The former lake of Texcoco located in the valley of Mexico City (Mexico) at an altitude of 2240 m above sea level with a mean annual temperature of 16°C and annual precipitation of 705 mm was drained from the 17th century onwards to avoid flooding in Mexico City (O’Hara and Metcalfe, 1997). The drainage of the lake left a soil with  a   high   Ph   and    salinity    and    little   vegetation (Dendooven et al., 2010). During the dry season, the wind erosion was high. However, during the rainy season, flooding occurred frequently. The groundwater, which is highly alkaline and saline, was just under the soil surface and after heavy rainfall the area flooded. The national water authority (Commission Nacional de Agua, CNA) installed  drainage   pipes   so   that   the  area  could  be vegetated to stop wind erosion and dust storms in Mexico City (Luna-Guido et al., 2000). However, previous research showed that the soil of the former lake Texcoco is N depleted (Conde et al., 2005). It can be speculated that the high pH and salinity inhibits N₂ fixing microorganisms thereby limiting the amount of mineral N that enters the soil (Barua et al., 2011; Welsh et al., 2007).

Urea is cheap and often applied to fertilize crops. It could easily be applied to grass, shrubs and trees that might be used to vegetate the former lake bed. Hydrolysis of urea can occur in three ways, biotic (Burton and Prosser, 2001), abiotic in soil with a high pH (Ghandi and Paliwal, 1976) and abiotic through extracellular ureases (Conrad, 1996) generating two NH₃ molecules. As such, concentration of NH₄⁺ will increase in urea-amended soil (Burton and Prosser, 2001). However, if most of the process is biological then the high pH and salinity might inhibit the release of NH₄⁺. Additionally, application of urea is known to increase emissions of nitrous oxide (N₂O) and might affect emissions of carbon dioxide (CO₂) and methane (CH₄) (Wang et al., 2011). Hydrolysis of urea will generate ammonium ions that can be oxidized by nitrifiers, first to NO₂^- and subsequently to NO₃^-. However, high electrolytic conductivity (EC) is known to inhibit the activity of nitrifiers so emissions of N₂O after application of urea might not increase (Zhu et al., 2011). Therefore, four soils of the former lake Texcoco with different pH and EC were amended with urea, with or without acetylene (C₂H₂), known to inhibit nitrification (Bateman and Baggs, 2005), while emissions of CO₂, CH₄ and N₂O were monitored in an aerobic incubation. Acetylene was applied to half of the soil samples so that the importance of the nitrification process in the emissions of N₂O could be determined. The objective of this study was to investigate the effect of high alkalinity and salinity on hydrolysis of urea, emissions of GHG and dynamics of mineral N (ammonium (NH₄⁺), nitrite (NO₂^-) or nitrate (NO₃^-)).

Soil sampling

Details of the soil of the former lake Texcoco can be found in Dendooven et al. (2010) and mineralogy in Gutiérrez-Castorena et al. (2005). Part of the former lake bed has been drained and irrigated with sewage effluent from a waste water plant to remove excess of salt (Luna-Guido et al., 2000). For instance, the concentration of the sodium ions (Na⁺) decreased from 21 g kg^-1 dry soil to 3 g kg^-1 dry soil after 8 years of flooding and the chloride ions (Cl^-) from 21 g kg^-1 dry soil to undetectable amounts (Luna-Guido et al., 2000). The Distichlis spicata, an indigenous grass with a high tolerance to salt and Na⁺ and tamarix (Tamarix species) have been introduced since the early 1970s to control erosion, and they now cover much of the area. More details on the vegetation and the effluents used to drain the plots can be found in Luna-Guido et al. (2000).

