Pools and distributions of soil phosphorus in China
Chi Zhang,1Hanqin Tian,1 Jiyuan Liu,2Shaoqiang Wang,2 Mingliang Liu,1,3 Shufen Pan,1 and Xuezheng Shi4
Received 12 May 2004; revised 6 December 2004; accepted 18 January 2005; published 16 March 2005.
 We have investigated the pools and distributions of soil phosphorus (P) in the top 50 cm of soil in China by using a combination of total and available P information from more than 2400 soil profiles and a map of soil types at a resolution of 1:1,000,000. Our estimates indicate that the average total P density and available P density in China are about 8.3102 g/m3and 5.4 g/m3, respectively. The total national soil P pool in the surface half meter is 3.5 Pg (1015g). The available P density ranges from 0.7 g/m3in the Lithosols to 16.7 g/m3 in the Irrigated Silting Soils. The total P density ranges from 1.2102g/m3in the Lithosols to 19102g/m3in the Frigid Desert Soils. The ratio of available P to total P density ranges from 0.6103in Aeolian Soils to 21.6103in Coastal Solonchaks. The available P content and its vertical distribution show a complex pattern among soil orders of different development stages, possibly indicating the important role of biota’s control over soil available P content. There are large variations of P content in different climatic regions. The tropical and subtropical region has the lowest available P density (4.8 g/m3) and the second lowest total P density (8.2 102 g/m3) among all climatic regions. The large variation in the soil P content suggests that further study is needed to investigate climatic and land-use controls over the soil P content.
Citation: Zhang, C., H. Tian, J. Liu, S. Wang, M. Liu, S. Pan, and X. Shi (2005), Pools and distributions of soil phosphorus in China, Global Biogeochem. Cycles,19, GB1020, doi:10.1029/2004GB002296.
 In many terrestrial ecosystems, accumulation of C and
N seems to be regulated by the pool of biologically active P [Walker and Adams, 1958;Tate and Salcedo, 1988;Vitousek, 1998; Lloyd et al., 2001]. The supply of P to plants is generally constrained by both low total quantity of this element in the soils and by the very low solubility of the scarce quantity [Sanchez, 1976;Uehara and Gillman, 1981;
Onthong et al., 1999; Neufeldt et al., 2000]. Nevertheless, compared to widely comprehensive evaluations of the C and N cycles, there are only a few studies which have examined the large-scale P cycles in terrestrial ecosystems [Tiessen, 1995;Reddy et al., 1999;Smil, 2000]. From both scientific and ecosystem management perspectives, it is important to investigate the pool and density of soil P in terrestrial ecosystems at regional and global scales.
 Estimates on the pool and density of P in terrestrial
ecosystems at regional and global scales remain largely
uncertain [Taylor, 1964; Lerman et al., 1975; Pierrou, 1976; Richey, 1983; Smil, 1990; Mackenzie et al., 1998]. The most frequently used global soil P database, the WISE (World Inventory of Soil Emission) global soil profile database [Batjes, 2002], contains P data for only 924 soil profiles of the entire globe. The P is primarily rock derived, and its spatial distribution is highly hetero-geneous. Besides parent material, other state factors such as climate and biota also play an important role in controlling soil P content. Together, they generate a very complex temporal and spatial pattern of different P fractions during soil development. Although it is gener-ally agreed that the total soil P gradugener-ally decreases as the result of weathering [Walker and Syers, 1976], the avail-able soil P content, which can be utilized by plants, may not change in the same pattern [Crews et al., 1995;
Frizano et al., 2002]. The plants’ uptake, litterfall and decomposition rates, and mycorrhizal symbioses all can modify the available P fraction of the soil [Lajtha and Harrison, 1995]. In addition, human activity is becoming an important factor in controlling soil P content. However, there is a shortage of the quantitative assessment of controls on the soil P content at a regional scale. To reduce this uncertainty, it is necessary to have a more accurate estimate of the pool and density of soil P under different climates and land uses.
 China has an area of about 9,600,000 square
kilo-meters, covering about 50° of latitude and 60° of longi-tude. Within the vast territory, there are various soil types, which are developed under different bioclimatic
School of Forestry and Wildlife Sciences, Auburn University, Auburn, Alabama, USA.
Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China.
Also at Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China.
State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China.
