Materials and Methods

Oral rabies vaccine—Vaccinia-rabies glycoprotein recombinant virus vaccine^a was used to orally vaccinate raccoons. The bait was available in 2 formats: an FP and a CS. For both, 1.5 to 2.0 mL of vaccine (a suspension with a concentration of 10^8.0 TCID₅₀/dose) was contained in a plastic sachet. Each FP bait contained 150 mg of tetracycline as a time-specific hard tissue biomarker (marking occurs when bone and teeth undergo mineralization), whereas the CS contained no tetracycline.

Target area—Anne Arundel County is on Maryland's western shore of the Chesapeake Bay and has several peninsulas that project into the Bay (Figure 1). Physiographically, it lies in the Maryland coastal plain.⁸ The topography varies from flat to sharply rolling, with most of the land in gentle slope. The county is laced with many inlets and streams. The county consists of 1,078 km² with a 2008 estimated human population of 512,790,^c or approximately 476 people/km². A variety of wild terrestrial mammals, including raccoons, gray (Urocyon cinereoargenteus) and red (Vulpes vulpes) foxes, skunks (Mephitis mephitis), opossums (Didelphis virginianus), several species of squirrels, Eastern cottontails (Sylvilagus floridanus), white-tailed deer (Odocoileus virginianus), and many rodents, including groundhogs (Marmota monax), inhabit the county. The county is connected to Maryland's Eastern Shore by 2 parallel bridges, each approximately 5 miles in length. Four other counties and the City of Baltimore border Anne Arundel County.

View Full Size

Initiation of ORV campaigns by year and areas targeted in Anne Arundel County, Md, 1998 through 2007. Bait distribution occurred throughout the county from 2003 through 2007 at a rate of 75 baits/km².

Citation: Journal of the American Veterinary Medical Association 241, 6; 10.2460/javma.241.6.725

• Download Figure
• Download figure as PowerPoint slide

Bait distribution—In 1998, 1999, and 2000, FP baits were placed on the Annapolis Peninsula during October at a rate of 100/km². In October 2000, Gibson Island was baited for the first time at a rate of 100 FP baits/km². In September 2001, baiting continued on the Annapolis Peninsula and Gibson Island at 100 FP baits/km², and Broadneck Peninsula was included for the first time at 75 FP baits/km². In September 2002, FP baits were placed on Annapolis Peninsula and Gibson Island at 100 baits/km² and continued on Broadneck Peninsula at 75 baits/km². Starting in 2003 and continuing through 2004, FP baits were distributed throughout the entire county at 75 baits/km² during August. In 2005, CS baits were used for the first time and were distributed in the southwestern part of the county during September at 75 baits/km². Fish meal polymer baits were placed in the remainder of the county during August and September at 75 baits/km². During September 2006 and 2007, the entire county was baited at 75 FP baits/km², but the southwestern county area again received CS baits at that same rate. The bait density of 75 baits/km² in 2003 and subsequent years was chosen because of the results on Broadneck Peninsula and because other similar ORV projects were using this bait density.^9,10

The bait was shipped to Anne Arundel County in a refrigerated truck, and refrigeration was maintained prior to distribution. On the day of ground distribution, the bait was kept in a chest cooler until distributed. For aircraft distribution, bait was kept in a shaded area until used.

In heavily populated residential areas, ground teams hand distributed baits. Heavily populated residential areas were landscapes where it was subjectively determined that the density of homes did not allow the bait to be safely dropped by aircraft. Team members threw baits from a slow-moving vehicle, or they parked and walked to prime habitat areas to release the bait. In less heavily populated residential areas and rural parts of the county, helicopter and fixed-wing aircraft teams were used. When the fixed-wing aircraft was used to drop the CS baits, the flight line spacing distance was 0.5 km. With either method, efforts were made to avoid placing bait in areas such as lawns, roads, driveways, waterways, pastures, and hayfields. Digital orthophoto maps of the county were reviewed to identify areas where only ground baiting could be accomplished because of the high concentration of residential developments. Each digital orthophoto map included approximately 575 acres subdivided into 25 equal grids and detailed streets and roads, buildings, waterways, and vegetation. Ground teams were provided individual digital orthophoto maps with probable habitat areas identified for ground baiting. The total number of baits intended for each area was indicated on each map. Helicopter drop zones were also identified on each map. Teams were asked to place an x in the approximate location that each bait was placed to maximize the likelihood that the targeted bait density would be achieved and to assist in estimating actual bait density.

