Abiotic and Biotic Contributors to Support Inter-Epidemic Francisella tularensis in an Agricultural Peri-Urban Environment

Abstract

To characterize the inter-epidemic ecology of Francisella tularensis, we surveyed vertebrates and invertebrates for the abundance, spatial distribution, and status of infection at a site in northern California that had evidence of endemic type B tularemia. We collected 2910 mosquitoes, 77 biting flies, 704 ticks, 115 mammals, and 1911 aquatic invertebrates in 2013–2014. Real-time PCR on all mosquitoes, 40 biting flies, 113 aquatic invertebrates, and 650 ticks did not detect F. tularensis DNA. Indirect enzyme linked immunosorbent assay (ELISA) on 109 mammals revealed 2 (of 2, 100%) seropositive feral cats, 1 (of 24, 4.5%) seropositive black rat, and 5 (of 10, 50%) seropositive Virginia opossums. A riparian reserve, ∼1 km from the primate research center, had the highest seroprevalence in mammals and the highest capture success for invertebrate vectors whereas opossums, cats, and ground squirrels in close proximity to the primate center had high seroprevalence and abundant fleas. Well-vegetated regions with standing water appeared to be ideal habitats for biotic components of tularemia enzootic persistence. Mesocarnivores may facilitate the spread of F. tularensis, and high densities of rodents and their fleas may be a mechanism for amplification and spillover.

Introduction

F rancisella tularensis, the bacterium that causes tularemia, is highly infectious, potentially lethal, and easily disseminated by vectors, through air, or via contaminated water, soil, or meat (Foley and Nieto 2010). Of the four recognized subspecies, two (F. tularensis tularensis and F. tularensis holarctica) are considered pathogenic to humans, with case fatality rates of up to 30% or 5%, respectively, if untreated (Center for Food Safety and Public Health 2009, Petersen et al. 2009). The agent of type B tularemia, F. tularensis holarctica, is distributed across the Holarctic (Ellis et al. 2002). It has been detected in more than 300 species of vertebrates and invertebrates (Keim et al. 2007, Foley and Nieto 2010, Akimana and Kwaik 2011) and may persist as a biofilm (Thelaus et al. 2009, van Hoek 2013). In North America, Dermacentor and Amblyomma spp. ticks, Tabanus and Chrysops spp. flies, and muskrats (Ondatra zibethicus), beavers (Castor canadensis), and voles (Microtus spp.) are implicated in maintenance of F. tularensis holarctica (Francis and Mayne 1921, Jellison et al. 1942, Foley and Nieto 2010). Feline cases occur with high rates of mortality as well as possible zoonotic risks as well (Gliatto et al. 1994, Woods et al. 1998). With so many possible vectors and hosts, the mechanisms for persistence are poorly understood.

In California, 2010 was notable for seven reported human cases of tularemia, compared with a yearly average of 2.8 (California Department of Public Health 2010). An outbreak of 20 cases of type B tularemia occurred in outdoor-housed rhesus macaques (Macaca mulatta) at the California National Primate Research Center (CNPRC) (Sammak et al. 2012). Tissue samples from nonhuman primates (NHP) at the CNPRC from the 1980s and 1990s revealed past cases, suggesting that tularemia is endemic (Sammak et al. 2012). We aimed at describing the inter-epidemic ecology at the CNPRC through surveys of vectors and hosts in the surrounding region.

Materials and Methods

Site description

Our study site encompasses 1.9 km² from 121°48′45.02″W to 121°48′12.25″W and from 38°32′48.41″N to 38°31′34.68″N (Fig. 1), including the CNPRC. The fence surrounding the CNPRC deters large wildlife but allows entry of raccoons (Procyon lotor), opossums (Didelphis virginiana), cats (Felis catus), and small mammals. There is a colony of ∼32 feral cats in the center of the CNPRC that is treated annually with selamectin (Zoetis, Florham Park, New Jersey), imidacloprid, or pyriproxyfen (Bayer AG, Leverkusen, Germany), and feeding stations are dusted weekly with flea powder containing phenothrin and pyriproxyfen (Hartz Mountain Corporation, Secaucus, NJ).

FIG. 1.

Map of five transects surrounding the California National Primate Research Center showing relevant features and placement of stations and transects as well as capture success rates of mosquitoes, biting flies, and vertebrates (mesocarnivores and rodents), and seroprevalence among vertebrates at each station.

Russell Marsh, north of the CNPRC, has willow (Salix sp.), sedges, and rushes fed by agricultural run-off, and it contains water seasonally during wet years. Immediately north, west, and south of the CNPRC are row crops whereas the eastern edge is bordered by a road. Field and road edges are populated with annual grasses and valley oaks (Quercus lobata). South of the fields is the solid waste facility that was not sampled and farthest south is the Putah Creek Riparian Reserve, with willows, valley oak, black walnut (Juglans hindsii), eucalyptus (Eucalyptus spp.), grasses, thistles, Himalayan blackberry (Rubus armeniacus), and wild oats (Avena fatua) (Fulks 2005). Wildlife include mule deer (Odocoileus hemionus), beavers, raccoons, opossums, squirrels (Sciurus and Otospermophilus spp.), and other small mammals (Fulks 2005).

