How plant inputs from above- versus below-ground affect long term carbon (C) and nitrogen (N) retention and stabilization in soils is not well known. We present results of a decade-long field study that traced the decomposition of ¹³C- and ¹⁵N-labeled Pinus ponderosa needle and fine root litter placed in O or A soil horizons of a sandy Alfisol under a coniferous forest. We measured the retention of litter-derived C and N in particulate (>2 mm) and bulk soil (<2 mm) fractions, as well as in density-separated free light and three mineral-associated fractions. After 10 years, the influence of slower initial mineralization of root litter compared to needle litter was still evident: almost twice as much root litter (44% of C) was retained than needle litter (22–28% of C). After 10 years, the O horizon retained more litter in coarse particulate matter implying the crucial comminution step was slower than in the A horizon, while the A horizon retained more litter in the finer bulk soil, where it was recovered in organo-mineral associations. Retention in these A horizon mineral-associated fractions was similar for roots and needles. Nearly 5% of the applied litter C (and almost 15% of the applied N) was in organo-mineral associations, which had centennial residence times and potential for long-term stabilization. Vertical movement of litter-derived C was minimal after a decade, but N was significantly more mobile. Overall, the legacy of initial litter quality influences total SOM retention; however, the potential for and mechanisms of long-term SOM stabilization are influenced not by litter type but by soil horizon.

Microbial biomass represents a substantial source of labile C contributing to soil organic matter (SOM) maintenance. Microbial residues may associate with the soil matrix through a variety of mechanisms, reducing its bioavailability and increasing its persistence in soil. Our objective was to examine soil matrix effects on the stability of non-living microbial C inputs in two contrasting forest ecosystems by following microbial residues (Fungi, Actinobacteria, Gram-positive bacteria (Gm +), Gram-negative bacteria (Gm −)) into SOM fractions in a temperate forest in California (CA) and a tropical forest in Puerto Rico (PR) for 3 and 2 years, respectively. We isolated 3 SOM fractions: (i) free light fraction (FLF), (ii) occluded light fraction (OLF), and (iii) dense fraction (DF). Additionally, we characterized SOM fraction chemistry to infer quality and source of native fraction SOM. Our results showed greater stabilization as mineral-associated microbial C (i.e., as DF and OLF), compared with loose detrital C (i.e., FLF). There was no microbial group effect (i.e., differences in fraction C recovery among different microbial cell types). Our findings suggest that mineral association is more important for stabilizing non-living microbial C in soil than the cellular structure of the initial source of microbial inputs, with site specific edaphic factors as the major controllers of the amount of microbial residues stabilized.