Rapid identification of source material levels in coral sand ornithogenic sediments by reflectance spectroscopy

Ecological Indicators 23 (2012) 517–523

Contents lists available at SciVerse ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Rapid identification of source material levels in coral sand ornithogenic sediments by reflectance spectroscopy Li-Qiang Xu a,b , Xiao-Dong Liu a,∗ , Li-Guang Sun a,∗∗ , Wen-Qi Liu c a

Institute of Polar Environment, University of Science and Technology of China, Hefei, Anhui 230026, China USTC-CityU Joint Advanced Research Center, Suzhou, Jiangsu 215123, China c Instruments’ Center for Physical Science, University of Science and Technology of China, Hefei, Anhui 230026, China b

a r t i c l e

i n f o

Article history: Received 24 May 2011 Received in revised form 10 March 2012 Accepted 7 May 2012 Keywords: South China Sea Coral island Ornithogenic sediments Reflectance spectroscopy Guano

a b s t r a c t Ornithogenic sediments are ideal materials for studying past eco-environmental changes. We collected sediment cores from Xisha archipelago of the South China Sea. The relative ratios of source materials (plant, guano and coral sand) of such sediments from Guangjin, Jinqing, Jinyin and Ganquan Islands in the Xisha archipelago, were reconstructed by reflectance spectroscopy. The consistence between the results obtained from reflectance spectroscopy and those obtained from traditional chemical analysis suggests that reflectance spectroscopy can be an ideal indicator for source material levels in the ornithogenic coral sand sediments. The possible implications of changes in source materials from the bottom toward top of the sediment cores are also discussed in our study. Seabird activities appear central to the ecological development of these island ecosystems. © 2012 Elsevier Ltd. All rights reserved.

1. Introduction The South China Sea is a region with one of the world’s highest levels of biodiversity and provides an ideal window for observing and predicting ecological changes in response to local and global climate oscillations (Morton and Blackmore, 2001; Liu et al., 2006; Zhao, 1996). There are many dotted and geographically isolated islands in the South China Sea. These pristine islands provide specific perspectives into species evolution and environmental changes, as well as ecological responses to climate changes. Due to the long distance from mainland China and the restrictions imposed by the Chinese military on travel to this area, the Xisha archipelago is an ideal place to study developmental history of tropical coral island ecosystem in the ocean, but our knowledge of past eco-environmental changes on these insular islands is limited. So far, the long-term interaction between plant development and seabird activity on these islands of the South China Sea has not been well documented. Our systematic field studies suggest that there were abundant seabirds on these islands, but they are now almost completely extinct. The development of such insular ecosystems and the reason for the rapid decline of seabird ecology in recent years remain unclear. Changes in source materials are essential in examining the development of coral

∗ Corresponding author. Tel.: +86 551 3606051. ∗∗ Corresponding author. Tel.: +86 551 3607583; fax: +86 551 3601415. E-mail addresses: [email protected] (X.-D. Liu), [email protected] (L.-G. Sun). 1470-160X/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ecolind.2012.05.004

island ecosystems and environmental stress on these islands. Thus, an investigation into the composition of the widely distributed ornithogenic sediments in such insular coral islands is necessary. However, traditional chemical analysis is sample destructive, time consuming and, normally, costly. Developing a rapid and effective approach to reconstructing source material levels will help in obtaining a rapid overview of the sedimentary environment and potential eco-environmental implications. Due to its rapidity, convenience and accuracy, reflectance spectroscopy within the visible–near-infrared region (380–2500 nm) has been widely used as a marker for predicting chemical constituents in soils, sediments and biological materials (Malley and Williams, 1997; Rosén et al., 2000; Wu et al., 2005; Roumet et al., 2006; Hay et al., 2010; Showers et al., 2006). Also, Nearinfrared reflectance spectroscopy (NIRS) can be used to quantify the ecological properties of large sample numbers, and thus this technique can offer ecologists enormous analytical power (Kleinebecker et al., 2009; Foley et al., 1998; Stolter et al., 2006). In our previous study, we have investigated the reflectance spectral properties of Antarctic ornithogenic sediments and successfully reconstructed bio-element levels, as well as penguin population, using a two-component model (Liu et al., 2010, 2011a). While our previous studies mainly focused on a two-component mixing model, whether this technique is applicable to complicated three-component systems in the tropical islands remains unclear. One of our aims in the present study is to test the feasibility of predicting constituents on a three-component basis. We test the potential use of reflectance spectroscopy to reconstruct

