3.1 Water samples
3.1.1 pH value
The pH analysis at all sampling points revealed pH values between 6.84 and 7.53 (test series 1) and between 6.83 and 7.48 (test series 2), which were within the WHO guidelines (WHO 1996). Along the flow route of the Tigris River, pH decreased to 6.8-6.9 within the city passage compared to unpolluted conditions (test site 1; 7.48-7.53) due to the discharge of organic pollutant loads and biodegradation of these contaminants (Al-Saffawi 2018) effecting partial acidification due to oxygen-limiting conditions. Due to a combination of higher water temperature and thus lower oxygen solubility, higher biodegradation activity, and lower water flow velocity due to low river levels during dry months, oxygen limitation is more prominent in summer months. However, the drop did not exceed 0.69 pH units because the carbonate hardness of the river water resulted in a high buffering capacity. If this buffering capacity was not present, a significant fluctuation in pH would be expected with implications for aquatic life (Ali et al. 2019, Kevat et al. 2016).
3.1.2 Electric conductivity, total dissolved solids and salinity
Electrical conductivity of surface waters is highly regional and can range from 50 to 1500 µS/cm (EPA 2012). Studies of US inland fresh waters indicate that streams supporting good mixed fisheries have a range between 150 and 500 µS/cm (EPA 2012). Conductivity outside this range may indicate non-suitable water for certain species of fishes or bugs. High conductivity up to 10,000 µS/cm in inland surface water strongly indicates water contamination by industry or crises (ENR). According to the WHO, a mid-range conductivity (200 to 1000 µS/cm) is the regular level for most major rivers (ENR), but suggests a limit of 400 µS/cm for drinking water (Meride and Ayenew 2016). Measurements in the Tigris River were within WHO target values for rivers in both test series. As these values were also within the range of electrical conductivities of North American streams, this parameter is not an exclusion criterion for the use of Tigris water. However, it can be seen over the flow length that the conductivity increases over all 10 measuring points from originally 405 µS/cm (test series 1) and 258.3 µS/cm (test series 2) as mean value to 750 µS/cm (test series 1) and 593 µS/cm (test series 2). Thus, there is a discharge of saline wastewater or contaminated surface runoff through the urban area. Also, infiltration of saline groundwater cannot be excluded, since hydrogeologically the Lazzaga area is part of the Tigris hydrogelogical basin in which water flow converges towards the River Tigris (Jassim S. Z. et al. 1999, Alsam et al. 1990, Araim 1991). Correspondingly, the detected TDS curve correlates with the electrical conductivity, which also increases continuously between the measuring points S1 to S6 (series 1: 405 µS/cm to 750 µS/cm; series 2: 258 µS/cm to 521 µS/cm; mean values in each case), drops by approx. 10 % to S7 in both series and then remains nearly unchanged. The TDS concentrations remain below the WHO specifications, whereby high TDS values are not fundamentally harmful to humans. However, the consumption of water with high TDS values can have an increased negative effect on people with kidney and heart diseases in particular. In general, this water may also increase laxative but also constipating effect (Meride and Ayenew 2016, Sasikaran et al. 2012).
The same effect was observed for the salt content, which increased from 0.39% (series 1) or 0.31% (series 2) at reference point S1 to measuring point S5 or S7 to a maximum of 1.0% (series 1) or 0.66% (series 2) and dropped by 25-30 % to S10. The WHO limit value is just met at S7 during wintertime. Comparing both series an inconsistent picture is drawn for these three parameters. Due to the lack of precipitation during the summer period and thus lower levels in the Tigris, an increase in electrical conductivity, TDS value and salinity due to anthropogenic discharges would be expected. However, direct comparison of Tables 2 and 3 shows that electrical conductivity outside the urban area remained almost unchanged, while within the urban area remained around 30 % lower than during wintertime (series 1).
