General situation of phosphorus resource in the world
Phosphorus (P), an essential element for plants, animals, and humans, plays a central role in soil fertilization and food security around the world. The P represents 0.12% of Earth crust composition (ranking the 11th). The lithosphere is thus the main source for the biosphere. Phosphorus cycles among lithosphere (bedrocks, sediments), pedosphere (soils), hydrosphere (river, lake and ocean), atmosphere and living matters (Fig. 2-1). However, the P cycle in biosphere is misbalanced because P tends to accumulate at a rate of 1 to1.55×10-7 t P per year in soils (Pédro 2012), mainly due to the fertilization to sustain crop yield.
The main source of P for mineral P fertilizers is phosphate rock (PR) since the mid-to-late 19th century (Fig. 2-2). According to the International Fertilizer Industry Association, almost 53.5 M t of P2O5, viz. 175M t of PR averaging 30.7% P2O5, has been mined in 2008. In 2011, agriculture accounted for 89% of the PR consumed worldwide (55.96 Mt of P2O5 from 191 Mt PR), with 82% as fertilizers and 7% as feed additives (Ulrich 2013). It is widely agreed that over the last decades the use of P increased consistently (Van Vuuren et al. 2010) with the explosive growth of the world population. According to the estimation of FAO (Food and Agriculture Organization of the United Nations) in 2015, the use of P fertilizer could still be soaring during 2014-2018 period with growth rate from 0.1% to 6.3% in different regions of the world (Table 2-1) (FAO: World fertilizer trends and outlook to 2018). There is no alternative to the limited PR resources (Weikard and Seyhan 2009). The PR reserves are concentrated in a few countries such as Morocco, China and the United States (Dawson and Hilton 2011). Mineral P fertilizers could be partially replaced by the recycling of animal manure, crushed animal bones, human and bird excreta, city waste composts and ash (Van Vuuren et al. 2010) with a share of about 19.2 Mt P year-1 (Cordell et al. 2009). Nontheless, it is widely accepted that the existing high-value PR ores could be exhausted by the next 50-100 years (Cordell et al. 2009).
Besides, over-fertilization or inappropriate P fertilization could both increase P losses by runoff and lixiviation, enhancing the risk of eutrophication of surface water. Sustainable P management should recycle P efficiently through improved uptake and use efficiency, and reduce P contamination.
Soil phosphorus availability to plants
Soluble P is present at low concentration in the soil solution due to strong and multiple interactions with soil constituents (Hinsinger et al. 2011). On the one hand, the majority of added P will be also highly retained. Balemi and Negisho (2012) reported that more than 80% of added mineral P could be sorbed, as found elsewhere (Holford 1988; Zhou and Li 2001; Mehdi et al. 2007; Binner et al. 2015). On the other hand, the transport of phosphate ions in soil relies on mass flow and diffusion. Low concentration and small diffusion coefficient of phosphate ions (Krom and Berner 1980) in the soil solution would make it difficult for plants to acquire soil P. Hence, P is partitioned into several pools according to its mobility in soil (Fig. 2-3) (Barber 1995).
Phosphorus in soil solution pool is in the form of di- and mono- hydrogen phosphate ions (H2PO4-, HPO42-) depending on soil pH, and contributes directly to root absorption. However phosphate ions in the liquid phase could be easily drawn out into solid phase of soil constituents by physical and chemical adsorption, then be released as environmental conditions change. The dynamic process of P adsorption and desorption is affected by pH, the ionic force of soil solution and temperature. The P sorption can be could be modelled by the Langmuir isotherm equation (Hartikainen et al. 2010; van Rotterdam et al. 2012; Binner et al. 2015) (Eq 2-1) or the Freundlich isotherm equation (Idris and Ahmed 2012; Messiga et al. 2012a; Wolde and Haile 2015) (Eq 2-2) as follows: 𝑸= 𝑸𝒎𝒂𝒙𝑲𝑪1+𝑲𝑪 Eq 2-1.
