This study examined the temporal and spatial variations in the occurrence of major algae and algae-related taste and odour (T&O) compounds in Qingcaosha Reservoir, China. Water samples were collected monthly from seven sites in the reservoir and analysed using solid phase microextraction combined with gas chromatography-mass spectrometry. Results showed that green algae were dominant over the 9-month study. In addition, cyanobacteria in the summer and early autumn peaked at 3.2 ∼ 9.6 × 106 cell/L. Among the four most common algae-related T&O compounds, dimethyl trisulphide was undetectable at any of the sampling sites, and 2-methylisoborneol (2-MIB), geosmin and β-cyclocitral were frequently over their respective odour threshold levels from August through October. β-cyclocitral concentration was linearly correlated with the concentration of cyanobacteria (R2 = 0.9831). The peak β-cyclocitral concentrations varied from 452 to 1,293 ng/L and were observed close to the water table. In contrast, high concentrations of 2-MIB and geosmin occurred in the water overlying sediments, probably because some of the 2-MIB and geosmin came from the micro-organisms in the sediments. These results provide important information regarding algae and algae-related T&O compounds in a typical reservoir in estuary and coastal areas, and will be helpful for developing appropriate strategies to minimize undesirable T&O issues in water sources.

INTRODUCTION

Since 2010, the Qingcaosha Reservoir in Shanghai, China has been used as a new drinking water source for China's largest city with a population of over 10 million. The reservoir with an area of 66 km2 has a maximum storage capacity of 4.35 × 108 m3. The designed raw water supply rate of this reservoir is 7.19 × 106 m3/d (Lin et al. 2009) that accounts for over 50% of the current water supply in Shanghai (Zhou et al. 2012). Given that the Qingcaosha Reservoir is located in the estuary of the Yangtze River, the longest river in China, its retention time should be carefully selected to minimize saltwater intrusion. A too short residence time does not allow for completing of an adequate water self-purification prior to entering local water treatment plants, while a too long residence time may lead to undesirable degradation of water quality due to numerous physical, chemical and biological processes such as eutrophication. Among the common water quality issues in reservoirs, taste and odour (T&O), a key aesthetic indicator, is the most frequently complained about, as a result of extremely low odour thresholds of many T&O-induced compounds (typically at a ppt level) and ineffectiveness of traditional drinking water treatment for removal of the T&O compounds (Suffet et al. 1996). Apart from the potentially negative health impacts, T&O present in finished water can cause the public to lose their trust in the local water industry. Therefore, there is an urgent demand to rapidly and efficiently mitigate the T&O problems. To develop appropriate T&O mitigation strategies for a specific reservoir, a clear understanding of occurrence of T&O-induced compounds in the raw water is critically important. Algae-related compounds are a major T&O source. Typical algal T&O-induced compounds include geosmin, 2-methylisoborneol (2-MIB), β-cyclocitral and dimethyl trisulphide (DMTS). Geosmin and 2-MIB are characterized with their well-known earthy-musty flavours (Deng et al. 2012). β-cyclocitral is derived from the carotenoid degradation, and gives off grassy, wooden or tobacco odour at different levels, respectively. Its threshold concentrations in many lakes and reservoirs are within the range of 100–3,000 ng/L (Zhang et al. 2011a). DMTS, a very common T&O-causing compound in China, is mainly responsible for the septic/swampy odour. Recently, DMTS has been found to be produced from Microcystis (Zhang et al. 2011b), particularly during the degradation of Microcystis (Jüttner 1984). In analysis of these T&O-induced compounds in water sources, an accurate and sensitive analytical method of gas chromatography (GC)-mass spectrometry (MS) is applied. Prior to GC/MS analysis, a headspace solid phase microextraction (SPME) technique is often used to collect volatile organic compounds emitted from water (Morales-Valle et al. 2010). The absorbance values at different wavelengths are important indexes to specify the viable algae species, and cell count with the phytoplankton analyser.

Most of the previous studies in the Qingcaosha Reservoir focussed solely upon the relationship of water quality and local climatic and hydrological conditions (Zhang et al. 2009; Zhou et al. 2009; Lu et al. 2011). Data on the occurrence of algae and major algal T&O-induced compounds in the reservoir are highly limited. The objective of this study was to determine the temporal and spatial patterns of occurrence in the predominant algae species and algal-induced T&O compounds in the Qingcaosha Reservoir, and evaluate the impacts of key factors such as nutrients and organic matters upon them.

