Background

Mosquitoes are the deadliest animals in the world, spreading many pathogens to their hosts [1, 2]. Although most vectors feed on endothermic hosts such as mammals or birds, some species are known to be ectotherm specialists [3, 4]. A primary example is Culex territans Walker 1856, a species of mosquito found in temperate and subtropical zones across the Northern Hemisphere [5,6,7,8,9] that feeds preferentially on amphibians, especially anurans (i.e., frogs and toads) [10,11,12]. Despite its widespread distribution, little is known about this mosquito. Culex territans is a potential vector of pathogens such as giant anuran trypanosomes, hepatozoon parasites [12,13,14], and viruses [15, 16], but its role in transmitting other pathogens, such as fungi, is unknown. Batrachochytrium dendrobatidis (Bd) is a topical pathogenic fungus affecting the keratinized tissue in amphibians and is a major contributor to declines in amphibian populations worldwide [17,18,19,20]. Chytridiomycosis, the disease caused by Bd, affects anuran amphibians more than other amphibian orders, with pathogenic effects varying significantly among species [21, 22]. As an aquatic fungus, Bd can be spread environmentally through water or by physical contact with other infected frogs [23]. Fish, crustaceans, and reptiles are known carriers of Bd [24,25,26,27,28]; however, the roles that vectors might play in transmitting this deadly fungus remains poorly understood [29, 30]. Toledo et al. [29] found Bd DNA present on biting midges (Corethrella spp.), showing that arthropod vectors can play a role in the transmission of this fungus. Mosquitoes are known to harbor fungi [31, 32], and Cx. territans has a long period of contact time with the skin surface of the frog during blood-feeding (minimum of 30 min to feed to repletion [33]). This amount of time could allow the proboscis and the legs to acquire any fungal spores present on the skin surface of their host. Interestingly, a closely related mosquito species, Cx. quinquefasciatus, which feeds on birds and mammals, has been shown to be able to acquire the fungus and transport it to a new surface [30]. Our study tests the ability of Cx. territans, the northern frog-biting mosquito, to transfer Bd to amphibian hosts, as Cx. quinquefasciatus is not known to feed on amphibians.

In the present study, we hypothesized that Cx. territans could acquire Bd spores and transmit them to a new surface, thus demonstrating the ability to vector this fungal pathogen to anuran hosts through mechanical transmission (i.e., by spreading spores present on the mosquito legs to the skin of their hosts). To test this hypothesis, we first conducted blood meal analyses to determine the primary hosts for this specific population of Cx. territans. We then screened field-caught mosquitoes and frogs for the presence of Bd at our field site, Mountain Lake Biological Station (MLBS, Pembroke, VA, USA), and carried out a controlled experiment to evaluate the ability of Cx. territans to transmit the fungus from one medium surface to another. This work will provide a deeper insight into mosquito-borne disease ecology as well as understanding the potential Cx. territans has for vectoring Bd, which will enhance our understanding of the spread of this deadly fungus.

Methods

Mosquito sample collection

Mosquitoes were collected at MLBS (37.375654°–80.522140°, 1160 m above sea level [ASL]) in Giles County, VA, USA. Mosquitoes were collected while at rest around midday from May through September at Sylvatica Pond (37.377079, −80.522245) using a giant bug aspirator (John W. Hock Company, Gainesville, FL, USA) [34]. Following collection, the mosquitoes were brought to the MLBS laboratory, cold-anesthetized, and sorted under a microscope based on species, sex, and reproductive stage [6]. Blood-fed Cx. territans females were put in individual 2-ml Eppendorf tubes (Cat #24-283S, Genesee Scientific, Rochester, NY, USA) and were immediately frozen at −80 °C until use for blood meal analysis and pathogen screening for the presence of Bd (Fig. 1A), and any bycatch was released.

Fig. 1
figure 1

A Mosquitoes were collected using a giant bug aspirator around a pond occupied by R. clamitans and other anurans. Female Cx. territans were sorted out, and blood meal analysis was performed. B Blood meal sources of Cx. territans mosquitoes by year between 2018 and 2022

