Frass drop for monitoring relative abundance of large arboreal invertebrates in a New Zealand mixed beech forest

Biodiversity managers need robust and cost-effective tools to monitor ecosystem health and outcomes of management actions. Large arboreal invertebrates are important components of forest ecosystem food webs, but can be difficult and expensive to monitor because of their inaccessibility. Frass drop has been used extensively in Europe and North America to index the abundance of arboreal invertebrates, but has rarely been used in an ecological context in New Zealand. We assessed the practicality of, and developed protocols for, the frass-drop method to monitor the abundance of tree weta (Hemideina crassidens), stick insects (Phasmatodea) and molluscs (Mollusca) under three tree species at two sites within a New Zealand native forest. Raw material costs for litter trays were low ($NZ 16.60 per tray) and frass (faecal material) of all three groups was distinctive and could be easily and quickly extracted from litter samples by manual sorting. Stick insect frass drop weight varied by one to two orders of magnitude between tree species and site (21–95% occurrence), but tree weta and molluscs frass was common (65–90% occurrence) under all tree species at both sites. Within site spatial variation in frass drop was large and dominated by differences between sample trays, regardless of tree species. Sampling using clusters of three trays as sampling units was more efficient than single-tray sampling at achieving target levels of power for simple site-mean indices of abundance. The method has several advantages over other low-cost methods that are currently used, but requires further validation of the relationship between frass drop and invertebrate abundance in a New Zealand context.