Quantification of soil C inputs from organic fertilizers in tropical long-term field experiments: potential of stable carbon isotopes
Résumé
<p>Soil C stocks can be increased by spreading organic fertilizer (OF) in crop fields. OF-derived C (OF-C) is usually estimated according to the differential with and without OF inputs [1-4]. But OF applications may boost crop production or induce initial-C mineralization due to an indirect effect (e.g. priming) [5-6]. Therefore, crop derived C and native-C (before plot testing) may change during the experiment link to OF additions. Thus, the differential method might not be the suitable one for quantifying OF-C. In this study, we used stable <sup>13</sup>C isotopes to avoid such OF-C estimation biases and compared two isotopic methods to the differential method. Both isotopic methods were set up with synchronous controls (e.g. soil δ<sup>13</sup>C signature compared to plot with and without OF inputs) and diachronous control (e.g. soil δ<sup>13</sup>C signature compared to the soil at the beginning of the experiment). In order to assess the all three methods, this study was implemented on an Arenosol and an Andosol with a 13-year history of compost or slurry amendment. The differential and synchronic isotopic methods gave similar OF-C estimations for the Arenosol, while for the Andosol both isotopic methods estimated twofold higher OF-C levels compared to the differential method. Changes in crop-C production or priming as a result of OF applications might explain this gap. Moreover, the control isotopic signature (without OF) slightly changed due to crop-C integrated during the experiment. Which is why the isotopic synchronic method was the most suitable compared to diachronic isotopic method. According to this method, OF-C retention was OF-nature dependent (21% for compost, 8% for slurry), and soil type and climate dependent (42% compost retention in the Andosol and 21% in the Arenosol), highlighting the recent carbon input retention capacity of Andosols. This method is also relevant to quantify the priming effect in field trials, in our case it was not possible due to the δ<sup>13</sup>C evolution of the soil without OF input.</p><p>[1] Y. Lou et al., CATENA, vol. 87, no 3, p. 386‑390, déc. 2011, doi: 10.1016/j.catena.2011.07.006.</p><p>[2] L. Paetsch et al., Agriculture, Ecosystems & Environment, vol. 223, p. 211‑222, mai 2016, doi:10.1016/j.agee.2016.03.008.</p><p>[3] H. Liu et al., Agriculture, Ecosystems & Environment, vol. 265, p. 320‑330, oct. 2018, doi: 10.1016/j.agee.2018.06.032.</p><p>[4] F. Liang et al., Geoderma, vol. 337, p. 853‑862, mars 2019, doi: 10.1016/j.geoderma.2018.10.033.</p><p>5] Y. Kuzyakov et al., Soil Biology and Biochemistry, vol. 32, no 11, p. 1485‑1498, oct. 2000, doi:10.1016/S0038-0717(00)00084-5.</p><p>[6] S. Fontaine et al., Soil Biology and Biochemistry, vol. 35, no 6, p. 837‑843, juin 2003, doi: 10.1016/S0038-0717(03)00123-8.</p>