Q is the amount of P bound to the soil in equilibrium with a certain concentration in solution, C is the concentration in solution, Qmax is the adsorption maximum and K is an effective soil P affinity constant (van Rotterdam et al. 2012). 𝑸=𝒂𝑪𝒏 Eq 2-2.
Q is the amount of P sorbed per unit weight of soil, C is the P concentration in the equilibrium solution, a is a constant related to sorption capacity, n is phosphate sorption energy (Idris and Ahmed 2012).
Phosphorus cycle in field
The soil phosphorus cycle illustrates the P turnover in cultivated soils (Fig. 2-4). External P added to soil is firstly retained by soil constituents via sorption and precipitation reactions. Reacted retained phosphorus will be made slowly available to plants and micro-organisms, lost by leaching, erosion and runoff, or converted to more stable forms. The P assimilated by living organisms could be recycled. At field scale (generally in the tilled soil layer in conventional agroeco-systems), the P cycle is simplified as several P input and output flows (Morel 2002). On the long run, P inputs comprise P fertilizers (mostly) and atmospheric depositions (mostly ignored). Phosphorus outputs are often P export through harvest, P leached below rooting zone and P runoff on soil surface. In addition, the transformation of P species (AlMaarofi et al. 2014) and the residual value of P application (Beck and Sanchez 1996) could be regarded as P flows considering seasonal changes in soil P stocks. To have a closer view of soil P cycle and soil P status, several factors must be addressed.
Phosphorus fertilizer application
In general, P fertilizer could be divided into chemical and natural P fertilizers. The chemical P fertilizers refer to commercial phosphate fertilizers; animal manure, crushed animal bones, human and bird excreta, city waste composts and ash are regarded as natural fertilizers (Van Vuuren et al. 2010).
The common commercial-grade granular phosphate fertilizers are OSP or SSP (ordinary-superphosphate or single-superphosphate), TSP (triple-superphosphate), MAP (mono-ammonium phosphate), and DAP (di-ammonium phosphate) (Chien et al. 2011). The manufacturing of most commercial phosphate fertilizers starts with the phosphoric acid.
Fertilizer P content is measured as total phosphorus oxide (P2O5). The P content and the fertilizer solubility (in water and in ammonium citrate or citric acid solution) determines the efficiency of P fertilizers in soil (Kratz et al. 2010). Compared to the chemical mineral P fertilizers, the natural P fertilizers, especially the organic P fertilizers, are more soluble because of the competing reactions with organic acids and of the micro-organism activities. But very little difference of fertilizer solubility effect has been reported on crop yields (Prochnow et al. 1998; Prochnow et al. 2001) as soil interactions control phosphorus uptake more than the physical form of fertilizers (Roper et al. 2004).
The P fertilization increases soil P content (Sims et al. 1998b). Most added P is fixed by soil constituents. The effective volume of P fertilizers is thus constrained in a limited zone around fertilizer. Stecker et al. (2001) observed that the zone affected by fertilizer did not exceed 5-cm radical distance from fertilizer in one year (Fig. 2-5). High P fixation also explains why the seasonal use of added P fertilizers is only about 10-25%. The performance of P fertilizers also depends on soil test P. On ryegrass grown optimally in greenhouse pot experiments using the isotope labeling method, Linères and Morel (2004) found that effective P recovery (REC) and relative contribution of applied P to plant nutrition (Pdff) values were 14.0 and 17.8% in at high Olsen-P level (52 mg kg-1) compared to 30.9% and 57.0% at low Olsen-P level (6.1 mg kg-1) (Table 2-3). At high P status, plants mainly took up P from the soil and the P fertilizer mainly remained in the soil.
Phosphorus losses by erosion, runoff and leaching
Intensive P fertilizer application often results in elevated P accumulation in soil. High soil P status increases the risk of P loss through wind and water erosion. Jørgensen (2010) mentioned the world soil erosion from agriculture areas (including cropland and pasture) amounts to 72.9 Gt yr-1 causing P losses at 19.3 and 17.2 Mt P yr-1 for cropland and pasture, respectively. There are two pathways of P transfer from agricultural soils to receiving waters: runoff and leaching (Zhang et al. 2003). Both pathways of P losses are quantified by the concentrations of various P forms (dissolved reactive P (DRP), dissolved un-reactive P (DURP), and particulate P (PP)) (Tan and Zhang 2011) and affected by a number of factors such as rainfall, field topography, soil types, soil preparation, fertilizer application, irrigation method and drainage system.