MATERIALS AND METHODS

Ethics statement

Although this study involved phytoplankton, algae species, or non-living materials, it did not include the use of non-human primates in research. The water samples were collected from the Qingcaosha Reservoir in 2009, when the reservoir was not put into operation as drinking water source. No specific permits were required for the described field studies. Moreover, the field studies involving phytoplankton and algae species did not involve any endangered or protected species. During the field study, the survey was known by and got the oral permission of the local authority (The People's Government of Shanghai City and Shanghai Academy of Environmental Science).

Reagents

All chemicals were at least of analytical grade if not otherwise noted. A total of 100 μg/mL geosmin and 100 μg/mL 2-MIB were prepared in a mixed standard methanol solution (Sigma-Aldrich, USA). DMTS (98%), β-cyclocitral (90%) and internal standard 2-isobutyl-3-methoxypyrazine (IB, 99%) were also purchased from Sigma-Aldrich (USA). Sodium chloride (NaCl) was obtained from Sinopharm Chemical Reagent Co., China, and baked at 450 °C for 2 hours prior to use to remove organic impurities.

Site description

Water samples were collected from seven sites (# 1–7) in the Qingcaosha Reservoir by yacht, as shown in Figure 1. The collection frequency was once a month from April to December 2009. The sampling sites were defined by a global positioning system every time. Locations, water depths, water quality, and sampling times at the seven sites are summarized in detail in Table 1.

Table 1

Locations, water depths, water quality, and sampling times of seven sites, 2009

Sampling site Location Mean depth (m) pH Dissolved organic carbon (DOC, mg/L) Sampling schedule (month/day) 
1# 31 °29′24.8″N, 121 °32′39.5″E 1.3 8.04–8.64 2.82–5.97 04/15, 05/18, 06/15, 07/13, 07/28, 08/13, 08/27, 09/16, 09/28, 10/29, 
2# 31 °28′59.4″N, 121 °36′12.4″E 3.9 8.06–8.59 2.52–4.72 
3# 31 °26′19.9″N, 121 °39′08.4″E 2.4 8.19–8.74 2.52–3.37 
4# 31 °25′05.7″N, 121 °42′33.7″E 7.9 8.15–8.23 1.93–4.34 
5# 31 °27′17.8″N, 121 °34′31.8″E 1.5 8.17–8.51 4.22–9.99 
6# 31 °26′57.4″N, 121 °36′29.5″E 10.2 8.28–8.74 2.73–5.61 
7# 31 °28′20.7″N, 121 °34′36.3″E 2.1 8.36–8.68 2.80–4.53 
Sampling site Location Mean depth (m) pH Dissolved organic carbon (DOC, mg/L) Sampling schedule (month/day) 
1# 31 °29′24.8″N, 121 °32′39.5″E 1.3 8.04–8.64 2.82–5.97 04/15, 05/18, 06/15, 07/13, 07/28, 08/13, 08/27, 09/16, 09/28, 10/29, 
2# 31 °28′59.4″N, 121 °36′12.4″E 3.9 8.06–8.59 2.52–4.72 
3# 31 °26′19.9″N, 121 °39′08.4″E 2.4 8.19–8.74 2.52–3.37 
4# 31 °25′05.7″N, 121 °42′33.7″E 7.9 8.15–8.23 1.93–4.34 
5# 31 °27′17.8″N, 121 °34′31.8″E 1.5 8.17–8.51 4.22–9.99 
6# 31 °26′57.4″N, 121 °36′29.5″E 10.2 8.28–8.74 2.73–5.61 
7# 31 °28′20.7″N, 121 °34′36.3″E 2.1 8.36–8.68 2.80–4.53 
Figure 1

Sampling sites in Qingcaosha Reservoir (sites 1–7#).

Figure 1

Sampling sites in Qingcaosha Reservoir (sites 1–7#).

Sample collection and preparation

In the field sampling, water from 0.5 m below the surface was collected and stored in a 4 L glass bottle that was sealed with a teflon cap with a zero headspace. The bottles were preserved in an icebox packed with ice packs until they were delivered to the environmental analytical laboratory and then stored in a refrigerator at 4 °C.

