Observations began at each facility's yearly fumigation application with ProFume™ (sulfuryl fluoride) gas. Fumigations were conducted as part of annual treatments performed by professional, commercial fumigators. Treatments were considered to have met the standards of the professionals conducting them and so trapping observations were interpreted through this lens of successful application. Treatments occurred in July (Facility A) and August 2011 (Facility B) in the first year and in early August 2012 for Facility B in the second year. Less than 24 h before treatment, vials containing adult and egg Tribolium castaneum (Herbst) were set out in each facility (Facility A: 181 adult vials, 63 egg vials; Facility B: 90 adult vials, 90 egg vials). Each vial contained five individuals of each life stage. Mortality after exposure was used as a measure of treatment effectiveness, with adult mortality measured by the number of dead following exposure and egg mortality measured by the number of individuals that failed to reach maturity after 42 d post-fumigation. Following treatment, Storgard™ Dome monitoring traps targeting T. castaneum were set out on each floor of the facility with a total of 50 traps setup in Facility A (∼12.5 traps/floor) and a total of 100 traps placed in Facility B (∼10 traps/floor). Adult beetle trap capture was recorded monthly for a period of 5 months in Facility A and 4 months in Facility B. Due to the working environment of both facilities, some traps were occasionally disturbed or removed from the environment by facility employees. To limit the effect of these disturbances, trap capture averages only used undisturbed traps and trap capture rates were standardized as “Number of beetles per trap per day.”
We evaluated each facility on its level of sanitation, based on a standardized set of criteria developed by Tsai (2010). Facility managers were also interviewed regarding the costs associated with sanitation and other pest control methods, as well as their knowledge or attitudes toward both. Cost of pest control was estimated from the sum total of investments for all classes of pest control used in the facility (fumigation, residual sprays, etc.) and cost of sanitation was estimated from average employee wage multiplied by the number of man-hours devoted to cleaning (Zagurramurdi et al., 2007). When an exact value was not available (i.e. if the information was the responsibility of another department), managers were asked to provide a reasonable estimate, based on prior experience.
2.2. Online survey
An online survey was developed using Qualtrics™. A request for participants was advertised in two milling journals, with a link directing the readers to the online survey. Respondents were asked a series of 31 questions relating to their facility, its pest management and sanitation programs, and their perceptions and knowledge of both. All responses were collected anonymously to protect respondents and their employers' identities.