全 文 ：佛山地区存留自然林和存留人工林的土壤无机磷
组分
侯恩庆1,2,3, CHEN Cheng-rong2, 黎建力4, 左伟东4, 王瑜1,3, 汪学金1,3,
温达志1*
(1. 中国科学院华南植物园，广州 510650； 2. Environmental Futures Centre, Griffith School of Environment, Griffith University, Brisbane 4111,
Australia； 3. 中国科学院研究生院， 北京 100049； 4. 佛山市南海区农林技术推广中心，广东 佛山 528222)
摘要： 为探讨热带亚热带森林，尤其城市及其周边地区残存森林土壤磷的有效性，对佛山地区 14 个残存林(7 个自然林和 7 个
人工林）的 0~3 cm 和 3~23 cm 矿质土壤的 P 有效性进行研究。结果表明，铁结合态无机 P 和还原剂可溶解无机 P 是土壤无机
P 的主要组分。在 0~3 cm 矿质层中，自然林土壤铝结合态无机 P、Bray 1 提取无机 P 和总无机 P 含量显著高于人工林；而在
3~23 cm 矿质土层中，自然林土壤钙结合态无机 P 含量显著高于人工林。其它土壤营养指标在自然林和人工林间差异不显著。
相关分析结果表明，土壤有机质含量与钙结合态无机 P 除外的其它无机 P 组分含量均成显著正相关。聚类分析结果表明 14
个残存林土壤 P 有效性可分成 3 组，整体上人工林土壤 P 有效性比自然林低。这有助于认识城市化影响下城市及其周边地区
残存森林土壤营养状况及加强养分管理。
关键词： 无机磷组分； 残存林； 磷有效性； 佛山地区
doi: 10.3969/j.issn.1005–3395.2012.06.002
Soil Inorganic Phosphorus Fractions of Remnant Native and Plantation
Forests in Foshan Region
HOU En-qing1,2,3, CHEN Cheng-rong2, LI Jian-li4, ZUO Wei-dong4, WANG Yu1,3, WANG Xue-jing1,3,
WEN Da-zhi1*
(1. South China Botanical Garden, Chinese Academy of Sciences, Guangzhou 510650, China; 2. Environmental Futures Centre, Griffith School
of Environment, Griffith University, Brisbane 4111, Australia; 3. Graduate University of Chinese Academy of Sciences, Beijing 100049, China; 4.
Agriculture and Forestry Technology Extension Centre of Nanhai District in Foshan, Foshan 528222, China)
Abstract: Soil phosphorous (P) availability is always thought to be a limiting factor for ecosystem primary
productivity in the highly weathered tropical and subtropical areas, but our knowledge about soil P availability
in tropical and subtropical forests is still poor, particularly in remnant native forests and frequently disturbed
plantation forests in the urban and suburban areas. The soils at 0–3 cm and 3–23 cm mineral depths from 14 forest
patches in Foshan region (seven native forest patches and seven plantation forest patches) were collected, and pH
in water and concentrations of organic carbon (C), total nitrogen (N), total P, Bray-1 Pi and sequential extractable
inorganic P fractions to estimate soil P availability were determined. The results showed that Fe-Pi and reductant
soluble inorganic P were major inorganic P fractions at the study forest patches. Most of the selected soil nutrient
Recieved： 2012–03–12　　　　Accepted： 2012–05–07
Supported by National Natural Science Foundation of China (No. 31070409) and the Agricultural and Forestry Promotion Fund of Nanhai Agro-
forestry Extension Centre, Guangdong Province (No. 084101001).
HOU En-qing (1986~ ). E-mail: houeq@scib.ac.cn
* Corresponding author. E-mail: dzwen@scib.ac.cn
热带亚热带植物学报　2012， 20(6)： 546~554
Journal of Tropical and Subtropical Botany
第6期 547
Phosphorus (P) is considered as one of the most
common nutritional constraints for forest ecosystem
productivity, especially in the strongly weathered
tropical and subtropical regions[1–3]. In subtropical
China, soil P availability is generally low[4], as soils in
this region are developed on the strongly weathered
parent materials [5–6]. Moreover, enhanced soil
acidification by continous acid deposition may further
reduce the soil P availability by its precipitation with
free iron (Fe) and aluminum (Al)[3,7]. Phosphorus has
been thought to be one of the major limiting nutrients
for forest growth in this region according to previous
studies which showed high ratio of nitrogen (N)
to P in leaf or low total P and available P contents
in soils[8–9]. Therefore, both scientists and forest
managers have become increasingly concerned about
the potential effects of the P limitation on the regional
forest growth in recent years[8–10].
Soil P availability exists as a continuum, and
different fractions of soil P are inter-changeable and
reach an equilibrium state eventually[11–12]. Different
fractions of P might be available to plants at different
time scales[11]. Separating total P into fractions of
different solubility was found to be helpful to improve
our understanding about the P supply in forest soils
both in short and long terms[11,13]. The inorganic P
sequential extraction of Chang and Jackson (1957)
and its modified methods have been widely used
to investigate the forms of inorganic P and soil P
availability in the past decades[13–15]. Inorganic P
was generally divided into four fractions by these
methods, including P fractions that associated with
soluble Al, Fe and calcium (Ca), and P fraction that
occluded by free oxide coatings[14,16]. These methods
are considered to be useful to investigate the form
and nature of soil P and its availability[10,13], and to
provide useful information for indentifying status
of soil weathering[5,17]. Studies on soils across long
chronosequence found that Ca-associated P fraction
was changing into Al-associated P and Fe-associated
P fractions with increasing soil development (i.e.
increasing weathering and leaching)[5,17].
