Computational protocol: Assessing the efficacy of fathead minnows (Pimephales promelas) for mosquito control

Similar protocols

Protocol publication

[…] We considered any dip-cup sample that contained either live mosquito larva or exoskeletons positive for the presence of mosquitoes. We then calculated the proportion of positive dips at each site and for each sampling occasion as Xpositive dips / Ndips. To estimate the influence of minnows on mosquito larvae density, we developed generalized linear mixed models using the nlme package in R (version 3.2.2) with the proportion of positive dips at each site as the response variable. Site (pond_id) was included as a random effect (intercept) in each model to account for random variation among sites. We specified an autoregression covariance structure (AR1) to account for temporal correlations among samples within each year and included year as a fixed effect in each model. We calculated Pearson’s correlation coefficient among all continuous covariates (fixed effects, ) and did not include variables correlated r > | 0.65 | in the same model. The total model set included 12 candidate models to assess the relative influence of covariates on the proportion of positive dips. Models were compared using Akaike’s information criteria adjusted for small sample sizes (AICc). Model fit was assessed through inspection of model residuals. We applied the control site trend models to the initial values (i.e., sampling occasion 1) in the treatment ponds to predict the proportion of positive dips over time in the absence of a treatment. Essentially, these models predict the likely level of larva densities in the treatment ponds in the hypothetical absence of treatment, assuming these ponds would have followed the same seasonal trajectory as the control sites.We analyzed the effect of pond morphology and water chemistry on the abundance of minnows (measured as catch per unit effort) in each treatment pond using a similar modeling approach. We modeled the average number of fish per transect (rounded to the nearest integer) using Poisson linear mixed effects models including a random intercept for each pond. All covariates were standardized prior to analysis and we assessed the level of correlation among all potential covariates for both the pond morphology and water chemistry model sets. We did not include any covariates correlated r > | 0.65 | in the same model. We used AICc to rank models. Models were estimated using the lme4 package (Bates et al. 2014, 2015) in R software.Minnows were obtained from a state-authorized Commercial Hatchery (Pleasant Valley Fish Farm, Nebraska). Research was conducted under a Chapter 33 Permit for scientific research issued by the Wyoming Game and Fish Department (#916). Minnows were euthanized for isotope analysis by decapitation. This research was conducted in compliance with the Animals for Research Act of Ontario (Revised Statutes of Ontario), the Guide to the Care and Use of Experimental Animals from the Canadian Council on Animal Care and the University of Waterloo’s Guidelines for the Care and Use of Animals in Research and Teaching (AUPP # 13–12). […]

Pipeline specifications

Software tools nlme, lme4
Application Mathematical modeling
Organisms Pimephales promelas, Homo sapiens, Human poliovirus 1 Mahoney, Drosophila melanogaster, Danio rerio, Caenorhabditis elegans
Diseases Q Fever, HIV Infections