In case of the TDS values, there was an increase by a factor of 1.85 in both series over the city passage, with summer levels also being around 35-40% lower. The situation is similar with the salinity, which is also about 30% lower in summer than in wintertime, but rises by 120 % (winter) or 25% (summer) over the city passage. It temporarily increases by a factor of 2-3 within the urban area. The most likely explanation for this scenario is that due to the lack of precipitation in summer and thus the lack of surface runoff, a significantly lower pollutant load is flushed into the Tigris. In addition to anthropogenic discharges, the increase in contaminants in the urban area is due to the infiltration of contaminated groundwater into the Tigris (Jassim S. Z. et al. 1999, Alsam et al. 1990, Araim 1991).
3.1.3 Chemical oxygen demand
The thesis of increased infiltration of contaminated groundwater can be confirmed using COD data. In both series, an increase in the background COD (about 25-30 mg O2/L) by a factor of 2-3 to 89 mg O2/L (series 1) and 58 mg O2/L (series 2), respectively, occurred across the urban area. Within the city, even higher COD values of 98.4 mg O2/L (S5, Series 1) and 74.33 mg O2/L (S5, Series 2) were temporarily achieved, detectable at nearly identical levels at S10 during the winter period, while a decrease of about 35% due to biodegradation processes was observed during the summer period. Increased direct or indirect discharge of sanitary wastewater and thus increased organic loads into the river due to war-related damages to the sewage system were previously described (Jankeer and Mustafa 2017, Mustafa 2002).
3.1.4 Nitrate, phosphate, and sulfate
There are a variety of potential emission sources for nitrate, phosphate, and sulfate. For example, these anions can be released during the exploration of minerals, agricultural use of fertilizer, industrial effluents, general leakages in the destroyed municipal sewage network, but also through direct use of weapons and ammunition. In the latter case, nitrate is directly used in munitions as ammonium nitrate.
Conflict-related phosphate emissions in the urban area of Mosul are particularly known from the use of phosphorus bombs during the liberation in the summer of 2017 (Aljazeera 2016). However, this type of bomb was also used by U.S. forces during the second Gulf War (Reuters et al. 2005). Sulfate emissions can result from the combustion of antimony sulphides when munitions are fired (Mariussen et al. 2021, Jovanovic et al. 2018). However, in the case of Mosul, the burning of the Mishraq industrial complex (SO2, SO3) (UNEP and OCHA 2016, BBC 2016) 45 km south of Mosul in October 2016 in particular, where the resulting gases entered the river through wet deposition into the soil and surface drainage (UNEP 2018), and the sulfur springs extending from Mosul in the north to Qayara in the south for 60 km along the Tigris River, where the River Tigris is also a zone of discharge (Jassim S. Z. et al. 1999), played a substantial role. Accordingly, the analyses northwestern measuring point, showed in some cases significant exceedances of the WHO limits. Thus, phosphate concentrations increase by a factor of 1.8-1.9 through the urban passage in both series, regardless of the season. In some cases, detected phosphate levels of 0.81 mg P/L (series 2) and 1.20 mg P/L (series 1) exceeded WHO limit by a factor of 2-3 within the urban region.
Sulfate levels exceeded limits from S2 to S10 in R1 series, and from S5 to S10 in R2 series, at the highest value of 495 mg/l (S6 in R1). Again, a key aspect of lower sulfate levels, but also other anion levels, was the lack of surface drainage in combination with groundwater infiltration to Tigris. However, biodegradation represents a second, possibly even dominant factor, where both sulfate and nitrate may be used as electron acceptors instead of oxygen. This aspect could be easily verified due to the severe odor emissions of H2S gas near the valleys. Direct drainage into the river was described previously (Al-Saffawi 2018, Al-Sarraj 2020). In addition to above factors, the detected sulfate levels were further increased by dissolution of gypsum rock due to impoundment of Tigris water in Mosul Lake (Fadhel 2020).