Full size image

Blood meal analysis

Blood-fed mosquitoes collected in the field were thawed, and the abdomens (containing the blood meals) were removed using fine forceps. Abdomens were homogenized using a homogenizer and pestle, and DNA was extracted using Qiagen DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany), following the manufacturer’s instructions (without the purification step, eluted in 20 µl AE buffer). We amplified the mitochondrial 16S ribosomal RNA (rRNA) gene from each mosquito blood meal via polymerase chain reaction (PCR) using a 50 µl solution containing 25 µl DreamTaq Green PCR Master Mix (2×) (Cat #FERK1081 Thermo Fisher Scientific, Waltham, MA, USA), 21 µl PCR-grade water (CAS: 7732-1-8-5, Sigma Aldrich, St. Louis, MO, USA), 1 µl of forward and reverse primers (L2513: 3′-GCCTGTTTACCAAAAACATCAC-5′; H2714: 3′-CTCCATAGGGTCTTCTCGTCTT-5′) [35] and 2 µl DNA. Amplification was carried out using a thermal cycler (MyCycler, Bio-Rad, Hercules, CA, USA) with 35 cycles of 95 °C for 30 s, 57 °C for 15 s, and 72 °C for 30 s. PCR products were run on a 1% agarose gel containing 0.1% GelRed Nucleic Acid Gel Stain (Biotium, Fremont, CA, USA) for 30 min at 110 V to confirm successful amplification (Bio-Rad), with the amplified product sent to Genewiz (South Plainfield, NJ, USA) for Sanger sequencing. Sequences were aligned using the Qiagen CLC Genomics Workbench (Qiagen, Hilden, Germany) [36], and analyzed using the National Center for Biotechnology Information (NCBI) Basic Local Alignment Search Tool (BLAST). Identification of the source of mosquito blood meals was determined for sequences with a sequence identity of 85% or higher [37] (Fig. 1A).

Anuran collection and swabbing

For taxonomic stability, we follow the nomenclature of AmphibiaWeb (https://amphibiaweb.org). Anurans (frogs) were collected by hand at three permanent artificial ponds (naturalized for over 50 years) at MLBS: Riopel (37.374552, −80.552109), Sylvatica (37.377079, −80.522245), and Horton (37.378483, −80.522078). The largest of these ponds, Riopel, is approximately 80 × 90× 4 m, whereas the smaller two fluctuate in depth (≤ 1 m) depending on rainfall and measure approximately 30 × 30 m. None of the ponds contain fish. Frog sampling was conducted under Virginia Department of Game and Inland Fisheries collection permit nos. 064750 and 070446 and Virginia Tech Institutional Animal Care and Use Committee (IACUC) protocols (#19-003 and #22-066). Based on the results obtained for the mosquito blood meal analysis, collections were limited to male and female Rana (Aquarana) clamitans and Rana (Aquarana) catesbeiana. Individuals were placed and kept in unused polyethylene bags (25.4 × 40.64 cm 4 Mil Industrial Poly Bags, Uline, Pleasant Prairie, WI, USA) with spring water and tied to prevent escape. Larger frogs were kept in individual terrariums (23.5 × 15.24 × 17.78 cm Kritter Keeper) (Lee’s Aquarium & Pet Products, San Marcos, CA, USA), which were cleaned using a 10% bleach solution between uses to prevent contamination. During handling, nitrile gloves were worn and changed between each individual to prevent potential cross-contamination. Adult frogs were swabbed for Bd (described below), measured (snout–vent length), weighed using a spring scale (Pesola, Schindellegi, Switzerland), and marked using the toe clipping method described by Martoff [38] prior to release.

Adult frogs of the two species and sexes were swabbed using rayon-tipped swabs (MW113 series urethral swab, Medical Wire & Equipment Co Ltd, Corsham, UK). The skin of each individual was swabbed five times each on the venter, each medial thigh, and between the digits of each hind leg (Fig. 2A). Swabs were frozen at −80 °C and sent for Bd screening at the Genetics Core Facility of the Sam Noble Museum, University of Oklahoma.

Fig. 2
figure 2

A Frogs were swabbed five times each on the (1) abdomen, (2) both medial hind thighs, and (3) between the hind digits of both feet. Swabs were (4) sent out for qPCR to confirm Bd presence or absence. B The (1) proboscis and legs of field-caught mosquitoes were removed and (2) placed on a sterile agar plate, which was divided to separate the two body parts. After 4 days, the plates were (3) swabbed and (4) sent out for qPCR and analysis. C Numbers of frogs caught and swabbed per species between 2020 and 2022. D Representation of the numbers of frogs that tested positive for Bd from 2020 to 2022

