Nitrogen cycling, river-floodplain system, modelling approach.
This chapter reviews studies about nitrogen cycling in the soil-plant-atmosphere system, the nitrate pollution in the aquatic ecosystem, the characterization and limit factors of denitrification function, the definition of river-floodplain system and the biogeochemical cycling occurs in this system. The models simulating the hydrologic and biogeochemical processes at the river-floodplain interface are introduced, and the objectives of the thesis are described also.
Nitrogen is a fundamental component in living organisms, is a component in all amino acids and is present in the bases that make up nucleic acids such as RNA (Ribonucleic Acid) and DNA (Deoxyribonucleic Acid). It is also a key element that controls the functioning of lots of terrestrial, freshwater and marine ecosystems (Vitousek et al., 1997). Nitrogen exists in both organic and mineral form in the environment. The inorganic forms of nitrogen are shown in Table 1.
Large numbers of transformation of nitrogen exist in the soil-plant-atmosphere system, both biological and physico-chemical process are included. The movement and transform of nitrogen in the soil-plant-atmosphere system is shown in Figure 1.
Nitrogen enters into ecosystems main through biological fixation and artificial fertilization. Some free living or symbiotic bacteria could combine gaseous nitrogen with hydrogen to produce ammonia. The chemical fertilizers produced by industrial Haber-Bosch process which uses high temperature and pressure to convert nitrogen gas and a hydrogen source into ammonia is globally used in agricultural systems.
NH4+ and NO3- could be uptake by plants and microorganisms and incorporate into organic forms, this process is called immobilization. Organic nitrogen is the main form of nitrogen in the soil, up to 90% of the nitrogen stored as organic form, either in living organisms or in humus originating from decomposition of organism residues. The organic nitrogen could be decomposed by microorganisms through mineralization and ammonium is produced by enzymatic degradation in this process. Fertilized and mineralized NH3 would escape to atmosphere by volatilization or convert to NO3- by nitrification. Nitrification is the biological process that converts NH3 or NH4+ to NO2- followed by the oxidation of NO2- to NO3-, and NH4+ is rapidly converted to NO3- in the majority of agriculture soil. NH4+ could also be converted into dinitrogen by anammox. Anammox is an important microbial process which was found in 1990s, NO2- and NH4+ are converted directly into dinitrogen gas in this process in the anaerobic condition. NO3- could convert to N2 through the biological process called denitrification. The detail process of denitrification is described in 2.1.3.
Nitrogen is usually a limiting element for primary production, fertilizer produced by the industry process has sustained the global increased population over the past century (Gruber and Galloway, 2008). However, the excessive chemical N-fertilization during the past century caused environmental and health problems, like more greenhouse gases – nitrous oxide and ammonia released to the atmosphere, soil and water body acidification and eutrophication caused by the excessive inorganic nitrogen in aquatic ecosystem.
Nitrate pollution in the aquatic ecosystems
Nitrate pollution that enter aquatic ecosystems via point and nonpoint sources had drawn worldwide attention for a long time (Bijay-Singh et al., 1995; Carpenter et al., 1998; Jalali, 2011). The excess nitrate content in the water body could cause eutrophication, it could also cause health problems such as methaemoglobinaemic and cancer (Camargo and Alonso, 2006; McIsaac et al., 2001). European Union and World Health Organization had both set the standard for nitrate concentration at 50 mg•L-1 for drinking water.
Nitrate, as a negatively charged ion, is repelled by the negative charged clay mineral surfaces in soil. Nitrate is the primary form of nitrogen leached into groundwater and the most common contaminant of nitrogen in aquifer systems (Freeze and Cherry, 1979). The leaching of nitrate away from soil profile is a problem to both agricultural production and environment quality. The main two factors that impact the leaching of nitrate from the root zone to shallow groundwater are the amount of nitrate in the soil profile that above the amount required by plant and the vertical drainage volume. Crop system condition (crop type, rotation, irrigation and fertilization) and climate condition are regarded as have influence on nitrate leaching (Meisinger and Delgado, 2002; Simmelsgaard, 1998).