Copyright 2005 by the American Geophysical Union. 0886-6236/05/2004GB002296$12.00
conditions and derived from various parent materials in diversified topographical environments. The large-scale land transformation in China in recent years has signif-icantly influenced biogeochemical and hydrological cycles [Liu et al., 2002; Tian et al., 2003]. Previous studies have indicated that land-use induced loss of soil P could result in land degradation and then affected forest regeneration and reforestation [He and Yue, 1984]. Concerns regarding a P limitation of net primary production in terrestrial ecosystems, particularly tropical forests, also arise. In addition, lakes throughout the country are commonly undergoing eutrophication partly because of the input of excessive P from the land [Jin, 2003]. To understand the P cycle and its impacts on both terrestrial and aquatic ecosystems in China, baseline estimates on the pool and density of both total and available soil P across the nation need to be developed. Since the 1960s, the Chinese government has conducted several soil surveys nation-wide, including total and available P. The major objec-tives of our study are to (1) estimate both total and available soil P content for different soil types in China and (2) investigate patterns of soil P over soil develop-ment stages and under different climatic zones. Special attention is paid to the tropical and subtropical region where terrestrial ecosystem productivity is suggested to be generally limited by soil P nutrient availability [Sanchez, 1976; Lu, 1990;Brady and Weil, 2002]. We also identify uncertainty in existing information that needs to be investigated in the future to improve our understanding
of the P cycle and its control on the C and N cycles in this region.
2. Materials and Methods
2.1. Data Sources
 The data we used for this study are from the second
national soil survey of China [National Soil Survey Office, 1993, 1994a, 1994b, 1995a, 1995b, 1996, 1998]. This database includes 2473 typical soil profiles, each of which represents soil species in the Genetic Soil Classification of China (GSCC) [Shi et al., 2004b]. Each soil profile is divided into A, B, and C horizons as depth increased. Each horizon may be further divided into several subhorizons. Soil properties examined include the thickness of the horizons, soil total P content, soil available P content, and soil bulk density. Perchloric acid digestion followed by molybdate colorimetric test was used for total P analysis [Smith, 1965]. The Olsen method was used for available P analysis [Olsen et al., 1954]. Bray I method was recom-mended for acidic soil samples [Bray and Kurtz, 1945]. Of all the 2473 soil profiles, 2451 have total P content records and 2174 have available P content records. Soil bulk density was determined by obtaining a known volume of soil, drying it to remove the water, and weighing the dry mass [Brady and Weil, 2002]. There are 1594 profiles with geographic location information and 2335 soil profiles with information on the area of each soil species according to the estimation of national soil survey [National Soil Survey
Figure 1. Distribution of soil profiles in five climatic zones across China: zone A, temperate desert;
zone B, cool temperate zone; zone C, warm temperate zone; zone D, frigid highland; and zone E, tropical and subtropical zone.
Office, 1993, 1994a, 1994b, 1995a, 1995b, 1996, 1998]. The soil data were integrated into a geographical informa-tion system (GIS) database with their geographical locainforma-tions (Figure 1). China’s soil P pools were estimated by combin-ing the P content information of each soil type with the distribution information of each soil type in China. After removal of the land area associated with water bodies,
glaciers, perennial snow, bare rock, and gravel hills the total investigated soil area is 839.36 M ha in this study.
 Since the GSCC [National Soil Survey Office, 1998]
was used in the soil surveys, we utilized the same soil classification system in this study. In addition, we have compared our system with the USDA soil taxonomy [Shi et al., 2004b] (Table 1). The Chinese soil taxonomy system we
Table 1. P Density, Total Soil P Pool Size, and Available P:Total P Ratio of Soil Great Groups
Great Groups in Chinese Soil Taxonomy U.S. Soil Taxonomy Number of Profiles Available P, g/m3 10Total P,2g/m3 Available P:Total P,
103 P Pool, 10China’s Soil2Pg
Brown coniferous forest soils Inceptisols 9 5.9 10.0 5.9 5.8
Yellow brown soils Alfisols 29 6.0 9.4 6.4 8.4
Yellow cinnamon soils Inceptisols 22 4.6 5.7 8.1 1.0
Brown soils Alfisols/Inceptisols 80 5.5 5.5 10.0 5.5
Dark-brown soils Alfisols/Inceptisols 57 2.1 9.3 2.2 18.8
Bleached Bejiang Soils Alfisols 20 6.6 8.6 7.7 2.3
Torrid red soils Inceptisols 15 2.3 2.9 7.9 0.1
Cinnamon soils Inceptisols 116 7.2 7.6 9.4 9.7
Gray-cinnamon soils Alfisols 19 4.6 2.4 3.7 3.8
