Environmental impact of utilization of “produced water” from oil and gas operations in turfgrass systems

This study attempted to use produced water (PW) to irrigate turfgrass species, Cynodon dactylon and Paspalum sp. Assessment on established grasses, heavy metal accumulation and germination tests for weeds and turf grass seeds were conducted to evaluate the impact of PW irrigation. C. dactylon depicted lower tolerance while Paspalum sp. showed better tolerance capacity towards PW. C. dactylon grown from seeds under greenhouse conditions were not able to tolerate more than 30% concentration of PW (4.5% salinity). In comparison to tap water irrigated turf grass, Paspalum sp. was found to accumulate higher concentrations of V and Pb in shoots and Cr, Ni and As in roots. The results of seed germination tests recommended that irrigation with PW is to be performed after turfgrass establishment. Germination tests also revealed that PW could encourage growth of the weed—Chloris virgata while it could discourage growth of Amaranthus viridis and Launaea mucronata. This study suggests that PW could be used for turfgrass as an alternative water resource but only after further research on the long-term scale.

crops-Switch grass (Panicum virgatum L.) and Rapeseed (Brassica napus L.). They reported that the highest concentrations tested (salinity and TOC) significantly lowered the growth, health and physiological characteristics of both species. The authors concluded that the removal of organic matter to keep the TOC concentration less than 5 mg/L is required to maintain a sustainable biomass production rate 11 . Another study used produced water sourced from Wyoming and Montana in the U.S. on switch grass and corn, and four biofuel specieslemongrass, Japanese corn mint, common wormwood and spearmint. They concluded that prolonged use of produced water could have long-term deleterious effects on the soil and the plants, except if the produced water was treated or diluted with clean water 12 .
Treatment of produced water has been performed widely through the use of various physical and biological treatments 13 . However, in the past decades, common treatment methods for produced water have mainly consisted of membrane processes. These include forward osmosis (FO), nanofiltration (NF), ultrafiltration (UF) and microfiltration (MF) 13 . Prior treatment of the produced water before its use as an irrigation source provides a means to minimize the negative effects of the produced water's components on the irrigated crops.
However, due to risks associated with using produced water for growing food crops, non-food crops such as turfgrasses make a better candidate for wastewater treatment. Turfgrasses provide areas for leisure and sport activities and, more importantly, contribute to carbon sequestration. In addition, they are also used for land reclamation activities in areas that have been contaminated or areas that are industrial sites. Turfgrass covers very large areas around the world and the use of wastewater to irrigate turfgrasses has been conceptualized in the last decades and it has been applied in various parts of the world. For instance, in the state of Nevada in the U.S., more than 30 of the 53 golf courses utilize recycled water to irrigate greens, fairways and landscape plants 14

. A study in
Portugal assessed the quality of turfgrass (Cynodon dactylon) as a response to varied irrigation regimes including wastewater 15 . They concluded that treated final effluents could be considered as an alternative to potable water for irrigation of golf courses. It has also been reported that Cynodon dactylon has a high ability to grow in high sodium containing-severely sodic-soils 16 . Hence, it is a potential candidate for receiving high salt containing produced water as an irrigation source. Bioaccumulation is a challenge posed by produced water usage. Bioaccumulation of components sourced from irrigation water is a common phenomenon and it hinders the application of wastewaters like produced water. The accumulation of heavy metals in soils due to irrigation sources has been reported in the Middle East and around the world [17][18][19][20] . Thus, understanding the bioaccumulation capacity of produced water and the deleterious effects it could bring about to the soil and to turf grasses is essential.
Weeds are a nuisance to fields of crops and to areas where turf grasses are grown, competing with the cultivated crop for nutrients, water and space. Weeds are known to have rapid growth and maturation 21 . A change in the nutrient availability, soil structure and the soil microbial community can induce a change in the weed community of the concerned area 22 . Since produced water contains organic matter and metals, the weed community structure is expected to change based on these parameters as they could lead to the enhancement of weed growth 22 . Hence, germination tests based on parameters such as effect of salinity, heavy metal concentration and hydrocarbons allow for the pre-understanding of weed dynamics.