Campaign evaluation—No serosurvey or raccoon density study was conducted on the Annapolis Peninsula before the initiation of the first campaign in 1998. Pre-ORV serosurveys were accomplished, however, on Gibson Island (2000), on Broadneck Peninsula (1999 and 2001), and in the southern part of Anne Arundel County (2003). These serosurveys did not follow a specific protocol in terms of required sample size, trap numbers, and placement or number of trap nights. The trapper had considerable latitude in selecting a trapping schedule, specific trap locations, bait attractants, and number of traps placed. Raccoons were trapped via collapsible live traps,^d and canned cat food was usually the attractant. Trapping locations were determined primarily by sites where permission could be obtained. In 2003, a raccoon density study was conducted in the northwestern part of the county according to a USDA protocol.^e

Following each bait distribution campaign during 1998 through 2007, free-ranging raccoons were trapped at various sites, and blood and tooth samples were collected. As with the pre-ORV serosurveys, no specific protocol was followed, but trapping locations were standardized in 2005 through 2007 to assist in making among-years comparisons for each site. The interval between bait distribution and trapping was generally 4 weeks (range, 2 to 6 weeks). The raccoon trapping lasted from 4 to 9 weeks.

Captured raccoons were sedated with a 5:1 mixture of ketamine^f (100 mg/mL) and xylazine^g (100 mg/mL); 0.1 mL of the mixture/kg (0.045 mL/lb) was administered IM (ie, 8.3 mg of ketamine/kg [3.8 mg/lb] and 1.7 mg of xylazine/kg [0.8 mg/lb]) on the basis of estimated body weight. Once sedated, raccoons were examined, and the sex, approximate age, weight, location and other data were recorded. An ear tag^h was placed in the medial aspect of the ear. During 1998 through 2000, only 1 ear tag was used, but from 2001 through 2007, an ear tag was placed in both ears to increase the likelihood that at least 1 ear tag would be present upon recapture. Five to 7 mL of blood was collected from the jugular or femoral vein from each raccoon to determine the presence of RVNA. Beginning in 2001, tooth samples were collected to evaluate the efficacy of bait placement. A tooth, usually the first premolar, was analyzed for tetracycline biomarking, a surrogate measure of bait acceptance. Beginning in 2003, captured raccoons were vaccinated IM with an inactivated rabies vaccine.ⁱ Each raccoon was allowed to recover and was released near the site of capture. If raccoons were recaptured within 30 days after the first capture, no additional workup was completed.

Notable efforts were made during the initial campaign (1998) to obtain raccoons that were found dead along roads to determine their rabies status and to collect the mandible for tetracycline analysis. However, very few samples were obtained (n = 3). Attempts during ensuing years (1999 and 2000) to acquire road-killed animals and other specimens were also not productive. Consequently, surveillance for terrestrial wildlife rabies was dependent on the results of specimens routinely submitted by Anne Arundel County residents, typically in situations where there was human or domestic animal exposure to a potentially rabid wild animal. The number of rabies cases reported in surrounding counties and the state of Maryland was compared with the number of rabies cases in Anne Arundel County.