Black soils Udolls 32 2.3 9.8 2.4 3.6
Gray forest soils Alfisols 6 1.0 9.2 1.0 1.4
Chernozems Cryolls 58 3.9 7.8 5.0 5.2
Castanozems Ustolls 47 3.1 8.7 3.6 15.4
Castano-cinnamon soils Ustolls 3 2.8 10.6 2.7 1.0
Dark loessial soils Calci-great groups 18 2.9 10.2 2.8 0.4
Brown caliche soils Calci-great groups 13 7.2 9.0 8.0 11.6
Sierozems Inceptisols 32 4.5 10.7 4.2 2.9
Gray desert soils Calci-great groups 11 5.4 8.7 6.2 1.3
Gray-brown desert soils Calci-great groups 10 3.9 6.8 5.8 10.4
Brown desert soils Salic great groups 8 4.7 5.1 9.2 1.1
loessial soil Orthents, Psamments 33 3.6 7.7 4.7 4.7
Red primitive soils Alfisols 21 2.7 7.4 3.7 0.7
Neo-alluvial soils Fluvents 48 5.2 7.6 6.9 1.6
Aeolian soils Psamments 45 1.0 15.4 0.6 45
Limestone soils Orthents, Psamments 47 2.8 7.0 4.1 3.8
Volcanic soils Andisols 12 4.6 16.0 2.9 0.2
Purplish soils Orthents, Psamments 83 2.5 7.5 3.4 7.1
Lithosols Orthents, Psamments 14 0.7 1.2 5.6 1.1
Skeletal soils Orthents, Psamments 42 2.7 5.1 5.4 6.6
Meadow soils Aquic suborders/Udolls 95 3.4 8.2 4.2 10.3
Sajiang black soils
(Lime concretion black soils)
Vertisols 22 3.8 6.6 5.7 1.2
Mountain meadow soils Inceptisols 14 4.1 8.4 4.8 1.8
Shrub meadow soils Inceptisols 2 1.4 7.7 1.8 0.2
Fluvi-aquic soils Fluvents 182 4.7 17.7 2.6 22.8
Bog soils Aquic suborders 34 3.4 9.4 3.6 6.0
Peat soils Histosols 8 5.0 8.0 6.3 0.5
Solonchaks Salic great groups 19 3.5 8.9 4.0 4.6
Desert solonchaks Salic great groups 5 8.0 6.9 11.6 0.9
Coastal solonchaks Salic great groups 20 15.0 7.0 21.6 0.7
Frigid plateau solonchaks Salic great groups 4 5.0 9.3 5.4 0.3
Solonetzs Salic great groups 11 5.2 5.4 9.8 0.2
Paddy soils Fluvents/Inceptisols 421 6.2 6.6 9.4 9.8
Irrigated silting soils Fluvents 24 16.7 12.8 13.0 0.8
Irrigated desert soils Fluvents 16 7.1 9.7 7.4 0.4
Felty soils (Alpine meadow soils) Inceptisols 20 3.2 6.9 4.6 18.6
Dark felty soils (subalpine meadow soils) Inceptisols 20 3.4 8.0 4.2 7.8
Frigid calcic soils Inceptisols 8 3.0 8.0 3.8 27.7
Cold calcic soil Inceptisols 24 7.9 8.9 9.0 4.9
Cold brown calcic soils Inceptisols 4 9.1 11.2 8.0 0.1
Frigid desert soils Psamments 3 4.1 19.1 2.2 8.6
Cold desert soils Psamments 2 6.5 5.2 12.5 1.4
Frigid frozon soils Psamments 5 3.8 5.6 6.7 8.7
Latosols Ultisols 23 4.1 4.0 10.2 0.8
Latosolic red soils Ultisols 29 3.2 5.0 6.4 4.4
Red soil Ultisols 122 6.4 6.1 10.6 17.3
used has a hierarchical structure, with 12 orders, 61 great groups, 235 sub-great groups, 909 families, and more than 2473 soil species.
2.2. Estimation of the Soil P Density and Pools
 We estimated the size of China’s soil P pool to the
depth of 50 cm based on the P density of each soil great group. Soil depth of each horizon based on thickness and vertical locating sequence of horizons was calculated as
where His the depth of the horizon; his the thickness of each horizon above the current horizon, including the thickness of the current horizon; and i is the number of profile layers. Then the P density of the soil was calculated as
H ; ð2Þ
where SPDHis the average soil P density of the soil to the depth of a certain horizon,His the depth of the horizon; and
hi, BDiandPiare the thickness, bulk density, and P content of each horizon above current horizon, respectively. In this way, we calculated the soil P density of different depths. In addition, for each profile we further selected the soil P density above the horizon whose depth (H) mostly approximates to 50 cm as the density of the soil species (SPDsp) it represents. Then we estimated the P density of
different soil great groups with the distributional area and P density of each soil species as following:
SPDgt¼X Asp Agt SPDsp ð3Þ Agt¼ X Asp; ð4Þ
where SPDgtis the soil P density of the great group;Aspis
the area of each soil species that belong to current great group, andAgtis the total area of the great group.
 China’s soil P pool of 50 cm depth is the summary of
the P pool of all the soil great groups. The pool size of each great group is calculated by multiply its soil P density to its area,
where SPCNis the soil P pool of China. Soil P density of
China is calculated by dividing SPCN with the total
investigated soil area of 839.36 M ha in this study.  Finally, using GIS technology, we assigned the soil P
density value to each sub-great group on the 1:1 million soil map of China [Shi et al., 2002, 2004a], and then generated the China’s soil P density map.
2.3. Estimation of Soil P Density Under Different
 To investigate the soil P density under different
climatic conditions, we first divided China into five
biocli-matic zones: Frigid Highland, Cool Temperate Zone, Warm Temperate Zone, Temperate Desert, and Tropical and Sub-tropical Zone. Then, by using a geographic information system, we located the soil profiles on the map of climatic zones, and calculated the average soil P density for each bioclimatic region (Figure 1).
