Spread Of Musca De Granjas Para Cidades

SAMPLING, DISTRIBUTION, DISPERSAL

Spread of Musca domestica (Diptera: Muscidae), from Two Caged Layer Facilities to Neighboring Residences in Rural Ohio KIM A. WINPISINGER,1 AMY K. FERKETICH,2 RICHARD L. BERRY,1 2 AND MELVIN L. MOESCHBERGER

J. Med. Entomol. 42(5): 732Ð738 (2005)

ABSTRACT Complaints of excessive numbers of ßies were reported by citizens living in a rural area surrounding a large (⬎2 million laying hens) egg-layer facility in northwestern Ohio. Sticky cylinder traps and hanging ßy strips were used at outdoor and indoor locations, respectively, at known distances from the layer farm and from control sites to determine the most likely source of the ßies and to determine the severity of the problem compared with ßy populations in nearby rural settings. House ßies, Musca domestica (L.), were the predominant ßies captured on ßy traps located within 6.4 km of the poultry operations. There was a signiÞcantly greater number of M. domestica trapped in the study area surrounding the layer facility than in the control areas. The quantity of house ßies captured decreased with increased distance from the layer farm. Two years into the study, a second egg-layer facility opened in an area that was originally a control site. With regard to this second farm, after 4 yr of study, there was a signiÞcant difference shown between the population of house ßies during the 2-yr control phase and the 2-yr period when the egg-layer facility was operational. This study documented that large egg layer facilities can signiÞcantly increase the house ßy population in the surrounding community up to 6.4 km from the source of the ßies and may result in a severe nuisance up to 3.2 km away. KEY WORDS house ßy, Musca domestica, poultry, egg, Ohio

IN SEPTEMBER 1995, A NEW commercial egg-layer facility housing 2.5 million chickens opened in Hardin County, Ohio, between the towns of LaRue and Mt. Victory. In early March 1996, local, county, and state health departments as well as the Ohio Environmental Protection Agency began receiving ßy nuisance complaints from citizens living and working near this facility. A nuisance with respect to ßies is a condition where citizens cannot enjoy the normal activities of living because of unusual numbers of ßies interfering with those activities by landing on their person, on food, and infesting their vehicles, yards, and buildings. The complaints alleged that the ßies were coming from the new egg-layer facility. Preliminary observations at residentsÕ homes and identiÞcation of submitted specimens indicated that house ßies, Musca domestica (L.), were causing the nuisance. Very few studies of house ßy dispersal and populations have been done in the United States, and much of the existing research is ⬎40 yr old. House ßy populations have been monitored inside poultry houses by using baited jug-traps and spot cards (Burg and Axtell 1984, Lysyk and Axtell 1985). There also have been studies of ßies marked with ßorescent dye and re1 Vector-Borne Disease Program, Ohio Department of Health, 900 Freeway Dr., Columbus, OH 43229. 2 School of Public Health, Division of Epidemiology and Biostatistics, The Ohio State University, Columbus, OH 43210.

leased to determine the distribution and movement between breeding sites, such as poultry houses and dairies (Lysyk and Axtell 1986). The greatest populations of ßies occur in the commercial caged layer houses used for producing eggs (Axtell and Arends 1990). This is a type of concentrated animal feeding operation (CAFO). The objective of this investigation was to determine whether ßy populations at residences close to poultry operations were greater than ßy populations at residences farther from poultry operations. A second objective of this study was to compare the ßy populations at a poultry operation site before the site opened to the ßy populations after the site was fully operational. A ßy population monitoring program was implemented around two egg farms to examine these objectives. Materials and Methods Study Design. The study was conducted from 1997 to 2000. Each year, the study ran for 18 wk, beginning the Þrst week of May and ending in early September. In 1997, 16 farm/residential sites were monitored surrounding the LaRue/Mt. Victory egg farm. Four sites were chosen in each compass direction (northeast, northwest, southeast, and southwest) from the egg farm at distances of ⬇0.8 km (0.5 mile), 1.6 km (1 mile), 3.2 km (2 miles), and 6.4 km (4 miles) from the facility. Two additional sites, hereafter referred to

