Climate change across the western US has increased air temperature, resulting in decreased snow and lengthening of the summer drought. Recent studies have also highlighted the positive feedback loops between soil moisture and vapor pressure deficit (VPD), which can exacerbates aridity in water limited ecosystems. These interactions can make it difficult to untangle the influences of soil moisture and VPD on tree growth, and yet our ability to untangle these parameters is important for building a predictive framework for forests vulnerability to drought. In montane ecosystems, the complexity of watershed characteristics, such as aspect, microtopography, soil properties, and plant species composition can further complicate the interactions between source water use and VPD, and influence how these parameters are recorded in the d18O of wood cellulose (d18Owc). Thus, a primary objective of this study was to understand if there were consistent and predictable patterns in d18Owc that were controlled by watershed characteristics and/or species composition. Through d18Owc analysis of three dominant tree species, including Ponderosa pine, Douglas fir, and Engelmann spruce, we found that topographic mediation of the microclimate, such as soil moisture and VPD, had a larger influence on the d18Owc than species specific traits. We also found that accurate estimates of source water d18O was important for modeling d18Owc, and that we could not assume d18O of source water was the same as d18O of precipitation. Finally, we found that including the influence of the Peclet effect was more important in the d18O model in trees that were growing in more mesic landscape positions, compared to drier landscape position.