3. Results and Discussion
3.1. Density and Pool of Total and Available Soil P
 Our results showed that the available P density
ranges from 0.7 g/m3 in the Lithosols (Lithic suborders, Entisols) to 16.7 g/m3in the Irrigated Silting Soils (Table 1). The total P density ranges from 1.2 102 g/m3 in the Lithosols to 19.1102g/m3in the Frigid Desert Soils. The ratio of available to total P ranges from 0.6 103 in Aeolian Soils to 21.6 103in Coastal Solonchaks. The averaged total and available P densities are 8.3102g/m3 and 5.4 g/m3, respectively.
 According toWalker and Syers, most of the P
in the Lithosols, a Lithic Entisol at its early stage of soil development, is in the primary mineral form, and the available P content is quite low. This prediction was proved by several chronosequence studies, in which available P in the youngest soil was more than an order of magnitude lower than the one in the intermediate-aged soil [Crews et al., 1995;Frizano et al., 2002]. The long cultivation history on Irrigated Silting Soils may have improved the soil productivity and increased its available P content [Zhu and He, 1992].
 Total P pool of the top 50 cm soil in China is
estimated at 3.5 Pg, which is much lower than Richey’s estimation of 13.4 Pg P in Chinese soil [Richey, 1983]. The difference between the two may be due to several reasons: (1) In our research, all of the water bodies, glaciers, perennial snow, bare rock, and gravel hills were excluded from the calculation of the total P pool, so the total terrestrial area (839.36 M ha) in our calculation is only about 77% of the area Richey used; (2) our estimate of soil P density is only two thirds of Richey’s value of 12.31
102 g/m3; and (3) the soil depth used in our estimation is
only half of what Richey used. Our estimation of total soil P density is 8.3102g/m3, or 0.42 kg P per square meter in the top 50 cm soil, which is close to Smil’s estimation of 7.5 102g/m3 [Smil, 2000].
3.2. Changes in Soil P Content Over Different Soil
 To further analyze the changes of soil P content over
time, we assigned the soils to different USDA soil orders (Figure 2). According to the weathering regime [Smeck, 1985; Brady and Weil, 2002], we divided soil orders into three weathering stages: slight (Entisols, Gelisols, Histosols, Inceptisols, Andisols), intermediate (Aridisols, Vertisols, Alfisols, Mollisols), and strong (Ultisols, Spodosol, Oxi-sols). Results indicated that as the soil ages, the total P content decreases from 9102g/m3(or about 0.64 mg P/g) to 4.9102g/m3(or 0.35 mg P/g) (Table 2), which is in the range of total soil content of eight orders (from 0.684 to 0.200 mg P/g) in another study based on 88 soil profiles
[Cross and Schlesinger, 1995]. The available P content, however, remains unchanged with soil age.
 The pattern of decreased total P content with soil age
agrees with the Walker and Syers model [Walker and Syers, 1976], which predicts that as soil ages, the loss of P through weathering will cause a decrease of total soil P content. The unchanged pattern of soil available P with soil aging is unexpected (Table 2). We attributed this pattern of available P fraction during soil development to the biota’s control over the soil P content [Lee et al., 1989; Schneider et al., 2001]. Many researchers have suggested that biological processes regulate the movement and distribution of labile forms of P, and intrasystem P cycle is important to the availability of soil P [Smeck, 1985; Stewart and Tiessen, 1987; Schlesinger, 1991]. The biota tunes the P minerali-zation rate depending on phosphorus availability and tries to maintain soil available P content to meet the nutrient requirements of the organisms [Cross and Schlesinger, 1995]. It is also suggested that the biota’s control over the P cycle becomes more and more important with soil age.
 As ecosystems develop, more soil available P is
needed to support the higher productivity of the vegetation and the increased amount of soil organisms. Also, more energy can be allocated by plants to exploit P in the deep soil by enhancing root system activity, or through symbio-ses with mycorrhizal fungi [Tiessen et al., 1994]. We further calculated the ratio between the available P density of the top 50 cm and the available P density of the total sample depth, and found this ratio increased from 1.1 in slightly weathered soil types to 1.2 in intermediately weathered soil types to 1.4 in strongly weathered soil types. This pattern suggests that biota may modify the soil P fractions by uptaking P from the deep soil and then moving it to the surface during soil development. Also, this control mecha-nism may generate a considerable upward translocation of available P that counteracts the downward P movement
through leaching during soil development. Furthermore, a productive vegetation cover can prevent loss of P in the top layer of soil by reducing soil erosion rate and by storing P in living biomass. The results of the chronosequence-landslide-scars study conducted in the Luquillo Mountains, Puerto Rico, also showed that available P in the topsoil increased with an increase in landslide age while the total P tended to decrease over time [Frizano et al., 2002]. Frizano and his colleagues further suggested that vegetation growth on highly leached soils may release some occluded P to the available pool. Similarly, investigators in the Hawaiian Chronosequence study suggested that the occlusion of P by secondary Fe and Al minerals in strongly weathered soil is not entirely permanent [Crews et al., 1995].