Roughly, the world total phosphate fertilizer application may lead to a loss of 0.5 Mt P yr-1 by surface runoff (Jørgensen 2010). Heavy rainfall (Simard et al. 2000) and steep field slope (Carroll et al. 2000) promote water flow which could increase soil water erosion and hence losses of PP. The concentration of DRP in runoff varies with different soil types (Wang et al. 2011). The DRP loss over a period of time is estimated based on the measurement of soil test P or degree of P saturation. The degree of P saturation (DPS) is defined as the ratio of sorbed phosphorus and phosphorus sorption capacity (number of sites available for phosphorus binding), and used as a potential phosphorus loss risk indicator (Casson et al. 2006). The P sorption capacity (PSC) of soils varies widely according to clay content, clay mineralogy, organic matter content, exchangeable aluminum, iron, and calcium concentrations, and pH. In non-calcareous soils, DSP, represented by the P saturation index (PSI), is defined as the ratio of extracted P to extracted Al and Fe. Various alternatives of PSI are presented in Beauchemin and Simard (1999).
Because of the high P-fixation capacity of many mineral soils, the vertical P transport (leaching) in relatively high clay content (>10%) soils is often assumed to be a minor contributor to surface-water P enrichment, compared to P loss by runoff and erosion (Makris et al. 2006). This assumption was challenged by Heckrath et al. (1995) and Zhang et al. (2003), pointing out that leaching is also a significant pathway for P loss in high P soils. Heckrath et al. (1995) reported that the concentration of total P in drainage water increased rapidly when soil P-Olsen exceeded a critical value (about 60 mg P kg-1). Soils saturated with P increase the risk of P losses through leaching as well as runoff. Besides, soil tillage and the drainage system influence P leaching. The NT could reduce P loss by runoff and erosion but may increase P leaching through preferential flow (Tan and Zhang 2011). Simard et al. (2000) found a higher P transfer concentration in drained soils (232 μg L-1) compared to undrained soils (152 μg L-1), because rapid flow through the drainage pathway reduced the contact time between the percolating water and the subsoil, hence opportunities for P sorption.
Phosphorus uptake by plants
The largest P output in field is crop P exportation (generally P in the grain). The P uptake occurs mainly through root absorption whereby nearly all P is absorbed in form of phosphate ions. The dynamic process for absorption of dissolved inorganic phosphate ions through root surface unite has been described by Michaelis-Menten kinetics (Barber 1995) as follows: 𝑰𝒓=𝑰𝒎𝒂𝒙(𝑪𝒍−𝑪𝒎𝒊𝒏)𝑲𝒎+𝑪𝒍−𝑪𝒎𝒊𝒏 Eq 2-5.
where Ir is P influx into the root, Imax is maximum influx, Cl is concentration of phosphate ions in the liquid phase, Cmin is minimum concentration of phosphate ions in the liquid phase for net influx of zero, and Km is concentration of phosphate ions in the liquid phase where influx is 1/2Imax. Therefore, the P absorption capacity of plant species (Imax and Km), P (phosphate ion) concentration in soil solution (Cl) and plant root biomass (e.g. total root surface) are three important factors controlling P uptake by plants.
The P absorption capacity refers to P uptake rate at root surface. It mainly depends on how fast phosphate ions can be transported through root cells to reach xylem vessels. For most mineral salts, there are two ways to cross root cells: an apoplastic route between the cells and a symplastic route within the cells. Because phosphate ion concentration in root cell is much higher than that in the soil solution, phosphate ions enter plant mainly through the symplastic route using specialized transporters at the root/soil interface to extract phosphate ions from solution at micro-molar concentrations (generally less than 10 μM P), and other mechanisms allowing phosphate ions to move across membranes between intracellular compartments, where the concentrations of phosphate ion may be 1000-fold higher than in the external solution (Schachtman et al. 1998). This process is controlled by thermodynamic parameters and requires energy. And it varies for different crop species. In a hydroponic study, Bhadoria et al. (2004) observed that corn had a 6 times higher Imax value and a 2 times higher Km value for phosphorus uptake compared to groundnut.
