaining the litter under them than
hemispheroidal crowns. Consistent with this, our results
showed that there was the highest amount of litter mass
accumulation beneath Ribes with its low branching and
hemispheroidal crown, medium under Cerasus and Rosa
with their medium branching, and the lowest beneath
Salix with its high branching (Fig. 4). This resulted in the
highest SOM and TN under Ribes, medium under Cerasus
and Rosa, and the lowest under Salix (Fig. 1a,b). However,
there was no difference in TP content beneath the different
shrubs, and between the shrubs and contrasts (Fig. 1c). This
is in contrast to the results of Zhao et al. (2007), who found
that TP content within a shrub canopy was higher than
outside, and that there were significant differences between
two species of different canopy shapes. Surprisingly, we
found that TP content also did not vary with shrub area for
all four shrubs evaluated. Further investigation will be
required to explain this result.
4.2 Herb community effects
We found that greater herb species richness existed
beneath shrubs, and many species were associated
exclusively with the shrubs. Furthermore, the richness
of the herb species increased significantly and positively
with shrub area for all four species examined (Fig. 2).
Most notably, we found that grass cover was reduced
significantly with increasing shrub size, while forb cover
increased (Fig. 3b,c), suggesting a shift from grass- to
forb-domination under shrubs as area-size increased.
Similar relationships have been found for Mediterranean
shrubs such as Ephedra fragillis and Quercus coccifera
(Maestre and Cortina 2005). We should note that the
different effects of shrub on herb development with the
increasing area are correlated closely to shrub species
identity, resulting in stronger effects for Salix than other
three species (Table 3).
The herb community effects can be related closely to
microhabitat effects. Throop and Archer (2008)sug-
gested that there are more complex spatial patterns of
microclimate under larger shrubs. Such microhabitat
heterogeneity could promote differentiation of niches,
which increase in number and availability as shrub size
increases, leading to establishment of more species
(Pugnaire et al. 1996). In present study, shrubs reduced
light transmission, so more light-intolerant herb species
could occur under larger shrubs with lower light trans-
mission. The effects may also be ascribed to the different
responses of forbs and grasses to microhabitat alternation
with increasing shrub area according to their adaptation
strategies. Forbs generally have higher nitrogen and water
requirements than grasses, so the increased availability of
these resources may have a greater positive effect on forbs
than on grasses, and could enhance individual perfor-
mance, and result in greater population size and/or
biomass production (Turner and Knapp 1996). Moreover,
low light conditions, especially in the early growing
season, may be less limiting to forbs than grasses, because
photosynthesis in forbs saturates with respect to light at a
much lower photosynthetic photon flux density (Turner
and Knapp 1996). The increase in soil nitrogen and
moisture, along with the reduction of available light may
thus favor forbs and/or reduce the competitive capability
of grasses (Seastedt et al. 1991; Wedin and Tilman 1993).
In addition to plant adaptation strategy, the herb commu-
nity effects of shrubs can be explained according to the
indirect facilitation mechanism (Levine 1999;Brookeret
al. 2008;Cuestaetal.2010). The occurrence of shrub
species may also convert competition between grasses and
forbs into indirect facilitation. Levine (1999) proposed
that apparent indirect facilitation may be more likely in
assemblages where the different pairs of competitors have
significantly different mechanisms to acquire resources.
Cuesta et al (2010) provides testimony that the nurse
shrub indirectly facilitated seedling growth by reducing
the competitive capacity of herbs. Shrubs, forbs and
grasses are quite different growth entities, and thus would
differ in their pathways to acquire resources; consequently
shrubs produced a strong indirect facilitative effect on
forbs by suppressing the competing grasses, resulting in a
shift from grass- to forb-domination composition with
increasing shrub island area.