518

L.-Q. Xu et al. / Ecological Indicators 23 (2012) 517–523

Fig. 1. Map of the Xisha Archipelago showing sampling sites.

the relative abundance of source materials (plant residue, guano and coral sand) in the ornithogenic coral sand sediments from the Xisha archipelago, South China Sea. Moreover, possible ecoenvironmental implications of reconstructed source material levels by reflectance spectroscopy are also discussed in the present study.

2. Materials and methods 2.1. Study area The Xisha archipelago (15◦ 47 –17◦ 08 N, 110◦ 10 –112◦ 55 E), located in the central South China Sea, consists of more than 40 islets, sandbanks and reefs (Fig. 1). The archipelagic water and land areas cover ∼500 000 km2 and 8 km2 , respectively. The Xisha archipelago can be divided into two groups: the eastern and the western islands. The group of Xuande archipelago in the east consists of Zhaoshu Island, North Island, Middle Island, South Island, Yongxing Island, Shidao Island, Dongdao Island and many sand cays. The southwestern group, Yongle archipelago, is composed of Shanhu Island, Ganquan Island, Jinyin Island, Guangjin Island, Jinqing Island, Chenhang Island and Zhongjian Island, in addition to sandbanks and reefs (Hainan Ocean Administration, 1999). According to observations from Yongxing meteorological station, the annual mean air temperature and annual rainfall of the Xisha Islands are 26–27 ◦ C and 1500 mm respectively. From June to November, the Xisha Islands are subject to the effects of the southwest monsoons, tropical cyclones of high frequency from intense convergent convection, and heavy precipitation; about 87% of the

total annual precipitation occurs in these months. The dry season runs from December to May due to the influence of northeast monsoons, and the precipitation during this period is generally 13% of the total annual precipitation (Lin et al., 1999). Most of the Xisha Islands are covered with thriving vegetation, and the typical plant community has circular-zonary growth around the islets. The interiors of most of the islets are covered by trees such as Pisonia grandis, and bordered by shrubs such as Scaevola sericea, Messerschmidia argentea and Guettarda speciosa. The abundant vegetation provides a good habitat for seabirds and there were once more than 60 species of birds on these islands. This has resulted in the accumulation of abundant guano in soils, which provides a major nutrient source for plant growth and enhances soil development (Exploration Group of Xisha Islands of Institute of Soil Science of CAS, 1977; Gong and Huang, 1997).

2.2. Sample collection All the sediment cores used in this study, except for core JY2 (collected during an earlier expedition in 2003), were collected in 2008. Using PVC gravity pipes, sediment cores GQ (16◦ 30 15.1 N, 111◦ 35 6.2 E), GJ3 (16◦ 27 07 N, 111◦ 42 05 E), JQ (16◦ 27 50 N, 111◦ 44 27 E) and JY2 (16◦ 26 57 N, 111◦ 30 24 E) were collected under woodland and shrubs of Ganquan Island, Guangjin Island, Jinqing Island and Jinyin Island, respectively (Fig. 1). Detailed parameters of the islands and the sampling sites are given in Table 1. The outer shores of these four islands are surrounded by several-meter-high sand barriers, which are covered with shrubs.

L.-Q. Xu et al. / Ecological Indicators 23 (2012) 517–523

519

Table 1 Parameters of the studied islands and sediment profiles. Island Jinyin Island Guangjin Island Jinqing Island Ganquan Island

Sampling site location ◦





◦





16 26 57 N, 111 30 24 E 16◦ 27 07 N, 111◦ 42 5 E 16◦ 27 50 N, 111◦ 44 27 E 16◦ 30 15.1 N, 11◦ 35 6.2 E

Elevation (m)

Length × width (m × m)

Area (km2 )

Profile code

Length (cm)