Crop response to applied phosphorus
The Mitscherlich function is commonly used to relate P fertilization (or soil test P) to crop yield (Morel 2002; Pukhovskiy 2013) (Eq 2-6). The model provides a critical value of soil test P to obtaion maximum crop production. The model is defined as follows: 𝜹𝒀𝜹𝒙⁄=𝒄(𝑨−𝒀) Eq 2-6.
where Y is crop yield at a given P dosage or soil test P, A is maximum crop yield, x is P fertilization or soil test P, c is a constant named efficiency factor related to crop growth rate.
The parameters of Mitscherlich function are site-specific and controlled by Liebscher’s law of optimum (the relationship depends on how close to their optima are other factors) (Harmsen 2000). Bai et al. (2013) obtained different determination coefficients at three sites (Yangling: R2=0.79–0.91, Chongqing: R2=0.87–0.93, Qiyang: R2=0.69–0.71) for the relationship between Olsen-P and relative crop yield using the Mitscherlich function. Pukhovskiy (2013) showed that the Mitscherlich function could be expressed as: 1.Mitscherlich-Spillman model; 2.Mitscherlich-Spillman modified by Boguslawski; 3. linear function with a plateau; 4. empirical quadratic model; 5. a linear approximation of the argument in logarithmic form with a plateau; 6. empirical quadratic model of in-transformed arguments; 7. empirical quadratic model of log-transformed argument. The relationship also depends on soil test P. Smethurst (2000) relating wheat yield to 8 concentrations of extracted P (Lactate, Truog, fluoride, Bray-1, Bray-2, Olsen, Colwell, and Mehlich) at 57 sites, reported different correlation coefficients. Thus, the curve fitting of Mitscherlich function should be ascertained by the proper choice of the underlying function form and soil test P.
The annual P budget is generally described as the difference between P inputs and outputs at field scale as follows (Obour et al. 2011): 𝑷 𝒃𝒖𝒅𝒈𝒆𝒕=𝑨𝒑+𝑭𝒑+𝑺𝒑−𝑼𝒑−𝑳𝒑−𝑹𝒑 Eq 2-7.
where Ap is atmospheric P deposition (always ignored), Fp is fertilizer P, Sp is soil test P, Up is P exportation in crop grains by P uptake, Lp is leached P, Rp is runoff P.
Studies showed that the P budget (the difference between P inputs and P outputs) was closely related to soil test P. In a 13-year field trial, Beck and Sanchez (1996) observed that a positive P budget increased the Hedley NaOH extractable inorganic phosphorus from 61.3 kg P ha-1 to 191.6 kg P ha-1. In addition, in a 5-year experiment at 56 sites, Linquist and Ruark (2011) found a positive but weak relationship (R2=0.27) between the average annual P budget and Olsen-P likely because change rate of Olsen-P widely strongly among soil types.
On the other hand, a strong relationship was obtained between P budgets and soil test P (by Messiga et al. (2012a) in a 7-year experiment established on a clay loam soil. The coefficients were R2=0.67 for Mehlich3-P and R2=0.70 for Olsen-P. Similar results were obtained in a luvic arenosol (Messiga et al. 2010). Such results (Fig. 2-7) indicated that the correlation between the cumulative P budget (especially positive P budget) and soil test P was linear and robust but site-specific. Several studies with similar results are listed in Table 2-4.