To analyse the T&O compounds in water, a 50 mL sample without filtration was withdrawn from the sample bottles, and transferred into 80 mL screw capped vials for headspace extraction of dissolved and particle-bound T&O compounds. Then, IB at a concentration of 20 ng/L was added as the internal standard for the quantitative analysis. Subsequently, 15 g of dry NaCl was added and dissolved in the water sample. Thereafter, the vials were incubated in a water bath at 65 °C for 30 minutes before GC/MS analysis. To determine the algal species, 1 L samples from the sample bottles were filtered with 0.45 μm cellulose acetate membranes (50 mm i.d.). The filter residue was again dissolved in 5 mL Milli-Q water. Then, the phytoplankton species in the samples were determined and counted according to a phytoplankton analyser (PHYTO-PAM, Walz, Germany).

Analysis

The T&O compounds (DMTS, 2-MIB, geosmin and β-cyclocitral), as well as the internal standard IB, in water were quantified using the headspace SPME (the fibre No. 57328 U) coupled with a gas chromatograph (GC, QP2010, Shimadzu, Japan) and a mass spectrometer (MS, QP2010S, Shimadzu, Japan). The detailed analysis procedure is available elsewhere (Zhang et al. 2012). In the selected ion monitoring mode, mass-to-charge ratio (m/z) of 126 and 79 for DMTS, m/z of 95 and 108 for 2-MIB, m/z of 112 for geosmin, m/z of 137 and 152 for β-cyclocitral, and m/z of 124 and 94 for IB were monitored, respectively. Total nitrogen (TN) and total phosphorus (TP) were measured according to the national standard methods (GB11894-89, GB11893-89, respectively).

RESULTS AND DISCUSSION

Algae species and TN/TP

Figure 2 presents data on the temporal variations of different algae species in the study period. Over the 9 months of this study, the average monthly algal cell count was approximately 107 cell/L, but the cell counts were highly time specific. The monthly cell counts during July–October 2009 were higher than those observed in other months. Particularly, the monthly cell count at 1.8 × 107 cell/L peaked in August 2009. At any specific month, diatoms, green algae and cyanobacteria were three dominant algal species. When the temperature was low, such as in April, November and December 2009, diatoms accounted for 53–62% of overall algal cells, probably because diatoms have a low requirement in solar irradiation and temperature compared with other species (Hakanson et al. 2003). In contrast, during August–October 2009, cyanobacteria gradually prevailed with a fraction of 32–53% in the overall algal cells. The finding suggests that a warm climatic condition favours the growth of cyanobacteria. In any month of this study, green algae were popular in the Qingcaosha Reservoir, with a fraction of >20% in the overall algae.

Figure 2

Variations in algae populations in Qingcaosha Reservoir from April to December 2009.

Figure 2

Variations in algae populations in Qingcaosha Reservoir from April to December 2009.

Nitrogen (N) and phosphorus (P) are two principal limiting nutrients for aquatic algal production due to their short supply compared to demands of cellular growth (Lv et al. 2011). TN at most of the seven sites peaked in April (1.38–1.89 mg/L) and August (1.54–2.08 mg/L), as a result of the nitrogen fixation and denitrification in the ecosystems (Gao et al. 2009). In August, the algae blooming maximized the nitrogen fixation dominant. In April, phosphorus became the limiting major nutrient for the algae growth, and dissolved phosphate could be absorbed by algae (Chen et al. 2009). The concentrations of TP in all the sampling sites were below 0.2 mg/L, and thus the water quality met with the Class III of Chinese surface water environmental quality standards.

The mean ratios of TN to TP (TN/TP) at all the seven sampling sites at different months are shown in Figure 3. Over the study period, TN/TP varied between 11.3 and 40.9. During August–November when the overall algae were abundant, TN/TP was approximately 15, lower than TN/TP in other months. The observation was in agreement with findings in many other studies that a high concentration of phosphorus, along with a low N/P supply ratio, favoured algae blooming (Liu et al. 2011). Sakamoto (1966) proposed that the phytoplankton biomass relied upon TP at TN/TP > 17, upon TN at TN/TP < 10, and upon both TN and TP at TN/TP = 10–17. Based on the criteria, the limiting nutrients for the phytoplankton production were both TN and TP in September, October, and November, while only TP was the control nutrient in other months.

