Thank you for visiting nature.com. The browser version you are using has limited CSS support. For the best experience, we recommend that you use the latest browser version (or turn off compatibility mode in Internet Explorer). Additionally, to ensure continued support, this site will not include styles or JavaScript.
       Dust storms pose a serious threat to many countries around the world due to their destructive impact on agriculture, human health, transportation networks and infrastructure. As a result, wind erosion is considered a global problem. One of the environmentally friendly approaches to curb wind erosion is the use of microbial induced carbonate precipitation (MICP). However, the by-products of urea-degradation-based MICP, such as ammonia, are not ideal when produced in large quantities. This study presents two formulations of calcium formate bacteria for the degradation of MICP without producing urea and comprehensively compares their performance with two formulations of non-ammonia-producing calcium acetate bacteria. The bacteria considered are Bacillus subtilis and Bacillus amyloliquefaciens. First, the optimized values ​​of the factors controlling CaCO3 formation were determined. Wind tunnel tests were then conducted on sand dune samples treated with the optimized formulations, and wind erosion resistance, stripping threshold velocity, and sand bombardment resistance were measured. Calcium carbonate (CaCO3) allomorphs were evaluated using optical microscopy, scanning electron microscopy (SEM), and X-ray diffraction analysis. Calcium formate-based formulations performed significantly better than acetate-based formulations in terms of calcium carbonate formation. In addition, B. subtilis produced more calcium carbonate than B. amyloliquefaciens. SEM micrographs clearly showed the binding and imprinting of active and inactive bacteria on calcium carbonate caused by sedimentation. All formulations significantly reduced wind erosion.
       Wind erosion has long been recognized as a major problem facing arid and semi-arid regions such as the southwestern United States, western China, Saharan Africa, and much of the Middle East1. Low rainfall in arid and hyper-arid climates has transformed large parts of these regions into deserts, sand dunes, and uncultivated lands. Continued wind erosion poses environmental threats to infrastructure such as transportation networks, agricultural land, and industrial land, leading to poor living conditions and high costs of urban development in these regions2,3,4. Importantly, wind erosion not only impacts the location where it occurs, but also causes health and economic problems in remote communities as it transports particles by wind to areas far from the source5,6.
       Wind erosion control remains a global problem. Various methods of soil stabilization are used to control wind erosion. These methods include materials such as water application7, oil mulches8, biopolymers5, microbial induced carbonate precipitation (MICP)9,10,11,12 and enzyme induced carbonate precipitation (EICP)1. Soil wetting is a standard method of dust suppression in the field. However, its rapid evaporation makes this method of limited effectiveness in arid and semi-arid regions1. The application of oil mulching compounds increases sand cohesion and interparticle friction. Their cohesive property binds sand grains together; however, oil mulches also pose other problems; their dark color increases heat absorption and leads to the death of plants and microorganisms. Their odor and fumes can cause respiratory problems, and most notably, their high cost is another obstacle. Biopolymers are one of the recently proposed eco-friendly methods for mitigating wind erosion; they are extracted from natural sources such as plants, animals and bacteria. Xanthan gum, guar gum, chitosan and gellan gum are the most commonly used biopolymers in engineering applications5. However, water-soluble biopolymers can lose strength and leach out of soil when exposed to water13,14. EICP has been shown to be an effective dust suppression method for a variety of applications including unpaved roads, tailings ponds and construction sites. Although its results are encouraging, some potential drawbacks must be considered, such as cost and the lack of nucleation sites (which accelerates the formation and precipitation of CaCO3 crystals15,16).
       MICP was first described in the late 19th century by Murray and Irwin (1890) and Steinmann (1901) in their study of urea degradation by marine microorganisms17. MICP is a naturally occurring biological process involving a variety of microbial activities and chemical processes in which calcium carbonate is precipitated by the reaction of carbonate ions from microbial metabolites with calcium ions in the environment18,19. MICP involving the urea-degrading nitrogen cycle (urea-degrading MICP) is the most common type of microbial-induced carbonate precipitation, in which urease produced by bacteria catalyzes the hydrolysis of urea20,21,22,23,24,25,26,27 as follows:
       In MICP involving the carbon cycle of organic salt oxidation (MICP without urea degradation type), heterotrophic bacteria use organic salts such as acetate, lactate, citrate, succinate, oxalate, malate and glyoxylate as energy sources to produce carbonate minerals28. In the presence of calcium lactate as a carbon source and calcium ions, the chemical reaction of calcium carbonate formation is shown in equation (5).