0.2 6-8 6.0 0.2

1020 × 350 320 × 200 880 × 230 1020 × 350

0.36 0.06 0.20 0.3

JY2 GJ3 JQ GQ

55 95 55 107

When sampling, we inserted 11 cm diameter PVC pipes with lengths of 107 cm (GQ), 95 cm (GJ3), 55 cm (JY2) and 55 cm (JQ) into the soft substrate, and then excavated sediments around the pipes to retrieve those cores. These cores were opened, photographed and then sectioned at 1 cm intervals in the laboratory. According to field and laboratory observation, these cores have a similar lithology. All the loose sediments were a mixture of plant humus, guano and coral sand, with a few well-preserved remains of bird and fish bones, as well as fish scales. Large quantities of black humus were observed in the upper layers, likely indicating a high level of organic matter. The color changed gradually from black to light yellow down the cores. Large numbers of guano pellets were observed overall, but the bottom layers consisted primarily of medium-large size coral sands with few yellow guano particles, reflecting a remarkable decrease in organic matter. Guano pellets were observed throughout all the profiles, and could be easily distinguished from white coral sand by their dark yellow color and relatively soft structure. Several plant residue, guano pellet and coral sand samples were extracted manually with stainless steel tweezers in duplicate profiles, and were considered to be source materials of the ornithogenic sediments. Our chronological analysis of these sediment cores have shown that ages of the sediments increased with depth and there was no abrupt change in sedimentation rates (Xu et al., 2010, 2011a,c; Liu et al., 2011b, 2012), suggesting that the sampling sites have not been disturbed by direct human activities and not subjected to floods during tropical cyclones. In addition, according to the lithological observation, the variation in sediment patterns throughout the core is progressive in nature, with few sharp boundaries, and therefore the sedimentation in these profiles is interpreted as continuous. The changes in source material compositions thus reflect the gradual development of such coral islands.

2.3. Analytical methods According to our previous study (Liu et al., 2006), phosphorous is a robust marker for seabirds’ activity due to its high level in both fresh and ancient bird droppings. To investigate the influence of guano on the texture of the sediments, we analyzed the level of phosphorous in the bulk sediments by ICP-AES, after the sediments had been digested with multi-acids (HClO4 /HF/HNO3 /HF). Reagent blanks and standard reference materials (GBW07108/GBW07303) were used as quality controls in the analysis. The CaCO3 level, an indicator of coral sand, was determined by subtracting the TOC from the total carbon (Chang et al., 2005; Bernárdez et al., 2008). Prior to spectral analysis, the extracted plant residues were placed in a 50 ml centrifuge tube and treated with excess 5% hydrochloric acid in an ultrasonic environment with acid changed at regular intervals until all the carbonate was completely removed. The residues were then repeatedly washed in de-ionized water to remove excess acid and ultimately air dried in the lab. The bulk sediments, as well as source materials, were processed as follows: subsamples at intervals of 1 cm were homogenized with a pestle and mortar after being dried to a constant weight at a temperature of 105 ◦ C, and were then passed through a 200 mesh sieve. The powder sample was packed into a measuring cell and the diffuse reflectance spectra of the bulk sediments were then acquired on a Shimadzu DUV-3700 UV–Vis–NIR recording spectrophotometer