 Because of multiple controls from the parent material,
soil weathering regime, and biota [Smeck, 1985], and because of changes in the relative importance of these controls during soil development [Cross and Schlesinger, 1995], the pattern of soil available P content is usually quite complex. Figure 2 shows that during the soil development, the available P fraction increases from Entisols to Incepti-sols, reaching the first peak at Aridisols. Then the fraction decreases at Vertisols and Mollisols. After that, its fraction increases again at Alfisols and finally reaches the peak at
Figure 2. Changes in soil P content among seven soil orders in China. The soil orders were arranged in
an increased development sequence from left to right.
Table 2. P Density of Soils of Different Development Stages in China Weathering Status Available P, g/m3 Total P, 102g/m3 Available P:Total P, 103 Slighta 4.7 ± 1.2 9.0 ± 1.6 5 Intermediateb 4.8 ± 1.3 8.4 ± 0.7 6 Strongc 5.1 ± 1.4 4.9 ± 0.7 10 a
Including soil orders Entisols, Gelisols, Histosols, Inceptisols, Andisols; b
Including soil orders Aridisols, Vertisols, Alfisols, Mollisols; cIncluding soil orders Ultisols, Spodosol, Oxisols.
the Ultisols. This pattern is consistent with the observation of resin and bicarbonate fraction among soil orders on 88 soil profiles in another study [Cross and Schlesinger, 1995], implying that there may be some general patterns of soil available P fraction among soil orders of different weather-ing stages. This pattern, however, cannot simply be de-scribed as being increased or decreased with soil age.
3.3. Variations in Total and Available Soil P Across
 The spatial distribution of total soil P density shows
substantial variation across China, but the total P content in the tropical and subtropical soils of the southeast China is generally lower than other parts of the nation (Figure 3). The content of soil P density is dependent on the degree of soil weathering, which is controlled by the hydrothermal conditions, soil age, and parent material [Gardner, 1990]. China is characterized by a great spatial variability in
climate, ranging from tropical to cool temperate zones [Tian et al., 2003;Wu et al., 2003]. The southern part of China is strongly humid due to the influences of the Asian monsoon circulations [Zhang, 1991; Tian et al., 2003], while in northwest China, the barrier effect of the Tibetan Plateau to moisture and the long distance from the ocean result in an arid climate. Thus the mean annual temperature of China increase from about6.5°C to 23.5°C with the decrease of latitude, and the annual precipitation decreases from about 2500 mm to 15 mm along southeast to northwest China. Cold conditions prevail across the Tibetan Plateau owing to the high elevation.
 To investigate the effects of climate on soil P pools,
we analyzed the soil P distribution patterns in five different climatic zones (Table 3, Figure 1). The results showed considerable variation of soil P density in the top 50 cm of soils among different climatic zones. The warm temper-ate zone has the highest soil available phosphorus density
Figure 3. Spatial distribution of total soil phosphorus density in China (g/m3). See color version of this figure at back of this issue.
Table 3. Soil P Density of Different Climatic Zones of China (g/m3)
Climatic Zone Total P Available P Number of Profiles Mean, 102g/m3 95% Confidence Interval Number of Profiles Mean, g/m3 95% Confidence Interval Frigid Highlands 113 9.7 1.0 102 5.8 1.4
Cool Temperate Zone 237 8.0 1.0 193 5.2 2.1
Warm Temperate Zone 461 8.8 1.3 413 8.9 1.4
Temperate Desert Zone 92 10.0 1.7 82 6.3 2.1
(8.9 ± 1.4 g/m3). The temperate desert zone in northwest China has the highest total P density (10 ± 1.7102g/m3). There may be many factors that control the total soil P density. The hydrothermal conditions in the temperate desert zone (low precipitation and temperature) may con-tribute to the preservation of total soil P content by limiting the degree of soil weathering and P loss through surface soil erosion.
 The tropical and subtropical regions in southeast
China have the second lowest total soil phosphorus density (8.2 ± 1.4 102 g/m3), and the lowest soil available phosphorus density (4.8 ± 0.9 g/m3). The low soil P density in southeast China can also be observed from the total soil P density map (Figure 3). Many previous investigators have suggested that high temperature and precipitation in tropical and subtropical regions enhance the soil weathering and the P loss through soil erosion [Vitousek and Walker, 1987;Lu, 1990;Onthong et al., 1999;Neufeldt et al., 2000;Lehmann et al., 2001]. In the Hawaii tropical forests, the P fertiliza-tion increased the forest net primary productivity by about 25% [Herbert and Fownes, 1995]. There were also reports that the ecosystem productivity of tropical and subtropical China is limited by its low soil P content [Peng and Zhao, 2000].
