Prevalence and Seasonality of Fleas Associated With California Ground Squirrels and the Potential Risk of Tularemia in an Outdoor Non-Human Primate Research Facility

Prevalence and Seasonality of Fleas Associated With California Ground Squirrels and the Potential... Abstract Ectoparasites at primate research centers may be difficult to control, e.g. without exposing non-human primates (NHPs) to toxicants, but their impact on NHP health is poorly understood. In 2010, there was an epizootic of tularemia at the California National Primate Research Center (CNPRC) in Yolo County, California that resulted in 20 confirmed and suspect clinical cases in outdoors housed rhesus macaques (Macaca mulatta [Zimmermann]) and a 53% seroprevalence in the southern section of the colony. We studied ectoparasite burdens at the CNPRC in order to understand possible conditions at the time of the epizootic and provide data for the management of ectoparasites for the future. In 2015, we recorded 52 California ground squirrel (Otospermophilus beecheyi [Richardson]) burrow systems in the southern colony and collected 560 fleas. The largest number of fleas (n = 184) was collected in October and the most common species were Hoplopsyllus anomalus (Baker) (n = 331), Oropsylla montana (Baker) (n = 158), Echidnophaga gallinacea (Westwood) (n = 60), and Ctenocephalides felis (Bouché) (n = 11), all of which are opportunistically anthropophilic. Free, non-host-associated fleas included 12 H. anomalus, 9 C. felis, 6 O. Montana, and 1 E. gallinacea. We collected 1 H. anomalus from a rhesus macaque. Our results suggest a high potential for the rapid spread of zoonotic infectious diseases via flea transmission in primate facilities with ground squirrels and that flea control measures should be given a high priority. fleas, Hoplopsyllus, anomalus, Echidnophaga, gallinacea, Oropsylla, montana, Ctenocephalides, felis Tularemia is a serious febrile illness caused by the intracellular bacterium Francisella tularensis (Foley and Nieto 2010). As few as 10 colony-forming units are needed to induce fatal septic or pulmonary infection. The bacterium is easily disseminated and has a potentially high fatality rate of up to 40%, leading it to be classified as a Category A select agent by the US Department of Health and Human Services. In 2010, there was an abnormally high incidence of tularemia in the western United States, with seven clinical cases reported compared with a yearly average of 2.8 (California Department of Public Health 2010). In addition to the increase in human cases, non-human primate (NHP) research centers, including the Oregon National Primate Research Center and the California National Primate Research Center (CNPRC) in Yolo County, experienced localized epizootics in NHPs in the same year (Ferrecchia et al. 2012, Sammak et al. 2012). The CNPRC had 20 confirmed and suspected clinical cases of tularemia and a 53% seroprevalence in rhesus macaques (Macaca mulatta) in an area of the CNPRC known as ‘South Colony’. Serological testing confirmed that all of these cases had occurred in 2010 (Sammak et al. 2012). Disease in research NHP populations may destabilize social structure, affect animal welfare, and affect research outcomes, potentially incurring expense through increased veterinary care and the loss of valuable animal genotypes. Although ticks are an important biological vector for F. tularensis (Foley and Nieto 2010), the spatial distribution of cases and wildlife surveys at CNPRC were not supportive of tick transmission (Sammak et al. 2012) and the infection pathway at the CNPRC is unknown. There are a wide variety of vertebrate taxa that can host F. tularensis (Foley and Nieto 2010); however, the agent was first isolated from California ground squirrels (Otospermophilus beecheyi) in Tulare county, California in 1911 (McCoy and Chaplin 1912). California ground squirrels host multiple zoonotic bacterial pathogens including Yersinia pestis, the agent of plague, and Bartonella washoensis which has been associated with human endocarditis and meningitis (McCoy and Chaplin 1912, Evans et al. 1943, Kartman and Prince 1956, Preiksaitis et al. 1979, Nelson 1980, Nelson et al. 1986, Chomel et al. 2003, Kosoy et al. 2003, Hubbart 2012). California ground squirrels have been directly implicated in the amplification and spread of F. tularensis to NHPs (Preiksaitis et al. 1979). They also serve as hosts of several species of flea that are opportunistically anthropophilic (such as Oropsylla montana) and that are capable of mechanically transmitting a variety of pathogens, such as plague (Eskey and Haas 1939, Prince and McMahon 1946, Hubbard 1968). Ground squirrel burrows are often heavily infested with fleas (Rutledge et al. 1979), and each squirrel may support 20 or more fleas at any given time (Metzger and Rust 1999, Hubbart 2012). Fleas are considered to be poor vectors for F. tularensis transmission (Kartman and Prince 1956) but it is possible that repeated bites and large numbers of this vector could overcome their poor vector efficiency. Flea loads on ground squirrels and in ground squirrel burrows have been studied in rural and semi-rural landscapes (Hubbart et al. 2011), but there is little available information on flea communities and load in locations with frequent pest management as occurs at the CNPRC. Our study was performed at the CNPRC South Colony and aimed to determine flea species, whether fleas were in burrows, above-ground, or host-associated, their seasonality, and association of burrow risk factors (evident rodent activity, depth, location, and pest control) with flea presence. Data from this study could inform whether fleas associated with facilities that regularly control for rodent pests, such as the CNPRC, pose a risk for disease transmission and whether pest control methods could better be targeted to mitigate this risk. Materials and methods Study Site The CNPRC is a large research facility in Davis, Yolo County, California that houses an average of 4,000 NHP research animals (predominantly rhesus macaques in indoor and outdoor enclosures). Outside, macaques are housed either in large enclosures of 20–200 individuals in the North Colony or in small group pens of 2–30 individuals in the South Colony. The floor substrate, which could affect ectoparasite presence, in the South Colony enclosures was gravel at the time of the study, while the North Colony enclosures had patches of grass and gravel. Beginning 9 September 2015, the South Colony of the CNPRC was being renovated and multiple NHP enclosures along the eastern edge of the site were removed in order to begin construction on new enclosures with concrete bases. All enclosures were supplied with enrichment items such as barrels, swings, platforms, and ropes and daily dietary items such as fruits and vegetables. Row-crop fields border the CNPRC to the north, west, and south. The eastern edge of the CNPRC is valley oak (Quercus lobata) savannah with non-native annual grasses and weeds. Seasonal standing water collects in multiple reservoirs and drainage ditches in and near the CNPRC. The center of the CNPRC, between the North and South Colony, houses a managed colony of ~32 feral cats. There were three feeding stations in the feral cat colony which were dusted weekly with flea powder (Hartz Mountain Corporation, Secaucus, NJ). Feral cats were trapped and treated annually for fleas with selamectin (Zoetis, Florham Park, NJ) or imidacloprid and pyriproxyfen (Bayer AG, Leverkusen, Germany). A vertebrate pest control program for the CNPRC consisted of live trapping and removal of rodents followed by filling burrow openings, particularly those nearest to primate cages and buildings, with gravel and soil weekly to bi-weekly. Toxicants were deployed rarely and only in ground squirrel burrows located outside the perimeter fence or in a disused fenced-in field located 10 meters from the nearest NHP enclosures. Flea Sampling NHPs were systematically examined for fleas during routine bi-annual health examinations from March to October 2015 and opportunistically from monkeys admitted to the hospital from 1 March 2015 until 31 October 2015. Special attention was paid to each animal’s back, groin, under arms, and behind the ears. The single flea recovered was stored in 70% ethanol. We collected fleas from burrow systems biweekly from 1 March to 31 October 2015 between 8:00 a.m. and 12:00 p.m. A burrow system was defined as a single underground unit with single or multiple entrances. Entrances were considered to be a part of the same burrow system if they were within 3 m and were oriented toward each other, determined by insertion of a 1.5-m long dowel. A global positioning system (GPS) coordinate was taken at the burrow entrance for single-entrance systems or at the center of the burrow system for multiple-entrance systems, determined from directionality indicated by dowels in entrances. Fleas were collected by swabbing burrows using a 20 × 20 cm piece of white flannel attached to the end of a 1.5m long dowel (Gage 1999). Burrows were only swabbed if the dowel could be inserted at least 10 cm. One entrance was swabbed if the burrow system had one to three entrances, two entrances were swabbed for burrow systems with four to five entrances, and three entrances were swabbed for burrow systems with six or more entrances. Fleas were also collected from 3-m linear transects near high human-traffic areas, including near the South Colony office, at a feral cat feeding station, between two storage buildings, between two buildings where NHP transport cages were stored, and between two laboratory buildings (Fig. 1), using a white-sock collection technique (Borchert et al. 2012). An individual wearing white medical scrub tucked into white socks and shirt tucked into pants walked through each transect for 3 min. Fleas were removed from socks and placed in 1.5-ml tubes of 70% ethanol as spotted. Sampling was