4.3 Effects on woody seedling recruitment
Shrubs facilitate woody seedling recruitment by enhancing
the richness of shrub species and increasing the number of
shrub seedlings (Fig. 5). Such results can be attributed to
the shrub function of the recruitment foci. Shrubs can
attract seed-dispersing animals by providing perches, fruits,
shade and nesting sites, thus increasing the number and
diversity of seeds arriving under their canopies (Duarte et
al. 2006). Consequently, the richness and number of
seedlings under the shrubs were obviously higher than in
the open meadow, and were positively related to shrub area
1136 Y. Wang et al.
Author's personal copy
for all evaluated shrub species except for Ribes (Fig. 5).
Shrubs can also improve the number and richness of woody
plant seedlings indirectly by reducing the competition
intensity from herbs (Cuesta et al. 2010). In our present
work, the decreasing cover and biomass of the herbaceous
community meant less resource utilization (e.g., amount of
soil nutrients) with increasing shrub area (Fig. 3); therefore,
better resource availability probably promoted more rich-
ness and a greater number of woody seedlings under larger
shrubs (Fig. 5). Kunstler et al (2006) found that the
promotion of woody seedling establishment beneath shrubs
can also be due to indirect facilitation by shade and
competition release from herbs, supporting the above
speculation.
Slocum (2001) suggested that woody species with
fleshy fruits will probably attract more seed dispersers
than those with dry fruits, resulting in favorite woody
seedling renewal. However, Salix with dry fruits gave rise
to a greater richness and number of woody seedlings than
the other three species with fleshy fruits (Fig. 5), indicat-
ing that barriers to seedling establishment after seed
dispersal were also probably working. Thus, litter may
be responsible for the different effects on woody seedling
recruitment, as it can inhibit seed germination and
seedling establishment of woody species (Xiong and
Nilsson 1999). In present work, beneath Ribes presented
greater litter cover and biomass but lower richness and
number of woody seedlings; conversely, Salix has smaller
litter cover and biomass, but higher richness and amount
of shrub seedlings (Table 3).
Our results also displayed the difference in the facilita-
tion effects from four shrub species for herbaceous species
and woody seedlings. The ranking in species effect on the
slopes of increasing effects with area was different for the
herbaceous species and the woody seedlings (Table 3).
Indeed, for herbaceous species, the slope of Salix was lower
than that of the three other species (lower effect of Salix),
whereas for woody seedlings the slope of Salix was higher
(higher effect of Salix) (Table 3). Possible explanations
were the shade-tolerance difference for herbaceous species
and woody seedlings, and the microhabitat resource
difference beneath various shrub islands. Forb species,
which account for the major proportion of total herbaceous
species, have higher shade-tolerance than woody seedling
species, which are composed of early-succession shrubs
such as several species from Rubus and Lonicera genera
(Appendix 2). The indirect positive effect of the shrubs on
both forbs and woody seedlings is due to the competition
release of the grasses mentioned above. However, the
facilitative effects can be modulated by changing spatio-
temporal resources (Soliveres et al. 2010; García-Palacios
et al. 2011), and these effects can therefore vary with
increasing shade of the shrubs (with area) from Salix to
Ribes (Table 3; Figs. 2,5). We presume there is a threshold
of shade above which the effect switches from positive to
negative, and that this threshold is earlier for woody
seedlings than for forbs, likely because woody seedlings
are less shade-tolerant than forbs. Thus, the shade of Salix
is at the threshold of this switch for woody seedlings but
not for forbs. Although some studies indirectly support this
standpoint (Kunstler et al. 2006; Brooker et al. 2008;
Cuesta et al. 2010), further research to obtain direct
evidence with which to explore this question still is
required.
5 Conclusion
There were clear effects of shrub islands on microcli-
mate (light transmission, air temperature and VPD), soil
nutrients (SOM and TN), herbaceous community (spe-
cies richness, cover and biomass) and woody seedling
recruitment (richness and number) in alpine forest
cutovers. These effects were greater for large shrubs
than small ones, and also varied among shrubs with
different crown architectures, with the result that shrub
effects were species- and size-dependent. We also found
that differences in species-dependency of the shrub
effect for the responses of the herbaceous (grasses and
forbs) and woody seedling species, suggested that
shrubs also indirectly facilitate forbs and seedling
regeneration through competition release of grasses.