0022-2585/05/0732Ð0738$04.00/0 䉷 2005 Entomological Society of America

September 2005

WINPISINGER ET AL.: SPREAD OF HOUSE FLIES

733

Table 1. Mean number of house flies per trap per week, standard deviation, sample size, and median on cylinder traps and hanging fly strips at various distances in Larue/Mt. Victory and in the control group, 1997–2000 Trap type Cylinder trap 1997 nb Mean (SD) Median 1998 n Mean (SD) Median 1999 n Mean (SD) Median 2000 n Mean (SD) Median Hanging strip 1997 n Mean (SD) Median 1998 n Mean (SD) Median 1999 n Mean (SD) Median 2000 n Mean (SD) Median

Egg farm

0.8 km

1.6 km

3.2 km

6.4 km

Controla

2 28.5 (28.2) 28.5 4 120.5 (42.2) 120.5c 4 382.6 (261.8) 364.9c 4 180.2 (107.3) 184.7c

4 7.9 (3.5) 7.7c 3 27.3 (16.6) 22.5c 4 40.8 (23.9) 32.7c 4 11.8 (5.1) 11.8c

4 2.7 (1.6) 2.5c 4 10.6 (5.7) 11.2c 4 13.2 (10.3) 12.8c 4 4.3 (4.1) 3.3c

4 0.7 (0.4) 0.8 4 1.6 (0.7) 1.6c 4 1.3 (0.8) 1.4c 4 0.5 (0.4) 0.5c

4 1.3 (1.3) 1.0 4 3.1 (3.5) 2.3 4 1.6 (2.9) 0.3 4 0.4 (0.4) 0.4

8 0.6 (0.5) 0.4 22 0.3 (0.4) 0.1 16 0.1 (0.1) 0.1 16 0.1 (0.1) 0.0

0 NA NA 0 NA NA 0 NA NA 0 NA NA

4 74.2 (140.1) 6.3 3 92.3 (163.8) 13.2c 4 142.8 (153) 92.1c 4 69.1 (102.8) 20.1c

4 18.5 (6.6) 17.8c 4 29.9 (29.3) 24.5c 4 22.0 (24.4) 10.9c 4 15.7 (15.0) 10.7c

4 6.3 (10.0) 1.1 4 6.6 (10.1) 0.9 4 13.4 (18.8) 6.0 4 8.5 (10.7) 4.9

4 9.6 (12.6) 5.7 4 3.7 (4.6) 2.1 4 10.3 (11.9) 7.3 4 8.8 (14.7) 2.1

8 4.1 (5.8) 0.4 20 1.2 (3.0) 0.2 16 0.9 (1.6) 0.3 16 1.1 (2.0) 0.2

NA, not applicable. Control sites in 1997 and 1998, Marseilles and Jackson; control site in 1999 and 2000, Jackson. n is number of sample sites. c SigniÞcantly different from control group (P ⬍ 0.05). a

b

as farm trap sites, were located at the egg farm at the LaRue/Mt. Victory location on the northeastern and northwestern ends of the property. Additional farm/ residential sites were monitored surrounding two proposed egg farm locations in Wyandot County. One of these locations was in Jackson Township, and the other was near the town of Marseilles, where an egg farm was subsequently built and made operational after the 1998 study ended. In 1997, each of these additional sites was ⬇0.8 km in each compass direction (as above) from the location of the proposed facilities. Tables 1 and 2 present the number of locations participating in each study year. All sites, except the egg farm property, were residential or farm/residential properties. Other small area farms, in both the study area and the control area(s), had livestock, including small numbers of poultry, up to 100 sheep, 50 cattle, several horses, and some smaller animals. The study areas were similar in that most of the acreage is devoted to ßat land crop production for corn, wheat, oats, and soybean. There are also small, scattered woodlots, ponds, and streams. In all years, participating locations were chosen by distances and compass directions as described above. Additional sites were added in successive years. Participating locations were chosen by health department personnel based on their location and the willingness of the residents to participate. Fly Traps. Each year, two Olson Biting Fly Traps (Olson Products Inc., Medina, OH) were placed outdoors at each site. These traps are large and cylindrical