conducted biweekly from 1 March to 31 October 2015 between 8:00 a.m. and 12:00 p.m. Fleas were also opportunistically collected from scrubs during burrow sampling if they were observed and placed in a separate vial from those collected from the burrow. In regions with high flea activity (such as highly infested burrows), opportunistic collection from scrubs was terminated after 5 min. The collecting individual moved from the area of activity, checked their scrubs a final time, and then proceeded with other activities. Fig. 1. View largeDownload slide The location of transects walked during white-sock collection at the CNPRC from March to November 2015. Fig. 1. View largeDownload slide The location of transects walked during white-sock collection at the CNPRC from March to November 2015. For processing, fleas were cleared in a 10% solution of potassium hydroxide for 24 h and serially dehydrated for 30 min, each in 75, 80, 90, and 100% ethanol. Fleas were then mounted on glass slides using Euparal (BioQuip Products, Rancho Dominguez, CA) and identified to species using keys (Lewis et al. 1988). Analysis GPS points were mapped using Arc GIS 10.2.x (ESRI, Redlands, CA). Data analyses were performed in R (R Core Team 2015) and a P ≤ 0.05 was used a cutoff to infer statistical significance. We calculated the average number of fleas recovered from burrows with the same number of entrances in order to determine whether burrows with more entrances had greater numbers of fleas. In order to account for variation in the number of entrances being swabbed, we also calculated the average numbers of fleas collected per swab (Table 2). Potential risk factors for fleas were recorded as whether the burrow appeared to be in recent active use by a rodent (indicated by soil disruption, feces, food caches or an entryway cleared of plant debris or spider webs, designated ‘Active’ or ‘Inactive’), whether the swab could be inserted ≥0.5 m into the burrow (indicating the presence of a straight/wide passage, designated as ‘Deep’ or ‘Shallow’), whether the burrow was located under a permanent structure such as a tree or building (‘Under’ or ‘Out’), whether the burrow had been filled in with soil or gravel during the previous sampling attempt (‘Open’ or ‘Closed’), and the number of burrow system entrances. We performed Mann–Whitney U tests to determine whether mean numbers of fleas per swab were different between paired risk factors. Results We collected a total of 589 fleas including 560 from ground squirrel burrows. H. anomalus was the most common species collected (n = 331), followed by O. montana (n = 158), Echidnophaga gallinacea (n = 60) and Ctenocephalides felis (n = 11) (Table 1). We collected 28 fleas (12 H. anomalus, 9 C. felis, 6 O. montana, and 1 E. gallinacea) using the modified white-sock collection technique (Table 1). Of the fleas collected on white socks, a single H. anomalus and all C. felis were collected from a region between two buildings where primate cages were stored that was being used by feral cats for shelter (Fig. 1). Despite more than 650 person-hours dedicated to searching NHPs admitted to the hospital for fleas, only one H. anomalus was found. Table 1. Fleas collected from California ground squirrel (O. beecheyi) burrows and from clothing during white-sock collections at the CNPRC from 26 March 2015 to 28 October 2015   3/26  4/9  5/28  6/11  6/24  7/9  7/27  8/9  8/26  9/10  9/23  10/12  10/28  Total  Fleas from burrows  C. felis  0  0  0  0  8  0  0  0  0  2  0  1  0  11  E. gallinacean  0  0  0  0  0  1  4  4  2  5  10  6  28  60  H. anomalus  0  1  1  7  2  49  12  19  33  50  23  82  52  331  O. montana  1  0  19  50  10  38  14  7  5  2  4  1  7  158  Free fleas  C. felis  0  0  0  9  0  0  0  0  0  0  0  0  0  9  E. gallinacean  0  0  0  0  0  0  0  0  0  0  0  0  1  1  H. anomalus  0  0  0  1  0  0  0  0  3  3  0  1  4  12  O. montana  0  0  3  1  0  0  0  0  1  0  0  0  1  6  Total  1  1  23  68  20  88  30  30  44  62  37  91  93  588    3/26  4/9  5/28  6/11  6/24  7/9  7/27  8/9  8/26  9/10  9/23  10/12  10/28  Total  Fleas from burrows  C. felis  0  0  0  0  8  0  0  0  0  2  0  1  0  11  E. gallinacean  0  0  0  0  0  1  4  4  2  5  10  6  28  60  H. anomalus  0  1  1  7  2  49  12  19  33  50  23  82  52  331  O. montana  1  0  19  50  10  38  14  7  5  2  4  1  7  158  Free fleas  C. felis  0  0  0  9  0  0  0  0  0  0  0  0  0  9  E. gallinacean  0  0  0  0  0  0  0  0  0  0  0  0  1  1  H. anomalus  0  0  0  1  0  0  0  0  3  3  0  1  4  12  O. montana  0  0  3  1  0  0  0  0  1  0  0  0  1  6  Total  1  1  23  68  20  88  30  30  44  62  37  91  93  588  No fleas were collected on sampling dates 11 March, 23 April, and 14 May. View Large The month in which we recovered the largest number of fleas was October, with 184 individuals, mostly H. anomalus (n = 134) followed by E. gallinacea (n = 34). The number of O. montana peaked in early June (n = 50) but dropped rapidly and remained low from late July till the end of the collection period. In contrast, H. anomalus numbers rose in July and remained high throughout the remainder of the study period, peaking at 82 individuals collected during a single-collection period in early October. E. gallinacea remained at very low levels until late October when they peaked at 28 individuals collected. To assess the association of rodent burrow characteristics with flea numbers, we recorded 52 distinct rodent burrow systems throughout the South Colony of the CNPRC (Fig. 1). We described the temporality of burrow attributes (e.g., whether they were in use by rodents), the spatial organization of burrows (Fig. 2), whether burrows had been closed by pest control staff the previous collecting period, and characteristics of the burrow entrance such as whether the entrance was located under a permanent structure. The month of June had the highest number (n = 38) of open burrow systems, while March had the fewest (n = 6) (Fig. 2 shows the distribution of burrows during June). About 73% of all burrow systems were single-entrance and were sampled a total of 189 times over the study period. This was followed by two-entrance (11% of all systems) and three-entrance (6.6% of all systems) burrows that were sampled a total of 29 and 17 times, respectively (Table 2). Burrow systems with six or more entrances were the most uncommon (5% of all systems). In general, burrow systems with six or more entrances had higher loads of fleas per swab than burrow systems with less than six entrances (P = 0.006). These categories were determined by plotting a histogram of numbers of burrows with given numbers of openings, and utilizing the natural cutoff between the two bimodal peaks in the data. The average number of fleas recovered per swab was significantly higher from burrows that had been opened the previous sampling period (1.33) compared with burrows that had been closed during the previous sampling period (1.05, P = 0.014). Burrows where the probe could be inserted deeply had a higher average fleas per swab (1.38) compared with from shallow burrows (0.99, P = 0.048). There were no significant differences in the average number of fleas per swab for burrows located under a permanent structure versus open soil areas or lawns or active versus inactive burrows (Table 3). The number of H. anomalus fleas from open burrows (0.69) was greater than those collected from closed burrows (0.62, P = 0.032). We did not detect any effect of these risk factors on the average number of O. montana or E. gallinacea recovered per swab (Table 3). Fig. 2. View largeDownload slide The distribution of open burrows recorded on GPS at the South Colony of the CNPRC on 11 June 2015. Fig. 2. View largeDownload slide The distribution of open burrows recorded on GPS at the South Colony of the CNPRC on 11 June 2015. Table 2. Summary statistics of fleas swabbed from open burrow systems recorded in the South Colony area of the CNPRC in 2015 Number of openings  Number of burrows  Average number of fleas per burrow (SE)  Average number of fleas/swab  Minimum and maximum number of fleas on swabs  Number of burrows with no fleas  1  189  1.322 (0.27)  1.322  0.35  126  2  29  1.24 (0.3)  1.24  0.4  16  3  17  0.7 (0.3)  0.7  0.4  12  4  8  3 (0.93)  1.5  1.9  0  5  3  1.2 (0.97)  0.6  0.5  1  6  1  0 (n/a)  0  0.0  1  7  4  10.5 (10.5)  3.5  0.42  3  8  1  0 (n/a)  0  8.8  0  9  2  9.5 (8.5)  3.167  1.18  0  10  2  37 (35)  12.33  2.72  0  11  1  4 (n/a)  1.33  4.4  0  15  2  28.5 (12.5)  9.5  16.41  0  Number of openings  Number of burrows  Average number of fleas per burrow (SE)  Average number of fleas/swab  Minimum and maximum number of fleas on swabs  Number of burrows with no fleas  1  189  1.322 (0.27)  1.322  0.35  126  2  29  1.24 (0.3)  1.24  0.4  16  3  17  0.7 (0.3)  0.7  0.4  12  4  8  3 (0.93)  1.5  1.9  0  5  3  1.2 (0.97)  0.6  0.5  1  6  1  0 (n/a)  0  0.0  1  7  4  10.5 (10.5)  3.5  0.42  3  8  1  0 (n/a)  0  8.8  0  9  2  9.5 (8.5)  3.167  1.18  0  10  2  37 (35)  12.33  2.72  0  11  1  4 (n/a)  1.33  4.4  0  15  2  28.5 (12.5)  9.5  16.41  0  View Large Table 3. The effects of various risk factors on the mean number of fleas per swab collected at the CNPRC from March to November 2015 Risk factor pairing  Means  P-value  Open vs. Closed  1.33, 1.05  0.014*  Deep vs. Shallow  1.38, 0.99  0.048*  Under vs. Out  1.59, 0.97  0.158  Active vs. Inactive  1.57, 0.86  0.089  O. montana Open vs. O. montana Closed  0.49, 0.33  0.128  O. montana Deep vs. O montana Shallow  0.42, 0.47  0.144  O. montana Under vs. O. montana Out  0.66, 0.27  0.154  O. montana Active vs. O. montana Inactive  0.36, 0.53  0.372  H. anomalus Open vs. H. anomalus Closed  0.699, 0.62  0.032*  H. anomalus Deep vs. H. anomalus Shallow  0.81, 0.44  0.130  H. anomalus Under vs. H. anomalus Out  0.76, 0.61  0.787  H. anomalus Active vs. H. anomalus Inactive  0.15, 0.08  0.286  E. gallinacea Open vs. E. gallinacea Closed  0.13, 0.096  0.344  E. gallinacea