Thus, the established shrubs can improve hash micro-
habitats, alter the properties of the communities beneath,
facilitate light-tolerant plant diversity development, and
accelerate the natural succession process from the herb
stage to the shrub stage in patches on cutovers in the
high-altitude region of the eastern Tibetan Plateau.
Our results, which will be of use to inform practices in
alpine forest restoration and cutover management, suggest
that shrub preservation and their utilization as nurse plants
for reforestation should be applied to cutover vegetation
management prescription in alpine regions. Also, it is better
to select large shrubs like Salix as nurse microhabitats for
reforestation because they have stronger positive effects on
woody seedling nursing.
Acknowledgments This work was funded by a grant from the
National Natural Science Foundation of China (No.30570333,
30972350) to W.K.B. We greatly thank D. Yang, Z.H. Se, B. Cheng,
C. Wang and X Liu. for their help in the field, and the Key Laboratory
of Mountain Ecological Restoration and Biological Resources
Utilization of Chinese Academy of Sciences for logistical support.
Two anonymous reviewers provided valuable suggestions and com-
ments that improved our explanation of the results.
Shrub island effects on a high-altitude cutover 1137
Author's personal copy
Appendix 1
Table 4 List of herb species identified beneath shrub islands on the
clear-cut sites in Rangtang, eastern Tibetan Plateau. A “+”indicates
that the species was found in the Cerasus (C. trichostoma), Ribes (R.
glaciale), Rosa (R.omeiensis) and Salix (S.sphaeronymphe)orCK
(open meadow field)
Species Family Cerasus island Ribes island Rosa Island Salix Island CK (Meadow)