Transects and stations

We ran five linear, east-west transects across the study site (Fig. 1). The Russell Marsh transect ran through agricultural fields and marsh, North CNPRC transect encompassed lawn between large NHP enclosures, and the South transect included the smaller NHP enclosures and a fallow field with fruit trees. The Dump transect ran along a gravel road to a large wood pile. The Putah Creek transect ran along a gravel road parallel to Putah Creek and then down to the water. Each transect had a CDC/CO₂-baited mosquito trap and five vertebrate traps at each end and in the center.

Arthropod collection

Three Townes traps (John W. Hock Co., Gainesville, Florida) loaded with 70% ethanol were run for 3 days/month from January 2013 to December 2014 at Russell Marsh, South CNPRC, and Putah Creek. Organisms were identified by using published references (Middlekauff and Lane 1980, MacAlpine et al. 1981, Meyer and Durso 1993, Triplehorn and Johnson 2005). CDC CO₂/light and dry ice-baited traps were utilized during periods of no rain, each sheltered from wind 1.0–1.5 m off the ground for 24 h monthly from January 2013 to December 2014. Arthropods were frozen at −80°C, identified to species, and finally preserved in 70% ethanol.

Relative tick abundance was measured monthly by dragging a 1 m² white flannel cloth for 10 min across vegetation along trails or roads at Russell Marsh, North CNPRC, South CNPRC, the Dump, and Putah Creek. Additional flagging (not included in quantitative comparisons) was conducted to increase sample size. Ticks were placed in 70% ethanol and identified to species by using a key (Furman and Loomis 1984).

We sampled aquatic larvae in 350 mL of water from every 1–2 m along the periphery of Putah Creek and Russell Marsh. Samples were filtered across a 0.5 mm mesh and suspended in 70% isopropanol. Invertebrates were identified to family, categorized as shredder, scraper, filterer, parasite, gatherer, piercer, or predator (Pennak 1989), and placed in 70% ethanol. If identification required microscopy, a sample was mounted on a slide by using Euparal (BioQuip Products, Rancho Dominguez, California). Filterers, scrapers, gatherers, and invertebrates that are hematophagous as adults were retained for PCR testing for F. tularensis DNA.

Vertebrate trapping

Animals were trapped for 3 days monthly from January 2013 to October 2013 at each station and then for 5 days in March, June, September, and December from January 2014 to December 2014. Three small Sherman traps (HB Sherman Traps, Tallahassee, Florida) and two small wire traps (Tomahawk Live Traps, Hazelhurst, Wisconsin) were baited with oatmeal and peanut butter. Raccoon-sized traps (Tomahawk Live Traps) were placed in shade near mesocarnivore pathways or fecal latrines at stations 1, 3, 10, 12, 13, 14, and 15 and baited with cat food. Rodents were anesthetized with 22 mg/kg ketamine and 2.4 mg/kg xylazine intramuscularly (IM), and blood was collected via retro-orbital abrasion into ethylenediaminetetraacetic acid (EDTA). Ectoparasites were removed with forceps, and an individually numbered metal ear tag was placed (National Brand and Tag Company, Newport, Kentucky). Mesocarnivores were anesthetized by using 10 mg/kg ketamine and 0.5 mg/kg midazolam IM, and blood was collected via cephalic or femoral venipuncture into EDTA. Ectoparasite collection and ear tagging were conducted as for rodents. Notes on species, sex, age, body condition score (Ullman-Cullere and Foltz 1999), and lesions or injuries were taken. Animals were released at the trap location. Animals that were recaptured within a trapping period were released without sampling.

Carcasses obtained from CNPRC pest control or vehicular strikes were examined for ectoparasites and sampled for blood, spleen, liver, and kidney. These samples were not included in comparing station or transect capture numbers.

Molecular testing

DNA from individual ticks and tabanid flies was extracted by using an ammonium hydroxide method (Humair et al. 2007). Mosquitoes, simulid and ceratopogonid flies, and aquatic invertebrates were pooled in groups of 1–12 individuals for ammonium hydroxide extraction. Real-time PCR for F. tularensis tul4 (Versage et al. 2003), which would amplify agents of both Type A and type B tularemia, was conducted on all samples incorporating water negative controls and a positive control of DNA from a culture-positive macaque. Cycle threshold values below 40 with characteristic amplification curves were considered positive.

Serology was conducted by enzyme linked immunosorbent assay (ELISA) (Virion/Serion, Würzberg, Germany), a commercially available test kit with >95% sensitivity and specificity (Chaignat et al. 2014) following manufacturer's instructions except using alkaline phosphatase-conjugated species-specific secondary antibodies from KPL (Gaithersburg, Maryland) or Bethyl Laboratories (Montgomery, Texas). Anti-dog secondary antibodies were used for raccoons and skunks, and anti-rat was used for voles. An absorptive value of >0.42 nm was considered positive, and 0.29–0.42 nm was considered borderline.