The climate impacts the water cycle directly. The soil water content and percolated water volume are significantly influenced by precipitation, temperate and humidity. The climate change may cause changes in temperature and precipitation, and will impact the agricultural nitrate cycling through changes in both soil processes and agricultural productivity (Stuart et al., 2011).
In the agriculture system, irrigation is a general action that provides water to crops during draught stress period. Irrigated water accounts for more than 60% of the freshwater use in southern Europe. After the irrigated water excess soil storage capacity, the nitrate would percolate to shallow groundwater along with soil water during heavy irrigation periods. Wang et al. (2010) found that the leached nitrate under heavy irrigation could arrive 60% of the accumulated N in the soil profile.
Fertilizers were widely applied in the agriculture system to stimulate the crop production, preferentially in regions where irrigation is available, and soil and climatic conditions are favorable for the growth of crop plants (Bijay-Singh et al., 1995). The massive fertilization could lead to nitrate accumulation in the soil profiles after successive cropping rotation (Westerman et al., 1994; Zhao et al., 2006). In agricultural region, especially in the irrigation areas, fertilizer is the main source of nitrate contamination of groundwater (Mishima et al., 2010).
Crop type has influence on the nitrate uptake, soil water drainage and soil microbial communities composition (Bending et al., 2002; Canter, 1996). The land cover and crop type in crop rotation have important influence on the leaching of nitrate (Beaudoin et al., 2005; Justes et al., 1999). Johnson et al. (2002) compared standard, intermediate and protective systems, the leaching of nitrate were found significantly different in these three systems, the protective system was proved able to substantially decrease nitrate losses.
The characteristics and depths of soils have impact on the transfer of the water and the solute nitrate in the soil profile. The water storage capacity increased along with the rise of soil depth, and the clay soil has a greater water holding capacity than sandy soil. Beaudoin et al. (2005) found that the nitrate concentration in the shallow groundwater was lowest in deep loamy soils and greatest in shallow loamy sand soils in the agricultural area.
Denitrification is a biological process that transforms nitrate into N2 gas in the anaerobic environment, microaerophilic, and occasionally aerobic conditions. In this process, N oxides rather than the general preferred oxygen are the terminal electron acceptors. It occurs in four steps, nitrate (NO3-) to nitrite (NO2-), NO2- to nitric oxide (NO), NO to nitrous oxide (N2O) and N2O to N2, the process can be arrested at any of the intermediate stages. Each step is catalyzed by different enzymes, and not all the denitrifying bacterium have the capacity to be involved in the sequence of the four steps (Bothe et al., 2006; Zumft, 1997) (Figure 2).
The conditions required by denitrification are the presence of N oxides as the electron acceptors, and then the presence of organic carbon which is the electron donor, finally the anaerobic environment which controls the activity of denitrifying enzymes. Except the three conditions, pH and temperature are regularly identified as the limit factors also.
Denitrification usually occurs at low dissolved oxygen concentration condition, Körner and Zumft (1989) found that the dissolved oxygen concentration thresholds of nitrate reductase, nitrite reductase were 5 mg•L-1 and 2.5 mg•L-1 independently. Rivett et al. (2008) reviewed the studies of the dissolved oxygen concentration threshold in the groundwater (Table 2).
The relationship between pH and denitrification are complicate, both the rate of the process and the ratio of its gaseous products depend on pH. Studies found that the liberate of N2O and the ratio N2O:N2 increased along with the decrease of soil pH (ŠImek and Cooper, 2002). The optimum temperature range of denitrifying enzymes is between 25 and 35 °C, but denitrification could occur in the range 2–50 °C (Rivett et al., 2008).
Organic carbon as the ‘energy’ of denitrifying bacteria is necessary in the denitrifying process. The complexity compositions of organic carbon in the ecosystems makes it difficult to identify the effective carbon source (Dodla et al., 2008; Hume et al., 2002). Dissolved organic carbon (DOC) or Bioavailable dissolved organic carbon (BDOC) are taken for carbon sources of denitrification in most studies (Hill et al., 2000; Inwood et al., 2005; Peterson et al., 2013). Furthermore, particulate organic carbon (POC) could enhance denitrification rate in both aquatic and terrestrial ecosystems (Arango et al., 2007; Stelzer et al., 2014; Stevenson et al., 2011). The dominant limit factors of denitrification in terrestrial and aquatic systems are nitrate and organic carbon (McClain et al., 2003). The BDOC content is around 4-54 % of DOC in the surface water (Servais et al., 1989; Wickland et al., 2012; Wiegner et al., 2006) and only around 8 % in the groundwater (Shen et al., 2014). Compared with the surface water, the denitrification limitation caused by organic carbon availability in groundwater is more important.
