Assessing the Effect of Bacillus Thuringiensis Var. Israelensis on Nontarget Chironomidae Emergence

Abstract

Bacillus-derived larvicides, which selectively target mosquito (Diptera: Nematocera: Culicidae) populations to reduce nuisance and health risks, were applied in the South March Highlands Conservation Forest near residential neighbourhoods in Ottawa, Ontario. The objective was to assess effects of application on the nontarget mosquito relative, Chironomidae (Diptera: Nematocera: Chironomidae), and other nontarget aquatic taxa captured using emergence traps. A secondary objective was to assess physicochemical variables that influence Chironomidae emergence. Study ponds received an application of Bacillus thuringiensis var. israeliensis, a subset also received an application of Bacillus sphaericus, and a group of control ponds were left untreated over 3 years (2016-2018). Weekly sampling included trap collections and measurements of water temperature, pH, water depth, conductivity, dissolved oxygen, ammonia, nitrate, and sulphate. Drought in 2016, high precipitation throughout 2017, and seasonal precipitation in 2018 influenced variable physicochemical conditions. Principal component analyses identified differences between sampling groups and between years. Redundancy analyses correlated insect emergence with pond pH, average water depth and water temperature and indicated a strong relationship between Chironomidae emergence and average water depth. Although significantly less Chironomidae annual emergence was observed at treated sites in 2017 and 2018, zero-inflated negative binomial generalized linear mixed modelling failed to detect a significant Bti treatment effect when controlling for within group variation. Rather, variations in pH, mean water depth and water temperature were identified as drivers of Chironomidae emergence. Culicidae emergence was reduced to zero briefly following treatment in 2017 and 2018. The model detected a marginal negative treatment effect on Culicidae in 2017 only, and a positive treatment effect in 2018 at the onset of a secondary hydroperiod, in the absence of treatment. Variations in pH and water temperature were also identified to be drivers of Culicidae emergence. Modelling failed to detect treatment effects on any of the nontarget taxa abundance, including Diptera, Lepidoptera, Ephemeroptera, Odonata, Coleoptera, Hymenoptera, and Arachnida. An inverse relationship between insectivore and prey taxa abundance was observed. In 2018, taxa richness increased between years and trended higher at treated sites and a positive relationship between insectivore and prey taxa richness was observed. In 2017, Shannon-Weiner index and Simpson’s index of diversity were higher at untreated sites, and in 2018 diversity indices were higher at treated sites, with taxa richness increasing between years and higher evenness trending at treated sites. Our data suggest that treatment effects were potentially shrouded by natural variability of physicochemical variables, especially due to the varying hydroperiod observed over the three years of sampling. Additional work is needed to capture average conditions and separate confounding variables from treatment effects. This study provides an inventory of the current wetland insect community in the South March Highlands Conservation Forest landscape that offers a reference for ongoing mosquito management.