Statistical methods

Statistical analyses were conducted in R (R Core Team, Vienna, Austria), with a value of p < 0.05 considered statistically significant. Maps were created by using ArcGIS version 10.3 (ESRI, Redlands, CA) (Fig. 1). The distance from each station center to the nearest standing water was estimated by using Google Earth (Mountain View, CA). We created a color code for “high” or “low” capture success, PCR prevalence, and seroprevalence by using natural breaks observed on scatterplots. Biting flies other than mosquitoes were grouped together for analysis. Ticks and arthropods collected from Townes malaise traps were not included in this analysis. Chi-square tests were performed to determine whether numbers of mosquitoes, biting flies, ticks, vertebrates, and seropositive vertebrates differed among stations. We performed logistic regression to assess association of seropositivity by transect with mean distance to standing water and number of mesocarnivores, mosquitoes, biting flies, or vertebrates.

Results

Invertebrates

CO₂/light-baited traps at five transects recovered 2699 mosquitoes over 22 sampling days, predominantly from Putah Creek (1667) (Table 1). Culex tarsalis was the most abundant species (n = 2287) followed by C. pipiens (n = 109). Five Simulium donovani were collected at South CNPRC, and 28 ceratopogonids were collected primarily from Russell Marsh. Putah Creek had the greatest species richness (Table 1). Townes traps recovered 120 mosquitoes, 11 Tabanus similis, 1 T. punctifer, 34 S. donovani, and 3 Culicoides spp. over 72 sampling days. Mosquitoes and S. donovani (n = 30) in Townes traps were significantly more common at South CNPRC (X² = 3156.2, p < 2.2 × 10⁻¹⁶) whereas biting flies were most common at Russell Marsh (n = 9, X² = 53.52, p = 6.6 × 10⁻¹¹), the only site from which T. punctifer was collected (Table 2).

Table 1.

Number and Species Richness of Invertebrates Collected from CO₂-Baited Traps Placed at Five Transects at the Study Region in Yolo County, CA, from 2013 to 2014

Transect no.	1			2			3			4			5
Station no.	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	Species total
Aedes nigromaculus	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	1
Aedes sierriensis	0	0	0	0	0	0	0	0	1	0	0	0	1	1	0	3
Aedes washinoi	0	1	0	0	0	0	0	0	2	0	0	0	0	0	1	4
Anopheles franciscanus	0	3	0	0	0	0	0	0	0	0	1	0	1	0	1	6
Anopheles freeborni	3	1	4	2	0	1	0	0	0	1	0	0	3	0	7	22
Ceratopogonidae	23	0	0	0	0	0	0	0	0	1	1	2	1	0	0	28
Culex erythrothorax	0	0	13	5	0	1	0	0	0	0	0	0	4	3	0	26
Culex pipiens	3	3	18	5	3	19	0	0	27	2	0	5	8	4	12	109
Culex reevesi	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	1
Culex sp.	3	94	4	0	0	4	0	0	4	0	1	0	5	1	22	138
Culex stigmatosoma	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1
Culex tarsalis	77	78	186	15	11	164	9	1	182	10	4	8	551	180	811	2287
Culiseta incidens	0	0	0	0	1	0	0	0	2	0	0	1	1	2	8	15
Culiseta inornata	7	7	6	1	1	1	1	1	5	3	3	1	1	0	2	40
Males (not identified)	0	0	4	0	0	0	0	0	6	0	0	0	1	0	35	46
Simulium donovani	0	0	0	0	0	0	0	0	5	0	0	0	0	0	0	5
Total Collected_STA	117	188	235	28	16	190	10	2	234	17	10	17	578	191	899	2732
Species Richness_STA	7	8	7	5	4	6	2	2	9	5	5	5	12	6	9
Total Collected_TRAN	540			234			246			44			1668
Species Richness_TRAN	11			7			9			8			13

STA, the values reported are by station; TRAN, the values are reported by transect.

Table 2.

Number and Species Richness of Invertebrates Collected from Malaise Traps Placed on the Russell Marsh, South California National Primate Research Center and Putah Creek Transects at the Study Region in Yolo County, CA, from 2013 to 2014

Transect name	Russell Marsh	South CNPRC	Putah Creek	Species totals
Aedes washinoi	1	2	0	3
Anopheles freeborni	3	4	9	16
Culex erythrothorax	0	4	0	4
Culex pipiens	8	8	8	24
Culex sp.	0	3	0	3
Culex tarsalis	2	24	10	36
Culicoides copiosus	2	0	0	2
Culicoides varipennis	0	1	0	1
Culiseta incidens	0	0	2	2
Culiseta inornata	2	1	1	4
Male Culicidae	2	23	3	28
Simulium donovani	4	30	0	34
Tabanus punctifer	1	0	0	1
Tabanus similis	9	0	2	11
Total by Transect	34	100	35	169
Species Richness	10	10	7	—

CNPRC, California National Primate Research Center.

While performing timed tick flagging, ticks were recovered at Russell Marsh and Putah Creek in 2013 (n = 2 and 44 respectively) and at Putah Creek in 2014 (n = 50). Supplementary collection yielded 604 ticks −282 from hosts and 322 from flags. Among all ticks, 585 were Dermacentor variabilis (50 from opossums, 1 from a cat, 14 from striped skunks [Mephitis mephitis], 197 from raccoons, 1 from a black rat, 1 from a western harvest mouse [Reithrodontomys megalotis], and the rest flagged), 3 were D. occidentalis (1 each from a raccoon, researcher, and flag), 1 was Ixodes pacificus (from a raccoon), and 15 were Haemaphysalis leporispalustris (all from brush rabbits [Sylvilagus bachmani]). There were significant differences in the numbers of ticks between transects (X² = 364.4, p < 2.2 × 10⁻¹⁶), with most found at Putah Creek.