Sample processing—Serum was separated and removed from the collected blood sample and stored at ≤ 0°C until tested for RVNAs. For the first 3 campaigns (1998 through 2000), the Laboratories Administration used the RFFIT¹¹ to determine the presence or absence of RVNAs. During subsequent campaigns, the frozen sera were forwarded to the CDC for testing by the RFFIT. The positive cutoff titer for samples tested was ≥ 1:32 for the Maryland Department of Health and Mental Hygiene's results and ≥ 1:5 for the CDC's results. Starting in 2004, the CDC reported results in international units, and a positive cutoff of ≥ 0.05 IU was used. The Maryland Department of Health and Mental Hygiene used the direct fluorescent antibody test¹² to determine the presence of rabies virus in submitted brain tissue. Tooth samples collected from 2001 through 2007 were allowed to dry, stored in a cool location, and submitted to Matson's Laboratory^j for tetracycline analysis and age estimation.

Data analysis—The outcomes of interest for project evaluation were the serologic response to rabies vaccine and the presence of tetracycline biomarkers in trapped raccoons and the number of reported rabid terrestrial animals. A master database containing essential information from each campaign year's results was created. A χ² test with Yates correction for analysis via statistical software^k was performed when appropriate.

Because the prevalence of RVNA in sampled raccoons following each annual campaign was being used as an index to population immunity, prior captures, meaning raccoons that were initially captured in a previous year, were not eliminated from the statistical analyses presented in this paper unless noted. The trapper estimated the ages of raccoons for the entire 10 years of the study. Starting in 2001, a tooth was collected to allow estimation of the age of the raccoon by laboratory analysis. Since the trapper's age estimates span the entire study period, these were used in the data analysis unless noted otherwise.

Results

Pre-ORV data—Of the 182 raccoons tested in all pre-ORV surveys, 19 (10%) were seropositive for rabies antibodies. Thirteen of the 70 (19%) raccoons tested on Broadneck Peninsula in 1999 and 2001 were seropositive. No seropositive raccoons, of 29 trapped, were identified on Gibson Island (2000), whereas 4 of 57 (7.0%) raccoons from southern Anne Arundel County (2002) were seropositive. The raccoon density study conducted in a state park in the northwestern area of the county in 2003 resulted in an estimate of 9 raccoons/km². Two of the 26 (7.7%) tested raccoons in the density study were seropositive. Overall, males had a slightly higher seropositivity than females. Ten percent of adult raccoons were seropositive, compared with 12.5% for subadults and 0% for juvenile raccoons. Subadults raccoons have passed through the juvenile period but not yet attained identifiable adult characteristics and are not sexually mature. This classification was used by the trapper only and only for the first 5 trapping years. Because the laboratory that started estimating age by tooth samples in 2001 did not have a comparable category, the trapper used juvenile and adult categories from 2003 through 2007. Of the 46 teeth samples collected during the pre-ORV survey in southern Anne Arundel County, none had evidence of tetracycline biomarking.

Bait distribution—During the 1998–2007 period, 475,653 baits were distributed in Anne Arundel County. All areas of the county were baited for the first time in 2003 (Figure 1). Once countywide distribution started, approximately 13% of the bait was placed by hand and the remainder by fixed- and rotary-wing aircraft.

Post-ORV results—During the 10 annual bait-distribution campaigns, 1,314 raccoons were trapped (median, 124; range, 46 to 261) at various trapping areas, and 1,310 of these had serologic positivity rates determined as part of the post-ORV bait distribution evaluation studies (Table 1). Tooth samples from 1,016 of the 1,314 raccoons were analyzed for tetracycline biomarkers. Of the 1,314 raccoons, 110 (8.4%) were classified as prior captures. The trapper determined that 761 of 1,311 (58%) were males and 320 of 1,214 (26%) were juveniles; however, 382 of 1,021 (38%), which had their age estimated by tooth analysis (starting in 2001), were juveniles.

