Approaches to compare the strength of pelagic trophic cascades often use singular sampling programs for measuring trophic variables, thus potentially neglecting the spatial and temporal heterogeneity in the distribution of fish, zooplankton and phytoplankton. Here, we compared the composition of six trophic variables from three trophic levels in a deep oligotrophic lake within temporal (diel and seasonal) and spatial (horizontal and vertical) sampling resolutions. Mean values and ratios between the variables were compared between day and night, in three sampling months, four lake basins, and three water depths. Factor analysis was used to determine abiotic variables which may explain the heterogeneous distribution of the trophic variables. All six trophic variables were strongly heterogeneously distributed between the sampling months and the water depths, whereas horizontal and day–night differences were lower. Distribution of fish, zooplankton and phytoplankton correlated with water temperature and nutrient concentrations. Accordingly, for the use in comparative and meta-analyses, singular sampling programs in deep lakes have to integrate the entire water depth and are best repeated over several seasons. Alternatively, mean water temperature and nutrient concentrations may be used as covariates to diminish the unexplained variance between samples from different lakes.