Aconitum liljestrandii Ranunculaceae + + +
Adenophora liliifolioides Campanulaceae + + +
Agrostis rupestris Gramineae + +
Agrostis perlaxa Gramineae + + + +
Ajania tenuifolia Compositae + + + +
Ajuga ciliata Labiatae + + + +
Ajuga ciliata Bunge var. hirta Labiatae + + +
Allium prattii Amaryllidaceae + + + +
Anaphalis lactea Compositae + + +
Anemone cathayensis Ranunculaceae + + + + +
Anemone rivularis Ranunculaceae + + +
Anemone rivularis Ranunculaceae + + + + +
Aquilegia ecalcarata Ranunculaceae + + + +
Artemisia globosoides Compositae + + +
Artemisia lancea Van Compositae + + +
Aster diplostephioides Compositae + + + + +
Astragalus membranaceus Leguminosae + + +
Athyrium dentigerum Athyriaceae + + + +
Bromus epilis Keng Gramineae + + + +
Bromus japonicus Gramineae + +
Caltha palustris Ranunculaceae + + +
Cardamine tangutorum Brassicaceae + + +
Carex lehmaii Cyperaceae + + + + +
Carex tristachya Cyperaceae + + + +
Cerastium fontanum Caryophyllaceae + + + +
Chrysosplenium griffithii Saxifragaceae + + +
Circaea alpina Onagraceae + + + +
Circaeaster agrestis Ranunculaceae + + + +
Clematis florida Ranunculaceae + + +
Codonopsis pilosula. Campanulaceae + + + + +
Corydalis impatiens Papaveraceae + + + +
Corydalis linarioides Papaveraceae + + + +
Corydalis curviflora Papaveraceae + + +
Corydalis edulis Papaveraceae + + +
Corydalis laucheana Papaveraceae + +
Cystopteris moupinensis Athyriaceae + + + + +
Delphinium potaninii Ranunculaceae + + +
Deschampsia sp. Gramineae + + + +
Deyeuxia arundinacea Gramineae + + +
Deyeuxia scabrescens Gramineae + + + + +
Doronicum thibetanum Compositae + +
Draba borealis Brassicaceae + + + +
Dracocephalum heterophyllum Labiatae + + +
Elsholtzia ciliata Labiatae +
1138 Y. Wang et al.
Author's personal copy
Table 4 (continued)
Species Family Cerasus island Ribes island Rosa Island Salix Island CK (Meadow)
Elymus tangutorum Gramineae + + + + +
Elymus nutans Gramineae + + + +
Epilobium angustifolium Onagraceae + + + + +
Epilobium palustre Onagraceae + + +
Epilobium tibetanum Onagraceae + + + +
Equisetum arvense Equisetaceae + + + +
Euphrasia Tenore Scrophulariaceae + + + +
Festuca ovina Gramineae + + +
Fragaria orientalis Rosaceae + + +
Galium paradoxum Rubiaceae + + + + +
Galium trifidum Rubiaceae + + +
Gentiana syringea Gentianaceae + + + + +
Gentianopsis paludosa Gentianaceae + + + +
Geranium pylzowianum Geraniaceae + + + + +
Geum aleppicum Rosaceae + + +
Halenia elliptica Gentianaceae + + +
Heteropappus altaicus Compositae + + + +
Impatiens noli-tangere Balsaminaceae + + +
Inula linariifolia Compositae + + + +
Inula japonica Compositae + + + + +
Juncus allioides Juncaceae + + + + +
Juncus potaninii Juncaceae + + + +
Leontopodium haplophylloides Compositae + + + +
Ligularia sagitta Compositae + + + +
Melica przewalskyi Gramineae + + +
Microula younghusbandii Boraginaceae + + +
Microula trichocarpa Boraginaceae + + + +
Microula turbinata Boraginaceae + + + +
Notholirion bulbuliferum Liliaceae + + + + +
Notopterygium incisum Umbelliferae + + + +
Parasenecio deltophyllus Compositae + + +
Pedicularis chenocephala Scrophulariaceae + + + +
Pedicularis kansuensis Scrophulariaceae + + + +
Pedicularis rudis Scrophulariaceae + + + +
Pilea racemosa Urticaceae +
Plantago major. Plantaginaceae
Poa annua Gramineae + + + + +
Poa elanata Gramineae + + + +
Poa malaca Gramineae + + + + +
Polygonatum verticillatum Liliaceae + + + +
Polygonum polystachyum Polygonaceae + + +
Polygonum sparsipilosum Polygonaceae + + + +
Polygonum viviparum Polygonaceae + + + + +
Polystichum shensiense Dryopteridaceae + + +
Primula polyneura Primulaceae + + + + +
Pseudostellaria sylvatica Caryophyllaceae + + + +
Pternopetalum longicaule Umbelliferae + + +
Ranunculus tanguticus Ranunculaceae + +
Rheum palmatum Polygonaceae + +