Qatar's turfgrass area accounts for 701,628 m 2 (Ministry of Municipality and Environment, Qatar, personal communication, 2017) which implicates the high water requirement for its maintenance. Statistical data has already proved the water insecurity of the country. Given that Qatar has an oil and gas based economy, we have an alternate source of water that is not only abundant but is also readily available-produced water. Although previous studies have evaluated and applied irrigation of turf grass with wastewater, a study testing produced water is novel and has been assessed in this study with three major objectives: Assessing turfgrass seed germination and turfgrass establishment and investigating seed germination of weed species as influenced by produced water irrigation.
Methods produced water samples. The produced water tested in this study was provided by TOTAL Qatar. The water was sourced from Al-Khalij offshore operation and transferred in a big container (1 m 3 ) to be kept in storage outdoors, after the thorough consideration of all the safety issues that may take place due to any possible leaking. In each sampling, the salinity and pH were determined before any experimental work took place. Produced water (PW) samples were analyzed for different metals and other chemical constituents 23,24 . Table 1 shows TDS values, pH, chemical constituents, metal concentrations and other parameters for the (PW) samples.
Prior to any assessment, the stock produced water samples were diluted according to the respective solution using tap water. The dilutions were prepared frequently and stored for short periods.
Given the safety issues of directly applying produced water to a land, experiments were set up using pots both inside and outside the greenhouse. Based on preliminary results, the subsequent experiments considered lower produced water dilutions and salinity as a factor. In addition, a further experiment was set up outdoors to simulate natural conditions. Germination tests of weed seeds were conducted to understand the effect of produced water on weed dynamics.
Assessing the effect of different concentrations of produced water on turf grass (Cynodon dactylon) establishment. 2 g of turfgrass seeds, Cynodon dactylon (Cȇsped gobi, Semillas fito, Spain) were sown in 30 pots (20 cm diameter) containing 60% sandy loam soil and 40% Peat moss soil. The pots were left for a period of 2 months to allow the grass to be established prior to treatment. The experiment included one factor with a completely randomized design (CRD) and it was set up inside a greenhouse (24 ± 2 °C with 15 h of light/ day at photon flux density minimum of 350 ± 50 µmol m −2 s −1 ). Six replicates were designated for each treatment and the pots were irrigated once a week with 200 mL of the assigned treatment. Treatment volume was determined based on irrigation regimes during the establishment period. The treatment levels were-0% (tap water), www.nature.com/scientificreports/ 25% produced water, 50% produced water, 75% produced water and 100% produced water concentrations. The dilutions were prepared using tap water. The dilutions were prepared frequently and stored for short periods in the refrigerator. Two weeks post treatment, the irrigation was alternated with tap water to de-stress the turf system and reduce the accumulative effects of heavy metals, salinity and other produced water components. The percentage of green biomass was observed and estimated every week using a scale rank of 0-100%, 0% for no green grass in the pot, and 100% for complete green and healthy grass in the whole pot and a percentage of green grass compared to control treatments for other results. After a period of 14 weeks, the green biomass for each pot was collected and put into a labeled paper bag. The root system of each pot was carefully cleaned from soil, washed with tap water and then put in a separate paper bag. All paper bags were placed inside an oven for 3-4 days 80 °C. The dry weight of the above ground biomass and the below ground root biomass were determined for each pot.
Assessing the effect of different produced water and saline water concentrations on turf grass (Cynodon dactylon) establishment. Based on the results obtained from the previous experiment (i.e. see "Assessing the effect of different concentrations of produced water on turf grass (Cynodon dactylon) establishment" section), lower concentrations of produced water were tested in order to reach the proper dilution of produced water needed to maintain healthy grass. In addition, different concentrations of saline water treatments were also added in order to determine if the negative effect was due to the salinity of the produced water. 2 g of turf grass seeds Cynodon dactylon (Cȇsped gobi, Semillas fito, Spain) were sown in 44 pots (20 cm diameter) containing a mixture of 60% sandy loam soil and 40% peat moss soil. The pots were left for a period of 2 months to let grass established prior to treatment. The experiment was one factor and four replications with a completely randomized design (CRD) and set up inside a greenhouse at similar conditions like experiment (see "Assessing the effect of different concentrations of produced water on turf grass (Cynodon dactylon) establishment" section). The treatment levels were: 0% (tap water), 10% produced water, 20% produced water, 30% produced water, 40% produced water and 50% produced water.