At four locations with different EC and pH due to different periods of  drainage,   five  approximately  400 m²  plots  were  defined  and sampled by augering the 0 to 15 cm layer 30-times with a stony soil auger with diameter 7 cm (Eijkelkamp, Nl) at random on 7 March 2011. The 30 soil samples taken from each site (n = 4) and plot (n = 5) were pooled, 5 mm sieved and characterized (Table 1). As such, 20 soil samples were obtained. Details of the sampled soils can be found in Table 1. The soil with an EC 3.3 dS m^-1 was denoted soil A, with EC 88.3 dS m^-1 soil B, with EC 96.9 dS m^-1 soil C and with EC 121 dS m^-1 soil D. This field based replication was maintained in the incubation study.

Treatments and experimental set-up

The experimental design was a completely randomized 2 ´ 4 factorial with five replications (maintained from the field site replications for each soil). The factors were four soil types and four soil amendments which were: 1) 200 mg N kg^-1 applied as urea; 2) 200 mg N kg^-1 applied as urea plus acetylene (C₂H₂) at 0.1%; 3) unamended soil; and 4) unamended soil plus C₂H₂ at 0.1%.

Sixteen sub-samples of 10 g dry soil from each of the four soils and five sampled plots were added to 120 ml flasks. Eight soil samples were adjusted to 40% water holding capacity (WHC) by adding distilled water and eight by adding an urea solution. The flasks were air-tight sealed with a Suba-seal. Four of the flasks amended with distilled water and four with the urea solution were injected with 0.1 ml acetylene (C₂H₂) to inhibit nitrification (Bateman and Baggs, 2005). Additionally, 15 flasks without soil were air-tight sealed and incubated in the same way to determine the concentration of CO₂, N₂O and CH₄ in the atmosphere. The flasks were incubated in the dark at 25 ± 2°C. After 0, 1, 3 and 7 days, one flask was selected from each soil and treatment at random and the headspace was analyzed for CO₂, CH₄ and N₂O. Additionally, three flasks without soil were selected at the same time and the headspace analyzed. The flasks were opened, the soil removed and extracted for mineral N (ammonium (NH₄⁺), nitrate (NO₃^-) and nitrite (NO₂^-)) with 0.5 M K₂SO₄.

Chemical analyses

Details of the techniques used to measure WHC, pH, EC, total N and soil particle size distribution can be found in Ruíz-Valdiviezo et al. (2010). The extracted NH₄⁺, NO₃^- and NO₂^- were measured colourimetrically with a San Plus System-SKALAR automatic analyzer (Breda, The Netherlands).

The headspace of the vials was analyzed for CO₂ and N₂O on an Agilent Technologies 4890D gas chromatograph fitted with an electron capture detector (ECD) and CH₄ on an Agilent Technologies 4890D gas chromatograph fitted with a flame ionization detector (FID). Details of the columns used, gas flow, and oven, detector and injector temperatures can be found in Ruíz-Valdiviezo et al. (2010). Concentrations of CO₂, N₂O and CH₄ were calculated by comparing peak areas against a standard curve prepared from known concentrations, 10 ppm N₂O in N₂, 5 ppm CH₄ in N₂ and 2500, 20000 and 40000 ppm CO₂ in N₂, every time samples were analysed.

Statistical analysis

The experimental design was a completely randomized 2 ´ 4 factorial with five replications. The factors were four soil types with different EC and four soil treatments, that is, unamended soil, urea-amended soil, C₂H₂ applied soil, and urea + C₂H₂. Emission of CH₄, N₂O and CO₂ was regressed on elapsed time using a linear regression model, which was forced to pass through the origin, but allowed different slopes (production rates) for each treatment. Significant differences  for  the  production  of  CH₄,  N₂O  and  CO₂ production between treatments, soil and their interactions were determined using PROC MIXED (SAS Institute Inc. 1989).

Concentrations of NH₄⁺, NO₂^- and NO₃^- were subjected to an analysis of variance using PROC GLM (SAS Institute Inc. 1989) to test for significant differences between soils, treatments and their interactions with Tukey’s Studentized Range test. All data presented were the mean of five replicates, that is, n = 5.

