Biogeoquímica del fósforo en el suelo: optimización de criterios para un uso agronómico eficiente y ambientalmente aceptable de un recurso no renovable

A relevant issue for accurate phosphorus fertilization management is to estimate phytoavailable phosphorus (P), which is defined as the amount of P present in soil that can be used by plants in successive crops until evident P deficiency symptoms appear in crop. It can be supposed that there is a starvation of this available pool when P release from solid phase does not maintain a high enough P concentration to maintain P adsorption by plants. This plant available P is normally assessed by "P availability indexes" or “soil P test” (SPT). However, the accuracy of these different methods, over a wide range of soils, may be limited and poor relationships are frequently observed between phosphorus availability indexes, such as Olsen P, and phosphorus absorption by plants. An efficient use of phosphorus in agriculture should be based on an accurate estimationof bio-available phosphorus in soils. In this thesis, focused on soils from Mediterranean environments, the Olsen P will be taken as a reference for SPT. In the first chapter of this PhD thesis, a general introduction to the problem of P in agriculture, with information regarding its relevance as non-renewable and strategic resource necessary in agriculture, was done. In this chapter, objectives of this thesis were defined, all focused on the sustainable use of P in agriculture. In the second chapter, factors affecting the accuracy of Olsen phosphorus index (Olsen-P) were studied. To this end, a P depletion experiment in 17 Mediterranean soils was carried out. Two samples differing widely in Olsen-P were used for each soil, named "low-P" and "high-P". The soil was mixed with silica sand to achieve 1 mg of P as Olsen-P per pot. Thus, the proportion of unavailable P to Olsen-P, either inorganic or organic P in the pot, was higher for "low- P" samples than for "high-P" samples. In "high-P" samples, P uptake by crop (cucumber) increased with increased affinity factor of soil phase, estimated from the P sorption curves (R2 = 0.76; P < 0.001) and with increased P adsorption capacity of soil, estimated as the concentration of Fe in oxides in soils (R2 = 0.56; P < 0.001). On the other hand, in "low-P" soil samples, total organic P and phosphatase activity in the rhizosphere contributed to explain P absorption by crops(R2 = 0.37, P < 0.01). These results revealed that both organic P and the hydrolytic activity in rhizosphere have implications in explaining P availability to plants when the ratio of organic P to Olsen P is high (low-P samples). On the other hand, the physical and chemical properties controlling the dynamics of inorganic P played an important role in the uptake of P from high-P samples. third chapter of this of this PhD Thesis was focused on the identification of soil properties affecting threshold values of Olsen P, below which a response by crop to P fertilization can be expected. There are recent evidences that, even in soils where a given SPT such as Olsen P is recommended, threshold values for SPT may range widely. This significantly affects the practical implications of SPT. Based on the soil properties affecting threshold values for Olsen P, models were proposed for the accurate estimation of threshold values in the interpretation of Olsen-P as SPT in these soils. This is basic for the identification of P responsive sites in geographical areas where this SPT is recommended. Threshold values for Olsen P and P in soil solution estimated by an extraction with 0.01 MCaCl2 varied widely between soils. Both indexes were positively correlated with each other (P <0.001). Clay content was the soil property affecting to a greater extend threshold values for Olsen P, explaining a 60 % of variation (P < 0.001). Overall, Olsen P threshold values decreased with increased buffer capacity (estimated with adsorption isotherms at 1 mg P L-1) and P sorption capacity (estimated as Fe in Fe oxides or clay). Threshold values for Olsen P can be predicted in studied soils by multiple regressions involving clay content, pH, and phosphatase activity in the rhizosphere (R2 = 0.87, P < 0.001); when only pH and clay were taken into account, it was explained 81 % of variation (P <0.001) with an average of the absolute error of 1.12. This means that a single model based on pH and clay content can accurately estimate threshold values in the studied soils. In the case of PCaCl2 threshold values can be estimated as a