In the case of nitrate, the course over the river flow showed only a small increase at S3, which is mainly due to agriculture upstream, but remains at an almost constant level over the course of the river. As microbiological H2S formation occurred, and thus inevitably microbiological nitrate respiration as a sink of nitrate, a continuous discharge of nitrate over the urban area took place. The discharge was caused by defects in the sewage system and infiltration processes as consequence. Nitrate concentration was clearly below the WHO limit value at all locations, irrespective of the season.
3.1.5 Heavy metals
Due to international contracts prohibiting the use of expanding bullets, as well as the higher cost of hollow point ammunition, full metal jacket bullets are the most used cartridges by soldiers as well as militias (Deese 2021). They consist of a soft core, usually lead, and a harder jacket metal alloy such as gilding metal (CuZn), CuNi, or steel encasing the core (Mariussen et al. 2021). The purpose of this design is to provide improved trajectory as well as greater penetration of the projectiles into soft tissue (Coget et al. 2021, Deese 2021). Supplementary galvanic cadmium coatings are applied to improve some properties like corrosion resistance, improved tribological properties, and chemical stability (Jovanovic et al. 2018). Chromatic layers are applied as a supplementary passivation (Jovanovic et al. 2018). Therefore, the use of weapons in conflict zones leads to an increase in heavy metal concentrations.
Concentrations of Cd, Pb, Cr, and Ni exceeded WHO limits even before entering the city (S1), while increasing further especially through the Old Town passage as a direct conflict zone (S5 - S7) in both series. With maximum concentrations (in ppm) of 0.47 Cd, 0.89 Pb, 0.39 Cr, and 0.98 Ni (series 1) and 0.40 Cd, 0.63 Pb, 0.32 Cr, and 0.73 Ni (series 2), the WHO limits were exceeded here in part by a factor of 100. For Zn, a comparable concentration pattern occurred over the urban passage, but with maximum values of 2.03 ppm (Series 1) and 1.52 ppm (Series 2), the limits were not exceeded.
3.1.6 Comparison of water samples in conflict zone vs. urban zone
The pollution of the Tigris River can be divided into a total of four zones over its flow length in the urban area. Zone 1 (S1) as a reference before entering the densely populated urban area illustrates the background pollution of the river. Zone 2 (S2-S4) essentially represents contamination from agricultural drainage. Since the zone is otherwise characterized by recreation areas, the influence of municipal drains is still low here. Between sampling points S4 and S5, the Al-Khosr valley drains into the Tigris, which carries a heavy pollutant load on its flow route from the northeastern valleys through densely populated areas (Al-Safawi 2007) and therefore leads to a strong increase in the parameters E.C., TDS, salinity, COD, nitrate, phosphate, and sulfate at S5.
The military operations to liberate Mosul focused on the Old City of Mosul, where sampling points S5-S7 (Zone 3) are located. Therefore, within this zone, an increase in heavy metal concentrations is expected and has been observed as an indicator. Destruction of the drainage infrastructure also leads to a further increase of the above parameters. Zone 4, represented by the S8-S10, again represents an area with strong agricultural use and residential development outside the direct conflict zone, so that the discharge situation here is limited to agricultural and municipal parameters and the heavy metals decrease in concentration due to (bio)chemical processes such as adsorption, precipitation, uptake in tissues as well as dilution effects.
The differences in concentration between summer and winter periods caused by the lack of surface drains as well as existing groundwater infiltration have already been pointed out. In addition, the role of entrained sediments should be mentioned, which will be further discussed in section 3.2.
3.1.7 Statistical analysis
An adequate approach for statistical analysis of deviations in analytical data and their significance is the t-test. The results for E.C, TDS, COD, nitrate, phosphate, sulfate, Cd, Pb, Zn, Cr, as well as Ni showed levels below the critical value (P <0.05), which is why they are considered as statistically significant differences. Thus, based on the null hypothesis that the sample mean, and population mean are statistically different at the 0.05 significance level due to the time lag between the two analytical series, a significant statistical difference results between these two series (P<0.05, see Table 4).
