This study aimed to investigate the relationships between atmospheric parameters, seawater quality and green mussel production which were cultured in pond, estuary and coastal areas. Seawater and mussel samples were collected from mussel farms in the inner Gulf of Thailand from January to December 2019. Climate data were obtained from the Thai Meteorological Department. The correlations between selected atmospheric and seawater parameters were developed using linear and nonlinear models. The influence of seawater quality on mussel production was evaluated using principal component analysis and stepwise multiple linear regression. The effects of atmospheric variation on green mussel productivity were simulated. The results showed that high air temperature and rainfall caused an increase in seawater temperature and a decrease in salinity, respectively. It was observed that the most influential factors affecting mussel production were nutrients and dissolved oxygen in ponds, temperature and salinity in estuaries, and nutrients and pH in coastal areas. The simulation indicated that mussel production can deteriorate when air temperature reaches 34°C and rainfall is higher than 200 mm per month. Our results suggest that under climate change events, locations with less riverine influence can provide higher mussel productivity. These results can be used as a guideline for farmers during a climate change event.
Esta investigación tuvo como objetivo investigar las relaciones entre los parámetros atmosféricos, la calidad del agua de mar y la producción de mejillones verdes que se cultivaron en estanques, estuarios y áreas costeras. Se recolectaron muestras de agua de mar y mejillones de granjas de mejillones en el interior del Golfo de Tailandia de enero a diciembre de 2019. Los datos climáticos se obtuvieron del Departamento Meteorológico de Tailandia. Las correlaciones entre parámetros atmosféricos y de agua de mar seleccionados se desarrollaron utilizando modelos lineales y no lineales. La influencia de la calidad del agua de mar en la producción de mejillones se evaluó mediante análisis de componentes principales y regresión lineal múltiple paso a paso. Se simularon los efectos de la variación atmosférica sobre la productividad del mejillón verde. Los resultados mostraron que la alta temperatura del aire y las precipitaciones provocaron un aumento de la temperatura del agua de mar y una disminución de la salinidad, respectivamente. Se observó que los factores más influyentes que afectaron la producción de mejillón fueron los nutrientes y el oxígeno disuelto (OD) en los estanques; temperatura y salinidad en estuarios; y nutrientes y pH en zonas costeras. La simulación indicó que la producción de mejillón puede deteriorarse cuando la temperatura del aire alcanza los 34°C y la precipitación supera los 200 mm por mes. Nuestros resultados sugieren que, bajo eventos de cambio climático, los lugares con menos influencia fluvial pueden proporcionar una mayor productividad de mejillón. Estos resultados se pueden utilizar como guía para los agricultores durante un evento de cambio climático.
A higher frequency and magnitude of extreme weather events were reported during the 20^{th} century (
Many researchers have documented that surface water quality is highly sensitive to climate. Its physical, chemical and biological properties respond rapidly to climaterelated changes (
Green mussel (
We hypothesize that disturbances in atmospheric parameters such as air temperature and rainfall cause changes in seawater quality and subsequently a decrease in green mussel productivity. The novelty of the research is that it combines and links atmosphere variation, seawater quality and mussel production. A literature review reveals that no reported scientific research combines all three parameters and investigates impacts on mussel production. Moreover, differences in cultivation methods (for pond, estuary and coastal areas) can also affect green mussel production in different ways.
Therefore, this study emphasizes how climate change affects green mussels cultivated by different methods. The objectives of the study are 1) to investigate correlations between atmospheric parameters and seawater quality; 2) to evaluate the influence of seawater quality on green mussel production; and 3) to simulate the effect of atmospheric variations (air temperature and rainfall) on green mussel productivity with mussels cultured in pond, estuary and coastal areas. The results of our study can provide information to farmers on how the cultivation method used is sensitive to climate change events. They can also assist farmers and provide adaptation strategies for green mussel production under extreme weather events occurring due to global warming and climate change.
In this study, mussel farms in the inner Gulf of Thailand, which is an important green mussel production area in Southeast Asia, were selected. Three cultivation areas and methods which are practised in Thailand (pond, estuarine and coastal areas) were chosen to investigate how each method will be affected by climate change and in which of them extreme events will have the most impact. To compare the climate change effect between these methods, Bangkhuntein (BKT), Maeklong river mouth (MK), and Sriracha (SRI) were chosen as a case study for green mussel cultivation in pond, estuary and coastal areas, respectively (
A, inner Gulf of Thailand and location map showing sampling stations for each green mussel farm; B, Bang Khun Tien (BKT); C, Maeklong river mouth (MK); and D, Sriracha (SRI). Panels B, C and D represent the pond, estuary and coastal areas, respectively.
Station  Cultivation method  Farm area (m^{2})  No. of cultivation ropes  Depth (m) 