by scanning from 380 to 2500 nm, at intervals of 1 nm. Thus, 2121 data points were obtained for a single spectrum. To calculate the relative abundance of source materials, the spectra of three endmembers (plant humus, guano and coral sand) were also investigated. An external, polyethylene (zero absorbance) reference standard was read alternately with the samples. The resultant reflectance (r) data were then transformed to absorbance (log 1/r) data using UV.Prove software. 3. Results and discussion 3.1. Reconstruction of source materials A large number of well-preserved guano pellets were observed throughout all the profiles and thus the influence of guano on the spectral property could be significant and cannot be ignored. It is widely accepted that coral islands are derived from coral clasts and coral sands are the dominant constituents in coral island ornithogenic sediments. In addition to guano and coral sand, both field investigation and laboratory analyses have shown that plant residue is also an important constituent of these sediments (Xu et al., 2010). Thus, we hypothesize that plant humus, guano and coral sand are three source materials for the bulk ornithogenic sediments. The physical/chemical properties of these three endmembers are quite different and this enables a potential use of reflectance spectroscopy to reconstruct their ratios in the sediments. The absorbance spectra data (380–2500 nm) of these three components (coral sand, guano and plant humus) obtained from reflectance spectrophotometer are plotted in Fig. 2. We investigated Antarctic ornithogenic sediments by spectroscopy and reconstructed penguin bio-element levels, or penguin population, using a two-component model. We obtained a result consistent with that obtained from chemical analyses (Liu et al., 2010, 2011a), validating the reliability of using reflectance to reconstruct penguin dropping levels. In the ornithogenic sediments from the Xisha Islands, we observed different characteristic absorbance spectra (1/r) for the three components and there was good repeatability for each endmember (Fig. 2). This is helpful in distinguishing the signal from each component. Due to the difference in spectra, we anticipate that the relative abundance of these three components can be calculated from the differences in their absorbance spectra on a three-endmember basis. The spectral signal of samples is a combination of the characteristic spectrum of each endmember, and the ratio of each endmember is calculated using the software Matlab 7.1 from the equation below: A = XAplant + YAguano + ZAcoral where Aplant , Aguano and Acoral are characteristic absorption spectrums of plant humus, guano and coral sand from 380 to 2500 nm and X, Y, Z are respectively the estimated ratios of each component, conforming to the condition X + Y + Z = 1. Aplant , Aguano and Acoral at any wavelength between 380 and 2500 nm were constant in the equation. We assigned different values with increments of 0.01 to X, Y and Z and synthesized a total of 104 spectra curves, plotting the guano, plant and coral sand proportion in the range from 0 to 100%. For each spectrum, we hypothesize that X, Y and Z are the estimated

520

L.-Q. Xu et al. / Ecological Indicators 23 (2012) 517–523

0.25

1 (A) Coral sand (n=4)

(B) Guano (n=5)

0.8

Absorbance

Absorbance

0.2 0.15 0.1 0.05

0.6 0.4 0.2

0

0 500

0

1000

1500

2000

2500

500

0

Wave length (nm)

1500

2000

2500

1.8

2

1.6

(C) Plant humus (n=5)

1.6

(D) Mean

1.4

Absorbance

Absorbance

1000

Wave length (nm)

1.2 0.8

1.2

Plant humus (n=5)

1

Guano (n=5)

0.8

Coral sand (n=4)

0.6 0.4

0.4

0.2 0

0 500

0

1000

1500

2000

0

2500

500

Wave length (nm)

1000

1500

2000

2500

Wave length (nm)

Fig. 2. Characteristic spectra of three endmembers. A–C are spectra of different environmental mediums, and D is the mean of each.

1

0.5 Plant

0.4

0.9

0.3

0.8

0.2

0.7

0.1

0.6 0.5

0 10

0

20

30

1

1

Coral

40

50

GQ

Guano

0.8

0.6

0.6

0.4

0.4

0.2

0.2 0

0

60

20

0

Coral

0.3 0.8 0.2

0.7

0.1

0.6

0

0.5 10

20

30

Depth (cm)

40

50

60

Plant, Guano proportion

Plant

0.9

0

60

80

100

Depth (cm)

Coral proportion

Plant, Guano proportion

Guano

40

0.5

1

JY2

Coral

0.8

Depth (cm) 0.4

Plant

1 Guano

GJ3

0.4

Plant

Coral 0.9

0.3

0.8

0.2

0.7

0.1

0.6

0 0

20

40

60

80

Depth (cm)

Fig. 3. Down-core distributions of guano, coral sand and plant proportions on Jinqing, Jinyin, Ganquan and Guangjin Islands.

0.5 100

Coral proportion

Guano

Plant, Guano proportion

JQ

Coral proportion

Plant, Guano proportion

The pedogenic parent material in the Xisha archipelago is coral debris and thus CaCO3 is the dominant constituent in the ornithogenic sediments. To validate our approach to reconstructing component ratios, we performed statistical analyses on the coral percentages reconstructed by reflectance spectroscopy and

Coral proportion

ratios of plant humus, guano and coral sands in the ornithogenic sediments when the calculated A best fits Asample (tested by least square method). Using the equation given above, we reconstructed the relative proportions of guano, plant and coral sand in each subsample of the profiles GQ, GJ3, JQ and JY2 (Fig. 3).