Denitrification is an essential branch of the global N cycle, is also the main biological process in charge of emissions of nitrous oxide. The spatially distributed global models of denitrification applied by Seitzinger et al. (2006) propose that the largest portion occurs in continental shelf sediments, which account for around 44% of total global denitrification.
Freshwater systems (groundwater, lakes, rivers) taken about 20% and estuaries 1% of total global denitrification. The denitrification in terrestrial soils and oceanic oxygen minimum zones represented 22% and 14% respectively.
Floodplain is the flat land area adjacent to a stream or river that experience flooding during high discharge period, is the result of both erosion and deposition (Brown, 1997). The size, form and vegetation type of floodplains are highly variable depend on the size, location and hydraulic condition of the rivers. Floodplain take a significant proportion of the earth’s surface, around 2% of African, 3% of South America and a greater proportion of tropical Asia (Gerrard, 1992).
Alluvial soils are generally associated with floodplain, originated from ancient riverbeds and the deposited sediment taken by the river water during flooding periods. The soil materials and drainage condition are greatly influenced by the natural of alluviums, since alluviums vary in different sectors the basin, the characteristics of soils are different also. Except the origin alluvium, water dynamic, living organisms and time all have influence on local soil characteristics (Gerrard, 1992; Piégay et al., 2003).
The hydrologic connectivity links floodplains and rivers into integrated ecosystems. Particulate and dissolved matter exchanged between those two systems via both surface flow and groundwater flow (Tockner et al., 1999). The concept-flood pulse is developed and flood pulses are regarded as the principle drive force for the existence of the system (Junk, 1989).
The hydrologic connections have significant influence on the biotic communities and ecosystem process on both river and floodplain ecosystems (Bayley, 1995; Thomaz et al., 2007). Large floodplains have an important role in the hydrologic cycle of watershed. In the flooding period, the water storage function of the floodplain could modify the river water and sediment transport (Frappart et al., 2005).
Hyporheic zone (HZ)
In the river-floodplain system, except flooding, the lateral flow also linked via beneath surface water and groundwater exchange also (Boulton et al., 2010). In recent decades, numerous studies indicated the importance of interactions between groundwater and surface water as they represent a substantial control on exchange of water, nutrients and organic matter in the connection area (Sophocleous, 2002). One of the most promising linkage concepts has been the development of what is known as the hyporheic zone (HZ) (Figure 4). It was first presented by Orghidan (1959) as a special underground ecosystem, but numerous different definitions by ecologists, hydrologists and biogeochemists have since been proposed. The HZ is viewed as a special benthic dynamic ecotone by ecologists and the zone of saturated sediment beneath and lateral to a stream or river channel that receive surface water input by hydrologists (Storey et al., 2003; Sophocleous, 2002; Hancock et al., 2005). The most important character of HZ is the mixture of surface water and subsurface water. Depends on the local hydrologic conditions, the HZs could be categorized into three broad typologies: 1) groundwater-dominated, 2) surface water-dominated and 3) sites exhibiting transient water table features (Malcolm et al., 2005).
The size and the flux of surface water groundwater (SW-GW) exchanges of the HZs vary both spatially and temporally according to the local conditions (Cardenas et al., 2004). The factors that regulatory SW-GW exchange in the hyporheic zone are different at different spatial scales. The key factors controlling HZ water exchange at riffle scale are found to be the hydraulic conductivity, the hydraulic gradient between upstream and downstream ends of the riffle and the flux of groundwater (Storey et al., 2003). Cardenas et al. (2004) found that streambed heterogeneity, stream curvature and bed form dynamically determined HZ geometry, fluxes, and residence time distributions. At the reach scale, the main factors are found to be the channel bed form, sediment permeability and particle size (Boano et al., 2007; Cardenas and Wilson, 2007a, 2007b). At the catchment scale, valley widths, depths of bedrock and aquifer properties are proved have influence on the HZ water flux (Bardini, 2013; Brunke and Gonser, 1997a; Malcolm et al., 2005).
