3.1. Irrigation water quality parameters
The concentration of all physicochemical parameters of water in the study area are presented in the table4.
Water temperature is one of the most important factors in aquatic environment since it regulates physicochemical as well as biological activities (Kumar and Siddiqui, 1997). Minimum and maximum surface water temperatures at the Ghrib Dam are observed during January and September respectively (table 4). The water temperature of the dam is affected by the change in season and temperature fluctuations. pH is inseparable from temperature values, dissolved oxygen, and electrical conductivity, because during the day the intense absorption of CO2 leads to an evolution of pH and precipitation of carbonates (Arrignon,1991, El Houati et al., 2013). The pH values range from 8.49 (P1) to 8.37 (P3) during January 2016 and between 8.13 (P5) and 6.23 (P4) during May 2016 and from 8.3 (P7) to 7.8 (P11) during April 2017 and April 2018 respectively (Table 4). The observed values indicate that the pH is slightly alkaline neutral at all points of the sample, this is due to the presence of carbonates that allow to buffer the water that flows towards the dam, pH values are, therefore, acceptable for irrigation as they meet FAO standards (6 à 8,5. Electrical conductivity is proportional to the amount of dissolved ionizable salts and is a bio-indicator of the degree of water mineralization (Nisbet and Verneaux 1970). The more the conductivity of water the lesser is its resistance to electric flow thereby indicating higher concentration of dissolved salts and higher trophic status of the system (Kumar and Siddiqui, 1997). The results show that the values of electrical conductivity fluctuate between 2.75 to 3.80 ds/m for the waters of January 2016 and 2.74 to 3.20 ds/m for the waters of the second campaign, between 2.4 and 2.2 ds/m during April and September 2017 respectively and between 2.5 and 2 ds/m during February and May 2018, which does not meet the standards demanded by FAO. These values indicate a strong mineralization of the waters in the Ghrib Dam. This excess mineralization is due to contact with rocks during the water path through the various geological formations of the Ghrib Dam.
Tableau 4. Descriptive statistics of Dam water parameters in comparison with FAO (1996)
Statistique
|
Unit
|
SAMPLE 1 (P1, P2, P3)
|
SAMPLE 2 (P4, P5, P6)
|
SAMPLE 7
|
SAMPLE 8
|
SAMPLE 9
|
SAMPLE 10
|
SAMPLE 11
|
SAMPLE 12
|
FAO
|
Min
|
Max
|
Mean
|
SD
|
Min
|
Max
|
Mean
|
SD
|
standard
|
HCO3−
|
mmg.L− 1
|
93
|
152
|
107.8
|
28.17
|
155
|
178.6
|
158.6
|
10.56
|
212
|
202
|
216
|
216
|
218
|
142
|
610
|
NH4+
|
mmg.L− 1
|
0.144
|
0.239
|
0.203
|
0.052
|
0.02
|
0.24
|
0.17
|
0.12
|
1
|
0.58
|
0.3
|
0.38
|
0.4
|
0.25
|
5
|
EC
|
dS.m− 1
|
2.75
|
3.8
|
3.3
|
0.42
|
2.74
|
3.2
|
2.93
|
0.11
|
2.4
|
2.2
|
2.5
|
2.2
|
2.4
|
2
|
0–3
|
Ca2+
|
mmg.L− 1
|
188.58
|
322
|
249.8
|
40.08
|
162
|
243
|
200.4
|
32.73
|
136
|
126
|
116
|
117
|
138
|
162
|
400
|
Cl−
|
mmg.L− 1
|
270.03
|
280
|
268.2
|
13.66
|
197
|
240
|
218.65
|
20.26