Table of contents :
Chapter I Phosphorus cycle in agroeco-systems and conservation agriculture: General Introduction
Chapter II Phosphorus in agro-system
2.1. General situation of phosphorus resource in the world
2.2. Phosphorus in soil
2.2.1. Phosphorus species in soil
2.2.2 Soil phosphorus availability to plants
2.2.3. Soil phosphorus test
2.3. Phosphorus cycle in field
2.3.1. Phosphorus fertilizer application
2.3.2. Phosphorus losses by erosion, runoff and leaching
2.3.3. Phosphorus uptake by plants
2.3.4. Other phosphorus flows
2.4. Phosphorus management
2.4.1. Crop response to applied phosphorus
2.4.3. Phosphorus in nutrient balance
2.4.4 Simulation models of phosphorus cycle in agroecosystems
2.5 Conservation agriculture and no-till system
2.5.1. Conservation agriculture
2.5.2. No-till and simplified tillage methods
2.5.3. Advantages and disadvantages of no-till
2.5.4. Soil properties and nutrient distribution under no-till
2.6 Hypotheses and objectives
Chapter III The Long-term (23 years) Effects of Tillage Practice and Phosphorus Fertilization on the Distribution and Morphology of Corn Root
3.4. Materials and Methods
3.4.1. Site description
3.4.2. Root sampling and analysis
3.4.3. Soil sampling and analysis
3.4.4. Crop harvest
3.4.5. Statistical analysis
3.5.1. Root mass and its distribution in soil profile
3.5.2. Root morphological traits: root surface density, root length density, average root diameter, root diameter distribution and root P content
3.5.3. Soil properties
3.5.4. Corn yield, and cumulative P budget and relation with corn roots
3.6.1. Spatial distribution of corn roots
3.6.2. Corn roots affected by tillage practice
3.6.3. Corn roots affected by P fertilization
3.6.4. Corn roots affected by the interaction of tillage practice and P fertilization
Chapter IV Soybean root traits after 24 years of different soil tillage and mineral phosphorus fertilization management
4.4. Materials and Methods
4.4.1. Site description
4.4.2. Root sampling and analysis
4.4.3. Soil sampling and analysis
4.4.4. Crop harvest
4.4.5. Statistical analysis
4.5. Results and discussion
4.5.1. Soil nutrient stratification
4.5.2. Lateral and vertical root characteristics
4.5.3. Root quantity (biomass) affected by tillage and P fertilization
4.5.4. Vertical root length proportion affected by tillage and P fertilization
4.5.5. Root morphology affected by tillage and P fertilization
4.5.6. Soybean yields
Chapter V Validation of an operational phosphorus cycling model CycP for a long-term ploughed and soybean-corn cropped agroecosystem in eastern Canada
5.4. Materials and methods
5.4.1. Field experiment
5.4.2. Laboratory determination of soil properties needed to calculate plant-available soil P
5.4.3. Validation of the CycP model
5.5. Results and Discussion
5.5.1. Dynamics of diffusive Pi at the solid-to-solution interface of soil
5.5.2. Simulations vs. field-observations of Cp
5.5.3. Simulations of P fertilization scenarios with CycP model
Chapter VI Spatial assessment of the contribution of subsoil P to crop P nutrition
6.4. Materials and Methods
6.4.1 Site description
6.4.2 Soil sampling for phosphate ion transfer kinetics
6.4.3 Soil sampling for Cp along soil profile
6.4.4 Root sampling and analysis
6.4.5 Phosphorus uptake weighting factor and proportion
6.4.6 Statistical analysis
6.5.1 Soil P status (Cp) data
6.5.2 Crop root distributions
6.5.3 Phosphate ion transfer kinetic and soil P availability estimation
6.5.4 Phosphate uptake proportions
6.6.1 P uptake effected by tillage practice
6.6.2 P uptake proportion from subsoil P stock
6.7. Summary and conclusion
Chapter VII Simplified model for phosphorus cycle in long-term no-till agroecosystems
7.4. Materials and methods
7.4.1. Model structure
7.4.2. P flux estimation
7.6. Results and discussion
7.6.1. P stock and flux
7.6.2. Soil P status evolution in NT
7.6.3. Comparison with measured data
Chapter VIII Conclusion and perspectives