Agronomic role and environmental concern

Get Complete Project Material File(s) Now! »

Soil organic matter in croplands

Agronomic role and environmental concern

Soil Organic Matter (SOM) is the largest reservoir of terrestrial carbon (C), the amount of which is estimated between 1500 to up to 2400 Gt (Batjes, 1996; Ciais et al., 2013). It can behave as a sink or a source of CO2 (Arrouays et al., 2001) depending on climate, land use and soil management (Malhi et al., 1999; Bolin and Sukumar, 2000; Guo and Gifford, 2002) and has a great potential to sequester C and offset CO2 emissions (Lal et al., 2007; Ciais et al., 2013). Presently, the contemporary rise of the atmospheric CO2 concentration and its consequence on global warming is commonly admitted (IPCC, 2013): 191 countries have ratified the Kyoto protocol and are thus committed to abate their GHG emissions, including CO2, to fight against this global threat. SOM in arable soils and its management, especially in croplands, can play an important role for this purpose. Arable soils cover 1708 million hectares in the world (Fischer et al., 2000) and represent more than 15 million hectares in France (Agreste, 2012). One fourth to one third of the world annual increase of CO2 emissions could be offset by means of those agricultural practices that favour SOC increase (Lal, 2004). More, the use of fossil fuel annually causes 8.9 Gt of C emissions (Le Quéré et al., 2014), which represent about 4‰ of the world soil C stocks. In France, an international research program called “4 per 1000” was launched in 2015 to develop agronomical research following the purpose of an annual C storage in cropped soils.
For croplands, carbon storage into soil has also many other benefits, such as improvement of soil chemical, physical and biological properties (Kundu et al., 2007) leading to enhanced agricultural productivity (Lal et al., 2004) Indeed, an optimum level of Soil Organic Carbon (SOC) content is needed to hold water and nutrients, decrease risks of erosion and degradation, improve soil structure and tilth, and provide energy to soil microorganisms (Lal et al., 2004). Low level of SOC for croplands has become a European threat for soil sustainability (Ciais et al., 2010; European Commission, 2006a), a SOC content of 2% being considered critical for soil in temperate regions (Loveland and Webb, 2003). A European directive aiming at protecting soils and their SOM was even projected (European Commission, 2006b).