L.-Q. Xu et al. / Ecological Indicators 23 (2012) 517–523

110 CaCO3

JQ

R = 0.86, n=37, p<0.01

0.9

0.8

Coral sand

90

0.7

80

JQ 0.6

0.5

70 60 20

30

40

50

100 CaCO3

JY2

90

Coral sand

80 70 0 94 92 90 88 86 84 82

10

20

30

40

50

0.3

0.4

1 0.9 0.8 0.7 0.6 0.5 0.4

1

70

60

60 0.9

CaCO3

0.8

Coral sand

0.7

0

20

40

60

80

CaCO3

90

20

0

40

60

80

100

70

80

90

95

0.9 R = 0.85, n=38, p<0.01

0.8 0.7

GJ3

0.6

0.7

0.7

0.5

0.5

0.3

0.3

0.1 120

0.1

GQ 30

100

0.4

0.4

50

90

JY2

0.5 100 0.9

70

100

0.6

0.5

Coral sand

90

R = 0.93, n=54, p<0.01

0.6

GJ3

80

0.8

Coral proportion

10

Coral proportion

0

CaCO3 (%)

1

1.1

100

521

80

85

0.9 R = 0.96, n=54, p<0.01

GQ

20

40

60

80

100

CaCO3 (%)

Depth (cm)

Fig. 4. Correlations between coral proportions from reflectance spectroscopy and CaCO3 levels in the profiles JY2, JQ, GJ3 and GQ.

determined CaCO3 contents by chemical method for the whole profile. The CaCO3 levels in each profile and the correlation analysis results are shown in Fig. 4. The coral percentages are significantly correlated with CaCO3 levels (R > 0.85, p < 0.01) for all four profiles. Phosphorous in ornithogenic sediments is a typical marker of seabird activity (Sun et al., 2000; Xu et al., 2010; Huang et al., 2009). Generally, the down-core record of phosphorous is consistent with guano that obtained from spectrum analysis (Fig. 5), with minor differences in the upper layers of each profile (Fig. 5, shaded area). The reconstructed guano ratios in the sediments below a critical depth of each profile (11 cm, 16 cm, 9 cm and 7 cm for cores GQ, GJ3, JQ and JY2, respectively) are significantly correlated with phosphorous (Fig. 5). However, the correlation in the top sediments of each core seems to be less pronounced. The discrepancy between guano proportion and phosphorous in the top sediments of each profile is perhaps caused by plant development and further migration of phosphorous. According to both field observation and published data, seabird population density has decreased significantly in recent years (Cao et al., 2007). The guano proportion in recent sediments, as reconstructed by spectrum analysis, has also decreased (Fig. 5). Since plant development needs nutrients, a small part of guano in upper layer of the sediments may be consumed. The consumption of guano by plants may be also responsible for the regression of guano in parallel to bird disappearance. As guano has a much higher level of nutrient phosphorous (∼15%) than plant (<0.5%), a small quantity of guano can thus support a large

number of plants. Thus, we suggest that consumption of guano by plant development may not significantly affect the level of guano in the ornithogenic sediments and guano proportion reconstructed by reflectance spectrum can be approximately considered as relative seabird population. In comparison with the rapid decrease in guano ratios in upper layers of each profile, the phosphorous levels in the bulk sediments seem to be constant in recent sediments (Fig. 5, shaded area). This discrepancy may be attributed to development of vegetation on these islands. Although seabird abundance has declined substantially over the past 100 years, nutrient-rich guano possesses a sustainable ability to support vegetation on the islands. The development of vegetation may have resulted in the remigration of bio-elements, e.g. phosphorous, after deposition, and these elements may be further enriched in humus. Indeed, we find that phosphorous levels in plants and humus of the islands are relatively high (Table 2). In terms of seabird population reconstruction, reflectance spectroscope performed better. Thus, reflectance spectroscope can be used as a rapid, non-destructive and effective way to reconstruct seabird population and plant development history on these islands. 3.2. Possible implication for development of tropical coral island ecosystem The changes in source materials in the ornithogenic sediments probably reflect the gradual development of coral island

Table 2 Phosphorous level in plants and humus. Sample

Plant A

Plant D

Humus 1

Humus 2

Humus 3

Humus 4

Humus 5

P (mg kg−1 )