Figure 3

Changes in TN/TP ratios in Qingcaosha Reservoir, 2009.

Figure 3

Changes in TN/TP ratios in Qingcaosha Reservoir, 2009.

Temporal variation of T&O compounds

Quantitative analysis of the selected T&O compounds, including DMTS, 2-MIB, geosmin and β-cyclocitral, was performed using SPME combined with GC/MS. Over the study period, DMTS was not detectable in all the sites, thereby indicating that algae rarely decayed because dissolved oxygen (DO) was always maintained at a level as high as 6–9 mg/L.

Total 2-MIB, geosmin and β-cyclocitral concentrations over time in the seven sampling sites are shown in Figure 4(a)4(c), respectively. Generally, high concentrations of 2-MIB, geosmin and β-cyclocitral typically occurred in summer and early autumn. In particular, the maximum concentrations of 2-MIB (277.7 ng/L) and geosmin (29.2 ng/L) were observed in site #5 in September. These findings were significantly greater than their respective odour threshold levels of 6–42 ng/L (2-MIB) and 4–20 ng/L (geosmin) (Jones & Korth 1995). As shown in Figure 1, site #5 near the bank is located in a downwind area. 2-MIB and geosmin are produced primarily by cyanobacteria, and incidentally by others such as actinomycete (Watson 2003; Zaitlin & Watson 2006). In this study, it was noted that the concentrations of 2-MIB and geosmin were positively correlated with the concentrations of organic matters (e.g. DOC = 4.22–9.99 mg/L at site #5) and algae 5# (7.5 × 107 cell/L at site #5).

Figure 4

Variations of T&O compound concentrations in sampling sites of reservoir, 2009 ((a) 2-MIB, (b) geosmin, (c) β-cyclocitral).

Figure 4

Variations of T&O compound concentrations in sampling sites of reservoir, 2009 ((a) 2-MIB, (b) geosmin, (c) β-cyclocitral).

In contrast, the maximum concentrations of β-cyclocitral (244.9–1293.3 ng/L) were observed in August 2009. Among the seven sampling sites, the highest β-cyclocitral level occurred at site #6 where the fraction of cyanobacteria reached 53% of the overall algae though the total algae concentration was not the highest. The relationship between the three T&O compounds and cell densities of different algae is shown in Table 2. As seen, the concentrations of 2-MIB, geosmin and β-cyclocitral were linearly correlated with the densities of cyanobacteria with R2 = 0.7447, 0.6378 and 0.9831, respectively. The production quota of 2-MIB, geosmin and β-cyclocitral were approximately 4, 0.6 fg/cell and 50 fg/cell, respectively, in the Qingcaosha Reservoir. The strongest correlation between cyanobacteria and β-cyclocitral further demonstrated that β-cyclocitral was produced principally by cyanobacteria, particularly Microcystis. Jüttner & Hoflacher (1985) revealed that β-cyclocitral was the product of an oxidative cleavage reaction of β-carotene bound on Microcystis. In our Taihu Lake samples, the strong correlation was also observed between the Microcystis cells and β-cyclocitral concentrations (R2 = 0.95) (data not shown here). Therefore, under field conditions, a β-cyclocitral concentration may be an excellent indicator to estimate the cell concentrations of Microcystis that are not conveniently and rapidly identified and analysed, when the cells are roughly in the same growth phase.

Table 2

Significant regression relationships between the T&O compounds and different algae

Related coefficient (R22-MIB Geosmin β-cyclocitral 
Diatoms 0.0006 0.0232 0.0138 
Green algae 0.4753 0.2244 0.2941 
Cyanobacteria 0.7447 0.6378 0.9831 
Related coefficient (R22-MIB Geosmin β-cyclocitral 
Diatoms 0.0006 0.0232 0.0138 
Green algae 0.4753 0.2244 0.2941 
Cyanobacteria 0.7447 0.6378 0.9831 