Foshan region has experienced a rapid economic
growth with associated expansion of urban areas
during the last three decades[18]. More than 80% of
the population in this area now lives in urban areas[19].
Rapid urbanization has resulted in a loss of more
than 30% of the forest areas during the 1988–2003,
which were mainly transformed to farmland, dike-
pond or built-up areas[18]. In the urban and suburban
areas, only small catchments of remnant native forests
nearby villages or graveyards are left, which benefit
from the geomantic culture of China[20–21]. Some
forests in hill lands are also left, as these lands are
not suitable for farming or construction. These two
types of forest sites are currently major remnant forest
sites in urban and suburban areas in Foshan region.
Their important roles in adjusting urban climate,
fixing carbon (C) and preserving biological diversity
were gradually recognized by local government,
forest managers, and scientists[21–22]. However, little is
known about the availability of soil nutrients in these
measures did not differ significantly between forest types in both soil layers. The concentrations of Al-Pi, Bray
1-Pi and total Pi in the 0–3 cm mineral soil layer and Ca-Pi in the 3–23 cm mineral soil layer were significantly
lower at the plantation forest patches than those at native forest patches. Concentrations of organic C, total N, total
P and all P fractions were significantly higher in 0–3 cm mineral soil layer than those in the 3–23 cm mineral soil
layer for both forest types. Correlation analysis indicated that soil organic matter concentration was significantly
and positively correlated with soil concentrations of all inorganic P fractions except Ca-Pi. Fourteen selected
forest patches could be divided into three groups according soil P availability by Cluster analysis. Generally, the
plantation forest patches were lower in soil P availability compared to native forest patches. These could help us
for understanding the soil nutrient status and strengthen nutrient management at remnant forest patches in the
urban and suburban areas.
Key words: Inorganic phosphorus fractions; Remnant forest; Phosphorus availability; Foshan region
侯恩庆等：佛山地区存留自然林和存留人工林的土壤无机磷组分
548　　　　　　　　　　　　　　　　　　　　 　 热带亚热带植物学报　　　　 　　　　　　　 　 第20卷
forest sites, though soil nutrients of these forest sites
were thought to be different, as different tree species,
community composition, stand ages and forest
managements were observed or documented between
or within these two types of forest sites[21,23–24].
In this study, we collected soils at 0–3 cm
and 3–23 cm mineral depths from 14 forest
patches (including seven native forest patches and
seven plantation forest patches) in the urban and
suburban areas in Foshan region, and analyzed the
concentrations of organic C, total N and total P
and inorganic P fractions, to investigate the soil P
availability at the remnant forest patches and study the
impacts of reforestation on soil P availability at the
plantation forest patches in the urban and suburban
areas in Foshan region. We hypothesized that soil P
availability was lower at the plantation forest patches
than that at the native forest patches.
1 Materials and methods
1.1 Site descriptions
This study was carried out in Foshan region
(22º38′–23º34′ N, 112º22′–113º23′ E), Guangdong
Province, China. This district is characterized by
a typical subtropical monsoon climate, with mean
annual precipitation of about 1600 mm, of which
nearly 80% falls in the hot-humid season (from April
to September) and 20% in the cool-dry season (from
October to March)[18,25]. The mean annual temperature
is about 22℃, with the coldest and warmest monthly
mean temperature of 13℃ in January and 27℃ in
July, respectively[18,25]. The soil is lateritic red earth
developed from granite[25].
Native forest patches are woodlands that are
older than 60 years and located nearby villages or
graveyards and maintained by local residents for
the geomantic culture of China (called fengshui
woods)[20–21]. These culturally protected forests are
the only type of the forests that is similar to the zonal
vegetation in the urban and suburban areas in Foshan
region[21]. Plantation forest patches are woodlands that
experienced two rotations of reforestation during the
last 60 years, and remained due to their unsuitability
for farming and construction. Before the first rotation
of reforestation, the vegetation at the plantation forest
patches is believed to be the same as that at the native
forest patches. The first rotation was reforested with
pine and/or eucalyptus species later after the harvest
of native forests during the 1970s–1980s[23,26–27]. The
second rotation was reforested with saplings of native
broadleaf tree species by imitating the tree species and
community composition of the native forest patches
around the year 2000[21].
1.2 Soil sampling and chemical analysis
For both native and plantation forests, seven
patches were randomly selected in Foshan region.
Basic information of the 14 selected forest patches
are shown in Table 1. Soils were sampled during Dec.
2008–Jan. 2009. At each forest patch, four sample
areas were randomly selected with a distance of at
least 50 m between each other. In each sample area,
after forest floor materials were removed, soils at
0–3 cm and 3–23 cm mineral depths were sampled
from down to up after a soil profile was excavated.