|
322
|
290
|
320
|
360
|
315
|
271
|
1065
|
BOD
|
mmg.L− 1
|
4
|
14
|
7.33
|
4.93
|
3
|
4
|
2.67
|
0.58
|
5.1
|
4.4
|
3
|
1.1
|
1.1
|
4.2
|
< 10
|
DCO
|
mmg.L− 1
|
18
|
28
|
21
|
5.57
|
24
|
28
|
24.33
|
2.52
|
34
|
19
|
15.1
|
9.0
|
20
|
19.2
|
< 20
|
Mg2+
|
mmg.L− 1
|
43
|
114
|
77.64
|
36.16
|
66
|
205
|
128.48
|
68.21
|
82
|
242
|
74
|
73
|
86
|
19
|
60
|
NO3−
|
mmg.L− 1
|
0.86
|
1.84
|
1.26
|
0.49
|
1.32
|
3.48
|
2.46
|
1.12
|
3.85
|
10.42
|
0.6
|
0.16
|
0
|
7.84
|
0–10
|
DO
|
mmg.L− 1
|
10.92
|
11.59
|
11
|
0.18
|
9.27
|
9.48
|
9.25
|
0.09
|
10.06
|
8.40
|
10.2
|
10.73
|
9.95
|
7.85
|
|
PO43−
|
mmg.L− 1
|
0.352
|
0.507
|
0.406
|
0.09
|
0
|
0.5
|
0.17
|
0.29
|
0.474
|
0.1
|
0.12
|
0.2
|
0.1
|
0.08
|
2
|
K+
|
mmg.L− 1
|
5.65
|
7.62
|
6.57
|
0.9
|
6.98
|
7.90
|
7.9
|
0.48
|
5.6
|
6.1
|
4.2
|
5
|
7.2
|
6.97
|
0–2
|
SAR
|
Meq.L− 1
|
0.006
|
0.05
|
0.026
|
0.018
|
0.05
|
0.02
|
0.042
|
0.003
|
0.021
|
0.018
|
0.023
|
0.024
|
0.022
|
0.023
|
< 0.2
Serious problem
|
Na+
|
mmg.L− 1
|
0.75
|
3.55
|
2.22
|
1.5
|
2.57
|
3.27
|
2.8
|
0.41
|
2.27
|
2.42
|
2.25
|
2.40
|
2.2
|
2.19
|
620
|
SO42−
|
mmg.L− 1
|
154
|
275
|
202
|
65.35
|
108
|
134
|
119
|
12.43
|
472
|
488
|
540
|
520
|
475
|
490
|
960
|
Tur.
|
NTU
|
9.92
|
10.33
|
10.04
|
0.18
|
5.02
|
11
|
6.39
|
3.35
|
5
|
12.3
|
3.4
|
5
|
4
|
17.1
|
|
pH
|
-
|
8.54
|
8.74
|
8.46
|
0.08
|
6.44
|
8.35
|
7.33
|
0.98
|
8.3
|
7.95
|
8.2
|
8.2
|
7.8
|
8.1
|
6.5–8.5
|
Dissolved oxygen is one of the most important parameters in assessing the quality of water, which affects the survival and distribution of flora and fauna. Oxygen content is important for direct need of many organisms and affects solubility of many nutrients and therefore, productivity of aquatic ecosystem (Wetzel, 1983). Variations in dissolved oxygen are related to several factors, mainly temperature and salinity (Lacaze 1996). Low concentration promotes the development of algae and parasites, which are responsible for the presence of toxic substances. Based on the results, the observed values of dissolved oxygen range from 10.92 to 11.59 mmg.L− 1 for the 1st campaign and 9.27 to 9.48 mmg.L− 1 for the 2nd campaign. Also, for the year 2017, the value of dissolved oxygen was between 8.40 mmg.L-1 during the month of September 2017 and 10.06 mmg.L-1 during the month of April 2017, while the values oscillated between 7.85 mmg.L− 1 (month of May) and 10.73 mmg.L− 1 (month of March) in 2018. It noticed that the values of dissolved Oxygen of the 2nd campaign, April 2017, and March 2018are low (below 10 mmg.L-1 minimum acceptable value) which shows the presence of organic pollution. These values are low compared to the 1st campaign (January 2016), September 2017 et May 2018, this is due to the decrease in flow and the increase in temperature, therefore the dilution process at the Ghrib Dam is decreasing. These values of dissolved oxygen make the water of the Ghrib Dam of poor quality. Overall, the closer the concentration of dissolved oxygen is to saturation, the greater the river's ability to absorb pollution (Rodier et al., 2009).
