The hotspot of biodiversity in the Andes of Southern Ecuador has been severely threatened by climate change and unsustainable land use. The high biodiversity requires strategies for conservation and management of natural resources to be developed at both individual and area-wide levels. In this paper we focus on the development of an automatic treecrown detection and classification approach, which is in line individual-based investigations in the tropical mountain environment. Airborne laser scanning of discrete type was used with a very high granularity (<10 returns per square meter). The individual tree crown detection reached an accuracy of 51% while supervised classification of palm-trees reached an accuracy of 69%. Accuracy measurements are given in the paper. The detection and characterization of individual tree crowns is the first step in the development of a monitoring approach for the tropical mountain forest.