1095

3987

25,652

18239

70,769

2901

3046

522

L.-Q. Xu et al. / Ecological Indicators 23 (2012) 517–523

0.5

JQ

0.4 0.3 0.2

Guano

0.1

P

0 0

10

20

30

40

50

0.5

3

3

2

2

1

1

0

0 -0.1 4

R = 0.90, n=45, p<0.01

JQ Upper 9cm 0.1

0.3

0.5

R = 0.88, n=30, p<0.01

3

3

2

0.3

2

0.2

Guano

0.1

0

0 0

10

20

30

40

50

60 2

0.5

GJ3

0.4

1.5

0.3

1

0.2

Guano

0.1

0.5

P

0 0

20

40

60

0.7

0 100 12 10 8 6 4 2 0 -2

80

GQ

0.5 Guano

0.3

P 0.1 0

20

40

60

80

100

P (%)

P

0

JY2

1

1

P (%)

Guano proportion

4

60 4

JY2

0.4

4

Upper 7cm

-1 0 2

0.1

0.2

0.3

0.4

R = 0.79, n=33, p<0.01

1.6 1.2 0.8

Upper 16cm

GJ3

0.4 0 0

0.1

0.2

0.3

0.4

0.5

15 R = 0.92, n= 48, p<0.01 10

GQ

5

Upper 11cm 0 0

0.2

Depth (cm)

0.4

0.6

0.8

Guano proportion

Fig. 5. Correlations between guano proportions from reflectance spectroscopy and phosphorous levels in the four profiles (shaded area: upper 11 cm, 16 cm, 9 cm and 7 cm for cores GQ, GJ3, JQ and JY2, respectively).

ecosystems. Seabirds act as an effective biological pump, transferring marine-derived nutrients (P, N, etc.) to their surrounding habitats, and nutrient-rich guano can greatly improve soil structure and thus enhance plant development and bioavailability. From Fig. 3, we can see that there is little part of plant, but a relatively high level of guano, in the deep deposits of the cores. Plant proportion increased significantly in upper layer of the sediments (Fig. 3). This suggests that the source materials of the four profiles gradually changed from a two-component system (coral sand and guano) to a three-component mixture (coral sand, guano and humus) in each of the four sediment profiles, reflecting the increasing pedogenesis towards the top sediment layers. We also attempted to examine the possible ecological development of these islands by multi-element geochemical analysis of the bulk sediments from these cores, and found a slow development of vegetation following seabird occupation (Xu et al., 2011b). The results of lithological characteristic and element geochemical study showed a shift of the sediments structure from a two-component to a three-component mixture at the critical depth of each profile (Xu et al., 2011b). Our finding in the present study is thus consistent with our previous study. The soil-forming materials on the coral islets of Xisha archipelago are mainly derived from coral sand, and the pedogenesis on the cay islands has a high affinity with guano input and plant growth (Wang, 2001). The four islands followed similar evolutionary patterns, suggesting that they were probably affected by the same controlling factor(s). We believe that seabird activities played an essential role in soil and sediment developments, by making the island ecosystem more favorable for plant growth. Tropical coral cays with high seabird densities received considerable quantities of inputs of nitrogen (N) and phosphorus (P) from

bird guano, which then enhanced nutrient availability, resulting in an increase in plant productivity and development of insular ecosystems (Anderson and Polis, 1999; Ellis, 2005; Sigurdsson and Magnusson, 2009; Schmidt et al., 2010). Seabird activities appear central to the ecological development of these island ecosystems. 4. Conclusions From well-preserved ornithogenic sediments from Ganquan, Guangjin, Jinqing and Jinyin Islands in the South China Sea, we analyzed the changes in source materials and concluded that: (1) Reflectance spectrum is an ideal indicator for source material levels, and thus could be used as a rapid, non-destructive and effective approach to reconstructing constituent levels in the coral sand ornithogenic sediments on a three-component basis. There is little plant residue, but a relatively high level of guano in deep sediments of the sediment cores. (2) From reconstructed sequential source material changes, it is inferred that plant development lagged behind seabird inhabitation in the tropical coral islands. Seabirds play an essential role in the development of coral reef ecosystems in the Xisha Islands. Seabirds act as a bio-vector, transferring considerable quantities of marine-derived nutrients to coral island ecosystems. Acknowledgments This work was funded by the National Basic Research Program of China (973 Program, No. 2010CB428902), Knowledge Innovation