BKT 1  floating raft  26935  336  2.0 
BKT 2  floating raft  26077  3280  2.5 
BKT 3  floating raft  179  48  1.8 
MK 1  horizontal rack  8064  33660  6.6 
MK 2  horizontal rack  360  1680  3.0 
MK 3  longline  2000  4200  3.0 
SRI 1  floating raft  1344  891  3.9 
SRI 2  floating raft  1224  1460  4.5 
SRI 3  floating raft  1716  1152  4.5 
The first sampling location was in Bang Khun Thian district, Bangkok (UTM unit zone 47P 655299 m E, 1497465 m N). At this location, green mussels are cultured in ponds. Juvenile mussels (shell length 2 cm) were brought from Chonburi province, Thailand. They were attached to cultivation ropes and hung on floating rafts. Seawater was circulated by opening the gate at high tide and closing the gate at low tide each day. The mussel farm is located inland at a distance of 4.69 km from the seashore. As shown in
Another sampling location, where mussels are cultivated in an estuarine area, was the Maeklong river mouth, Samut Songkhram province (UTM unit zone 47P 612675 m E, 1475015 m N). This mariculture site has riverine influences. The juvenile mussels were obtained from this site. Three survey locations were selected across a wide area of the Maeklong river mouth (
Green mussel farms in Sriracha, Chonburi province (UTM unit zone 47P 708208 m E, 1458524 m N) were selected for mussel cultivation in a coastal area. This cultivation site is outside of estuarine influences and is about 26 to 35 km south of Bang Pakong estuary. As shown in
Green mussels and seawater samples were collected monthly during the low tide from January to December 2019. All mussels that were attached to one cultivation rope were collected from each sampling station. Epifauna and attachments were removed from the mussels. The fresh weight of the flesh of each mussel was measured using a 4digit analytical balance. Then, the mussel production (kg m^{2}) for each mussel farm was determined by multiplying the fresh mussels’ weight by the number of cultivations ropes and dividing by the total farm area (
At each station, a 1 L sample of seawater was collected in a glass bottle. Then, the seawater temperature, pH, salinity, conductivity and dissolved oxygen (DO) were measured in situ using a handheld multiparameter probe (YSI 600QS, USA). Salinity was measured in practical salinity units (psu). Transparency was analysed using a Secchi disc. Seawater samples were brought to the environmental laboratory at Bansomdejchaopraya Rajabhat University, Bangkok, Thailand for chlorophyll
In this study, seawater quality includes only the water variables that may affect green mussel production. To investigate the correlations between atmospheric parameters and seawater quality in the study area, atmospheric data were used. The average of the highest air temperature (HAT) and the monthly total rainfall were obtained from the Thai Meteorological Department (TMD). Then, in each station, the atmospheric data were correlated with seawater parameters (temperature, pH, salinity, DO, NH_{4} ^{+}N, NO_{2} ^{}+NO_{3} ^{}N, PO_{4} ^{3}P and Si(OH)_{4}Si) from monthly field observations for one year (n=12) by model fitness (linear and nonlinear) using Microsoft Excel 365 software.
Analysis of variance (ANOVA) and TurkeyHSD were used to test for the spatial and temporal variations of green mussel production and seawater quality. The sample size (n) was 108 and the degree of freedom (df) was 8 for spatial analysis and 11 for temporal analysis. Principal component analysis (PCA) was used to determine the main parameters that cause changes in the environment at each sampling site (BKT, MK and SRI). Only principal components (PCs) with an eigenvalue exceeding 1 were analysed with Varimax rotation. Then, multiple linear regressions were performed to evaluate the influence of seawater quality on green mussel production. Based on the results from PCA (PC 1) and stepwise analysis, only significant variables were used in the regression models. Moreover, cultivation duration and operation methods were different at all stations, resulting in a different age of the mussels between stations. Thus, in addition to seawater quality, cultivation time is another important parameter affecting green mussel density. Hence, to obtain reliable results, cultivation time was included in the regressions analysis. All statistical parameters were computed using IBM SPSS Statistic Version 23 for Windows.
To simulate the effects of atmospheric variation on green mussel productivity, seawater parameters were correlated with the variation in monthly rainfall or the HAT using a multiple linear regression model to evaluate the mussel production. In this study, the models were simulated under two different scenarios: 1) drought conditions; and 2) storms or rapid heavy rainfall. In the first scenario, the variation in HAT was considered with fixed rainfall of 0.1 mm (lowest data used in the models). In the second scenario, the variation in monthly rainfall was considered with fixed HAT. For both scenarios, simulations were done for 12 months of cultivation time, and the effects were assessed on the healthiest mussels that were ready for harvest.
The results indicated that variations in air temperature and rainfall caused changes in seawater quality (
Results from statistical analysis showed that green mussel production and seawater parameters in the three cultivation areas (BKT, MK and SRI) were significantly different, except for seawater temperature and Chl
Oneway ANOVA  TurkeyHSD  

