A rapid method of estimating catches of abundant fruit flies species in modified steiner traps

Abstract

Males of many fruit fly species (Diptera: Tephritidae) are responsive to lures and these are used in modified Steiner traps for the regular monitoring of populations of fruit flies. Unfortunately, male lures are so attractive to many species of fruit fly that even weekly collections can produce catches of many hundreds of flies from a single trap. To help facilitate the assessment of trap catches in cases such as this, we compared total fly counts from real trap catches with various manual and computer-simulated subsamples of those catches. Based on computer simulations, subsamples of 75, 100 and 200 flies accurately reflected the frequency of different fruit fly species in the original trap catch (mean trap catch size = 617.42 ᠵ3.72 SD flies, n = 26 weekly trap catches). However, based on the Chi-squared analysis, subsample sizes of less than 200 individuals run the risk of being statistically unreliable. Subsamples based on weight were found to be just as accurate as subsamples where flies were drawn individually from the parent sample and thus offer a rapid method of producing accurate subsamples. Rare species were not well represented at any of the three subsample sizes and if the object of a program is to record rare species, then subsampling is not recommended. Where abundant species are the focus of a trapping program, subsampling offers a statistically reliable and time saving approach for assessing trap-catch.