Dynamics of organic matter in soils

Organic matter (OM) in cropped soils comes from organic inputs such as crop residues including roots or rhizodeposition or exogenous organic matter applications (see §2.1). These organic inputs are decomposed and used by soil microorganisms, transformed into stabilised organic material on one side and mineralised into CO2 on the other side. The SOM composition and dynamics of evolution depend on many factors such as OM inputs characteristics, soil characteristics, climatic conditions, farming practices. Over time, SOM quantity tends toward some equilibrium determined by rates of inputs and decomposition. A maximum level of SOC seems to exist based on clay and fine silt contents in a soil as used for the concept of SOC saturation (Angers et al., 2011; Six et al., 2002).
Decomposition rate of SOM depends on its quantity: the more SOM content, the quicker the decomposition. Very often the decompositions of organic pools in soil are described by first-order kinetics which rates of degradation depend on their biochemical characteristics (see §1.3). Indeed, chemical and biochemical composition of OM inputs are important drivers of decomposition. Biochemical composition, for example determined by sequential fractionation based on Van Soest’s method (Van Soest, 1963; Van Soest et al., 1991; Van Soest and Wine, 1967), describe the composition of an organic input in several fractions: water soluble, neutral detergent soluble, hemi-cellulose-like, cellulose-like, lignin-like components. Such biochemical characterisation has been used for OM inputs rather than for SOM directly. Van Soest’s fractions are generally associated with decreasing degradability, and the OM proportion of each component determines the degradability of the whole OM (Morvan et al., 2006). For modelling, Van Soest’s fractions are associated with decreasing decomposition rates (Nicolardot et al., 1994; Corbeels et al., 1999), The OM and its degradability in soil can also be characterised through its organic C/N ratio (Nicolardot et al., 2001; Giacomini et al., 2007).
As for OM inputs, SOM can be chemically or physically fractionated, distinguishing homogeneous fractions that can be characterised separately, the proportion of the fractions and their specific properties such as their degradability characterising the whole SOM. Chemical composition can be determined by separation of OM into humic acids, fulvic acids and humins (Stevenson, 1994; Piccolo, 1996), or by separation of non-hydrolysable SOM fraction with acid hydrolysis (Paul et al., 1997; Augris et al., 1998). The OM stability in soils also depends on its association with soil constituents through aggregation and physical protection and interactions with minerals (Chenu and Plante, 2006). The particle size and density fractionation methods consist in dispersing soil more or less strongly and break the aggregates and separate free OM, physically protected OM and OM protected physico-chemically.
Physical parameters such as soil humidity and temperature also drive SOC decomposition. These effects can be described by a number of mathematical functions (Rodrigo et al., 1997). An increase in soil temperature enhances SOC decomposition rate. Temperature effect was first described by an exponential function as by Balesdent and Recous (1997). However, this trend seemed to be valid until a specified temperature (25°C) only and (Valé, 2006) demonstrated that logistical functions were preferable, showing that from a certain level, a temperature increase leads to decrease SOC decomposition. As well as for temperature, soil moisture effect was first described within a too narrow interval of soil moisture and was thus described by a linear or logarithmic increasing function. It has been further described with a logarithmic function in which decomposition rate was maximum at an optimum moisture value that varied from a soil to another (Rodrigo et al., 1997; Leirós et al., 1999).
Soil texture and mineralogy composition are other factors influencing OM decomposition rate: the more clay and calcareous content, the lower this rate. Functions were found by Delphin et al. (1986) and Chaussod et al. (1986) and later validated by Valé (2006) and Saffih-Hdadi and Mary (2008). Soil pH may also have effects on SOC decomposition. A decrease in pH (from 8.3 to 4.5) results in lowering global microbial activity and hence SOC mineralisation (Rousk et al., 2009). Crop production system also influences SOM dynamics: (i) the grown crops and their yields control biomass production and therefore organic inputs into soil either through root exudates (Rasse et al., 2005) or though shoot and leaf residues left on soils (Abiven et al., 2005; Redin et al., 2014) , (ii) cultivation practices determine shoot and/or leaf residues exportation and soil management. Among farming practices likely to enhance carbon storage in soil, we can mention the following: fertilisation compared to no fertilisation (Alvarez, 2005), more intensive cropping (> 1 crop/year; Young et al., 2009), no-till or reduced till practices compared to tillage (Angers and Eriksen-Hamel, 2008; Luo et al., 2010), residues incorporation compared to exportation (Dersch and Böhm, 2001; Wilhelm et al., 2004) and Exogenous Organic Matter (EOM) applications (Valé, 2006; Diacono and Montemurro, 2010). The latter, in combination with crop type and crop rotation, is decisive for SOM dynamics (Valé, 2006).
SOM is composed of about 58% of SOC (Nelson and Sommers, 1982), the rest being essentially organic N, the quantity of which being defined by the soil organic C/N ratio. Soil C/N ratios generally fall in the range of 9-13 (Kirkby et al., 2011). Indeed, SOM allows storage and destocking through mineralisation of many nutrients such as N that can be then used by plants and soil microorganisms (Rasool et al., 2008). Depending on the rate of OM degradation and its N contents, N is mineralised and becomes available for plant uptake or other N losses in the environment.