We collected 122 fleas, including 15 from cats, 55 from California ground squirrels (Otospermophilus beecheyi), 10 from rats, 7 from raccoons, and 35 from opossums (Table 3). The greatest number was from Russell Marsh (n = 17), followed by Putah Creek (n = 15). Cats were parasitized by Ctenocephalides felis, Echidnophaga gallinacea, and Pulex sp.; ground squirrels by C. felis, E. gallinacae, Hoplopsyllus anomalus, and Oropsylla montana; rats by Nosopsyllus fasciatus; raccoons by C. felis and Pulex sp.; and opossums by all species of flea. Fleas in the CNPRC included C. felis, E. gallinacea, H. anomalus, N. fasciatus, P. simulans, and O. montana.

Table 3.

The Number and Species of Fleas and Their Corresponding Hosts Collected from Five Transects Surrounding the California National Primate Research Center in Yolo County, CA, from January 2013 to December 2014

	Didelphis virginiana	Otospermophilus beecheyi	Procyon lotor	Rattus rattus	Felis catus	Total
Ctenocephalides felis	19	1	5	0	12	37
Echidnophaga gallinacea	4	1	0	0	2	7
Hoplopsyllus anomalus	1	34	0	0	0	35
Nosopsyllus fasciatus	3	0	0	10	0	13
Pulex simulans	2	0	0	0	0	2
Pulex sp.	5	0	2	0	1	8
Oropsylla montana	1	19	0	0	0	20
Total	35	56	7	10	15	122

Aquatic invertebrate sampling yielded 1911 individuals from 68 families, most commonly Chironomidae (n = 949), Corixidae (n = 161), and Physidae (n = 100).

Vertebrates

Over 56 sampling days, we trapped: 14 opossums, 1 striped skunk, 5 California voles (Microtus californicus), 38 house mice (Mus musculus), 3 California ground squirrels, 1 brush mouse (Peromyscus boylii), 24 black rats, 11 western harvest mice, and 1 eastern gray squirrel (Sciurus carolinensis) (Table 4). Species richness is reported in Table 4. There were significantly more animals trapped at Russell Marsh (n = 28) and Putah Creek transect (n = 24) than other stations (X² = 42.7, p = 1.18 × 10⁻⁸). Harvest, brush, and house mice were most commonly trapped at the Dump, mesocarnivores and rats at Putah Creek, and ground squirrels at the CNPRC. Pest control activities and vehicular strikes yielded opossums, a cat, jackrabbits (Lepus californicus), ground squirrels, and raccoons (Table 5). The highest overall capture success of all animal and invertebrate taxa was at stations 1 at Russell Marsh, and 13 and 15 at Putah Creek.

Table 4.

Number and Species Richness of Live-Trapped Animals on Each Station and Transect Collected from 2013 to 2014 from Five Transects Surrounding the California National Primate Research Center in Yolo County, CA

	Transect 1			Transect 2			Transect 3			Transect 4			Transect 5			CNPRC
Stations	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	CEN
American possum (D. virginiana)	1	1	1	0	0	1	0	1	0	0	0	0	4	0	5	2
Domestic cat (F. catus)	0	0	1	0	0	0	0	0	0	0	0	1	0	0	0	1
Black-tailed jackrabbit (L. californicus)	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2
Striped skunk (Mephitis mephitis)	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
California vole (Microtus californicus)	0	0	0	0	0	0	0	0	0	1	1	0	0	1	2	0
House mouse (Mus musculus)	16	0	1	0	1	0	0	0	0	9	7	2	1	1	0	0
California ground squirrel (O. beecheyi)	0	0	0	0	1	0	0	0	2	0	0	0	0	0	0	9
Brush mouse (P. boylii)	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0
Raccoon (P. lotor)	0	0	7	0	0	0	0	1	0	0	0	0	3	0	4	2
Black rat (R. rattus)	4	0	1	0	0	0	0	0	0	2	0	0	6	3	8	0
Western harvest mouse (R. megalotis)	3	1	1	0	0	0	0	0	0	2	1	2	0	0	1	0
Eastern gray squirrel (S. carolinensis)	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0
Total number_STA	24	2	12	0	2	2	0	2	2	14	9	6	14	6	20	16
Total number_TRAN	38			4			4			29			40			—
Species richness_STA	4	2	6	0	2	2	0	2	1	4	3	4	4	4	5	5
Species richness_TRAN	12			4			3			11			13			—

Table 5.