Deep vs. E. gallinacea Shallow  0.14, 0.08  0.184  E. gallinacea Under vs. E. gallinacea Out  0.16, 0.09  0.227  E. gallinacea Active vs. E. gallinacea Inactive  0.15, 0.08  0.286  Risk factor pairing  Means  P-value  Open vs. Closed  1.33, 1.05  0.014*  Deep vs. Shallow  1.38, 0.99  0.048*  Under vs. Out  1.59, 0.97  0.158  Active vs. Inactive  1.57, 0.86  0.089  O. montana Open vs. O. montana Closed  0.49, 0.33  0.128  O. montana Deep vs. O montana Shallow  0.42, 0.47  0.144  O. montana Under vs. O. montana Out  0.66, 0.27  0.154  O. montana Active vs. O. montana Inactive  0.36, 0.53  0.372  H. anomalus Open vs. H. anomalus Closed  0.699, 0.62  0.032*  H. anomalus Deep vs. H. anomalus Shallow  0.81, 0.44  0.130  H. anomalus Under vs. H. anomalus Out  0.76, 0.61  0.787  H. anomalus Active vs. H. anomalus Inactive  0.15, 0.08  0.286  E. gallinacea Open vs. E. gallinacea Closed  0.13, 0.096  0.344  E. gallinacea Deep vs. E. gallinacea Shallow  0.14, 0.08  0.184  E. gallinacea Under vs. E. gallinacea Out  0.16, 0.09  0.227  E. gallinacea Active vs. E. gallinacea Inactive  0.15, 0.08  0.286  See text for classifications of risk factors. *Significant value View Large Discussion Outdoor housing of NHPs encourages natural behavior and lowers stress (Novak and Suomi 1988) but may also expose the animals to wildlife-associated diseases (Matz-Rensing et al. 2007, Sammak et al. 2012) which can destabilize social structure, affect animal welfare, and affect research outcomes, potentially incurring expense through increased veterinary care. Disease control measures in NHP research centers must avoid NHP exposure to toxins, whereas studies on ground squirrels and their fleas have focused on areas where toxicants (Baldwin et al. 2014) and burrow destruction (Salmon et al. 1987) have been viable methods for pest control. Here, we characterized the flea community on NHPs and off-host at an NHP facility with a history of tularemia epizootics and found abundant fleas associated with ground squirrel burrows at the facility despite active pest control. We recovered fleas including O. montana, H. anomalus, and C. felis that are aggressive parasites of humans and likely will feed from NHP as well (Lewis et al. 1988, Bitam et al. 2010). O. montana and H. anomalus are vectors of sylvatic plague in the western United States (Stewart and Evans 1941, Wimsatt and Biggins 2009). Ground squirrel fleas have also been implicated in the spread of Bartonella washoensis, a pathogen that has been reported to cause meningitis in humans and mitral valve endocarditis in dogs (Chomel et al. 2003, Probert et al. 2009). C. felis are pests of domestic cats and dogs as well as wildlife such as Virginia opossums (Didelphis virginiana) and are important vectors of Rickettsia felis, the agent of cat-flea typhus (Reif and Macaluso 2009), B. henselae, the agent of cat scratch fever (Higgins et al. 1996), and R. typhi, the agent of murine typhus (Noden et al. 1998). They also act as an intermediate host for the tapeworm Dipylidium caninum (Hubbard 1968, Lewis et al. 1988). Although fleas tend to have low vector competence for F. tularensis (Parker and Johnson 1957, Bibikova 1977), the low infectious dose of this pathogen could contribute to a relatively high vectorial capacity for fleas in a situation where there is a high flea burden (Volfertz and Kolpakova 1946, Kartman and Prince 1956). Despite the large number of fleas collected from the CNPRC grounds, we only recovered one flea from a hospitalized NHP directly. It is unlikely that the fleas cannot access the NHPs or are unwilling to feed on them. More likely, the low number of fleas is due to our inability to detect fleas on them or social grooming. All NHPs are anesthetized before examination and fleas often leave anesthetized hosts due to a dropping body temperature (Gross and Bonnet 1949). If there were substantial delays between initiation of anesthesia and examination, it is possible we could have overlooked fleas. If fleas were removed by grooming, this could still be a transmission route for some pathogens as monkeys often consume groomed ectoparasites. This is supported by the fact that the majority of tularemia cases reported in 2010 were oropharyngeal (Sammak et al. 2012) likely acquired via consumption of contaminated materials. Consistent with this result, oral acquisition of infection is well-documented, with natural outbreaks of food- and water-borne tularemia in both Old and New Worlds (Foley and Nieto 2010), although with a higher infectious dose in animal models than may be required for other transmission routes (Tulis et al. 1969). Most literature regarding arthropods and tularemia, however, suggests transmission via arthropod bite, possibly overlooking the mechanism of arthropod ingestion. In addition to the possibility of oral transmission to monkeys via flea ingestion, the disease may also have been acquired if monkeys ate infected squirrels as well. Flea burden in ground squirrel burrows was associated with the number of burrow openings, which is correlated with the number of squirrels living in the burrow (Owings and Borchert 1975). Our study may not have accurately classified all burrow systems if, e.g., new entrances were built greater than 3 m from the central mound (Grinnell 1918) and we inaccurately classified it as a different burrow system. Some studies solve this problem by pumping smoke into the burrows and recording all entrances from which the smoke exits (Owings and Borchert 1975). As this was infeasible at the CNPRC, we used proximity and common orientation of entrances toward a single source as evidence of being within a single system. Large burrows appeared to be in use at CNPRC for multiple years (Roth, unpublished data). Consistent occupation of burrow systems by squirrels ensures a food source for fleas (Lang 1996), which may explain the association between flea burden and burrow size. Our flea phenology results are somewhat atypical when compared with other studies (Hubbard 1968, Lang 1996, Hubbart et al. 2011). While we did see replacement of O. montana with H. anomalus during the hottest months of the year (Lang 2004), the onset of O. montana activity occurred relatively late in the year. O. montana requires a relative humidity of 65% to complete maturation (Metzger and Rust 1999) and typically reaches highest population numbers on ground squirrels when ambient temperatures are <18.4°C (Lang 1996, Krasnov et al. 2004). 2015 was a drought year for California (National Climatic Data Center 2015b) with an average regional high temperature of 28.9°C in Davis, CA for the month of October (National Climatic Data Center 2015a). Water is readily available at the CNPRC through irrigation and mouth-activated spigots for the NHP’s. Perhaps as a result, we recorded high numbers of H. anomalus late into October (Lang 1996). H. anomalus and E. gallinacea are more xeric-adapted than O. montana (Sammak et al. 2012) and burrow temperature may be as much as 12°C lower than day-time ambient temperatures (Baudinette 1972). Burrow humidity is also relatively independent of ambient surface values (Baudinette 1972). Such conditions may provide an environment that is capable of maintaining active O. montana or other species deep in burrows. Some flea species may survive up to 20 d without a bloodmeal (Krasnov et al. 2002) and often come to the entrance of the burrow to seek a new host once their host has left or perished (Gage 1999). This behavior improves the sensitivity of the swabbing method and thus we expected to see higher flea loads from inactive burrows than active burrows (Gage 1999). We did not observe this result which may be because ground squirrels use some entrances preferentially over others (Grinnell 1918, Salmon et al. 1987), so while an entrance may appear inactive, it does not mean the burrow has been vacated. We also found that the swab technique tended to be more productive when the swab could be inserted ≥0.5 m into the burrow system, possibly because a deeply inserted probe covers more surface area, or because fleas prefer to rest in deeper, cooler tunnels closer to their hosts. There may be a substantial number of fleas associated with ground squirrel burrows even in heavily managed locations. The flea species we recovered are capable of transmitting zoonotic diseases and while fleas are poor vectors for tularemia, they may be present in high enough numbers to compensate for their low vector competence. Closing the burrows reduced H. anomalus burden for at least the next month but did not affect O. montana and C. felis. Filled burrows are often rapidly re-excavated by squirrels, especially near food sources. Previous studies have demonstrated that bait laced with the neonicotinoid insecticide imidacloprid had a nearly 100% efficacy at reducing flea abundance on ground squirrels for at least 29 d (Borchert et al. 2009) and could be an example of a safe and effective option for outdoor research facilities with a history of wildlife-associated diseases. Acknowledgments This project was funded by the Center for Vector Borne Diseases at UC Davis. We thank Brett Farnham and Jaleh Janatpour and the veterinary and the animal care staff at the CNPRC for assistance with this study. We thank Jeffrey Roberts and Christopher Barker for their feedback on the text. References Cited Baldwin R. A. Salmon T. P. Schmidt R. H., and Timm R. M.. 2014. Perceived damage and areas of needed research for wildlife pests of California agriculture. Integrative Zool . 9: 265– 279. Google Scholar CrossRef Search ADS   Baudinette R. V. 1972. Energy metabolism and evaporative water loss in the California ground squirrel. J. Comp. Physiol . 81: 57– 72. Google Scholar CrossRef Search ADS   Bibikova V. A. 1977. Contemporary views on the interrelationships between fleas and the pathogens of human and animal diseases. Ann. Rev. Entomol . 