L.-Q. Xu et al. / Ecological Indicators 23 (2012) 517–523

Program of CAS (KZCX2-EW-QN50), the National Natural Science Foundation of China (No. 41076123) and the Fundamental Research Funds for the Central Universities (WK2060190007). All members of field study team, including the Chinese People’s Liberation Army, are acknowledged for their help in sample collection. Two anonymous reviewers are acknowledged for their constructive comments on earlier version of the manuscript. References Anderson, W.B., Polis, G.A., 1999. Nutrient fluxes from water to land: seabirds affect plant nutrient status on Gulf of California islands. Oecologia 118, 324–332. Bernárdez, P., González-Álvarez, R., Francés, G., Prego, R., Bárcena, M.A., Romero, O.E., 2008. Late Holocene history of the rainfall in the NW Iberian peninsula—evidence from a marine record. J. Marine. Syst. 72, 366–382. Cao, L., Pan, Y.L., Liu, N.F., 2007. Waterbirds of the Xisha Archipelago, South China Sea. Waterbirds 30 (2), 296–300. Chang, Y.P., Wu, S.M., Wei, K.Y., Masafumi, M., Hodaka, K., Chen, M.T., 2005. Foraminiferal oxygen isotope stratigraphy and high-resolution organic carbon, carbonate records from the Okinawa Trough (IMAGES MD012404 and ODP Site 1202). Terr. Atmos. Ocean. Sci. 16 (1), 57–73. Ellis, J.C., 2005. Marine birds on land: a review of plant biomass, species richness, and community composition in seabird colonies. Plant Ecol. 181, 227–241. Exploration Group of Xisha Islands of Institute of Soil Science of Chinese Academy of Sciences (CAS), 1977. Soil and Guano Phosphorus Mine in Xi-Sha. Science Press, Beijing. Foley, W.J., Mcllwee, A., Lawler, I., Aragones, L., Woolnough, A.P., Berding, N., 1998. Ecological applications of near infrared reflectance spectroscopy—a tool for rapid, cost-effective prediction of the composition of plant and animal tissues and aspects of animal performance. Oecologia 116, 293–305. Gong, Z.T., Huang, B., 1997. Pedogeochemistry and bioavailability of soils in South China Sea Islands. Acta Pedol. Sin. 34 (1), 10–27. Hainan Ocean Administration, 1999. The Professional Proceedings of the Integrated Investigation Research on Sea Islands Resource of Hainan Province. China Ocean Press, Beijing (in Chinese with English abstract). Hay, K.B., Millers, K.A., Poore, A.G.B., Lovelock, C.E., 2010. The use of near infrared reflectance spectrometry for characterization of brown algal tissue. J. Phycol. 46, 937–946. Huang, T., Sun, L.G., Wang, Y.H., Liu, X.D., Zhu, R.B., 2009. Penguin population dynamics for the past 8500 years at Gardener Island, Vestfold Hills. Antarct. Sci. 21 (6), 571–578. Kleinebecker, T., Schmidt, S.R., Fritz, C., Smolders, A.J.P., Hölzel, N., 2009. Prediction of ␦13 C and ␦15 N in plant tissues with near-infrared reflectance spectroscopy. New Phytol. 184, 732–739. Lin, X., Chen, X.L., Wang, F., 1999. The Report on Investigation of Sea Island of Hainan Province. China Ocean Press, Beijing. Liu, X.D., Zhao, S.P., Sun, L.G., Luo, H.H., Yin, X.B., Xie, Z.Q., Wang, Y.H., Liu, K.X., Wu, X.H., Ding, X.F., Fu, D.P., 2006. Geochemical evidence for the variation of historical seabird population on the Dongdao Island of South China Sea. J. Paleoliminol. 36, 259–279. Liu, X.D., Sun, J., Sun, L.G., Liu, W.Q., Wang, Y.H., 2010. Feasibility of using reflectance spectroscopy for the analysis of bio-element concentrations in Antarctic ornithogenic sediments. Chin. J. Polar Sci. 21 (1), 31–45.