Vertical profiles of algae and T&O compounds

The vertical profiles of algae and their T&O compounds over the water depth are shown in Figure 5(a)5(d) (site #6, July–December 2009). As seen in Figure 5(a), the cell concentrations of algae decreased with the increasing depth in the reservoir over all of the study period. The finding is likely because the growth of algal cells was limited by low water temperature and reduced light intensities in deep water. In Figure 5(b), the vertical patterns of β-cyclocitral concentrations were synchronized with the profiles of algal densities, cyanobacteria in particular. The highest concentration was observed at a depth of 0.5 m, and varied within 452–1293 ng/L from August to October. The vertical profiles of 2-MIB concentrations in this study are shown in Figure 5(c). Of note, the concentrations of 2-MIB at 0.5 m were extremely high from the July to the November relative to the concentrations observed at the other two depths, with a peak at 39 ng/L in the November. The trend was correlated with the patterns of algae blooms in surface water. The 2-MIB concentrations at 5.5 and 10 m depths were very close at any specific month, all below 10 ng/L. Finally, geosmin exhibited a different vertical distribution over the water depth, as shown Figure 5(d). The highest geosmin concentrations were observed at 10 m (0.7–5.7 ng/L), followed by geosmin at 0.5 m (0.7–5.5 ng/L). The lowest concentrations occurred at the depth of 5.5 m (0.2–2.3 ng/L). Different from β-cyclocitral, the sediments closest to the 10 m depth could contribute 2-MIB and geosmin into water. Previous studies showed that actinomycetes were mainly found in terrestrial environments and could produce 2-MIB and geosmin that might enter the overlying water near sediments (Zaitlin & Watson 2006). Another study reported that the concentrations of 2-MIB and geosmin near to sediments were greater than those in surface water in the Xionghe Reservoir, China (Zuo et al. 2010). As seen from Table 2, less of a correlation between 2-MIB (R2 = 0.7447)/geosmin (R2 = 0.6378) and cyanobacteria was observed than that between β-cyclocitral and cyanobacteria (R2 = 0.9831). The observation reflected that 2-MIB and geosmin might derive from a diversity of micro-organisms, fungi and phytoplankton. Similarly, Dzialowski et al. found that cyanobacterial biovolume is not a consistent predictor of dissolved geosmin (Dzialowski et al. 2009). Zuo et al. (2010) found cyanobacteria, actinomycetes, and other odour-producers were not responsible for 2-MIB occurring in one reservoir. Therefore, the sources of geosmin and 2-MIB continue to remain more complicated than β-cyclocitral.

Figure 5

Concentration profiles of algae and T&O compounds in reservoir, 2009 ((a) algae, (b) β-cyclocitral, (c) 2-MIB, (d) geosmin).

Figure 5

Concentration profiles of algae and T&O compounds in reservoir, 2009 ((a) algae, (b) β-cyclocitral, (c) 2-MIB, (d) geosmin).

CONCLUSIONS

This study provided key information regarding the temporal and spatial variations in the occurrence of major algal species and algae-induced T&O compounds in Qingcaosha Reservoir, China. Three principal algae species, including diatoms, green algae and cyanobacteria, were observed in the 9-month study. Over all of the study period, green algae were present. From the August to the October of 2009, cyanobacteria gradually became dominant, with a 32–53% fraction in the overall algae. Among the common algae-induced T&O compounds, DMTS was not detected at any of the sampling sites due to a high DO in the water, but three other T&O compounds were observed, including β-cyclocitral, 2-MIB, and geosmin. β-cyclocitral concentrations were strongly correlated with the cell number of cyanobacteria (R2 = 0.9831). Therefore, a β-cyclocitral concentration may be a good indicator to estimate the cell number of cyanobacteria in a reservoir. High concentrations of 2-MIB and geosmin were found near sediments of the reservoir, thereby indicating that 2-MIB and geosmin might originate from not only phytoplankton, but also from micro-organisms and fungi in the sediments. The results added to our knowledge of the occurrence and fate of major algae and algal-related T&O compounds in the reservoirs located in estuary and coastal areas, and provided a foundation to develop appropriate strategies to minimize the algae-caused T&O issues in the raw water.

ACKNOWLEDGEMENTS

This project was supported jointly by National Natural Science Foundation of China (51208456); The National Major Science and Technology Project of China (2012ZX07403-003); The Programme of Zhejiang Leading Team of S&T Innovation (2010R50037); and Foundation of Key Disciplines of Zhejiang University of Technology (No. 20130307).