and have been used successfully in other studies of nuisance ßies (Broce 1988). These traps consist of an alsynite plastic cylinder, ⬇30 cm in height and 20 cm in diameter and are covered with a removable propylene sleeve with a sticky surface on the outer side. The cylinder is mounted on a slotted wooden stake ⬇60 cm in length. For purposes of later data analysis, cylinder traps at each residence were arbitrarily labeled cylinder #1 and cylinder #2. Each year, these traps were located in the same place for the duration of the study, i.e., cylinder #1 was always in its respective place and cylinder #2 in its respective place, and so on. In 1997, two clear propylene sleeves were used. These sleeves are labeled as attractive to stable ßies, Stomoxys calcitrans (L.), but we used them to see what type of ßies were in the environment. In 1998 only, a third outdoor trap, cylinder #3, was added with a white sleeve. In 1999 and 2000, only cylinders #1 and #2 with white sleeves were used. At residences, cylinder traps were placed on the side of the yard that faced the direction of the egg farm or proposed egg farm. The traps were placed in such a manner as to be convenient for the homeowner to cut the lawn, cultivate the garden, trim bushes, and so on. At the egg farm itself, cylinder traps were placed at the four corners of each farm. During the 1997 study, Olson Stiky Traps for Windows (Olson Products Inc.) were used in participantsÕ kitchens. Nine weeks into the 1997 study, TAT ßy strips (Walco-Linck Co., Valley Cottage, NY), commonly used in homes, stables, and barns, also were

734

JOURNAL OF MEDICAL ENTOMOLOGY

Vol. 42, no. 5

Table 2. Mean number of house flies per trap per week, standard deviation, sample size, and median on cylinder traps and hanging fly strips at various distances in Marseilles and in the control group, 1997–2000 Trap type Cylinder trap 1997 nb Mean (SD) Median 1998 n Mean (SD) Median 1999 n Mean (SD) Median 2000 n Mean (SD) Median Hanging strip 1997 n Mean (SD) Median 1998 n Mean (SD) Median 1999 n Mean (SD) Median 2000 n Mean (SD) Median

Egg farm

0.8 km

1.6 km

3.2 km

6.4 km

Controla

0 NA NA 2 0.1 (0.0) 0.1 4 236.3 (151.0) 209.2c 4 354.7 (357.6) 348.3c

4 0.7 (0.7) 0.6 4 0.5 (0.6) 0.3 4 15.1 (8.6) 16.0c 4 9.1 (7.9) 8.0c

0 NA NA 4 0.4 (0.6) 0.1 4 4.5 (4.5) 3.4c 4 1.0 (0.7) 0.9c

0 NA NA 4 0.1 (0.1) 0.0 4 0.4 (0.6) 0.1 4 0.8 (0.8) 0.6c

0 NA NA 4 0.6 (0.6) 0.6 4 0.9 (1.1) 1.0 4 1.3 (1.1) 1.2c

NA NA NA NA NA NA 16 0.1 (0.1) 0.1 16 0.1 (0.1) 0.0

0 NA NA 0 NA NA 0 NA NA 0 NA NA

4 3.4 (6.8) 0 4 0.1 (0.3) 0.0 4 24.0 (25.8) 19.9 4 10.8 (11.9) 7.8

0 NA NA 4 0.2 (0.3) 0.2 4 17.3 (15.0) 12.6c 4 8.8 (7.5) 6.2c

0 NA NA 4 1.1 (1.5) 0.5 4 12.5 (4.4) 12.4c 4 14.0 (6.9) 15.3c

0 NA NA 4 0.1 (0.1) 0.1 4 2.7 (1.9) 3.1 4 3.6 (3.6) 3.6

NA NA NA NA NA NA 16 0.9 (1.6) 0.3 16 1.1 (2.1) 0.2

NA, not applicable. a Marseilles was part of the control group in 1997Ð1998. b n is number of sample sites. c SigniÞcantly different than control group (P ⬍ 0.05).

given to participants to hang inside their kitchens. In subsequent years, Victor ßy strips (Woodstream Corporation, Lititz, PA) were used indoors during all weeks of the study. Both brands were typical hanging ribbon-type strips, and the brand was only changed due to a change in vendors. A single hanging ßy strip was provided to each participant during each week of the study. Fly traps and ßy strips from participantsÕ homes were collected and replaced the same time each week for the duration of each study year. Study participants collected the ßy strip from inside their home, and placed it in a gallon-sized Ziploc bag, which was then placed in a tote bag and hung on an outside door knob on the designated collection day each week. Captured ßies were identiÞed and counted on each trap by employees of the Ohio Department of HealthÕs Vector-Borne Disease Program who were trained in ßy identiÞcation. Data Analyses. Data were analyzed using True EpiStat (Gustafson 1987), SAS (SAS Institute 1999), and EpiInfo (CDC 1997, 2000) software packages. We explored the possibility of using parametric data analysis early in the study. Each year, the data were highly skewed, and we were not able to Þnd a transformation that could be used to normalize the residuals or meet the equality of variances assumption. When these assumptions are violated, the F-tests, P values, and conÞdence intervals are not accurate (Dean and Voss 1999). Therefore, nonparametric methods were used to analyze the data.