function of clay and the ratio of Fe in poorly crystalline to that in crystalline Fe oxides(R2 = 0.57, P <0.001). In addition, accurate predictions of PCaCl2 threshold values can be also achieved based on multiple regressions with pH and phytase hydrolysable P in some P fractions. It can be concluded that, in the group of soils studied, soil properties related with P buffering capacity of soils are the most relevant explaining threshold values for both SPT tested. However, it should be highlighted the contribution of factors related to organic P dynamics affecting threshold values. The fourth chapter of this PhD Thesis was aimed to define more precise methods to estimate total plant available P (TAP).To this end, different P extraction methods were tested: (i) Olsen P (ii) P concentration in 0.01 M CaCl2, (iii) 0.27 M Na citrate + 0.11 M NaHCO3 (CB), (iv) the two first sequential extractions in the Ruíz et al. (1997) fractionation scheme (NaOH +CB), (v) extraction with anion exchange resins (AER) in Cl- and HCO- 3 forms. Total available P (TAP) values in soils were obtained by P depletion experiment by growing plants successively in soils. TAP was significantly related to HCO3 --AER but also with Cl-_ AER (R2 = 87% and 77% respectively). In addition, variance of TAP explained by Cl-AER increased till 86% if phytase hydrolysable P in NaOH and CB extracts was taken into account. AERs have proved to be sensitive not only to Q, but also to buffer capacity (BC). When Olsen P was considered as the only predictive variable, it only explained a 53 % of the TAP variation (mean value for wheat and sunflower, P < 0.001). When the ratio Feca/Fecbd and the buffer capacity (BC) were also included as predictive variables in the model, variance explained increased to 69 and 61 %, respectively. When clay and CCE were included in the predictive model with Olsen P, variance explained increased till 80 %. Overall, better results obtained with AERs than with Olsen P or other single chemical extraction and were explained by their better relationship with most of the soil properties conditioning TAP. In the fifth chapter of this PhD Thesis, organic P forms and their potential contribution to available P to plants in Mediterranean soils were studied. Organic P in soil is the less understood aspect of the P cycle in soils, particularly in those from semiarid and arid lands. Organic P forms and how may be affected by P status and soil properties was dealt in this chapter. The study was carried out by NMR spectroscopy of 31P (31P NMR) in NaOH-EDTA extracts. From the soil collection used in the study described in chapter 1, eight representative Mediterranean soils were selected, and for each one, a "low-P" and "high-P" sample was studied. Organic P concentration in “high-P” and “low-P” samples was not significantly different. Beside this, the concentration of orthophosphate monoesters and their relative contribution to extracted P or OP according to the dry combustion method did not significantly differ between low- and high-P samples. Monoesters were the dominant fraction of OP accounting for half of the total organic P. The differences in monoesters phosphates between low- and high-P soil samples increased with increasing differences in Olsen P (R2 = 0.61, P < 0.05). The mono- to diesters ratio increased with increased Olsen P in soil (R2 = 0.49; P < 0.01), also monoesters and IP6, as OP proportion, increased as increased Olsen P in soil in vineyards and olive orchards, where almost no crop residue accumulation occurs. All these results reveal that, under P starvations conditions, some hydrolysis of these OP forms and potential contribution to P supply to plants may be expected. Myo- IP6 seems to stabilize by adsorption, as revealed by its positive relationship with the ratio of Fe in oxide to clay content and its decreased concentration in soil with increasing pH. On the other hand, precipitation of Ca phosphates may explain the retention of other IP6 stereoisomers in these soils. A general discussion of results is done in the sixth chapter, and finally, exposition of main conclusions of the present work is presented in the seventh chapter.

Author dc:creator

• Recena Garrido, Ramiro

Advisor dc:contributor.advisor

• Delgado García, Antonio

Statement dc:rights

• Attribution-NonCommercial-NoDerivatives 4.0 Internacional

Licence dc:rights.uri

• http://creativecommons.org/licenses/by-nc-nd/4.0/

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11441/45228

OAI identifier oai:identifier

oai:idus.us.es:11441/45228