We divided the soils into 0–3 cm and 3–23 cm
depths because most of the excavated soil profiles
with an apparent color change at the mineral depths
of about 3 cm. Four soils at the same depth of each
forest patch were equally mixed as one composite
sample. And finally a total of 28 soil sample (2 forest
types × 7 patches × 2 soil depths) were prepared for
laboratory analysis. Soils were air dried and then
sieved through 10 mesh sieve to remove roots, gravel
and stones. Subsamples of the sieved soils were
ground to pass 100 mesh sieve for the determination
of concentrations of organic C, total N and total P.
All samples were stored in sealed plastic containers
before analysis.
Soil pH was measured using a soil∶water ratio
of 1∶2.5. Organic C concentration was measured
by dichromate oxidation methods[28]. Total N
concentration was determined with the semimicro-
Kjeldahl digestion followed by the detection of
ammonium[28]. Total P was analyzed by the molydate
第6期 549
blue method after digested with HF/HClO4
[28].
Bray 1-Pi was extracted with 0.03 mol L-1 NH4F –
0.025 mol L-1 HCl and analyzed by the molybdate blue
method[29–30]. Inorganic P fractions were sequentially
extracted with 1 mol L-1 NH4Cl (represents the soluble
and loosely bound inorganic P, SLPi in abbreviation),
0.5 mol L-1 NH4F (Al associated inorganic P, Al-Pi),
0.1 mol L-1 NaOH (Fe associated with inorganic P, Fe-
Pi), 0.3 mol L-1 Na3C3H6O7
-1 mol L-1 NaHCO3-Na2S2O4
(reductant soluble inorganic P, RSPi) and 0.25 mol L-1
H2SO4 (Ca associated inorganic P, Ca-Pi), following
the procedures of Kovar and Pierzynski[16]. Since
the color of the extracts of Al-Pi, Fe-Pi and RSPi
were dark and could affect the colorimetric analysis,
activated carbon was used to eliminate the color of
these extracts before the colorimetric analysis[30]. Total
inorganic P (total Pi in abbreviation) concentration
was calculated as the sum of concentrations of all
inorganic P fractions (SLPi, Al-Pi, Fe-Pi, RSPi and
Ca-Pi). Total organic P (total Po) concentration
was calculated as the difference between the total P
concentration and the total Pi concentration.
1.3 Data analysis
Independent-samples t test was used to compare
the differences in soil properties between the native
and plantation forests for each soil layer. Paired-
samples t test was used to compare the differences
in soil properties between the 0–3 cm and 3–23 cm
mineral soil layers for each forest type. Pearson
correlation was used to analysis the relationships
between soil properties for all soils. Hierarchical
cluster analysis using the Furthest neighbor method
and Z-scores transformation was carried out to
classify the soil P availability of 14 selected forest
patches. All these analyses were carried out using the
SPSS version 16.0 for Windows. The ratios of C∶N,
C∶P and N∶P were all on a mass basis.
2 Results and discussion
2.1 Overall soil characteristics
Soil pH values (3.8–4.3) were low at all studied
forest patches (Table 2), indicating a general strong
soil acidity in the forested areas in Foshan region.
Soil total P concentration and C∶P and N∶P ratios
are indicators of soil P availability frequently used
in previous studies[4,31]. In this study, mean total P
concentration of all soils (338 mg kg-1; Table 2) was
lower than the mean value of soil total P concentration
in tropical and subtropical China (589 mg kg-1)[31],
and mean C∶P and N∶P ratios (80.7 and 7.0,
respectively; Table 2) were both more than twice of
those in tropical and subtropical China (30.2 and 2.9,
Table 1 Site information of the 14 selected forest patches in Foshan region
Forest type Patches Topography Elevation (m) Area (hm2) Major tree species
Native Lunyong Slope 16–65 9.5 Schima superba, Indocalamus tessellates, Alpinia chinensis
Shukeng 60–87 5.9 Machilus chinesis, Lasianthus chinensis, Indocalamus tessellates
Linyue 10–32 20.8 Syzygium hancei, Symplocos lancifolia, Desmos chinensis
Shanbu 10–27 9.0 Phoebe namu, Ardisia hanceana, Ixora chinensis
Kengmei Flat 5 4.6 Schefflera octophylla, Desmos chinensis, Alocasia macrorhiza
Yangao 15 3.0 Syzygium hancei, Bambusa stenostachya, Ardisia hanceana
Yuantou 5 2.9 Helicia cochinchinensis, Desmos chinensis, Ardisia hanceana
Plantation Xialiang Slope 15–36 10.7 Cinnamomum camphora, Castanopsis hystrix, Liquidambar formosana
Zhanqigang 16–108 83.7 Cinnamomum camphora, Liquidambar formosana, C. burmannii
Sanguigang 40–69 42.2 Ficus altissima, Polyspora axillaris, Bombax malabaricum
Zhongxingang 6–73 32.8 Ficus altissima, Cinnamomum camphora, Schima superba
Longtou 10–26 5.6 Cinnamomum camphora, Albizia falcataria, Delonix regia
Xian 5–33 38.7 Schima superba, Liquidambar formosana, Ficus microcarpa
Xinjing 30–65 104.0 Ficus altissima, F. microcarpa, Bischofia javanica
侯恩庆等：佛山地区存留自然林和存留人工林的土壤无机磷组分
550　　　　　　　　　　　　　　　　　　　　 　 热带亚热带植物学报　　　　 　　　　　　　 　 第20卷
respectively)[31], indicating a relative low supply of
soil P to vegetation growth in the forested areas in
Foshan region compared to many other tropical and
subtropical areas of China.