 The dramatic land-use change in recent years may be
an important control in the soil P content of tropical and subtropical China where soil P content is relatively low [Liu et al., 2002; Tian et al., 2003, 2005]. Previous analyses showed that soil erosion could lead to a low P content in the dry cultivated field, despite the P fertilizer applied into the soil [He and Yue, 1984]. The nutrient loss through soil erosion is significant in tropical and subtropical China where annual precipitation is high [Yu and Peng, 1995]. In Hainan, a tropical province of China, the annual soil erosion in a fallowed land is as high as 3200 g/m2while the erosion rate of a forest land is only 5 g/m2per year [Peng and Zhao, 2000]. The high soil erosion after deforestation may deplete the total P budget. In Xiaoliang Tropical Chinese Ecological Research Station, the P content of forest soil is 9 times higher than that of deforested soil. From 1995 to 2000, about 90,724 hectares of forests are converted to agriculture lands in southeast, most of which occurred in the tropical-subtropical zone of China [Liu et al., 2002]. The high erosion rate of deforested lands may threaten soil productivity [Saugier et al., 2001; Davidson et al., 2002] and even cause reforestation to be impossible without P fertilization on such poor-P soils [Peng and Zhao, 2000]. Furthermore, the P input from the land has become a major problem of eutrophication of aquatic ecosystem in China [Jin, 1990, 2003].
 Our findings show that the average total P density
and available P density of the China’s soils are about 8.3
102g/m3and 5.4 g/m3, respectively. The total national soil
P pool in the surface half meter is 3.5 Pg. A general pattern of decreased total soil P density with increased soil weath-ering state was apparent while little difference in available soil P density was observed. A comparison of total soil P
content in the upper 50 cm relative to total P in soils below 50 cm also varied with soil weathering state, demonstrating an increasing relative amount of P in the upper soil with weathering. General patterns of P content with climatic zone also were apparent in the analysis, with tropical-subtropical zone being relatively lower than most of the other zones. Deforestation in the tropical-subtropical zone could lead to a decrease in soil P.
 It is also recognized that since the structure of soil is
highly heterogeneous, estimations and comparisons of soil P density may contain noticeable errors. The large variation in the soil P content suggests that further study is needed to determine the climatic and land-use controls over the soil P content.
 Acknowledgments. This study has been supported by grants from the Chinese Academy of Sciences (CAS), National Science Founda-tion of China (40128005), and NASA Interdisciplinary Science Program (NNG04GM39C). We thank A. Chappelka, H. Chen, M. Andreae, and two anonymous reviewers for very helpful comments and suggestions.
Batjes, N. H. (2002), A homogenized soil profile data set for global and regional environmental research (WISE, version 1.1),Rep. 2002/01, Int. Soil Ref. and Inf. Cent., Wageningen, Netherlands.
Brady, N. C., and R. R. Weil (Eds.) (2002),The Nature and Properties of Soils, 13th ed., Pearson Educ. Group, Upper Saddle River, N. J. Bray, R. H., and L. T. Kurtz (1945), Determinations of total, organic and
available forms of phosphorus in soils,Soil Sci.,125, 49 – 54. Crews, T. E., K. Kitayama, J. H. Fownes, R. H. Riley, D. A. Herbert, M.-D.
Dieter, and P. M. Vitousek (1995), Changes in soil P fractions and eco-system dynamics across a long chronosequence in Hawaii,Ecology,76, 1407 – 1424.
Cross, A. F., and W. H. Schlesinger (1995), A literature review and evalua-tion of the Hedley fracevalua-tionaevalua-tion: Applicaevalua-tions to the biochemical cycle of soil phosphorus in natural ecosystems,Geoderma,64, 197 – 214. Davidson, E. A., C. J. R. de Carvalho, I. C. G. Vieira, R. O. Figueiredo,
P. Moutinho, F. Y. Ishida, M. T. P. dos Santos, J. B. Guerrero, K. Kalif, and R. T. Saba´ (2002), Nutrient limitation of biomass growth in a tropical secondary forest: Early results of a nitrogen and phosphorus amendment experiment,Ecol. Appl.,14, S150 – S163.
Frizano, J., A. H. Johnson, D. R. Vann, and F. N. Scatena (2002), Soil P fractionation during forest development on landslide scars in the Luquillo Mountains, Puerto Rico,Biotropica,34, 17 – 26.
Gardner, L. R. (1990), The role of rock weathering in the phosphorus budget of terrestrial watersheds,Biogeochemistry,11, 97 – 110. He, S., and Z. Yue (1984), The studies on the reconstruction of vegetation
in tropical coastal eroded land in Guangdong,Trop. Subtrop. For. Eco-syst.,2, 87 – 90.
Herbert, D. A., and J. H. Fownes (1995), P limitation of forest leaf area and net primary production on a highly weathered soil,Biogeochemistry,29, 223 – 235.
Jin, X. (1990),Eutrophication of Lakes in China, China Environ. Sci. Press, Beijing.