The saline water concentrations were matched to the produced water concentrations. The calculations were made using 100% salinity as 150 g/L (average of total measured salinity of produced water). The saline water concentrations tested were-15 g/L saline solution (designated as 1.5% S), 30 g/L saline solution (designated as 3% S), 45 g/L saline solution (designated as 4.5% S), 60 g/L (designated as 6% S) and 75 g/L (designated as 7.5% S). The irrigation was accomplished using 200 mL per pot of the assigned treatment, once a week. Two weeks post treatment, the irrigation was alternated with tap water to de-stress the turf system.
The percentage of green biomass was visually estimated every week using a scale rank of 0-100%. 0% stood for no green grass in the pot, 100% for complete green and healthy grass in the whole pot, and a % in between of green grass when compared to the control treatments for the other results. After a period of 14 weeks, the green biomass for each pot was collected and put into a labeled paper bag. The root system of each pot was carefully cleaned from the soil, washed with tap water and then put into a separate paper bag. All paper bags were placed Table 1. Chemical analysis of produced water in comparison with tap water. Adapted from 23,24 . ND not defined. www.nature.com/scientificreports/ inside an oven for 4 days at 80 °C. The dry weight of the aboveground biomass and the belowground root biomass were thus determined for each pot.

Assessing the effect of different produced water concentrations and saline water concentrations on established turf grass (Paspalum sp.) on a larger scale and outdoor conditions. Ready
turf grass rolls of Paspalum sp. which are commonly used in Qatari turfgrass systems were obtained upon request from the Ministry of Environment, Public parks department. 21 large pots of the size-65 cm × 25 cm × 20 cm were procured. Although they had drainage systems, the pots were designed in a way to prevent water leakage and thus had no outlets. Seven treatment levels were designated-0% (tap water), 10% produced water, 20% produced water, 30% produced water, 15 g/L saline solution (designated as 1.5% S), 30 g/L saline solution (designated as 3% S) and 45 g/L saline solution (designated as 4.5% S). Three replications were assigned for each treatment. The experiment followed completely randomized design (CRD) with three replications and was set up outdoors to simulate natural conditions. The pots were irrigated twice weekly with 250 mL/pot of assigned treatment each time. This treatment volume and schedule was determined based on water requirements of the grass observed during the establishment period. The irrigation was alternated with tap water (in the third week) to de-stress the turf system. The percentage of green biomass was visually estimated biweekly using the aforementioned method. After a period of 10 weeks, the green biomass for each pot was collected and put into a labeled paper bag. The root system of each pot was carefully cleaned from soil, washed with tap water and then put into a separate paper bag. All paper bags were placed inside an oven for 4 days at 80 °C. The dry weight of the aboveground biomass and the belowground root biomass were thus determined for each pot.

Metal digestion and icp analysis. Shoot and root samples from treatments-0%, 10%, 20% and 30%
PW were collected at the end of the previous experiment (i.e. Assessing the effect of different produced water concentrations and saline water concentrations on established turf grass (Paspalum sp.) on a larger scale and outdoor conditions). The samples were dried at 100 °C overnight after which they were manually ground to a powder form. 0.25 g of the sample was weighed and added to heat resistant tubes. To this, 5 mL of concentrated nitric acid was added. SRM 15151-apple leaves were used as reference material. One sample (10% PW roots) was duplicated to ensure a validation of the measurements. A spike was prepared by taking 0.1 mL of standard 100 ppm (ICUS-2959) and diluting it to 100 ppb. Two blanks were also prepared containing the acid only. All samples were capped and placed on a hot block set at 105 °C. After 2 h, the sample tubes were uncapped and the hot block temperature was increased to 130 °C to allow for evaporation. Following evaporation, the residue was mixed with 3 mL of conc. nitric acid and 1 mL of hydrogen peroxide and the sample tubes were allowed to boil at 155 °C in the hot block. The samples were then transferred to measuring flasks and their volume was completed to 50 mL using distilled water. The digested samples were then filtered twice using 0.25 µm filters to remove the precipitates. The filtered samples were then analyzed for metals-vanadium, chromium, manganese, nickel, cobalt, zinc, arsenic, cadmium and lead through ICP Analysis.