       In the MICP process, bacterial cells provide nucleation sites that are particularly important for the precipitation of calcium carbonate; the bacterial cell surface is negatively charged and can act as an adsorbent for divalent cations such as calcium ions. By adsorbing calcium ions onto bacterial cells, when the carbonate ion concentration is sufficient, calcium cations and carbonate anions react and calcium carbonate is precipitated on the bacterial surface29,30. The process can be summarized as follows31,32:
       Biogenerated calcium carbonate crystals can be divided into three types: calcite, vaterite, and aragonite. Among them, calcite and vaterite are the most common bacterially induced calcium carbonate allomorphs33,34. Calcite is the most thermodynamically stable calcium carbonate allomorph35. Although vaterite has been reported to be metastable, it eventually transforms into calcite36,37. Vaterite is the densest of these crystals. It is a hexagonal crystal that has better pore filling ability than other calcium carbonate crystals due to its larger size38. Both urea-degraded and urea-undegraded MICP can lead to the precipitation of vaterite13,39,40,41.
       Although MICP has shown promising potential in stabilizing problematic soils and soils susceptible to wind erosion42,43,44,45,46,47,48, one of the by-products of urea hydrolysis is ammonia, which can cause mild to severe health problems depending on the level of exposure49. This side effect makes the use of this particular technology controversial, especially when large areas need to be treated, such as for dust suppression. In addition, the odor of ammonia is intolerable when the process is carried out at high application rates and large volumes, which may affect its practical applicability. Although recent studies have shown that ammonium ions can be reduced by converting them into other products such as struvite, these methods do not completely remove ammonium ions50. Therefore, there is still a need to explore alternative solutions that do not generate ammonium ions. The use of non-urea degradation pathways for MICP may provide a potential solution that has been poorly explored in the context of wind erosion mitigation. Fattahi et al. investigated urea-free MICP degradation using calcium acetate and Bacillus megaterium41, while Mohebbi et al. used calcium acetate and Bacillus amyloliquefaciens9. However, their study was not compared with other calcium sources and heterotrophic bacteria that could ultimately improve wind erosion resistance. There is also a lack of literature comparing urea-free degradation pathways with urea degradation pathways in wind erosion mitigation.
       In addition, most wind erosion and dust control studies have been conducted on soil samples with flat surfaces.1,51,52,53 However, flat surfaces are less common in nature than hills and depressions. This is why sand dunes are the most common landscape feature in desert regions.
       To overcome the above mentioned shortcomings, this study aimed to introduce a new set of non-ammonia producing bacterial agents. For this purpose, we considered non-urea degrading MICP pathways. The efficiency of two calcium sources (calcium formate and calcium acetate) was investigated. To the best of the authors’ knowledge, carbonate precipitation using two calcium source and bacteria combinations (i.e. calcium formate-Bacillus subtilis and calcium formate-Bacillus amyloliquefaciens) has not been investigated in previous studies. The choice of these bacteria was based on the enzymes they produce that catalyze the oxidation of calcium formate and calcium acetate to form microbial carbonate precipitation. We designed a thorough experimental study to find the optimal factors such as pH, types of bacteria and calcium sources and their concentrations, the ratio of bacteria to calcium source solution and curing time. Finally, the effectiveness of this set of bacterial agents in suppressing wind erosion through calcium carbonate precipitation was investigated by conducting a series of wind tunnel tests on sand dunes to determine the wind erosion magnitude, threshold breakaway velocity and wind bombardment resistance of the sand, and penetrometer measurements and microstructural studies (e.g. X-ray diffraction (XRD) analysis and scanning electron microscopy (SEM)) were also performed.
       Calcium carbonate production requires calcium ions and carbonate ions. Calcium ions can be obtained from various calcium sources such as calcium chloride, calcium hydroxide, and skim milk powder54,55. Carbonate ions can be produced by various microbial methods such as urea hydrolysis and aerobic or anaerobic oxidation of organic matter56. In this study, carbonate ions were obtained from the oxidation reaction of formate and acetate. In addition, we used calcium salts of formate and acetate to produce pure calcium carbonate, thus only CO2 and H2O were obtained as by-products. In this process, only one substance serves as a calcium source and a carbonate source, and no ammonia is produced. These characteristics make the calcium source and carbonate production method that we considered very promising.