Emissions of CO₂, CH₄ and N₂O

In the unamended soil, the emission of CO₂ was largest in soil A and lowest in soil D (Figure 1a). Application of C₂H₂ to the unamended soil had no significant effect on the CO₂ emission rates. The emission of CH₄ was similar in all unamended soils and was not affected by the application of C₂H₂ (Figure 1b). In the unamended soil, the emission of N₂O was significantly larger in soil A than in the other soils (P<0.05) (Figure 1c). Application of C₂H₂ had no significant effect on the N₂O emission rate.

Application of urea increased the emission of CO₂ in soils A, B and D significantly, but not in soil C (P<0.05) (Figure 1d). Application of C₂H₂ to the urea-amended soil decreased the emission of CO₂ significantly in soils A and B, but not in soils C and D (P<0.05). The emission of CO₂ was similar in the C₂H₂-amended soils applied with or without urea. Application of urea did not affect the emission of CH₄ and was similar in the C₂H₂-amended soils applied with or without urea (Figure 1e). Application of urea increased the emission of N₂O significantly in soil A, but not in the other soils (P<0.05) (Figure 1f). Application of C₂H₂ to the urea-amended soil decreased the emission of N₂O in soils A and B, but not in soils C and D.

The emission of CO₂ was significantly affected by soil and the interactions between urea ´ C₂H₂ and soil ´ urea ´ C₂H₂ (P<0.05) (Table 2). The emission of N₂O was significantly affected by the different interactions between urea, soil and C₂H₂, and the emission of CH₄ was affected significantly only by soil (P<0.05).

Dynamics of inorganic N

The concentration of NH₄⁺ was similar in the unamended soils and soils applied with C₂H₂ (Figure  2a).  Application of C₂H₂ to the unamended soil reduced the concentration of NO₂^- significantly in soils B and C, but not in soils A and D (P<0.05) (Figure 2b). The concentration of NO₃^- was similar in the unamended soils and soils with applied C₂H₂ (Figure 2c).

Application of urea increased the concentration of NH₄⁺ significantly in all soils and the increase was most accentuated in soil A (P<0.05) (Figure 2d). Application of C₂H₂ to the urea-amended soil decreased the amount of NH₄⁺ significantly in soil A, but not in the other soils (P<0.05). Application of urea increased the concentration of NO₂^- significantly in soil A, but not in the other soils (P<0.05) (Figure 2e). Application of C₂H₂ to the urea-amended soil decreased the amount of NO₂^- significantly in soil A, but not in the other soils (P<0.05). Application of urea increased the concentration of NO₃^- significantly in soil A, but not in the other soils (Figure 2f). Application of C₂H₂ to the urea-amended soil decreased the amount of NO₃^- in soil A, but not in the other soils.

The concentrations of NH₄⁺, NO₂^- and NO₃^- were affected significantly by urea, C₂H₂, soil and their interactions, except for the effect of urea and its interaction with C₂H₂ on the concentration of NO₃^-  (Table 2).

Emissions of CO₂, CH₄ and N₂O

The emission of CO₂ decreased with increased EC in the soils. It is well known that increased salinity reduces the soil microbial biomass and inhibits microbial activity (Setia et al., 2011a, b). However, it has to be remembered that other characteristics, such as soil organic matter content, clay content and pH, are also known to affect microbial activity and thus emissions of CO₂ (Setia et al., 2011a, b).

The application of urea increased the CO₂ emission rate significantly in soils A, B and D compared to the unamended soil. It is well known that application of an N fertilizer to an N depleted soil can increase emission of CO₂ as microbial activity is stimulated (Wang et al., 2011). The high salt content in the Texcoco soils will inhibit N₂ fixation, which will limit the N content of the soil (Barua  et  al.,  2011;  Welsh  et  al.,  2007).  Additionally,  hydrolysis of urea will release CO₂ (Snyder et al., 2009). The emission of CO₂ will thus increase with 86 mg CO₂ after the application of 200 mg N kg^-1 soil if all urea was hydrolyzed.