The solute load capacity of HZs is highly dynamic, in the small channels, stream water is often completely exchanged with water storage in hyporheic zone within several kilometres (Jones and Mulholland, 1999). However, the ratio of HZ exchange flow compared with stream water discharge decreased as stream size increased (Wondzell, 2011).
Riparian zones are known as the buffer zones that located between the terrestrial and the aquatic ecosystems (Gregory et al., 1991) (Figure 5). As ecotones, the ecosystem services values of riparian zones had been noticed for a long time. The relative function of riparian zones are different and depends on the size of the stream, the position of the stream within the drainage network, the hydrologic regime and the local geomorphology (Naiman and Decamps, 1997). Riparian zone as an efficient BMP (Best Management Practice) had been used all over the world (Ice, 2004; Lee et al., 2004; Matteo et al., 2006).
The environmental services of riparian zone main fall into three major categories:
1) Hydrology and sediment storage:
The high hydraulic roughness of riparian vegetation contribute to the decrease of both runoff from upland to river during no-flood period and the kinetic energy of flood from river to upland during flood period (Tabacchi et al., 2000). Since water flow through riparian zone before enters the channel, riparian zone is also known the area to trap sediment erosion from the agricultural land. The buffer width and slope of the riparian zone were proved have the most important factors on sediment trap (Cooper et al., 1987; Liu et al., 2008). The storage of water and sediment in the riparian zone could reduce damage from flood, stabilize river bank and reduce channel erosion also.
2) Biogeochemistry and nutrient cycling intercept, cycle and accumulate
The locations and the hydrologic conditions of riparian zones ensure their high biogeochemical activities. The saturate soil caused by inundation with river water and shallow confining layer make riparian zones are hot spots of anaerobic processes like denitrification (Hoffmann et al., 2009). The phosphorus attached on the sediment is trapped together with the sediment, however, the trapped phosphorus also could be release and enrich runoff waters in available soluble phosphorus (Surridge et al., 2007). Riparian vegetation could remove nutrients dissolved in the water and accumulate in the plants, lead to a short term accumulation of nutrients in non-woody tissues and a long term accumulation in woody tissues (Groffman et al., 1992; Peterjohn and Correll, 1984). At the same time, litter of riparian vegetation is source of organic matters to aquatic organisms (Jardine et al., 2011).
3) Biodiversity maintenance
As ecotones, riparian zone could increase regional specie richness (Sabo et al., 2005). Variations of flood duration and frequency and the succession of riparian vegetation created complex and shifting habitats to support high biodiversity in the riparian system (Malanson, 1993).
Except the hydrologic impact, the connection in the floodplain-river systems also has great influence on the biogeochemistry process. Depend on the hydrograph and floodplain topography, the natural floodplains serve as sinks, sources, or transformers of dissolved and particulate organic matter, inorganic nutrients, and biota (Tockner et al., 1999).
The sedimentation and the buffer function of riparian zone in floodplain could lead to sediment and nutrients retention in floodplain. Noe and Hupp (2005) compared sediment and C, N, and P accumulation rates in floodplains with different watershed land use and hydrogeomorphology conditions, it is found that watershed land use types have significant impact on sediment and nutrient retention in floodplains. The hydrologic disconnections between river channels and floodplains minimize material retention by floodplains.
Floodplains usually are large pools of POC exist as litter or coarse woody debris (Robertson et al., 1999). The occurrence of flooding is proved increasing the turnover rates of organic matter and nutrients in the river-floodplain system (Bayley, 1995). The flooded water was found led to increasing nutrient mobilization (Banach et al., 2009). Flood, especially the duration of flood was proved can increase leaf decomposition rate in the floodplain (Langhans and Tockner, 2006).
