Limited effect of organic matter addition on stabilised organic carbon in four tropical arable soils

Abstract. Arable soils are generally characterised by a low soil organic carbon (SOC) content, with negative consequences for soil health, crop yield and global climate. Thus, over the past decades, there has been a focus on how agricultural management practices, such as organic resource addition, can increase the amount of SOC. To sustainably increase SOC stocks, a portion of the organic amendments added to the soil has to be stabilised in persistent fractions such as mineral-associated organic carbon (MAOC). However, there is a lack of research on the magnitude of changes in MAOC in tropical agroecosystems in response to organic resource additions. Here, we show for four long-term field trials in Kenya that the addition of large amounts of organic amendments (farmyard manure or Tithonia diversifolia biomass at 4 t C ha^-1 yr^-1 for 16 to 19 years) to maize monocropping systems had variable effects on topsoil MAOC stocks (0–15 cm depth), and no significant effect on subsoil MAOC stocks (15–50 cm depth) compared to a control treatment. The addition of mineral N fertiliser did not affect MAOC stocks at any site. Using stable carbon isotopes δ¹³C, we found that the portion of topsoil MAOC originating from Tithonia biomass was larger in the sandy (25–40 %) compared to the clayey soils (0.5–12 %), while the portion of total added Tithonia biomass that was stabilised over a time period of 16–19 years was below 7 % across all sites, or a SOC stabilisation rate of 0.8–27 g C m^-2 yr^-1. Using these results, we conclude that while in sandy soils the stabilisation of added OC contributed substantially to limiting SOC losses upon cultivation, this was not the case for clayey soils. These differences were due to the much lower SOC stocks in the sandy soils, compared to the clayey soils. Our results underline the challenges associated with improving soil health in sub-Saharan Africa and stress the need for more research to reliably assess if and how additional SOC can be stabilised over decadal time scales in highly weathered tropical soils.