Seed germination experiments. Samples of produced water analysis showed that it contained-nickel 3.2 ppb, zinc 49.7 ppb, cobalt 0.75 ppb, lead 48.86 ppb (Table 1) in addition to a high salinity of 150 g/L. Therefore, germination tests on turf grass seeds (Cynodon dactylon) and weed species that are known to be associated with turf grass systems were performed using concentrations of the above mentioned heavy metals in addition to produced water and saline water treatments. Chloride salts of each of the metals was used to prepare the following concentrations nickel 3.2 ppb, zinc 49.7 ppb, cobalt 0.75 ppb, lead 48.86 ppb.
Germination test of turf grass (Cynodon dactylon) seeds. Seeds of turfgrass, Cynodon dactylon (Cȇsped gobi, Cynodon dactylon, Semillas fito, Spain) were surface sterilized using 5% sodium hypochlorite solution for 2 min and were followed by a washing with distilled water 3 times. 10 seeds of turfgrass (Cynodon dactylon) were placed in a Petri dish layered with a cheesecloth that was priorly soaked with 3 mL of the assigned treatment solution and then sealed with parafilm. 4 replicates were prepared for each treatment. The treatments tested were-0% control (distilled water), 1% produced water, 5% produced water, 10% produced water, 20% produced water, 1.5 g/L saline water (designated as 0.15% S), 7.5 g/L saline water (designated as 0.75% S), 15 g/L saline solution (designated as 1.5% S) and 30 g/L saline solution (designated as 3% S), nickel chloride 3.2 ppb, zinc chloride 49.7 ppb, cobalt chloride 0.75 ppb and lead chloride 48.86 ppb. The Petri dishes were placed in a growth chamber at 28 °C for 14 days. Distilled water was added (~ 2 mL) upon requirement to prevent dry out. The plates were observed daily for seed germination. The appearance of the white radicle was used as an indicator of germination. The germinated seeds were counted on a daily basis for 14 days. Accumulative of percent germination was documented for every treatment in all the experiments. The whole experiment was repeated once with statistically similar results.
Germination test of seeds of weed species associated with turf grass. Seeds of 6 common turf grass weeds were collected from turf grass fields inside Qatar University's campus. The seeds were cleaned from husk and other residues, and then kept in labeled paper bags in a refrigerator. Seed germinability tests revealed that seeds of 3 out of 6 species were not able to germinate. The tested species were Amaranthus viridis, Launea mucronata, and Chloris virgata. The seeds were soaked in 5% sodium hypochlorite solution for 2 min and were followed by a washing with distilled water 3 times. 10 seeds were placed in a Petri dish layered with a cheesecloth that was priorly soaked with 3 mL of the assigned treatment solution and then sealed with parafilm. Four replicates were prepared for each treatment. The treatments tested were the same as in above-mentioned experiment (i.e. Ger-

Scientific RepoRtS
| (2020) 10:15051 | https://doi.org/10.1038/s41598-020-72069-5 www.nature.com/scientificreports/ mination test of turf grass (Cynodon dactylon) seeds). The Petri dishes were placed in a growth chamber at 28 °C for 14 days. Distilled water was added (~ 2 mL) upon requirement to prevent dry out. The plates were observed daily for seed germination. The appearance of the white radicle was used as an indicator of germination. The germinated seeds were counted on a daily basis for 14 days. Accumulative of percent germination was documented for every treatment in all experiments. During the experiment seeds of Malva neglecta, Launaea capitata and Oligomeris subulata were found to be non-viable and the experiment was then continued with Amaranthus viridis, L. mucronata and Chloris virgata.
Data analysis. One-way ANOVA at P ≤ 0.05 was used to test the significance among treatments and the significance in each measured parameter. Tukey's test was used for mean comparisons at P ≤ 0.05. SigmaStat 4.0 was used to perform the data analysis. Data from the two experimental trials of turfgrass seed germination experiment were subjected to the Bartlett test for the homogeneity of variances. Data for all measured variables were homogeneous and the two experiments were pooled and analyzed as one with 8 replications.