Like the two pole water interaction, while sediment, nutrients and biota are transported from river channel to floodplain during flooding, the released DOC and nutrients during flooding and the POC storage in the floodplain tend to transported back into the river channel during flood recession. The study in the floodplain of Amazon river by Moreira-Turcq et al. (2013) illustrated that floodplains are important sources of organic carbon of the river main channels. Organic matter in the river is imported to floodplain during rising water period and the OM (organic materials) produced in the floodplain is exported to the river during high and falling water periods.
Baldwin and Mitchell (2000) illustrated the effects of the wetting-drying regime on nutrient cycles in floodplain system, four wetting-drying regimes were studied: partial drying sediment, complete desiccation of sediments, rewetting of desiccated soils and sediments and inundation of floodplain soils. It was proved that they all have different impact on nutrient cycling in floodplain.
Denitrification in riparian soil
Floodplains support intensive agricultural activities, in Europe and North America, up to 90% of floodplains are cultivated (Tockner and Stanford, 2002). Floodplains are important source of nutrients like N, P, pesticides and sediment (Bainbridge et al., 2009). Since recharged groundwater in cultivated fields is a major source of the nitrate contamination of surface water, the nitrate level in groundwater can have a major influence on the quality of surface water (Cey et al., 1999).
Table of contents :
1. Introduction générale
1. General introduction
Chapter 2. Nitrogen cycling, river-floodplain system, modelling approach.
2.1.2 Nitrate pollution in the aquatic ecosystems
2.2 River-Floodplain system
2.2.1 Hyporheic zone (HZ)
2.2.2 Riparian zone
2.2.3 Biogeochemistry cycling
2.3 Modelling approach of river-floodplain system
2.3.1 Models of river-floodplain system
2.3.2 Catchment scale model and denitrification
Chapter 3. Material and methods.
3.1 Initial SWAT model
3.1.1 Hydrology in SWAT model
22.214.171.124 The hydrologic processes in HRU
126.96.36.199 The hydrologic processes in the channel
3.1.2 Nitrogen cycling in SWAT model
188.8.131.52 The nitrogen cycling processes in HRU
3.1.3 Organic Carbon in SWAT model
3.1.4 Denitrification in SWAT model
3.2 From SWAT to SWAT-LUD
3.2.1 SWAT-LU model
3.2.2 SWAT-LUD model
3.3 Study site
3.3.1 The Garonne River
3.3.2 Floodplain area in the middle section of the Garonne River
Chapter 4. Improved simulation of river water and groundwater exchange in an alluvial plain using the SWAT model.
4.2.1 SWAT model
4.2.2 Model development
4.2.3 Study area
4.2.4 Landscape Unit parameters
4.2.5 Calibration and validation
4.3.1 Calibrated parameters
4.3.2 Groundwater levels
4.3.3 Water exchange between surface water and groundwater
Chapter 5. Assessment of the denitrification process in alluvial wetlands at floodplain scale using SWAT model.
5.2.1 SWAT-LUD model
5.2.2 Nitrogen cycle in the SWAT-LUD model
5.2.3 Organic carbon in SWAT-LUD
5.2.4 Study site
5.2.5 LU parameters
5.2.6 Calibration and evaluation of the SWAT-LUD model
5.3.1 Calibrated parameters
5.3.2 STICS and SWAT-LUD model comparison
5.3.3 Nitrate concentrations in shallow aquifer
5.3.4 DOC concentrations in the shallow aquifer
5.3.5 Denitrification rate
5.3.6 The influence of hydraulic conditions and nitrate content on denitrification
5.4.2 Influence of hydraulic conditions on denitrification
5.4.3 Nitrate dynamics
5.4.4 Influence of POC and DOC on denitrification
Chapter 6. Assessment of the surface water – groundwater exchange and shallow aquifer denitrification in the floodplain area using the SWAT-LUD model.
6.2.2 Definition of LUs and distribution of HRUs
6.2.3 Study site
6.2.4 Subbasins and LUs parameters
6.2.5 Calibration and evaluation
6.3.1 Calibrated parameters
6.3.2 Surface water – groundwater exchange
6.3.4 The influence of hydraulic conditions and nitrate content on denitrification
6.3.5 Channel nitrate balance
6.6 Assessment of surface water-groundwater exchanges in alluvial floodplain at the catchment scale using SWAT model
6.6.3 Main results and discussion
7. Conclusions and perspectives