The percentage difference of heavy metal analyses between the two series at all sites showed that the values of Cd, Pb, Zn, Cr and Ni in S1, S2, S3, S8, S9, S10 were higher than the values in S4, S5, S6, S7, which is due to the fact that the pollutants at these sites are discharged throughout the year via the drains from the valleys and are not only discharged via rainwater. Rather, surface drains from rainwater additionally enter the Tigris in addition to existing drains, which significantly increases the percent of pollution in the river at these sites. The lowest percent differences between both runs were seen at S5 and S6 for Cd (12.5 % vs. 20.4 %), Pb (38.9 % vs. 22.6 %), Ni (24.9 % vs. 23.4 %), Zn (24.9 % vs. 28.2 %), and Cr (22.0 % vs. 12.6 %) as shown in Table 5 and Figure 3. Differences at other locations were higher in this case, since the main pollution source of the Tigris River is the winter rainwater drain, therefore the difference in heavy metal concentrations between both series is even more evident. The highest values were obtained for Cd at S1 (83.3%) > Pb at S10 (69.7%) > Cr at S3 (51.4%) > Ni at S3 (50.0%) > Zn at S2 (45.2%) as shown in Table 5.
Figure 4 shows that the ranking in the percent differences in site contaminants between upstream S1 and downstream S10 in both runs was as follows: Cd>Pb>Cr>Ni>Zn> COD>TDS>>E.C>pH. Except for pH, the percentage increase of all parameters between upstream and downstream was higher in series 1 (winter, spring) than in series 2 (summer, fall). The pH was lower in series 1 than in series 2, because the increased biodegradation of organics in summer combined with oxygen limitation led to an enhanced decrease in pH.
3.2 Sediment samples
3.2.1 Chemical parameters and heavy metals
Sediments are an important sink for pollutants, but they can also lead to contamination of the water body through desorptive processes. In particular, if the sediment is primarily of mineral nature, ions (e.g. heavy metals) may be adsorbed (Miranda et al. 2022). However, interaction strength of heavy metals and geochemical fractions of sediments strongly depends on exchangeable, reducible, oxidizable, and residual fractions of the sediment (Buyang et al. 2019, Keshavarzifard et al. 2019). The exchangeable and, therefore, weakly-bound geochemical sediment fractions include fixation of heavy metals to the solid surface by weak electrostatic interactions (Brady et al. 2014, Jayarathne et al. 2019). Hence, changes in environmental parameters like pH, E.C., or salinity can promote desorption of heavy metals by breakdown of these weak electrostatic interactions (Chon et al. 2012, Keshavarzifard et al. 2019, Xia et al. 2020). If the sediment is primarily of organic nature, organic contaminants (e.g., pharmaceuticals, flame retardants) may be adsorbed (Dobslaw et al. 2021, Xu 2021). In addition to the complexity of the sediment composition, the interpretation of occurring distribution equilibria between sediment phase and liquid phase is particularly complicated by the downstream transport of sediment or small-particle and thus suspended silt fractions. Hence, it is to be expected that even outside the direct conflict zone or the urban area increased pollutant concentrations may occur in the water phase as well as in the river sediment due to sediment transport and subsequent desorption.