Jin, X. (2003), Analysis of eutrophication state and trend for lakes in China, J. Limnol.,62, 60 – 66.
Lajtha, K., and A. F. Harrison (1995), Strategies of P acquisition and con-servation by plant species and communities, inPhosphorus in the Global Environment, edited by H. Tiessen, pp. 139 – 147, John Wiley, Hoboken, N. J.
Lee, D., X. Han, and C. F. Jordan (1989), Soil P fractions, aluminum, and water retention as affected by microbial activity in an Ultisol,Plant Soil, 121, 125 – 136.
Lehmann, J., D. Gu¨nther, M. S. Mota, M. P. Almeida, W. Zech, and K. Kaiser (2001), Inorganic and organic soil P and sulfur pools in an Amazonian multistrata agroforestry system,Agrofor. Syst.,53, 113 – 124. Lerman, A., F. T. Mackenzie, and R. M. Garrels (1975), Modeling of geochemical cycles: P as an example, Geol. Soc. Am. Mem., 142, 205 – 218.
Liu, J., M. Liu, D. Zhuang, Z. Zhang, and X. Zheng (2002), The spatial patterns of recent Chinese land use change,Sci. China,32, 1031 – 1040. Lloyd, J., M. I. Bird, E. M. Veenendaal, and B. Kruijt (2001), Should P availability be constraining moist tropical forest responses to increasing
CO2concentrations?, inGlobal Biogeochemical Cycles in the Climate
System, edited by S. Schulze et al., pp. 96 – 109, Elsevier, New York. Lu, R. (1990), Soil phosphorus, in Soils of China, 2nd ed., edited by
Y. Xiong and K. Li, pp. 577 – 602, Sci. Press, Beijing.
Mackenzie, F. T., L. M. Ver, and A. Lerman (1998), Coupled biogeochem-ical cycles of carbon, N, P, and sulfur in the land-ocean-atmosphere system, inAsian Change in the Context of Global Change, edited by J. N. Galloway and J. M. Melillo, pp. 42 – 100, Cambridge Univ. Press, New York.
National Soil Survey Office (Ed.) (1993),Soil Species of China, vol. I, Chin. Agric. Press, Beijing.
National Soil Survey Office (Ed.) (1994a),Soil Species of China, vol. II, Chin. Agric. Press, Beijing.
National Soil Survey Office (Ed.) (1994b),Soil Species of China, vol. III, Chin. Agric. Press, Beijing.
National Soil Survey Office (Ed.) (1995a),Soil Species of China, vol. IV, Chin. Agric. Press, Beijing.
National Soil Survey Office (Ed.) (1995b),Soil Species of China, vol. V, Chin. Agric. Press, Beijing.
National Soil Survey Office (Ed.) (1996),Soil Species of China, vol. VI, Chin. Agric. Press, Beijing.
National Soil Survey Office (Ed.) (1998),Chinese Soils, Chin. Agric. Press, Beijing.
Neufeldt, H., J. E. da Silva, M. A. Ayarza, and W. Zech (2000), Land-use effects on phosphorus fractions in Cerrado oxisols,Biol. Fertil. Soils,31, 30 – 37.
Olsen, S. R., C. V. Cole, F. S. Watanabe, and L. A. Dean (1954), Estimation of available phosphorus in soils by extraction with sodium bicarbonate, Circ. 939, U.S. Dep. of Agric., Washington, D. C.
Onthong, J., M. Osaki, C. Nilnond, and T. Tadano (1999), Phosphorus status of some highly weathered soils in peninsular Thailand and avail-ability in relation to citrate and oxalate application,Soil Sci. Plant Nutr., 45, 627 – 637.
Peng, S., and P. Zhao (2000), Restoration of degraded ecosystems and its ecological effect in low subtropical China, paper presented at Interna-tional Restoration Conference, Soc. for Ecol. Restoration, Liverpool, UK. Pierrou, U. (1976), The global phosphorus cycle,Ecol. Bull.,22, 57 – 88. Reddy, K. R., G. A. O’Connor, and C. L. Schelske (Eds.) (1999),
Phos-phorus Biogeochemistry in Sub-Tropical Ecosystems, CRC Press, Boca Raton, Fla.
Richey, J. E. (1983), The P cycle, inThe Major Biogeochemical Cycles and Their Interaction, edited by B. Bolin et al., pp. 51 – 65, John Wiley, Hoboken, N. J.
Sanchez, P. A. (Ed.) (1976),Properties and Management of Soils in the Tropics, John Wiley, Hoboken, N. J.
Saugier, B., J. Roy, and H. A. Mooney (2001), Estimations of global terrestrial productivity: Converging toward a single number?, in Terres-trial Global Productivity, edited by J. Roy et al., pp. 543 – 557, Elsevier, New York.
Schlesinger, W. H. (1991),Biogeochemistry: An Analysis of Global Change, Elsevier, New York.