The application of C₂H₂ had no effect on emission of CO₂. Acetylene can be used by certain organisms, e.g. Rhodococcus opacus, Rhodococcus ruber and Gordona species, as C substrate thereby increasing emission of CO₂ (Rosner et al., 1997). Soil characteristics are known to affect C₂H₂ degradation (Brzezinska et al., 2011), although the limited time that the soil microorganisms were exposed to C₂H₂ (7 days) might have reduced the possibility that they use C₂H₂ as C source, that is, they were not yet adapted.

Agricultural soils can be a source or a sink for CH₄, but they are normally a sink and fluxes are normally small (Wang et al., 2011). Large amounts of CH₄ are only emitted from paddy soils or wetlands (Wright et al., 2011). Production of CH₄ occurs under anaerobic and oxidation under aerobic conditions. Although, the soils were incubated aerobically, emission of CH₄ occurred in all soils. Anaerobic micro-sites exist even in a soil at 40% WHC that will stimulate production of CH₄, and oxidation of CH₄ did not match production. The high salt content might have inhibited methanotrophic activity.

Application  of  urea  or C₂H₂ did not affect emissions of CH₄. Application of NH₄⁺, released after the hydrolysis of urea, is known to inhibit oxidation of CH₄, but not in the Texcoco soils (Stiehl-Braun et al., 2011; Xu et al., 2011). Aronson and Helliker (2010) reported after a meta-analysis that not only the amount of N applied, but also the history of the soil affected the inhibitory effect. They reported that managed soil and soil with a longer duration of fertilizer application showed greater inhibition of CH₄ uptake with added N. The Texcoco soil was not fertilized and is N depleted so it can be assumed that N fertilizer would not inhibit CH₄ oxidation. Bronson and Mosier (1994) reported a strong inhibitory effect of C₂H₄ on oxidation of CH₄ (76 to 100% inhibition) in two soils. No such inhibitory effect was found in the Texcoco soil, so it can be speculated that little or no CH₄ oxidation occurred as stated before.

Application of urea increased emission of N₂O in soil A compared to the unamended soil, but not in the other soils. It is well known that application of urea to soil increases emission of N₂O (Wang et al., 2011). Emission of N₂O from soil is mainly due to nitrification, that is, the oxidation of NH₄⁺ to NO₂^- and NO₂^- to NO₃^- under aerobic conditions and denitrification, that is, the reduction of NO₃^- to NO₂^-, N₂O and N₂ under anaerobic conditions (Wrage et al., 2001). As the soil was incubated under aerobic  conditions and the concentration of NH₄⁺ sharply increased after application of urea and its subsequent hydrolysis, emission of N₂O was most likely due to oxidation of NH₄⁺. Application of C₂H₂ (as an inhibitor of the oxidation of NH₄⁺) to soil A sharply reduced the emission of N₂O confirming that oxidation of NH₄⁺ was the main source of N₂O emission. The emission of N₂O in soil A amended with urea plus C₂H₂, however, was still higher than in the unamended control soil. As such, although the soil was incubated aerobically, it is likely that some anaerobic microsites were formed in soil stimulating denitrification and thus emission of N₂O.

Emission of N₂O also increased when urea was added to soil B, and C₂H₂ decreased it. As such, nitrification contributed to the emission of N₂O in soil B. No increase in emission of N₂O occurred in soils C and D amended with urea. As such, the high salt content inhibited the nitrification process in soils C and D. Application of urea might be used as N fertilizer for a pioneering vegetation to minimise N₂O emission in the alkaline saline Texcoco soil. However, it would have to be injected into the soil as the high pH will favour NH₃ volatilization.

The soils were incubated at a constant water content in this experiment. In the field, water content will fluctuate continuously thereby changing soil conditions constantly. These constantly changing conditions will put a further strain on the microbial population. The soil micro-organisms will have to adapt strategies to survive a dried out or flooded environment, to altering anaerobic and aerobic conditions, and salt concentrations that decrease in the rainy season but increase in the dry season when evaporation will concentrate the salt ions mostly in the upper soil layer.