Full size image

Experimental transmission assays

Culex territans mosquitoes were caught as larvae in the ponds at MLBS and reared in the laboratory under conditions as described in Reinhold et al. [39]. Female Cx. territans were tested for transmission using a modified experimental setup after Gould et al. [30]. All experiments were conducted in a biosafety cabinet at 25–27 °C using standard sterile practices to maintain the sterility of the plates. Batrachochytrium dendrobatidis Global Panzootic Lineage (GPL) strain JEL0423 (Collection of Zoosporic Eufungi at the University of Michigan [CZEUM], Ann Arbor, MI, USA) was plated on sterile 1%T media using CZEUM protocols (10 g Fisher BioReagents tryptone BP1421-500; Fisher Scientific, Waltham, MA, USA) and 10 g bacteriological-grade agar [IBI Scientific CAS #9002-18-0, Dubuque, IA, USA], in 1 L deionized water). Mosquitoes were briefly cold-anesthetized and tethered by the thorax to tungsten rods with ultraviolet (UV)-cured glue (Bondic, Niagara Falls, NY, USA), which held the mosquitoes stationary when exposed to a plate containing Bd (cell count: 5.251 × 105 ± 9.727 × 105) (50 × 11 mm petri dishes, #EW-14005-24, Advantech, Taipei City, Taipei, Taiwan). These plates were swabbed to confirm the presence of Bd after mosquito exposure. An average of ten mosquitoes (8 < n < 12) were held on the Bd containing plate for 30 min (N = 123, 13 replicates, with some mosquitoes lost during transfer). These mosquitoes were then transferred to new individual sterile 1%T agarose plates for 30 min of contact time to simulate the average time needed for this mosquito species to feed to repletion on a frog host [33]. Plates were incubated at room temperature for 96 h before being swabbed. Swabs were sent to the Genetics Core Facility of the Sam Noble Museum, University of Oklahoma, to confirm the transfer of Bd. To control for environmental or potential previous Bd exposure (during larval or pupal stage), the same experiment was conducted by exposing 50 mosquitoes (in 5 sets of 10) to sterile plates instead of Bd-inoculated plates (N = 50). These mosquitoes were also transferred to new individual plates after 30 min and allowed to maintain contact for 30 min. Plates were swabbed after 96 h and sent out for processing as described previously (Fig. 3A).

Fig. 3
figure 3

A Mosquitoes were (1) tethered to a tungsten rod and (2) placed on a Bd plate with legs touching the agar. After 30 min, the mosquitoes were (3) transferred to a new, individual sterile plate with limbs in contact with the agar for another 30 min and (4) let sit for 4 days. After 4 days, the plates were (5) swabbed and stored at −80 °C until they were (6) sent out for qPCR and analysis. B Representation of negative controls (n = 56). C Representation of Bd-infected plates to confirm the presence of Bd (n = 6). D The number of plates that were positive for Bd 4 days after contact with an inoculated mosquito (n = 123)

Full size image

Field mosquito Bd sampling

A total of 82 field-caught, blood-fed Cx. territans mosquitoes [57 and 25 from 2021 and 2022, respectively] were sampled for Bd. Sterile 1%T agar plates (50 × 11 mm petri dishes, #EW-14005-24, Advantech, Taipei City, Taipei, Taiwan) were made following CZEUM instructions to simulate the landing onto the surface of an amphibian’s skin. Each plate was divided into 2 sides for the proboscis and legs respectively. The proboscis was clipped using fine-tipped forceps and put on one side of the plate, and the legs from the same mosquito were pulled and placed on the other side of the plate. Forceps were cleaned and dried before removing any parts of the mosquitoes and in between individuals. Plates were incubated at room temperature for 96 h before being swabbed with MW113 swabs. Swab samples were stored at −80 °C until being sent out for processing as described previously (Fig. 2B).

Screening samples using qPCR

Pathogen screening for Bd was conducted at the Genetics Core Facility of the Sam Noble Museum at the University of Oklahoma. Swabs from both agar plates and frogs were first extracted using the PrepMan Ultra Sample Preparation Reagent protocol [40] (Applied Biosystems, Thermo Fisher Scientific). Quantitative PCR (qPCR) techniques were used to determine the presence/absence of Bd genetic material and to estimate the number of gene copies per sample, or Bd load, using QuantStudio software v3 (Applied Biosystems, Thermo Fisher Scientific). The Bd assay primers target the internal transcribed spacer (ITS-1) rRNA gene (forward primer: ITS1-3 Chytr; reverse primer: 5.8S) from Boyle et al. [41]. All samples were run in triplicate, with positive controls of known Bd gene copies (gBlock DNA quantities 1e1–1e4) and a negative control (molecular grade sterile water). Samples were considered positive for Bd (Bd+) if amplification occurred in at least two of the three wells and if the mean Bd gene copy number was greater than 1.0, with samples rerun as needed [42, 43].