Parameter  BKT  MK  SRI  Cultivation Areas  BKT  MK  SRI  
GM (kg m^{2})  0.11  ±  0.13  12.64  ±  8.29  1.03  ±  0.67  ***  a  b  a 
Temp (^{o}C)  29.27  ±  2.20  29.13  ±  2.19  29.76  ±  1.80  ns       
pH  7.68  ±  0.55  8.16  ±  0.26  8.21  ±  0.33  ***  a  b  b 
Sal  21.56  ±  3.52  23.91  ±  5.60  28.14  ±  3.10  ***  a  a  b 
Conduct (ms)  25.43  ±  2.18  25.85  ±  4.13  30.12  ±  3.00  ***  a  a  b 
DO (mg L^{1})  4.87  ±  1.74  5.74  ±  0.95  5.95  ±  0.90  **  a  b  b 
Trans (cm)  131.17  ±  57.32  119.09  ±  77.90  230.45  ±  98.17  ***  a  a  b 
TSS (mg L^{1})  32.75  ±  32.36  26.54  ±  15.08  19.48  ±  5.91  *  b  a,b  a 
NH_{4} ^{+}N (µg L^{1})  200.01  ±  143.17  90.78  ±  73.91  67.23  ±  66.97  ***  b  a  a 
NO_{2} ^{}+NO_{3} ^{}N (µg L^{1})  50.12  ±  31.44  23.63  ±  19.63  7.59  ±  8.18  ***  a  c  b 
PO_{4} ^{3}P (µg L^{1})  77.23  ±  59.76  13.39  ±  12.41  5.75  ±  4.83  ***  b  a  a 
Si(OH)_{4}Si (µg L^{1})  1151.98  ±  716.70  492.08  ±  261.29  657.71  ±  300.80  ***  b  a  a 
Chl 
10.31  ±  16.21  8.24  ±  7.05  5.62  ±  7.20  ns       
Significance level of oneway ANOVA: ***, P <0.001; **, P <0.01; *, P <0.05; ns = not significant Groups of data from TurkeyHSD are shown as a, b and c
Results from PCA revealed that the three cultivation areas were differently affected by seawater variables. The PCs with eigenvalues higher than 1 and the variables whose loading exceeded 0.5 were mentioned. As shown in
For mussel cultivation in an estuarine area (MK), salinity, conductivity, and temperature were identified as the most important component (PC 1), with 25.35% of the variance, followed by TSS, transparency and Chl
For the coastal area (SRI), nutrient concentrations (NH_{4}
^{+}N, PO_{4}
^{3}P and Si(OH)_{4}Si) were positively correlated and pH was negatively correlated. The most important components explained 32.32% (PC 1) and 20.56% (PC 2) of the variance (
PCA and stepwise analysis were employed to determine the most important parameters affecting green mussel production in each cultivation area. The selected parameters were used in the multiple linear regression model. As shown in
Model equation  Multiple R  Rsquared  Adjusted Rsquared  Standard error 