Prevalence of RVNA—Four hundred thirty-one of the 1,310 (33%) raccoons tested from 1998 through 2007 were RVNA positive (range, 21% to 47%; Table 2). A cumulative increase in the proportion of seropositive raccoons was not observed over successive annual ORV campaigns. In 2005, the year in which bait was distributed in 2 different months, 19 of 67 (28%) of raccoons trapped after the August campaign were seropositive, compared with 22 of 128 (17%) following September's effort. Except in 2000, when only 1 juvenile raccoon was trapped, adults had higher seropositivity rates than juveniles. Overall, 38% of adults (trapper determined) were seropositive, compared with 18% for juveniles (χ² with Yates correction = 41.48; P < 0.01). There was no significant difference in the proportion seropositive by the sex of the raccoon. For the years 2005 through 2007, when both bait types were distributed, the proportion of seropositive raccoons did not vary by bait format (33% [34/103] in the CS-baited area vs 35% [184/527] for the FP-baited areas). However, a greater proportion of adults were seropositive in the FP-baited areas (42% [155/369] vs 34% [29/86]; χ² with Yates correction = 1.66; P = 0.198), and a greater proportion of juveniles were seropositive in CS-baited areas (29% [5/17] vs 18% [29/158]; χ² with Yates correction = 0.60; P = 0.440; Fisher exact test, P > 0.213), although neither of these differences was significant. This observed age difference remained when the age determined by the laboratory's tooth analysis was used. For each of the 3 years, the age difference remained, except for 2006, when juvenile seropositivity in CS-baited areas (14%) was less than in FP-baited areas (26%). For 2001 and 2002, when 100 baits/km² were placed on the Annapolis Peninsula and 75 baits/km² on Broadneck Peninsula, the seropositivity rate was 41% (47/114) for the Annapolis Peninsula versus 34% (31/91) for Broadneck Peninsula, which was not a significant difference.

A cumulative yearly increase in the proportion of previously captured raccoons was not observed. Overall, 88 of the 110 (80%) prior captures were seropositive (yearly seropositivity ranged from 60% to 100%). From 1999 through 2002, trapped raccoons were not vaccinated with an injectable rabies vaccine, whereas from 2003 through 2007, a parenteral inactivated rabies vaccineⁱ was administered. The mean seropositivity rate was 67% among prior captures for 1999 through 2002 and 84% for 2003 through 2007, which is not a significant difference. Of 33 initially seronegative raccoons that received injectable rabies vaccine and were recaptured 1 year later, 27 (82%) were seropositive and 6 (18%) were seronegative. Interestingly, approximately 70% of prior captures were female, whereas females represented approximately 40% of newly captured raccoons (χ² with Yates correction = 32.76; P < 0.01).

The overall seroprevalence of trapped raccoons did not vary significantly by the start month of the yearly campaign (Table 2). The seroprevalence was 28.5% for August, 35.0% for September, and 31.3% for October. However, when stratified by raccoon age category, juvenile raccoons had a seroprevalence of only 7.6% in August, compared with 22.3% for September and 25.0% for October (P < 0.01). Adult raccoon seroprevalence did not vary significantly by start month.

Bait acceptance—Two hundred eighty-four of 1,016 (28%) raccoons analyzed for tetracycline biomarker had teeth marked with tetracycline. The yearly percentage positive ranged from 20% to 34%. When results from CS-baited areas are excluded (CS baits do not contain tetracycline), 278 of 922 (30%) tested were tetracycline positive with a range of 22% to 35% (Table 2). No cumulative yearly increase was noted in the proportion of raccoons marked with tetracycline. Juveniles (trapper determined) had a slightly higher marking rate than adults, and females had a slightly higher rate than males. However, these differences were not significant. Only 114 of the 285 (40%) seropositive raccoons were also tetracycline positive, whereas 470 of the 634 (74%) seronegative raccoons were also tetracycline negative. Of the 641 tetracycline-negative raccoons, 171 (27%) were seropositive, whereas of the 278 tetracycline-positive raccoons, 114 (41%) were seropositive.

