The use of long-term observations of soil respiration for calculations of net carbon balance in Chernozem zone

Long-term (2017–2023) year-round observations of soil CO₂ efflux (soil respiration, SR) in the area of typical Chernozems distribution (Kursk region, Medvensky district, Kursk Biosphere Station of the Institute of Geography RAS) were carried out using a unified technique based on field measurements with IRGA. Annual estimates of SR were calculated in the most representative local anthropogenic and natural ecosystems that form the forest–steppe agro–landscape in Chernozem zone of European Russia (permanently used arable land, non–fertilized vegetable garden, 2–8 yr fallow, meadow steppe, and broadleaved forests). According to estimates averaged by various models and calculation schemes, SR in the study area varies from 650.6 ± 88.2 (non–fertilized vegetable garden) to 1550 ± 215 (2–8 yr abandoned land) g C–CO₂ m^–2 yr^–1, amounting to 895 ±91 on permanently used arable land and 1040 ± 55 g C–CO₂ m^–2 yr^–1 in natural ecosystems. The share of microbial (heterotrophic) respiration (MR) in local soils averaged 55.5% (broadleaf forests), 77.6% (abandoned land), and 72.4% (meadow steppe) during the growing season. The ecosystem net carbon balance of the three broadleaf forest sites, calculated by the difference between net primary production (NPP estimated by CBM–CFS3 model and taxation data) and heterotrophic respiration (SR and CO₂ losses from decomposition of coarse woody debris, CWD) vary from atmospheric carbon uptake +130 g C–CO₂ m^–2 yr^–1 to its source –112 g C–CO₂ m^–2 yr^–1. In the latter case, the net carbon source is due to the accumulation of significant stocks of CWD as a result of tree mortality from xylophages. On arable lands, similarly, calculated by NPP and MR net carbon balance, according to the expert assessment of MR share, vary from +186 to –143.7 g C–CO₂ m^–2 yr^–1, with almost ultimate contribution of the crop type grown in a given year to inter–annual variations of SR (87.9%). It is concluded that the above approach is applicable to estimates of the net C balance in rather different ecosystems, however, the greatest uncertainty is introduced by estimates of the soil MR share.