0.9885  0.9772  0.9202  0.0021 

0.9187  0.8439  0.7139  0.0183 

0.9334  0.8711  0.7101  0.0741 

0.9128  0.8332  0.7359  0.2487 

0.7742  0.5994  0.2656  4.6909 

0.8425  0.7098  0.4776  1.8867 

0.9676  0.9363  0.8725  0.2975 

0.8562  0.7331  0.5106  0.3273 

0.9185  0.8437  0.7135  0.0797 
Note:
Results from model simulations on atmospheric parameters affecting seawater quality (
It was found that increased air temperature resulted in a higher seawater temperature, and the correlations fitted well with simple linear regression (
Rainfall can increase runoff, increasing the cations in seawater and subsequently increasing the pH level. In the pond area (BKT), intense rainfall corresponded to high pH in seawater, and the plateau was found with rainfall of over 50 mm per month (
This study found that high rainfall caused a decrease in salinity that can be described by simple linear regression (
Rainfall can cause an increase or decrease in DO. Our results revealed that at sampling stations MK 1 and SRI 3 intense rainfall resulted in high DO in seawater. This is because there were more wind and waves than at other stations. This finding is in line with a study on gas exchange across an airwater interface by
The results of this study also revealed that rainfall causes a decrease or increase in nutrient concentrations. During high rainfall periods, low nutrient contents were observed in the pond area (BKT), with the exception of silicatesilicon.
It was found that seawater quality in pond, estuary and coastal areas was significantly different and affected green mussel production in different ways. Research on greenlipped mussel
In an estuarine area (MK), our results from PCA demonstrated that variation of temperature and salinity were the most important parameters. Salinity and temperature are affected by freshwater discharges through river runoff. In line with our findings,
In this study, the dominant parameters affecting environmental variability in a coastal area (SRI) were pH, NH_{4}
^{+}N, PO_{4}
^{3}P and Si(OH)_{4}Si. These variables were found to have a significant effect on green mussel production. This is because the study area is located farther from the river mouth. There was less input of freshwater, and the riverine influences were reduced along the coast. Therefore, temperature and salinity become less impactful factors. Seawater quality in this area is associated with precipitation and water circulation inside the inner Gulf of Thailand. The amount of rainfall affects the variation of pH levels in seawater (
The results of the simulations indicated that at all sampling stations variations in air temperature and rainfall influence green mussel production in different ways. For mussels cultured in a pond, drought strongly affected green mussel production (
Based on simulation results, drought did not influence green mussel production in estuary (MK) and coastal (SRI) areas (
Scenarios of green mussel production under variations of rainfall showed that intense rainfall resulted in a decrease in production of green mussel cultured in a pond (
Surprisingly, the simulations revealed that heavy rainfall did not affect green mussels that were cultured in the estuary area (
For a coastal area, green mussel production was not affected by heavy rainfall, except at SRI 1 (
The results of field observation demonstrated that the estuary area (MK) had significantly the highest green mussel production (
Although some simulations showed that variations in air temperature or rainfall do not affect green mussel production, there are threshold limits. To estimate atmospheric parameters that cause the death of green mussels, the threshold levels of seawater quality (
Seawater parameter  Maximum lethal level^{*}  Upper survival level  Lower survival level  Minimum lethal level^{*} 