       The corresponding reactions of calcium formate and calcium acetate to form calcium carbonate are shown in formulas (7)-(14). Formulas (7)-(11) show that calcium formate dissolves in water to form formic acid or formate. The solution is thus a source of free calcium and hydroxide ions (formulas 8 and 9). As a result of the oxidation of formic acid, the carbon atoms in formic acid are converted into carbon dioxide (formula 10). Calcium carbonate is ultimately formed (formulas 11 and 12).
       Similarly, calcium carbonate is formed from calcium acetate (equations 13–15), except that acetic acid or acetate is formed instead of formic acid.
       Without the presence of enzymes, acetate and formate cannot be oxidized at room temperature. FDH (formate dehydrogenase) and CoA (coenzyme A) catalyze the oxidation of formate and acetate to form carbon dioxide, respectively (Eqs. 16, 17) 57, 58, 59. Various bacteria are capable of producing these enzymes, and heterotrophic bacteria, namely Bacillus subtilis (PTCC #1204 (Persian Type Culture Collection), also known as NCIMB #13061 (International Collection of Bacteria, Yeast, Phage, Plasmids, Plant Seeds and Plant Cell Tissue Cultures)) and Bacillus amyloliquefaciens (PTCC #1732, NCIMB #12077), were used in this study. These bacteria were cultured in a medium containing meat peptone (5 g/L) and meat extract (3 g/L), called nutrient broth (NBR) (105443 Merck).
       Thus, four formulations were prepared to induce calcium carbonate precipitation using two calcium sources and two bacteria: calcium formate and Bacillus subtilis (FS), calcium formate and Bacillus amyloliquefaciens (FA), calcium acetate and Bacillus subtilis (AS), and calcium acetate and Bacillus amyloliquefaciens (AA).
       In the first part of the experimental design, tests were conducted to determine the optimum combination that would achieve maximum calcium carbonate production. Since the soil samples contained calcium carbonate, a set of preliminary evaluation tests was designed to accurately measure the CaCO3 produced by the different combinations, and mixtures of culture medium and calcium source solutions were evaluated. For each combination of calcium source and bacteria solution defined above (FS, FA, AS, and AA), optimization factors (calcium source concentration, curing time, bacteria solution concentration measured by optical density of the solution (OD), calcium source to bacteria solution ratio, and pH) were derived and used in the sand dune treatment wind tunnel tests described in the following sections.
       For each combination, 150 experiments were conducted to study the effect of CaCO3 precipitation and evaluate various factors, namely calcium source concentration, curing time, bacterial OD value, calcium source to bacterial solution ratio and pH during aerobic oxidation of organic matter (Table 1). The pH range for the optimized process was selected based on the growth curves of Bacillus subtilis and Bacillus amyloliquefaciens in order to obtain faster growth. This is explained in more detail in the Results section.
       The following steps were used to prepare the samples for the optimization phase. The MICP solution was first prepared by adjusting the initial pH of the culture medium and then autoclaved at 121 °C for 15 min. The strain was then inoculated in a laminar air flow and maintained in a shaking incubator at 30 °C and 180 rpm. Once the OD of the bacteria reached the desired level, it was mixed with the calcium source solution in the desired proportion (Figure 1a). The MICP solution was allowed to react and solidify in a shaking incubator at 220 rpm and 30 °C for a time that reached the target value. The precipitated CaCO3 was separated after centrifugation at 6000 g for 5 min and then dried at 40 °C to prepare the samples for the calcimeter test (Figure 1b). The precipitation of CaCO3 was then measured using a Bernard calcimeter, where CaCO3 powder reacts with 1.0 N HCl (ASTM-D4373-02) to produce CO2, and the volume of this gas is a measure of the CaCO3 content (Figure 1c). To convert the volume of CO2 to CaCO3 content, a calibration curve was generated by washing pure CaCO3 powder with 1 N HCl and plotting it against the evolved CO2. The morphology and purity of the precipitated CaCO3 powder were investigated using SEM imaging and XRD analysis. An optical microscope with a magnification of 1000 was used to study the formation of calcium carbonate around the bacteria, the phase of the formed calcium carbonate, and the activity of the bacteria.
       The Dejegh Basin is a well-known highly eroded region in the southwestern Fars Province of Iran, and the researchers collected wind-eroded soil samples from the area. The samples were taken from the soil surface for the study. Indicator tests on the soil samples showed that the soil was poorly sorted sandy soil with silt and was classified as SP-SM according to the Unified Soil Classification System (USC) (Figure 2a). XRD analysis showed that the Dejegh soil was mainly composed of calcite and quartz (Figure 2b). In addition, EDX analysis showed that other elements such as Al, K, and Fe were also present in smaller proportions.