Dynamics of mineral N

Concentration of NH₄⁺ increased in all urea-amended soils. Hydrolysis of urea can occur in three ways, biotic (Burton and Prosser, 2001), abiotic in soil with a high pH (Ghandi and Paliwal, 1976) and abiotic through extracellular ureases (Conrad, 1996) generating two NH₃ molecules. As such, concentration of NH₄⁺ will increase in urea-amended soil (Burton and Prosser, 2001) as found in this study. The increase in the concentration of NH₄⁺ was lower in soil with EC ≥88.3 dS m^-1 than in soil with EC 3.3 dS m^-1, so the high EC had an inhibitory effect on the hydrolysis of urea (Wilson et al., 1999).

Application of C₂H₂ reduced the concentration of NO₂^- in the unamended or urea-amended soils A, B and C, but not in soil D. Consequently, oxidation of NH₄⁺ occurred in soil with EC ≤96.9 dS m^-1. Ammonium oxidizing organisms have been found in extreme environments (Sorokin and Kuenen, 2005) so it would come as no surprise that oxidation of NH₄⁺ occurred in soil with EC≤96.9 dS m^-1. Oxidation of NO₂^- only occurred in the soil with the lowest EC. Although NO₂^- oxidizing bacteria have been isolated from  alkaline  environments  (Sorokin et al., 1998), it might well be that the extreme high EC inhibited NO₂^- oxidation as the energetic gain from this process is low (Oren, 2011).

Under the experimental conditions, NO₂^- oxidation seems to be absent in soils with EC ≥88.3 dS m^-1. However, NH₄⁺ oxidation occurred even at 96.9 dS m^-1 suggesting that NH₄⁺ oxidation is less sensitive to salinity than NO₂^- oxidation. From a biological point of view, there are several possible explanations for this phenomenon. First, aerobic ammonium oxidation yields more energy for growth than nitrite oxidation (Bock and Wagner, 2006). Second, aerobic NH₄⁺ oxidation is done not only by a restricted group of Bacteria, but also by Archaea belonging to the phylum Thaumarchaeota (Leininger et al., 2006). These Archaea have a different physiology than NH₄⁺ oxidizing Bacteria. In some environments, these Archaea can even be the major NH₄⁺ oxidizers (Prosser and Nicol, 2008). Furthermore, even when aerobic NH₄⁺ and NO₂^- oxidizers form tight associations, that is, the NO₂^- produced by NH₄⁺ oxidizers is consumed by NO₂^- oxidizers, there seems to be ecophysiological differences between both groups that might be in part the consequence of the evolution of their metabolic life styles. Nitrite oxidizers can be heterotrophic/mixotrophic or strict chemotrophics (Bock and Wagner, 2006), and even between them there are differences since Nitrospira species can be K-strategists with high substrate affinity and low growth rate, while Nitrobacter species might be r-strategists (Schramm et al., 1999). However, more studies need to be done as new nitrite oxidizers groups are emerging and their physiological must still be studied.

In this study, the microbial population was not investigated. It would be interesting to study the microbial population in each of the treatments and investigate which organisms were involved in each of the processes discussed, e.g. nitrifiers, methanogens and methanotrophs. A transcriptomics analysis would surely reveal genes that are relevant in these extreme environments, but absent in more normal soil conditions.

It was found that urea increased emission of CO₂ in all soils and emission of N₂O in soil with EC ≤88.3 dS m^-1, but emission of CH₄ was not affected. Hydrolysis of urea occurred in all soils although it was significantly slower in soil with EC ≥88.3 dS m^-1. Oxidation of NH₄⁺ occurred in soil with EC≤96.9 dS m^-1, but oxidation of NO₂^- only in soil with EC 3.3 dS m^-1_.

The authors have not declared any conflict of interests.

The research was funded by Cinvestav. C.V.-E. received grant-aided support  by  ‘Consejo  Nacional  de  Ciencia y Tecnología’ (CONACyT).