Temperature (°C)  38.00  36.00  14.00  8.00 
pH  9.00  8.65  7.54  3.50 
Salinity  80.00  65.00  25.00  24.00 
DO (mg L^{1})    8.00  4.00  0.50 
Note: *Lethal level is the level of seawater parameters that causes 50% to 100% dieoff of green mussels.
It can be concluded that climate change affects green mussel productivity differently according to the cultivation area. Atmospheric parameters affect seawater quality in pond, estuary and coastal areas in different kinetics. A high air temperature significantly increased the seawater temperature. Intense rainfall results in a decrease in salinity and an increase in pH except at MK 3, where the pH of water can be lowered by a high quantity of river runoff. Increased DO was found during heavy rainfall. However, it decreased with excessive suspended sediments and dissolved organic matter caused by water discharges. Rainfall caused variations in nutrient concentrations that were diluted under unpolluted conditions and concentrated by the dispersion of domestic waste and soil erosion. Additionally, results also indicated that the most influential seawater parameters on green mussel production in pond, estuary and coastal areas were different. This can be explained by the results of PCA and the multiple linear regression analysis. Nutrients and DO significantly affected the mussels cultured in a pond. In an estuary, temperature and salinity were influential factors due to freshwater input caused by river runoff. Nutrient and pH levels were found to be more important in coastal waters. Finally, the cultivation areas most sensitive to droughts and heavy storms were ponds, followed by estuary and coastal areas. Drought conditions with high air temperature and negligible rainfall strongly affected the mussel production in the pond. Tropical storms cause intense rainfall that significantly affects mussel production close to the river mouth. This study suggests that cultivating in a coastal area or a large distance from the river mouth can effectively produce green mussels under extreme weather events caused by climate change.
This research was financially supported by the Thailand Research Fund (TRF) and the Office of the Higher Education Commission (OHEC), Grant No. MRG6180223. The authors are also thankful to the Environmental Science and Technology Programme, Faculty of Science and Technology, Bansomdejchaopraya Rajabhat University for providing laboratory space, and the Sriracha Fisheries Station, Faculty of Fisheries, Kasetsart University for providing the green mussels and samples of seawater in this study.
The following supplementary material is available through the online version of this article and at the following link:
Table S1.  Simple linear regression and logarithm model equations and coefficient of determination (R2) show correlations between atmospheric parameters and seawater parameters in each sampling station at Bangkhuntein (BKT), the Maeklong river mouth (MK) and Sriracha (SRI).
Table S2.  Spatial and temporal variation of green mussel production and seawater parameters in the mussel cultivation areas at Bangkhuntein (BKT) by using oneway ANOVA and TurkeyHSD.
Table S3.  Spatial and temporal variation of green mussel production and seawater parameters in the mussel cultivation areas at the Meaklong River Mouth (MK) by using oneway ANOVA and TurkeyHSD.
Table S4.  Spatial and temporal variation of green mussel production and seawater parameters in the mussel cultivation areas at Sriracha (SRI) by using oneway ANOVA and TurkeyHSD.
Equation  R^{2}  Equation  R^{2} 