       To prepare the laboratory dunes for wind erosion testing, the soil was crushed from a height of 170 mm through a 10 mm diameter funnel to a firm surface, resulting in a typical dune of 60 mm in height and 210 mm in diameter. In nature, the lowest density sand dunes are formed by aeolian processes. Similarly, the sample prepared using the above procedure had the lowest relative density, γ = 14.14 kN/m³, forming a sand cone deposited on a horizontal surface with an angle of repose of approximately 29.7°.
       The optimum MICP solution obtained in the previous section was sprayed onto the dune slope at application rates of 1, 2 and 3 lm-2 and then the samples were stored in an incubator at 30 °C (Fig. 3) for 9 days (i.e. the optimum curing time) and then taken out for wind tunnel testing.
       For each treatment, four specimens were prepared, one for measuring calcium carbonate content and surface strength using a penetrometer, and the remaining three specimens were used for erosion tests at three different velocities. In the wind tunnel tests, the amount of erosion was determined at different wind speeds, and then the threshold breakaway velocity for each treatment specimen was determined using a plot of erosion amount versus wind speed. In addition to the wind erosion tests, the treated specimens were subjected to sand bombardment (i.e., jumping experiments). Two additional specimens were prepared for this purpose at application rates of 2 and 3 L m−2. The sand bombardment test lasted 15 min with a flux of 120 g m−1, which is within the range of values ​​selected in previous studies60,61,62. The horizontal distance between the abrasive nozzle and the dune base was 800 mm, located 100 mm above the tunnel bottom. This position was set so that almost all the jumping sand particles fell on the dune.
       The wind tunnel test was conducted in an open wind tunnel with a length of 8 m, a width of 0.4 m and a height of 1 m (Figure 4a). The wind tunnel is made of galvanized steel sheets and can generate a wind speed of up to 25 m/s. In addition, a frequency converter is used to adjust the fan frequency and gradually increase the frequency to obtain the target wind speed. Figure 4b shows the schematic diagram of the sand dunes eroded by wind and the wind speed profile measured in the wind tunnel.
       Finally, to compare the results of the non-urealytic MICP formulation proposed in this study with the results of the urealytic MICP control test, dune samples were also prepared and treated with a biological solution containing urea, calcium chloride and Sporosarcina pasteurii (since Sporosarcina pasteurii has a significant ability to produce urease63). The optical density of the bacterial solution was 1.5, and the concentrations of urea and calcium chloride were 1 M (selected based on the values ​​recommended in previous studies36,64,65). The culture medium consisted of nutrient broth (8 g/L) and urea (20 g/L). The bacterial solution was sprayed on the dune surface and left for 24 hours for bacterial attachment. After 24 hours of attachment, a cementing solution (calcium chloride and urea) was sprayed. The urealytic MICP control test is hereinafter referred to as UMC. The calcium carbonate content of urealytically and non-urealytically treated soil samples was obtained by washing according to the procedure proposed by Choi et al.66
       Figure 5 shows the growth curves of Bacillus amyloliquefaciens and Bacillus subtilis in the culture medium (nutrient solution) with an initial pH range of 5 to 10. As shown in the figure, Bacillus amyloliquefaciens and Bacillus subtilis grew faster at pH 6-8 and 7-9, respectively. Therefore, this pH range was adopted in the optimization stage.
       Growth curves of (a) Bacillus amyloliquefaciens and (b) Bacillus subtilis at different initial pH values ​​of the nutrient medium.
       Figure 6 shows the amount of carbon dioxide produced in the Bernard limemeter, which represents precipitated calcium carbonate (CaCO3). Since one factor was fixed in each combination and the other factors were varied, each point on these graphs corresponds to the maximum volume of carbon dioxide in that set of experiments. As shown in the figure, as the calcium source concentration increased, the production of calcium carbonate increased. Therefore, the concentration of the calcium source directly affects the production of calcium carbonate. Since the calcium source and the carbon source are the same (i.e., calcium formate and calcium acetate), the more calcium ions are released, the more calcium carbonate is formed (Figure 6a). In the AS and AA formulations, calcium carbonate production continued to increase with increasing curing time until the amount of precipitate was almost unchanged after 9 days. In the FA formulation, the rate of calcium carbonate formation decreased when the curing time exceeded 6 days. Compared with other formulations, formulation FS showed a relatively low calcium carbonate formation rate after 3 days (Figure 6b). In formulations FA and FS, 70% and 87% of the total calcium carbonate production was obtained after three days, while in formulations AA and AS, this proportion was only about 46% and 45%, respectively. This indicates that the formic acid-based formulation has a higher CaCO3 formation rate at the initial stage compared with the acetate-based formulation. However, the formation rate slows down with increasing curing time. It can be concluded from Figure 6c that even at bacterial concentrations above OD1, there is no significant contribution to calcium carbonate formation.
