Soil carbon and nitrogen emissions under farmer managed conservation agriculture in Zimbabwe

Abstract Worldwide agriculture operates under the threefold challenge of adapting to climate change and mitigating its effects while aiming for sustainable agricultural intensification to meet the food demands of a growing population. Conservation agriculture (CA), a combination of reduced tillage, diversified crop rotations, and mulching, claims to target all three challenges at the same time. However, major knowledge gaps regarding CA’s mitigation potential remain. This study used a mobile, closed chamber system to determine soilborne, greenhouse gas (GHG) emissions from rainfed, farmer-managed CA- and conventional agriculture (CONV), in northern Zimbabwe. Measurements were carried out in locations of contrasting soil fertility (Arenosols and Luvisols) and under contrasting environmental conditions (cold-dry, cold-moist, warm-dry, warm-moist). Additionally, a horizon-specific soil fractionation with consecutive soil carbon and nitrogen quantification was conducted. The GHG emissions from a total of 8 farms depended on soil temperature and moisture and tended to be higher in CONV fields, although differences were statistically not significant. Field emissions were highest under warm-moist conditions, which are prevailing for large parts of the growing season. Mean carbon dioxide (CO 2 ) emissions from Luvisols were 3.0% lower in CA fields (583 mg CO 2 m 2 h −1 ) than under CONV (601 mg CO 2 m 2 h −1 ), respectively 7.6% lower in CA fields (464 mg CO 2 m 2 h −1 ) than under CONV (502 mg CO 2 m 2 h −1 ) in Arenosols. Conservation agriculture reduced mean nitrous oxide (N 2 O) emissions by 17.5% from 0.27 mg N 2 O m 2 h −1 (CONV) to 0.23 mg N 2 O m 2 h −1 (CA) in Luvisols and by 54.7% from 1.16 mg N 2 O m 2 h −1 (CONV) to 0.53 mg N 2 O m 2 h −1 (CA) in Arenosols. The upper soil horizons of Luvisols had higher concentrations of particulate- and mineral-associated organic matter compared with Arenosols and lower soil horizons but no differences were noted between management systems. Our data indicate that the mitigation effects of CA are highly site-specific and that CA management practices can have unexpected negative effects on GHG fluxes. The unimodal rainfall distribution with a long dry winter period of 7 months and recurrent dry spells in northern Zimbabwe may prevent a net carbon sequestration under CA management that would have occurred in the humid tropics.