Rabid animals—The number of reported raccoon rabies cases and other terrestrial animal rabies cases has decreased in Anne Arundel County since the first ORV campaign in 1998 (Figure 2). In 1997 (the year prior to the first ORV distribution), there were 79 raccoons, 3 cats, 8 foxes, and 6 skunks reported as rabid throughout the county, compared 6 raccoons, 1 fox, and 1 groundhog in 2007. No terrestrial animal rabies cases were reported on the Annapolis Peninsula from February 1999 through August 2003, but interestingly, most of the cases reported during 2004 through 2008 centered on the Annapolis Peninsula. The incidence of rabid terrestrial animals for this 5-year period (2004 through 2008) was 0.24 rabid terrestrial animals/km² on Annapolis Peninsula, 0.09 rabid terrestrial animals/km² on Broadneck Peninsula, 0.00 rabid terrestrial animals/km² on Gibson Island, and 0.05 rabid terrestrial animals/km² for the remainder of the county.

View Full Size

Laboratory-confirmed rabid terrestrial animals by year and ORV activities, Anne Arundel County, Md, 1984 through 2007. The first ORV campaign occurred in October 1998 on the Annapolis Peninsula, but the entire county was not baited until August 2003. Coated sachet baits were used only in the rural southwestern area during 2005, 2006, and 2007, whereas FP baits continued to be used in the remainder of the county.

Citation: Journal of the American Veterinary Medical Association 241, 6; 10.2460/javma.241.6.725

• Download Figure
• Download figure as PowerPoint slide

Comparison of Anne Arundel County's data with adjoining jurisdictions (Baltimore, Calvert, Howard, and Prince Georges counties and the City of Baltimore), which have not had an ORV program, failed to show a similar decrease for these counties during 1997 through 2007. Two of these are smaller counties and consequently report relatively few rabies cases. Frederick and Montgomery counties, usually in the top 5 Maryland counties for reporting terrestrial rabies cases, were also selected to enhance this comparison. Both these counties, also without an ORV project, did not have a similar decrease, compared with Anne Arundel County (Figure 3). The State of Maryland rabid animal reports for the same 10 years had a mean of 408, a median of 404, and a range from 337 to 505. From 2004 through 2007, Maryland cases increased each year. Therefore, statewide experience did not parallel the monotonic decrease observed in Anne Arundel County. The Maryland Department of Health and Mental Hygiene reported that rabid animal cases adjusted for the human population in Anne Arundel County were the lowest of any Maryland county in 2007.¹³

View Full Size

Laboratory-confirmed rabid terrestrial animals by year in Anne Arundel County (diamonds), Frederick County (squares), and Montgomery County (triangles), Md, 1997 through 2007. Frederick and Montgomery typically fall in the top 5 Maryland counties in terms of the incidence of terrestrial animal rabies. Neither is contiguous with Anne Arundel County or ever had an ORV campaign.

Citation: Journal of the American Veterinary Medical Association 241, 6; 10.2460/javma.241.6.725

• Download Figure
• Download figure as PowerPoint slide

Discussion

Following yearly ORV campaigns conducted from 1998 through 2007, annual seropositivity rates in trapped raccoons varied between 21% and 47%. The number of case of rabies in terrestrial animals decreased 92% between 1997 (the year before the first campaign) and 2007, from 96 cases to 8. Rabies cases in Maryland were reported to have periodic upsurges possibly every 3 to 5 years.¹⁴ Historically, Anne Arundel County has had a 5- to 6-year cycle from 1984 through 1998.^b The fact that there was not an upswing in reported rabid animals but rather a monotonic decrease in cases since the first campaign in 1998 provides evidence that the yearly ORV campaigns with an overall bait density of 75 to 100 baits/km² have had a major impact on reducing terrestrial animal rabies reported in Anne Arundel County.

One of the shortcomings of the evaluation of the project was the inability to conduct active surveillance of raccoons and other terrestrial animals for presence of the rabies virus. Reliance on passive surveillance for determining occurrence of rabid animals may have resulted in the underreporting of cases.¹⁵ For example, during a 55-month period, from February 1999 through August 2003, no cases of rabies were reported on Annapolis Peninsula, whereas unbaited areas of the county continued to report rabid terrestrial animals. This lack of cases caused the reduction in bait placement on the southern part of Annapolis Peninsula for the year 2002 and its elimination in 2003. Also, the bait density in the northern parts of Annapolis Peninsula was reduced in 2003 from 100 to 75/km². In 2004, a resurgence of cases occurred on Annapolis Peninsula, with cases in 5 raccoons and 1 groundhog being reported. This raises the possibility that, during the putative rabies-free period, undetected rabid animals were present but not identified by passive surveillance, resulting in the erroneous conclusion that the southern area of the Annapolis Peninsula was probably free of cases of rabies in terrestrial animals. Perhaps the reduction and subsequent elimination of baiting during that 2-year period reduced immunity in the population to a point that allowed rabies to be reintroduced and spread. Alternatively, rabies may have been present but not detected because of an inadequate surveillance system.