Seasonal Distribution of Vertebrates and Arthropods Collected from Five Transects Surrounding the California National Primate Research Center in Yolo County, CA, from January 2013 to December 2014

	December–February	March–May	June–August	September–November	Total
Vertebrates
American opossum (D. virginiana)	0	9	7	0	16
Domestic cat (F. catus)	0	2	0	0	2
Black-tailed jackrabbit (L. californicus)	0	1	1	0	2
Striped skunk (Mephitis mephitis)	0	1	0	0	1
California vole (Microtus californicus)	0	0	5	0	5
House mouse (Mus musculus)	3	14	20	0	37
California ground squirrel (O. beecheyi)	0	3	9	0	12
Brush mouse (P. boylii)	1	0	0	0	1
Raccoon (P. lotor)	2	9	6	0	17
Black rat (R. rattus)	7	7	10	0	24
Western harvest mouse (R. megalotis)	3	2	6	0	11
Eastern gray squirrel (S. carolinensis)	0	0	1	0	1
Total by season	16	48	65	0	129
Arthropods
Biting flies	1	7	43	30	81
Mosquitoes	22	157	1560	526	2265
Ticks	0	51	45	0	96
Fleas	0	41	70	8	119
Total by season	23	256	1718	564	2561

Diagnostic testing

We tested 650 ticks, 519 pools of mosquitoes (n = 2477), 10 T. similis, 18 pools of S. donovani (n = 39), and 13 pools of ceratopogonids (n = 28) for F. tularensis DNA with no positive results. All of the aquatic 42 culicids, 13 ceratopogonids, 60 physids, 1 planorbid, 12 daphnids, and 23 corbiculids were also PCR negative.

Of 109 vertebrates tested for antibodies to F. tularensis, seropositive animals included 1 rat (4.54% of rats, 95% CI: 0.002–0.249), 2 cats (100%, 95% CI: 0.198–1.0), and 5 opossums (50%, 95% CI: 0.237–0.763) from Putah Creek (4 of the opossums and the rat) and CNPRC and adjacent Dump (both cats) (Table 6). The number of seropositive animals varied significantly by station (X² = 287.18, p = 2.2 × 10⁻¹⁶), with the majority coming from Putah Creek.

Table 6.

Indirect ELISA Results for Francisella tularensis Antigen on Serum Collected from Small Mammals and Mesocarnivores Collected from Five Transects Surrounding the California National Primate Research Center in Yolo County, CA, in 2013 and 2014

	Number tested	Borderline	Positive	Seroprevalence, %
Station 1, transect 1, all species	24	2	0	0.00
Virginia possum (D. virginiana)	1	1	0	0.00
House mouse (Mus musculus)	16	1	0	0.00
Station 3, transect 1, all species	11	0	1	9.09
Virginia possum (D. virginiana)	1	0	1	100
Station 12, transect 4, all species	3	0	1	33.33
Domestic cat (F. catus)	1	0	1	100
Station 14, transect 5, all species	5	1	0	0.00
Black rat (R. rattus)	3	1	0	0.00
Station 15, transect 5, all species	18	0	5	27.78
Virginia possum (D. virginiana)	4	0	4	100
Black rat (R. rattus)	7	0	1	14.3
Collected by CNPRC pest control, all species	12	1	1	8.33
Domestic cat (F. catus)	1	0	1	100
California ground squirrel (O. beecheyi)	9	1	0	0.00
Overall	109	4	8	7.34

Seroprevalence values do not include borderline results.

ELISA, enzyme linked immunosorbent assay.

Logistic regression indicated that the number of mosquitoes (OR = 1.001, 95% C.I. 1.000–1.003, p = 0.05) and the number of ticks (OR = 1.019, 95% C.I. 1.000–1.039, p = 0.04) were very slightly but significantly associated with a higher probability of recovering a seropositive animal from a transect.

Discussion

Identification of a source of risk for F. tularensis is complicated by the very low prevalence of the agent in the environment during an inter-epidemic period as well as the variety of species it infects. Here, we examine the ecology of F. tularensis holarctica during an inter-epidemic period at an NHP research site that experiences recurrent, although poorly characterized, epidemics. The 2010 epidemic featured a clinical incidence of 2.7%, and a retrospective examination of records also documented cases three times between 1987 and 1990 (Sammak et al. 2012). Although we did not confirm an arthropod reservoir, we identified opossums and cats as important sentinels. Nearby locations with a permanent water source and natural, dense vegetation may be functional reservoirs, and small mammals and their fleas may have been amplifying and spillover hosts, increasing risk to the NHPs and people.

The tick D. variabilis is the most common source of human cases in California (Reese et al. 2010) but was only present in our study in peripheral marsh and creek vegetation and, thus, unlikely to have directly infected the NHPs. All ticks were PCR negative, which was not surprising given reported infection prevalence as low as 0.7% in endemic areas (Goethert et al. 2004). Mosquitoes were common at both the Russell Marsh and Putah Creek transects, with Putah Creek having higher numbers and higher species richness. Mosquitoes are important vectors of tularemia in Europe (Thelaus et al. 2014) but in the United States, infection in mosquitoes has only been reported from Alaska (Triebenbach et al. 2010). Tabanid flies that mechanically vector tularemia (Philip et al. 1955, Krinsky 1976) were primarily at Russell Marsh, across the street from cattle and horses; and black flies were common near the CNPRC South Colony, probably associated with the birds at the site. From vertebrate hosts, we recovered multiple species of fleas that will parasitize different hosts (Dryden and Rust 1994). The possible role of fleas in tularemia ecology is debated. They are reported to be poor vectors of F. tularensis (Volfertz and Kolpakova 1946), but, because they retain infection for many weeks and the infectious dose is so low, high numbers may contribute to vectorial capacity (Keim et al. 2007, Foley and Nieto 2010).