Modelling carbon and nitrogen cycle from SOM turn over

One of the oldest models describing C dynamics is the model proposed by Hénin and Dupuis as early as 1945 (Hénin and Dupuis, 1945), (Fig. 1). It is very simple and requires few parameters only, the SOM content following equation [1].
where y: the SOM content (t.ha-1) : the annual C input to soil (t.ha-1.yr-1)
K1: the fraction of that enters into SOM, also called iso-humic coefficient, the remaining being assumed to be mineralised instantaneously, K2 (yr-1) is the decay rate of SOM.
The K2 parameter varies according to soil and climatic conditions (Le villio et al., 2001).
From the simple Hénin-Dupuis model, a number of other models have been built to simulate organic matter turnover. Some of them remain simple, focused on carbon, with a limited number of OM compartments (or model (Andriulo et al., 1999) relies on the Hénin-Dupuis model but the soil OM compartment is subdivided into an active organic carbon and a stable organic carbon pool, the latter having a limited role in C dynamics as its size remains constant over time. Other models as Roth-C (Coleman and Jenkinson, 1996; Jenkinson et al., 1990, 1987), still focused on C, are more complex and composed with a higher number of pools and therefore parameters. They may be a priori capable of better accounting for OM dynamic complexity in soil.
SOM models have then allowed to predict N dynamics by coupling N to C dynamics. Among them, NCSOIL (Hadas and Molina, 1993; Molina et al., 1983; Nicolardot et al., 1994; Molina, 1996) or CANTIS (Garnier et al., 2003) had first been designed for simulating OM dynamics of laboratory measurements, and SOMM (Chertov, 1990; Chertov and Komarov, 1997) for SOM dynamics of forest soil. These first categories of models generally consider effect of climate, soil or cropping systems with simple functions to adapt decomposition rates to various contexts. They do not integrate a plant model but predictions of SOM dynamics over cultivated soils is allowed by manually setting organic inputs.
To better take into account of varying factors influencing organic inputs and SOM decomposition, models simulating SOM dynamics are coupled with deterministic crop models. An AMG-equivalent model (Nicolardot et al., 2001) is incorporated in the crop model STICS (Brisson et al., 2009), NCSOIL in CERES (Gabrielle et al., 2004), CANTIS in PASTIS (Garnier et al., 2003). These crop models simulate processes in the Soil, Water, Atmosphere and Plant compartments also called SWAP models. They are quite similar in their crop growth and development modelling however they may differ in the types of crop they can model. They simulate fluxes more or less precisely and especially water and solutes fluxes. (Addiscott and Wagenet, 1985) distinguished two types of models of water fluxes in soil: (i) mechanistic flux-gradient models and (ii) functional reservoir or capacity based model. DAISY (Hansen et al., 1991; Mueller et al., 1996) and PASTIS are directly based on Richard’s equation (i), while STICS and CERES-EGC use semi-empirical Darcy’s law i.e. they mixed water tipping budget modelling (ii) and fluxes based on Darcy’s law (i). The first modelling (i) allows a fine description of water and solutes fluxes however it requires many parameters to be set and time-consuming calculations. The second is usually sufficient for simulating water and N fluxes between soil horizons in order to simulate crop growth and major N transformation correctly and its parameterisation is more robust than for the first types of models (Gabrielle et al., 1995). Eventually, such models differ in modelling OM dynamics.
Smith et al., (1997) also tested some of these models along with long-term experiments: CANDY (CArbon-Nitrogen-Dynamics; Franko et al., 1995); DNDC (DeNitrification and DeComposition; Li et al., 1992, 1997); DAISY; CENTURY (Parton et al., 1987; Parton, 1996). These models were tested and generally performed better for the ecosystem they were designed for. The SOMM model was better adapted for forest and grasslands, while CENTURY and DNDC performed better for grasslands. CANDY, DAISY seemed to perform well on all datasets tested. Gabrielle et al. (2002) compared the ability of STICS and NCSOIL to correctly simulate long-term SOC evolution and short-term soil mineral N dynamics in crop fields and they highlighted a trade-off between the two predictions quality. While NCSOIL well simulated long-term SOC evolution, STICS rather performed well for short term simulations of N cycle.
Models also differed on the temporal scale they are designed for. For instance, Roth-C and CENTURY simulate SOM dynamics on a monthly time step being able to predict SOM evolution over centuries, as SOMM. The other models presented here generally simulate on a daily time step being better adapted to predict SOM evolution over months. NCSOIL first devised for short term N and C dynamics can simulate on hourly time step, but it performed well for simulating SOM dynamics in a 4-y field experiment (Gabrielle et al., 2005).

READ  International Sustainability Reporting Guidelines


The CERES-EGC model

The CERES-EGC model (Gabrielle et al. , 1995, 2002, 2006, 2005) was adapted from CERES (Jones et al., 1986), a mechanistic model, simulating dynamics of water, carbon and nitrogen in soil-crop systems. CERES-EGC runs on a daily time step and is composed of three main modules. First, a “physical” module for heat, water and solute transfers in soil, as well as plant water uptake and evapo-transpiration governed by climatic demand. Second, a module addresses SOM dynamics in the ploughing layer including microbiological mineralisation and immobilisation being assured by the coupling with the NCSOIL model. Third, a module addresses crop growth through crop net photosynthesis.

