In the Northern Ecuadorian Andes, deforestation severely lowered the natural Upper Forest Line (UFL). Thus, to ascertain the natural position of the UFL, e.g. To support -sustainable reforestation, the RUFLE1 program investigates the local vegetation history. Andisols could provide great opportunities for this paleoecological study, since they are characterized by the accumulation of vast amounts of Organic Matter (OM) and are dominant in the region. However, the use of these soils requires an understanding of the chronostratigraphy and preservation of Soil Organic Matter (SOM). To investigate SOM chronostratigraphy and preservation, we selected soil profiles at altitudinal intervals of 100 m between 3300 and 4100 m.a.s.l. Within the Guandera Biological Reserve. The Mean annual temperatures range from 4–10°C but daily fluctuations are considerable (up to 20°C) with a mean annual precipitation of 2000 mm. The current forest reaches approximately 3650 m.a.s.l., above this UFL tropical alpine grasslands dominate (páramo). The Andisols in Guandera contain high amounts of organic carbon (average 16 ± 4 % in topsoil), are acidic (pHCaCl2 = 4 in topsoil) and contain some allophane in the subsurface horizons (Alp/Alo: 0.3 -1.20 from subsurface to surface). The chronostratigraphy of OM in soils is less straightforward than in deposits traditionally used in paleoecological studies (e.g. Peat sequences and lake sediments). Volcanic activity determines the sedimentary stratification of Andisols and sets boundary conditions for the distribution of SOM. However pedogenetic processes, such as bioturbation and leaching, finally determine its vertical distribution in the soil. Preliminary interpretation of the major and trace elements indicate that the Andisols are formed in at least 3 different parent materials, in the field recognizable as the current soil, a paleosol and a truncated paleosol. Further examination will reveal if differences within these ashes are related to upward fining or point to the presence of even more layers. This knowledge then supports the interpretation of the radiocarbon dates. In the current soil, the radiocarbon age increases with depth (preliminary sample distance 30–40 cm). If the current soil is formed in multiple ash layers, the sedimentation rate mainly determines the age-depth relation of SOM. On the other hand, if the current soil is formed in one ash deposit, two scenarios may explain this age to depth relation. In the first scenario, without bioturbation, the young SOM continuously buries the older SOM. In the second scenario, with bioturbation, an age-depth correlation can only exist if the zone of active bioturbation shifts upward during SOM accumulation. In the forest soils, the presence of a thick (up to 50 cm) fermentation horizon implies that OM incorporation is slow and deep bioturbation is virtually absent. In the páramo soils, ectorganic horizons were not observed, suggesting higher rates of OM incorporation into the soil i.e. Higher bioturbation rates. However, litter accumulates within the grass tussocks instead of the soil surface, preventing formation of a true ectorganic layer. We applied micromorphological analysis to investigate bioturbation and to substantiate these field observations. The effect of leaching on the vertical distribution of OM is limited due to the high OM-metal ratios. Using pollen and plant-specific organic molecules (biomarkers) in paleoecological studies requires preservation in soil for long periods of time. Pollen in Colombian Andisols appeared to be protected against degradation due to its incorporation in biological and/or physico-chemical aggregates. We applied micromorphological techniques to find out if and how pollen was preserved in our soils, in order to rule out species-specific degradation. Lignin may be a potential biomarker to distinguish grasses (páramo) from trees (forest), because its composition is related to plant taxonomy. Unfortunately, in andic mineral horizons, lignin is rapidly degraded and can thus not be used as a biomarker of past vegetation. By contrast, extractable lipids, with differing compositions between plant species, appear to be well preserved in the studied Andisols. Our preliminary results showed that they can be used as biomarkers of past vegetation. 1 RUFLE = Reconstruction of the Upper Forest Line in Ecuador