For each year of the study, the total number of house ßies per residence was divided by the number of traps used at that residence to get a “seasonal average.” Because data were not normally distributed, the median seasonal averages were compared between groups (locations). This analysis used the Wilcoxon Rank Sum test (Hollander and Wolfe 1999) with BonferroniÕs adjustment for multiple comparisons (Madsen and Moeschberger 1986). These tests were done for both cylinder traps and hanging ßy strips. Differences between the distance groups within a given study area and the control group were determined using an exact Wilcoxon Rank Sum comparing the medians of each distance group to the entire control group. The JonckheereÐTerpstra test was used to compare average weekly house ßy count differences within each of the groups by distance from the nearest egg farm. When the data follow a natural order, such as increasing distance from the egg farm, this test is more powerful than the KruskalÐWallis test (Hollander and Wolfe 1999). In both 1997 and 1998, the Marseilles and Jackson control sites were grouped together for analysis due to the small number of participants. This grouping was done for both cylinder traps and hanging ßy strips. However, not all participants hung a ßy strip every week of the study, so averages were computed for each household. They were deÞned as the total number of house ßies on all strips divided by the number of weeks of observation. Therefore, if a house-

48

362

Marseilles

2.0

4.0

0

Control

0.5 1.0

2.0

4.0

Distance (miles)

Distance (miles)

LaRue / Mt. Victory

Marseilles

Control

0

60 40 0

20

Flies per strip per week

80 60 40 20

80

100

0.5 1.0

100

0

Flies per strip per week

735

0

24

48

Flies per cylinder per week

362

LaRue / Mt. Victory

0

Flies per cylinder per week

WINPISINGER ET AL.: SPREAD OF HOUSE FLIES

24

September 2005

0.5 1.0

2.0

4.0

Control

Distance (miles)

0.5 1.0

2.0

4.0

Control

Distance (miles)

Fig. 1. Plot of median weekly house ßy counts by method of capture (strips vs. cylinders), location (LaRue/Mt. Victory vs. Marseilles vs. control), year, and distance within location. (F, 1997; Œ, 1998; }, 1999;
, 2000.)

hold had one missing week, the denominator was reduced by 1. The median ßy counts in Marseilles in 1997 and 1998, before the egg farm existed, were compared with the counts in 1999, after the Marseilles egg farm was established, by using the Wilcoxon sign rank test. Similarly, the results from 1999 were compared with the 2000 results. Results Fly Traps. The cylinder traps captured a greater average number of house ßies than did the hanging ßy strips. In 1997, 1999, and 2000, the average of the two cylinder traps at each residence was generated. However, in 1998, the average of all three cylinder traps was used. The total number of house ßies captured on all outdoor cylinder traps ranged from a low of 4,011 in 1997 to a high of 100,167 in 1999. During the 4-yr study, ⬎38,257 house ßies were captured on indoor hanging ßy strips in participantsÕ homes. In 1997 during the 9 wk that the indoor hanging ßy strips were used, they caught 3,138 house ßies at the LaRue/Mt. Victory sites compared with 96 in Jackson and 95 in Marseilles. The Stiky Window Traps used in 1997 proved to be a poor collector of house ßies. These traps caught a total of just 252 house ßies in 18 wk: 242 in LaRue/Mt. Victory, eight in Jackson, and two in Marseilles. Use of these

traps was discontinued after the Þrst year of the study, because these traps were deemed an ineffective method of capturing house ßies. Distance Groups in LaRue/Mt. Victory. In 1997, cylinder traps at 0.8 and 1.6 km from the LaRue/ Mt. Victory egg farm had a signiÞcantly higher median number of house ßies per trap compared with the control group. In 1998 through 2000, cylinder traps at sites on and 0.8, 1.6, and 3.2 km from the LaRue/ Mt. Victory egg farm had a signiÞcantly higher median number of house ßies per trap compared with the control group (Table 1). Although hanging ßy strips showed one difference between 1.6 km and the control groups in 1997, there were signiÞcantly higher ßy counts per hanging ßy strip from 1998 through 2000 at sites 0.8 and 1.6 km from the LaRue/Mt. Victory egg farm compared with the control group (Table 1). Tests comparing cylinder trap captures at various distances within the LaRue/Mt.Victory group showed signiÞcant differences within the distance groups in all 4 yr of the study (P ⬍ 0.01) (Fig. 1). When comparing these distances for the hanging ßy strips, there was no decrease in the average number of ßies caught with increasing distance from the egg farm in 1997 (P ⫽ 0.33) or 1998 (P ⫽ 0.09). In 1999 and 2000, there was a signiÞcant decrease in the average number of house ßies per hanging ßy strip with increasing distance from the egg farm (P ⬍ 0.01 and P ⫽ 0.04, respectively).

