Tree Risk Assessment: Component analysis of six visual methods applied in an urban park, Montevideo, Uruguay

Abstract

The urban trees must be properly evaluated to ensure their presence with acceptable risk levels, so that they provide ecosystem services. However, most Latin American countries lack their own or adaptations methods for this evaluation. To identify the applicability of some methods in Montevideo (Uruguay), the following were evaluated: A Photographic Guide to the Evaluation of Hazard Trees in Urban Areas (ISA/M&C); Best Management Practices - Tree Risk Assessment (ISA/BMP); Quantified Tree Risk Assessment (QTRA); Tree Hazard: Risk Evaluation and Treatment System (THREATS); Guide to Hazard Tree Management (USDA1); A Guide to Identifying, Assessing and Managing Hazard Trees in Developed Recreational Sites of the Northern Rocky Mountains and the Intermountain West (USDA2). The methods were organized by the components: likelihood of failure, likelihood of impact, consequence of failure and risk rating, and were applied by two types of assessors to trees of Eucalyptus and Tipuana. A generalized linear mixed model was applied, identifying significant differences between methods, assessors, genus, and their interactions. The results showed that there were significant differences amidst the methods, for all components. There were significant differences between assessors for the likelihood of failure and the risk rating, and a "method x type of assessor" interaction, associated with the consequence of failure. There were no significant differences between genus, however, there was a "genus x method" interaction for the likelihood of impact. The part size interpretation on the consequence analysis explained the differences between assessors for the QTRA, USDA1 and ISA/M&C methods; the location of trees presented greater influence in the likelihood of impact on the USDA2 method. In this context, to monitor and carry out tree management for urban parks of Montevideo, it is recommended to apply the unchanged ISA/BMP method. The QTRA and USDA1 methods also presented good applicability, but they should be adjusted.