To help determine whether the decrease in reported rabid animals was related to the ORV campaigns, Anne Arundel County rabid animal reports were compared with neighboring and other Maryland counties' results during the same time periods. The finding that these counties did not have similar decrease in terrestrial rabies cases and the Maryland Department of Health and Mental Hygiene reported that rabid animal cases adjusted for the human population in Anne Arundel County were the lowest of any Maryland county in 2007¹³ support the conclusion that the decrease in terrestrial rabid animals in Anne Arundel County was most likely due to the ORV program.

It was thought that use of this new ORV technology in a mainly suburban and urban setting presented the opportunity to gain additional knowledge about the best methods to achieve a successful outcome. Collection of post-ORV serologic and other data revealed several interesting observations. A recent study¹⁶ found no differences in RVNA prevalence on the basis of age or sex in raccoons. Other studies^17,18 have found an association between age and seropositivity rates. Results of the study reported here supported the similar conclusion regarding the raccoon's sex, and post-ORV biomarker results consistently provided evidence that juvenile raccoons accessed FP bait at approximately the same rate as adults (tetracycline marking, approx 30%). However, from 1998 through 2007, significantly fewer juveniles were seropositive (18% vs 38%; P < 0.000; Table 2). Factors such as an immature immune system and possible maternal antibody interference may have influenced the development of RVNAs in young raccoons. Also, the behavior patterns of very young raccoons may have resulted in fewer opportunities for juveniles to find the bait. However, if our observation that juvenile raccoons access the baits at the same rate as adults is correct, another contributory factor may exist. Most juvenile raccoons are born in April in Maryland,⁸ and therefore, many may not have reached sufficient physical size in August to be able to bite into the FP and penetrate the sachet. Interestingly, 36% of the juveniles trapped in August weighed < 2.0 kg (4.4 lb), compared with 18% in September.

The initial campaigns (1998 through 2000) were conducted during several weeks in October with post-ORV trapping being conducted in November and December. It was found that the trapping success diminished in December probably because of onset of colder weather. Consequently, ORV campaigns for 2001 and 2002 were conducted in September, but for aircraft access issues and convenience, campaigns in 2003 and 2004 were conducted in August. When RVNA prevalence was compared on the basis of the start month, it was found that juvenile raccoons fared more poorly during August than September and October.

Although the inability for some juveniles to be able to puncture the FP sachet may be present in September and October campaigns, it is probably more prevalent in August. This problem would not be expected for CS, and when seroprevalence for 2005 through 2007 was compared between FP- and CS-baited areas for all raccoons (35% and 33%, respectively; Table 2), there was no significant difference (χ² with Yates correction = 0.07; P = 0.796). However, seropositivity among juveniles was similar to adults (29% vs 34%; χ² with Yates correction = 0.00; P = 0.950) from CS-baited areas, whereas the adult seroprevalence was significantly (P < 0.01) higher than that of juveniles (42% vs 18%; χ² with Yates correction = 26.20) in FP-baited areas. These data indicate that, although juveniles access the FP bait at approximately the same rate as adults (30% vs 35%, respectively), their seropositivity rate was significantly less. This age differential was not observed for the CS bait format. The data for bait type comparison were collected from 2005 through 2007, years during which almost all the bait, except for the FP bait on Broadneck Peninsula in 2005, was distributed in the month of September (Table 1), removing the start month of the campaign as an important confounding factor in this analysis.