736

JOURNAL OF MEDICAL ENTOMOLOGY

Distance Groups in Marseilles. The data for the cylinder traps and hanging strips for the different distance groups in Marseilles are presented in Table 2. In 1999 and 2000, there were signiÞcantly more house ßies on the cylinder traps at the egg farm at Marseilles, and at sites 0.8 and 1.6 km from the egg farm at Marseilles compared with the Jackson control group. In 2000, this difference also was noted at sites located 3.2 and 6.4 km from the egg farm at Marseilles. Hanging ßy strip data at sites 1.6 and 3.2 km from the egg farm at Marseilles had a signiÞcantly higher median number of house ßies per hanging strip compared with the Jackson control site in 1999 and 2000. Sites 0.8 and 6.4 km from the egg farm at Marseilles were not statistically different from the Jackson group. With regard to cylinder traps, there was a difference within the distance groups in the Marseilles area in 1999 and 2000, but there was no difference in 1998 (Fig. 1). For hanging strips, in 1999 there was a signiÞcant decrease in the average number of house ßies caught on the hanging strips as the distance from the egg farm increased. However, in 1998 and 2000, the average number of ßies per hanging strip did not vary within the distance groups surrounding the Marseilles egg farm. Marseilles: Comparisons of 1997–2000. In both 1997 and 1998 when Marseilles was a control site, house ßies per cylinder trap averaged less than one house ßy per trap per week, and house ßies per hanging ßy strip averaged less than four house ßies per trap per week. The median number of house ßies per trap increased signiÞcantly in Marseilles in 1999 compared with 1998, the Þrst year with full participation at each distance group in the study. This was true for both the cylinder traps and hanging ßy strips (P ⬍ 0.01). The Þrst year that the Marseilles egg farm was operational was in 1999. There was no difference in the median number of house ßies per trap from 1999 to 2000 for both the cylinder traps as well as the hanging ßy strips at Marseilles. Discussion Two Þndings stand out in this study: 1) house ßy counts closer to a CAFO egg farm were signiÞcantly higher compared with counts in similar rural areas further from an egg farm; and 2) the building of a new CAFO egg farm was associated with a signiÞcant, dramatic increase in the number of house ßies in the area. Our arrangement of ßy traps was designed to determine the source of the majority of ßies. Our hypothesisÑthat ßies from a point source disperse radiallyÑcan be proven if catch rates diminish by increasing distance from that source. If there were another source of ßies within the study area, the distribution of ßies in the concentric circles surrounding the egg farms would have shown greater numbers away from the center. This did not happen, strongly suggesting that the egg farm was the principle source. Although ßies were certainly coming from other farms with livestock and poultry, their contribution was negligible in this study. There was no other source of

