Figure 1

Figure 1. Locations of intensive sampling for SARS-CoV-2 samples collected across 28 species in 8 states during community-scale surveillance of SARS-CoV-2 and influenza A viruses in wild mammals, United States, October 2022–June...

Figure 2

Figure 2. Locations of opportunistic sampling for SARS-CoV-2 samples collected across 17 species in 13 states and Puerto Rico during community-scale surveillance of SARS-CoV-2 and influenza A viruses in wild mammals, United...

We collected 1,172 samples from wildlife communities across the United States during September 2022–November 2023. Postmortem samples were collected opportunistically from 36 species in 20 states and Puerto Rico (Appendix 1 Table 1) by US Department of Agriculture Wildlife Services personnel during ongoing permitted management activity (Figure 1) and by the Association of Fish and Wildlife Agencies during Best Management Practices Trap Training (Figure 2); samples were taken from a variety of sympatric mammals in disparate locations. Where possible, we used continuous intensive sampling at the same location for >1 month to improve detection within a given mammal community.

Personnel collected swab samples and Nobuto strip blood samples from each animal (S. Bevins et al., unpub. data, https://doi.org/10.1101/2023.04.14.533542); we performed SARS-CoV-2 RNA preparation and subsequent detection using quantitative reverse transcription RT PCR (qRT-PCR) as previously described (8). US Department of Agriculture Animal and Plant Health Inspection Service National Veterinary Services Laboratories subjected nonnegative samples from novel hosts to confirmatory testing. We prepared SARS-CoV-2–specific neutralizing antibodies (NAbs) from Nobuto strips and detected using the Genscript cPass SARS-CoV-2 neutralization antibody detection kit (Thermo Fisher Scientific, https://www.thermofisher.com) as described (S. Bevins et al., unpub. data). We further investigated nonnegative samples from novel hosts by conventional viral neutralization testing (cVNT). We screened Nobuto eluates from intensively sampled sites (n = 747) (Figure 1) for IAV antibodies using a commercial blocking ELISA Influenza A MultiS-Screen Ab test (IDEXX Laboratories, https://www.IDEXX.com) (12) as described previously (13). We used the manufacturer’s recommended sample-to-negative ratio threshold of <0.5 to determine detection of IAV antibodies in serum (Appendix 2).

qRT-PCR testing detected SARS-CoV-2 RNA in 1 white-tailed deer (Odocoileus virginianus) sample (n = 45) and 2 nutria (Myocastor coypus) samples (n = 41) (Appendix 1 Table 2; Appendix 2 Figure 1). SARS-CoV-2 was not previously documented in nutria; because cycle threshold values were high, we used Sanger sequencing to verify the samples contained nutria host nucleic acid and were not contaminated by a sample from another species. After retesting, we did not have sufficient material for confirmatory testing. Additional sampling and testing are required to confirm nutria susceptibility to or SARS-CoV-2 presence in nutria populations.

Figure 3

Figure 3. Results for testing conducted on a seropositive mink sample collected in Pennsylvania during community-scale surveillance of SARS-CoV-2 and influenza A viruses in wild mammals, United States, October 2022. A) Conventional...

We found serologic evidence of SARS-CoV-2 exposure in 14 samples from 5 species (Appendix 1 Table 3): 1 coyote (Canis latrans; n = 25) (Appendix 2 Figure 2), 1 muskrat (Ondatra zibethicus; n = 41) (Appendix 2 Figure 1), 1 woodchuck (Marmota monax; n = 18) (Appendix 2 Figure 1), 1 domestic American mink (n = 13) (Appendix 2 Figure 1), and 10 white-tailed deer (n = 45) Appendix 2 Figure 1). Sample quality issues prevented the coyote Nobuto sample from cVNT testing. cVNT testing of other Nobuto eluates detected SARS-CoV-2 NAbs at dilution factors of 1:32 for muskrat and 1:8 for woodchuck (Appendix 1 Table 3). Because the mink sample was an escaped domestic mink from a farm that vaccinated its mink, we tested the sample for the Omicron BA.1 and B1 (variant D614G) strains of SARS-CoV-2. We detected NAbs for the B1 (variant D614G) strain at a dilution factor of 1:8 from the mink Nobuto eluate (Figure 3, panel A) but no response to Omicron BA.1 strain. Finally, N luciferase immunoprecipitation assay screening showed reactivity against the N protein (Figure 3, panel B; Appendix 1 Table 4), which indicates the animal was likely exposed to a pre-Omicron variant instead of or in addition to being vaccinated.

ELISA screening for IAV antibodies detected positive results in 7 raccoons (Procyon lotor; n = 270 across sites) (Appendix 2 Figure 3) and 1 Virginia opossum (Didelphis virginiana; n = 112 across sites) (Appendix 1 Table 5; Appendix 2). All positive animals were from the same site in Iowa within the Mississippi Flyway during 2 time periods (October 2022 and March 2023). Samples collected during October 2022 included 3 raccoon detections (n = 88; seroprevalence 3.4%) and the Virginia opossum detection (n = 40; seroprevalence 2.5%), whereas samples collected during March 2023 included 4 raccoon detections (n = 98; seroprevalence 4.1%). Previous opportunistic surveillance of avian IAVs in raccoons reported a similar seroprevalence in Maryland during 2004 (2.4%) but a higher seroprevalence in some western states: 25% in Wyoming during 2004 and 12.8% in Colorado during 2006 (14).

Figure 4

Figure 4. Posterior estimates of disease freedom varied based on prior input and the number of samples collected from each species in March 2023 in study of SARS-CoV-2 and influenza A viruses...

In sites where no detections occurred, predictions of disease freedom were strongly influenced by the prior probability of disease freedom (i.e., site-level disease risk), but that influence was weakened by the sample size collected from each species within a site (Appendix 1 Table 6). We analyzed and illustrated the dependence between disease freedom estimates, sample size, and site-level disease risk at 1 site sampled in Iowa (Figure 4). Assumptions about site-level disease risk strongly determined disease freedom probability for species with <3 samples, such the eastern cottontail rabbit (Sylvilagus floridanus) with 1 sample. By comparison, disease freedom probability did not depend as greatly on site-level disease risk when >30 samples for a single species per site were collected, such as for raccoons or Virginia opossum.

We did not find widespread SARS-CoV-2 occurrence in the wildlife communities, even for wildlife species sympatric with deer. We found evidence for infrequent incidence of SARS-CoV-2 exposure in novel species, highlighting the importance of appropriate site-level sample sizes for detection of rare exposure events. Our disease freedom analysis provides sampling guidance for detection of rare events in future surveillance programs. We did not find evidence of co-circulation of SARS-CoV-2 and IAVs in the same animals or species but did find sympatric exposure to IAVs in raccoons and Virginia opossum in the Mississippi Flyway. Community-scale wildlife disease surveillance is important for monitoring changing host ranges that can be realized given local ecologic contexts for rapidly evolving viruses and for refining risk-based surveillance designs.