Schneider, K., M. B. Turrion, P. F. Grierson, and J. F. Gallardo (2001), Phosphatase activity, microbial P, and fine root growth in forest soils in the Sierra de Gata, western central Spain,Biol. Fertil. Soils,34, 151 – 155.
Shi, X., D. Yu, X. Pan, W. Sun, Z. Gong, E. D. Warner, and G. W. Petersen (2002), A framework for the 1:1,000,000 soil database of China, in paper presented at 17th World Congress of soil Science, Int. Union of Soil Sci., Bangkok.
Shi, X., D. Yu, X. Pan, W. Sun, H. Wang, and Z. Gong (2004a), 1:1,000,000 soil database of China and its application (in Chinese), in Proceedings of 10th National Congress of Soil Science of China, pp. 142 – 145, Sci. Press, Beijing.
Shi, X., D. Yu, W. Sun, H. Wang, Q. Zhao, and Z. Gong (2004b), Reference benchmarks relating to great groups of genetic soil classification of China with soil taxonomy,Chin. Sci. Bull.,49, 1507 – 1511.
Smeck, N. E. (1985), Phosphorus dynamics in soils and landscapes, Geo-derma,36, 185 – 199.
Smil, V. (1990), N and P, inThe Earth as Transformed by Human Action, edited by B. L. Turner, pp. 428 – 435, Cambridge Univ. Press, New York. Smil, V. (2000), P in the environment: Natural flows and human
interfer-ences,Annu. Rev. Energy Environ.,25, 53 – 88.
Smith, G. F. (1965),The Wet Chemical Oxidation of Organic Compositions Employing Perchloric Acid, G. Frederick Smith Chem. Co., Columbus, Ohio.
Stewart, J. W. B., and H. Tiessen (1987), Dynamics of soil organic phos-phorus,Biogeochemistry,4, 41 – 60.
Tate, K. R., and I. Salcedo (1988), P control of soil organic matter accu-mulation and cycling,Biogeochemistry,5, 99 – 107.
Taylor, S. R. (1964), Abundance of chemical elements in the continental crust: A new table,Geochim. Cosmochim. Acta.,28, 1273 – 1285. Tian, H. Q., J. M. Melillo, D. W. Kicklighter, S. Pan, J. Liu, A. D. McGuire,
and B. Moore III (2003), Regional carbon dynamics in monsoon Asia and its implications to the global carbon cycle,Global Planet. Change,37, 201 – 217.
Tian, H. Q., J. Liu, J. Melillo, M. Liu, D. Kicklighter, X. Yan, and S. Pan (2005), The terrestrial carbon budget in East Asia: Human and natural impacts, inChanges in the Human-Monsoon System of East Asia in the Context of Global Change, edited by C. Fu, J. Freney, and J. Stewart, Island Press, Washington D. C., in press.
Tiessen, H. (Ed.) (1995), Phosphorus in the global environment: Transfers, cycles and management, inSCOPE 54, 460 pp., John Wiley, Hoboken, N. J.
Tiessen, H., E. Cuevas, and P. Chacon (1994), The role of soil organic matter in sustaining soil fertility,Nature,371, 783 – 785.
Uehara, G., and G. Gillman (1981), The mineralogy, chemistry, and physics of tropical soils with variable charge clays, inWestview Tropical Agri-culture Series, vol. 4, Westview, Boulder, Colo.
Vitousek, P. M. (1998), The structure and functioning of montane tropical forests: Control by climate, soils, and disturbance,Ecology,79, 1 – 2. Vitousek, P. M., and L. R. Walker (1987), Colonization, succession and
resource availability: Ecosystem-level interactions, inColonization, Suc-cession and Stability, edited by A. J. Gray, M. J. Crawley, and P. J. Edwards, pp. 207 – 223, Blackwell Sci., Malden, Mass.
Walker, T. W., and A. F. R. Adams (1958), Studies on soil organic matter: I. Influence of phosphorus content of parent materials on accumulations of carbon, nitrogen, sulfur and organic phosphorus in grassland soils,Soil Sci.,85, 307 – 318.
Walker, T. W., and J. K. Syers (1976), The fate of P during pedogenesis, Geoderma,14, 1 – 19.
Wu, H., Z. Guo, and C. Peng (2003), Distribution and storage of soil organic carbon in China,Global Biogeochem. Cycles, 17(2), 1048, doi:10.1029/2001GB001844.
Yu, Z., and S. Peng (1995), The artificial and natural restoration of tropical and subtropical forests,Acta Ecol., Sin.,15, 1 – 17.
Zhang, J. (Ed.) (1991),Climate of China, Sci. Press, Beijing.
Zhu, H., and Y. He (Eds.) (1992),Pedogeography(in Chinese), Higher Ed., Beijing.
J. Liu and S. Wang, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China.
M. Liu, S. Pan, H. Tian, and C. Zhang, School of Forestry and Wildlife Sciences, Auburn University, Auburn, AL 36849, USA. (tianhan@auburn. edu)
X. Shi, State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008, China.