Biological oxygen demand (BOD), an organic pollution criterion based on the amount of oxygen consumed at 20oC, is considered as an important parameter in an aquatic ecosystem to establish the status of organic pollution (Jain, Dhanija, (2000). Adakole (2000) categorized water based on BOD levels into unpolluted (BOD < 1.00 mmg.L− 1), moderately polluted (2–9 mmg.L− 1) and heavily polluted (BOD > 10 mmg.L− 1). The BOD levels in all Ghrib dam during study period fluctuated between 1 and 7.33 mmg.L− 1 indicating that the lake is moderately polluted. BOD values range from 4 mmg.L− 1 to 14 mmg.L− 1 for the first campaign, from 3 to 4 mmg.L− 1 for the 2nd campaign, from 4.4mmg.L− 1 to 5.1mmg.L− 1 for April and September 2017 and from 1mmg.L-1 in mars 2018 to 4.2mmg.L− 1 in May 2018, which may be due to the small amount of organic matter available in the medium. The BOD, in the study, area indicates that the waters of Ghrib Dam are moderately polluted according to the classification of Adakole (2000), but during the 1st sampling campaign, the BOD was 14mmg.L− 1 which shows that the waters during this period are severely polluted. This may be due to input of organic wastes and enhanced bacterial activity (Prasannakumari et al., 2003). Chemical oxygen Demand (COD) determines the amount of oxygen required for chemical oxidation of most organic matter and oxidizable inorganic substances with the help of strong chemical oxidant. Based on the results obtained, the variation in the concentration of DCO at the level of the campaigns is from 18 to 28 mmg.L− 1 for the first campaign and 24 to 28 mmg.L− 1 for the 2nd campaign, from 19 to 34mmg.L-1 for September and April2017 respectively and from 9 to 20mmg.L− 1 for mars and April 2018 respectively, which is explained by the fact that the amount of oxygen provided by the oxidation of organic matter is therefore low biodegradability.The higher levels may be due to higher decomposition activities and lower water levels. Similar trends have been observed in several studies (Hallouz et al., 2014; Touhari et al., 2018). On the other hand, the highest values probably correspond to a high content of organic matter that has a biodegradable character, and therefore has degraded in this environment. Finally, these values do not meet FAO standards. Sulphate plays an important role in soft water systems where complex metal ions prevent reacting with other substances (Wetzel, 1983). The main source of sulphates is runoff from catchment area rich in mineral and organic Sulphur. The sulphate values recorded in the waters of these campaigns vary from point to point, they are moderately high by FAO standards, this is due to the nature of the rocks crossed during their displacement. Indeed, crops like sugar cane requires higher sulfate concentrations (94 kg. ha− 1) than other crops (corn: 47 kg. ha− 1, rice: 20 kg. ha− 1) (Schueneman, 2001; De La Mora-Orozco et al., 2017). Sulphate can be naturally occurring from gypsum or pyrite, of industrial origin or from agricultural processing products with gypsum land, or from runoff or infiltration into gypsum fields. They are also the result of the activity of certain bacteria (chlorothiobacteria, rhodothiobacteria, etc.).
The presence of nitrate nitrogen in lakes is governed by the activity of nitrifying bacteria on nitrogen source of domestic or agricultural origin. The most widespread contaminants are nitrogen compounds in sub-surface areas, mainly originated from decaying organic matter, leakage of septic tanks, sewage wastes, and fertilizers, as well as the infiltration of nitrate with the leaching water (Sirajudeen and Mubashir, 2013; Adam Khalifa et al., 2019). Nitrates stimulate the development of aquatic flora and increase the productivity of a lake ecosystem (Arrignon, 1976). Ganapati (1960) pointed out that concentration of NO3–N > 150 µgL-1 is an indicative of eutrophication. The concentration of nitrogen in water is quite high, due to equilibrium with nitrogen in the atmosphere. Based on the results, a remarkable variation exists between nitrate values for the campaigns from 0.86 to 1.84 mmg.L− 1 for the first campaign, 1.32 to 3.48 mmg.L− 1 for the 2nd campaign, 3.85 to 10.42 mmg.L− 1 for the 3rd campaign (April and September 2017) and from 0 to 7.84 mmg.L− 1 for the last campaign (April and May 2018), this variation is mainly due to leaching of agricultural land by runoff, this was observed during the period of the 1st campaign. In the end, nitrate values remain below the allowable values given by FAO standards, except for September 2017, which shows a value above the FAO standard.