22: 23– 32. Google Scholar CrossRef Search ADS   Bitam I. Dittmar K. Parola P. Whiting M. F., and Raoult D.. 2010. Fleas and flea-borne diseases. Int. J. Infect. Dis . 14: e667– e676. Google Scholar CrossRef Search ADS PubMed  Borchert J. N. Davis R. M., and Poché R. M.. 2009. Field efficacy of rodent bait containing the systemic insecticide imidacloprid against the fleas of California ground squirrels. J. Vector Ecol . 34: 92– 98. Google Scholar CrossRef Search ADS PubMed  Borchert J. N. Eisen R. J. Holmes J. L. Atiku L. A. Mpanga J. T. Brown H. E. Graham C. B. Babi N. Montenieri J. A. Enscore R. E.et al.  . 2012. Evaluation and modification of off-host flea collection techniques used in northwest Uganda: laboratory and field studies. J. Med. Entomol . 49: 210– 214. Google Scholar CrossRef Search ADS PubMed  California Department of Public Health. 2010. Yearly summaries of selected general communicable diseases in California, 2001–2010, http://www. cdph.ca.gov/data/statistics/Pages/YearlySummariesofSelectedGeneralCommunicableDiseasesinCA 2001-2010.aspx, pp. 1– 6 (accessed 14 January 2017). Chomel B. B. Wey A. C., and Kasten R. W.. 2003. Isolation of Bartonella washoensis from a dog with mitral valve endocarditis. J. Clin. Microbiol . 41: 5327– 5332. Google Scholar CrossRef Search ADS PubMed  Eskey C., and Haas V.. 1939. Plague in the western part of the United States: infection in rodents, experimental transmission by fleas and inoculation tests for infection. Publ. Health Rep . 54: 1467– 1481. Google Scholar CrossRef Search ADS   Evans F. Wheeler C., and Douglas J.. 1943. Sylvatic plague studies III. An epizootic of plague among ground squirrels (Citellus beecheyi) in Kern County, California. J. Infect. Dis . 72: 68– 76. Google Scholar CrossRef Search ADS   Ferrecchia C. E. Colgin L. Andrews K. R., and Lewis A. D.. 2012. An outbreak of tularemia in a colony of outdoor-housed rhesus macaques (Macaca mulatta). Comp. Med . 62: 316– 321. Google Scholar PubMed  Foley J., and Nieto N. C.. 2010. Tularemia. Vet. Microbiol . 140: 332– 338. Google Scholar CrossRef Search ADS PubMed  Gage K. L. 1999. Plague surveillance. Plague manual: epidemiology, distribution, surveillance, and control . World Health Organization, Geneva, Switzerland: 135– 165. Grinnell J. 1918. Natural history of the ground squirrels of California . California State Printing Office, Sacramento, CA. Gross B., and Bonnet D. D.. 1949. Snap traps versus cage traps in plague surveillance. Publ. Health Rep . 64: 1214– 1216. Google Scholar CrossRef Search ADS   Higgins J. A. Radulovic S. Jaworski D. C., and Azad A. F.. 1996. Acquisition of the cat scratch disease agent Bartonella henselae by cat fleas (Siphonaptera: Pulicidae). J. Med. Entomol . 33: 490– 495. Google Scholar CrossRef Search ADS PubMed  Hubbard C. A. 1968. Fleas of Western North America. Their relation to Public Health . Hafner Publishing Co., New York, NY. Hubbart J. A. 2012. The California ground squirrel (Spermophilus beecheyi): Characterizing an adaptive fossorial vertebrate for improved science-based management decisions. J. Biol. Life Sci . 3: 1– 12. Hubbart J. A. Jachowski D. S., and Eads D. A.. 2011. Seasonal and among‐site variation in the occurrence and abundance of fleas on California ground squirrels (Otospermophilus beecheyi). J. Vector Ecol . 36: 117– 123. Google Scholar CrossRef Search ADS PubMed  Kartman L., and Prince F. M.. 1956. Studies on Pasteurella pestis in fleas. V. The experimental plague-vector efficiency of wild rodent fleas compared with Xenopsylla cheopis, together with observations on the influence of temperature. Am. J. Trop. Med. Hygiene  5: 1058– 1070. Google Scholar CrossRef Search ADS   Kosoy M. Murray M. Gilmore R. D. Bai Y., and Gage K. L.. 2003. Bartonella strains from ground squirrels are identical to Bartonella washoensis isolated from a human patient. J. Clin. Microbiol . 41: 645– 650. Google Scholar CrossRef Search ADS PubMed  Krasnov B. Khokhlova I., and Shenbrot G.. 2004. Sampling fleas: the reliability of host infestation data. Med. Vet. Entomol . 18: 232– 240. Google Scholar CrossRef Search ADS PubMed  Krasnov B. Khokhlova I. Fielden L., and Burdelova N.. 2002. Time of survival under starvation in two flea species (Siphonaptera: Pulicidae) at different air temperatures and relative humidities. J. Vector Ecol . 27: 70– 81. Google Scholar PubMed  Lang J. D. 1996. Factors affecting the seasonal abundance of ground squirrel and wood rat fleas (Siphonaptera) in San Diego County, California. J. Med. Entomol . 33: 790– 804. Google Scholar CrossRef Search ADS PubMed  Lang J. D. 2004. Rodent-flea-plague relationships at the higher elevations of San Diego County, California. J. Vector Ecol . 29: 236– 247. Google Scholar PubMed  Lewis R. E. Lewis J. H., and Maser C.. 1988. The Fleas of the Pacific Northwest . Oregon State University Press, Corvallis, OR. Matz-Rensing K. Floto A. Schrod A. Becker T. Finke E. Seibold E. Splettstoesser W., and Kaup F.. 2007. Epizootic of tularemia in an outdoor housed group of Cynomolgus monkeys (Macaca fascicularis). Vet. Pathol . 44: 327– 334. Google Scholar CrossRef Search ADS PubMed  McCoy G. W., and Chaplin C. W.. 1912. Further observations on a plague-like disease of rodents with a preliminary note on the causative agent, Bacterium tularense. J. Infect. Dis . 10: 61– 72. Google Scholar CrossRef Search ADS   Metzger M. E., and Rust M. K.. 1999. Abiotic factors affecting the development of fleas (Siphonaptera) of California ground squirrels (Rodentia: Sciuridae) in southern California, USA, pp. 235– 239. In Proceedings, 3rd International Conference on Urban Pests, 19–22 July 1999, Prague, Czech Republic. National Climatic Data Center. 2015a. Record of climatological observations (October 1–31, 2015). National Oceanic and Atmosphere Administration, http://www.ncdc.noaa.gov/cdo-web/datasets (accessed 14 January 2017). National Climatic Data Center. 2015b. Monthly climatological summary (2015). National Oceanic and Atmosphere Administration, http://www.ncdc.noaa.gov/cdo-web/datasets (accessed 14 January 2017). Nelson B. C. 1980. Plague studies in California—the roles of various species of sylvatic rodents in plague ecology in California. In, Proceedings of the 9th Vertebrate Pest Conference (1980) Paper 30, 4–6 March 1980, Lincoln, Nebraska. Nelson B. C. Madon M. B., and Tilzer A.. 1986. The complexities at the interface among domestic/wild rodents, fleas, pets, and man in urban plague ecology in Los Angeles, County, California, p. 46. In, Proceedings of the 12th Vertebrate Pest Conference (1986), 4–6 March 1986, San Diego, CA. Noden B. H. Radulovic S. Higgins J. A., and Azad A. F.. 1998. Molecular identification of Rickettsia typhi and R. felis in co-infected Ctenocephalides felis (Siphonaptera: Pulicidae). J. Med. Entomol . 35: 410– 414. Google Scholar CrossRef Search ADS PubMed  Novak M. A., and Suomi S. J.. 1988. Psychological well-being of primates in captivity. Amer. Psychologist  43: 765– 773. Google Scholar CrossRef Search ADS   Owings D. H., and Borchert M.. 1975. Correlates of burrow location in Beechey ground squirrels. Great Basin Naturalist  35: 402– 404. Parker D. D., and Johnson D. E.. 1957. Experimental transmission of Pasteurella tularensis by the flea, Orchopeas leucopus (Baker). J. Infect. Dis . 101: 69– 72. Google Scholar CrossRef Search ADS PubMed  Preiksaitis J. Crawshaw G. Nayar G., and Stiver H.. 1979. Human tularemia at an urban zoo. Canad. Med. Assoc. J . 121: 1097– 1099. Prince F., and McMahon M.. 1946. Tularemia: attempted transmission by each of the species of fleas: Xenopsylla cheopis (Roths.) and Diamanus montanus (Baker). Publ. Health Rep . 61: 79– 85. Google Scholar CrossRef Search ADS   Probert W. Louie J. K. Tucker J. R. Longoria R. Hogue R. Moler S. Graves M. Palmer H. J. Cassady J., and Fritz C. L.. 2009. Meningitis due to a “Bartonella washoensis”-like human pathogen. J. Clin. Microbiol . 47: 2332– 2335. Google Scholar CrossRef Search ADS PubMed  R Core Team 2015. R: A language and environment for statistical computing computer program, version By R Core Team, Vienna, Austria. http://www.R-project.org/ (accessed 14 January 2017). Reif K. E., and Macaluso K. R.. 2009. Ecology of Rickettsia felis: a review. J. Med. Entomol . 46: 723– 736. Google Scholar CrossRef Search ADS PubMed  Rutledge L. Moussa M. Zeller B., and Lawson M.. 1979. Field studies of reservoirs and vectors of sylvatic plague at Fort Hunter Liggett, California. J. Med. Entomol . 15: 452– 458. Google Scholar CrossRef Search ADS PubMed  Salmon T. P. Marsh R. E., and Stroud D. C.. 1987. Influence of burrow destruction on recolonization by California ground squirrels. Wildl. Soc. Bull . 15: 564– 568. Sammak R. L. Rejmanek D. Roth T. M. Christie K. L. Chomel B. B., and Foley J. E.. 2012. Francisella tularensis outbreak investigation following natural infection of outdoor housed rhesus macaques (Macaca mulatta). Comp. Med . 63: 183– 190. Stewart M., and Evans F.. 1941. Comparative study of rodent and burrow flea populations. Proc. Soc. Exper. Biol. Med . 47: 140– 142. Google Scholar CrossRef Search ADS   Tulis J. Eigelsbach H., and Hronick R.. 1969. Oral vaccination against tularemia in the monkey. Proc. Soc. Exper. Biol. Med . 132: 893– 896. Google Scholar CrossRef Search ADS   Volfertz A., and Kolpakova S.. 1946. The epizootology of tularaemia. Third communication. The role of the fleas Ctenophthalmus orientalis Wagn., in enzootology of tularaemia. Meditsinskaia Parazitologiia i Parazitarnye Bolezni  15: 83– 87. Wimsatt J., and Biggins D. E.. 2009. A review of plague persistance with special emphasis on fleas. J. Vector Borne Dis . 46: 85– 99. Google Scholar PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of Entomological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Medical Entomology Oxford University Press

Prevalence and Seasonality of Fleas Associated With California Ground Squirrels and the Potential Risk of Tularemia in an Outdoor Non-Human Primate Research Facility

Loading next page...
