A comparative field trial, conducted without pre-treatment census baiting, of the rodenticides zinc phosphide, thallium sulphate and gophacide against Rattus norvegicus

Field trials of a new sub-acute rodenticide flupropadine, against wild Norway rats (Rattus norvegicus)

Fifteen experimental treatments with rodenticidal baits containing 0.1, 0.15 or 0.2% flupropadine were conducted on farmsteads against Rattus norvegicus infestations. Eight treatments were completely successful and the others gave kills ranging from 36 to 72% in 28 days. Treatments with 0.1 and 0.15% flupropadine were less successful against large infestations than against small ones. Flupropadine was most successful at 0.2% but still gave incomplete kills on farms where abundant alternative food was available. The compound was more effective than acute poisons in achieving complete control of Norway rat infestations, but was less reliable in doing so than anticoagulants. On the other hand, many flupropadine treatments gave quicker control and used smaller quantities of bait than anticoagulant treatments.

Laboratory and field evaluation of pyriminyl as a poison for Rattus norvegicus

Laboratory tests indicated that the optimum concentration for pyriminyl in rat baits was between 1% and 3%. In field trials in which 0.5% pyriminyl (the concentration in commercial use) was compared with 2.5% zinc phosphide for the control of rats on farms, the pyriminyl treatments were significantly less effective than the zinc phosphide even when the poisoned baits were left down for 7 days instead of 1 day after prebaiting. Both poisons were as effective in medium oatmeal bait as they were in medium oatmeal containing 5% corn oil and 5% sugar.

Statistical analysis of comparative field trials of acute rodenticides

The present method of analysis of data from comparative field trials of acute rodenticides was compared with two alternative methods of analysis, using computer simulation techniques. One of the proposed alternative analyses, the use of post-treatment census takes as a percentage of pre-treatment census takes in an analysis of variance, was found to be more accurate, to avoid a theoretical difficulty associated with the present method, and is computationally much simpler.