Elucidating the complex interrelationships of soil organic carbon fractions with land use/land cover types and landform attributes in a montane ecosystem

Soil organic carbon (SOC) consists of different fractions characterized by different physical sizes, chemical composition, and turnover rates. As these fractions might respond differently to the environmental factors, deciphering their relationships with environmental factors is useful to better understand carbon (C) dynamics. Therefore, this study was aimed to elucidate the complex interrelationships of SOC fractions with land use/land cover (LULC) types and landform attributes in a montane ecosystem of Bhutan. Soil samples were collected based on genetic horizons from the sites predetermined by condition Latin hypercube sampling (cLHS) scheme. Samples were then fractionated into > 53 and < 53 mu m fractions and their C contents determined using Vario Max CNS Analyzer. The C content in each of the fractions > 53 and < 53 mu m corresponds to particulate organic carbon (POC) and humic organic carbon (HOC), respectively. A spline function was fitted to interpolate the POC and HOC concentrations of the genetic horizons to determined depths (0-20, 20-40, 40-60, 60-80, and 80-100 cm). The resulting data were analyzed using multivariate ordination techniques and linear mixed model followed by a posteriori mean separation tests. POC in the upper depths was positively correlated to slope and altitude, while it was negatively correlated to multi-resolution valley bottom flatness (MR-VBF), SAGA wetness index (SWI), and normalized difference vegetation index (NDVI). POC was also positively associated with forests and northern aspect in the upper depths and shrublands in the lower depths but was negatively associated with agricultural lands and southern aspect in the upper depths and grasslands and western aspect in the lower depths. Conversely, HOC was positively correlated to MR-VBF, SWI, and NDVI but was negatively correlated to slope and altitude in the lower depths. While HOC was positively associated with agricultural lands, southern and western aspect, and grasslands, it was negatively associated with northern aspect, shrublands, and forests at some depths. Both POC and HOC concentrations were significantly higher under forests and shrublands, 4000-5520 and 3500-4000 m altitudinal zones and north-facing slope than other LULC types, altitudinal zones, and aspect directions. This study shows that POC and HOC respond differently to various LULC types and landform attributes depicting their unique interrelationships with the environmental attributes. Therefore, this has implications on land management practices to maintain optimum levels of both POC and HOC under different LULC types to enhance sustainable agriculture, ecosystem services, and C sequestration.