In fact, the sediment analysis confirmed continuous increase of heavy metal concentrations over the flow distance, with the highest concentrations occurring at the southern urban border (S10). A similar result was already presented by Al-Sarraj et al (Al-Sarraj et al. 2019). With detected heavy metal concentrations as mean of 6.23 ppm Cd, 45.33 ppm Pb, 21.33 ppm Zn, 81.33 ppm Cr, 64.00 ppm Ni (Series 1) and 9,200 ppm Cd, 62,800 ppm Pb, 30,000 ppm Zn, 123,600 ppm Cr, 79,000 ppm Ni (Series 2), respectively, WHO limits were permanently exceeded at all locations except Zn. In the case of S10, the limit values for Cd, Pb, Cr and Ni were exceeded by a factor of 3.1, 1.3, 3.3 and 3.2 (series 1) and by a factor of 4.6, 1.8, 4.9 and 4.0 (series 2). For zinc, the limits were met despite an increase in concentration over the urban passage (see Tables 6 and 7). However, due to the interaction between dissolved heavy metals in the water phase and sorbed heavy metals at the solid phase, severe local variations in concentration, in contrast to the water phase, were not detectable.
In the case of the parameters pH, E.C., and salinity, no significant change occurred for the pH value during the winter series. During the summer series, however, a moderate decrease of the pH value by 0.77 units in absolute terms was observed over the urban area. This decrease is related to the increased degradation of organic pollutants induced by higher temperatures with partial oxidation of the components. The pH drop turns out to be comparatively small, since both the morphological sediment composition and the existing lime-carbonic acid equilibrium in the water phase as well as the sediment phase lead to a pH buffering (Zarraq 2012). This pH buffering prevents enhanced pH-induced mobilization of heavy metal ions. This effect has already been described for Cr, Ni, and Zn (Miranda et al. 2022). The mobility of heavy metals is further influenced by E.C. as well as salinity, with enhanced mobilization at increasing E.C. and decreasing salinity, respectively. Both parameters increased by a factor of 1.6 - 2.4 over the urban passage, and just exceeded the WHO limits by a factor of 6.4 for E.C. and by a factor of 3.3 for salinity in the summer measurement, while in the winter period only salinity exceeded the WHO limit by a factor of 1.8.
3.2.2 Comparison of sediment samples in conflict zone vs. urban zone
The direct comparison of these parameters between both series showed, with exception of the pH value (see above) and the E.C. (almost unchanged), a significant increase of 1.6-1.8 (salinity), 1.2-1.5 (Cd), 1.2-1.4 (Pb), 1.4 (Zn), 1.2-1.5 (Cr) as well as 1.1-1.2 (Ni) in the summer period (series 2, see tables 6 and 7). This increase during dry weather is probably related to a combination of low flow velocity of the Tigris during the dry weather phase and thus also an increased deposition of small particles, which either infiltrated directly into the Tigris from the conflict zones or were lifted up by wind and settled down above the water surface, as well as an increased particle-bound heavy metal input during the precipitation phase in winter and spring (series 1, see Figure 8). These particles, some even forming conglomerates, are slowly transported downstream as sediment mixtures, resulting in the high concentrations detected throughout the entire sampling stretch of the river (see Figure 5).
3.2.3 Statistical analysis
The statistical T-test analysis shows that the T-test values for salinity, Cd, Pb, Zn and Cr were below the critical value of P <0.05. This means that there is a significant statistical difference between series 1 and series 2 (P<0.05). Conversely, the numerical values of Ni and E.C. were greater than 0.05 (P>0.05), which means that there is no significant statistical difference between series 1 and series 2 for these two parameters. This indicates different pollution sources or input sources of nickel to the sediments (see Table 9).
The percent difference in sediment analysis data between the locations of both series indicated that there is no regular variation between sites, as the rates of variation between both series were different in all sites depending on the specific element. This is due to differences in the river´s flow rate and the movement of sediments in the river, as well as the different sources of pollution. The highest percent difference in heavy metals was as follows: Cr in S1 (36.8 %) > Cd in S10 (32.2 %) > Pb in S2 (26.6 %) > Zn in S7 (23.6 %) > Ni in S1 (23 %) (see Table 8).
There was a general increase in contaminant concentrations between upstream S1 and downstream S10 in both test series, with the highest percentage increase as follows: Cr> Cd >Pb>Zn > E.C > salinity. In the case of pH, an enhanced decrease was observed during summer tests due to present biodegradation.