Vol. 42, no. 5

house ßy breeding in any of the three study communities even close to the magnitude of the egg farms, and there was little difference in the number of farm animals between the three study communities. Although some participants in the control areas had 15Ð50 large farm animals, including cows, sheep, and horses, the number of house ßies captured on these farms was signiÞcantly less than the LaRue/Mt. Victory participantsÕ sites with no farm animals. Outdoor cylinder traps provided much more reliable information than the indoor hanging ßy strips because we knew exactly when and where the traps were placed and collected. There was no chemical attractant on these traps, so the ßies that were collected were those that came close enough to be visually attracted to them. These traps still caught ßies after rainy and windy weather. The only drawback to the cylinder traps was that they were affected by dust and vegetation that grew near them. Dust particles made the traps nonsticky, and weeds blocked insectsÕ access to the traps, both leading to an underestimate of the ßy population. Beginning in 1998, weeds were cut weekly in a circle ⬇8 feet (2.4 m) in diameter around each cylinder to help prevent underestimates. The Þrst year of the study, nonspecies-speciÞc clear cylinder traps were used to determine which types of ßies were causing the complaints. We compared average counts of house ßies, stable ßies, horse ßies (Tabanidae), deer ßies (Tabanidae: Chrysops spp.), ßesh ßies (Sarcophagidae), and bottle ßies (Calliphoridae) on cylinder traps between each of the three study locations. In subsequent years, these species were noted but not counted. Numerous other ßies captured on the cylinder traps, such as Anthomyidae, Syrphidae, Anispopdidae, and Stratiomyidae, were not considered pertinent to the study. Average counts of only one species (M. domestica), differed signiÞcantly between the three study groups (P ⫽ 0.04), and those counts were signiÞcantly higher in the LaRue/ Mt. Victory group than in either the Marseilles or Jackson group, conÞrming that house ßies were indeed the ßies that were prompting complaints (K.A.W. and R.L.B., unpublished data). In an attempt to get a more accurate estimate of the population, in 1998 we began using white cylinder traps, marketed as attractive to house ßies. In 1998, the only year that both white and clear cylinder traps were used, the correlation between them was strong, P ⬍ 0.01 (K.A.W. and R.L.B., unpublished data). In the LaRue/Mt. Victory area that year, the white traps caught more house ßies than did the clear traps. Although we did not use clear traps in any subsequent year, it is possible that the greater ability of the white cylinder traps to capture house ßies contributed to higher numbers of house ßies on these traps from 1998 to 2000. We found the white cylinder traps to be an especially useful tool in capturing house ßies outdoors. Hanging ßy strips were the trap of choice indoors and were useful in determining the principal pest ßy species. The Stiky Window traps easily caught many gnat-like insects, but house ßies often escaped these

September 2005

WINPISINGER ET AL.: SPREAD OF HOUSE FLIES

traps, thus their counts reßected an artiÞcially low number. The hanging ßy strips captured house ßies well and participants who had a ßy problem did not mind using them. Some of the participants with ßy problems used multiple ßy strips in their homes, but only the ßy strip provided weekly for the study was collected and counted. Because more than one ßy strip was sometimes used in homes with ßy problems, but only one was counted, the number of house ßies in some of noncontrol homes was likely underestimated. There were limitations to using hanging ßy strips, and the number of ßies captured on them was subject to variables related to human activity. These variables included the number of people entering and exiting a home, use of air conditioners, amount of time away from home, and how ßytight the home was considered. Also, because researchers did not enter participantsÕ homes, the participants had control over where and how long the strips were hung, within the week that they had possession of each strip. The analysis of Marseilles comparing 1998 and 1999 data provided an internal control to determine whether there was a true association between the number of house ßies on a farm/residential property and whether there is an egg farm nearby. The significant increase in house ßy numbers in Marseilles in 1999 was consistent with the addition of ⬎3 million chickens, and their manure, to the egg farm at Marseilles after the completion of the 1998 study. The same analysis for 1999 versus 2000 showed no difference in house ßy counts, indicating that the house ßy population had reached a statistical equilibrium at a higher level. During the last 2 yr of the study, this analysis also showed that within the Marseilles group, the farther the study site was from the egg farm at Marseilles, the fewer ßies were captured on cylinder traps. During both these years, the chicken population was wellestablished and the ßy counts were high. The same was true in LaRue/Mt. Victory in all 4 yr of the study. Jackson, a control group for the duration of the study, had consistently low house ßy counts all 4 yr. With a larger sample size, the differences both within and between study groups would likely have been even more pronounced. In both study areas with an established egg farm, the average number of house ßies on cylinder traps decreased signiÞcantly as distance from the respective egg farm increased. With regard to judging nuisance house ßy populations related to poultry CAFOs in other areas, the average number of house ßies captured weekly in the control areas, as presented in Table 1, could be used as a baseline. In our study, the upper limit would be 1 for cylinder traps and 5 for hanging ßy strips. These numbers represent the highest control averages over the 4-yr study, rounded to the next whole number. These Þgures did not result in ßy complaints from study participants and could be considered normal rural “background” ßy levels. Averages signiÞcantly above these would be indicative of a ßy nuisance, needing remediation. The best way to investigate complaints of a house ßy problem may come from