The NCSOIL model

NCSOIL simulates plant residues and EOM decomposition after incorporation into soil, nitrification, immobilisation and mineralisation of nitrogen, along with SOM decomposition and formation. It considers four endogenous soil compartments or pools, resulting in 6 pools involved in OM turnover: (i) pool 0, the “zymogenous biomass”, (ii) pool I, the “microbial biomass”, (iii) pool II, the “mineralisable SOM”, (iv) pool III, composed of highly-humified OM, (v) pool EOM, (vi) pool plant residues. EOM and soil pools (except pool III) are divided into 2 “fractions”, one being only decomposed slowly, referred to as “resistant”, and the second which is labile, i.e. readily decomposed. Following the same scheme plant residues are divided into 3 fractions, carbohydrates, cellulose and lignin. Each pool fraction is characterised by its C content (mg, a decomposition rate (day-1), an organic C to N ratio (C/N). C transformation rates follow a first-order kinetics and the rate of N flows follow the C flows respecting the C/N ratio of a given pool being decomposed, mineral N being added or withdrawn by the following pool to maintain its C/N ratio.

Exogenous organic matter use in agriculture

EOM definition and sources of production

EOM are organic materials, i.e. from degraded vegetal or animal constituents, resulting from agriculture, agro-industries or city activities, The EOM definition generally includes crop residues because of agronomic use and characterisation methods similarities (Marmo et al., 2004), however, here we clearly distinguish EOM and crop residues. EOM can be used in agriculture through applications on cropped soils either directly or after treatments. Various treatments exist: (i) biological treatments, being either composting or anaerobic digestion and (ii) physical or chemical treatments as drying or liming applied to sludges, digestates, slurries or poultry litters.
Composting is an aerobic decomposition of OM into a more stable OM, a process composed of a heating fermentation phase then a maturation phase. Composting results in a loss of weight of about 50% due to water evaporation and organic matter mineralisation, and to a modification of the chemical and biochemical composition of OM. The heating reaching up to 70°C allows pathogens, pests and diseases destruction. The compost obtained is characterised by a higher homogeneity than the raw material, and higher stability. Digesting is an anaerobic decomposition of OM that emits biogas (CO2 and usable CH4). The process can be in a liquid or dry way, at various temperatures and the residue named “digestate” can be applied either directly or after post-treatments such as “phase separation” and/or “composting” (Peltre, 2010).
We can classify the different EOM into 3 main types (Dhaouadi, 2014):
EOM from agriculture, mainly livestock effluents, with manures (mix of effluents and litter, being usually straw), slurries (liquid effluents), poultry litters;
EOM from cities, with two main sub-types: sewage sludge and waste from cities and household. Urban wastes are often composted and various composts can be found according to the type of composted waste and/or waste sorting: green waste compost, co-compost of sludge and greenwastes, biowaste compost obtained from the fermentescible fraction of household waste and often mixed with green waste, and municipal solid waste from household waste after the removing of packaging.
EOM from industries, such as agro-industries or chemical industries, sludges issued of industrial waste water treatment plants as for example paper-mill sludges.
In France, EOM produced by livestock farming amount to 297.8 Mt.yr-1 (Biomasse Normandie, 2002), from which about 50% is not recyclable because remaining in pastures where animals are breeded. The recyclable part is almost totally reused in agriculture with 109 Mt.yr-1 applied in 2012 (Agreste, enquêtes de pratiques culturales 2011; FranceAgriMer, 2012). In comparison to agricultural effluents, less EOM from industries and cities which represent lower quantities are reused. Indeed, industries generate an organic waste production of 15,600 Mt.yr-1, urban sewage sludge represents a production of 1.2 Mt Dry Matter (DM).yr-1 and 38 Mt.yr-1 for municipal solid waste and they are reused in agriculture in the following proportions: 63%, 75 % and 14% respectively (ADEME, 2012, 2009; Biomasse Normandie, 2002). In 2008, 1.8 Mt DM of organic industrial wastes were applied on fields while 0.3 Mt DM were sent for composting or anaerobic digestion. In 2011, among the 73% of sewage sludge from cities applied on fields, 31% was composted. In 2011, about 40 Mt Fresh Matter (FM) of household waste was used to make 2.2 Mt FM of compost and digestates, 80% of which were applied on field in 2006 (Houot et al., 2014).

Regulatory framework of EOM use: the French case

EOM and their uses are regulated by superposition of regulations summarised in Table 1 depending on:
Their status as “waste”, “by-product” or “product”; their source either from establisment classified for an environmental protection in accordance with French Law No 76-663 of 18 July 1976 (“Installation Classée pour la Protection de l’Environnement” (ICPE)) or from whatever other origin;
the location of field application: either in a nitrate-vulnerable zone (“Zone Vulnérable, ZV”), in a nitrate-vulnerable zone with additional remedial measures(“Zone à Action Renforcée, ZAR”) under the nitrates Directive of December 1991 (91/676/EEC) or outside such zones.