 
/lp/ou_press/prevalence-and-seasonality-of-fleas-associated-with-california-ground-AjvJ2c5ZTC
Publisher
Entomological Society of America
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of Entomological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0022-2585
eISSN
1938-2928
D.O.I.
10.1093/jme/tjx201
Publisher site
See Article on Publisher Site

Abstract

Abstract Ectoparasites at primate research centers may be difficult to control, e.g. without exposing non-human primates (NHPs) to toxicants, but their impact on NHP health is poorly understood. In 2010, there was an epizootic of tularemia at the California National Primate Research Center (CNPRC) in Yolo County, California that resulted in 20 confirmed and suspect clinical cases in outdoors housed rhesus macaques (Macaca mulatta [Zimmermann]) and a 53% seroprevalence in the southern section of the colony. We studied ectoparasite burdens at the CNPRC in order to understand possible conditions at the time of the epizootic and provide data for the management of ectoparasites for the future. In 2015, we recorded 52 California ground squirrel (Otospermophilus beecheyi [Richardson]) burrow systems in the southern colony and collected 560 fleas. The largest number of fleas (n = 184) was collected in October and the most common species were Hoplopsyllus anomalus (Baker) (n = 331), Oropsylla montana (Baker) (n = 158), Echidnophaga gallinacea (Westwood) (n = 60), and Ctenocephalides felis (Bouché) (n = 11), all of which are opportunistically anthropophilic. Free, non-host-associated fleas included 12 H. anomalus, 9 C. felis, 6 O. Montana, and 1 E. gallinacea. We collected 1 H. anomalus from a rhesus macaque. Our results suggest a high potential for the rapid spread of zoonotic infectious diseases via flea transmission in primate facilities with ground squirrels and that flea control measures should be given a high priority. fleas, Hoplopsyllus, anomalus, Echidnophaga, gallinacea, Oropsylla, montana, Ctenocephalides, felis Tularemia is a serious febrile illness caused by the intracellular bacterium Francisella tularensis (Foley and Nieto 2010). As few as 10 colony-forming units are needed to induce fatal septic or pulmonary infection. The bacterium is easily disseminated and has a potentially high fatality rate of up to 40%, leading it to be classified as a Category A select agent by the US Department of Health and Human Services. In 2010, there was an abnormally high incidence of tularemia in the western United States, with seven clinical cases reported compared with a yearly average of 2.8 (California Department of Public Health 2010). In addition to the increase in human cases, non-human primate (NHP) research centers, including the Oregon National Primate Research Center and the California National Primate Research Center (CNPRC) in Yolo County, experienced localized epizootics in NHPs in the same year (Ferrecchia et al. 2012, Sammak et al. 2012). The CNPRC had 20 confirmed and suspected clinical cases of tularemia and a 53% seroprevalence in rhesus macaques (Macaca mulatta) in an area of the CNPRC known as ‘South Colony’. Serological testing confirmed that all of these cases had occurred in 2010 (Sammak et al. 2012). Disease in research NHP populations may destabilize social structure, affect animal welfare, and affect research outcomes, potentially incurring expense through increased veterinary care and the loss of valuable animal genotypes. Although ticks are an important biological vector for F. tularensis (Foley and Nieto 2010), the spatial distribution of cases and wildlife surveys at CNPRC were not supportive of tick transmission (Sammak et al. 2012) and the infection pathway at the CNPRC is unknown. There are a wide variety of vertebrate taxa that can host F. tularensis (Foley and Nieto 2010); however, the agent was first isolated from California ground squirrels (Otospermophilus beecheyi) in Tulare county, California in 1911 (McCoy and Chaplin 1912). California ground squirrels host multiple zoonotic bacterial pathogens including Yersinia pestis, the agent of plague, and Bartonella washoensis which has been associated with human endocarditis and meningitis (McCoy and Chaplin 1912, Evans et al. 1943, Kartman and Prince 1956, Preiksaitis et al. 1979, Nelson 1980, Nelson et al. 1986, Chomel et al. 2003, Kosoy et al. 2003, Hubbart 2012). California ground squirrels have been directly implicated in the amplification and spread of F. tularensis to NHPs (Preiksaitis et al. 1979). They also serve as hosts of several species of flea that are opportunistically anthropophilic (such as Oropsylla montana) and that are capable of mechanically transmitting a variety of pathogens, such as plague (Eskey and Haas 1939, Prince and McMahon 1946, Hubbard 1968). Ground squirrel burrows are often heavily infested with fleas (Rutledge et al. 1979), and each squirrel may support 20 or more fleas at any given time (Metzger and Rust 1999, Hubbart 2012). Fleas are considered to be poor vectors for F. tularensis transmission (Kartman and Prince 1956) but it is possible that repeated bites and large numbers of this vector could overcome their poor vector efficiency. Flea loads on ground squirrels and in ground squirrel burrows have been studied in rural and semi-rural landscapes (Hubbart et al. 2011), but there is little available information on flea communities and load in locations with frequent pest management as occurs at the CNPRC. Our study was performed at the CNPRC South Colony and aimed to determine flea species, whether fleas were in burrows, above-ground, or host-associated, their seasonality, and association of burrow risk factors (evident rodent activity, depth, location, and pest control) with flea presence. Data from this study could inform whether fleas associated with facilities that regularly control for rodent pests, such as the CNPRC, pose a risk for disease transmission and whether pest control methods could better be targeted to mitigate this risk. Materials and methods Study Site The CNPRC is a large research facility in Davis, Yolo County, California that houses an average of 4,000 NHP research animals (predominantly rhesus macaques in indoor and outdoor enclosures). Outside, macaques are housed either in large enclosures of 20–200 individuals in the North Colony or in small group pens of 2–30 individuals in the South Colony. The floor substrate, which could affect ectoparasite presence, in the South Colony enclosures was gravel at the time of the study, while the North Colony enclosures had patches of grass and gravel. Beginning 9 September 2015, the South Colony of the CNPRC was being renovated and multiple NHP enclosures along the eastern edge of the site were removed in order to begin construction on new enclosures with concrete bases. All enclosures were supplied with enrichment items such as barrels, swings, platforms, and ropes and daily dietary items such as fruits and vegetables. Row-crop fields border the CNPRC to the north, west, and south. The eastern edge of the CNPRC is valley oak (Quercus lobata) savannah with non-native annual grasses and weeds. Seasonal standing water collects in multiple reservoirs and drainage ditches in and near the CNPRC. The center of the CNPRC, between the North and South Colony, houses a managed colony of ~32 feral cats. There were three feeding stations in the feral cat colony which were dusted weekly with flea powder (Hartz Mountain Corporation, Secaucus, NJ). Feral cats were trapped and treated annually for fleas with selamectin (Zoetis, Florham Park, NJ) or imidacloprid and pyriproxyfen (Bayer AG, Leverkusen, Germany). A vertebrate pest control program for the CNPRC consisted of live trapping and removal of rodents followed by filling burrow openings, particularly those nearest to primate cages and buildings, with gravel and soil weekly to bi-weekly. Toxicants were deployed rarely and only in ground squirrel burrows located outside the perimeter fence or in a disused fenced-in field located 10 meters from the nearest NHP enclosures. Flea Sampling NHPs were systematically examined for fleas during routine bi-annual health examinations from March to October 2015 and opportunistically from monkeys admitted to the hospital from 1 March 2015 until 31 October 2015. Special attention was paid to each animal’s back, groin, under arms, and behind the ears. The single flea recovered was stored in 70% ethanol. We collected fleas from burrow systems biweekly from 1 March to 31 October 2015 between 8:00 a.m. and 12:00 p.m. A burrow system was defined as a single underground unit with single or multiple entrances. Entrances were considered to be a part of the same burrow system if they were within 3 m and were oriented toward each other, determined by insertion of a 1.5-m long dowel. A global positioning system (GPS) coordinate was taken at the burrow entrance for single-entrance systems or at the center of the burrow system for multiple-entrance systems, determined from directionality indicated by dowels in entrances. Fleas were collected by swabbing burrows using a 20 × 20 cm piece of white flannel attached to the end of a 1.5m long dowel (Gage 1999). Burrows were only swabbed if the dowel could be inserted at least 10 cm. One entrance was swabbed if the burrow system had one to three entrances, two entrances were swabbed for burrow systems with four to five entrances, and three entrances were swabbed for burrow systems with six or more entrances. Fleas were also collected from 3-m linear transects near high human-traffic areas, including near the South Colony office, at a feral cat feeding station, between two storage buildings, between two buildings where NHP transport cages were stored, and between two laboratory buildings (Fig. 1), using a white-sock collection technique (Borchert et al. 2012). An individual wearing white medical scrub tucked into white socks and shirt tucked into pants walked through each transect