523

Liu, X.D., Sun, J., Sun, L.G., Liu, W.Q., Wang, Y.H., 2011a. Reflectance spectroscopy: a new approach for reconstructing penguin population size from Antarctic ornithogenic sediments. J. Paleoliminol. 45, 213–222. Liu, X.D., Xu, L.Q., Sun, L.G., Liu, F., Wang, Y.H., Yan, H., Liu, Y., Luo, Y.H., Huang, J., 2011b. A 400-year record of black carbon flux in the Xisha archipelago, South China Sea. Mar. Pollut. Bull. 62, 2205–2212. Liu, X.D., Xu, L.Q., Chen, Q.Q., Sun, L.G., Wang, Y.H., Yan, H., Liu, Y., Luo, Y.H., Huang, J., 2012. Historical change of mercury pollution in remote Yongle archipelago, South China Sea. Chemosphere 87, 549–556. Malley, D.F., Williams, P.C., 1997. Use of near-infrared reflectance spectroscopy in prediction of heavy metals in freshwater sediment by their association with organic matter. Environ. Sci. Technol. 31, 3461–3467. Morton, B., Blackmore, G., 2001. South China Sea. Mar. Pollut. Bull. 42, 1236–1263. Rosén, P., Dåbakk, E., Renberg, I., Nilsson, M., Hall, R., 2000. Near-infrared spectrometry (NIRS): a new tool for inferring past climatic changes from lake sediments. Holocene 10 (2), 161–166. Roumet, C., Picon-Cochard, C., Dawson, L.A., Joffre, R., Mayes, R., Blanchard, A., Brewer, M.J., 2006. Quantifying species composition in root mixtures using two methods: near-infrared reflectance spectroscopy and plant wax markers. New Phytol. 170, 631–638. Schmidt, S., Mackintosh, K., Gillett, R., Pudmenzky, A., Allen, D.E., Rennenberg, H., Mueller, J.F., 2010. Atmospheric concentrations of ammonia and nitrogen dioxide at a tropical coral cay with high seabird density. J. Environ. Monitor. 12, 460–465. Showers, S.E., Tolleson, D.R., Stuth, J.W., Kroll, J.C., Koerth, B.H., 2006. Predicting diet quality of white-tailed deer via NIRS fecal profiling, rangeland. Ecol. Manage. 59, 300–307. Sigurdsson, B.D., Magnusson, B., 2009. Ecosystem respiration, vegetation development and soil nitrogen in relation to breeding density of seagulls on a pristine volcanic island, Surtsey, Iceland. Biogeosci. Discuss. 6, 8393–8409. Stolter, C., Julkunen-Tiitto, R., Ganzhorn, J., 2006. Application of near infrared reflectance spectroscopy (NIRS) to assess some properties of a sub-arctic ecosystem. Basic Appl. Ecol. 7, 167–187. Sun, L.G., Xie, Z.Q., Zhao, J.L., 2000. A 3,000-year record of penguin populations. Nature 407, 858. Wang, G.Z., 2001. Sedimentology of the Coral Reef in South China Sea. Ocean Press, Beijing (in Chinese). Wu, Y.Z., Chen, J., Ji, J.F., Tian, Q.J., Wu, X.M., 2005. Feasibility of reflectance spectroscopy for the assessment of soil mercury contamination. Environ. Sci. Technol. 39, 873–878. Xu, L.Q., Liu, X.D., Sun, L.G., Yan, H., Luo, Y.H., Huang, J., Wang, Y.H., 2010. Distribution of radionuclides in the guano sediments of Xisha Islands, South China Sea and its implication. J. Environ. Radioactiv. 101, 362–368. Xu, L.Q., Liu, X.D., Sun, L.G., Chen, Q.Q., Yan, H., Liu, Y., Luo, Y.H., Huang, J., 2011a. A 700-year record of mercury in avian eggshells of Guangjin Island, South China Sea. Environ. Pollut. 159, 889–896. Xu, L.Q., Liu, X.D., Sun, L.G., Yan, H., Liu, Y., Luo, Y.H., Huang, J., 2011b. Geochemical evidence for the development of coral island ecosystem in the Xisha Archipelago of South China Sea from four ornithogenic sediment profiles. Chem. Geol. 286, 135–145. Xu, L.Q., Liu, X.D., Sun, L.G., Yan, H., Liu, Y., Luo, Y.H., Huang, J., 2011c. A 2200-year record of seabird population on Ganquan Island, South China Sea. Acta Geol. Sin. Engl. 85 (4), 957–967. Zhao, H.T., 1996. History of expeditions to Xisha Islands. Geogr. Res. 15 (4), 55–65.