Shrub island effects on a high-altitude cutover 1139
Author's personal copy
Appendix 2
Table 4 (continued)
Species Family Cerasus island Ribes island Rosa Island Salix Island CK (Meadow)
Rhodiola kirilowii Crassulaceae + + + + +
Rhodiola quadrifida Crassulaceae + + +
Rhodiola eurycarpa Crassulaceae + + +
Roegneria parvigluma Gramineae + + +
Roegneria kamoji Gramineae + + + +
Rumex crispus Polygonaceae + + +
Salvia prattii Labiatae + + + +
Sanicula chinensis Umbelliferae + + +
Saussurea cana Compositae + + + +
Saussurea japonica Compositae + + + +
Saxifraga egregia Saxifragaceae + + +
Scutellaria baicalensis Labiatae + + +
Sibbaldia tenuis Rosaceae + + + +
Sinocarum coloratum Umbelliferae + + + +
Souliea vaginata Ranunculaceae + + +
Stipa penicillata Gramineae + + + +
Taraxacum maurocarpum Compositae + + + +
Thalictrum przewalskiii. Ranunculaceae + + +
Trigonotis peduncularis Boraginaceae + + + +
Trigonotis tibetica Boraginaceae + + + +
Trisetum clarkei Gramineae + + + + +
Urtica fissa Urticaceae + + + +
Valeriana officinalis Valerianaceae + + + +
Valeriana tangutica Valerianaceae + + +
Veronica eriogyne Scrophulariaceae + + + + +
Veronica szechuanica Scrophulariaceae + + + +
Veronica didyma Scrophulariaceae + + + +
Veronica rockii Scrophulariaceae + + + +
Viola biflora Violaceae + + + +
Viola rockiana Violaceae + + + + +
Table 5 Species list of woody seedlings identified beneath shrub
islands on clear-cut sites in the eastern Tibetan Plateau. A “+”
indicates that the species was found in Cerasus (Cerasus trichostoma),
Ribes (Ribes glaciale), Rosa (Rosa omeiensis)orSalix(Salix
sphaeronymphe) or CK (open meadow field)
Species Family Cerasus Ribes Rosa Salix CK
Berberis diaphana
a
Berberidaceae + +
Cerasus trichostoma Rosaceae + +
Lonicera hemsleyana Caprifoliaceae + + + +
Lonicera hispida Caprifoliaceae + + + +
Lonicera webbiana Caprifoliaceae + +
Potentilla glabra
a
Rosaceae + + + +
Ribes glaciale Saxifragaceae + + + +
Ribes meyeri Saxifragaceae + + +
1140 Y. Wang et al.
Author's personal copy
References
Bao WK (2004) Environmental degradation on forest logging blank,
vegetation restoration, and management in Tibetan Plateau. In:
Zheng D, Yao TD (eds) Uplifting of Tibetan Plateau with its
environmental effects. Science, Beijing, pp 537–542
Brooker RW, Maestre FT, Callaway RM, Lortie CL, Cavieres LA,
Kunstler G, Liancourt P, Tielbörger K, Travis JMJ, Anthelme F,
Armas C, Coll L, Corcket E, Delzon S, Forey E, Kikvidze Z,
Olofsson J, Pugnaire F, Quiroz CL, Saccone P, Schiffers K, Seifan
M, Touzard B, Michalet R (2008) Facilitation in plant commu-
nities: the past, the present, and the future. J Ecol 96:18–34
Cuesta B, Villar-Salvador P, Puértolas J, Benayas JM, Michalet R
(2010) Facilitation of Quercus ilex in Mediterranean shrubland is
explained by both direct and indirect interactions mediated by
herbs. J Ecol 98:687–696
Dona AJ, Galen C (2007) Nurse effects of alpine willows (Salix)
enhance over-winter survival at the upper range limit of fireweed
Chamerion angustifolium. Arc Antarct Alp Res 39:57–64
Duarte LDS, Dos-Santos MMG, Hartz SM, Pillar VD (2006) Role of
nurse plants in Araucaria forest expansion over grassland in south
Brazil. Austr Ecol 31:520–528
Endo M, Yamamura Y, Tanaka A, Nakano T, Yasuda T (2008) Nurse-
plant effects of a dwarf shrub on the establishment of tree
seedlings in a volcanic desert on Mt, Fuji, central Japan. Arct
Antarct Alp Res 40:335–342
Franco AC, Nobel PS (1989) Effect of nurse plants on the
microhabitat and growth of cacti. J Ecol 77:870–886
García-Palacios P, Maestre FT, Gallardo A (2011) Soil nutrient heteroge-
neity modulates ecosystem responses to changes in the identity and
richness of plant functional groups. J Ecol 99:551–562
Garner W, Steinberger Y (1989) A proposed mechanism for the formation