Among nutrients, the importance of phosphates in water bodies is well studied (Vollenweider, 1968); Vaithiyanathan and Subramanian, 1993). Niswander and Mitsch (1995) pointed out that the addition of phosphate to water brings about eutrophication by increasing the bacterial content, increase in oxygen demand, and increase in production of growth factors for algae thus resulting in increased algal growth. Phosphate is essential for the development of suspended micro-algae and the results obtained show low values than those given by the FAO standards, where the presence of excess phosphorus can provoke the proliferation of plants. This parameter informs us about the degradation that is due to eutrophication. High concentrations of chloride (Cl−) gives a salty taste to water (Adam Khalifa et al., 2019). The chloride dosing results show that the values of the first campaign range from 270.03 to 280 mg.L− 1, from 197 to 240 mg.L− 1 for the second campaign, 290 mmg.L− 1 to 322 mmg.L− 1 for the 3rd campaign (2017) and from 271 mmg.L− 1 to 360 mmg.L− 1 for the last campaign (2018). These values are in accordance with FAO standards. Then, the chloride levels of the water are extremely variable and are mainly related to the nature of the land crossed. Sodium can come from several origins, namely: the decomposition of mineral salts such as silicates, the leaching of NaCl-rich geological formations from saltwater in the slicks, and the discharge of industrial and domestic wastewater. Indeed, for all samples, sodium concentrations are low for all samples with values ranging from 0.75 to 3.55 mg.L− 1 for the first campaign, from 2.57 to 3.27 mg.L− 1 for the second campaign, from 2.27 to 2.42 mg.L− 1 for the 3rd campaign and from 2.19 to 2.40 mmg.L− 1 for the last samples, these values are in line with FAO standards. Like sodium, potassium is also a naturally occurring element. Potassium levels were high during all samples. Low potassium concentrations lead to lower growth rate and photosynthesis of blue-green algae and increased respiration (Wetzel, 1983). The variation in potassium values between the two companions varies from 5.65 to 7.62 mg.L− 1 for the first companion, from 6.98 to 7.9 mg.L− 1 for the 2nd campaign., from 5.6 to 6.1 mg.L− 1 for the 3rd campaign and from 4.2 to 7.2 mg.L− 1 for last campaign. Potassium values remain high by input to FAO standards. Potassium can come from fertilizers, clays, and volcanic rocks and from industrial origin. Also, calcium and magnesium are often present in significant concentrations of natural water and are directly related to hardness (Adam Khalifa et al., 2019). The results obtained show that the waters of the Ghrib Dam are rich in calcium at all sampling points, these values are between 188.58 and 322 mg.L− 1 for the waters of the first campaign, vary between 162 and 243 mg.L− 1 for the second campaign, from 126 to 136 mg.L− 1 for the 3rd campaign and from 116 to 162 mg.L− 1 for the last sampling points. These variations in calcium levels remain in line with FAO standards. Le magnésium affiche des valeurs supérieures à la norme FAO pour l’ensemble des échantillons. For all samples, the bicarbonate concentrations (HCO− 3) decrease, this applies to all samples with values ranging from 93 and 152 mg.L− 1 for the first campaign, 155 to 178.6 mg.L− 1 for the second campaign, 202 to 212 mg.L− 1 for the 3rd campaign and from 142 to 218 mg.L− 1 for the last campaign. These values are high but remain in line with FAO standards, this increase is probably related to the increase in pH. On the other hand, magnesium levels vary from season to season, both surface and depth. The level of Mg+ for the waters of the first campaign varies between 43 and 114 mg.L− 1, for the second campaign, it varies from 66 to 205 mg.L− 1, for the 3rd campaign, the values of magnesium are from 82 to 242 mg.L− 1 and from 19 to 86 mg.L− 1 for the last sampling points. These values remain moderately high by FAO standards, this is due to a mineral origin, the main source is the mineral complex of sedimentary rocks and to the terrain crossed. Ammonia nitrogen is present in two forms in solution, ammonia NH3 and ammonium NH4, whose relative proportions depend on pH and temperature. Ammonium is often dominant; therefore, this term is used to design ammonia nitrogen (Aminot and Chaussepied, 1983). In an oxidizing medium, ammonium turns into nitrites and then nitrates, which induces oxygen consumption (Gaujous, 1995). Based on the results obtained, there is a remarkable variation between the values of the Ammonia Nitrogen for all campaigns. The values obtained are low and do not meet FAO standards, or this water ranks in the wrong class. Turbidity reflects the presence of suspended particles in water (organic debris, clays, microscopic organisms, etc.). It is important to know the content of turbidity when considering treating water because it facilitates the development of germs that indicate contamination, reduces the effectiveness of disinfectants, and increases chlorine consumption while decreasing its effectiveness. (Miquel, 2003). The results obtained revealed that turbidity is slightly variable for all campaigns during the analysis period, these values are between 9.92 and 10.33 NTU for the waters of the first campaign, they are, therefore, murky waters and this is due to the presence of finely divided suspended matter: clays, silt, and organic matter. As for the 2nd campaign, the values oscillate 5.02 and 11 NTU which indicates that this water is slightly cloudy, so presence of some particles suspended. During the months of September 2017 and May 2018, turbidity increased dramatically, reaching the values of 12.3 and 17.1 NTU respectively. In addition, the values of the 1st campaign as well as those of September 2017 and May 2018 remain higher compared to the waters of the other campaigns, this is due to the difference in the harvest dates, during the 1st period, the flow was abundant due to heavy rains during January 2016, September 2017, and May 2018 (which correspond to 88mm, 36mm and 58mm respectively) but during the other campaigns, the flow was zero.