The most abundant small mammals were rats and mice, especially at Putah Creek. Rodents are amplification hosts, typically experiencing high mortality rates (McCoy and Chapin 1912, Burroughs et al. 1945, Origgi et al. 2015). Given their short lifespans and high population turnover, our finding of a seropositive rat as previously reported (Reintjes et al. 2002) suggests endemic tularemia at Putah Creek. Although lagomorphs are an important reservoir (Burroughs et al. 1945), we could not evaluate them because of the species' reluctance to enter box traps. The highest capture rate of ground squirrels (one of which was borderline seropositive) was at the CNPRC South Colony. Ground squirrels are a known amplification host for tularemia (McCoy and Chapin 1912) and support ectoparasites that readily take blood meals from other species, including humans (Hubbart 2012).

Mesocarnivores, particularly raccoons and skunks, are important sentinels for tularemia with seroprevalence as high as 60% (Taylor et al. 1991, Berrada et al. 2006), are commonly parasitized by D. variabilis ticks, and may share fleas with other hosts (Hubbard 1968). Human cases have been linked to exposure to raccoon carcasses (Francis 1937). Serology does not allow us to differentiate whether animals had been exposed during the 2010 epidemic or whether there was ongoing exposure. However, in our study, only one of the 17 raccoons was seropositive, which may reflect high turnover in this species, possibly associated with attraction to monkey chow and local pest control. In contrast, feral cats and opossums had 100% and 50% seroprevalence, respectively, although again, animals could have been exposed in 2010 or subsequently. Cats are frequently associated with human cases of tularemia (Baldwin et al. 1991, Eliasson et al. 2002) and may transmit the disease through infectious bites (Arav-Boger 2000). Opossums have been identified as a sentinel in the past (Mease 1929, Francis 1937, McKeever et al. 1958).

There were marked differences among microhabitats in the total numbers of animals and seroprevalence. Although our logistic regression showed that higher numbers of mosquitoes and ticks were associated with higher seroprevalence in vertebrates, these results were subject to low power. Station 15 at Putah Creek had a seroprevalence of 28%, the highest capture results for vector species, and the second highest capture results for both sentinel and amplifying vertebrate hosts. This station was located among dense trees and blackberries. Moisture supports vector persistence, whereas standing water may become contaminated with F. tularensis and allow semi-aquatic mammals and water-associated vectors to amplify the pathogen (Jellison et al. 1942, Forsman et al. 1995, Tarnvik et al. 1996, Gyuranecz et al. 2011).

Climatic conditions during our study were somewhat different than during the latest tularemia epidemic at CNPRC with high temperatures and drought in 2013 and 2014, whereas 2010 was a relatively cool, rainy year (National Climatic Data Center 2011). Warmer temperatures likely created a longer active season for many arthropods and reduced the lengths of their reproductive cycles (Reisen et al. 1993, Delatte et al. 2009). In contrast, hard ticks would have reduced survival and questing behavior (Lane et al. 1995, Zahler and Gothe 1995), although tree cover at Putah Creek mitigates effects of high temperature and low humidity. Reduced water during our study period may also have concentrated susceptible vertebrates around resources, whereas in 2010, vertebrates may have been more broadly distributed.

In this study, we have given evidence that riparian sites may be at high risk for endemic tularemia and serve as possible sources for spread and spillover of the pathogen. Mesocarnivores, particularly opossums and cats, may serve as sentinels and transport F. tularensis either directly or via their ectoparasites into populated areas. High densities of rodents and their fleas located adjacent to human settlements represent a risk for disease spillover into human or outdoor-housed captive animal populations. Our results can be used to structure future surveillance efforts to maximize the chance of early detection and spillover prevention.

Footnotes

Acknowledgments

The authors thank the Sacramento/Yolo Vector Control for the resources, expertise, and advice regarding vector trapping for this project; Nicole Stephenson and Mary Straub for providing veterinary support; Christina Ball for her assistance in collecting insects in the field; Kenny Lou for his assistance in the laboratory; Patrick Foley and many other members of the Foley lab for their input and assistance on this project; and Jeff Roberts, Chris Barker, and Rebecca Sammak. Vertebrates and invertebrates were collected in accordance with the regulations under USGS Scientific Collecting Permit #12178 and University of California Davis IACUC Permit #16555. This work was funded by the Karen C. Drayer Wildlife Health Center.

Author Disclosure Statement

No competing financial interests exist.

References

Akimana

, Kwaik

. Francisella-arthropod vector interaction and its role in patho-adaptation to infect mammals. Front Microbiol, 2011; 2:1–15.

Arav-Boger

. Cat-bite tularemia in a seventeen-year-old girl treated with ciprofloxacin. Ped Inf Dis J, 2000; 19:583.

Baldwin

, Panciera

, Morton

, Cowell

, et al. Acute tularemia in three domestic cats. J Am Vet Med Assoc, 1991; 199:1602–1605.