of fertile in the desert ecosystem. J Arid Environ 16:257–262
Jones HG (1992) Plants and microclimate, a quantitative approach to
environmental plant physiology, 2nd edn. Cambridge University
Press, Cambridge, UK
Körner C (2003) Alpine plant life—function plant ecology of high
moutain ecosystems. Springer, Heidelberg
Kunstler G, Curt T, Bouchaud M, Lepart J (2006) Indirect facilitation
and competition in tree species colonization of sub-
Mediterranean grasslands. J Veg Sci 17:379–388
Levine JM (1999) Indirect facilitation: evidence and predictions from
a riparian community. Ecology 80:1762–1769
Li J, Zhao C, Zhu H, Li Y, Wang F (2007) Effect of plant species on
shrub fertile isalnd at an oasis-desert ecotone in the South
Junggar Basin, China. J Arid Environ 71:350–361
Liu GS (1996) Soil physical and chemical analys and description of
soil profiles. Standards Press of China, Beijing, China
Maestre FT, Cortina J (2005) Remnant shrubs in Mediterranean semi-
arid steppes: effects of shrub size, abiotic factors and species
identity on understorey richness and occurrence. Acta Oecol
27:161–169
Mattson KDM, Putz FE (2008) Sand pine (Pinus clausa) seedling
distribution and biomechanics in relation to microsite conditions
and proximity to potential nurse plants. For Ecol Manag
255:3778–3782
Pugnaire FI, Haase P, Puigdefábregas J, Cueto M, Clark SC, Incoll LD
(1996) Facilitation and succession under the canopy of a
leguminous shrub, Retama sphaerocarpa, in a semi-arid envi-
ronment in south-east Spain. Oikos 76:455–464
Reisman-Berman O (2007) Age-related change in canopy traits shifts
conspecific facilitation to interference in a semi-arid shrubland.
Ecography 30:459–470
Seastedt TR, Briggs JM, Gibson DJ (1991) Controls of nitrogen
limitation in tallgrass prairie. Oecologia 87:72–79
Shumway SW (2000) Facilitative effects of a sand dune shrub on species
growing beneath the shrub canopy. Oecologia 124:138–148
Slocum MG (2001) How tree species differ as recruitment foci in a
tropical pasture. Ecology 82:2547–2559
Soliveres S, DeSoto L, Maestre FT, Olano JM (2010) Spatio-temporal
heterogeneity in abiotic factors can modulate multiple ontoge-
netic shifts between competition and facilitation. Persp Plant Ecol
Evol Syst 12:227–234
Suzán H, Nabhan GP, Patten DT (1996) The importance of Olneya tesota
as a nurse plant in the Sonoran Desert. J Veg Sci 7:635–644
Throop HL, Archer SR (2008) Shrub (Prosopis velutina) encroachment
in a semidesert grassland: spatial-temporal changes in soil organic
carbon and nitrogen pools. Global Change Biol 14:2420–2431
Turner CL, Knapp AK (1996) Responses of a C4 grass and three C3
forbs to variation in nitrogen and light in tallgrass prairie.
Ecology 77:1738–1749
Wang JX, Xu JD, Hou GW, Liu JB, Pu JT, Shi LX, Liao GY (1995)
Ecology and regeneration of cut-overs in high-altitude and
plateau region of the upper reach of Yangtze river. Chinese
Publication House of Forestry, Beijing
Wedin D, Tilman D (1993) Competition among grasses along a
nitrogen gradient—initial conditions and mechansims of compe-
tition. Ecol Monogr 63:199–229
Xiong SJ, Nilsson C (1999) The effects of plant litter on vegetation: a
meta-analysis. J Ecol 87:984–994
Zhao HL, Zhou RL, Su YZ, Zhang H, Zhao LY, Drake S (2007) Shrub
facilitation of desert land restoration in the Horqin sand land of
Inner Mongolia. Ecol Eng 31:1–8
Table 5 (continued)
Species Family Cerasus Ribes Rosa Salix CK
Ribes tenue Saxifragaceae + + +
Ribes alpestre Saxifragaceae + +
Ribes himalense Saxifragaceae + + +
Rosa omeiensis Rosaceae + + + +
Rosa sericea Rosaceae + +
Rubus lutescens Rosaceae + +
Rubus pungens Rosaceae + + + + +
Salix sphaeronymphe
a
Salicaceae + +
Spiraea omeiensis
a
Rosaceae + + + +
a
Not vetebrate-dispersed
Shrub island effects on a high-altitude cutover 1141
Author's personal copy
