4.3 Drivers of soil organic matter fractions
The uniform decline in POC and MAOC in the same relative proportion among forest states except arid land indicated similar effects of degradation on both fractions. POM and MAOM are assumed to have different C turnover rates and degrees of physical-chemical protection owing to distinct origins (Lavallee et al., 2020; Nicolás et al., 2012). Comparing these fractions between different ecosystems, Yu et al. (2022) found that 90% of grassland/shrubland soils stored more SOC as MAOM than POM, whereas 40% of forest soils exhibited a higher POM fraction, which was attributed to the contribution of woody vegetation. Similarly, POC was shown to be highly sensitive to soil changes resulting from the conversion of dry forest to crops in the semiarid Chaco (Villarino et al., 2017). Focusing on the C component in POM and MAOM, we thus expected a proportionally larger fraction of POC in less degraded SDTF states given its origin primarily from plants (Baldock and Skjemstad, 2000). Contrary to our expectations, however, the POC/MAOC ratio was constant across natural, semi-natural, simplified and shrub-dominated forests in both seasons, despite large differences in vegetative biomass and structure among these states. This suggests that POC and MAOC were equally susceptible to forest degradation and, in general, degradation resulted in SOC loss despite mineral protection in soil aggregates. In contrast, studies from other ecosystems show a decoupling of POC and MAOC. For example, soils with high SOC tend to store more C as POC given that MAOC reaches saturation (Cotrufo et al., 2019). In disturbed soils, MAOC is more stable and less sensitive to disturbance (Sokol and Bradford, 2019). However, our results support the hypothesis that the two fractions are coupled, suggesting that microbial and plant residues contribute to both components to a similar extent and with the same dynamics (Angst et al., 2021; Córdova et al., 2018; Yu et al., 2022). It is also possible that soil texture, SOC content, and precipitation play important roles in the distribution of organic C within physically and chemically defined pools (Haddix et al., 2020; Yu et al., 2022). Despite significant progress toward understanding the intrinsic mechanisms of POC and MAOC distribution and functions in soils, further research is needed to identify the primacy of biotic versus abiotic controls on their dynamics in SDTFs subjected to different degrees of disturbance.
The higher POC/MAOC ratio in arid land compared to the other forest states implies a functional tipping point with extreme degradation (Ghazoul and Chazdon, 2017). This anomaly in arid land is likely the outcome of several related factors. First, the microbial C pump conceptualized by Liang (2020) that transforms plant residues to POC and finally to MAOC through microbial anabolism was probably disrupted by the harsh biotic and abiotic conditions of this state. Functioning of this mechanism occurs via microbial transformation of POC through microbial biomass and ultimately conversion into MAOC. Low inputs of leaf litter and root exudates as well as water stress inhibit microbial activity and biomass and constrain the C pump (Liang 2020), which could lead to changes in the proportions of POC and MAOC. Likewise, low vegetative cover means that arid land is prone to erosion, which leads to nutrient leaching and depletion of SOC stocks (Villarino et al., 2017). Second, lower quality litter tends to favour the formation of POC over MAOC (Cotrufo et al., 2013). Therefore, low litter quantity and quality combined with water stress in arid land likely contributed low decomposition rates (Machado et al., 2019) which, in turn, inhibited the conversion of POC into MAOC. Third, sandier soils in arid land could inhibit the formation of MAOC given that it tends to associate with fine silt and clay particles (Lavallee et al., 2020; Machado et al. 2019; Schulz et al. 2016). Any or all of these factors could result in a lower proportion of MAOC compared to less degraded states, altering the delicate balance between these two fractions and disrupting an important function of SOC storage and turnover. Given the implications of this tipping point for the resilience of the STDF ecosystem, we urgently need to better understand the mechanism underlying this result.