Berrada

, Goethert

, Talford

. Raccoons and skunks as sentinels for enzoonotic tularemia. Emerg Infect Dis, 2006; 12:1019–1021.

Burroughs

, Holdenried

, Longanecker

, Meyer

. A field study of latent tularemia in rodents with a list of all known naturally infected vertebrates. J Infect Dis, 1945; 76:115–119.

California Department of Public Health. Yearly summaries of selected general communicable diseases in California, 2001–2010. 2010. Available at www.cdph.ca.gov/data/statistics/Pages/YearlySummariesofSelectedGeneralCommunicableDiseasesinCA2001-2010.aspx, pp. 1–6.

Center for Food Safety and Public Health. Tularemia Factsheet. 2009. Available at www.cfsph.iastate.edu/Factsheets/pdfs/tularemia.pdf, Iowa State University.

Chaignat

, Djordjevic-Spasic

, Ruettger

, Otto

, et al. Performance of seven serological assays for diagnosing tularemia. BMC Infect Dis, 2014; 14:234.

Delatte

, Gimonneau

, Triboire

, Fontenille

. Influence of temperature on immature development, survival, longevity, fecundity, and gonotrophic cycles of Aedes albopictus, vector of chikungunya and dengue in the Indian Ocean. J Med Entomol, 2009; 46:33–41.

10.

Dryden

, Rust

. The cat flea: Biology, ecology and control. Vet Parasitol, 1994; 52:1–19.

11.

Eliasson

, Lindback

, Nuorti

, Arneborn

, et al. The 2000 tularemia outbreak: A case-control study of risk factors in disease-endemic and emergent areas, Sweden. Emerg Infect Dis, 2002; 8:956–960.

12.

Ellis

, Oyston

, Green

, Titball

. Tularemia. Clin Microbiol Rev, 2002; 15:631–646.

13.

Foley

, Nieto

. Tularemia. Vet Microbiol, 2010; 140:332–338.

14.

Forsman

, Nyren

, Sjostedt

, Sjokvist

, et al. Identification of Francisella tularensis in natural water samples by PCR. FEMS Microbiol Ecol, 1995; 16:83–92.

15.

Francis

. Sources of infection and seasonal incidence of tularemia in man. Publ Health Rep, 1937; 52:103–123.

16.

Francis

, Mayne

. Experimental transmission of tularemia by flies of the species Chrysops discalis . Publ Health Rep, 1921; 36:1738–1746.

17.

Fulks

. 2005. Putah Creek Riparian Reserve Management Plan. University of California at Davis Office of Resource Management and Planning. UC Davis, 46.

18.

Furman

, Loomis

. The Ticks of California (Acari: Ixodida), vol. 25. Berkeley, CA: University of California Press, 1984.

19.

Gliatto

, Rae

, McDonough

, Dasbach

. Feline tularemia on Nantucket Island, Massachusetts. J Vet Diagn Invest, 1994; 6:102–105.

20.

Goethert

, Shani

, Telford

. Genotypic diversity of Francisella tularensis infecting Dermacentor variabilis ticks on Martha's Vineyard, Massachusetts. J Clin Microbiol, 2004; 42:4968–4973.

21.

Gyuranecz

, Rigo

, Dan

, Foldvari

, et al. Investigation of the ecology of Francisella tularensis during an inter-epizootic period. Vector Borne Zoonotic Dis, 2011; 11:1031–1035.

22.

Hubbard

. Fleas of Western North America. Their Relation to Public Health. New York, NY: Hafner Publishing Co., 1968.

23.

Hubbart

. The California ground squirrel (Spermophilus beecheyi): Characterizing an adaptive fossorial vertebrate for improved science-based management decisions. J Biol Life Sci, 2012; 3:1–12.

24.

Humair

, Douet

, Moran Cadenas

, Schouls

, et al. Molecular identification of bloodmeal source in Ixodes ricinus ticks using 12S rDNA as a genetic marker. J Med Entomol, 2007; 44:869–880.

25.

Jellison

, Kohls

, Butler

, Weaver

. Epizootic tularemia in the beaver, Castor canadensis, and the contamination of stream water with Pasteurella tularensis . Am J Epidemiol, 1942; 36:168–182.

26.

Keim

, Johansson

, Wagner

. Molecular epidemiology, evolution and ecology of Francisella . Ann N Y Acad Sci, 2007; 1105:30–66.

27.

Krinsky

. Review article: Animal disease agents transmitted by horse flies and deer flies (Diptera: Tabanidae). J Med Entomol, 1976; 13:225–275.

28.

Lane

, Kleinjan

, Schoeler

. Diel activity of nymphal Dermacentor occidentalis and Ixodes pacificus (Acari: Ixodidae) in relation to meteorological factors and host activity periods. J Med Entomol, 1995; 32:290–299.

29.

MacAlpine

, Peterson

, Shewell

, Teskey

, et al. Ceratopogonidae. Manual of Nearctic Diptera Vol. 1. Ottawa, Ontario: Canandian Government Publication Center, 1981:393–421.

30.

McCoy

, Chapin

. Further observations on a plague-like disease of rodents with a preliminary note on the causative agent, Bacterium tularense . J Infect Dis, 1912; 10:61–72.

