Flies Impacting Livestick, Poultry and Food Safety

NON-TECHNICAL SUMMARY: Stable flies bite livestock and people, and house flies spread pathogenic microbes. Both kinds of flies are difficult to control. Source reduction is a non-chemical strategy for preventing outbreaks of stable flies, but knowledge is needed to understand where the flies come from, and how to anticipate outbreaks. While it is likely that house flies moving from and among animal confinement facilities spread pathogenic organisms, distance of spread is not currently understood. Objectives in this project will develop understanding of the mechanisms that determine how many stable flies will occur seasonally in different US locations, and whether or not the flies repopulate northern locations by weather assisted spread from southern sources. Other objectives will determine how risk of fly-borne pathogen transmission declines with distance from individual farms, and the extent to which insecticide resistance has developed among different US locations.

<p>

APPROACH: <ol>
<li> Records of numbers of flies caught per trap per day at 60 different combinations of sites and years throughout the US will be analyzed using mixed regression models to assess relations between population growth and density and weather effects. Growth rates will be analyzed with a mixed model response surface approach, using preceding densities, temperatures, and precipitation as candidate predictors. First dates of stable fly appearance will be evaluated to estimate when northward dispersal of stable flies has historically occurred. Locally, stored feed and manure handling systems will be surveyed to locate overwintering stable fly larvae, and emergence traps will be placed on potential overwintering sites to collect emerging flies. Samples from MN and other northern regions of the US will be evaluated for genetic similarity using mitochondrial and microsatellite markers in serial samples collected in spring until populations disappear in the fall. <li> Quantitative models for spread of house flies and associated pathogens from point sources will be developed from measures of microbe and fly densities at source points, frequency distributions of microbes acquired per fly, rates of loss per unit time, and spread of flies from point sources into surrounding landscapes. The model will be used to analyze sensitivity of spread distance to component parameters, and to project relative risk of spread as a function of radial distance from the source <li> As part of a national survey for insecticide resistance, house flies will be netted from confinement facilities, and F1 offspring assayed for susceptibility to selected insecticides, using a standardized protocol to be replicated in all participating states. </ol>

PROGRESS: 2007/01 TO 2007/12 <BR>
OUTPUTS: We analyzed a 16-year data set from a beef facility in Iowa to gain insights into whether founding stable fly populations overwinter locally or immigrate from the south, and if density dependence and weather could account for variation in population growth during the breeding season. Abundance of stable flies was measured with white sticky traps operated continuously at fixed stations. Trapping each year was done at 1-3 d intervals from late winter until autumn when weather became too cold. Abundance was indexed as no. flies caught per 10 day-degrees (DDs) above a flight threshold of 5 C. A grand total of 56,550 flies were trapped on the 10 traps during 2,233 trapping intervals spanning the 16 years. Analysis was done in log scale. Temperature records were used to convert calendar to generation time, assuming a generation required 317 DD above a base of 10.8 C. Daily wind trajectories were examined with HYSPLIT 4.8. A day was counted as having a southerly wind event if its 12-hr backwards daytime air parcel trajectory was at least 50 km south of Ames, and its temperature exceeded 15 C for at least 1 h. Population growth was analyzed by grouping dates into "slices" of time one-third generation (106 DD) wide. Log catch rates were averaged within slices, as were matching air temperatures and precipitation rates. We fit mixed regression models with lme in R to analyze growth between successively paired mothers (g) and daughters (g+3). Dates when adults were first trapped ranged from 4 April (1986) to 25 May (1995), and averaged 25 April (SD = 14 d). Initial numbers were always low. Average - from 1 Jan to first date was 98.4 (SD = 56). Frequency of days with southerly wind events increased from none 2 or more weeks before first date to ~ 40% during week of first date. Mean no. southerly wind events up to first date was 3.6 (SD = 2.6). First dates were mildly correlated with dates when DDs reached 98.4, but were more strongly correlated with interpolated dates when no. wind events reached 3.6. By both measures, appearance in 1991 was exceptionally late. Evidence for southern immigration was stronger than for local overwintering. Hierarchical mixed models indicated population growth depended negatively on density of mothers, positively on rainfall when daughters were pupae, positively on temperatures when daughters were pupae, and negatively on temperatures when daughters were nulliparous females. Temperature and rainfall experienced by mothers and daughters were otherwise uninformative. Trends in growth rates also varied significantly among years. The latter may have reflected variation in supply and quality of larval breeding media, which was not assessed during this study. <BR> PARTICIPANTS: One graduate student, K. Borewicz, completed her Master of Science degree on this project. Numerous undergraduate students and veterinary students helped conduct mosquito surveillance and pcr blood meal identifications. <BR>TARGET AUDIENCES: Public health authorities, horse owners. Results are being conveyed through publications and horse owner meeetings. <BR>PROJECT MODIFICATIONS: No major changes were made.
<p>
IMPACT: 2007/01 TO 2007/12<BR>
To extent results from Ames can be generalized regionally, livestock managers can expect stable flies to be continuously present from date of appearance to late autumn every year. Date of first appearance can be predicted to occur when southerly wind events exceed three. Densities thereafter will rise most quickly and remain high when spring and summer weather is relatively warm and wet.