When space can’t replace time: spatial vs temporal inference on soil organic matter after afforestation

Land-use change from natural grasslands to tree plantations is widely promoted as a climate-mitigation strategy due to increased aboveground biomass, yet its effects on soil organic matter remain uncertain. Because SOM changes slowly, most studies rely on spatial comparisons between afforested stands and adjacent natural grasslands, implicitly assuming that reference grasslands remain stable. Here, we compare spatial and temporal approaches to quantify changes in SOM stock and carbon source following afforestation in the Río de la Plata grasslands in South America. We analyzed surface soil samples from eight grassland sites converted to pine or eucalyptus plantations, using both a spatial comparison with adjacent grasslands and a temporal comparison based on archived pre-afforestation soil samples from the same sites. We quantified soil organic carbon (SOC) and δ¹³C to determine carbon input sources in particulate (POM) and mineral-associated organic matter fractions. Spatial and temporal approaches yielded divergent results for SOC in POM. Declines appeared under the spatial approach but not temporally, because POM increased in reference grasslands, challenging the assumption of temporal stability. In contrast,  δ¹³C shifts were more consistent across approaches, though weaker under the spatial approach, as declining atmospheric  δ¹³C from fossil-fuel emissions biased the reference grassland and lowered the estimated contribution from afforestation in both fractions. Our results demonstrate that spatial and temporal approaches are not interchangeable. While spatial comparisons may be more sensitive for assessing stock changes, temporal data are critical to infer SOM dynamics and C sources following afforestation, with implications for carbon accounting and paired-site designs.