for 3 min. Fleas were removed from socks and placed in 1.5-ml tubes of 70% ethanol as spotted. Sampling was conducted biweekly from 1 March to 31 October 2015 between 8:00 a.m. and 12:00 p.m. Fleas were also opportunistically collected from scrubs during burrow sampling if they were observed and placed in a separate vial from those collected from the burrow. In regions with high flea activity (such as highly infested burrows), opportunistic collection from scrubs was terminated after 5 min. The collecting individual moved from the area of activity, checked their scrubs a final time, and then proceeded with other activities. Fig. 1. View largeDownload slide The location of transects walked during white-sock collection at the CNPRC from March to November 2015. Fig. 1. View largeDownload slide The location of transects walked during white-sock collection at the CNPRC from March to November 2015. For processing, fleas were cleared in a 10% solution of potassium hydroxide for 24 h and serially dehydrated for 30 min, each in 75, 80, 90, and 100% ethanol. Fleas were then mounted on glass slides using Euparal (BioQuip Products, Rancho Dominguez, CA) and identified to species using keys (Lewis et al. 1988). Analysis GPS points were mapped using Arc GIS 10.2.x (ESRI, Redlands, CA). Data analyses were performed in R (R Core Team 2015) and a P ≤ 0.05 was used a cutoff to infer statistical significance. We calculated the average number of fleas recovered from burrows with the same number of entrances in order to determine whether burrows with more entrances had greater numbers of fleas. In order to account for variation in the number of entrances being swabbed, we also calculated the average numbers of fleas collected per swab (Table 2). Potential risk factors for fleas were recorded as whether the burrow appeared to be in recent active use by a rodent (indicated by soil disruption, feces, food caches or an entryway cleared of plant debris or spider webs, designated ‘Active’ or ‘Inactive’), whether the swab could be inserted ≥0.5 m into the burrow (indicating the presence of a straight/wide passage, designated as ‘Deep’ or ‘Shallow’), whether the burrow was located under a permanent structure such as a tree or building (‘Under’ or ‘Out’), whether the burrow had been filled in with soil or gravel during the previous sampling attempt (‘Open’ or ‘Closed’), and the number of burrow system entrances. We performed Mann–Whitney U tests to determine whether mean numbers of fleas per swab were different between paired risk factors. Results We collected a total of 589 fleas including 560 from ground squirrel burrows. H. anomalus was the most common species collected (n = 331), followed by O. montana (n = 158), Echidnophaga gallinacea (n = 60) and Ctenocephalides felis (n = 11) (Table 1). We collected 28 fleas (12 H. anomalus, 9 C. felis, 6 O. montana, and 1 E. gallinacea) using the modified white-sock collection technique (Table 1). Of the fleas collected on white socks, a single H. anomalus and all C. felis were collected from a region between two buildings where primate cages were stored that was being used by feral cats for shelter (Fig. 1). Despite more than 650 person-hours dedicated to searching NHPs admitted to the hospital for fleas, only one H. anomalus was found. Table 1. Fleas collected from California ground squirrel (O. beecheyi) burrows and from clothing during white-sock collections at the CNPRC from 26 March 2015 to 28 October 2015   3/26  4/9  5/28  6/11  6/24  7/9  7/27  8/9  8/26  9/10  9/23  10/12  10/28  Total  Fleas from burrows  C. felis  0  0  0  0  8  0  0  0  0  2  0  1  0  11  E. gallinacean  0  0  0  0  0  1  4  4  2  5  10  6  28  60  H. anomalus  0  1  1  7  2  49  12  19  33  50  23  82  52  331  O. montana  1  0  19  50  10  38  14  7  5  2  4  1  7  158  Free fleas  C. felis  0  0  0  9  0  0  0  0  0  0  0  0  0  9  E. gallinacean  0  0  0  0  0  0  0  0  0  0  0  0  1  1  H. anomalus  0  0  0  1  0  0  0  0  3  3  0  1  4  12  O. montana  0  0  3  1  0  0  0  0  1  0  0  0  1  6  Total  1  1  23  68  20  88  30  30  44  62  37  91  93  588    3/26  4/9  5/28  6/11  6/24  7/9  7/27  8/9  8/26  9/10  9/23  10/12  10/28  Total  Fleas from burrows  C. felis  0  0  0  0  8  0  0  0  0  2  0  1  0  11  E. gallinacean  0  0  0  0  0  1  4  4  2  5  10  6  28  60  H. anomalus  0  1  1  7  2  49  12  19  33  50  23  82  52  331  O. montana  1  0  19  50  10  38  14  7  5  2  4  1  7  158  Free fleas  C. felis  0  0  0  9  0  0  0  0  0  0  0  0  0  9  E. gallinacean  0  0  0  0  0  0  0  0  0  0  0  0  1  1  H. anomalus  0  0  0  1  0  0  0  0  3  3  0  1  4  12  O. montana  0  0  3  1  0  0  0  0  1  0  0  0  1  6  Total  1  1  23  68  20  88  30  30  44  62  37  91  93  588  No fleas were collected on sampling dates 11 March, 23 April, and 14 May. View Large The month in which we recovered the largest number of fleas was October, with 184 individuals, mostly H. anomalus (n = 134) followed by E. gallinacea (n = 34). The number of O. montana peaked in early June (n = 50) but dropped rapidly and remained low from late July till the end of the collection period. In contrast, H. anomalus numbers rose in July and remained high throughout the remainder of the study period, peaking at 82 individuals collected during a single-collection period in early October. E. gallinacea remained at very low levels until late October when they peaked at 28 individuals collected. To assess the association of rodent burrow characteristics with flea numbers, we recorded 52 distinct rodent burrow systems throughout the South Colony of the CNPRC (Fig. 1). We described the temporality of burrow attributes (e.g., whether they were in use by rodents), the spatial organization of burrows (Fig. 2), whether burrows had been closed by pest control staff the previous collecting period, and characteristics of the burrow entrance such as whether the entrance was located under a permanent structure. The month of June had the highest number (n = 38) of open burrow systems, while March had the fewest (n = 6) (Fig. 2 shows the distribution of burrows during June). About 73% of all burrow systems were single-entrance and were sampled a total of 189 times over the study period. This was followed by two-entrance (11% of all systems) and three-entrance (6.6% of all systems) burrows that were sampled a total of 29 and 17 times, respectively (Table 2). Burrow systems with six or more entrances were the most uncommon (5% of all systems). In general, burrow systems with six or more entrances had higher loads of fleas per swab than burrow systems with less than six entrances (P = 0.006). These categories were determined by plotting a histogram of numbers of burrows with given numbers of openings, and utilizing the natural cutoff between the two bimodal peaks in the data. The average number of fleas recovered per swab was significantly higher from burrows that had been opened the previous sampling period (1.33) compared with burrows that had been closed during the previous sampling period (1.05, P = 0.014). Burrows where the probe could be inserted deeply had a higher average fleas per swab (1.38) compared with from shallow burrows (0.99, P = 0.048). There were no significant differences in the average number of fleas per swab for burrows located under a permanent structure versus open soil areas or lawns or active versus inactive burrows (Table 3). The number of H. anomalus fleas from open burrows (0.69) was greater than those collected from closed burrows (0.62, P = 0.032). We did not detect any effect of these risk factors on the average number of O. montana or E. gallinacea recovered per swab (Table 3). Fig. 2. View largeDownload slide The distribution of open burrows recorded on GPS at the South Colony of the CNPRC on 11 June 2015. Fig. 2. View largeDownload slide The distribution of open burrows recorded on GPS at the South Colony of the CNPRC on 11 June 2015. Table 2. Summary statistics of fleas swabbed from open burrow systems recorded in the South Colony area of the CNPRC in 2015 Number of openings  Number of burrows  Average number of fleas per burrow (SE)  Average number of fleas/swab  Minimum and maximum number of fleas on swabs  Number of burrows with no fleas  1  189  1.322 (0.27)  1.322  0.35  126  2  29  1.24 (0.3)  1.24  0.4  16  3  17  0.7 (0.3)  0.7  0.4  12  4  8  3 (0.93)  1.5  1.9  0  5  3  1.2 (0.97)  0.6  0.5  1  6  1  0 (n/a)  0  0.0  1  7  4  10.5 (10.5)  3.5  0.42  3  8  1  0 (n/a)  0  8.8  0  9  2  9.5 (8.5)  3.167  1.18  0  10  2  37 (35)  12.33  2.72  0  11  1  4 (n/a)  1.33  4.4  0  15  2  28.5 (12.5)  9.5  16.41  0  Number of openings  Number of burrows  Average number of fleas per burrow (SE)  Average number of fleas/swab  Minimum and maximum number of fleas on swabs  Number of burrows with no fleas  1  189  1.322 (0.27)  1.322  0.35  126  2  29  1.24 (0.3)  1.24  0.4  16  3  17  0.7 (0.3)  0.7  0.4  12  4  8  3 (0.93)  1.5  1.9  0  5  3  1.2 (0.97)  0.6  0.5  1  6  1  0 (n/a)  0  0.0  1  7  4  10.5 (10.5)  3.5  0.42  3  8  1  0 (n/a)  0  8.8  0  9  2  9.5 (8.5)  3.167  1.18  0  10  2  37 (35)  12.33  2.72  0  11  1  4 (n/a)  1.33  4.4  0  15  2  28.5 (12.5)  9.5  16.41  0  View Large Table 3. The effects of various risk factors on the mean number of fleas per swab collected at the CNPRC from March to November 2015 Risk factor pairing  Means  P-value  Open vs. Closed  1.33, 1.05  0.014*  Deep vs. Shallow  1.38, 0.99  0.048*  Under vs. Out  1.59, 0.97  0.158  Active vs. Inactive  1.57, 0.86  0.089  O. montana Open vs. O. montana Closed  0.49, 0.33  0.128  O. montana Deep vs. O montana Shallow  0.42, 0.47  0.144  O. montana Under vs. O. montana Out  0.66, 0.27  0.154  O. montana Active vs. O. montana Inactive  0.36, 0.53  0.372  H. anomalus Open vs. H. anomalus Closed  0.699, 0.62  0.032*  H. anomalus Deep vs. H. anomalus Shallow  0.81, 0.44  0.130  H. anomalus Under vs. H. anomalus Out  0.76, 0.61  0.787  H. anomalus Active