One evaluation area on Annapolis Peninsula, Quiet Waters Park, was a trapping site during all 10 campaigns (1998 through 2007), and similar to the total results for the 10-year period, there was not a cumulative increase in the percentage of seropositive raccoons observed following each annual bait distribution. This site was not baited in 2003, and the seropositivity rate was 29% (15/51) for that year (only 1 juvenile was seropositive), which compares to its mean of 44% (62/141) for the preceding 5 years (bait density, 100/km²). These results, based on relatively small samples, suggest that the decrease in the seropositivity rate in 2003 was due to no bait distribution in Quiet Waters Park that year. Baiting resumed in 2004 (bait density, 75 baits/km²), and the mean seropositivity rate for 2004 through 2007 was 41% (70/170). The fact that terrestrial animal rabies continued to occur (even at a relatively low incidence) on Annapolis Peninsula with an estimated 41% seropositivity rate would suggest that a higher RVNA prevalence may be needed to consistently interrupt terrestrial animal rabies transmission.

On the basis of our project's results with 2 age-determination methods, it estimated that the juvenile proportion of Anne Arundel County's raccoons lies somewhere between 26% and 38% during 1998 through 2007. Applying our study's observed seropositivity rates based on the 2 bait formats and the raccoon's age to a theoretical population of 1,000 raccoons with an assumed 33% proportion of juveniles, the FP bait would vaccinate 341 raccoons, whereas CS would reach 324, a difference of 17 (5.0%). Each bait format has different logistic and cost factors that warrant consideration in deciding which format to use, but from an RVNA-prevalence perspective, FP did slightly better overall than CS in this theoretical exercise.

Refinement in mapping efforts has permitted improved targeting of bait placement and greater logistic efficiency, which are felt to be reflected in the 2007 results. For this year, a 47% seropositivity rate was observed overall. This compares to the prior 9-year mean of 30% (χ² with Yates correction = 27.14; P < 0.01). Unexpectedly, a comparable improvement in tetracycline positivity rates was not observed in FP-baited areas (35% in 2007 vs 29% for the 6-year mean; χ² with Yates correction = 2.51; P = 0.113). Moreover, other reported ORV studies^6,18–20 that have used tetracycline as a biomarker to evaluate bait acceptance have typically found that tetracycline positivity rates are higher than seropositivity rates, which is contrary to our findings (30% vs 33%, respectively, for FP-baited areas). In fact, 60% of the seropositive raccoons from FP-baited areas had no biomarker. It has been reported that manufacturing methods may affect the concentration and stability of the tetracycline in the baits (ie, tetracycline is converted to epitetracycline, which is a poorer biomarker, during the manufacturing process), thereby lowering the marking potential of the baits.²¹

One study^l suggested that the animal's age may influence tetracycline marking ability, but the reported study²¹ found that the age of the test animals had no effect on tetracycline marking. We found essentially no overall significant difference in tetracycline positivity between juvenile and adult age categories in FP-baited areas. In our ORV project, tooth samples (usually the first premolar) from 913 raccoons were analyzed by a commercial laboratory that determined age in terms of years. Of the juveniles, 116 of 349 (33%; Table 2) were tetracycline positive, as were 160 of 564 (28%) adults 1 year of age or older (χ² with Yates correction = 2.20; P = 0.138). However, as the raccoon age increased, the tetracycline positivity decreased. In fact, only 2 of 26 (8%) of raccoons 7 years of age or older were tetracycline positive. Given that there is probably a dose-dependent response between tetracycline and body weight, older raccoons, which typically weigh more, may not ingest a sufficient amount of tetracycline to cause marking of the teeth. Another explanation may be that adults deposit tetracycline in tissues at a slower rate than juveniles.²² Other factors such as the simultaneous consumption of other foods and water may affect the biomarking capability of the tetracycline in the FP bait matrix. Environmental factors, such as time and temperature, may influence the tetracycline's stability, but the reported relatively rapid uptake of baits once distributed^5,6 would lessen the likelihood that this would explain our findings. As stated, baits were kept at refrigeration temperature prior to bait placement. On the day of distribution, ground baits were kept in insulated chest coolers, and for aircraft drops, these were kept in the shade. There were no recognized problems in maintaining the so-called cold chain while handling the baits.