Table 2 Statistics of soil properties of forest patches
Soil properties Mean SE Range
pH 4.0 0.0 3.8–4.3
Organic C (g kg-1) 26.2 3.0 6.4–51.7
Total N (g kg-1) 2.4 0.4 0.5–7.9
Total P (mg kg-1) 338.1 31.0 107.0 –754.0
C∶N 12.9 1.1 6.5–35.7
C∶P 80.7 8.2 23.1–188.9
N∶P 7.0 0.8 1.5–19.8
Bray 1-Pi (mg kg-1) 9.0 2.2 0.8–56.0
SLPi (mg kg-1) 0.7 0.2 0–3.1
Al-Pi (mg kg-1) 26.3 6.2 1.4–151.8
Fe-Pi (mg kg-1) 47.2 7.4 9.6–145.7
RSPi (mg kg-1) 67.2 6.8 18.2–165.7
Ca-Pi (mg kg-1) 6.0 0.8 0.8–16.8
Total Pi (mg kg-1) 147.4 17.3 31.9–399.6
Total Po (mg kg-1) 190.7 19.9 20.1–475.7
The composition of inorganic P fraction has been
extensively discussed during the past decades, and
was thought to be largely dependent on the parent
material, soil weathering and soil type[5,32–33]. In this
study, RSPi and Fe-Pi were the largest inorganic
P fractions, accounting for 21.1%–80.2% (mean
50.4%) and 13.7%–52.7% (mean 30.1%) of the
total Pi, respectively. This result is consistent with
two other studies carried out on forest soils in south
China[13,34]. The consistent composition pattern may be
because of the highly weathered and Al- and Fe-rich
characteristics of many forest soils in south China[5–6],
as Ca-P might have been gradually changed into Al-Pi
and Fe-Pi with increasing degrees of weathering and
leaching[5,33].
2.2 Comparison between forest patch types and
soil layers
None of the basic soil properties differed
significantly between the forest patch types in both
soil layers, though concentrations of organic C,
total N and total P were all tended to be lower at the
plantation forest patches than at the native forest
patches in both soil layers (Table 3). Concentrations of
most of the P fractions also did not differ significantly
between two types of forest patches in both soil layers
(Table 4). The exceptions were the concentrations of
Bray 1-Pi, Al-Pi and total Pi in 0–3 cm mineral soil
layer and Ca-Pi concentration in 3–23 cm mineral soil
layer that were all significantly lower at the plantation
forest patches than at the native forest patches (Table 4).
Table 3 Comparison of basic soil properties between forest types and soil layers
Soil depth
(cm)
Forest
type
pH
Organic C
(g kg-1)
Total N
(g kg-1)
Total P
(mg kg-1)
C∶N C∶P N∶P
0–3 Native 3.9±0.0aB 43.0±3.1aA 4.5±0.8aA 487.6±77.6aA 11.0±1.5aA 97.6±11.6aA 9.3±0.9aA
Plantation 3.9±0.1aB 38.6±2.4aA 3.2±0.6aA 311.7±34.7aA 15.1±3.6aA 130.3±12.9aA 10.6±2.0aA
3–23 Native 4.0±0.1aA 12.9±0.7aB 1.2±0.2aB 317.0±49.9aB 11.9±1.7aA 44.7±4.7aB 4.1±0.5aB
Plantation 4.2±0.0aA 10.3±1.0aB 0.8±0.1aB 236.3±43.9aB 13.6±1.5aA 50.2±7.3aB 4.1±0.9aB
n =7. Data followed different small and capital letters indicate significant difference at 0.05 level between forest types and soil layers, respectively.
In general , these resul ts suggested that
reforestation might not significantly reduce soil
nutrient concentrations or the reductions might
have been largely recovered during the past years
of recovery. Since wood harvest, burns of harvest
residues and thereafter reforestation with saplings
are all likely to reduce soil nutrient concentrations
at the plantation forest patches[35–37], the insignificant
difference of soil nutrient concentrations between
forest patch types is some surprising. We proposed
that widespread high atmospheric N deposition in
the Pearl River Delta area where Foshan region
located might have accelerated the recovery rate of
soil organic C and total N concentrations by applying
第6期 551
abundant N resource for vegetation growth[38–39].
Large variations of within forest patch type of the
selected properties may also contribute to the general
insignificant differences observed in this study, since
the study forest patches are widespread in Foshan
region.