Finally, in this study, the variation in Sodium Absorption Rates (SAR) values was 0.006 to 0.05 meq.L− 1 for the first campaign, which means that there was dilution during that month of January 2016 since the amount of rain during that month was 88mm. On the other hand, the values of the other campaigns ranged from 0.010 to 0.025 meq.L− 1, the latter do not meet FAO standards, which proves that this water has a high salinity.
3.3. Statistical analysis
3.3.1. Pearson’s Correlation Coefficient among Parameters
The correlations among water quality variables can reveal several important hydrochemical relationships (Wu et al., 2014). The Pearson’s correlation matrices were applied to identify the relationship between the variables. Correlation matrix of the 18 measured parameters was computed and presented in Table 5. Pearson’s correlation value ranges between 0 (in the case of no correlation) and 1 (when the correlation is perfect). Samples having a correlation coefficient greater than 0.7 are strongly correlated. The loading values are classified according to Liu et al. (2003); the absolute loading values of > 0.75, 0.75 − 0.50, and 0.50 − 0.30 are classified as strong, moderate, and weak, respectively. From the correlation matrix, many of the physicochemical parameters showed strong correlations with each other, indicating the close association of these parameters with each other. Results of correlation analysis showed that the high positive correlations between EC and Ca2+, Mg2+, Na+, K+, Cl−, HCO3−, SO42−, and PO43− was observed, and this reflects the great contribution of these ions in water salinity (Jahin et al., 2020). In addition, significant positive correlation of EC with NO3− indicates relative lower contribution in water salinity (Table….). High significant positive correlations observed among Na+, K+, Ca2+, Mg2+, Cl−, SAR, and SO42− demonstrate that these six constituents form the majority of soluble salts in waters. This also reflects the natural composition dominating surface water in the studied area (Chapman, 1996). The NO3− showed a strongly negative correlation with DO, moderate significant positive correlation with Turbidity (TUR), but weak significant positive correlations with Mg2+, EC, and Cl−. Furthermore, PO43− showed a significant positive correlation with pH, DO, and Ca2+, but weak significant negative correlations with Mg2+. The water temperature (T) had shown moderate significant and positive correlation between sulfate, NO3−, and Mg2+, and a negative high correlation with DO. The Turbidity had shown a weak significant and positive correlation with K+ and BOD. Lastly, NH4+ had shown a significant and negative correlation with BOD (Table 5). Also, pH has a significant negative correlation with most of the physicochemical parameters. Despite being a very effective tool, correlation analysis could only indicate the general insight into water-rock interactions [(Adam Khalifa et al., 2019).
Table 5
Pearson correlations of the physicochemical variables
Var.