737

using white cylinder traps outside homes and hanging ßy strips inside homes weekly. Professional judgment is needed in the case of a potential nuisance, especially if some legal action is anticipated, taking into account such factors as 1) tolerance of the local citizens for house ßies in their environment; 2) proximity to villages, cities, and urban areas; 3) professional inspections of complainant homes; 4) number of complainants and over what time period; 5) speciÞc identiÞcation of the nuisance ßies as house ßies; 6) other possible sources of the house ßies; 7) time of year; 8) number of samples taken and over what time period; and 9) existence of a properly executed house ßy control plan at the CAFO. For most local health departments that often receive the complaints, gathering and interpreting this information is a daunting task. The problem may become chronic. Outside professional help may be needed, including state health and environmental agencies and cooperative extension services at land grant universities. Conditions remain ideal for ßy breeding in commercial caged layer houses even during the cold months of the year. During these months, ventilation fans in the egg farm houses are used less, resulting in a buildup of moisture in the manure (Merchant et al. 1987). The manure generates heat, contributing to a suitable environment for house ßy breeding that can lead to an outbreak of ßies in the community during months when house ßies are not expected to be a nuisance, e.g., during a period of unseasonably warm weather. We draw two principle conclusions from this study. First, the closer a property is to a large CAFO housing millions of chickens, the greater the number of house ßies found there. Second, a tremendous increase in house ßy populations in rural areas can occur after a large poultry CAFO is established. This study has shown that large egg-laying facilities can signiÞcantly increase the house ßy population and thus the nuisance potential in the surrounding community. We found that a signiÞcant difference in house ßy levels can occur up to 3.2 km from the source of the ßies. Given the results of this study, we would recommend that a poultry CAFO not be built within 3.2 km of a residential area. There are some factors that may alter this setback distance, such as: the type and density of homes in the area, area topography as a natural barrier, and the quality of the manure management at the CAFO.

Acknowledgments We thank Jeff Ritchey and Karla Seigle (Wyandot County Health Department)and Michael Collart for assistance with collecting and setting ßy traps each week. We also thank Robert Restifo, Richard Gary, Chris Coudret, W. Adam Stout, Shelly Parker, Crystal Forrider, Kevin Klinect, Kayla Stewart, and Candace Sack for doing the same as well as identifying the ßies on the traps. Additional thanks go to Richard Gary and Chad Schone for assistance with editing this manuscript.

738

JOURNAL OF MEDICAL ENTOMOLOGY References Cited

Axtell, R. C., and J. J. Arends. 1990. Ecology and management of arthropod pests of poultry. Annu. Rev. Entomol. 35: 101Ð126. Broce, A. B. 1988. An improved alsynite trap for stable ßies, Stomoxys calcitrans (Diptera: Muscidae). J. Med. Entomol. 25: 406Ð409. Burg, J. G., and R. C. Axtell. 1984. Monitoring house ßy, Musca domestica (Diptera: Muscidae), populations in caged-layer poultry houses using a baited jug-trap. Environ. Entomol. 13: 1083Ð1090. [CDC] Centers for Diseases Control and Prevention. 1997. Epi Info 6.04 software. Centers for Disease Control and Prevention. Atlanta, GA. [CDC] Centers for Diseases Control and Prevention. 2000. Epi Info 2000 software. Centers for Disease Control and Prevention. Atlanta, GA. Dean, A., and Voss, D. 1999. Design and analysis of experiments. Springer, New York. Gustafson, T. L. 1987. True EpiStat software version 5.3. Epistat Services, Richardson, TX.

Vol. 42, no. 5

Hollander, M., and D. A. Wolfe. 1999. Nonparametric statistical methods, 2nd ed. Wiley, New York. Lysyk, T. J., and R. C. Axtell. 1985. Comparison of baited jug-trap and spot cards for sampling house ßy, Musca domestica (Diptera: Muscidae), populations in poultry houses. Environ. Entomol. 14: 815Ð 819. Lysyk, T. J., and R. C. Axtell. 1986. Movement and distribution of house ßies (Diptera: Muscidae) between habitats in two livestock farms. J. Econ. Entomol. 79: 993Ð998. Madsen, R. W., and M. L. Moeschberger. 1986. Statistical concepts. Prentice-Hall, Englewood Cliffs, NJ. Merchant, M. E., R. V. Flanders, and R. E. Williams. 1987. Seasonal abundance and parasitism of house ßy (Diptera: Muscidae) pupae in enclosed, shallow-pit poultry houses in Indiana. Environ. Entomol. 16: 716 Ð721. SAS Institute. 1999 –2001. The SAS system for Windows software. SAS Institute, Cary, NC. Received 13 August 2004; accepted 5 May 2005.