Another possible explanation for the relatively low tetracycline positivity may relate to the tooth chosen for tetracycline analysis. The first premolar was selected because of its ease of removal and the minimal impact its loss would have on survival of the raccoon. However, it has been recognized that tetracycline does not always fluoresce in this tooth as well as in second premolars.²³ Some researchers prefer the canine tooth, but for a live raccoon, this is felt to be an essential tooth. A better understanding of the availability and uptake of the tetracycline in the FP bait is needed.

Age was estimated for all 10 years by trappers and from 2001 through 2007 on the basis of the analysis of a tooth sample by a commercial laboratory. Based on both the trapper's and laboratory's determination during 2001 through 2007, the juveniles comprised 296 of 1,093 (27%) and 382 of 1,018 (38%) raccoons trapped, respectively. The trapper classified 86 raccoons as adults that the laboratory indicated were juveniles. The reverse of this (ie, the trapper classified a raccoon as juvenile that the laboratory classified as an adult) occurred only 16 times. There was 90% agreement between the trapper's and laboratory's classifications, and the 10% discrepancy in the age structure did not impact our interpretation of the results. For example, when the laboratory's age was used for the 1,018 analyzed teeth samples, the overall seropositivity was 19% for juveniles and 39% for adults, which differed slightly from the trapper's age-respective values of 18% and 38% for the 1,210 raccoons the trapper classified as either adult or juvenile (Table 2).

The results reported in this paper are based on the inclusion of the prior captures (8.4% of the sample of 1,314 raccoons) in our calculations unless otherwise noted. It was an interesting question whether to include or exclude this subgroup. Other reported studies of the trapping and serologic testing of raccoons over separate time periods differed in whether the results for prior captures were reported separately or included with the unique raccoon data.^3,16–20 Completing calculations without prior captures included did not alter the interpretation of our data. Raccoons that are seropositive may have acquired RVNAs because of natural infection, maternal RVNA transfer, inoculation with the oral rabies vaccine, or inoculation with an injectable killed rabies vaccine (eg, released pet or rehabilitated raccoons). Also, seropositive raccoons may be translocated or migrate into the county. With this in mind, the seroprevalence identified in this study are an estimation of the total number of raccoons with RVNA induced by any one or a combination of these events. In our study, how total RVNAs prevalence relates to the frequency of terrestrial animal rabies was a key question. Considering that first-time captures may have seroconverted in a prior year and not the current year and prior captures may have seroconverted during the current year's campaign, we decided to include the prior captures in the presented data so as not to underestimate the impact of the ORV program.

The pre-ORV serosurvey results, conducted over a 5-year period, varied widely by area but when combined indicated a background seropositivity rate of 10%, which is within the range (0% to 41%) reported by other researchers for raccoon rabies enzootic and rabies-free areas.^3,17,20,24 To gauge the impact of the ORV project on the prevalence of RVNA in an enzootic area, it is necessary to consider the background level. For example, it would be expected that perhaps 26 of the 259 (10%) raccoons trapped in 2007 would have acquired RVNA from a source other than the ORV program. If these 26 are eliminated from the total seropositives, the overall seropositivity rate for 2007 would decrease from 47% (121/259) to 41% (95/233). However, from a practical standpoint, it would seem that the important comparison is the relationship of the total seroprevalence (regardless of source of RVNA) to the frequency of terrestrial animal rabies.

In our study, an across-all-years RVNA prevalence of 33% was associated with a 9-fold reduction in terrestrial animal rabies. Given a raccoon population in which perhaps 40% turns over each year, the current efficacy of the ORV programs in reaching raccoons, and the presence of raccoon rabies strain in neighboring jurisdictions, it is expected that at least annual ORV campaigns at a bait density of 75 baits/km² will be required to maintain the current success of this program in Anne Arundel County.