Results and discussion
Effect of produced water on turfgrass. Second week post produced water treatment, Cynodon dactylon grass started to dry without future recovery. Only pots irrigated with 25% produced water concentration were tolerant and survived up to the end of experiment (14 weeks). Although the grass was able to survive at 25% PW irrigation (Fig. 1), the green biomass and the dry matter both above and below the ground were significantly reduced (Fig. 2) compared to the turf grass biomasses subjected to tap water irrigation. The decline in grass biomass might be due to salinity factors or other toxic substances available in the raw produced water. Therefore, the experiment was repeated under similar conditions however this time salinity treatments were added. Results obtained were almost similar to the results of the first experiment. Above 30% PW irrigation killed the grass ( Fig. 1-1). Interestingly, salinity treatment of 4.5% (equivalence to salinity in 30% PW) showed similar grass biomass like the 30% PW treatment ( Fig. 1-2). Almost similar trends were obtained on coverage percentage and again the salinity treatment effects almost matched the effects obtained from produced water treatments. After 14 weeks of irrigation, results from dry matter biomass, indicated a statistically significant (p ≤ 0.05) reduction in leaf dry weight starting from 20% PW and the matching 3% salinity treatments, while a significant reduction (p ≤ 0.05) in root biomass started from 30% PW or the matching equivalence of salinity treatment (4.5%) was observed (Fig. 3). Furthermore, another experiment was established in outdoor conditions using the turf grass species Paspalum sp., which is commonly used in Qatari turfgrass systems. The grass employed in the experiment was planted from rolls of seedlings and under outdoor conditions. In this experiment, we aimed to simulate the natural conditions of turf grass growth in Qatar. The obtained results indicated a better tolerance of Paspalum sp. to the produced water and salinity treatment. Figure 1-3 shows the appearance of the experimental turfgrass under all of the treatments following 10 weeks of irrigation. It is clear that there are neither negative effects on the green biomass due to the produced water nor due to the saline irrigation. However, Fig. 4 shows a statistically significant reduction (p ≤ 0.05) of dry weight of above and belowground parts under both treatments. Experimentation on C. dactylon and the similarity in the results obtained in produced water treatments and salinity treatments was an indication that negative effects on the treated grass could be primarily due to NaCl concentration in the raw produced water.
Based on obtained results, C. dactylon can tolerate up to 30% PW treatment but with loss in both aboveground biomass coverage% and biomass (leaf and root) while Paspalum sp. can at least tolerate 30% PW treatment with no loss in aboveground biomass coverage% but with reduction in biomass (leaf and root dry weight). A decrease in biomass of plants treated with produced water was an expected scenario. A change in irrigation regime is bound to bring about significant changes in the biomass of treated plants. Interestingly, similar biomass reduction results have been reported in studies that also tested the utilization of produced water for irrigation of varied plants 12,13 . It has been reported that, in general, turf grass species are efficient in utilizing nutrients from sources like wastewater 25 . The differences in growth patterns observed in C. dactylon and Paspalum sp. of the current study suggest that Paspalum sp. may have higher efficiency in utilizing the nutrient available in produced water as compared to C. dactylon. Different species have different tolerance capacities and have different growth requirements. The experimental conditions might also play a role in growth abilities of the turfgrasses. A study reported that difference in experimental conditions might relieve or worsen stress that is induced by salinity 26 . Since, the experiment concerned with C. dactylon was an indoor/greenhouse experiment and that concerned with Paspalum sp. was an outdoor experiment simulating natural conditions, differences in their response to produced water and its salinity was likely to occur. In addition, Paspalum sp. may have an inherent capacity of higher tolerance. It has been reported that Paspalum sp. such as Paspalum vaginatum have high salinity tolerance 27 and this can be an explanation for its ability to better withstand both produced water and saline water concentrations in comparison to C. dactylon. The higher salt tolerance of Paspalum sp. may originate from the possession of 'up-regulated' stress defensive proteins that are unique to them 28 . Hence, the results obtained indicate that in the application of produced water for irrigation in Qatar, Paspalum sp. could be the better choice. This is advantageous since Paspalum sp. happens to be the most commonly used turfgrass system in Qatar.