Table 4 Concentration (mg kg-1) of soil fractions in forest types and soil layers
Soil depth (cm) Forest type Bray 1-Pi SLPi Al-Pi Fe-Pi RSPi Ca-Pi Total Pi Total Po
0–3 Native 21.7±6.2aA 1. 6±0.4aA 65±17aA 79±20aA 98±10aA 7.5±1.9aA 251±39aA 237±49aA
Plantation 6.2±0.8bA 1.1±0.2aA 22±4bA 39±10aA 71±18aA 5.1±1.1aA 138±23bA 174±40aA
3–23 Native 6.3±2.8aB 0.2±0.1aB 14±5aB 50±13aA 53±8aB 8.1±1.6aA 126±22aB 191±32aA
Plantation 1.9±0.4aB 0.1±0.0aB 4±1aB 21±6aB 46±8aA 3.3±0.9bA 75±12aB 162±39aA
n=7. Data followed different small and capital letters indicate significant difference at 0.05 level between forest types and soil layers, respectively.
Although reforestation might not significantly
affect the total P concentrations of both 0–3 cm and
3–23 cm mineral soils at the study forest patches, it
probably have significantly reduced the concentrations
of available P fractions of the 0–3 cm mineral soil,
and the impact was probably still significant after
about 9 years of recovery. As Bray 1-Pi is always
taken as the available P fraction for plant growth in
acid soil[29,40–41], and is frequently found to be related
to the P uptake by plant growth[42–43]. Soil Al-Pi
concentration is also usually found to be available to
plants[15,44]. Soil Bray 1-Pi and Al-Pi concentrations
at the plantation forest patches were both only about
30% of those at the native forest patches in both soil
layers (Table 4). The slow recovery of concentrations
of available P fractions may be because of the low
atmospheric P deposition and low soil P weathering
rate in the study area[6,45]. This result is coincident
with our previous study that found soil concentration
of mineral nutrient with low atmospheric input
(potassium) was significantly lower at the plantation
forest patches than at the native forest patches,
while soil concentration of mineral nutrient with
high atmospheric input (calcium) did not differ
significantly between two types of forest patches[46].
Both of these two studies suggested the significant
role of atmospheric deposition in determining
the recovery rates of soil mineral nutrients at the
plantation forest patches of the study area, which need
further study to verify in future.
Most of nutrient concentrations were significantly
higher in the 0–3 cm mineral soil layer than in the
3–23 cm mineral soil layer for both types of forest
patches (Tables 3 and 4). This result is likely to reflect
the uplift of nutrients by plant cycling, especially
for the strongly cycled nutrients[47]. Differences in
SLPi and Al-Pi concentrations between soil layers
were larger than the differences in Fe-Pi and RSP
concentrations between soil layers, indicating a more
strongly cycling of soil SLPi and Al-Pi fractions than
soil Fe-Pi and RSPi fractions by vegetation growth.
2.3 Relationships between soil properties
Soil pH was significantly and negatively
correlated with the concentrations of Bray 1-Pi, SLPi,
Al-Pi, RSPi and total Pi (Table 5). The negative
relationships might be because of the increasing
activation of Al and Fe and thus releasing of inorganic
P by increasing soil acidity (decreasing soil pH),
since Aln+ and Fen+ are likely to be the major cations
buffering soil acidity at the study forest patches as
suggested by the soil pH (3.8–4.3)[48]. Soil organic
C concentration was significantly and positively
correlated with soil concentrations of all inorganic
P fractions except Ca-Pi (Table 5), indicating a
significant role of soil organic matter in maintaining
soil P availability at the study forest patches.
Positive relationships between soil organic C (or
organic matter) concentration and soil inorganic P
concentrations were also reported by some other
studies[49–50]. The relationship may be because of the
sorption of inorganic P by organic matter in the soil[49].
SLPi is always believed to be one of the most
available P fractions for plant growth[14,16,51], and
侯恩庆等：佛山地区存留自然林和存留人工林的土壤无机磷组分
552　　　　　　　　　　　　　　　　　　　　 　 热带亚热带植物学报　　　　 　　　　　　　 　 第20卷
Bray 1-Pi was proposed to be one of most effective
P fractions in predicting responsiveness and fertilizer
requirement of plants in acid soils[30,42–43]. Other P
fractions which are highly correlated with these two
fractions of P may be considered to be potentially
available to plants[11]. In this study, both Al-P and Fe-P
fractions were highly correlated with the Bray 1-Pi
and SLPi fractions (r = 0.58–0.96, P < 0.05; Table 5),
indicating the potential availability of Al-Pi and Fe-Pi
fractions to plants.