EOM legal qualification as « waste», « by-product» ou « product»

« Wastes »

A waste is first defined in article L. 541-1-1 of the Environment Code (“Code de l’environnement”). A non-exhaustiv waste list is mentioned in annex II article R. 541-8 of the Environment Code. Some of the organic wastes are regulated by specific dipositions: for instance, straws and agricultural effluents are not qualified as “waste” anymore if used in the farm where they were produced (L. 541-4-1). Sewage sludge are considered as “waste” (R. 211-27) but are also regulated by a supplementary directive (n° 86/278/CEE du 12 juin 1986) relative to the environment preservation and especially soil for sewage sludge use by agriculture. The directive determines maximum levels of heavy metals in soil and sludge. At a national level, sewage sludge regulations result from the law N°92-3 of January 3rd 1992 concerning water, decree N°97-1133 of december 8th 1997 (for the waste qualification principle) and the ministerial order of January 8th 1998 determining technical prescription for sludge application on agricultural soils.
Waste can stop being “waste” (Enckell, 2013) if recycled or treated, in a classified establishment (ICPE) or an establishment regulated by the water law, in order to be reused, as set out in the article L. 541-4-3 in the Environment Code, deriving from the directive transposition 2000/98/CE.

Table of contents :

Chapter 1. Literature Review
1. Soil organic matter in croplands
1.1. Agronomic role and environmental concern
1.2. Dynamics of organic matter in soils
1.3. Modelling carbon and nitrogen cycle from SOM turn over
1.4. The CERES-EGC/NCSOIL model
1.4.1. The CERES-EGC model
1.4.2. The NCSOIL model
2. Exogenous organic matter use in agriculture
2.1. EOM definition and sources of production
2.2. Regulatory framework of EOM use: the French case
2.2.1. EOM legal qualification as « waste», « by-product» ou « product»
2.2.2. EOM legal qualification
2.2.3. EOM use
2.3. EOM agronomic interests and environmental impact
2.3.1. EOM capacity to increase SOM, amending potential
2.3.2. EOM capacity to substitute mineral fertilisers
2.3.3. Potential environmental impacts
2.3.4. Modelling and parameterising EOM
3. Spatial distribution and temporal evolution of soil organic matter at the regional scale
3.1. Mapping soil organic matter and parameters driving its fate
3.1.1. Mapping methodology
3.1.2. Mapping issues
3.1.3. Mapping methods used
3.2. Spatially-explicit EOM impacts on C and N for regional concerns
3.3. Regional EOM management
4. Conclusion and objectives
5. Supplementary data
6. References
Présentation du territoire d’étude
Chapter 2. Modelling the long-term effect of urban waste compost applications on carbon and nitrogen dynamics in temperate cropland
1. Abstract
2. Introduction
3. Materials and methods
3.1. The QualiAgro long-term experiment
3.2. EOM characterisation
3.2.1. Physico-chemical analysis and biochemical fractionation
3.2.2. C and N mineralisation under laboratory controlled conditions
3.3. Model description
3.3.1. CERES-EGC model
3.3.2. NCSOIL model
3.4. Model parameterisation
3.4.1. CERES-EGC parameterisation for the QualiAgro experiment
3.4.2. NCSOIL parameterisation
3.4.3. Evaluation of model goodness of fitting
4. Results
4.1. Initial parameterisation of Pool II
4.1.1. Estimation of initial Pool II size with CERES-EGC
4.1.2. Estimation of Pool II C/N ratio with NCSOIL
4.2. EOM parameterisation based on lab-scale results and simulations with NCSOIL
4.2.1. Results obtained from the NLR procedure
4.2.2. Variation of C and N mineralisation with the EOM origin
4.2.3. Goodness of fitting of simulated EOM kinetics of mineralisation
4.3. Field-scale CERES-EGC simulations over a 13 y-time series
4.3.1. Soil organic matter
4.3.2. Crop yield and N uptake
4.3.3. Soil mineral N
4.3.4. N leaching
4.3.5. N fluxes and balances
5. Discussion
5.1. Relevancy of NCSOIL parameterisation methods
5.2. Scale-transfer between lab NCSOIL parameterisation and CERES-EGC long-term field simulations