Metal digestion and icp analysis. Preliminary investigation of the fate of heavy metals in the plant parts
after being subjected to produced water irrigation was performed and results are shown in Table 2 for aboveground biomass and roots. According to the results of the ICP analysis, some elements were mainly accumulated in shoots i.e. V and Pb and others were more accumulated in the roots i.e. Cr, Ni, and As (Fig. 5). Metal digestion and ICP analysis of shoots and roots of Paspalum sp. treated with 10% PW-30% PW depicted accumulation of certain metals in shoots while others accumulated in roots (Fig. 5). Vanadium (V) and lead (Pb) were found to www.nature.com/scientificreports/ accumulate in the shoots of 10% PW-30% PW in higher concentration as compared to 0% treatment. It was reported that in mine tailings having high concentrations of Pb, Paspalum sp. was found to naturally colonize the region, thus asserting its ability to tolerate high Pb concentration and its possible usage in revegetation of Pb  www.nature.com/scientificreports/ tailings 29 . In contrast, chromium (Cr), nickel (Ni) and arsenic (As) accumulated in the roots in higher quantities compared to the control (0%) treatments. This is supported by the fact that Paspalum sp. such as Paspalum racemosum, and Paspalum tuberosum are known to be hyper accumulators of arsenic (As) 30 . Interestingly, the 10-30% PW treated grass also lost certain metals, making their concentrations lower than the concentrations observed in the 0% tap water treatment. Their shoots had lowered concentration of manganese (Mn) cobalt (Co), and cadmium (Cd), while their roots had lowered concentration of manganese (Mn) and zinc (Zn). Our results are similar to the paper 31 that studied Ni accumulation in C. dactylon and to that also studied element accumulation in C. dactylon which was irrigated with produced water 32 . Another study also reported similar accumulation of metals including Pb, Cd and Ni in parts of plants that were irrigated with wastewater 33 . Due to their long-term persistence in nature, heavy metals are an issue of concern. Bioaccumulation of heavy metals in turfgrass may provide a way for heavy metal removal from produced water and avoid heavy metal accumulation  www.nature.com/scientificreports/ in soils. Phytoextraction, a strategy of phytoremediation allows heavy metal removal by their uptake through plant's shoots and roots, followed by removal of the plant from the site 34 . A longer study would completely ascertain the ability of Paspalum sp. to bio-accumulate heavy metals of produced water and act as a phytoextractor.
Effect on germination of turf grass seeds. Seed germination experiments were completed to investigate the effect of heavy metals found in produced water on the germination of turfgrass seeds. All kinds of heavy metals significantly decreased (p ≤ 0.05) the germination potential of the seeds of Cynodon dactylon turfgrass compared to control (Fig. 6), and hence, it can be concluded that the seeds were affected by both, the presence of NaCl and the presence of tested metals, in the produced water. Nickel is known to be toxic to plants even in low quantities 35 . Lead is known to be a phytotoxin 36 . Cadmium is also reported to be phytotoxic, while zinc has been reported to inhibit germination in certain species 37 and thus could have inhibited the germination capacity of C. dactylon. This data suggests that while already established C. dactylon may tolerate up to 30% PW concentrations, germinating C. dactylon using produced water is not recommended. Irrigation of C. dactylon with produced water should be performed after it is well established. However, removal of salinity and of metals through produced water treatments may enhance the germination capacity of C. dactylon seeds.
Effect on germination of weed seeds. Seeds of weeds encountered in the turf grass were collected to be assessed for germination under different heavy metal treatments. Different weeds showed different germination potentials after being treated with different heavy metals (Fig. 7). For example, the most common weed species grown in turf grass systems of Qatar is Amaranthus viridis. The germination of the species was significantly reduced under all treatments (PW, salinity and heavy metals). Weeds are capable of disrupting vital ecosystem processes and out compete native species. Having a successful germination is extremely crucial in the life cycle of seeds 38 . In modified environmental conditions, seeds that can modify their germination behavior are highly . Average percentage of germination (4 replications with 10 seeds/replicate) for three weed species that are common to Qatari turfgrass after being subjected to varied produced water (PW), salinity (S) and heavy metal treatments. Error bars refers to standard error of the means. Any common letter between treatments refers to no significance at P ≤ 0.05 using Tukey's test. www.nature.com/scientificreports/ likely to survive and establish themselves 38 . Thus, the germination capacity of weed seeds gives an indication of its survival rate when subjected to produced water irrigation. The weed species Amaranthus viridis was discovered to be tolerant of salinity between 0.15-1.5% S but with lowered germination capacity (Fig. 7). However, the seeds could not germinate in produced water concentrations higher than 1% the stock PW solution suggesting that factors other than salinity affect its germination. Metal treatments also reduced germination capacity of A. viridis. Based on the results, it can be assumed that fields that are irrigated with produced water would discourage the growth of Amaranthus viridis, thus decreasing competition between turfgrass and weed species. This is advantageous as Amaranthus viridis is an invasive species characterized by rapid growth with high seed production rate 39 . It is known to grow well in very sandy soils and heavy organic soils 40 , and hence, could have been a nuisance if its growth was encouraged by produced water irrigation. In addition, it would lower costs for its removal and management as a result of the combination of hand weeding, fallow land management, and pre-post emergence herbicide that is required to control unwanted A. viridis growth 40 .