Results

Blood meal analysis

A total of 538 mosquitoes were collected from Sylvatica Pond at MLBS. Among them, 98 were blood-fed Cx. territans. We identified a total of five host species, including R. clamitans (61%) (green frog), R. catesbeiana (35%) (American bullfrog), Rana sylvatica (2%) (wood frog), H. versicolor (1%) (gray treefrog), and Nerodia sipedon (1%) (northern water snake). We noted that Cx. territans preferentially fed on R. clamitans (61%) and R. catesbeiana (35%) (Fig. 1B).

Frog collection and swabbing

A total of 127 R. clamitans and 87 R. catesbeiana were caught and swabbed between 2020 and 2022 (Fig. 2A, C). Among the individuals swabbed, 17.8% were confirmed as Bd+. The prevalence of Bd infection among the surveyed populations varied between years and species: 9.1% prevalence in R. catesbeiana and 6.1% in R. clamitans during 2020 dropped to 4.5% and 0% in R. catesbeiana and R. clamitans, respectively, in 2021. During 2022 we found 38% and 50% prevalence among R. catesbeiana and R. clamitans, respectively (Fig. 2D). None of the 82 field-caught mosquitoes screened for Bd were positive.

Transmission assays

Of the 123 experimental mosquitoes tested, 80 were positive for Bd (65%) (cell counts 5.251 × 105 ± 9.727 × 105; gene copies 1.94499 × 105 ± 3.37174 × 105) (Fig. 3D). None of the five initial control plates or the 50 test controls were positive for Bd, showing no evidence of prior exposure to the fungus (Fig. 3B). All Bd-inoculated plates contained Bd, confirming the experimental mosquitoes were exposed to the fungus (Fig. 3C).

Discussion

This study was conducted to assess the prevalence of the pathogenic fungus at our field site, MLBS, and to ascertain whether the amphibian-biting mosquito, Cx. territans, is capable of vector transmission of Bd to anurans. We determined the primary hosts of Cx. territans and tested two focal hypotheses: (1) Pathogen screening efforts would detect Bd on Cx. territans and their anuran blood meal hosts in the field, and (2) Cx. territans is capable of Bd transmission from one surface to another in a laboratory setting.

Although Cx. territans has been known to feed on a wide variety of hosts [10,11,12], our results show that the mosquito population at MLBS exhibits a strong preference for R. clamitans and R. catesbeiana (Fig. 1). Both anuran species are commonly observed around the MLBS ponds. A plethora of other potential hosts have been observed at the MLBS ponds, including R. sylvatica, Rana palustris (pickerel frog) H. versicolor, and Pseudacris crucifer (spring peeper). Additional potential hosts present at MLBS include N. sipedon (northern water snake), Crotalus adamanteus (eastern diamondback rattlesnake), and numerous salamander species (in particular Notophthalmus viridescens, and several Desmognathus and Plethodon spp.) [44]. The strong host preference for R. clamitans and R. catesbeiana exhibited by Cx. territans led us to focus our Bd screening on these two species.

Overall, we found that the current prevalence of Bd is relatively low among the R. clamitans and R. catesbeiana populations at MLBS. Previous studies at MLBS found several species of amphibians carrying Bd. Rothermel et al. [45] and Wimsatt et al. [46] found Bd present in all Notophthalmus viridescens (red-spotted newt) individuals tested at MLBS, which is present in all three ponds at the station where R. clamitans and R. catesbeiana are found. While these studies only tested one species, they demonstrate that Bd has been present in the MLBS ponds for more than a decade. Hughey et al. [47] also reported Bd-infected individuals of N. viridescens, P. crucifer, and R. catesbeiana, and found lower rates of infection among R. catesbeiana than the other two species. None of these studies associated Bd infections at MLBS with die-offs or significant population declines. Collectively, the present and previous studies show that Bd is established at MLBS. Our results also suggest that infection prevalence and detectability vary from year to year (Fig. 2D). This kind of variability is not uncommon in areas where Bd is considered endemic [47,48,49,50,51,52] and may be part of natural long-term oscillations described by Talley et al. [49]. Several factors could be playing a role in these oscillations. The presence of carriers in these areas, including (but not limited to) N. viridescens, R. catesbeiana, and crayfish (e.g., Cambarus spp.) may be acting as fungal reservoirs and may partially account for the yearly fluctuations we detected in anuran populations [24, 26, 27, 45,46,47, 53,54,55]. Environmental factors could also affect Bd prevalence, with colder and drier years potentially causing a shift in Bd prevalence [20, 23, 49, 51, 56, 57]. Despite the sudden increase in Bd in 2022, no mass die-offs of amphibians were observed at the time, which suggests that Bd is not an immediate concern for the amphibians at the site, but future observations and surveillance will provide more insight.