REFERENCES

REFERENCES
Chen
S. N.
Chen
X. L.
Peng
Y.
Peng
K. D.
2009
A mathematical model of the effect of nitrogen and phosphorus on the growth of blue-green algae population
.
Applied Mathematical Modelling
33
(
2
),
1097
1106
.
Dzialowski
A. R.
Smith
V. H.
Huggins
D. G.
deNoyelles
F.
Lim
N. C.
Baker
D. S.
Beury
J. H.
2009
Development of predictive models for geosmin-related taste and odour in Kansas, USA, drinking water reservoirs
.
Water Research
43
(
11
),
2829
2840
.
Gao
Y. X.
Zhu
G. W.
He
R. R.
Wang
F.
2009
Variation of water quality and trophic state of Lake Tianmu, China
.
Environmental Science
30
(
3
),
673
679
.
Jones
G. J.
Korth
W.
1995
In situ production of volatile odour compounds by river and reservoir phytoplankton
.
Water Science and Technology
31
(
11
),
145
151
.
Jüttner
F.
1984
Characterization of Microcystis strains by alkyl sulfides and β-cyclocitral
.
Journal of Biosciences
39
(
9–10
),
867
871
.
Lin
W. Q.
Gu
Y. L.
Lu
S. Q.
Wang
X. P.
2009
Reasonable hydraulic retention time of Qingcaosha Reservoir to prevent algae bloom
.
Water and Wastewater Engineering
35
(
9
),
60
62
.
Lu
N.
Huang
X.
Gao
N. Y.
Liu
H. B.
Zhang
D.
2011
Bromonium ion and its bromate formation potential in raw water of Qingcaosha Reservoir
.
Water Purification Technology
30
(
3
),
10
12
,
19
.
Morales-Valle
H.
Silva
L. C.
Paterson
R. R. M.
Oliveira
J. M.
Venancio
A.
Lima
N.
2010
Microextraction and gas chromatography/mass spectrometry for improved analysis of geosmin and other fungal “off” volatiles in grape juice
.
Journal of Microbiological Methods
83
(
1
),
48
52
.
Sakamoto
M.
1966
Primary production by phytoplankton community in some Japanese Lakes and its dependence on lake depth
.
Archives of Hydrobiologica
62
,
1
28
.
Suffet
I. H.
Corado
A.
Chou
D.
McGuire
M. J.
Butterworth
S.
1996
AWWA Taste and odor survey
.
Journal of the American Water Works Association
88
(
4
),
168
180
.
Zhang
H. W.
Wu
J.
Che
Y.
Li
W.
Tong
C. F.
2009
Environmental impact assessment of Qingcaosha Reservoir construction project in the Yangtze Estuary
.
Journal of East China Normal University (Natural Science)
(
3
),
38
47
.
Zhang
K. J.
Gao
N. Y.
Yen
H. K.
Chiu
Y. T.
Lin
T. F.
2011a
Degradation and formation of wood odorant β-cyclocitral during permanganate oxidation
.
Journal of Hazardous Materials
194
(
30
),
362
368
.
Zhang
K. J.
Gao
N. Y.
Deng
Y.
Shui
M. H.
Tang
Y. L.
2011b
Granular activated carbon (GAC) adsorption of two algal odorants, dimethyl trisulfide and β-cyclocitral
.
Desalination
266
(
1–3
),
231
237
.
Zhang
K. J.
Gao
N. Y.
Deng
Y.
Zhang
T. Q.
Li
C.
2012
Aqueous chlorination of algal odorants: reaction kinetics and formation of disinfection by-products
.
Separation and Purification Technology
92
(
18
),
93
99
.
Zhou
J. J.
Gao
N. Y.
Zhao
S. J.
Gu
Y. L.
Le
Q.
Wang
X. P.
2009
Nitrogen and phosphorus dynamic changing features of raw water before Qingcaosha reservoir operation
.
Water and Wastewater Engineering
36
(
12
),
49
52
.
Zhou
C.
Gao
N. Y.
Zhao
S. J.
Chu
W. H.
2012
Evaluation of water quality in Qingcaosha Reservoir of Shanghai
.
Journal of Tongji University (Natural Science)
40
(
6
),
894
899
.
Zuo
Y. X.
Li
L.
Zhang
T.
Zheng
L. L.
Dai
G. Y.
Liu
L. M.
Song
L. R.
2010
Contribution of Streptomyces in sediment to earthy odor in the overlying water in Xionghe Reservoir, China
.
Water Research
44
(
20
),
6085
6094
.