T_{BKT1} = 2.6753(AHT)  52.986  R² = 0.7074  A_{BKT1} = 101.40ln(R) + 637.28  R² = 0.9090 
T_{BKT2} = 2.4945(AHT)  47.59  R² = 0.7454  A_{BKT2} = 69.46ln(R) + 478.61  R² = 0.8717 
T_{BKT3} = 2.634(AHT)  51.855  R² = 0.7494  A_{BKT3} = 75.00ln(R) + 508.88  R² = 0.8829 
T_{MK1} = 2.5276(AHT)  52.091  R² = 0.7621  A_{MK1} = 52.42ln(R) + 272.73  R² = 0.8908 
T_{MK2} = 2.4532(AHT)  49.874  R² = 0.8169  A_{MK2} = 43.53ln(R) + 242.78  R² = 0.8336 
T_{MK3} = 2.1928(AHT)  41.565  R² = 0.7790  A_{MK3} = 44.47ln(R) + 242.92  R² = 0.8814 
T_{SRI1} = 1.2173(AHT)  11.377  R² = 0.8820  A_{SRI1} = 29.76ln(R) + 171.12  R² = 0.7063 
T_{SRI2} = 1.6318(AHT)  24.948  R² = 0.9107  A_{SRI2} = 27.61ln(R) + 156.72  R² = 0.7975 
T_{SRI3} = 1.7192(AHT)  27.999  R² = 0.8137  A_{SRI3} = 32.21ln(R) + 174.79  R² = 0.7133 
pH_{BKT1} = 0.3787ln(R) + 6.1702  R² = 0.8532  N_{BKT1} = 8.232ln(R) + 79.446  R² = 0.7588 
pH_{BKT2} = 0.1556ln(R) + 7.3110  R² = 0.7038  N_{BKT2} = 2.982ln(R) + 53.669  R² = 0.1225 
pH_{BKT3} = 0.3040ln(R) + 6.7289  R² = 0.7649  N_{BKT3} = 20.84ln(R) + 158.95  R² = 0.8296 
pH_{MK1} = 0.0038(R) + 8.0392  R² = 0.8430  N_{MK1} = 0.2253(R) + 9.136  R² = 0.6540 
pH_{MK2} = 0.0041(R) + 7.9575  R² = 0.8761  N_{MK2} = 0.2648(R) + 24.129  R² = 0.7654 
pH_{MK3} = 0.0028(R) + 8.331  R² = 0.8208  N_{MK3} = 0.1947(R) + 10.276  R² = 0.6859 
pH_{SRI1}= 0.0047(R) + 7.9888  R² = 0.6688  N_{SRI1} = 1.501ln(R) + 9.8385  R² = 0.7181 
pH_{SRI2} = 0.0035(R) + 8.0932  R² = 0.6700  N_{SRI2} = 0.552ln(R) + 5.6547  R² = 0.7882 
pH_{SRI3} = 0.0046(R) + 7.8071  R² = 0.7854  N_{SRI3} = 2.325ln(R) + 12.577  R² = 0.6390 
S_{BKT1} = 0.0411(R) + 23.605  R² = 0.6379  P_{BKT1} = 31.99ln(R) + 200.02  R² = 0.8777 
S_{BKT2} = 0.0439(R) + 24.778  R² = 0.6474  P_{BKT2} = 38.33ln(R) + 243.27  R² = 0.7533 
S_{BKT3} = 0.0496(R) + 25.142  R² = 0.6118  P_{BKT3} = 21.57ln(R) + 158.26  R² = 0.9434 
S_{MK1} = 0.0274(R) + 30.415  R² = 0.6852  P_{MK1} = 3.345ln(R) + 19.965  R² = 0.6171 
S_{MK2} = 0.0665(R) + 28.316  R² = 0.7873  P_{MK2} = 9.063ln(R) + 49.76  R² = 0.7841 
S_{MK3} = 0.0402(R) + 29.865  R² = 0.7032  P_{MK3} = 1.312ln(R) + 12.297  R² = 0.6364 
S_{SRI1} = 0.0519(R) + 29.697  R² = 0.8428  P_{SRI1} = 1.625ln(R) + 11.086  R² = 0.7575 
S_{SRI2} = 0.044(R) + 30.138  R² = 0.7345  P_{SRI2} = 2.059ln(R) + 11.414  R² = 0.8379 
S_{SRI3} = 0.0481(R) + 30.517  R² = 0.8055  P_{SRI3} = 1.967ln(R) + 12.541  R² = 0.7388 
D_{BKT1} = 0.2497ln(R) + 3.2924  R² = 0.5765  Si_{BKT1} = 2.8139(R) + 318.30  R² = 0.7107 
D_{BKT2} = 0.4723ln(R) + 3.4519  R² = 0.6880  Si_{BKT2} = 5.9321(R) + 463.79  R² = 0.7265 
D_{BKT3} = 0.5528ln(R) + 2.2477  R² = 0.8659  Si_{BKT3} = 12.407(R) + 626.85  R² = 0.7802 
D_{MK1} = 0.0025(R) + 5.7695  R² = 0.0608  Si_{MK1} = 2.438(R) + 54.633  R² = 0.7759 
D_{MK2} = 0.0064(R) + 6.0290  R² = 0.7557  Si_{MK2} = 2.297(R) + 369.44  R² = 0.8922 
D_{MK3} = 0.0153(R) + 7.8423  R² = 0.9633  Si_{MK3} = 3.4005(R) + 168.67  R² = 0.7645 
D_{SRI1} = 0.0014(R) + 5.7808  R² = 0.0146  Si_{SRI1} = 125.6ln(R) + 951.33  R² = 0.7260 
D_{SRI2} = 0.0087(R) + 6.2643  R² = 0.6362  Si_{SRI2} = 145.2ln(R) + 1145.6  R² = 0.7047 
D_{SRI3} = 0.0139(R) + 5.7621  R² = 0.7842  Si_{SRI3} = 140.1ln(R) + 1155.2  R² = 0.8350 
Note: R is rainfall (mm); AHT is average the highest air temperature (°C); T is seawater temperature (°C); pH = pH level of seawater; S is salinity (psu); D is dissolved oxygen (DO) (mg L^{1}); A is ammoniumnitrogen (NH_{4} ^{+}N) (mg L^{1}); N is nitrite and nitratenitrogen (NO_{2} ^{}+NO_{3} ^{}N) (mg L^{1}); P is phosphatephosphorus (PO_{4} ^{3}P) (mg L^{1}); Si is silicatesilicon (Si(OH)_{4}Si) (mg L^{1})
TurkeyHSD  