       Change in CO2 volume (and corresponding CaCO3 content) measured by the Bernard calcimeter as a function of (a) calcium source concentration, (b) setting time, (c) OD, (d) initial pH, (e) ratio of calcium source to bacterial solution (for each formulation); and (f) maximum amount of calcium carbonate produced for each combination of calcium source and bacteria.
       Regarding the effect of the initial pH of the medium, Figure 6d shows that for FA and FS, the CaCO3 production reached a maximum value at pH 7. This observation is consistent with previous studies that FDH enzymes are most stable at pH 7-6.7. However, for AA and AS, the CaCO3 precipitation increased when the pH exceeded 7. Previous studies also showed that the optimal pH range for CoA enzyme activity is from 8 to 9.2-6.8. Considering that the optimal pH ranges for CoA enzyme activity and B. amyloliquefaciens growth are (8-9.2) and (6-8), respectively (Figure 5a), the optimal pH of AA formulation is expected to be 8, and the two pH ranges overlap. This fact was confirmed by experiments, as shown in Figure 6d. Since the optimum pH for B. subtilis growth is 7-9 (Figure 5b) and the optimum pH for CoA enzyme activity is 8-9.2, the maximum CaCO3 precipitation yield is expected to be in the pH range of 8-9, which is confirmed by Figure 6d (i.e., the optimum precipitation pH is 9). The results shown in Figure 6e indicate that the optimum ratio of calcium source solution to bacterial solution is 1 for both acetate and formate solutions. For comparison, the performance of different formulations (i.e., AA, AS, FA, and FS) was evaluated based on the maximum CaCO3 production under different conditions (i.e., calcium source concentration, curing time, OD, calcium source to bacterial solution ratio, and initial pH). Among the formulations studied, formulation FS had the highest CaCO3 production, which was approximately three times that of formulation AA (Figure 6f). Four bacteria-free control experiments were conducted for both calcium sources and no CaCO3 precipitation was observed after 30 days.
       The optical microscopy images of all the formulations showed that vaterite was the main phase in which calcium carbonate was formed (Figure 7). The vaterite crystals were spherical in shape69,70,71. It was found that calcium carbonate precipitated on the bacterial cells because the surface of the bacterial cells was negatively charged and could act as an adsorbent for divalent cations. Taking formulation FS as an example in this study, after 24 hours, calcium carbonate began to form on some bacterial cells (Figure 7a), and after 48 hours, the number of bacterial cells coated with calcium carbonate increased significantly. In addition, as shown in Figure 7b, vaterite particles could also be detected. Finally, after 72 hours, a large number of bacteria seemed to be bound by the vaterite crystals, and the number of vaterite particles increased significantly (Figure 7c).
       Optical microscopy observations of CaCO3 precipitation in FS compositions over time: (a) 24, (b) 48 and (c) 72 h.
       To further investigate the morphology of the precipitated phase, X-ray diffraction (XRD) and SEM analyses of the powders were performed. The XRD spectra (Fig. 8a) and SEM micrographs (Fig. 8b, c) confirmed the presence of vaterite crystals, as they had a lettuce-like shape and a correspondence between the vaterite peaks and the precipitate peaks was observed.
       (a) Comparison of X-ray diffraction spectra of formed CaCO3 and vaterite. SEM micrographs of vaterite at (b) 1 kHz and (c) 5.27 kHz magnification, respectively.