Our experimental transmission test confirmed that Cx. territans can transmit Bd from one surface to another (Fig. 3D). This suggests that Cx. territans can transfer the fungus to a new host after landing and feeding on an infected host. Ecologically, Cx. territans is a good candidate to vector a topical fungus, given that it feeds primarily on frogs and is very slow to feed to repletion [10,11,12, 33]. If the mosquito does not have the opportunity to feed to repletion (i.e., frog jumps into the water), they may seek a new host to complete their meal, allowing them to transmit Bd zoospores between hosts. They may also transmit the fungus between gonotrophic cycles, though further testing is needed to determine the maximum time the mosquito can carry the fungus. Because 35% of the mosquitoes did not transmit Bd to a new plate, there may be a threshold concentration necessary for transmission. The large range in cell counts on each inoculated Bd plate could also have contributed to the variability in individual mosquitoes’ ability to transmit zoospores between plates. Another source of variability could be the uneven nature of the way the Bd fungus grows. All mosquitoes were placed on visible fungus, but it was difficult to determine which plated area was actively releasing zoospores. This may provide a more ecologically relevant way of testing the ability of a mosquito to transmit Bd experimentally than has been previously shown in the literature [30], since the landing area of the mosquito on a frog may not be releasing zoospores. Because more than two-thirds of the mosquitoes transmitted Bd to uninfected surfaces, we conclude that Cx. territans can carry spores on their appendages and could thus be a competent vector of Bd in the field.

We hypothesized that we would find Bd on the mosquitoes caught in the field, but all field-caught mosquitoes were negative for Bd. These results are not unexpected, considering the low prevalence of the fungus in the frog population in 2020 and 2021 and the small mosquito sample size for 2022. However, based on the ability of other arthropod vectors to act as carriers of Bd in the wild [29] and the ability of Cx. territans to transmit the fungus in a laboratory setting, we would expect to find Bd-carrying mosquitoes in a larger sample or among samples from a more highly infected population of frogs. Because this mosquito species lives in close association with frogs throughout the Northern Hemisphere and given that Bd is also prevalent in these regions [5,6,7,8,9,10], transmission in areas of high Bd prevalence is likely. The lack of Bd found on field mosquitoes may also be due to the location of Bd on frogs. Rana clamitans and R. catesbeiana were often observed resting partially submerged in water, as noted in previous studies [38, 58]. This position results in their ventral surfaces being more likely to be exposed to the water-borne fungus, and consequently, unless the fungus has spread extensively, a mosquito feeding on the dorsum of a frog’s body may not make contact with the fungus. Mosquitoes were observed in the field clustering on the frogs’ back legs and lower back, possibly to avoid detection and antiparasitic behavior (e.g., grooming) from the host as seen in other mosquito species [59,60,61]. We also observed mosquitoes on the ventral surfaces of the body and on the toes when those areas were exposed. Culex territans is known to feed on more terrestrial amphibians and reptiles [10, 11]. It would therefore be beneficial for future studies to collect additional Cx. territans mosquitoes and test for field transmission, especially in areas where Bd prevalence is high.

Conclusion

In this study, we determined the host preference of Cx. territans, the prevalence of Bd at the MLBS field site, the ability of Cx. territans to transmit Bd in a laboratory setting, and Bd prevalence among blood-fed mosquitoes in the field. We found that, at MLBS, Cx. territans feeds preferentially on R. clamitans and R. catesbeiana. We observed varying Bd prevalence over the course of 3 years at MLBS within R. clamitans and R. catesbeiana populations. Most importantly, we found that Cx. territans can transmit the fungus to a clean surface. We did not, however, find mosquitoes in the field with Bd present on their proboscis or legs. Overall, the results of this study add to the growing understanding of Bd epidemiology and disease ecology both on a local and global scale. Future research will help determine whether mosquitoes in the field can carry the fungus and transmit it to a new host as well as the length of time the fungus can stay on the mosquito.