The canopy arthropod community in tropical montane cloud forests has rarely achieved research attention, but this remote, diverse community is likely under considerable threat from climate change. Modern techniques are synthesised in this thesis to investigate this community. Accurate climate knowledge is necessary for any climate change research, and this thesis utilises current and predicted climate data to aid in sampling, and validates these data using ground-truthed measurements. Modern molecular ecology techniques enable high-taxonomic-resolution studies such as this, but the pipelines need careful validation: this thesis presents an empirical study to examine biases in metabarcoding, new tools for easy validation of bioinformatic pipelines, and discusses taxonomic assignment of anonymous sequences. A new canopy sampling design is employed to sample canopy arthropods in two tree species in a rainforest, and metabarcoding is successfuly used to uncover high levels of spatial turnover and phylobetadiversity across a highly heterogenous landscape in a spatially explicit nested study.