vs. H. anomalus Inactive  0.15, 0.08  0.286  E. gallinacea Open vs. E. gallinacea Closed  0.13, 0.096  0.344  E. gallinacea Deep vs. E. gallinacea Shallow  0.14, 0.08  0.184  E. gallinacea Under vs. E. gallinacea Out  0.16, 0.09  0.227  E. gallinacea Active vs. E. gallinacea Inactive  0.15, 0.08  0.286  Risk factor pairing  Means  P-value  Open vs. Closed  1.33, 1.05  0.014*  Deep vs. Shallow  1.38, 0.99  0.048*  Under vs. Out  1.59, 0.97  0.158  Active vs. Inactive  1.57, 0.86  0.089  O. montana Open vs. O. montana Closed  0.49, 0.33  0.128  O. montana Deep vs. O montana Shallow  0.42, 0.47  0.144  O. montana Under vs. O. montana Out  0.66, 0.27  0.154  O. montana Active vs. O. montana Inactive  0.36, 0.53  0.372  H. anomalus Open vs. H. anomalus Closed  0.699, 0.62  0.032*  H. anomalus Deep vs. H. anomalus Shallow  0.81, 0.44  0.130  H. anomalus Under vs. H. anomalus Out  0.76, 0.61  0.787  H. anomalus Active vs. H. anomalus Inactive  0.15, 0.08  0.286  E. gallinacea Open vs. E. gallinacea Closed  0.13, 0.096  0.344  E. gallinacea Deep vs. E. gallinacea Shallow  0.14, 0.08  0.184  E. gallinacea Under vs. E. gallinacea Out  0.16, 0.09  0.227  E. gallinacea Active vs. E. gallinacea Inactive  0.15, 0.08  0.286  See text for classifications of risk factors. *Significant value View Large Discussion Outdoor housing of NHPs encourages natural behavior and lowers stress (Novak and Suomi 1988) but may also expose the animals to wildlife-associated diseases (Matz-Rensing et al. 2007, Sammak et al. 2012) which can destabilize social structure, affect animal welfare, and affect research outcomes, potentially incurring expense through increased veterinary care. Disease control measures in NHP research centers must avoid NHP exposure to toxins, whereas studies on ground squirrels and their fleas have focused on areas where toxicants (Baldwin et al. 2014) and burrow destruction (Salmon et al. 1987) have been viable methods for pest control. Here, we characterized the flea community on NHPs and off-host at an NHP facility with a history of tularemia epizootics and found abundant fleas associated with ground squirrel burrows at the facility despite active pest control. We recovered fleas including O. montana, H. anomalus, and C. felis that are aggressive parasites of humans and likely will feed from NHP as well (Lewis et al. 1988, Bitam et al. 2010). O. montana and H. anomalus are vectors of sylvatic plague in the western United States (Stewart and Evans 1941, Wimsatt and Biggins 2009). Ground squirrel fleas have also been implicated in the spread of Bartonella washoensis, a pathogen that has been reported to cause meningitis in humans and mitral valve endocarditis in dogs (Chomel et al. 2003, Probert et al. 2009). C. felis are pests of domestic cats and dogs as well as wildlife such as Virginia opossums (Didelphis virginiana) and are important vectors of Rickettsia felis, the agent of cat-flea typhus (Reif and Macaluso 2009), B. henselae, the agent of cat scratch fever (Higgins et al. 1996), and R. typhi, the agent of murine typhus (Noden et al. 1998). They also act as an intermediate host for the tapeworm Dipylidium caninum (Hubbard 1968, Lewis et al. 1988). Although fleas tend to have low vector competence for F. tularensis (Parker and Johnson 1957, Bibikova 1977), the low infectious dose of this pathogen could contribute to a relatively high vectorial capacity for fleas in a situation where there is a high flea burden (Volfertz and Kolpakova 1946, Kartman and Prince 1956). Despite the large number of fleas collected from the CNPRC grounds, we only recovered one flea from a hospitalized NHP directly. It is unlikely that the fleas cannot access the NHPs or are unwilling to feed on them. More likely, the low number of fleas is due to our inability to detect fleas on them or social grooming. All NHPs are anesthetized before examination and fleas often leave anesthetized hosts due to a dropping body temperature (Gross and Bonnet 1949). If there were substantial delays between initiation of anesthesia and examination, it is possible we could have overlooked fleas. If fleas were removed by grooming, this could still be a transmission route for some pathogens as monkeys often consume groomed ectoparasites. This is supported by the fact that the majority of tularemia cases reported in 2010 were oropharyngeal (Sammak et al. 2012) likely acquired via consumption of contaminated materials. Consistent with this result, oral acquisition of infection is well-documented, with natural outbreaks of food- and water-borne tularemia in both Old and New Worlds (Foley and Nieto 2010), although with a higher infectious dose in animal models than may be required for other transmission routes (Tulis et al. 1969). Most literature regarding arthropods and tularemia, however, suggests transmission via arthropod bite, possibly overlooking the mechanism of arthropod ingestion. In addition to the possibility of oral transmission to monkeys via flea ingestion, the disease may also have been acquired if monkeys ate infected squirrels as well. Flea burden in ground squirrel burrows was associated with the number of burrow openings, which is correlated with the number of squirrels living in the burrow (Owings and Borchert 1975). Our study may not have accurately classified all burrow systems if, e.g., new entrances were built greater than 3 m from the central mound (Grinnell 1918) and we inaccurately classified it as a different burrow system. Some studies solve this problem by pumping smoke into the burrows and recording all entrances from which the smoke exits (Owings and Borchert 1975). As this was infeasible at the CNPRC, we used proximity and common orientation of entrances toward a single source as evidence of being within a single system. Large burrows appeared to be in use at CNPRC for multiple years (Roth, unpublished data). Consistent occupation of burrow systems by squirrels ensures a food source for fleas (Lang 1996), which may explain the association between flea burden and burrow size. Our flea phenology results are somewhat atypical when compared with other studies (Hubbard 1968, Lang 1996, Hubbart et al. 2011). While we did see replacement of O. montana with H. anomalus during the hottest months of the year (Lang 2004), the onset of O. montana activity occurred relatively late in the year. O. montana requires a relative humidity of 65% to complete maturation (Metzger and Rust 1999) and typically reaches highest population numbers on ground squirrels when ambient temperatures are <18.4°C (Lang 1996, Krasnov et al. 2004). 2015 was a drought year for California (National Climatic Data Center 2015b) with an average regional high temperature of 28.9°C in Davis, CA for the month of October (National Climatic Data Center 2015a). Water is readily available at the CNPRC through irrigation and mouth-activated spigots for the NHP’s. Perhaps as a result, we recorded high numbers of H. anomalus late into October (Lang 1996). H. anomalus and E. gallinacea are more xeric-adapted than O. montana (Sammak et al. 2012) and burrow temperature may be as much as 12°C lower than day-time ambient temperatures (Baudinette 1972). Burrow humidity is also relatively independent of ambient surface values (Baudinette 1972). Such conditions may provide an environment that is capable of maintaining active O. montana or other species deep in burrows. Some flea species may survive up to 20 d without a bloodmeal (Krasnov et al. 2002) and often come to the entrance of the burrow to seek a new host once their host has left or perished (Gage 1999). This behavior improves the sensitivity of the swabbing method and thus we expected to see higher flea loads from inactive burrows than active burrows (Gage 1999). We did not observe this result which may be because ground squirrels use some entrances preferentially over others (Grinnell 1918, Salmon et al. 1987), so while an entrance may appear inactive, it does not mean the burrow has been vacated. We also found that the swab technique tended to be more productive when the swab could be inserted ≥0.5 m into the burrow system, possibly because a deeply inserted probe covers more surface area, or because fleas prefer to rest in deeper, cooler tunnels closer to their hosts. There may be a substantial number of fleas associated with ground squirrel burrows even in heavily managed locations. The flea species we recovered are capable of transmitting zoonotic diseases and while fleas are poor vectors for tularemia, they may be present in high enough numbers to compensate for their low vector competence. Closing the burrows reduced H. anomalus burden for at least the next month but did not affect O. montana and C. felis. Filled burrows are often rapidly re-excavated by squirrels, especially near food sources. Previous studies have demonstrated that bait laced with the neonicotinoid insecticide imidacloprid had a nearly 100% efficacy at reducing flea abundance on ground squirrels for at least 29 d (Borchert et al. 2009) and could be an example of a safe and effective option for outdoor research facilities with a history of wildlife-associated diseases. Acknowledgments This project was funded by the Center for Vector Borne Diseases at UC Davis. We thank Brett Farnham and Jaleh Janatpour and the veterinary and the animal care staff at the CNPRC for assistance with this study. We thank Jeffrey Roberts and Christopher Barker for their feedback on the text. References Cited Baldwin R. A. Salmon T. P. Schmidt R. H., and Timm R. M.. 2014. Perceived damage and areas of needed research for wildlife pests of California agriculture. Integrative Zool . 9: 265– 279. Google Scholar CrossRef Search ADS   Baudinette R. V. 1972. Energy metabolism and evaporative water loss in the California ground squirrel. J. Comp. Physiol . 81: 57– 72. Google Scholar CrossRef Search ADS   Bibikova V. A. 1977. Contemporary views on the interrelationships between fleas and the pathogens of human and animal diseases. Ann. Rev. Entomol . 