Table 5 Correlations between soil properties
Property pH OC TN TP C∶N C∶P N∶P Bray 1-Pi SLPi Al-Pi Fe-Pi RSPi Ca-Pi Total Pi
OC –0.52 　 　 　 　 　 　 　 　 　 　 　 　 　
TN –0.30 0.86 　 　 　 　 　 　 　 　 　 　 　 　
TP –0.14 0.58 0.69 　 　 　 　 　 　 　 　 　 　 　
C∶N –0.05 –0.06 –0.45 –0.29 　 　 　 　 　 　 　 　 　 　
C∶P –0.31 0.69 0.74 0.08 –0.40 　 　 　 　 　 　 　 　 　
N∶P –0.49 0.67 0.39 –0.16 0.27 0.74 　 　 　 　 　 　 　 　
Bray 1-Pi –0.49 0.56 0.49 0.63 –0.06 0.11 0.07 　 　 　 　 　 　 　
SLPi –0.51 0.80 0.60 0.65 0.03 0.28 0.33 0.79 　 　 　 　 　 　
Al-Pi –0.44 0.64 0.58 0.65 –0.12 0.23 0.15 0.96 0.82 　 　 　 　 　
Fe-Pi –0.36 0.51 0.56 0.76 –0.10 0.07 –0.01 0.75 0.58 0.73 　 　 　 　
RSPi –0.38 0.49 0.43 0.55 –0.12 0.13 0.11 0.38 0.54 0.37 0.45 　 　 　
Ca-Pi 0.01 0.22 0.39 0.71 –0.36 –0.09 –0.32 0.22 0.24 0.25 0.60 0.47 　
Total Pi –0.46 0.66 0.64 0.81 –0.15 0.16 0.08 0.83 0.77 0.83 0.90 0.74 0.58 　
Total Po 0.18 0.34 0.52 0.86 –0.32 –0.01 –0.32 0.26 0.34 0.29 0.40 0.21 0.61 0.38
Data in bold indicate significant correlations at 0.05 level.
2.4 Different P availability at different forest sites
As indicated by results in this study and some
previous studies, the Bray 1-Pi, SLPi, Al-Pi and Fe-
Pi are all likely to be available to plants[14–16,42–43].
We included concentrations of these four inorganic
P fractions of the 3–23 cm mineral soils in a cluster
analysis to group the 14 selected forest patches with
respect to the different P availability. Three major
groups of forest patches were identified from the
dendrogram derived from cluster analysis. The results
are shown as following (Fig. 1 and Table 6):
Fig. 1 Hierarchical cluster dendrogram of P availability in 3–23 cm
mineral soils of 14 selected forest patches.
Table 6 Statistics of inorganic P concentrations (mg kg-1) of three 3
groups
Variance
Group A* Group B**
Group C***
Mean SD Mean SD
Bray 1-Pi 1.7 0.8 4.5 3.6 21.3
NH4Cl-P 0.1 0.1 0.2 0.1 0.4
Al-P 3.0 1.2 12.4 2.0 42.3
Fe-P 14.3 5.5 55.4 15.2 104.8
* include Shukeng, Zhongxingang, Longtou, Xian, Linyue, Xialiang,
Sanguigang and Zhanqigang；** include Kengmei, Xinjing, Yuantou,
Shanbu and Lunyong; *** include Yangao.
Low P availability: Shukeng, Zhongxingang,
Longtou, Xian, Linyue, Xialiang, Sanguigang and
Zhanqigang;
Middle P availability: Kengmei, Xinjing,
Yuantou, Shanbu and Lunyong;
High P availability: Yangao.
All plantation forest patches except for the
Xinjing forest patch were in the low soil P availability
group, reflecting the generally low soil P availability
at the plantation forest patches. In contrast, soil
P availability at the native forest patches varied
第6期 553
more greatly, with two forest patches in the low P
availability group, four forest patches in the middle P
availability group and one forest patch in the high P
availability group.
3 Conclusions
Results in this study showed that soil P
availability was relative low at the remnant forest
patches in Foshan region, compared to many other
tropical and subtropical areas of China. Fe-Pi and
RSPi were major inorganic P fractions at the study
forest patches. Most of the selected soil nutrient
measures did not differ significantly between two
types of forest patches in both soil layers, suggesting
a possible large recovery of surface soil nutrients or
insignificant impacts of reforestation on surface soil
nutrients at the plantation forest patches. However,
concentrations of Al-Pi, Bray 1-Pi and total Pi in 0–
3 cm mineral soil layer were significantly lower at the
plantation forest patches compared to the native forest
patches, indicating a slow recovery of soil available P
at the plantation forest patches, which might be related
to the low atmospheric P deposition characteristic.
Concentrations of organic C, total N, total P and all P
fractions were significantly higher in 0–3 cm mineral
soil layer than those in 3–23 cm mineral soil layer for
both types of forest patches, reflecting a significant
uplift of soil nutrients by plant cycling at both types
of forest patches. Correlation analysis suggested that
soil organic matter probably played a significant role
in maintaining soil P availability at the study forest
patches, and Al-Pi and Fe-Pi fractions were potential
available P fractions for plant growth. Fourteen
selected forest patches were divided into three groups
with different soil P availability by cluster analysis.
Generally, the plantation forest patches were lower in
soil P availability compared to native forest patches.
Results in this study provided a scientific basis for the
soil nutrient managements at remnant forest patches
in the urban and suburban areas in Foshan region.
References
[1]　 Vitousek P M. Litterfall, nutrient cycling, and nutrient limitation
in tropical forests [J]. Ecology, 1984, 65(1): 285–298.
[2]　 Davidson E A, de Carvalho C J R, Figueira A M, et al.
Recuperation of nitrogen cycling in Amazonian forests following
agricultural abandonment [J]. Nature, 2007, 447(7147): 995–996.
[3]　 Vitousek P M, Porder S, Houlton B Z, et al. Terrestrial phosphorus
limitation: Mechanisms, implications, and nitrogen-phosphorus
interactions [J]. Ecol Appl, 2010, 20(1): 5–15.