       The results of the wind tunnel tests are shown in Figure 9a, b. It can be seen from Figure 9a that the threshold erosion velocity (TDV) of the untreated sand is about 4.32 m/s. At the application rate of 1 l/m² (Figure 9a), the slopes of the soil loss rate lines for fractions FA, FS, AA and UMC are approximately the same as for the untreated dune. This indicates that the treatment at this application rate is ineffective and as soon as the wind speed exceeds the TDV, the thin soil crust disappears and the dune erosion rate is the same as for the untreated dune. The erosion slope of fraction AS is also lower than that of the other fractions with lower abscissas (i.e. TDV) (Figure 9a). The arrows in Figure 9b indicate that at the maximum wind speed of 25 m/s, no erosion occurred in the treated dunes at the application rates of 2 and 3 l/m². In other words, for FS, FA, AS and UMC, the dunes were more resistant to wind erosion caused by CaCO³ deposition at the application rates of 2 and 3 l/m² than at the maximum wind speed (i.e. 25 m/s). Thus, the TDV value of 25 m/s obtained in these tests is the lower limit for the application rates shown in Figure 9b, except for the case of AA, where the TDV is almost equal to the maximum wind tunnel speed.
       Wind erosion test (a) Weight loss versus wind speed (application rate 1 l/m2), (b) Threshold tear-off speed versus application rate and formulation (CA for calcium acetate, CF for calcium formate).
       Figure 10 shows the surface erosion of sand dunes treated with different formulations and application rates after the sand bombardment test and the quantitative results are shown in Figure 11. The untreated case is not shown because it showed no resistance and was completely eroded (total mass loss) during the sand bombardment test. It is clear from Figure 11 that the sample treated with biocomposition AA lost 83.5% of its weight at the application rate of 2 l/m2 while all other samples showed less than 30% erosion during the sand bombardment process. When the application rate was increased to 3 l/m2, all treated samples lost less than 25% of their weight. At both application rates, compound FS showed the best resistance to sand bombardment. The maximum and minimum bombardment resistance in the FS and AA treated samples can be attributed to their maximum and minimum CaCO3 precipitation (Figure 6f).
       Results of bombardment of sand dunes of different compositions at flow rates of 2 and 3 l/m2 (arrows indicate wind direction, crosses indicate wind direction perpendicular to the plane of the drawing).
       As shown in Figure 12, the calcium carbonate content of all the formulas increased as the application rate increased from 1 L/m² to 3 L/m². In addition, at all application rates, the formula with the highest calcium carbonate content was FS, followed by FA and UMC. This suggests that these formulas may have higher surface resistance.
       Figure 13a shows the change in surface resistance of untreated, control and treated soil samples measured by permeameter test. From this figure, it is evident that the surface resistance of UMC, AS, FA and FS formulations increased significantly with the increase of application rate. However, the increase in surface strength was relatively small in AA formulation. As shown in the figure, FA and FS formulations of non-urea-degraded MICP have better surface permeability compared to urea-degraded MICP. Figure 13b shows the change in TDV with soil surface resistance. From this figure, it is clearly evident that for dunes with surface resistance greater than 100 kPa, the threshold stripping velocity will exceed 25 m/s. Since in situ surface resistance can be easily measured by permeameter, this knowledge can help to estimate TDV in the absence of wind tunnel testing, thereby serving as a quality control indicator for field applications.
       The SEM results are shown in Figure 14. Figures 14a-b show the enlarged particles of the untreated soil sample, which clearly indicates that it is cohesive and has no natural bonding or cementation. Figure 14c shows the SEM micrograph of the control sample treated with urea-degraded MICP. This image shows the presence of CaCO3 precipitates as calcite polymorphs. As shown in Figures 14d-o, the precipitated CaCO3 binds the particles together; spherical vaterite crystals can also be identified in the SEM micrographs. The results of this study and previous studies indicate that the CaCO3 bonds formed as vaterite polymorphs can also provide reasonable mechanical strength; our results show that the surface resistance increases to 350 kPa and the threshold separation velocity increases from 4.32 to more than 25 m/s. This result is consistent with the results of previous studies that the matrix of MICP-precipitated CaCO3 is vaterite, which has reasonable mechanical strength and wind erosion resistance13,40 and can maintain reasonable wind erosion resistance even after 180 days of exposure to field environmental conditions13.
       (a, b) SEM micrographs of untreated soil, (c) MICP urea degradation control, (df) AA-treated samples, (gi) AS-treated samples, (jl) FA-treated samples, and (mo) FS-treated samples at an application rate of 3 L/m2 at different magnifications.
       Figure 14d-f shows that after treatment with AA compounds, calcium carbonate was precipitated on the surface and between the sand grains, while some uncoated sand grains were also observed. For AS components, although the amount of CaCO3 formed did not increase significantly (Fig. 6f), the amount of contacts between sand grains caused by CaCO3 increased significantly compared with AA compounds (Fig. 14g-i).
