Amino acid ¹⁵N in long-term bare fallow soils: influence of annual N fertilizer and manure applications

Long-term dynamics of amino acids (AAs), from a bare fallow soil experiment (established in 1928 at INRA-Versailles, France), were examined in unamended control (Con) plots and plots treated with ammonium sulphate (Amsul), ammonium nitrate (Amnit), sodium nitrate (Nanit) or with animal manure (Man). Topsoil (0-25 cm) from 1929, 1963 and 1997 was analysed for C, N and ¹⁵N content and distribution of 18 amino acids recovered after acid hydrolysis with 6 m HCl. With time, soil N, C and AA content were reduced in Con, Amsul, Amnit and Nanit, but increased in Man. However, the absolute N loss was 3-11 times larger in Man than Nanit, Amsul, Amnit and Con, due to the much higher N annual inputs applied to Man. From 1929 to 1997 in Con, Amsul, Amnit and Nanit the whole soil and non-hydrolysable-N pool δ¹⁵N increased associated with the loss of N (indicative of Rayleigh ¹⁵N/¹⁴N fractionation). No δ¹⁵N change from 1929 to 1997 was found in the hydrolysable AA-N (HAN) pool. Fertilizer N inputs aided stabilization of soil AA-N, as AA half-life in the mineral N fertilizer treatments increased from 34 years in 1963 to 50 years in 1997. The δ¹⁵N values of alanine and leucine reflected both source input and ¹⁵N/¹⁴N fractionation effects in soils. The δ¹⁵N increase of ornithine (~6[per thousand]) was similar to the whole soil. The δ¹⁵N change of phenylalanine in Con (decrease of 7[per thousand]) was related to its proportional loss since 1929, whereas for Amsul, Amnit, Nanit and Man it was associated with isotope effects caused by the fertilizer inputs. However, the soil δ¹⁵N value of most individual amino acids (IAAs) did not significantly change over nearly 70 years, even with mineral or organic N inputs. We conclude for these bare fallow systems that: (i) δ¹⁵N changes in the whole soil and non-hydrolysable AA pool were solely driven by microbial processes and not by the nature of fertilizer inputs, and (ii) without plant inputs, the δ¹⁵N of the HAN pool and (most) IAAs may reflect the influence of plant-soil interactions from the previous (arable cropping) rather than present (fallow) land use on these soil δ¹⁵N values.