|
pH
|
EC
|
DO
|
TUR
|
Cl−
|
HCO3−
|
Ca+
|
Mg+
|
Na+
|
K+
|
NO3−
|
PO4−
|
SO4−
|
BOD
|
DCO
|
SAR
|
NH4+
|
T
|
pH
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
EC
|
0.183
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DO
|
0.418
|
-0.224
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
TUR
|
0.206
|
-0.145
|
-0.400
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Cl−
|
0.570
|
0.812
|
0.304
|
-0.203
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
HCO3−
|
-0.145
|
0.784
|
-0.200
|
-0.500
|
0.560
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
Ca+
|
-0.093
|
0.774
|
0.364
|
0.170
|
0.796
|
-0.682
|
1
|
|
|
|
|
|
|
|
|
|
|
|
Mg+
|
-0.328
|
0.797
|
-0.314
|
0.106
|
0.799
|
0.228
|
0.800
|
1
|
|
|
|
|
|
|
|
|
|
|
Na+
|
-0.416
|
0.868
|
-0.118
|
-0.068
|
0.801
|
0.151
|
0.848
|
0.811
|
1
|
|
|
|
|
|
|
|
|
|
K+
|
-0.256
|
-0.794
|
-0.286
|
0.467
|
0.762
|
-0.501
|
0.797
|
0.802
|
0.775
|
1
|
|
|
|
|
|
|
|
|
NO3−
|
-0.061
|
0.343
|
-0.773
|
0.579
|
0.313
|
0.017
|
-0.208
|
0.318
|
-0.062
|
0.126
|
1
|
|
|
|
|
|
|
|
PO4−
|
0.444
|
0.743
|
0.494
|
0.201
|
0.097
|
-0.293
|
0.331
|
0.139
|
0.032
|
-0.041
|
-0.313
|
1
|
|
|
|
|
|
|
SO42−
|
0.277
|
0.967
|
-0.160
|
-0.031
|
0.856
|
0.639
|
0.768
|
-0.164
|
0.773
|
0.759
|
0.258
|
-0.240
|
1
|
|
|
|
|
|
BOD
|
0.364
|
-0.308
|
0.330
|
0.353
|
-0.134
|
-0.350
|
0.698
|
-0.061
|
0.183
|
0.024
|
0.044
|
0.368
|
-0.289
|
1
|
|
|
|
|
DCO
|
-0.190
|
-0.276
|
-0.105
|
0.009
|
-0.436
|
-0.063
|
0.334
|
0.194
|
0.392
|
0.376
|
0.078
|
0.388
|
-0.436
|
0.421
|
1
|
|
|
|
SAR
|
-0.453
|
-0.503
|
-0.074
|
-0.157
|
-0.647
|
-0.160
|
0.236
|
0.233
|
0.848
|
0.570
|
-0.211
|
0.015
|
-0.636
|
-0.039
|
0.335
|
1
|
|
|
NH4+
|
0.162
|
0.648
|
-0.061
|
-0.241
|
0.551
|
0.511
|
-0.465
|
0.029
|
-0.182
|
-0.363
|
0.343
|
0.063
|
0.588
|
-0.483
|
0.239
|
-0.393
|
1
|
|
T(°C)
|
-0.417
|
0.279
|
-0.801
|
0.103
|
-0.150
|
0.436
|
-0.406
|
0.591
|
0.193
|
0.263
|
0.718
|
-0.410
|
0.158
|
-0.321
|
0.191
|
0.053
|
0.367
|
1
|
Values in bold are different from 0 at a significance level alpha = 0.05
3.3.2. Principal Component Analysis (PCA)
For the meaningful interpretation of all the data collected from the chemical analysis of the waters, we used the method of Principal Component Analysis (PCA) which allows to establish correlations between the different variables and to specify the relationships between the chemical variables. This method provides information on the most meaningful parameters which describe complete dataset allowing data reduction with minimum loss of original information (Helena et al., 2000, Singh et al., 2004).
The objective of this analysis is to describe or classify the data, to allow the interpretation of the hydrochemical functioning of the dam waters. A principal component analysis was carried out in order to complete the hydrochemical study.
Principal component analysis (PCA) is a method of reducing the number of variables to allow their geometric representation. This reduction is only possible if the initial variables are not independent and have non-zero correlation coefficients (Faye et al., 2020).
To identify the important water quality parameters and their controlling mechanisms, PCA is executed on 18 variables including T, EC, pH, DO, Turbidity, major ions (Ca2+, Mg2+, Na+, K+, NH4+, Cl−, NO3−, SO4−and PO4−), SAR, alkalinity, BOD and COD of the 12 different sampling sites which shows the major variables governing water physiochemical properties of the Ghrib dam.