       From Figures 14j-l and 14m-o it is clear that the use of calcium formate as a calcium source leads to a further increase in CaCO3 precipitation compared to the AS compound, which is consistent with the calcium meter measurements in Figure 6f. This additional CaCO3 appears to be mainly deposited on the sand particles and does not necessarily improve the contact quality. This confirms the previously observed behavior: despite the differences in the amount of CaCO3 precipitation (Figure 6f), the three formulations (AS, FA and FS) do not differ significantly in terms of anti-eolian (wind) performance (Figure 11) and surface resistance (Figure 13a).
       In order to better visualize the CaCO3 coated bacterial cells and the bacterial imprint on the precipitated crystals, high magnification SEM micrographs were taken and the results are shown in Figure 15. As shown, calcium carbonate precipitates on the bacterial cells and provides the nuclei required for the precipitation there. The figure also depicts the active and inactive linkages induced by CaCO3. It can be concluded that any increase in inactive linkages does not necessarily lead to further improvement in mechanical behavior. Therefore, increasing CaCO3 precipitation does not necessarily lead to higher mechanical strength and the precipitation pattern plays an important role. This point has also been studied in the works of Terzis and Laloui72 and Soghi and Al-Kabani45,73. To further explore the relationship between precipitation pattern and mechanical strength, MICP studies using µCT imaging are recommended, which is beyond the scope of this study (i.e., introducing different combinations of calcium source and bacteria for ammonia-free MICP).
       CaCO3 induced active and inactive bonds in samples treated with (a) AS composition and (b) FS composition and left an imprint of bacterial cells on the sediment.
       As shown in Figures 14j-o and 15b, there is a CaCO film (according to EDX analysis, the percentage composition of each element in the film is carbon 11%, oxygen 46.62% and calcium 42.39%, which is very close to the percentage of CaCO in Figure 16). This film covers the vaterite crystals and soil particles, helping to maintain the integrity of the soil-sediment system. The presence of this film was observed only in the samples treated with the formate-based formulation.
       Table 2 compares the surface strength, threshold detachment velocity, and bioinduced CaCO3 content of soils treated with urea-degrading and non-urea-degrading MICP pathways in previous studies and this study. Studies on the wind erosion resistance of MICP-treated dune samples are limited. Meng et al. investigated the wind erosion resistance of MICP-treated urea-degrading dune samples using a leaf blower,13 whereas in this study, non-urea-degrading dune samples (as well as urea-degrading controls) were tested in a wind tunnel and treated with four different combinations of bacteria and substances.
       As can be seen, some previous studies have considered high application rates exceeding 4 L/m213,41,74. It is worth noting that high application rates may not be easily applicable in the field from an economic point of view due to the costs associated with water supply, transportation and application of large volumes of water. Lower application rates such as 1.62-2 L/m2 also achieved fairly good surface strengths of up to 190 kPa and TDV exceeding 25 m/s. In the present study, dunes treated with formate-based MICP without urea degradation achieved high surface strengths that were comparable to those obtained with the urea degradation pathway in the same range of application rates (i.e., samples treated with formate-based MICP without urea degradation were also able to achieve the same range of surface strength values ​​as reported by Meng et al., 13, Figure 13a) at higher application rates. It can also be seen that at the application rate of 2 L/m2, the yield of calcium carbonate for wind erosion mitigation at a wind speed of 25 m/s was 2.25% for the formate-based MICP without urea degradation, which is very close to the required amount of CaCO3 (i.e. 2.41%) compared to dunes treated with the control MICP with urea degradation at the same application rate and the same wind speed (25 m/s).
       Thus, it can be concluded from this table that both the urea degradation pathway and the urea-free degradation pathway can provide quite acceptable performance in terms of surface resistance and TDV. The main difference is that the urea-free degradation pathway does not contain ammonia and therefore has a lower environmental impact. In addition, the formate-based MICP method without urea degradation proposed in this study seems to perform better than the acetate-based MICP method without urea degradation. Although Mohebbi et al. studied the acetate-based MICP method without urea degradation, their study included samples on flat surfaces9. Due to the higher degree of erosion caused by eddy formation around the dune samples and the resulting shear, which results in lower TDV, the wind erosion of the dune samples is expected to be more obvious than that of flat surfaces at the same speed.