On the contrary, the weed Chloris virgata was observed to germinate with no significant differences between 0% treatment, 0.15% S, 0.75% S, NiCl 2 , ZnCl 2 and CoCl 2 treatments. 10% PW lowered germination percentage as compared to all treatments (Fig. 7). No germination was observed in 3% Salinity and 20% PW treatments. It can be suggested that C. virgata was affected primarily by a salinity higher than 1.5% since the germination of seeds was reported in metal treatments. It can be concluded that C. virgata weed can germinate and grow in fields irrigated with produced water if the concentration used is below 20% PW for a duration of 14 weeks. Concentration of produced water higher than 20% PW may discourage their growth but it would also deter growth of the turf grass. Chloris virgata is recognized as a halophyte species commonly growing in saline areas and degenerated grasslands 38 . It follows a C4 photosynthetic pathway giving it ability to grow in desert conditions and be drought resistant 41 , Hence its ability to germinate in both produced water and saline water. Thus, produced water concentration used for irrigation of turfgrass needs to be chosen in a manner to maximize turfgrass growth and to minimize the growth of C. virgata.
Seeds of Launaea mucronata could not germinate in 0.75% S-3% S, 5% PW-20% PW and NiCl 2 treatments (Fig. 7). Interestingly, seeds germinated well in PbCl 2 solution with no significant differences as compared to 0% treatment while 0.15% S, 1% PW, and ZnCl 2 and CoCl 2 treatments significantly decreased germination in comparison to the 0% treatment. It can be said that fields that have been irrigated with produced water would have reduced growth of Launaea mucronata and hence decreasing competition between turf grass and the weed species. It would also lower costs for the weed's removal and management.
Change in weed dynamics due to irrigation sources is commonly observed. For instance, it was reported that the use of yeast wastewater exerted impact on weed communities 22 . When it's compared to a freshwater irrigated and rain fed site, the yeast wastewater irrigated site had significantly higher vegetation cover and increased species' frequency, dispersion, richness and density. The vegetation cover obtained with yeast wastewater was twice to that of fresh water. The authors thus reported that the irrigation source has the ability to affect the relative abundance and the species composition of the weeds 22 . Similarly, produced water can also affect weed composition and abundance as observed in the results provided above, by encouraging growth of some species while discouraging that of others. Hence, thorough evaluation of the effects that produced water would have on turfgrass associated weeds is critical.

conclusions
The two tested turf grass species depicted a varied degree of tolerance and growth ability. C. dactylon was reported to be able to withstand up to 30% PW maximum concentration. The incorporation of a salinity factor in the experiment gave insight into understanding that NaCl concentrations could be the primary cause of the observed effects on C. dactylon turfgrass and hence the treatment for salt removal can allow higher concentrations of produced water to be used. The experiments conducted on Paspalum sp. suggested that it has a much higher capacity to tolerate salinity as well as produced water as a whole. It can withstand at least 30% PW/4.5% salinity. If long term experiments, that take into account the accumulation of salt and metals in the soil, were to be conducted to confirm these results, Paspalum sp. could potentially be used in areas in Qatar that are to be irrigated with produced water. However, as mentioned, the grass requires that it should be well established prior to the treatment in order to maximize its growth. In addition, a study conducted over a period of two seasons would further allow the understanding of the ability of the studied turf grass species to withstand produced water treatments.
Accumulation of metals in Paspalum sp. could indicate its ability to be used as means of metal removal or to be utilized in bioremediation projects. Use of produced water is expected to discourage growth of Amaranthus viridis and Launaea mucronata but may have no effect on the growth and abundance of Chloris virgata. More species of weeds need to be analyzed for their response to produced water. This will allow for the development of proper weed management strategies and precise management of cost. The concentration of the produced water chosen for irrigation, hence, is the key determinant on the effect it could have on the growth of the turfgrass and the growth and abundance of weeds.. In conclusion, if confirmed in the long term, produced water could be a viable, alternative irrigation source that could be used for irrigating turf grass. That is, if the suitable turf species are chosen, the area-requiring irrigation is well studied, an appropriate concentration of produced water is used, and a public risk assessment is performed.