[4]　 Zhang C, Tian H, Liu J, et al. Pools and distributions of soil
phosphorus in China [J/OL]. Glob Biogeochem Cycle, 2005,
19(1): 1–8. doi:10.1029/2004GB002296.
[5]　 Adams J A and Walker T W. Some properties of a chrono-
toposequence of soils from granite in New Zealand: 2. Forms and
amounts of phosphorus [J]. Geoderma, 1975, 13(1): 41–51.
[6]　 Duan L, Hao J M, Ye X M, et al. Study on weathering rate of soil in
China [J]. Acta Sci Circumst, 2000, 20(Suppl.): 1–7.(in Chinese)
[7]　 Larssen T, Lydersen E, Tang D, et al. Acid rain in China [J].
Environ Sci Techn, 2006, 40(2): 418–425.
[8]　 Mo J M. Phosphorus availability of soils under degraded pine,
mixed and monsoon evergreen broad-leaved forests of subtropical
China [J]. Guihaia, 2005, 25(2): 186–192.(in Chinese)
[9]　 Liu X Z, Zhou G Y, Zhang D Q, et al. N and P stoichiometry of plant
and soil in lower subtropical forest successional series in southern
China [J]. Chin J Plant Ecol, 2010, 34(1): 64–71.(in Chinese)
[10]　 Chen H. Phosphatase activity and P fractions in soils of an 18-
year-old Chinese fir (Cunninghamia lanceolata) plantation [J].
For Ecol Manag, 2003, 178(3): 301–310.
[11]　 Guo F, Yost R S. Partitioning soil phosphorus into three discrete
pools of differing availability [J]. Soil Sci, 1998, 163(10): 822–833.
[12]　 Frossard E, Condron L M, Oberson A, et al. Processes governing
phosphorus availability in temperate soils [J]. J Environ Qual,
2000, 29(1): 15–23.
[13]　 Chen C R, Sinaj S, Condron L M, et al. Characterization of
phosphorus availability in selected New Zealand grassland soils
[J]. Nutr Cycl Agroecosys, 2003, 65(1): 89–100.
[14]　 Chang S C, Jackson M L. Fractionation of soil phosphorus [J].
Soil Sci, 1957, 84(2): 133–144.
[15]　 Smith A N. The supply of soluble phosphorus to the wheat plant
from inorganic soil phosphorus [J]. Plant Soil, 1965, 22(2): 314–316.
[16]　 Kovar J L, Pierzynski G M. Methods of Phosphorus Analysis for
Soils, Sediments, Residuals, and Waters [M]. 2nd ed. Virginia:
Southern Cooperative Series Bulletin No. 408, 2009: 50–53.
[17]　 Crews T E, Kitayama K, Fownes J H, et al. Changes in soil
phosphorus fractions and ecosystem dynamics across a long
chronosequence in Hawaii [J]. Ecology, 1995, 76(5): 1407–1424.
[18]　 Zhang H, Wang X R, Ma, W C, et al. Rapid urbanization and
implications for flood risk management in hinterland of the Pearl River
Delta, China: Foshan study [J]. Sensors, 2008, 8(4): 2223–2239.
[19]　 Guangdong Statistical Bureau. Guangdong Statistical Yearbook
[M]. Beijing: China Statistics Press, 2009: 70–122.(in Chinese)
侯恩庆等：佛山地区存留自然林和存留人工林的土壤无机磷组分
554　　　　　　　　　　　　　　　　　　　　 　 热带亚热带植物学报　　　　 　　　　　　　 　 第20卷
[20]　 Zhuang X Y, Corlett R T. Forest and forest succession in Hong
Kong, China [J]. J Trop Ecol, 1997, 13(6): 857–866.
[21]　 Liao Y H, Chen H Y, Wang Z, et al. Study on community of
fengshui woods and the value of application in construction of
ecological public welfare forest [J]. J Subtrop Resour Environ,
2008, 3(2): 42–48.(in Chinese)
[22]　 Huang P, Hou C M, Zhang C, et al. The value of forest
ecosystem services of Guangdong Province [J]. Ecol Sci, 2002,
21(2): 160–163.(in Chinese)
[23]　 Li J L, Fang Z L, Chen C G, et al. Investigation of tree species
composition of fengshui woods in Foshan City [J]. Guangdong
For Sci Techn, 2006, 22(1): 39–43.(in Chinese)
[24]　 Lü H R, Liu S S, Zhu J Y, et al. Effects of human disturbance on
understory woody species composition and diversity in fengshui
forests [J]. Biodiv Sci, 2009, 17(5): 458–467.(in Chinese)
[25]　 Sun L J, Chen H Y, Fang Z L, et al. Water-holding characteristic
of 0–20 cm depth soil in fengshui woods of Foshan City [J].
Guangdong For Sci Techn, 2007, 23(1): 47–52, 57.(in Chinese)
[26]　 Wen D Z, Kuang Y W, Liu S Z, et al. Evidences and implications
of vegetation damage from ceramic industrial emission on a
rural site in the Pearl River Delta of China [J]. J For Res, 2006,
17(1): 7–12.