Oneway ANOVA  Spatial Analysis  Temporal AnalysisBKT  
Parameter  Station  Month  BKT 1  BKT 2  BKT3  Jan  Feb  Mar  April  May  Jun  Jul  Aug  Sep  Oct  Nov  Dec 
GM (kg m^{2})  ***  ns  a  a  b                         
Temp (°C)  ns  ***        c  e,f  d,e  g  e,f,g  b  g  d  f,g  d,e  c  a 
pH  *  *  a  b  a,b  a,b  a  a,b  a,b  a,b  a,b  a,b  b  b  a,b  a,b  a,b 
Sal (psu)  ns  ***        c,d  c,d  b  a,b  a,b  a,b  a,b  a,b,c  a,b  a,b  b,c  d 
Conduct (ms)  ns  ***        a,b  b,c  b,c  b,c  a  c,d  a,b  a,b,c  a,b  a  a,b  c,d 
DO (mg L^{1})  ns  **        a  a,b  a,b  a,b,c  a,b,c  a,b,c  a,b,c  a,b,c  a,b,c  a,b  a,b,c  c 
Trans (cm)  ***  ns  a  b  c                         
TSS (mg L^{1})  ***  ns  a  a  b                         
NH_{4} ^{+}N (mg L^{1})  ns  ***        c  b  b  b  a  a  a  a  a  a  a  a 
NO_{2} ^{}+NO_{3} ^{}N (mg L^{1})  *  ns  a,b  a  b                         
PO_{4} ^{3}P (mg L^{1})  ns  ***        c  a,b,c  b,c  a,b,c  a,b  a  a  a  a  a  a  a 
Si(OH)_{4}Si (mg L^{1})  ns  *        a,b  a  a,b  a,b  b  a,b  a,b  a,b  a,b  a,b  a,b  a 
Chl 
ns  ns                               
Note: Significance level of oneway ANOVA: *** = P < 0.001; ** = P < 0.01; * = P < 0.05; ns = not significant
TurkeyHSD  

