RESUMO
BACKGROUND: Human population growth has led to biodiversity declines in tropical cities. While habitat loss and fragmentation have been the main drivers of urban biodiversity loss, man-made interventions to reduce health risks have also emerged as an unintentional threat. For instance, insecticide fogging to control mosquito populations has become the most common method of preventing the expansion of mosquito-borne diseases such as Dengue. However, the effectiveness of fogging in killing mosquitoes has been called into question. One concern is the unintended effect of insecticide fogging on non-target invertebrates that are crucial for the maintenance of urban ecosystems. Here, we investigate the impacts of fogging on: (1) target invertebrate taxon (Diptera, including mosquitoes); (2) non-target invertebrate taxa; and (3) the foraging behavior of an invertebrate pollinator taxon (Lepidoptera) within an urban tropical forest. METHODS: We carried out fogging with Pyrethroid insecticide (Detral 2.5 EC) at 10 different sites in a forest situated in the state of Selangor, Peninsular Malaysia. Across the sites, we counted the numbers of knocked-down invertebrates and identified them based on morphology to different taxa. We constructed Bayesian hierarchical Poisson regression models to investigate the effects of fogging on: (1) a target invertebrate taxon (Diptera) 3-h post-fogging; (2) selected non-target invertebrate taxa 3-h post-fogging; and (3) an invertebrate pollinator taxon (Lepidoptera) 24-h post-fogging. RESULTS: A total of 1,874 invertebrates from 19 invertebrate orders were knocked down by the fogging treatment across the 10 sites. Furthermore, 72.7% of the invertebrates counted 3-h post-fogging was considered dead. Our regression models showed that given the data and prior information, the probability that fogging had a negative effect on invertebrate taxa 3-h post-fogging was 100%, with reductions to 11% of the pre-fogging count of live individuals for the target invertebrate taxon (Diptera), and between 5% and 58% of the pre-fogging count of live individuals for non-target invertebrate taxa. For the invertebrate pollinator, the probability that fogging had a negative effect 24-h post-fogging was also 100%, with reductions to 53% of the pre-fogging count of live individuals. DISCUSSION: Our Bayesian models unequivocally demonstrate that fogging has detrimental effects on one pollinator order and non-target invertebrate orders, especially taxa that have comparatively lower levels of chitinisation. While fogging is effective in killing the target order (Diptera), no mosquitos were found dead in our experiment. In order to maintain urban biodiversity, we recommend that health authorities and the private sector move away from persistent insecticide fogging and to explore alternative measures to control adult mosquito populations.