5.3. Impacts of EOM applications
6. Conclusion
7. Supplementary data
8. References
Chapter 3. Parameterisation of the NCSOIL model to simulate C and N short-term mineralisation of exogenous organic matters in different soils.
1. Abstract
2. Introduction
3. Materials and methods
3.1. Soils and EOM
3.1.1. Soils
3.1.2. Exogenous Organic Matters
3.1.3. Incubation experiments
3.2. Modelling
3.2.1. The NCSOIL model
3.2.2. Model parameterisation
3.2.3. statistical evaluation of soil effect on EOM mineralisation
3.2.4. Factor analysis: relation between EOM parameters and characteristics
4. Results
4.1. Variation of C mineralisation kinetics of EOM with soil type
4.2. Classification of EOM for kEOM parameterisation
4.3. Simulation of EOM mineralisation
4.3.1. EOM mineralisation behaviours
4.3.2. Appropriateness of parameterisation methods
4.3.3. EOM characterisation requiring another parameterisation approach
4.4. Use of Van Soest fractionation for parameters and group characterisation
5. Discussion
5.1. Parameterisation methods
5.2. General scheme for relating model parameters to measured values
5.3. EOM types available in the “Plain of Versailles” region
6. Conclusion
7. Supplementary data
8. References
Chapter 4. Regional-scale scenarios of organic amendment use on cropped soils: impacts on soil organic carbon stocks and substitution of mineral fertilisation simulated with the CERES-EGC crop model
1. Abstract
2. Introduction
3. Materials and methods
3.1. Study Area
3.2. Set of geographical layers included in the modelling
3.3. Characterisation of regional EOM
3.4. The CERES-EGC model: overall presentation, parameterisation and running
3.4.1. The CERES-EGC model
3.4.2. The NCSOIL model
3.4.3. Model parameterisation
3.5. Building scenarios of EOM application for each cropping system
3.5.1. Crop residues and intercrops
3.5.2. EOM applications
3.5.3. Crop succession constraints for EOM application
3.6. Insertion of mineral N saving in the modelling
3.6.1. Principle
3.6.2. Some additional rules
3.7. Running the model
3.8. Analysis of the factors driving carbon storage, nitrogen saving and nitrate leaching
4. Results
4.1. EOM characteristics and consequences on their insertion in scenarios
4.2. C storage
4.2.1. C storage drivers
4.2.2. Differences in C storage
4.3. N savings
4.3.1. N savings drivers
4.3.2. Differences in N savings
4.4. Nitrate leaching
4.4.1. Nitrate leaching drivers
4.4.2. Differences in N Leaching
5. Discussion
5.1. Simulations of C and N cycles
5.2. Nitrogen Savings and other impacts
5.3. Uncertainties and how to reduce them
5.4. Feasibility of EOM use scenarios at crop plot and regional scale
6. Conclusion
7. Supplementary data
8. References
Chapter 5. Optimisation of Exogenous Organic Matter use by agriculture at the regional scale: case study of a peri-urban area near Paris, France
1. Abstract
2. Introduction
3. Material & methods
3.1. Study area
3.2. Spatially-explicit simulations of EOM applications
3.2.1. Spatially-explicit data for model parameterisation
3.2.2. Scenario simulations of EOM use
3.2.3. Simulations with the CERES-EGC crop model
3.3. Optimisation of EOM distribution
3.3.1. Preparation of optimisation entries
3.3.2. Optimisation model
3.3.3. EOM availability situations
3.3.4. Optimisation results analysis
4. Results
4.1. EOM availability situations
4.2. Potential impacts related to EOM use at the regional scale
4.2.1. Optimised distributions of current EOM (situation A)
4.2.2. Impact of composting development (situation B)
4.2.3. Impacts on NH3, N2O and N2
4.3. EOM distributions
4.3.1. EOM distribution per soil types
4.3.2. EOM distribution per crop succession
4.3.3. EOM application concentration
4.3.4. Spatial patterns of EOM distribution
5. Discussion
5.1. Regional strategies for EOM management
5.2. Regional potential benefits and risks
5.3. Method reproducibility
5.4. Multi-criteria analysis
5.5. Applicability
6. Conclusion
7. Supplementary data.
8. References
Discussion générale et conclusion


Related Posts