31.

McKeever

, Schubert

, Moody

, Gorman

, et al. Natural occurrence of tularemia in marsupials, carnivores, lagomorphs, and large rodents in southwestern Georgia and northwestern Florida. J Infect Dis, 1958; 103:120–126.

32.

Mease

. Tularemia from opossums. JAMA, 1929; 92:1042.

33.

Meyer

, Durso

. Identification of the Mosquitoes of California. Sacramento, CA: California Mosquito and Vector Control Association, 1993.

34.

Middlekauff

, Lane

. Adult and Immature Tabanidae (Diptera) of California, Vol. 22. Berkeley, CA: University of California Press, 1980.

35.

National Climatic Data Center. Monthly Climatological Summary (2010, 2013–2014). National Oceanic and Atmosphere Administration. 2011. Available at www.ncdc.noaa.gov

36.

Origgi

, Konig

, Lindholm

, Mayor

, et al. Tularemia among free-ranging mice without infection of exposed humans, Switzerland, 2012. Emerg Infect Dis, 2015; 21:133–135.

37.

Pennak

. Fresh-water invertebrates of the United States: Protozoa to Mollusca, 3rd ed. United States of America: Wiley Interscience, 1989.

38.

Petersen

, Mead

, Schriefer

. Francisella tularensis: An arthropod-borne pathogen. Vet Res, 2009; 40:07.

39.

Philip

, Bell

, Larson

. Evidence of infectious diseases and parasites in a peak population of black-tailed jack rabbits in Nevada. J Wildl Management, 1955; 19:225–233.

40.

Reese

, Dietrich

, Dolan

, Sheldon

, et al. Transmission dynamics of Francisella tularensis subspecies and clades by nymphal Dermacentor variabilis . Am J Trop Med Hygiene, 2010; 83:645–652.

41.

Reintjes

, Dedushaj

, Gjini

, Jorgensen

, et al. Tularemia outbreak investigation in Kosovo: Case control and environmental studies. Emerg Infect Dis, 2002; 8:69–73.

42.

Reisen

, Meyer

, Presser

, Hardy

. Effect of temperature on the transmission of western equine encephalomyelitis and St. Louis encephalitis viruses by Culex tarsalis (Diptera: Culicidae). J Med Entomol, 1993; 30:151–160.

43.

Sammak

, Rejmanek

, Roth

, Christie

, et al. Francisella tularensis outbreak investigation following natural infection of outdoor housed rhesus macaques (Macaca mulatta). Comp Med, 2012; 63:183–190.

44.

Tarnvik

, Sandstrom

, Sjostedt

. Epidemiological analysis of tularemia in Sweden 1931–1993. FEMS Immunol Med Microbiol, 1996; 13:201–204.

45.

Taylor

, Istre

, McChesney

, Satalowich

, et al. Epidemiological characteristcs of human tularemia in the southwest-central States, 1981–1987. Am J Epidemiol, 1991; 133:1032–1038.

46.

Thelaus

, Andersson

, Mathisen

, Forslund

A-L

, et al. Influence of nutrient status and grazing pressure on the fate of Francisella tularensis in lake water. FEMS Microbiol Ecol, 2009; 67:69–80.

47.

Thelaus

, Andersson

, Broman

, Backman

, et al. Francisella tularensis subspecies holarctica occurs in Swedish mosquitoes, persists through the developmental stages of laboratory-infected mosquitoes and is transmissible during blood feeding. Microb Ecol, 2014; 67:97–107.

48.

Triebenbach

, Vogl

, Lotspeich-Cole

, Sikes

, et al. Detection of Francisella tularensis in Alaskan mosquitoes (Diptera: Culicidae) and assessment of a laboratory model for transmission. J Med Entomol, 2010; 47:639–648.

49.

Triplehorn

, Johnson

. Borror and DeLong's Introduction to the Study of Insects, 7th ed. Belmont, CA: Brooks/Cole, Cenage Learning, 2005.

50.

Ullman-Cullere

, Foltz

. Body condition scoring: A rapid and accurate method for assessing health status in mice. Lab Anim Sci, 1999; 49:319–323.

51.

van Hoek

. Biofilms: An advancement in our understanding of Francisella species. Virulence, 2013; 4:833–846.

52.

Versage

, Severin

, Chu

, Petersen

. Development of a multitarget real-time TaqMan PCR assay for enhanced detection of Francisella tularensis in complex specimens. J Clin Microbiol, 2003; 41:5492–5499.

53.

Volfertz

, Kolpakova

. The epizootology of tularaemia. Third communication. The role of the fleas Ctenophthalmus orientalis Wagn., in enzootology of tularaemia. Meditsinskaia Parazitologiia i Parazitarnye Bolezni, 1946; 15:83–87.

54.

Woods

, Panciera

, Morton

, Lehenbauer

. Feline tularemia. Comp Cont Ed Pract Vet, 1998; 20:442–457.

55.

Zahler

, Gothe

. Effect of temperature and humidity on longevity of unfed adults and on oviposition of engorged females of Dermacentor reticulatus (Ixodidae). Appl Parasitol, 1995; 36:200–211.