Oneway ANOVA  Spatial Analysis  Temporal AnalysisMK  
Parameter  Station  Month  MK 1  MK 2  MK 3  Jan  Feb  Mar  April  May  Jun  Jul  Aug  Sep  Oct  Nov  Dec 
GM (kg m^{2})  ***  ns  b  b  a                         
Temp (°C)  ns  ***        f  b  b,e  g  e  d,e  c,d,e  b,c,d  b,c,d  d,e  b,c  a 
pH  ns  **        a,b  a  a,b  a,b  a  a,b  a,b  a,b  a,b  b  a,b  a,b 
Sal (psu)  ns  **        b,c  b,c  a,b,c  a,b  a,b  a,b,c  b,c  a,b,c  a,b  a  a,b,c  c 
Conduct (ms)  *  ns  b  a  a,b                         
DO (mg L^{1})  ns  **        a,b  a,b,c  a,b,c  b,c  a,b,c  b,c  a,b,c  a,b,c  a,b,c  a,b,c  a  c 
Trans (cm)  **  ns  b  a  a                         
TSS (mg L^{1})  ns  *        a  a,b  a,b  a,b  a,b  a,b  b  a,b  a,b  a,b  a,b  a,b 
NH_{4} ^{+}N (mg L^{1})  ns  ***        f  e,f  e  d  b,c,d  a,b,c  a,b  a,b  a,b  a  a,b,c  c,d 
NO_{2} ^{}+NO_{3} ^{}N (mg L^{1})  ns  ns                               
PO_{4} ^{3}P (mg L^{1})  ns  ns                               
Si(OH)_{4}Si (mg L^{1})  ns  ns                               
Chl 
**  ns  a  a,b  b  a  a,b  a,b  a,b  a,b  b  a,b  a,b  a,b  a,b  a,b  a,b 
Note: Significance level of oneway ANOVA: *** = P < 0.001; ** = P < 0.01; * = P < 0.05; ns = not significant
TurkeyHSD  

Oneway ANOVA  Spatial Analysis  Temporal Analysis  
Parameter  Station  Month  SRI 1  SRI 2  SRI 3  Jan  Feb  Mar  April  May  Jun  Jul  Aug  Sep  Oct  Nov  Dec 
GM (kg m^{2})  **  ns  a  a  b                         
Temp (°C)  ns  ***        b,c,d,e  d  c  f  f  e  b,c,d,e  e  b  b,c,d  b,c  a 
pH  ns  *        a  a,b  a  a  a,b  a,b  a,b  a,b  b  a,b  a,b  a,b 
Sal (psu)  ns  ***        c,d  d  c,d  c,d  b  b,c,d  b,c  b,c,d  a  d  c,d  c,d 
Conduct (ms)  ns  ***        b,c  b,c  c  b,c  b,c  b,c  b,c  a,b  a  c  b,c  c,d 
DO (mg L^{1})  ns  ***        a,b,c  a,b,c  a,b,c  a,b,c  a,b  b,c,d  a  a,b,c  d  a  a,b,c  c,d 
Trans (cm)  ns  ns                               
TSS (mg L^{1})  ns  ns                               
NH_{4} ^{+}N (µg L^{1})  ns  ***        b  b  c  b,c  a  a  a  a  a  a  a  a 
NO_{2} ^{}+NO_{3} ^{}N (µg L^{1})  ns  ns                               
PO_{4} ^{3}P (µg L^{1})  ns  ***        c,d  b,c,d  a  b,c,d  d  a  a  a  a,b,c  a,b  a  a 
Si(OH)_{4}Si (µg L^{1})  ns  ***        b  b  b  b  a  a  a  a  a  a  a  a 
Chl 
ns  ***        a  a,b  a,b  a,b  a,b  a,b  a,b  b,c  c  a,b  a  a 
Note: Significance level of oneway ANOVA: *** = P < 0.001; ** = P < 0.01; * = P < 0.05; ns = not significant