[27]　 Fang Z L. An investigation of fengshui woods at Wanshitou
village [J]. Guangdong For Sci Techn, 2006, 22(4): 45–48.(in
Chinese)
[28]　 Liu G S, Jiang N H, Zhang L D. Soil Physical and Chemical
Analysis and Description of Soil Profiles [M]. Beijing: Standards
Press of China, 1996: 31–130.(in Chinese)
[29]　 Bray R H, Kurtz L T. Determination of total, organic, and available
forms of phosphorus in soils [J]. Soil Sci, 1945, 59(1): 39–46.
[30]　 Murphy J, Riley J P. A modified single solution method for the
determination of phosphate in natural waters [J]. Anal Chem
Acta, 1962, 27(1): 31–36.
[31]　 Tian H Q, Chen G S, Zhang C, et al. Pattern and variation of
C∶N∶P ratios in China’s soils: A synthesis of observational
data [J]. Biogeochemistry, 2010, 98(1/2/3): 139–151.
[32]　 Uriyo A P, Kesseba A. Phosphate fractions in some Tanzania soils
[J]. Geoderma, 1973, 10(3): 181–192.
[33]　 Walker T W, Syers, J K. The fate of phosphorus during
pedogenesis [J]. Geoderma, 1976, 15(1): 1–19.
[34]　 Zhang D H, Tu C J, Shen P S, et al. Phosphorus status of main
soil groups in Fujian mountainous regions [J]. Sci Silv Sin,
2008, 44(8): 29–36.(in Chinese)
[35]　 Johnson D W, Curtis P S. Effects of forest management on soil
C and N storage: Meta analysis [J]. For Ecol Manag, 2001,
140(2/3): 227–238.
[36]　 Chen C R, Xu Z H, Mathers N J. Soil carbon pools in adjacent
natural and plantation forests of subtropical Australia [J]. Soil
Sci Soc Amer J, 2004, 68(1): 282–291.
[37]　 Burton J, Chen C, Xu Z, et al. Gross nitrogen transformations in
adjacent native and plantation forests of subtropical Australia [J].
Soil Biol Biochem, 2007, 39(2): 426–433.
[38]　 Liu X J, Duan L, Mo J M, et al. Nitrogen deposition and its
ecological impact in China: An overview [J]. Environ Pollut,
2011, 159(10): 2251–2264.
[39]　 Fang Y T, Yoh M, Koba K, et al. Nitrogen deposition and forest
nitrogen cycling along an urban-rural transect in southern China
[J]. Global Change Biol, 2011, 17(2): 872–885.
[30]　 Holford I, Crocker G. Efficacy of various soil phosphate tests for
predicting phosphate responsiveness and requirements of clover
pastures on acidic tableland soils [J]. Soil Res, 1988, 26(3): 479–488.
[41]　 Romanyà J, Khanna P K, Raison R J. Effects of slash burning
on soil phosphorus fractions and sorption and desorption of
phosphorus [J]. For Ecol Manag, 1994, 65(2–3): 89–103.
[42]　 Hons F, Larson-Vollmer L, Locke M. NH4OAc-EDTA-extractable
phosphorus as a soil test procedure [J]. Soil Sci, 1990, 149(5):
249–256.
[43]　 Akinnifesi F K, Mweta D E, Saka J D K, et al. Use of pruning
and mineral fertilizer affects soil phosphorus availability and
fractionation in a gliricidia/maize intercropping system [J]. Afr J
Agri Res, 2007, 2(10): 521–527.
[44]　 Zoysa A K N, Loganathan P, Hedley M J. A technique for
studying rhizosphere processes in tree crops: Soil phosphorus
depletion around camellia (Camellia japonica L.) roots [J]. Plant
Soil, 1997, 190(2): 253–265.
[45]　 Zhang N, Qiao Y N, Liu X Z, et al. Nutrient characteristics
in incident rainfall, throughfall, and stemflow in monsoon
evergreen broad-leaved forest at Dinghushan [J]. J Trop Subtrop
Bot, 2010, 18(5): 502–510.(in Chinese)
[46]　 Hou E Q, Wen D Z, Li J L, et al. Soil acidity and exchangeable
cations in remnant natural and plantation forests in the urbanized
Pearl River Delta, China [J/OL]. Soil Res, (2012-05-14) http://
dx.doi.org/10.1071/SR11344.
[47]　 Jobbágy E, Jackson, R. The distribution of soil nutrients
with depth: Global patterns and the imprint of plants [J].
Biogeochemistry, 2001, 53(1): 51–77.
[48]　 Ulrich B, Pankrath J. Effects of Accumulation of Air Pollutants
in Forest Ecosystem [M]. London: D. Reidel Publishing
Company, 1983: 127–146.
[49]　 Darke A K, Walbridge, M R. Al and Fe biogeochemistry
in a floodplain forest: Implications for P retention [J].
Biogeochemistry, 2000, 51(1): 1–32.
[50]　 Borggaard O K, Jorgensen S S, Moberg J P, et al. Influence of
organic matter on phosphate adsorption by aluminum and iron
oxides in sandy soils [J]. J Soil Sci, 1990, 41(3): 443–449.
[51]　 Fytianos K, Kotzakioti A. Sequential fractionation of phosphorus
in lake sediments of Northern Greece [J]. Environ Monit Assess,
2005, 100(1): 191–200.