3.3.3. Multivariate statistical analyses
Based on the Eigen values results, 4 factors explaining 78.96% of the variance or information contained in the original data set were retained. The factors correspond to the Eigen values (5.88, 4.02, 2.38, 1.93) respectively and are sufficient to give a good idea of the data structure. Any factor with an Eigen value greater than 1 is considered significant. The Eigen values for different factors, percentage variance, cumulative percentage variance and component loadings (unrotated and varimax rotated) are summarized in Table 6. The first component represented 32.69% of the variance with an eigenvalue of 5.88. It included the most significant variables controlling the water chemistry in the studied area, i.e., EC, Ca2+, Mg2+, Na+, K+, Cl−, SO42−, PO43−, SAR, and NH4+. As these parameters were well-correlated (Table 6). The second component represented 22.33% of the variance with an eigenvalue of 4.02. This component was dominated by trace elements of natural origin, including pH, DO, and T. The third component was dominated by variables Turbidity and NO3−.This third component represented 13.24% of the variance with an eigen value of 2.38. The fourth component accounted for 10.70% of the total variance with an eigenvalue of 1.93. It was dominated by well-correlated three variables, i.e., NH4+, BOD, and DCO.
Tableau 6. Varimax rotated component matrix for water quality parameters.
|
PC1
|
PC2
|
PC3
|
PC4
|
Eigenvalue
|
5.88
|
4.02
|
2.38
|
1.926
|
Variance, (%)
|
32.69
|
22.33
|
13.24
|
10.70
|
Cumulative, %
|
32.69
|
55.01
|
68.26
|
78.96
|
Indicator
|
Eigenvectors
|
|
|
|
pH
|
0.28
|
-0.65
|
0.41
|
0.19
|
EC
|
0.94
|
0.14
|
-0.02
|
0.13
|
DO
|
-0.11
|
-0.87
|
-0.33
|
0.19
|
TUR
|
-0.21
|
0.11
|
0.86
|
-0.06
|
Cl−
|
0.88
|
-0.39
|
-0.03
|
0.16
|
HCO3−
|
0.72
|
0.33
|
-0.44
|
0.27
|
Ca+
|
-0.80
|
-0.32
|
0.15
|
-0.01
|
Mg+
|
-0.60
|
0.48
|
-0.02
|
0.38
|
Na+
|
-0.54
|
0.42
|
-0.41
|
0.41
|
K+
|
-0.70
|
0.37
|
0.13
|
-0.11
|
NO3−
|
0.19
|
0.63
|
0.70
|
0.03
|
PO4−
|
-0.63
|
-0.48
|
0.14
|
0.42
|
SO4−
|
0.96
|
0.00
|
0.11
|
-0.01
|
BOD
|
-0.34
|
-0.38
|
0.42
|
0.59
|
DCO
|
-0.44
|
0.22
|
0.05
|
0.70
|
SAR
|
-0.67
|
0.35
|
-0.49
|
0.11
|
NH4+
|
0.64
|
0.13
|
0.08
|
0.52
|
T
|
0.21
|
0.91
|
0.15
|
0.11
|
Bold-face numbers indicates highly loaded variables.
The quality of irrigation water is highly variable depending upon both the type and the quantity of the salts dissolved in it. These salts originate from natural (i.e., weathering of rocks and soil) and anthropological (i.e., domestic, and industrial discharges) sources and once introduced, they follow the flow path of the water. It is commonly accepted that the problems originating from irrigation water quality vary in type and severity as a function of numerous factors including the type of the soil and the crop, the climate of the area as well as the farmer who utilizes the water. Nevertheless, there is now a common understanding that these problems can be categorized into the following major groups: (a) salinity hazard, (b) infiltration and permeability problems, (c) Specific ion toxicity and (d) miscellaneous problems (Simsek and Gunduz, 2007; Rasul and Hassan, 2013).
3.3.4. Hierarchical cluster analysis
The hierarchical cluster analysis HCA aims to classify objects into groups based on the similarity between samples with respect to the physicochemical elements (Noshadi and Ghafourian 2016). (Fig. 4). Hierarchical clustering of all sampling sites using Ward’s method with squared Euclidean distance resulted in dendrogram consisting of four statistically significant clusters (Fig. 3): cluster 1 (P1, P2, P3), cluster 2 (P4, P5, P6), cluster 3 (P7, P9, P10, P11) and cluster 4 (P8, P12). The result of cluster analysis has revealed different water quality exists in different zones of the lake depending upon their proximity to the water source. The agglomeration of these sites into different clusters may be due to the different environmental conditions they are exposed to. It is evident that the CA is advantageous in offering reliable classification of sampling sites and will support designing a more precise sampling strategy for future monitoring programs. The obtained results reinforce those already achieved through PCA.
