Phosphorus availability and its interaction with plant belowground carbon, nitrogen and phosphorus input into a Ferralsol
Closed access
Author
Date
2019Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
Tropical grasslands cover 44% of the worldwide grassland area and are mainly grown on Ferralsols and Acrisols. These highly weathered soils usually contain a low concentration of plant available phosphorus (P), due to the strong P sorption onto the soil solid phase and/or the small total P content, which often limits plant growth. More than half grass-alone pastures in South America are degraded due to over-grazing and low plant available nitrogen (N) and P in the soil. Fodder legumes integrated into grass-alone pastures can improve soil fertility and present a major contribution to alleviate pasture degradation. However, P availability affects the carbon (C) cycle by influencing its capture by plants and the N input in soil by limiting the symbiotic N2 fixation by legumes. The overall goal of this study was to achieve a quantitative understanding of the coupling of the C, N and P cycle in tropical pasture using a legume, Canavalia brasiliensis (canavalia).
The objectives of this thesis were (i) to design, develop and optimize a tri-isotope (13-C, 15-N, 33-P) plant labelling method in sand and assess the effect of the labelling homogeneity on BG input quantification, (ii) to quantify BG-C, N and P input, including roots and rhizodeposition, into a Ferralsol in response to N and P fertilization, and (iii) to examine their incorporation in different soil pools over time.
The 13-CO2 single-pulse labelling combined with the simultaneous 15-N and 33-P cotton-wick stem feeding could effectively label canavalia roots and facilitate the tracking of rhizodeposited C, N and P input from root systems. The tri-isotope pulse-labelling method did not provide a homogeneous root labelling, thus invalidating the assumptions on the quantification of rhizodeposition proposed by Janzen and Bruinsma (1989). The difference of measured 13-C:15-N and C:N ratios in percolates showed that the tracers did not reflect the elemental stoichiometry of the water-soluble rhizodeposits. Furthermore, soluble root exudates were always less enriched with isotopes than roots. This suggests an underestimation of the amount of rhizodeposition released during 2 weeks after labelling when using the equation of Janzen and Bruinsama (1989), as it assumes the rhizodeposits have the same isotopic enrichment as the roots.
When growing in a Ferralsol, the amount of BG input by canavalia determined at 4 weeks after labelling was dominated by the root biomass. Root biomass was closely related to the fertilzer supply of the soil. The proportion of water-soluble vs. insoluble rhizodeposition and their respective isotopic composition affected the quantification of rhizodeposited C, N and P. Nitrogen and P fertilization increased the BG proportion of rhizodeposited C, N and P by canavalia. In addition, roots and rhizodeposition showed similar C:N ratio, whereas C:P ratio in rhizodeposition were always higher than in roots, which ultimately influenced their incorporation in soil.
The incubation of pots over 8 weeks containing the entire labelled root system revealed that N and P fertilizer application promoted the root decomposition, reflecting the better quality of the root residues for microbial degradation. The presence of canavalia BG inputs enhanced the mineralization of the initial soil organic matter (SOM) compared to unplanted pots indicating a positive priming effect. Phosphorus addition increased the proportion of P derived from root degradation products in both microbial and soil available P pools
The multi-isotope approach allowed investigating the coupling of the C, N and P cycle in tropical soils. The great respiration rate and priming effect observed during the incubation with P fertilization revealed that SOM was rapidly mineralized because of the incorporation of fresh root residues. The quality (i.e. C:N:P ratio) of BG input controlled their incorporation into the soil pools and directly affected the C storage during the short-term of the experiment period.
The results obtained allow to better understand the C:N:P stoichiometry of legume BG inputs, which is a crucial factor to combat pasture degradation efficiently and to improve the fertility of pasture and cropping systems. Show more
Publication status
publishedExternal links
Search print copy at ETH Library
Contributors
Examiner: Frossard, Emmanuel
Examiner: Oberson, Astrid
Examiner: Abiven, Samuel
Examiner: Rao, Idupulapati M.
Examiner: Mayer, Jochen
Publisher
ETH ZurichOrganisational unit
03427 - Frossard, Emmanuel / Frossard, Emmanuel
More
Show all metadata
ETH Bibliography
yes
Altmetrics