22: 23– 32. Google Scholar CrossRef Search ADS   Bitam I. Dittmar K. Parola P. Whiting M. F., and Raoult D.. 2010. Fleas and flea-borne diseases. Int. J. Infect. Dis . 14: e667– e676. Google Scholar CrossRef Search ADS PubMed  Borchert J. N. Davis R. M., and Poché R. M.. 2009. Field efficacy of rodent bait containing the systemic insecticide imidacloprid against the fleas of California ground squirrels. J. Vector Ecol . 34: 92– 98. Google Scholar CrossRef Search ADS PubMed  Borchert J. N. Eisen R. J. Holmes J. L. Atiku L. A. Mpanga J. T. Brown H. E. Graham C. B. Babi N. Montenieri J. A. Enscore R. E.et al.  . 2012. Evaluation and modification of off-host flea collection techniques used in northwest Uganda: laboratory and field studies. J. Med. Entomol . 49: 210– 214. Google Scholar CrossRef Search ADS PubMed  California Department of Public Health. 2010. Yearly summaries of selected general communicable diseases in California, 2001–2010, http://www. cdph.ca.gov/data/statistics/Pages/YearlySummariesofSelectedGeneralCommunicableDiseasesinCA 2001-2010.aspx, pp. 1– 6 (accessed 14 January 2017). Chomel B. B. Wey A. C., and Kasten R. W.. 2003. Isolation of Bartonella washoensis from a dog with mitral valve endocarditis. J. Clin. Microbiol . 41: 5327– 5332. Google Scholar CrossRef Search ADS PubMed  Eskey C., and Haas V.. 1939. Plague in the western part of the United States: infection in rodents, experimental transmission by fleas and inoculation tests for infection. Publ. Health Rep . 54: 1467– 1481. Google Scholar CrossRef Search ADS   Evans F. Wheeler C., and Douglas J.. 1943. Sylvatic plague studies III. An epizootic of plague among ground squirrels (Citellus beecheyi) in Kern County, California. J. Infect. Dis . 72: 68– 76. Google Scholar CrossRef Search ADS   Ferrecchia C. E. Colgin L. Andrews K. R., and Lewis A. D.. 2012. An outbreak of tularemia in a colony of outdoor-housed rhesus macaques (Macaca mulatta). Comp. Med . 62: 316– 321. Google Scholar PubMed  Foley J., and Nieto N. C.. 2010. Tularemia. Vet. Microbiol . 140: 332– 338. Google Scholar CrossRef Search ADS PubMed  Gage K. L. 1999. Plague surveillance. Plague manual: epidemiology, distribution, surveillance, and control . World Health Organization, Geneva, Switzerland: 135– 165. Grinnell J. 1918. Natural history of the ground squirrels of California . California State Printing Office, Sacramento, CA. Gross B., and Bonnet D. D.. 1949. Snap traps versus cage traps in plague surveillance. Publ. Health Rep . 64: 1214– 1216. Google Scholar CrossRef Search ADS   Higgins J. A. Radulovic S. Jaworski D. C., and Azad A. F.. 1996. Acquisition of the cat scratch disease agent Bartonella henselae by cat fleas (Siphonaptera: Pulicidae). J. Med. Entomol . 33: 490– 495. Google Scholar CrossRef Search ADS PubMed  Hubbard C. A. 1968. Fleas of Western North America. Their relation to Public Health . Hafner Publishing Co., New York, NY. Hubbart J. A. 2012. The California ground squirrel (Spermophilus beecheyi): Characterizing an adaptive fossorial vertebrate for improved science-based management decisions. J. Biol. Life Sci . 3: 1– 12. Hubbart J. A. Jachowski D. S., and Eads D. A.. 2011. Seasonal and among‐site variation in the occurrence and abundance of fleas on California ground squirrels (Otospermophilus beecheyi). J. Vector Ecol . 36: 117– 123. Google Scholar CrossRef Search ADS PubMed  Kartman L., and Prince F. M.. 1956. Studies on Pasteurella pestis in fleas. V. The experimental plague-vector efficiency of wild rodent fleas compared with Xenopsylla cheopis, together with observations on the influence of temperature. Am. J. Trop. Med. Hygiene  5: 1058– 1070. Google Scholar CrossRef Search ADS   Kosoy M. Murray M. Gilmore R. D. Bai Y., and Gage K. L.. 2003. Bartonella strains from ground squirrels are identical to Bartonella washoensis isolated from a human patient. J. Clin. Microbiol . 41: 645– 650. Google Scholar CrossRef Search ADS PubMed  Krasnov B. Khokhlova I., and Shenbrot G.. 2004. Sampling fleas: the reliability of host infestation data. Med. Vet. Entomol . 18: 232– 240. Google Scholar CrossRef Search ADS PubMed  Krasnov B. Khokhlova I. Fielden L., and Burdelova N.. 2002. Time of survival under starvation in two flea species (Siphonaptera: Pulicidae) at different air temperatures and relative humidities. J. Vector Ecol . 27: 70– 81. Google Scholar PubMed  Lang J. D. 1996. Factors affecting the seasonal abundance of ground squirrel and wood rat fleas (Siphonaptera) in San Diego County, California. J. Med. Entomol . 33: 790– 804. Google Scholar CrossRef Search ADS PubMed  Lang J. D. 2004. Rodent-flea-plague relationships at the higher elevations of San Diego County, California. J. Vector Ecol . 29: 236– 247. Google Scholar PubMed  Lewis R. E. Lewis J. H., and Maser C.. 1988. The Fleas of the Pacific Northwest . Oregon State University Press, Corvallis, OR. Matz-Rensing K. Floto A. Schrod A. Becker T. Finke E. Seibold E. Splettstoesser W., and Kaup F.. 2007. Epizootic of tularemia in an outdoor housed group of Cynomolgus monkeys (Macaca fascicularis). Vet. Pathol . 44: 327– 334. Google Scholar CrossRef Search ADS PubMed  McCoy G. W., and Chaplin C. W.. 1912. Further observations on a plague-like disease of rodents with a preliminary note on the causative agent, Bacterium tularense. J. Infect. Dis . 10: 61– 72. Google Scholar CrossRef Search ADS   Metzger M. E., and Rust M. K.. 1999. Abiotic factors affecting the development of fleas (Siphonaptera) of California ground squirrels (Rodentia: Sciuridae) in southern California, USA, pp. 235– 239. In Proceedings, 3rd International Conference on Urban Pests, 19–22 July 1999, Prague, Czech Republic. National Climatic Data Center. 2015a. Record of climatological observations (October 1–31, 2015). National Oceanic and Atmosphere Administration, http://www.ncdc.noaa.gov/cdo-web/datasets (accessed 14 January 2017). National Climatic Data Center. 2015b. Monthly climatological summary (2015). National Oceanic and Atmosphere Administration, http://www.ncdc.noaa.gov/cdo-web/datasets (accessed 14 January 2017). Nelson B. C. 1980. Plague studies in California—the roles of various species of sylvatic rodents in plague ecology in California. In, Proceedings of the 9th Vertebrate Pest Conference (1980) Paper 30, 4–6 March 1980, Lincoln, Nebraska. Nelson B. C. Madon M. B., and Tilzer A.. 1986. The complexities at the interface among domestic/wild rodents, fleas, pets, and man in urban plague ecology in Los Angeles, County, California, p. 46. In, Proceedings of the 12th Vertebrate Pest Conference (1986), 4–6 March 1986, San Diego, CA. Noden B. H. Radulovic S. Higgins J. A., and Azad A. F.. 1998. Molecular identification of Rickettsia typhi and R. felis in co-infected Ctenocephalides felis (Siphonaptera: Pulicidae). J. Med. Entomol . 35: 410– 414. Google Scholar CrossRef Search ADS PubMed  Novak M. A., and Suomi S. J.. 1988. Psychological well-being of primates in captivity. Amer. Psychologist  43: 765– 773. Google Scholar CrossRef Search ADS   Owings D. H., and Borchert M.. 1975. Correlates of burrow location in Beechey ground squirrels. Great Basin Naturalist  35: 402– 404. Parker D. D., and Johnson D. E.. 1957. Experimental transmission of Pasteurella tularensis by the flea, Orchopeas leucopus (Baker). J. Infect. Dis . 101: 69– 72. Google Scholar CrossRef Search ADS PubMed  Preiksaitis J. Crawshaw G. Nayar G., and Stiver H.. 1979. Human tularemia at an urban zoo. Canad. Med. Assoc. J . 121: 1097– 1099. Prince F., and McMahon M.. 1946. Tularemia: attempted transmission by each of the species of fleas: Xenopsylla cheopis (Roths.) and Diamanus montanus (Baker). Publ. Health Rep . 61: 79– 85. Google Scholar CrossRef Search ADS   Probert W. Louie J. K. Tucker J. R. Longoria R. Hogue R. Moler S. Graves M. Palmer H. J. Cassady J., and Fritz C. L.. 2009. Meningitis due to a “Bartonella washoensis”-like human pathogen. J. Clin. Microbiol . 47: 2332– 2335. Google Scholar CrossRef Search ADS PubMed  R Core Team 2015. R: A language and environment for statistical computing computer program, version By R Core Team, Vienna, Austria. http://www.R-project.org/ (accessed 14 January 2017). Reif K. E., and Macaluso K. R.. 2009. Ecology of Rickettsia felis: a review. J. Med. Entomol . 46: 723– 736. Google Scholar CrossRef Search ADS PubMed  Rutledge L. Moussa M. Zeller B., and Lawson M.. 1979. Field studies of reservoirs and vectors of sylvatic plague at Fort Hunter Liggett, California. J. Med. Entomol . 15: 452– 458. Google Scholar CrossRef Search ADS PubMed  Salmon T. P. Marsh R. E., and Stroud D. C.. 1987. Influence of burrow destruction on recolonization by California ground squirrels. Wildl. Soc. Bull . 15: 564– 568. Sammak R. L. Rejmanek D. Roth T. M. Christie K. L. Chomel B. B., and Foley J. E.. 2012. Francisella tularensis outbreak investigation following natural infection of outdoor housed rhesus macaques (Macaca mulatta). Comp. Med . 63: 183– 190. Stewart M., and Evans F.. 1941. Comparative study of rodent and burrow flea populations. Proc. Soc. Exper. Biol. Med . 47: 140– 142. Google Scholar CrossRef Search ADS   Tulis J. Eigelsbach H., and Hronick R.. 1969. Oral vaccination against tularemia in the monkey. Proc. Soc. Exper. Biol. Med . 132: 893– 896. Google Scholar CrossRef Search ADS   Volfertz A., and Kolpakova S.. 1946. The epizootology of tularaemia. Third communication. The role of the fleas Ctenophthalmus orientalis Wagn., in enzootology of tularaemia. Meditsinskaia Parazitologiia i Parazitarnye Bolezni  15: 83– 87. Wimsatt J., and Biggins D. E.. 2009. A review of plague persistance with special emphasis on fleas. J. Vector Borne Dis . 46: 85– 99. Google Scholar PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of Entomological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

Journal of Medical EntomologyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off