3.4. Assessment of water quality using IWQI and SAR
The results of Table 7 demonstrated that IWQI in this section of the Ghrib dam decreases from downstream to upstream. Indeed, the IWQI values were less than 59 units for all sampling sites and in May 2018 (41 units). According to the classifications for the different uses of water, values below 60 indicate that the water is of poor quality for irrigation. Nevertheless, it is recommended that more data be obtained over several years to confirm the effect of environmental conditions on water quality at Ghrib Dam. It should be noted that several authors have reported the polluting effect of rain and climatic conditions on the dynamics and water quality of rivers and reservoirs (Roselli et al., 2009; Razmkhah et al., 2010; González-Ortegón, 2010; Chilundo, 2008; De La Mora-Orozco, 2017). However, rainy season runoff can introduce pollutants such as nitrogen and phosphorus, which contribute to the flowering of water hyacinths. Water treatment is recommended to make dam water more suitable for irrigation.
Added to these results, the variation in SAR values was 0.006 to 0.05 meq.L− 1 for the 1st companion, which means that there was dilution during this month of January 2016 since the amount of rain during that month was > 70mm.
On the other hand, the values of the 2nd campaign ranged from 0.05 to 0.019 meq.L− 1, the latter do not meet FAO standards, which proves that there is a high salinity. SAR values in the third campaign ranged from 0.021 in April 2017 to 0.018 in September 2017 and during the last campaign, the SAR values were between 0.022 in April 2018 and 0.023 in February and May 2018. Based on the results, the waters of the Ghrib Dam have a poor chemical quality.
Table 7
IWQI values from the Ghrib dam reservoir in all sampling
Date
|
Sampling
|
IWQI values
|
January
30th, 2016
|
P1
|
55
|
P2
|
55
|
P3
|
53
|
May
2nd, 2016
|
P4
|
59
|
P5
|
59
|
P6
|
59
|
April, 2017
|
P7
|
50
|
Sept. 2017
|
P8
|
55
|
Feb. 2018
|
P9
|
59
|
March, 2018
|
P10
|
58
|
April, 2018
|
P11
|
59
|
May, 2018
|
P12
|
41
|
There were no significant differences in IWQI values among the sampling sites during all campaigns along the dam over the study year (Table 7). According to the classifications for the different water uses, values below 60 indicate that the water requires treatment before use for irrigation or other purposes. However, the campaigns of 2016 and 2017 observe a quality index below 60%, which makes these waters of poor quality. We hypothesize that this improvement is the result of the rainy season in this particular year of study (2016, 2017).
It should also be noted that the campaigns of May 2017 and February and April 2018 have given values of IWQI equal to 59 units, these values indicate that we are in the presence of poor water quality, this is probably due to climatic conditions during this period, where rainfall of 57mm, 101 mm and 192mm were recorded during the months of May 2017 and February and April 2018 respectively. It should be noted that several authors have reported the polluting effect of rain and climatic conditions on the dynamics and water quality of rivers and reservoirs (Roselli et al., 2009; Razmkhah et al., 2010; González-Ortegón, 2010 ; Chilundo, 2008 ; De La Mora-Orozco, 2017).
Also, the minimum value of this quality index was recorded during the month of May 2018 although during this period a rainfall of 56mm was observed, and it was, also, the irrigation season of the perimeter (beginning April 1st of each year). This explains that the rainy season runoff can introduce pollutants such as nitrogen and phosphorous, which contribute to the blooming of water hyacinths. Water treatment is recommended to make dam waters more fit for irrigation or other purposes.
In the end, the values of the physical and chemical analyses of the Ghrib dam waters, intended for irrigation of farms and compared to the guide values (FAO) led to the following results:
- The waters of our study region have low alkalinity, as well as high mineralization for all sampling points.
- The indicators of pollution reveal the presence of acute pollution which is certainly caused by direct releases of either industrial or domestic origin, and this pollution remains variable depending on the sampling periods.
- Chloride-calcium and sulfate facies are the most dominant in the two harvest periods during all sampling periods for dam waters, resulting in poor water quality for irrigation. In addition, water is, therefore, highly mineralized and is likely to be suitable for irrigation of certain species (cucumbers...) that are well tolerant to salt and on well-drained and leached soils.